does competition affect financial stability?
MJS1994
journels.zipBranch Banking, Bank Competition, and Financial Stability.pdf
Branch Banking, Bank Competition, and Financial Stability Author(s): Mark Carlson and Kris James Mitchener Source: Journal of Money, Credit and Banking, Vol. 38, No. 5 (Aug., 2006), pp. 1293-1328 Published by: Ohio State University Press Stable URL: http://www.jstor.org/stable/3839007 .
Accessed: 13/02/2015 16:11
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Ohio State University Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of Money, Credit and Banking.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON
KRIS JAMES MITCHENER
Branch Banking, Bank Competition, and
Financial Stability
It is often argued that branching stabilizes banking systems by facilitating diversification of bank portfolios; however, previous empirical research on the Great Depression offers mixed support for this view. Using data on national banks from the 1920s and 1930s, we show that branch banking raises the level of competition and increases exit from the banking system. This consolidation strengthens the system as a whole without necessaiiLly strengthening the branch banks themselves. Our empirical results suggest that the effects that branching had on competition were quantitatively more important than geographical diversification for bank stability in the 1920s and 1930s.
JEL codes: G21, N22, E44 Keywords: branch banking, bank consolidation, financial
stability, Great Depression.
ONE OF THE FOUNDATIONS of the theoretical literature on banking regulation is that branch banking leads to more stable banking systems by enabling banks to better diversify their assets and widen their depositor base (Gart, 1994, Hubbard, 1994). This conventional wisdom has been used to argue that historical banking crises in the United States, especially those ofthe 1930s, wouldhave been less severe had the U.S. permitted widespread branch banking (Friedman and Schwartz, 1963, Calomiris, 2000). The empirical literature examining U.S. banking instability during the Great Depression, however, has not universally confirmed this
We thank Waiyi Poon for valuable research assistance and Joe Mason, Bill Sundstrom, David Wheelock, Eugene White, and conference and seminar participants at the AEA Annual Meetings, the Economic History Society Annual Meetings, NBER-DAE, and Yale University for comments and suggestions. The views presented in this paper are solely those of the authors and do not necessarily represent those of the Federal Reserve System or its staff.
MARK CARLSON is an Economist at the Federal Reserve Board (E-mail: Mark.A. [email protected]). KRIS JAMES MITCHENER is an Assistant Professor from the Department of Economics at Santa Clara University and Faculty Research Fellow NBER (E-mail: kmitchener@ scu.edu).
Received December 16, 2003; and accepted in revised form April 3, 2005.
Journal of Money, Credit, and Banking, Vol. 38, No. 5 (August 2006) Copyright 2006 by The Ohio State University
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1294 : MONEY, CREDIT, AND BANKING
prediction. In fact, this research presents a paradox. Studies using aggregate bank failure data from the Depression find that states that allowed branch banking had lower failure rates than those that only allowed unit (or single-office) banking (Wheelock, 1995, Mitchener, 2000, 2005); although the result is consistent with the diversification hypothesis, these studies do not test the precise channel through which branching reduced failures. In contrast, studies using individual bank data from the same period cast doubt on the common view that the stabilizing benefits of branching operated via increased diversification opportunities. Calomiris and Mason (2000) and Carlson (2004) find that banks with branches were more likely to fail than unit banks, in part because they pursued strategies to reduce reserves rather than to diversify their portfolios.
In this paper, we resolve this empirical puzzle by focusing on an additional channel through which branch banking could affect financial stability: increased competition. Our hypothesis is that, faced with heightened competition, banks that are only marginally profitable are forced out of the banking system either through merger or voluntary liquidation. As these weaker banks close, the overall stability of a state's banking system improves through consolidation. Thus, in the 1920s and 1930s, states allowing branch banking experienced lower failure rates without the branch banks themselves necessarily being the strongest banks.
Our hypothesis draws on the theoretical and empirical literature that examines the removal of legal restrictions on competition, and then links it to the literature on bank failures. Although policymakers often debate whether there are tradeoffs between competition and stability, surprisingly little theoretical or empirical research has analyzed these linkages in depth (Allen and Gale 2000).l Consistent with the hypothesis posited here, Berger and Hannan (1998) find that banks not exposed to competition are able to exercise monopoly power and tend to be less efficient than banks subject to more competition. When laws restricting competition are relaxed, bank profits generally decline. This has been found both within the United States (Amel and Liang 1997) and internationally (Claessens, Demirguc-Kunt, and Huizinga, 2001, Levine, 1996). Moreover, the increase in competition resulting from the removal of branching restrictions has been linked to the weeding out of weak banks (Jayaratne and Strahan, 1998, Stiroh and Strahan, 2003). We similarly argue that the expansion of branching in the 1920s facilitated an increase in competition. To help clarify the theoretical debate over the effects of competition on financial stability, we directly test how the growth of branching influenced bank competition and how this in turn affected bank failures.
Since our hypothesis emphasizes changes in the competitive environment induced by the onset of branch banking, it is necessary to test our model using data from a period when branch banking was expanding in scope. Moreover, because we want
1. "On the one hand, there are many models of competition in the literature including models of bank regulation in a competitive environment. On the other hand, there is a well-developed literature on bank crises... But there is little on the impact of competition on stability." (Allen and Gale, 2000, p. 268). Important exceptions are Koskela and Stenbacka (2000) and Matutes and Vives (2000); however, this newer theoretical literature presents conflicting views on how competition affects financial stability.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1295
uZ
m o
z
1900 1905 1910 1915 1920 1925 1930 Year
FIG. 1. Branches of U.S. Banks. Source: Federal Reserve Board of Governors (1931), Vol. 2. (Note: Home city indicates branches located in the same city as the bank's headquarters.)
to test how branching influences the stability of banking systems, we also need to examine a period when there were numerous failures. In this respect, the experience of the U.S. banking system from 1920 to 1930 is ideal since branching was expanding rapidly (Figure 1), and because the 1920s were characterized by a large number of bank failures (Figure 2).2 Finally, examining this period allows us to compare our
1 600
1400 * All Banks
e 1200 @ National Banks >
._ 1 000
L * L
O 800 | |
200 LIL LILILILILILII 1922 1923 1924 1925 1926 1927 1928 1929 1930
Year
FIG. 2. Bank Failures in the United States. Source: Federal Reserve Board of Governors (1943).
2. Chapman and Westerfield (1942) describe how the issue of branch banking gained national attention during the 1920s, in part because it was spreading rapidly in states such as California and prompting federal regulators to reconsider their longstanding prohibitions against it.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1296 : MONEY, CREDIT, AND BANKING
results to existing research on the Great Depression and resolve the paradox that currently exists in the literature.
Our hypothesis has several testable propositions. First, the expansion of branching should change the competitive environment. If branch banking removes weaker banks from a banking system, then states permitting branch banking should experi- ence higher merger and voluntary liquidation rates and lower entry rates (by new banks) than states prohibiting it. Second, over time more competition in states permitting branch banking should result in lower profit levels. Finally, if the competitive shakeout induced by branching stabilizes banking systems by removing weak banks from the system, then in the long run failure rates should be lower in states where branch banking was expanding. The link between branching, competition, and stability ought to be present even after controlling for any benefits to stability coming from improved geographical diversification of bank portfolios. We draw on the bifurcated nature of the dual banking system that existed in the 1920s to design a statistical test to discriminate between the effects of geographical diversification and competition due to branching.
Our empirical results support the predictions of the competition hypothesis. States in which branching was more prevalent experienced more mergers and voluntary liquidations during the 1920s. We also find that although there was significant consolidation in the banking sector in states allowing widespread branch banking, profits were lower on average in these states, suggesting that branching led to increased competition rather than monopoly power. To test whether branching re- duced failures, we first confirm that our data produce the usual state-level result that states allowing branching or those with more branch offices had lower failure rates. We then construct proxies for the portfolio diversification and competition channels of branching and test whether their inclusion affects this result. Our econo- metric evidence shows that, at least for national banks, the consolidation effects were quantitatively more important than increased portfolio diversification opportunities for banking stability during this period. These results suggest that, at the onset of the Great Depression, there were still many weak banks in states prohibiting branch banking; the real shock of the 1930s caused many of these to fail. However, in states that permitted branching, weak banks had been pruned from the system, and failures were consequently lower at the system-wide level. Thus, we resolve the paradox in the existing literature by showing that the expansion of branching improved stability at the statewide level through the competitive shakeout process without necessarily improving individual banks' ability to diversify away risk during a large shock such as the Great Depression.
The paper proceeds as follows. Section 1 discusses the previous literature on branching and financial stability. In Section 2, we present our hypothesis for resolv- ing the existing puzzle in the literature. The next section tests the consolidation hypothesis and some of its implications for bank competition and financial stability. Section 4 provides concluding remarks.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1297
1. THE EFFECTS OF BRANCH BANKING ON FINANCIAL STABILITY
An argument commonly articulated in the literature is that branch banking stabi- lizes banking systems by reducing their vulnerability to local economic shocks: branching enables banks to diversify their loans and deposits over a wider geo- graphical area or customer base.3 Restnctions on branching have been linked to the instability of banking systems. Calomins (2000) argues that bank failures were more prevalent in regions of the United States without branch banking as well as in countnes lacking it. Fnedman and Schwartz (1963) suggest that the absence of branching in the U.S. increased the seventy of the banking panics dunng the Great Depression. Moreover, they argue that the U.S. experience stands in contrast to Canada, which expenenced banking distress dunng the Depression but not wide- spread failures and a collapse of its banking system.4 The notion that branch banking stabilizes banking systems by increasing diversification opportunities is in fact an argument with old roots (Sprague 1903). In the 1920s, proponents of branch banking used this argument to encourage state legislatures to adopt laws legalizing branch banking (Preston, 1928, Southworth, 1928).
Research examining the effects of branching at the aggregate level generally supports the hypothesis that allowing branch banking increases systemic stability. Wheelock (1995) studies the eiTects of different state banking regulations on bank failures dunng the Depression and finds that states that allowed branch banking tended to have lower failure rates. Mitchener (2000, 2005) further examines state- and county-level bank failure rates. Controlling for economic fundamentals and differences in both state supervision and regulation, he also finds that states with legalized branching had lower failure rates dunng the Depression. Comparing 25 diiTerent countnes dunng the Great Depression, Grossman (1994) finds that countnes with large branching networks were less likely to expenence banking cnses.5 A1- though the studies that rely on aggregate data find a positive correlation between branch banking and financial stability, they do not establish the precise channel through which branching improved stability.6 It is therefore possible that the stability effects of branching are related to something besides or in addition to diversification.
3. Studies by Wacht (1968) and Lauch and Murphy (1970) find reduced variance in deposit flows for branch banks. Cherin and Melicher (1988) find that branching has moderating effects on asset returns.
4. Drummond (1991) and White (1983) make a similar argument. Kryzanowski and Roberts (1993), however, find that nationwide branch banking did not prevent banks in Canada from becoming techni- cally insolvent.
5. Many of these studies also include some measure of bank concentration as an explanatory variable. Although none of them provides an interaction term for branch banking and their measure of concentration, it is possible that these two effects worked together to influence failure rates. In Section 3, we relate our consolidation hypothesis to the issue of banking concentration. Our emphasis, however, differs in that we are testing whether changes in bank concentration are an outcome of branching laws and whether this in turn affected bank profitability.
6. Wheelock (1995) attributes the positive correlation between restrictions on branch banking and failure rates to limited diversification. Similarly Alston, Grove, and Wheelock (1994) also consider the impact of branching legislation on bank failures, and emphasize that branching may reduce a bank's susceptibility to distress in a particular area; however, they do not find that the ratio of (non-home- office) branches to total banks helps to explain the cross-state variation in failure rates during the 1920s.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1298 : MONEY, CREDIT, AND BANKING
Studies using data on individual banks operating in the 1920s and 1930s paint a diiTerent picture of the effects that branching has on the survivorship of individual banks, and they cast doubt on the common view that the stabilizing benefits of branching operated via increased diversification opportunities. Calomins and Mason (2000) find that, dunng the Depression, Federal Reserve members that were branch banks tended to fail sooner than unit banks. Also, using data on individual state banks from this period, Carlson (2004) examines three states where branch banking was relatively widespread and finds that branched banks were more likely to fail than unit banks. Furthermore, he rejects some potential reasons for this phenomenon, including insufficient diversification and over-expansion on the part of banks. Instead, he finds that branch banks used diversification to reduce their reserves rather than to lower the risk of their portfolios a strategy that worked poorly during the global shock of the Great Depression.
Because it is difficult to reconcile the findings based on aggregate data (which report a negative relationship between bank failures and statewide branch banking) with the empirical results from studies using individual bank data (which are inconsis- tent with the view that the source of stability was improved opportunities for diversification), this article proposes an additional channel through which branching could have affected stability: competition. Although they do not construct a formal model, Berger, Demsetz, and Strahan (1999) argue that increased competition results in the purging of ineEcient banks from the banking system. Consistent with these ideas, Jayaratne and Strahan (1998) and Stiroh and Strahan (2003) find that the branch-banking reform that began in the United States in the 1980s resulted in the removal of weaker banks from the system. Additionally, Koskela and Stenbacka (2000) suggest that greater competition decreases interest rates and increases the likelihood that borrowers are able to remain solvent and repay their loans. These stud- ies suggest that the introduction of competition (in our case, driven by the growth of branch banking) may improve the stability of banking systems. On the other hand, Matutes and Vives (2000) argue that raising the level of competition causes an increase in failures as lower profits resulting from competition encourage banks to take on more risk.7 We are not aware of any previous studies that systematically test the effects of competition on the stability of banking systems in particular.8
2. BRANCH BANKING AND COMPETITION IN THE 1920S
As Figure 1 shows, the total number of branches operated in the United States nearly tripled between 1920 and 1930, rising from 1281 to 3518. Many of these branches were located in home-oice cities and the number of these branches
7. Demsetz and Strahan (1997) find that consolidation for bank holding companies (BHCs) enhanced diversification following regulatory reform in 1994, but that larger BHCs then operated with lower capital ratios and increased their risky lending.
8. Kaminsky and Schmukler (2002), however, compare broad financial systems of different countries between the early 1970s and the late l990s and find that, although reducing barriers to external competition initially results in some turmoil, the long run effect of deregulation is increased stability.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1299
more than doubled, increasing from 508 in 1920 to 1131 in 1930 (Federal Reserve 1931, Vol. 2). We hypothesize that this expansion of branching networks increased the level of competition in states that allowed branching to occur. As a result of this dynamic process, banks that were only marginally profitable prior to the increase in competition would become unprofitable due to the increase in competition. In turn, these banks would likely merge with existing banks or voluntarily liquidate.9 Also, because it is less costly to open a branch than a new bank, it is likely that fewer new banks would be able to find an unexploited profitable niche and enter the market despite the fact that regulatory barriers to entry have been removed. With the exit of the weakest banks, the economic viability of the average bank would increase and the rate of failure for banks within that state would decline.l° The idea that the removal of bamers to competition would lead to a reduction in the number of banks in the banking system is consistent with the model by Economides, Hubbard, and Palia (1996).
Why should these competitive forces apply to the introduction of new branches (as a result of legal changes) and not simply to the emergence of new unit banks in the 1920s? First, banks with branches were more cost-effective, since some jobs at difFerent branches could be consolidated and performed at the head office, thus reducing employment costs (Federal Reserve 1931, Vol. 2, p. 224.) Also, start-up costs were lower, and in some states, regulators required less capital for new branches than for new unit banks (Southworth 1928). Second, new branches that were set up in previously restricted markets may have been more adept at realizing higher rates of return than comparable new unit banks since branches could transfer deposits out of the local market to regions where capital was in higher demand.ll The ability to obtain a cost advantage through branching and realize higher rates of return made entry into existing local markets easier for branch banks than new unit banks. Indeed, branch banking may have been instrumental in bringing banking and banking competition to small towns. Calomiris (2000, Chapter 1) makes a similar argument. In 1931 (the only year for which we have so far been able to locate the distribution of branches by town size), nearly half of all branches outside the home-office city were located in towns of less than 2500 people (Table 1). Figure 3 shows the locations of branches outside the home-office city.
Laws that permitted statewide branching applied only to state-chartered banks (which were regulated by state banking departments) while the data used below to test this hypothesis are for national banks (which were regulated by the C)ffice of the
9. Wheelock and Wilson (2000) find that, during the 1980s and 1990s, inefficiency reduced the likelihood that a bank would be acquired Carlson (2004), however, finds that during the early 1930s, acquired banks were generally weaker than other banks.
10. By limiting the development of secondary markets, entry barriers such as restrictions on branching could also prevent productive assets of weak banks from being digested or taken over by more efficient banks. Without the existence of local competition to absorb bank assets, weak banks may have been forced to sell productive assets in thin markets at fire sale prices or not at all, in turn increasing the likelihood of bank failures within the system.
11. Morgan, Rime, and Strahan (2003) provide another channel through which diversified banks improve stability shifting capital between regions to dampen economic shocks.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 1
THE DISTRIBUTION OF BANK BRANCHES BY TOWN POPULATION (1931)
Branches Outside Branches Inside All Branches Home-Office City Home-Office City
Town Population Number Percent of Total Number Percent of Total Number Percent of Total
Under 500 189 16.32 2 0.09 191 5.73 500-1000 173 14.94 0 0.00 173 5.19 1000-2500 207 17.88 7 0.32 214 6.42 2500-5000 134 11.57 7 0.32 141 4.23 5000-10,000 107 9.24 9 0.41 116 3.48 10,000-25,000 91 7.86 27 1.24 118 3.54 25,000-50,000 46 3.97 63 2.90 109 3.27 50,000-100,000 60 5.18 132 6.07 192 5.76 100,000+ 151 13.04 1929 88.65 2080 62.39 Total 1158 100 2176 100 3334 100
Source: Federal Reserve Board (1931), Vol. 2.
---*- v@E
1300 : MONEY, CREDIT, AND BANKING
Comptroller of the Currency). The logic of our hypothesis, however, still applies to national banks since they would be subject to increased competition from the state banks that were allowed to establish branches. 12 Indeed, it may actually be better
otet }t "N
vY lAg6 e
WTAXW<" - azs : i - r4} YK o - Y o
FIG. 3. Branches of National and State Banks outside the City of the Home Office (December 31, 1931) Source: Federal Reserve Board of Governors (1931, Vol. 2, p. 17). (Note: In California there are numerous branches in the metropolitan areas centering around San Francisco and Los Angeles, but technically outside their city limits. On the map the dots extend considerably beyond the territory in which the branches are actually located around these cities.)
12. Additionally, state banks that converted into national banks were permitted to keep branches they had established while they were state banks, enabling branch banks to become national banks through a legal technicality.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1301
for testing the competition hypothesis since the vast majority of national banks had no branches outside the home-office city over our period and therefore enjoyed no effective geographical diversification benefits. Excluding California, the country's nearly 8,000 nationally chartered banks operated 18 branches outside the home- office city in 1925. In 1930 this number was 27 (Federal Reserve Board of Governors 1943). Since our sample consists of national banks, and only a few of them branched outside the headquarter city, any observed or unobserved channels through which branching might affect both competition and diversification are likely to be less of a problem. Finally, using a sample consisting of national banks has the additional advantage that the regulatory environment and the attitudes of regulators regarding bank mergers were more uniform than those concerning state banks.
It has been shown by Calomiris and Mason (2000), among others, that less profitable banks were more likely to fail during the Great Depression. A key compo- nent of our hypothesis is that, in the long run, fewer of these banks would exist in states that allowed branch banking because the competitive pressures associated with the rise of branch banking networks prior to the Depression would have forced weak banks to exit from the banking system earlier in the decade. In states without the competitive pressures of branch banking, more weak banks would still exist at the start of the Depression and these states would therefore have more banks that would be likely to fail during the subsequent downturn. This interpretation is consistent with the findings in Mitchener (2000, 2005) and Wheelock (l99S): states allowing branch banking had lower failure rates than those prohibiting it. And it would also be true even though it was not necessarily the case that branch banks were the survivors.
The alternative hypothesis, that a banking system with branching is more diversi- fied and therefore more stable than a banking system with only unit banks, presup- poses that the two systems are in equilibrium. The banking system of the United States during the 1920s and 1930s, however, was in the process of transition. As noted above, branching was expanding rapidly in some states and the total number of banks was declining steadily from 29,715 in June 1920 to 23,855 by June 1930 (Federal Reserve 1943). The growth in branch banking during the 1920s was facilitated by a variety of legal and technological changes. In 1922, the Comptroller ruled that national banks could, "under the law, establish agencies, teller windows, or additional offices within the city of the parent bank provided state banks were permitted to operate branches in that state (Chapman and Westerfield, 1942, p. 97)," although these offices could not issue loans and were not full-fledged branches. Possibly seeing these offices as one step away from approval by the Comptroller of full-fledged branch banking for national banks, state banks may have responded by increasing their branching networks in order to compete with national banks. Relationships with correspondents were weakened due to amendments of the Federal Reserve Act in 1917 (which put check clearing in the hands of the Federal Reserve and required national banks to hold their entire reserve requirement at the Federal Reserve), possibly inducing banks to pursue the loss of deposits by buying banks and converting them to branches. Dramatic improvements in road networks and
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1302 : MONEY, CREDIT, AND BANKING
improvements in telephone networks likely improved the ability of managers to oversee branch networks. And rising urbanized populations in the newer cities of Los Angeles and Detroit (both augmented by the rise of the automobile, the former via the conversion of rails to roads and the latter which served as the industry's manufacturing hub) also led to an increase in demand for banking services. Both cities realized dramatic increases in the number of branch banking oices in the 1920s. It is also likely that more banks were willing to develop branch networks during the 1920s as they observed the success of branch banks in places such as California.l3 It is this structural change in the banking system wrought by these factors that motivates our hypothesis. The key role played by this change in the banking system suggests that our explanation is specific to the United States and may not apply to other economies during the 1920s, such as Canada and the United Kingdom, which had removed barriers to branching earlier and likely had completed the transition to a branch banking system. That is, as branching expanded rapidly in some U.S. states in the 1920s shakeout took place, initially causing exit and later reducing bank failures.
Because the growth in branching is attributable largely to shifts in the relationships of banks with each other, technological progress, and population/economic growth rather than changes in regulation specifically concerning branching, we use changes in actual branching activity over time, and across states, rather than changes in branching laws, to examine the effect of branching on a state's banking system. We argue that using actual branches provides a more complete picture of the effect that branching might have on the competitive environment because the laws regulating the establishment of branches varied substantially so that variables categorizing regu- lations capture quite different situations.l4
One notable change in the legal environment in the 1920s was the McFadden Act of 1927, which allowed national banks to establish local branches in the city of their home office if state law allowed branching. However, the Act imposed several restrictions: national banks could open no new branches in cities with fewer than 25,000 people, only two branches in cities with populations between 50,000 and 100,000, and at the discretion of the Office of the Comptroller of the Currency for cities of over 100,000 (Tippetts 1929).
It should be noted that the expansion of branching (and consequently the consolida- tion of the banking system) in the 1920s, which was driven by the establishment of the Federal Reserve, technological changes, population growth, and economic
13. California's branching network developed more quickly and extensively than any other state in the decade, in part due the financial entrepreneur, A.P. Giannini, who created an extensive branching network for the Bank of America. This spurred competing large banks in California to develop branching networks, especially in Los Angeles, to fight Bank of America's geographical expansion. Given the size of Bank of America's branching network in California, a change in its charter in 1927 (from state to national) could have a large impact on the results in our paper, so later we test whether our results on stability are sensitive to its inclusion.
14. For example: Massachusetts allowed trust companies to have one branch in the same city as the home office; New York allowed unlimited branching in the city of the bank's home office if the population of the city exceeded 50,000; Louisiana allowed banks to have up to two branches, which could be located in the parish of the home office; and California allowed statewide branching.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1303
growth, is quite different from expansion of branching in the 1980s and 1990s, which appears to have been more strongly influenced by changes in regulation (Stiroh and Strahan, 2003, Kroszner and Strahan, 1998). Thus, focusing on a measure of branching activity is likely to better capture the effect of branching on the banking system during the 1920s than the shifts in regulation that have played a prominent role in dealing with the expansion of branching in recent periods.
Our hypothesis, which emphasizes how the expansion of branch banking within a state increases the competitive pressures on inefficient banks and can induce them to merge or voluntarily liquidate, is consistent with recent research examining the effects of bank deregulation (DeYoung, Hasan, and Kirchhoff, 1998, Berger, Demsetz, and Strahan, 1999, Jayaratne and Strahan, 1998, Stiroh and Strahan, 2003). However, it stands in contrast to one of the longstanding populist arguments lodged against branch banking. Opponents of branch banking have often complained that it was a form of cartelization that would result in consolidation of the industry and reduced competition, and that its growth would reduce the viability of businesses in small communities by siphoning funds to urbanized areas. Such sentiments were widely expressed in the first quarter of the 20th century when branching was spreading rapidly.ls
While the growth of branch banking may lead to consolidation, the effects on competition are not as clear as opponents of branching suggest. In fact, the economic theory or private-interest view of regulation argues that branching restrictions are used to protect inefficient, local monopolies and restrict competition.l6 The result of these intrastate regulations was less than full-scale competition in local deposit and loan markets. Chapman and Westerfield (1942, p. 233) described the situation in the 1920s and 1930s:
"Country bankers foresee danger to themselves in the possibility of inroads into their areas of operation, should the larger institutions of the cities be permitted to establish branches and compete with them in their area on equal terms. They know that such a policy would result in a reduction of interest rates in their towns and that their chances for the profitable use of their funds might be somewhat diminished unless they were prepared to go as far as their new nvals in serving customers cheaply. The alleged apprehension of unit bankers as to the monopolistic character of branch banking is, to say the least, selfish. What really motivates them is their desire to preserve their local monopolies and escape the competition of the more effective branch banks."
Our hypothesis is sympathetic to this view of regulatory impediments. It is also analogous to one that has been made in the context of the global financial services industry: just as foreign banks have brought necessary competition to inefficient domestic markets (Levine, 1996, Claessens, Demirguc-Kunt, and Huizinga, 2001, Folkerts-Landau and Lindgren, 1998), branch banking can introduce greater competi- tion to local markets, improving the cost and delivery of services to customers
15. See, for example, the discussion in Chapman and Westerfield (1942, p. 10) for examples of this view.
16. See Kroszner and Strahan (1998, 2000), Mitchener (2000), and Chapman and Westerfield (1942).
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1304 : MONEY, CREDIT, AND BANKING
and the safety of the banking system by forcing ineEcient banks to merge or go out of business.
3. TESTING THE COMPETITION/CONSOLIDATION HYPOTHESIS
This section tests several predictions of our hypothesis using data on national banks from the 1920s and the first two years of the Depression. First, we test whether the number of mergers and voluntary liquidations was higher (and the number of entries by new banks was lower) in states where there was more branch-banking activity. We then test whether other factors related to competition and consolidation, including the number of banks per capita and the profitability of different banks, are related to the extent of branch banking in a state. Finally, we test whether competition, induced by branching, is better at accounting for the variation in failure rates across states than the diversification argument.
3.1 Mergers, Voluntary Liquidations, and Entry in Branch Banking States Our first test examines industry consolidation whether there were more mergers
and voluntary liquidations for national banks in states that allowed branch banking- and whether there were also fewer new banks established in these states.l7 Table 2 summarizes the rate of entry of new national banks and rates of exits of national banks over the sample period of 1922-30 and groups by whether states permitted branch banking. (Appendix A describes the sources for our data. Detailed information on state branching laws is shown in Appendix B.) As the last column of the table shows, during the 1920s, states permitting some form of branching averaged somewhat more mergers and voluntary liquidations than states that prohib- ited branching.
The tabular results for the effects of branching on consolidation are tested more formally by regressing the number of exits and entries of national banks on a measure for branching activity in the state. Using observations on each state i, we estimate the following function:
COMPETITIONi = f l i l BRANCHi + :2BANKS/CAPi ( 1 )
+ GRYi + 4DEPINSi + :5BANKSi},
where COMPETITION is specified as: (1) the number of mergers, (2) voluntary liquidations, or (3) new branches for national banks.l8 BRANCH is a measure of
17. Wheelock (1993, p. 815) suggests that these changes may have occurred in the 1920s, but does not formally test this notion: "Like most Midwestern states, Kansas was a unit banking state during the 1 920s, with over 1,000 small banks in operation...Had branching restrictions been removed those counties might have experienced greater consolidation, through either mergers or failures."
18. It should be noted that entry of national banks is an imperfect measure of total entry, as one would also expect that the entry of state banks would be affected by branching. Additionally, some banks may prefer to enter the banking system as state banks in states allowing branching in order to establish branches, which might lead to measurement error in the dependent variable.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 2
SUMMARY STATISTICS FOR NATIONAL BANKS
1922 1923 1924 1925 1926 1927 1928 1929 1930 Average
SOURCES AND NOTES: Rates are computed using data from the Annual Report of the Comptroller of the Currency (1922-1930). Rates are expressed as percentages of existing banks. Within-category rates are unweighted averages of state rates. The last column shows the average of the yearly observations over the entire sample period, 1922-30.
the extent of branching in the state; BANKS/CAP is the number of state and national banks per person, AGRY is the share of a state's income from agriculture in 1920, BANKS is the log number of national banks, and DEPINS is a dummy variable that equals one for states with deposit insurance systems.
Since exits and entries take on discrete integer values and are bounded below by zero, we analyze the number of exits and entries employing count data analysis, where the data are pooled across states. Count data analysis assumes that the dependent variable was generated through a Poisson process. Determining the effect of the independent variables shown in Equation (1) involves maximizing the log- likelihood function:
n
ln L(,) = , {yiXi ,B-exp(Xi 13)-ln Yi ! }, (2) i=l
MARK CARLSON AND KRIS JAMES MITCHENER : 1305
Unit Number of Banking states
Average number of banks
Average voluntary liquidation rate
Average merger rate
Average entry rate
Allows Number Some Form of states of Branching
Average number of banks
Average voluntary liquidation rate
Average merger rate
Average entry rate
29
157.9
2.2
2.1
1.9
19
169.1
2.6
2.4
1.9
30 30 28 28 28 29 30 29 28
175.8 173.2 165.8 165.6 155.4 147.6 143.3 142.3 138.2
1.4 1.7 1.8 2.4 1.7 2.0 2.2 3.5 4.7
1.2 2.0 1.8 2.2 1.9 2.3 2.2 3.5 4.0
2.4 1.8 1.5 2.3 1.9 1.2 1.3 2.3 2.2
18 18 20 20 20 19 18 19 20
168.6 169.1 171.6 172.9 167.3 172.8 176.7 165.9 157.5
1.8 1.5 2.6 1.6 3.4 2.0 2.8 4.0 4.9
1.4 1.6 2.6 1.2 2.8 2.2 3.1 4.4 3.8
2.6 1.9 2.5 2.5 1.8 1.2 1.5 2.0 0.8
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 3
EFFECT OF INITIAL LEVEL OF BRANCHING ON VOLUNTARY LIQUIDATIONS, MERGERS, AND ENTRY
Average Voluntary Liquidations Average Mergers Average New Banks
Coefficient Estimate S.E. Coefficient Estimate S.E. Coefficient Estimate S.E.
Intercept -6.80*** (1.17) -6.77*** (1.16) -4.63*** (1.09) Ratio of Branches to 5.02*** (1.56) 4.75*** (1-54) 1.47 (1.66)
Total Bank Offices
(1922 Value)
Banks Per Capita 0.58** (0.23) 1.26*** (0.18) 1.15*** (0.17) Share of State Income 0.03** (0.02) 0.04** (0.02) 0.00 (0.02)
from Agriculture
Log Number of Banks 1.26*** (0.18) 1.26*** (0.18) 1.15*** (0.17) Deposit Insurance -0.46 (0.39) -0.49 (0.38) 0.83** (0.40) Dispersion 0.12 (0.08) 0.10 (0.07) 0.12 (0.10)
Observations 48 48 48 Log-likelihood - 96.9 -91.1 -96.1
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Standard errors are in parentheses. Data on voluntary liquidations, mergers, and entnes of national banks are from the Annual Report of the Comptroller of the Currency (1923-30). Population and income shares are from 1920 Census report. The log of the number of banks is the average number of national banks for 1923-30. Branch offices and total bank offices include both state and national banks. The dependent variable is the average count of voluntary liquidations, mergers, and entries of national banks in a state for the period 1923-30.
where y is the dependent variable vector, X is the matrix of independent variables, a is the coefficient vector, and n is the number of observations.l9 When analyzing counts, it is important to control for the size of the population from which the counts are produced. In the Poisson and negative binomial distributions the rate of "arrival" is constant so that the number of arrivals depends on the population size. Increasing the population by a given percentage should increase the number of arrivals by the same percentage. To correct for this, we include the log of the number of banks to account for the size of the "at-risk population" of banks.20
Count data analysis has several advantages over other estimation strategies: it treats observations with the value of zero as containing important information and takes into account the granularity in the data (the fact that the number of exits and entries can only take on whole numbers). Moreover, since entries and exits are truncated at zero rather than censored at zero, using an alternative estimator such as Tobit would produce biased estimates. When there is some evidence of overdispersion (the variance significantly exceeds the mean) in our sample, we employ a negative binomial distribution rather than the Poisson distribution. The coefficients can be interpreted as the percentage change in the dependent variable (number of entries or exits) due to a one-unit change in the independent variable.
Table 3 displays the results from cross-sectional regressions of the average number of mergers, voluntary liquidations, and entries from 1923 to 1930 on the ratio of branch oices for state and national banks to total branch and bank offices for state and national banks in 1922 (BRANCH) and the other explanatory variables described
l9. The likelihood function is based on a Poisson process, although this can be modified; it involves taking the factorial of the dependent variable (noting that the factorial of zero is one).
20. For a detailed treatment of count data analysis see Cameron and Trivedi (1998).
1306 : MONEY, CREDIT, AND BANKING
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 4
EFFECT OF AN INCREASE IN BRANCHING ON VOLUNTARY LIQUIDATIONS, MERGERS, AND ENTRY
Count
Average Voluntary Average Liquidations Average Mergers New Banks
(192S30) (192S30) (192S30)
CoeEcient Estimate S.E. Coefficient Estimate S.E. Coefficient Estimate S.E.
Intercept - 6.18*** (0.82) - 6.03*** (0.82) - 4.79*** (1.28) Change in the Ratio of 6.62*** (1.03) 6.66*** (1.03) 3.15 (2.48)
Branches to Total
Bank Of fices (1922-25) Banks Per Capita 0.32 (0.23) 1.19*** (0.13) 1.13*** (0.23) Shre of State Income 0.04*** (0.01) 0.04*** (0.01) - 0.01 (0.02)
from Agriculture
Log Number of Banks 1.21*** (0.13) 1.19*** (0.13) 1.13*** (0.23) Deposit Insurance - 0 50* (0.26) - 0 49* (0.26) 1.39*** (0.50) Dispersion N.A.A N.A.A 0.26 (0.17)
Observations 48 48 48 Log-likelihood - 165.9 - 150.3 - 48.4
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1, 5, and 10 percent level, respectively. The symbol (^) indicates that we reject overdispersion in the data and estimate the regression using a Poisson distnbution. Standard errors are in parentheses. Data on voluntary liquidations, mergers, and entnes of national banks are from the Annual Report of the Comptroller of the Currency (1926-30). Population and income shares are from 1920 U.S. Census. Banks per capita is the 1920 value. The log of the number of banks is the average number of national banks for 192S30. Branch offices and total bank offices include both state and national banks. The dependent vanable is the average count of voluntary liquidations, mergers, and enties of national banks in a state for the penod 192S30.
in Equation (1). This specification enables us to examine whether cross-state variation is sufficient to identify a relationship between branching and our measures of exit. By holding our measure of branching at the 1922 level, we also alleviate concerns about the endogeneity of this variable. The estimated coefficients support the hypoth- esis that branch banking increases mergers and involuntary liquidations, leading to consolidation in a state's banking system. Testing these three coefficients jointly suggests that branching results in a net reduction of the number of banks (the Chi-squared test statistic has a p-value of 0.02). During the 1920s, states that had more extensive branch-banking networks had more voluntary liquidations than states that allowed only unit banking (Column 1). The point estimate suggests that an increase in the ratio branch offices to total offices by 0.1 (about one standard deviation) resulted in roughly a 50% increase in the number of voluntary liquidations. Column 2 also shows that states with more branching activity had significantly more mergers than unit banking states during the 1920s. The coefficient on the extent of branching suggests that its effect on the number of mergers is similar to its effect on voluntary liquidations. These results are consistent with White (1985), who describes the wave of bank mergers in the 1920s and suggests that the increase in mergers and the ability to branch may have been related. In contrast to the findings for voluntary liquidations and mergers, branching activity appears to have had little impact on entry by new national banks (Column 3).
We next examine whether there is a sufficient expansion of branching activity in the early 1920s to induce exit as our hypothesis suggests. As Appendix B indicates, some states had passed branching laws before 1920. So even though the evidence
MARK CARLSON AND KRIS JAMES MITCHENER : 1307
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 5
EFFECT OF BRANCHING ON VOLUNTARY LIQuIDATzoNs, MERGERS, AND ENTRY RJ8ING A POOLED SAMPLE
Voluntary Liquidations (1923-30) Mergers (1923-30) New sanks (1923-30)
Coeffscient Estimate s.E. Coefficient Estimate S.E. Coefficient Estimate S.E.
Intercept -5.2188* (0.39) -5.14*** (0.39) -4.08*** (0.44) Ratio of Branches to Total 2.41*** (0.39) 2.34*** (0.39) 0.51 (0.45)
Bank Offices (lagged) Banks Per Capita 0.66* (0.36) 0.67* (0.35) -0.99*** (0.07) Share of State Income 0.04*** (0.01) 0.04*** (0.01) ° °° (0.01)
from Agriculture Log Number of Banks 1.01*** (0.06) 1.02*** (0.06) 1.07*** (0.07) Deposit Insurance -0.72*** (0.18) -0.69*** (0.18) 0.81*** (0.20) 1924 0.20 (0.19) 0.01 (0.19) -0.21 (0.21) 1925 -0.02 (0.20) -0.25 (0.20) 0.13 (0.20) 1926 0.21 (0.19) 0.05 (0.19) -0.08 (0.21) 1927 0.25 (0.19) 0.18 (0.19) -0.38* (0.21) 1928 0.19 (0.19) 0.04 (0.19) -0.38* (0.22) 1929 O.59*** (0.19) 0.52*** (0.18) -0.02 (0.21) 1930 0.88*** (0.19) 0.55*** (0.18) -0.27 (0.22) Dispersion 0.39 (0.06) 0.39 (0.06) 0.54 (0.08)
Observations 384 384 384 Log-likelihood -1031 -964 -1086
NOTES: The symbols (***), (**) and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Standard elTors are in parentheses. Data on voluntary liquidations, mergers, and entnes of national banks are from the Annual Report of the Comptroller of the Currency (1922-30). Population and income shares are from 1920 Census report. The first year in each sample penod is the omitted year from the regression. Branch offices and total bank offices include both state and national banks. The dependent vanable is the count of voluntary liquidations, mergers, and entnes of national banks in a state per year.
from the tables and figures suggests that branching was expanding rapidly during the decade, in Table 4 we explicitly test whether the change in this branching ratio between 1922 and 1925 is associated with a greater average number of mergers and voluntary liquidations between 1926 and 1930. The statistically significant coeffi- cient on the change in the branching ratio shown in Columns 1 and 2 suggests that the expansion of branching in the early 1920s had a significant role in shaking out the banking system even though some branching laws had been passed earlier.
The results from Tables 3 and 4 are also consistent with Wheelock (1993), who showed that the overall percentage change in banks per capita in the 1920s was greatest in states where branching expanded the most.21 However, the decomposition presented here provides the further insight that branching encouraged banks to exit in a variety of ways. That voluntary liquidations increased suggests that the decline in banks per capita was not simply due to widespread purchases of banks by a few industry leaders. Rather the trends exhibited in the regressions are consistent with an overall increase in competition where branching was permitted.
We also estimated several alternative specifications in order to examine the ro- bustness of the results, including pooling the data and using the legal status of branching in a state rather than the extent of branching. Table 5 tests whether the
21. And they are also consistent with Berger, Kashyap, and Scalise (1995), who show that the introduction of nationwide banking beginning in the 1980s accelerated the reduction in the number and market share of small bank organizations for the period of recent regulatory liberalization.
1308 : MONEY, CREDIT, AND BANKING
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 6
EFFECT OF STATE BRANCHING LAWS ON VOLUNTARY LIQUIDATIONS, MERGERS, AND ENTRY USING
A POOLED SAMPLE
Voluntary Liquidations (1922-30) Mergers (1922-30) New Banks (1922-30)
Coefficient Estimate S.E. Coefficient Estimate S.E. Coefficient Estimate S.E.
lntercept -4.64*** (0 40) -4.79*** (0.41) -3.87*** (0.43) Law Allows Branching 0.24* (0.12) 0.21* (0.12) -0.10 (0.13) Banks Per Capita -0.12 (0.37) -0.04 (0.37) -1.61*** (0.07) Share of State Income 0.03*** (0.01) 0.038** (0.01) ° °° (0.01)
from Agriculture
Log Number of Banks 0.968** (0.06) 0.99*** (0.06) 1.08*** (0.07) Deposit Insurance -0.45** (0.19) -0.45** (0.19) 0.968** (0.19) 1923 0.19 (0.22) 0.31 (0.21) -0.06 (0.21) 1924 0.38* (0.22) 0.34 (0.21) -0.28 (0.22) 1925 0.18 (0.22) 0.09 (0.22) 0.05 (0.21) 1926 0.40* (0.22) 0.38* (0.21) -0.15 (0.22) 1927 0.46** (0.21) 0.55*** (0.21) -0.45** (0.22) 1928 0.43** (0.22) 0.43** (0.22) -0.45** (0.22) 1929 0.81*** (0.21) 0.89*** (0.21) -0.09 (0.22) 1930 1.07**8 (0.21) 0.89*** (0.21) -0.36 (0.22) Dispersion 0.56 (0.07) 0.55 (0.07) 0.60 (0.08)
Observations 480 480 480 Log-likelihood -1065 -968 -1603
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Standard errors are in parentheses. Data on voluntary liquidations, mergers, and entries of national banks are from the Annual Report of the Comptroller of the Currency (1922-30). Data on branching laws are from the Federal Reserve Board of Governors (1931) and population is from the 1920 Census. The first year in each sample penod is the omitted year from the regression. The dependent vanable is the count of voluntary liquidations, mergers, and entnes of national banks in a state per year.
annual number of mergers, voluntary liquidations, and entries from 1923 to 1930 are related to level of branching activity lagged one year, where the data are pooled over the entire sample period.22 The measure for the extent of branch banking in a state is lagged one year to reduce the potential for endogeneity.23 We include year dummies in Table S (and in other pooled specifications in the paper) to control for any time-specific effects. This allows us to account for the changes in the macroeconomic environment and any residual effects of the changes in the regulatory environment, such as the McFadden Act, not captured by our degree of branching measures. The coefficients on the extent of branching are again positive and statisti- cally significant at conventional levels for voluntary liquidations and mergers.
As a final robustness test regarding the exogeneity of our branching variable, in place of the measure based on the extent of branching, we use dummy variables indicating whether the state permitted branch banking. In a pooled regression, we find that states permitting statewide branching had more mergers and consolidations (Table 6). We note, however, that the rest of analysis in this article focuses on our
22. There is, however, substantial variation over time in our measure of the extent of branching. On average, it changes 75% between 1922 and 1930 and by 30% between 1926 and 1930.
23. As we indicated in the previous section, since technological and legal changes were driving the increase in branch banking in the 1920s, the lagged measure of the extent of branching seems plausibly exogenous.
MARK CARLSON AND KRIS JAMES MITCHENER : 1309
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1310 : MONEY, CREDIT, AND BANKING
preferred measure, the extent of branching activity in a particular state, because it allows us to sidestep two definitional problems which complicate and potentially muddle the interpretation of regression results based on the legal status of branching in a state: (1) some states in the 1920s permitted limited branching, but what they permitted differed markedly and (2) some states differed in terms of what was practiced by regulators (de facto) and what was on the books (de jure). (See Appendix B for more details.)
3.2 Consolidation and Increased Competition versus Monopoly Power Although the results presented in the previous section support the view that
branch-banking activity increases mergers and exits, it is unclear whether this consol- idation led to a more competitive banking system or to a monopolistic banking system. For example, when banks are forbidden from establishing branches, it artificially segments the market and enables unit banks to develop monopolies within particular localities. The immediate impact of branching is that it introduces competition into these geographically segmented markets. If banks that were pre- viously insulated from the cool winds of competition are acquired or forced out of the market as a result of this increased competition, it is possible that the surviving banks in turn acquire monopoly power. This was one of the concerns of those who lobbied against more permissive branching laws in the 1920s: the end result of liberalization would simply be further concentration in the banking industry. To investigate further how statewide branching affects the competitive environment of a state's banking system, we examine the relationship between branching and two additional measures: banking sector concentration and bank profits.
We first test whether states allowing branch banking had more concentrated banking sectors by regressing measures of bank concentration on the extent of branch banking. The first measure of concentration is essentially an annual Herfindahl index of national banks. The Federal Reserve published data on the number of banks in diiTerent size categories with different levels of profitability for each state, for each of the years 1926-30 (Federal Reserve Board of Governors 1931).24 The index is constructed as follows: the number of banks is multiplied by the means of the size category; these groups are aggregated to get the total value of loans and investments in the state. The mean of each size category is divided by total state assets, squared, and multiplied by the number of banks in each size category. These fractions are then added together to obtain the index.25 Since the Federal
24. The size categories, based on loans and investments, are: category 1, under $150,000; category 2, $150,000-$250,000; category 3, $250,000-$500,000; category 4, $500,000-$750,000; category 5, $750,000-$1,000,000; category 6, $1,000,00(}$2,000,000; category 7, $2,000,000-$5,000,000; category 8, $5,()00,000-$10,000,000; category 9, $10,000,000-$50,000,000; category 10, $50,000,000 and over. We employ the profit information below.
25. Mathematically, we can write this is as:
p meanvalueofloans (i) 1
i=siEroup l [Li (mean value of loans(i))(banks in category(i))] | g P
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 7
BRANCHING AND INDUSTRY CONCENTRATION
Herfindahl Index 4-Firm Index Banks Per Capita
Coefficient Estimate S.E. Coefficient Estimate S.E. Coefficient Estimate S.E.
Intercept 1084*** (101) 0.21*** (0.02) 38.61*** (2.44) Ratio of Branches to 1088*** (301) 0.44*** (0.06) -69.65*** (6.34)
Total Bank Offices
(lagged)
1924 0.01 (0.03) -2.18 (3.39) 1925 0.02 (0.03) -3.07 (3.39) 1926 0.02 (0.03) -4.20 (3.40) 1927 60 (139) 0.03 (0.03) -5.31 (3.40) 1928 50 (139) 0.04 (0.03) -6.27* (3.40) 1929 101 (139) 0.04 (0.03) -6.97** (3.40) 1930 150 (139) 0.05 (0.03) -7.98** (3.40)
Observations 239 383 383 Adjusted R-square 0.04 0.14 0.25
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5No, and 10% level, respectively. Estimated using a pooled sample. Data on the number of banks, size distnbution, and number of branches are from the Federal Reserve Board of Governors ( 1931 ) . Four-firm concentration index denved from Polk 's Bank Directory (various years) and the Comptoller of the Currency (various years) . Population is from the 1920 Census. The dummy for the first year in each time penod is the omitted year. Banks per capita and the four- firm index include state and national banks, whereas the Herfindahl index consists only of national banks. The dependent variables are concentration indexes.
Reserve data used to construct this index are only available for 1926-30, the results for the Herfindahl index are presented for this period. The second measure is a four-firm concentration ratio based on the deposits in the largest four commercial banks (state or national) in each state for each year of our sample. This measure allows us to examine a longer sample period (1923-30) as well as a measure based on the liabilities of banks. The third measure is simply the number of banks per capita (both state and national commercial banks), also calculated for 1923-30. The sample consists of annual observations on each state.
Analysis using the ratio of branch oEces to total bank offices of state and national banks in 1922 as our measure of branching indicates that states with more extensive branching had more concentrated banking sectors-higher Herfindahl scores, more deposits in the largest four banks, and fewer banks per capita.26 We find similar results if we pool the data and regress measures of bank concentration on the ratio of branch offices to total offices lagged one year (Table 7). Thus, regardless of the measure, states with more extensive branching tended to have more concentrated banking systems.
Next, we examine how branch banking affected bank profits. Since national banks were all regulated at the national level by the Office of the Comptroller of Currency and were largely restricted in their ability to have branches, differences in the level of competition that they faced would be a significant factor in creating differences
26. Our Herfindahl index is admittedly imperfect since it does not include state banks (comparable data do not exist), and state banks were on average smaller. The average assets (loans and investments) for state banks in 1926 was $1,272,183, whereas for national banks it was $2,403,295; state banks had 57% of total assets in the commercial banking system in this year. Nevertheless, since our other two measures of concentration also show that states with more branching activity had more concentrated banking systems, we are confident that our result is robust.
MARK CARLSON AND KRIS JAMES MITCHENER : 1311
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1312 : MONEY, CREDIT, AND BANKING
in profit levels between states. Finding lower profits in states that witnessed a growth in branch banking would indicate that the national banks faced greater competition from state banks in these locations.27
To test how branching laws affect profitability, we use profit data on national banks that are available from 1926 to 1930 and that are decomposed by both size category (based on loans and investments) and by state. Rather than listing the actual profit rates, the Federal Reserve (1931) grouped the data into seven profit ranges, using the percentage return to capital as the measure of profitability.28 The data set thus has up to 70 observations per state, per year because there are 7 profit ranges for each of the 10 bank size categories.29 Because the ranges are ordered but the distance between them is not constant, an ordered logit is estimated using weights equal to the number of national barlks in each size category s, state i, and year t, earning a given level of profits. In Tables 8-10, we estimate variations on the following function:
PROFIT RANGEist = f { D l BVNCHit_ 1 + 52 STBANKSHAREit
+53 DEPINSi+SYsSIZEs +zYt YEARt), (3)
with observations weighted by the number of national banks in category ist. STBANKSHARE is the share of commercial banks that are state chartered, DEPINS is a dummy variable indicating whether a state had a deposit insurance system, and YEAR and SIZE are time and size dummies, respectively. BRANCH is one of our two measures of the extent of branching. In Table 8, BRANCH is the ratio of branch offices to total bank offices of state and national banks in 1922 and in Table 9 BRANCH is the change in this branching ratio between 1922 and 1925.
Tables 8 and 9 show that banks located in states with relatively more branching activity had declining profitability over the period 1926-30, suggesting that there was more competition in branch banking states.30 A higher initial level of branching or a more significant expansion of branching in the early 1920s is found to have reduced profitability as competitive forces unleashed by branching likely eliminated the geographic monopolies enjoyed by unit banks and state banking systems moved toward a new equilibrium. This is shown in Tables 8 and 9 by the shift in the coefficient on branching from positive in the beginning of the sample period to
27. Lower profitability may also result from higher merger costs; however, due to branching restrictions on nationally chartered banks, this interpretation is more likely to apply to state banks than the national banks used in our sample. Alternatively, national banks may have higher profits if branching leads to so much industry consolidation that surviving banks are able to generate oligopoly profits. Higher profits may also result if competition removes a sufficient number of inefficient national banks from the banking system.
28. These ranges are: deficit of 6% or more, deficit of 5.9%-0%, profit of less than 3%, profit of 3%-5.9%, profit of 6%-8.9%, profit of 9%-11.9%, and profit of 12% and over.
29. There may be less than 70 observations for a state within a year if none of the banks in a particular size category had profits within a particular profit range.
30. Calomiris and Ramirez (2002) also find that states that prohibited branching had higher profit levels. They attribute this partly to the ability of unit banks to exert monopoly power and partly to the higher level of risk that being an undiversified unit bank would entail.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 8
EFFECT OF INITIAL LEVEL OF BRANCHING ON PROFITABILITY BY YEAR, 1926-30
1926 1927 1928 1929 1930
Coefficient Coefficient Coefficient Coefficient Coefficient Estimate S.E. Estimate S.E. Estimate S.E. Estimate S.E. Estimate S.E.
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Estimated using an ordered logit. Deposit insurance data are from Calomiris (1992) and White (1981). All other information is from the Federal Reserve Board of Governors (1931). The dependent variable is an ordered variable indicating the level of profitability weighted by the number of banks that reported earning that level of profits. Size category 6 is omitted.
1.2*** (0.2)
- 1.7*** (0.1)
-0.4*** (0.1) -1.6*** (0.1) -1.1*** (0.1) -0.6*** (0.1) -0.2*** (0.1) -0.1* (0.1)
0.2*** (0.1) 0.2 (0.1) 0.2* (0. 1) 0.2 (0.3)
-0.3*** (0.1) 0.7*** (0.1) 1.8*** (0.1) 2.7*** (0.1) 3.4*** (0.1) 4.4*** (0.1)
1752 - 13750
0.6** (0.2)
-2.5*8* (0.2)
-0.1** (0.1) -1.5*** (0.1) -o.g*** (0.1) -0.6*** (0.1) -0.2*** (0.1) -0.1 (0.1)
0.1 (0.1) 0.2* (0.1) 0.2 (0.1) 0.0 (0-3) 0.2* (0.1) 1.2*** (0.1) 2.3*** (0.1) 3.1*** (0.1) 3.8*** (0.1) 4.8*** (0.1)
1781 - 1 3634
0.1 (0.2)
- 1.9*** (0.2)
0.3*** (0.1) -1.6*** (0.1) -0.g*** (0.1) -o.5*** (0.1) -0.1** (0.1) -0.1 (0.1)
0.2*** (0.1) 0.3** (0.1) 0.3** (0.1) 0.7** (0.3)
-0.2** (0.1) 0.8*** (0. 1) 1.9*** (0.1) 2.8*** (0.1) 3.5*** (0.1) 4.4*** (0.1)
1737 - 1 3406
Ratio of Banks to Total Bank Offices (1922 value)
Ratio of State Banks to Total Banks
Deposit Insurance Size Category 1 Size Category 2 Size Category 3 Size Category 4 Size Category 5 Size Category 7 Size Category 8 Size Category 9 Size Category 10 Intercept 7 Intercept 6 Intercept 5 Intercept 4 Intercept 3 Intercept 2
Observations Log-likelihood
-0.4* (0.2)
-0.8*** (0.2)
0.2*** (0.1) -1.1*** (0.1) -0.7*** (0.1) -0.3*** (0.1) -0.1 (0.1) -0.2*** (0.1)
0.3*** (0.1) 0.3*** (0.1) 0.6*** (0.1) 1.1*** (0.3)
-1.1*** (0.1) -0.1 (0.1)
O.g*** (0.1) 1.8*** (0.1) 2.5*** (0.1) 3.6*** (0.1)
1715 - 1 3342
-1.1*** (0.2)
0.5*** (0.2)
0.3** (0.1) -0.8*** (0.1) -0.5*** (0.1) -0.1** (0.1)
o.o (0.1) o.o (0.1) 0.2** (0.1) 0.4*** (0.1) 0.6*** (0.1) 1.0*** (0-3)
-2.8*** (0.1) -1.9*** (0.1) -o.g*** (0.1)
0.2 (0.1) o.9*** (0.1) 2.1*** (0.1)
1697 - 12962
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 9
EFFECT OF GROWTH OF BRANCHING ON PROFITABILITY BY YEAR, 192S30
1926 1927 1928 1929 1930
Coefficient Coefficient Coefficient Coefficient Coefficient Estimate S.E. Estimate S.E. Estimate S.E. Estimate S.E. Estimate S.E.
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Estimated using an ordered logit. Deposit insurance data are from Calomiris (1992) and White (1981). All other information is from the Federal Reserve Board of Governors (1931). The dependent variable is an ordered variable indicating the level of profitability weighted by the number of banks that reported earning that level of profits. Size category 6 is omitted.
Change in Ratio of Branching to Total Bank Offices (1922-25)
Ratio of State Banks to Total Banks Deposit Insurance Size Category 1 Size Category 2 Size Category 3 Size Category 4 Size Category 5 Size Category 7 Size Category 8 Size Category 9 Size Category 10 Intercept 7 Intercept 6 Intercept 5 Intercept 4 Intercept 3 Intercept 2
Observations (weighted) Log-likelihood
2.5*** (0.6)
-1.8*** (0.1) -0.4*** (0.1) -1.6*** (0.1) -1.1*** (0.1) -0.6*** (0.1) -0.2*** (0.1) -0.1* (0.1)
0.2*** (0.1) 0.2 (0.1) 0.2* (0.1) 0.2 (0.3)
-0.3*** (0.1) 0.8*** (0.1) 1.8*** (0.1) 2.7*** (0.1) 3.4*** (0.1) 4.4*** (0.1)
1752 - 14369
-0.3 (0-7)
-2.6*** (0.2) -0.2*** (0.1) - 1 .5*** (0. 1) - 1 .0*** (O. 1 ) -0.6*** (0.1) -0.2*** (0.1) -0.1 (0.1)
0.1 (0.1) 0.2* (0.1) 0.2 (0.1) 0.1 (0.3) 0.3*** (0.1) 1.3*** (0.1) 2.4*** (0.1) 3.3*** (0.1) 4.0*** (0.1) 4.9*** (0.1)
1781 - 14101
-0.9 (0.7)
-2.0*** (0.2) 0.2*** (0.1)
- 1.6*** (0.1) -o.9*** (0.1) -0.5*** (0.1) -0.1** (0.1) -0.1 (0.1)
0.2*** (0.1) 0.3** (0.1) 0.3** (0.1) 0.7** (0-3)
-0.2 (0.1) 0.8*** (0. 1) 2.0*** (0.1) 2.8*** (0.1) 3.5*** (0.1) 4.5*** (0. 1)
1737 - 13758
-2.1*** (0.7)
-0.9*** (0.2) 0.2*** (0.1)
-1.1*** (0.1) -0.7*** (0.1) -0.3*** (0.1) -0.1 (0.1) -0.2*** (0.1)
0.3*** (0.1) 0.3*** (0.1) 0.6*** (0.1) 1.1 *** (0.3)
-1.0*** (0.1) -0.1 (0.1)
1.0*** (0.1) 1.9*** (0.1) 2.6*** (0.1) 3.7*** (0.1)
1715 - 135 1 1
-2.1*** (0.7)
o.s*** 0.3**
-0.8*** -0.5*** -0.1*
0.0 0.0 0.2*** 0.4*** 0.6*** 1.0***
-2.9*** -2.0*** -0.9***
0.1 0.9*** 2.1***
1697 - 1 3070
(0.2) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1) (0.3) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1)
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 10
BRANCHING AND PROFITABILITY, 1926-30
Coefficient Estimate S.E.
Ratio of Branches to -0.19* * * (0.07) Total Bank Offices (lagged)
Ratio of State Banks -1.29*** (0.06) to Total Banks
Deposit Insurance -0.16*** (0.03) Size Category 1 -1.36*** (0.05) Size Category 2 -0.84*** (0.04) Size Category 3 -0.45*** (0.03) Size Category 4 -0.15*** (0.03) Size Category 5 -0.12*** (0.04) Size Category 7 0.19*** (0.03) Size Category 8 0.27*** (0.05) Size Category 9 0.35*** (0.06) Size Categorz7 10 0.62*** (0.13) 1927 -0.02 (0.03) 1928 0.03 (0.03) 1929 -0.13*** (0.03) 1930 -0.93*** (0.03) Intercept 7 -0.59*8* (0.05) Intercept 6 0.38*** (0.05) Intercept 5 1.43*** (0.05) Intercept 4 2.34*** (0.05) Intercept 3 3.07*** (0.05) Intercept 2 4.12*** (0.06)
Observations 8640 Log-likelihood -69460
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Estimated using an ordered logit and pooled observations. Deposit insurance data are from Calomiris and White. All other information is from the Federal Reserve Board of Governors (1931). The dependent vaxiable is an ordered variable indicating the level of profitability weighted by the number of banks that reported earning that level of profits. Size category 6 and the year 1928 are also omitted.
negative by the end of the period. Depending on which measure of BRANCH we use, the effect on profits from branching turns negative in 1927 or 1928 and is significantly so starting in 1929. Using the 1929 regression, a one-standard-deviation increase in the ratio of branch offices to total offices or a one-standard-deviation increase in the growth of branching reduces the probability of being in a higher profit category by 4%-7%.
As a robustness check, we also considered a pooled regression specification (which includes year dummies) over the sample period 1926-30, and find that states with relatively more branching activity had lower profits (Table 10). We use our same measure of the extent of branch banking, but lag it one year. A one-standard- deviation increase in the ratio of branch offices to total bank offices reduces the probability of being in a higher profit category by about 3%. While this overall effect is not terribly large, it masks different effects for different size banks. A one- standard-deviation increase in the ratio of branch offices to total bank offices for small banks, which seem to have been most affected by branching activity, reduced the probability of being in a higher category by 20%. For the largest banks, the probability of being in a higher category was reduced 10%. There seems to have been little
MARK CARLSON AND KRIS JAMES MITCHENER : 1315
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1316 : MONEY, CREDIT, AND BANKING
effect on the profitability of mid-size banks. The other coefficients are as expected. Larger banks were more profitable, and the years 1929 and 1930 were particularly bad years for bank profitability.
The shift from positive profits to negative profits in later years suggest that the banking system was being transformed by the consolidation that branch banking brought with it. Our findings are consistent with studies of recent banking deregula- tion such as Amel and Liang (1997) and Berger, Kashyap, and Scalise (1995), who view the removal of geographic restrictions as likely reducing the exercise of market power (by unleashing more actual or potential customers into local markets) and improving allocative eEciency (by enabling resources to flow more easily toward ac- tivities that yielded higher returns and more efficient producers).
3.3 Branch Banking and Financial Stability We now examine whether branch banking is responsible for lower national bank
failures in the 1920s and 1930s. We consider two ways that branching may have reduced the incidence of failures: (1) by improving diversification opportunities through geographical expansion and (2) by weeding out weak banks via the process of competition. We use the same data from the Federal Reserve that were used to examine bank profitability. Since the dependent variable we use is the number of failures, we estimate the regressions using count data analysis.
We first verify that we are able to replicate the state-level results of Wheelock (1995) and Mitchener (2000, 2005). We aggregate the size categories into state level observations and regress the number of failures of national banks in a state on variables indicating whether the state allowed branching, year dummies, and (log) number of banks in the state.31 (We also include a specification that uses the extent of branching as an independent variable.) The results shown in Table 11 match the previous literature and indicate that states permitting branch banking had fewer bank failures from 1927 to 30, the period where the profits regressions suggest that
. . . competltlon was most lmportant. We now explore whether the lower number of failures from widespread branch
banking was due to increased competition, diversification, or both. To conduct this test, we develop proxies for the competition and diversification effects of branch banking by taking advantage of the bifurcated nature of the U.S. banking system. To proxy how branch banking could lower failures by increasing consolidation, we compute the ratio of the branches of state banks to total bank offices in the state (COMPETITION). In states where this ratio is larger, national banks will be competing with more bank offices that are the direct result of statewide branching. And since this variable is based on the expansion of state-chartered branching, it excludes any benefits to national banks from greater diversification opportunities.
31. We use a pooled sample. Following Bertrand, Duflo, and Mullainathan (2004), we adjust the standard errors for the fact that there are multiple observations for each state by clustering at the state level. In states where the de jure situation differed from the de facto situation (such as West Virginia, where the laws allowed branching but the state banking commissioner refused all applications for establishing branches) or where states do not have a law, the de facto situation is used.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 11
THE EFFECTS OF BRANCHING LAWS ON NATIONAL BANK FAILURES, 1927-30
Legal Environment Actual Branches
Coefficient Estimate S.E. Coefficient Estimate S.E.
Intercept -2.84*** (0.89) -2.79*** (0.62) Branch Banking Permitted -0 79* * * (0.21) Ratio of Branches to -4.46*** (1.28)
Total Bank Offices (1922 value) Log Banks 0.65*** (0.11) 0.63*** (0.11) 1928 -0.17 (0.30) -0.16 (0.30) 1929 0.16 (0.29) 0.16 (0.29) 1930 1.07*** (0.26) 1.04 (0.27) Dispersion 0.87 (0.20) 0.91 (0.20)
Observations 192 192 Log-likelihood -8.2 -7.8
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the lSo, 5%, and 10% level, respectively. Estimated using pooled data and count data analysis with a negative binomial distribution. Information on bank failures is from the Annual Report of the Comptroller. Information on the number of banks and branching laws is from the Federal Reserve Board of Governors (1931). The dummy for the year 1926 is omitted. The dependent variable is the number of failing banks in a state in a year.
To proxy for the diversification effects of branching on national banks, we compute both the percentage of national banks with branches (DIVERSIFYI) as well as the ratio of national bank branches to national banks (DIVERSIFY2). In states where DIVERSIFYI is larger, more national banks will have greater opportunities to diver- sify their portfolios across branch offices in the state. DIVERSIFY2 captures the extent to which national banks can diversify. Mathematically, these ratios are:
COMPETITION = (Branches of state banks)/ (4)
(Banks and branch offices of state and national banks)
DIVERSIFY1 = (Number of national banks with branches)/ (S)
(Number of national banks)
DIVERSIFY2 = (Branches of national banks)/ (6)
(Number of national banks)
Including these measures side-by-side in regressions will allow us to identify which channel was quantitatively more important in lowering national bank failures. We use the 1922 values for all three of theses ratios in order to ensure that they are plausibly exogenous.
For state i, year t, and size category s, we estimate the total national bank failures (FAILURE) by the following function:
FAILUREiSt = f {s1 COMPETITIONit_1 + 52DIVERSIFYit_1
+ 4DEPINSi + sBANKsit + 4BUSFAILit-l (7)
+ DsFARMFAILit_l + 6HERFt + S0iYEARt + SYsSIZEs} v
MARK CARLSON AND KRIS JAMES MITCHENER : 1317
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1318 : MONEY, CREDIT, AND BANKING
1 800
5 6 Size Categories
FIG. 4. Annual Average Number of National Banks and Bank Failures by Size Category. Sources and Notes: Comptroller of the Currency (various years) and Federal Reserve Board (1931). Annual averages are constructed over the period 1926-30. Size categories are category 1, under $150,000; category 2, $150,00(}$250,000; category 3, $250,000-$500,000; category 4, $500,000-$750,000; category 5, $750,00s$1,000,000; category 6, $1,000,000- $2,000,000; category 7, $2,000,00>$5,000,000; category 8, $5,000,00s$ 10,000,000; category 9, $10,000,000- $50,000,000; category 10, $50,000,000 and over.
where COMPETITION and DIVERSIFY are lagged, and DIVERSIFY is one of the two measures of diversification defined above (DIVERSIFYI or DlVERSlFY2). Since our dependent variable is disaggregated by size, we are also able to include bank size indicators (SIZE) as additional conditioning variables a factor that has been associated with the probability of failure in previous studies (Calomiris and Mason, 2000, Carlson, 2004, White, 1984).32 Figure 4 shows the distribution of banks and failures by bank size category. To control for differences in real shocks across states, we include a measure of lagged business failures (BUSFAIL) and compute agricultural distress as the lagged farm failure rate (FARMFAIL). We control for additional state level factors: the (log) number of national banks (BANKS); the Herfindahl index (HERF), described above; and whether a state had a deposit insurance system (DEPlNs).33 All 48 states are used and we include year dummies
32. Larger banks may be more stable because of their greater ability to spread risks and coordinate emergency assistance with other banks and their reduced propensity to spread contagion when the banking sector is subjected to external shocks.
33. We include the deposit insurance indicator variable for two reasons. First, the presence of deposit insurance in eight states has been linked to higher failure rates in the 1920s in these states (Wheelock 1992, Wheelock and Wilson, 1995, Calomiris, 1990). Moreover, increased competition can potentially exacerbate the moral hazard problem associated with deposit insurance (Keeley, 1990). There is however, little overlap between states that allowed branching and state that had a deposit insurance program.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1319
to capture any time-specific influences. Since each size category provides a unique observation, we have 10 observations for each state in each year.
Since our profit regressions suggest that the initial shakeout and consolidation process had taken place by the end of 1926, we pool our data over the period 1927-30, and estimate Equation (7) using count data analysis with a negative binomial distribution.34 However, because we observe each state multiple times, we cluster our standard errors as suggested by Bertrand, Duflo, and Mullainathan (2004).35
The relative effects of competition and diversification in producing lower failures rates in states allowing branch banking are shown in Table 12.36 The negative and statistically significant estimated coefficient on our ratio that proxies for competition supports the hypothesis that branch banking improves financial stability by weeding out inefficient banks through increased competition and consolidation. Increased competition stemming from branching significantly reduces failures. An increase in the competition ratio of 0.1 (about one standard deviation) results in a 35% decrease in the number of failures in a state. On the other hand, we find no evidence that either of our measures for diversification resulting from branching is associated with fewer national bank failures, where column 1 of Table 12 uses DIVERSIFY1 (the share of national banks with branches) and column 2 uses DIVERSIFY2 (average branches per national bank).
These results do not appear to be sensitive to choosing a diiTerent measure of the extent of branching. As a sensitivity test, we use one-year lagged values of COMPETITION, DIVERSIFYI, and DIVERSIFY2 (shown in Table 13). Here again, the sign on COMPETITION is negative and statistically significant. We find no evidence that either of our measures for diversification resulting from branching is associated with fewer national bank failures; indeed, the only statistically significant coefficient for either of the diversification measures has the wrong expected sign. It should be noted, however, that these results do not provide any indication about the importance of diversification through branching at state banks, where the bulk of the banking systems' branches were located.
The signs on the other explanatory variables in the failure regressions are largely as expected. Bank failures were significantly higher in 1929 and 1930 (as a result of the Depression). An increase in the banks at risk boosts failures. And consistent with the research using individual bank data from this period, larger banks were less
34. A few states have less than 10 observations per year because no banks exist in some size categories. 35. As an alternative, we estimated a panel regression with state-level fixed effects and found
similar effects. 36. Branching and bank size are generally correlated, and both allow a bank to diversify: size allows
the bank to make more total loans and branching permits the bank to make loans in different locations. In this analysis, we focus on the geographic diversification allowed by branching and control for bank size by including dummies for the size group of the bank. Another way of controlling for bank size is by repeating this regression using only banks of a particular size group. These regressions yield similar results to those shown here.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 12
EFFECTS OF INITIAL LEVEL OF BRANCHING ON NATIONAL BANK FAILURES, 1927-30
Coefficient Estimate S.E. Coefficient Estimate S.E.
Ratio of Branches of -3.46* (2.09) -3.40* (2.08) State Banks to Total Bank Offices (1922 value)
Share of National Banks - 8.00 (14.59) with Branches (1922 value)
Ratio of National Branches -2.94 (4.78) to National Banks (1922 value)
Log Banks 0.57*** (0.11) 057**t (0.11) Herfindahl Index -0.15 (1.68) 0.10 (1.80) Business Fail Rate 0.00 (0.38) -0.03 (0.37) Farm Failure Rate 23.84 (102.09) 29.67 (100.39) Deposit Insurance 0.38 (0.35) 0.35 (0.33) Size Category 1 0.68* (0.38) 0.708 (0.38) Size Category 2 0.59* (0.30) 0.59** (0.30) Size Category 3 0.78*** (0.21) 0.78**t (0.22) Size Category 4 0.16 (0.25) 0.16 (0.25) Size Category 5 -0.51 (0.38) -0.51 (0.37) Size Category 7 -0.61 (0.38) -0.61 (0.38) Size Category 8 - 1.31 *** (0.51) - 1.30*** (0.51) Size Category 9 -1.75** (0.85) -1.74** (0.85) Size Category 10 -18.12*** (0.55) -18.09*** (0.55) 1928 -0.29 (0.27) -0.29 (0.27) 1929 0.05 (0.27) O.05 (0.28) 1930 1.06*** (0.26) 1.06*** (0.26) Constant -3.52*** (0.74) -0.61 (4.77)
Observations 1614 1613 Log-likelihood -665.7 -665.6
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5So, and 10% level, respectively. Estimated using a negative binomial distribution. Standard errors are clustered to account for multiple observations from the same state. Data on bank failures are from the Annual Report of the Comptroller of the Currency ( 1932). Data on the number of bank, size distribution, branching laws, and number of branches is from the Federal Reserve Board of Governors (1931). Business failure rates are from the U.S. Department of Commerce and farm foreclosures from the Department of Agriculture ( 1936). Deposit insurance data are from Calomiris ( 1992) and White (1981). Size category 6 and the year 1927 are omitted. The dependent variable is the average number of failing banks in each group (where a group is a set of banks in a similar size category in a state).
1320 : MONEY, CREDIT, AND BANKING
likely to fail and smaller banks were more likely to fail.37 After controlling for other factors, farm failure rates and deposit insurance were statistically insignificant.
California had by far the most branches, about 660 in 1926. New York had the second most with approximately 489 branches. As a test of the robustness of our results, we eliminate California from the sample and repeat the estimation. As Table 14 shows, the results are similar in terms of the effects that the two channels had on failures. Interestingly, both the size of effect of competition and its significance are
37. We also tested an alternative hypothesis that branching decreases failures by facilitating a substitute to failure, namely merging with another bank. (We thank Joe Mason for suggesting this idea.) According to this view, branching reduces failures because it facilitates mergers by expanding the pool of possible merger partners from banks within the city to all state banks within the state. We tested this hypothesis by constructing an index of the ease with which banks might merge. This index is the number of mergers divided by the total number of bank exits (mergers, failures, and voluntary liquidations). However we do not find a significant relationship between this measure and the number of failures.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 13
EFFECTS OF BRANCHING ON NATIONAL BANK FAILURES USING LAGGED BRANCHING, 1927-30
Coefficient Estimate S.E. Coefficient Estimate S.E.
Ratio of Branches of -3.64** (1.61) -3.92** (1.72) State Banks to Total Bank Oices
Share of National Banks 1.77 (2.93) with Branches
Ratio of National Branches 0.36 (0.24) to National Banks
Log Banks 057*** (0.11) 0.56*** (0.11) Herfindahl Index -0.18 (1.71) -0 40 (1.77) Business Fail Rate -0.08 (0.37) -0.10 (0.38) Farm Failure Rate 33.46 (98.64) 23.57 (98.17) Deposit Insurance 0.32 (0.33) 0.31 (0.33) Size Category 1 0.69* (0.38) 0.67* (0.38) Size Category 2 0.59* (0.31) 0.57* (0.31) Size Category 3 0.77*** (0.22) 0.77*** (0.22) Size Category 4 0.15 (0.25) 0.15 (0.25) Size Category 5 -0.52 (0.38) -0.52 (0.38) Size Category 7 -0.62 (0.38) -0.63* (0.38) Size Category 8 - 1.30*8 (0.51) - 1.32*** (0.51) Size Category 9 -1.76** (0.85) - 1.77** (0.85) Size Category 10 -17.60*** (0.56) -17.64*** (0.55) 1928 -0.28 (0.27) -0.30 (0.27) 1929 0.06 (0.28) 0.03 (0.28) 1930 1.08*** (0.26) 1.05*** (0.26) Constant -3.50*** (0.73) -3.71*** (0.71)
Observations 1613 1613 Log-likelihood -656.0 -655.5
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Estimated using a pooled sample and a negative binomial distribution. Standard errors are clustered to account for multiple observations from the same state. Data on bank failures are from the Annual Report of the Comptroller of the Currency (1932). Data on the number of bank, size distribution, branching laws, and number of branches is from the Federal Reserve Board of Governors (1931). Business failure rates are from the U.S. Department of Commerce and farm foreclosures from the Department of Agriculture (1936). Deposit insurance data are from Calomiris (1992) and White (1981). Size category 6 and the year 1927 are also omitted. The dependent variable is the Number of failing banks in each group (where a group is a set of banks in a similar size category in a state and by year).
MARK CARLSON AND KRIS JAMES MITCHENER : 1321
greater when California is excluded from the sample. The same size increase in the ratio of branches of state banks to total bank offices (again about one-tenth of a standard deviation) now reduces the number of failures by about 60%. The reported coefficients on the other variables are quite similar to those shown in the previous table.
To check the sensitivity of our results to the choice of stability measures, we also considered an alternative dependent variable the share of assets at national banks affected by failures. The results (not reported) are similar to what is displayed in Tables 12 and 13. States with more extensive state branching had a smaller share of assets located in failing national banks.
4. CONCLUSIONS
This paper revises our understanding of the role that branching played in improving the stability of banking systems during the 1920s and 1930s. Diversification was not
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE 14
EFFECTS OF INITIAL LEVEL OF BRANCHING ON NATIONAL BANK FAILURES EXCLUDING CALIFORNIA,
1927-30
Coefficient Estimate S.E. Coefficient Estimate S.E.
Ratio of Branches of - 5.96*** (2.15) - 6.35*** (2.29)
State Banks to
Total Bank Offices
(1922 value)
Share of National Banks - 6.76 (16.40)
with Branches
(1922 value)
Ratio of National Branches - 0.40 (4.82)
to National Banks
(1922 value)
Log Banks 0.54*** (0.11) 0.54*** (0.11)
Herfindahl Index - 0.21 (1.69) - 0.13 (1.80)
Business Fail Rate - 0.08 (0.38) - 0.10 (0.37)
Fann Failure Rate 6.69 (103.47) 13.17 (102.09)
Deposit Insurance 0.34 (0.35) 0.31 (0.33)
Size Category 1 0.66* (0.39) 0.67* (0.39)
Size Category 2 0.53* (0.31) 0.54* (0.31)
Size Category 3 0.79*** (0.22) 0.79*** (0.22)
Size Category 4 0.15 (0.26) 0.15 (0.26)
Size Category 5 -0.49 (0.38) -0.49 (0.38)
Size Category 7 -0.65 (0.40) -0.65* (0.40)
Size Category 8 -1.30** (0.51) -1.30** (0.51)
Size Category 9 -1.75** (0.86) -1.74** (0.86)
Size Category 10 -17.83*** (0.58) -18.10*** (0.59)
1928 -0.29 (0.28) -0.28 (0.27)
1929 0.06 (0.28) 0.06 (0.29)
1930 1.06*** (0.27) 1.06*** (0.27)
Constant -3.27*** (0.75) -2.90 (4.80)
Observations 1573 1573
Log-likelihood - 635.4 - 635.6
NOTES: The symbols (***), (**), and (*) indicate statistical significance at the 1%, 5%, and 10% level, respectively. Estimated using a negative binomial distribution. Standard errors are clustered to account for multiple observations from the same state. Data on bank failures are from the Annual Report of the Comptroller of the Currency (1932). Data on the number of bank, size distribution, branching laws, and number of branches is from the Federal Reserve Board of Governors (1931). Business failure rates are from the U.S. Department of Commerce and farm foreclosures from the Department of Agriculture (1936). Deposit insurance data are from Calomiris (1992) and White (1981). Size category 6 and the year 1927 are omitted. The dependent variable is the average number of failing banks in each group (where a group is a set of banks in a similar size category in a state).
1322 : MONEY, CREDIT, AND BANKING
the primary channel through which branch banking made state banking systems more resistant to shocks. Instead, the expansion of statewide branch banking induced greater competition in states where it was permitted and improved the stability of their banking systems by removing weak and inefficient banks. Our results are largely consistent with recent literature that has examined the effects of deregulation in other settings. Like Amel and Liang (1997), who examine banks and branching in the 1980s, and Claessens, Demirguc-Kunt, and Huizinga (2001), who look at banks expanding internationally, we find that the growth of branch banking in the 1920s is associated with lower profits in the latter part of the decade. Our results put the well-documented response of unit bankers (particularly those located in rural areas) to the growth of branch banking in the 1920s in proper perspective. Because
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1323
the growth of statewide branching was eroding the monopoly profits that unit bankers had previously enjoyed, they responded (as predicted by the economic theory of regulation) by lobbying state and federal governments to legally limit it from spread- ing. As a result of their influence, many states consequently continued to prohibit branching during the 1920s.
Similar to what Jayaratne and Strahan (1998) and Stiroh and Strahan (2003) find for the branching deregulation of the 1980s, we hnd that this more competitive atmosphere is associated with a higher exit rate. This confirms what some economic historians have suggested, but not shown. As White (1985, p. 291) contends, "The number of small banks in rural areas needed to be reduced, and the mergers assisted the weaker institutions with less pain than the massive failures that followed. Unfortunately, this development was stifled by regulations in most states that forbade branch banking." The market upheaval and increase in competition resulting from the removal of legal barriers to entry, however, resulted in longer-run stability, with fewer failures in states where branching had spread. Our results also confirm the hypothesis of Berger, Demsetz, and Strahan (1999), that competition prompts weaker banks to leave the banking system, and parallels the findings of Kaminsky and Schmukler (2002), that international financial liberalization causes some initial turbu- lence in financial markets but over time results in reduced volatility.
APPENDIX A
Data Sources Two main sets of data are employed. The first consists of state-level aggregate
data on national banks, covering the period 1922-30. These data are used to compare the prevalence of entry and exits between states with different branching regimes. Information on number of banks, mergers, voluntary liquidations, and new banks are compiled using the Annual Report of the Comptroller of the Currency (1922-30).
The second contains additional information on national banks in each state between 1926 and 1930. These data are drawn from the Federal Reserve Board of Governors' (1931) Report on Branch, Chain and Group Banking, Volume 9: Bank Profts and further categorize banks by size (based on the sum of loans and investments). This source also includes the information on bank profits by size category. Data from this Federal Reserve report are also used for the construction of the Herfindahl index of banking concentration. The four-firm bank concentration index is calculated using data on the four largest national banks in each state from Polk's Bank Directory (various years).
Bank failures for 1926-30 are taken from the Annual Report of the Comptroller of the Currency (1932, Table 43, pp.208-228) and the Comptroller of the Currency's Statements of National Banks (1925-29), and are matched to the appropriate size category for the appropriate state using the information contained in Federal Reserve Board of Governors (1931).
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
State bank data are matched to branching laws and to indicators of state economic activity. Information on the branching laws for each state are from the Federal Reserve's Report on Branch, Chain and Group Banking, Volume 2: Branch Banking in the United States. This reports the developments in each state's branch banking laws from 1909 until 1931. Detailed information on state branching laws (including whether the de jure rather than the de facto situation was used) is shown in Appendix B1. Information on Deposit Insurance follows Calomiris (1992) and White (1981). Business failures and population estimates are from the U.S. Department of Com- merce, Statistical Abstract of the United States (various years). Farm foreclosure (bankruptcies) rates are computed using data from U.S. Department of Agricul- ture (1936) while income shares are derived from the 1920 Census.
APPENDIX B
TABLE B1
BRANCH BANKING LAWS
State Type Year Law Passed Notes
Alabama Unit 1911 A few banks have branches Arizona Statewide 1901 Branching had been practiced before law came
into effect Arkansas Unit 1923 State commissioner authorizes a few "exceptions" California Statewide 1909 Branching had been practiced before law came
into effect Colorado Unit 1909 Connecticut Unit 1902 Delaware Statewide 1895 If charter allows Florida Unit 1913 Georgia Multiple 1929 Branching allowed until 1927, banned until 1929
when it is permitted in Savannah and Atlanta Idaho Unit 1919 Illinois Unit 1923 Prior to 1923, branches had been "not authorized" Indiana Unit 1921 Iowa Unit 1927 Prior to 1927, branches had been "not authorized" Kansas Unit* 1929 Law against it enacted in 1929. Previously there
was no law Kentucky Statewide* 1895 In 1902 the banking authority stated that law did not
authorize, but "was not construed as prohibitive." In 1909 the courts say the banks cannot have branches but can have "offices to receive deposits and pay checks or transact other necessary duties not requiring special discretion or business acumen." Observers at the time noted little difference between these agencies and branches.
Louisiana Limited 1902 Two branches in the same parish Maine Limited 1895 Only in the county of the home office or
contiguous counties Maryland Statewide 1910 Branching had been practiced before law
came into effect Massachusetts Limited 1928 Trusts can have branches in the same city, prior to
1928 were limited to one branch Michigan Limited* 1895 No law, but a lot of branches in the city
of the home office
(continue)
1324 : MONEY, CREDIT, AND BANKING
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
TABLE B1
CONTINUED
State Type Year Law Passed Notes
NOTES: All types are as indicated by law (de jure) except were indicated by a * in which case the de facto type is used. Please consult the table for the reason the de jDsre is not used. Source: Federal Reserve Board of Governors (1931), lVeport of the Branchw, Chain, and Grosp Banking Comntittee, Volume 9.
LITERATURE CI 1 ED
Allen, Franklin, and Douglas Gale (2000). Comparing Financial Systems. Cambridge, MA: MIT Press.
Alston, Lee, Wayne Grove, and David Wheelock (1994). "Why Do Banks Fail? Evidence from the 1920s." Explorations in Economic History 31, 409-431.
MARK CARLSON AND KRIS JAMES MITCHENER : 1325
Minnesota M. . . .
SSlSSlppl Missouri Montana Nebraska Nevada New Hampshire
New Jersey New Mexico
New York North Carolina
North Dakota
Ohio Oklahoma Oregon Pennsylvania
Unit Limited Unit Unit Unit Unit Unit*
Limited Unit
Limited Statewide
Unit*
Limited Unit* Unit Limited
1923 1924 1899 1927 1927 1909 1895
1895 1915
1919 1921
1895
1923 1895 1921 1927
Prior to 1924, the law had prohibited branches
No law prior to 1927 No law prior to 1927
Commissioner reported that he was not aware of any law prohibiting or allowing branching
Allows "mercantile corporation which maintains a banking department to continue operations at its branches." A clause in the law specifically for a particular corporation.
Branching within cities of 50,000 Branching had been practiced before law came
into effect Branches were not specifically mentioned, but
the law was construed as not permitting them In contiguous communities No law
Branches are only permitted in places where national banks have branches already. Prior to 1927, it had been in home office-city only.
No law per se, but instead what the capital requirements would be should the bank have branches
No law Law unclear, seems to be allowed in the county of
the home office
Went from "not authorized" to able to establish "agencies" which are branches in all but name
In 1922 authorized for anywhere, in 1928 restricted to cities of 50,000
Technically allowed by law from 1925 to 29, however the commissioner of banking did not permit them
Law of 1921 has banks articles of incorp. (AoI) state the "place or places where its offices be located," while the 1926 law has the AoI state the "place where its office..."
Rhode Island Statewide 1908 South Carolina Statewide* 1895
South Dakota Tennessee
Texas Utah Vermont
Virginia
Washington West Virginia
Wisconsin Wyoming
Unit* Limited*
Unit Unit Multiple
Multiple
Unit Unitt
Unit Multiple
1895 1925
1905 1917 1929
1928
1920 1929
1909 1926
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1326 : MONEY, CREDIT, AND BANKING
Amel, Dean, and J. Nellie Liang (1997). "Determinants of Entry and Profits in Local Banking Markets." Review of Industrial Organization 12, 59-78.
Berger, Allen, Rebecca Demsetz, and Phillip Strahan (1999). "The Consolidation of the Financial Services Industry: Causes, Consequences, and Implications for the Future." Journal of Banking and Finance 23, 135-194.
Berger, Allen, and Timothy Hannan (1998). "The Efficiency Cost of Market Power in the Banking Industry: A Test of the 'Quiet Life' and Related Hypotheses." The Review of Economics and Statistics 80, 454465.
Berger, Allen, Anil Kashyap, and Joseph Scalise (1995). "The Transformation of the U.S. Banking Industry: What a Long, Strange Trip It's Been." Brookings Pupers on Economic Activity 55-218.
Bertrand, Marianne, Esther Duflo, and Sendhil Mullainathan (2004). "How Much Should We Trust Differences-in-Differences Estimates?" Quarterly Journal of Economics 119, 249-275.
Calomiris, Charles (1990). "Is Deposit Insurance Necessary? A Historical Perspective." Journal of Economic History 50, 283-295.
Calomiris, Charles (1992). "Do 'Vulnerable' Economies Need Deposit Insurance? Lessons from U.S. Agriculture in the 1920s." In If Texas Were Chile: A Primer on Banking Reform, edited by Philip Brock, pp. 237-349. San Francisco, CA: Institute for Contemporary Studies.
Calomiris, Charles (2000). U.S. Bank Deregulation in Historical Perspective. Cambridge, MA: The Cambridge University Press.
Calomiris, Charles, and Joseph Mason (2000). "Causes of U.S Bank Distress During the Depression." NBER Working Paper No. 7919.
Calomiris, Charles, and Carlos Ramirez (2002). "The Political Economy of Bank Entry Restrictions: Theory and Evidence from the U.S. in the 1920s." Unpublished Manuscript.
Cameron, A. Colin, and Pravin Trivedi ( 1998). Regression Analysis of Coant Data. Cambridge, MA: The Cambridge University Press.
Carlson, Mark (2004). "Are Branch Banks Better Survivors? Evidence from the Depression Era." Economic Inquiry 42, 111-126.
Chapman, John, and Ray Westerfield (1942). Branch Banking: Its llistorical and Theoretical Position in America and Abroad. New York: Harper and Brothers Publishers.
Cherin, Antony, and Ronald Melicher (1988). "Impact of Branch Banking on Bank Firm Risk via Geographic Market Diversification." Quarterly Journal of Business and Economics 27, 73-95.
Claessens, Stijn, Asli Demirguc-Kunt, and Harry Huizinga (2001). "How Does Foreign Entry Affect the Domestic Banking Market?" Journal of Banking and Finance 25, 891-911.
Comptroller of the Currency (various years). Annual Report. Washington, DC: United States Government Printing Office.
Comptroller of the Currency (various years). Statements of National Banks. Washington, DC: United States Government Printing Office.
Demsetz, Rebecca S., and Philip E. Strahan (1997). "Diversification, Size, and Risk at Bank Holding Companies." Journal of Money, Credit, and Banking 29, 30s313.
DeYoung, Robert, Iftekhar Hasan, and Bruce Kirchhoff (1998). "The Impact of Out-of-State Entry on the Cost Efficiency of Local Commercial Banks." Journal of Economics and Business 22, 191-203.
Drummond, Ian (1991). "Why Canadian Banks Did Not Collapse in the 1930s." In The Role of Banks in the Interwar Economy, edited by James et al. Cambridge, MA: Cambridge University Press.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
MARK CARLSON AND KRIS JAMES MITCHENER : 1327
Economides, Nicholas R., Glenn Hubbard, and Darius Palia (1996). "The Political Economy of Branching Restrictions and Deposit Insurance: A Model of Monopolistic Competition Among Small and Large Banks." Journal of Law and Economics 39, 667-704.
Federal Reserve Board of Governors (1931). Report of the Branch, Chain, and Group Banking Committee Volumes 2 and 9. Washington, DC: Board of Governors.
Federal Reserve Board of Governors (1943). Banking and Monetary Statistics. Washington, DC: The National Capital Press.
Folkerts-Landau, David, and Carl-Johan Lindgren ( 1998). Toward a Framework for Financial Stability. Washington, DC: International Monetary Fund.
Friedman, Milton, and Anna Schwartz (1963). A Monetary History of the United States, 1867-1960. Princeton, NJ: Princeton University Press.
Gart, Alan (1994). Regulation, Deregulation, Reregulation. New York: John Wiley and Sons, Inc.
Grossman, Richard (1994). "The Shoe That Didn't Drop: Explaining Banking Stability in the Great Depression." Journal of Economic History 54,654-682.
Hubbard, R. Glenn (1994). Money, the Financial System, and the Economy. Reading, MA: Addison-Wesley Publishing Company.
Jayaratne, Jith, and Philip E. Strahan (1998). "Entry Restrictions, Industry Evolution, and Dynamic Efficiency: Evidence from Commercial Banking." Journal of Law and Economics 41,239-273.
Kaminsky, Graciela, and Sergio Schmukler (2002). "Short-Run Pain, Long-Run Gain: The Effects of Financial Liberalization." World Bank Working Paper Series 2912.
Keeley, Michael C. (1990). "Deposit Insurance, Risk and Market Power in Banking." American Economic Review 80,1183-1200.
Koskela, Erkki, and Rune Stenbacka (2000). "Is There a Tradeoff Between Bank Competition and Financial Fragility?" Journal of Banking and Finance 24,1853-1873.
Kroszner, Randall, and Phillip Strahan (1998). "What Drives Deregulation? Economics and Politics of the Relaxation of Bank Branching Restrictions." Quarterly Journal of Economics 114,1437-1467.
Kroszner, Randall, and Phillip Strahan (2000). "Obstacles to Optimal Policy: The Interplay of Politics and Economics in Shaping Bank Supervision and Regulation Reforms." NBER Working Paper No. 7582.
Kryzanowski, Lawrence, and Gordon Roberts (1993). "Canadian Banking Solvency." Journal of Money, Credit, and Banking 25,361-376.
Lauch, Louis, and Neil Murphy (1970). "A Test of the Impact of Branching on Deposit Variability." Journal of Financial and Quantitative Analysis 5,323-327.
Levine, Ross (1996). "Foreign Banks, Financial Development, and Economic Growth." In International Financial Markets: Harmonization versus Competition, edited by Claude Barfield, pp. 224-254. Washington, DC: AEI Press.
Matutes, Carmen, and Xavier Vives (2000). "Imperfect Competition, Risk Taking, and Regula- tion in Banking." European Economic Review 44, 1-34.
Mitchener, Kris (2000). "Do Supervision and Regulation Influence Financial Stability? Evi- dence from the Great Depression." Unpublished Manuscript, University of California, Berkeley.
Mitchener, Kris (2005). "Bank Supervision, Regulation, and Instability During the Great Depression." Journal of Economic History 65, 152-185.
Morgan, Donald, Bertrand Rime, and Philip E. Strahan (2003). "Bank Integration and State Business Cycles." NBER Working Paper No. 9704.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
1328 : MONEY, CREDIT, AND BANKING
Polk's Bankers Encyclopedia Co (various years). Polk's Bankers Encyclopedia. Detroit, MI: Polk's Bankers Encyclopedia Co.
Preston, Howard (1928) "Recent Developments in Branch Banking." American Economic Review 14, 443v62.
Southworth, Shirley (1928). Branch Banking in the United States. New York: McGraw-Hill Book Company.
Sprague, Oliver M.W. (1903). "Branch Banking in the United States." Quarterly Journal of Economics 27, 242-260.
Stiroh, Kevin J., and Philip E. Strahan (2003). "Competitive Dynamics of Deregulation: Evidence from U.S. Banking." Journal of Money, Credit, and Banking 35, 801-828.
Tippetts, Charles S. (1929). State Banks and the Federal Reserve System. New York: D. Van Nostrand Company.
U.S. Department of Agriculture (1936). Circular No. 414 Farnt Bankraxptcies, 1898-1935. Washington, DC: U.S. Department of Agriculture.
U.S. Department of Commerce (various years). Statistical Abstract of the United States. Washington, DC: GPO.
Wacht, Richard (1968). "Branch Banking and Risk." Journal of Financial and Quantitative Anaylsis 3,97-107.
Wheelock, David (1992). "Regulation and Bank Suspensions: New Evidence from the Agricultural Collapse of the 1920s." Journal of Economic History 52,80S825.
Wheelock, David (1993). "Government Policy and Banking Market Structure in the 1920s." Journal of Economic History 53,857-879.
Wheelock, David (1995). "Regulation, Market Structure, and the Bank Failures of the Great Depression." Federal Reserve Bank of St. Louis Review 77,27-38.
Wheelock, David, and Paul Wilson (1995). "Explaining Bank Failures: Deposit Insurance, Regulation, and Efficiency." The Review of Economics and Statistics 77,689-700.
Wheelock, David, and Paul Wilson (2000). "Why do Banks Disappear? The Determinants of U.S. Bank Failures and Acquisitions." Review of Economics and Statistics 82,127-138.
White, Eugene (1981). "State-Sponsored Insurance of Bank Deposits in the United States, 1907-1929." Journal of Economic History 41, 537-557.
White, Eugene (1983). The Regulation and Reform of the American Banking System, 1900- 1929. Princeton, NJ: Princeton University Press.
White, Eugene (1984). "A Reinterpretation of the Banking Crises of 1930." Journal of Economic History 44, 119-138.
White, Eugene (1985). "The Merger Movement in Banking, 1919-33." Journal of Economic History 45, 285-291.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:11:13 PM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. [1293]
- p. 1294
- p. 1295
- p. 1296
- p. 1297
- p. 1298
- p. 1299
- p. 1300
- p. 1301
- p. 1302
- p. 1303
- p. 1304
- p. 1305
- p. 1306
- p. 1307
- p. 1308
- p. 1309
- p. 1310
- p. 1311
- p. 1312
- p. [1313]
- p. [1314]
- p. 1315
- p. 1316
- p. 1317
- p. 1318
- p. 1319
- p. 1320
- p. 1321
- p. 1322
- p. 1323
- p. 1324
- p. 1325
- p. 1326
- p. 1327
- p. 1328
- Issue Table of Contents
- Journal of Money, Credit and Banking, Vol. 38, No. 5 (Aug., 2006), pp. 1127-1403
- Front Matter
- The Predictive Content of the Output Gap for Inflation: Resolving In-Sample and Out-of-Sample Evidence [pp. 1127-1148]
- Has U.S. Monetary Policy Changed? Evidence from Drifting Coefficients and Real-Time Data [pp. 1149-1173]
- Taylor Rules and the Deutschmark: Dollar Real Exchange Rate [pp. 1175-1194]
- Do Bank Loan Relationships Still Matter? [pp. 1195-1209]
- Macroeconomic Dynamics and Credit Risk: A Global Perspective [pp. 1211-1261]
- Stock Market Reaction to Financial Statement Certification by Bank Holding Company CEOs [pp. 1263-1291]
- Branch Banking, Bank Competition, and Financial Stability [pp. 1293-1328]
- Shorter Papers, Discussions, and Letters
- Additional Evidence of Long-Run Purchasing Power Parity with Restricted Structural Change [pp. 1329-1349]
- A Portfolio View of Banking with Interest and Noninterest Activities [pp. 1351-1361]
- Multiple Regimes in U.S. Monetary Policy? A Nonparametric Approach [pp. 1363-1377]
- Does Political Instability Lead to Higher Inflation? A Panel Data Analysis [pp. 1379-1389]
- Euro-Illusion: A Natural Experiment [pp. 1391-1403]
- Back Matter
CENTRAL BANK COMMUNICATION ON FINANCIAL.pdf
CENTRAL BANK COMMUNICATION ON FINANCIAL STABILITY*
Benjamin Born, Michael Ehrmann and Marcel Fratzscher
Central banks regularly communicate about financial stability issues. This article asks how such communications affect financial markets, based on a unique dataset covering more than 1,000 releases of Financial Stability Reports (FSRs) and speeches by 37 central banks over the past 14 years. The findings suggest that optimistic FSRs lead to significant and potentially long-lasting positive abnormal stock market returns, whereas no such effect is found for pessimistic FSRs. Speeches and interviews, in contrast, have smaller effects on market returns during tranquil times but have been influential during the 2007–10 global financial crisis.
The global financial crisis has triggered heated discussions on how best to achieve financial stability in the future. An important role in that regard has been assigned to central banks, many of which have already had or have been given explicit financial stability mandates. In light of this, a large number of central banks have communicated extensively on financial stability-related matters, e.g. through the publication of Financial Stability Reports (FSRs) and financial stability-related speeches and interviews.
Such communications have at times triggered substantial reactions in financial markets. To give a prominent example, on 5 December, 1996, Alan Greenspan, then chairman of the Federal Open Markets Committee (FOMC), said
Clearly, sustained low inflation implies less uncertainty about the future, and lower risk premiums imply higher prices of stocks and other earning assets. We can see that in the inverse relationship exhibited by price/earnings ratios and the rate of inflation in the past. But how do we know when irrational exuberance has unduly escalated asset values, which then become subject to unexpected and prolonged contractions as they have in Japan over the past decade?
The phrase ‘irrational exuberance’ was interpreted as a warning that markets were overvalued, leading to substantial declines in stock markets worldwide.
* Corresponding author: Michael Ehrmann, European Central Bank, Kaiserstrasse 29, 60311 Frankfurt am Main, Germany, Email: [email protected].
We thank for comments Refet G€urkaynak, our discussant Anja Baum, as well as participants at seminars at Bonn University, HEI Geneva, the BIS, FU Berlin, University of St Gallen, the ECB, University of Navarra, the Bank of England, the 2010 Konstanz Seminar on Monetary Theory and Policy, the 2010 Finlawmetrics conference, the University of M€unster/Viessmann European Research Centre/NBP conference ‘Heteroge- neous Nations and Globalized Financial Markets: New Challenges for Central Banks’, the CEPR/ESI 14th Annual Conference and the BoK-BIS Conference on Macroprudential Regulation and Policy. We are also grateful to a large number of colleagues in various central banks for their help in identifying the release dates of Financial Stability Reports. Earlier versions of this article have been circulated under the title ‘Macroprudential policy and central bank communication’. This article presents the authors’ personal opinions and does not necessarily reflect the views of the European Central Bank.
[ 701 ]
The Economic Journal, 124 (June), 701–734. Doi: 10.1111/ecoj.12039 © 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
Published by John Wiley & Sons, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
The aim of the current article is therefore to shed light on the potential effects of central bank communication about financial stability. Of course, the objective of macroprudential policies is rather broad (and generally not very precisely defined) – it can be described as countering the decline in measured risks during booms and its rise in busts (Brunnermeier et al., 2009). This is also reflected in the aims of the corresponding central bank communication. The ECB’s FSRs, for instance, aim to promote awareness in the financial industry and among the public at large of issues that are relevant for safeguarding the stability of the euro area financial system. By providing an overview of sources of risk and vulnerability for financial stability, the Review also seeks to play a role in preventing financial crises (European Central Bank, 2010, p. 7).1
Given the breadth of this definition, policy makers might want to achieve a multitude of aims, such as affecting asset prices, their volatility, the skewness or kurtosis of returns. They might also want to change market trends, burst bubbles, or reduce the co-movement of asset prices. Furthermore, the objectives are clearly time-varying and depend on market conditions – during financial crises, central banks typically take measures to support the financial system, whereas they would try to lean against financial booms and prevent crises in normal times. For tractability, our article has a relatively narrow focus; it takes a financial market perspective and studies how financial sector stock indices react to the release of such communication, given that the financial sector is one of its main addressees.
We are in particular interested in the effects on returns and on volatility. The analysis of returns allows us to understand to what extent the views that a central bank expresses in its communications get reflected in the markets. For instance, if the central bank expresses a rather pessimistic view about the prospects for financial stability and this view gets heard in financial markets, we would expect that stock prices for the financial sector decline. In that sense, these communications ‘create news’ (Blinder et al., 2008). With regard to volatility, central banks might either have the desire to reduce uncertainty in financial markets (which should lead to a reduction in volatility), or alternatively to generate a two-way risk for financial market participants or to instil greater uncertainty about a market development that the central bank views as undesirable, and thus to increase volatility.
But why and through what channels should central bank communications have an effect on financial markets at all? A number of factors could come into play here. First, the central bank is obviously an important player in financial markets. For instance, if it is ready to change its policy rates, it can affect asset prices directly. Its communication can therefore exert effects through what has been labelled the ‘signalling channel’ in the literature on foreign exchange interventions (Kaminsky and Lewis, 1996). Second, the analyses that feed into the communications are potentially of high quality and there are few other institutions communicating about financial stability, such that a central bank publication might indeed contain news. Thus, a coordination channel might be at play, whereby communication by the central bank works as a coordination
1 In a similar vein, the Bank of England’s FSRs aim ‘to identify the major downside risks to the UK financial system and thereby help financial firms, authorities and the wider public in managing and preparing for these risks’ See http://www.bankofengland.co.uk/publications/fsr/index.htm.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
702 TH E E CONOM I C J O U RN A L [ J U N E
device, thereby reducing heterogeneity in expectations and information and thus inducing asset prices to more closely reflect the underlying fundamentals, a channel that has also been found to be important to explain the effect of foreign exchange interventions (Sarno and Taylor, 2001; Fratzscher, 2008). This channel might imply that communications have longer-lasting effects, as they might change the dynamics in financial markets.
Importantly, the way central banks reach out to financial markets is most likely time-varying and depends on market conditions – we would expect them to have a much more direct effect during financial crises, when it comes to taking measures to support the financial system, whereas central banks might have a harder time being heard in normal times, when they try to lean against financial booms and prevent crises.
To conduct the empirical analysis, the article constructs a unique and novel database on communication comprising more than 1,000 releases of FSRs and speeches/ interviews by central bank governors from 37 central banks and over the past 14 years. We not only identify the precise timing of these communications but we also determine their content. We employ a computerised textual-analysis software (called Diction 5.0), which allows us to grade each of the central bank financial stability statements, based on different semantic features, according to the degree of optimism that is expressed.
The article’s findings suggest that communication about financial stability has important repercussions for financial sector stock prices. Moreover, there are clear differences between FSRs, on the one hand and speeches and interviews, on the other. FSRs clearly create news in the sense that the views expressed in FSRs move stock markets in the expected direction. This effect is quite sizeable as, on average, FSR releases move equity markets by more than 1% during the subsequent month. Another important finding is that FSRs also reduce market volatility. These effects are particularly strong if the FSR contains an optimistic assessment of the risks to financial stability, when FSRs are found to move equity markets upwards in up to two thirds of the cases. Speeches and interviews, in contrast, have only modest effects on stock market returns and tend not to reduce market volatility.
However, the effects of FSRs and speeches crucially depend on market conditions and other factors. Importantly, during the financial crisis, FSRs were moving financial markets less than before the crisis, while speeches by governors started to move financial markets. Finally, the results indicate that financial stability communication of central banks influences financial markets primarily via a coordination channel, i.e. it provides relevant information which exerts a significant and persistent effect on markets. Our results generalise to overall stock market indices, i.e. are not confined to financial sector stocks.
The article shows that while the release schedule of FSRs is pre-scheduled, speeches and interviews are a much more flexible communication tool. For instance, their number is clearly positively correlated with financial market volatility. Given their flexibility, speeches and interviews by definition carry some surprise element. Since it is often at the discretion of the central bank governors whether or not to make statements about financial stability, the fact that a governor feels compelled to raise financial stability issues in a speech or an interview can therefore be an important
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 703
additional news component. In contrast, due to the fixed release schedule for FSRs, financial markets expect statements about financial stability issues on the release days. There might be surprising elements in their content, but the mere fact that the FSR is released does not come as a surprise. This difference might be at the heart of the different effects of the two instruments on market volatility.
The empirical findings of the article raise a number of policy issues. Communication on financial stability by a central bank has been watched closely and is likely to be watched even more closely in the future, and thus can potentially have an important influence on financial markets. Does this imply that central banks should limit transparency and their communication on financial stability, as argued by Cukierman (2009), or does this make the case for enhanced transparency and accountability, as argued by others (Born et al., 2011)? The findings of the article underline that communication by monetary authorities on financial stability issues can indeed influence financial market developments. Yet the findings also show that such communication may unsettle markets. Hence central bank communication on financial stability needs to be employed with utmost care, stressing the difficulty of designing a successful communication strategy on these matters.
The article proceeds in Section 1 by relating the current article to the existing literature. Section 2 explains the dataset underlying the empirical analysis. In particular, it reports how the measures for central bank communication have been extracted and quantified. It also presents the event study methodology that we employ. Section 3 discusses the empirical results and implications and presents robustness tests. Section 4 concludes.
1. Related Literature
This article relates to the recent literature on central bank communication (for a survey, see Blinder et al., 2008), which has identified three main objectives for monetary policy-related communication, namely
(i) to make central banks credible, (ii) to enhance the effectiveness of monetary policy and (iii) to make central banks accountable.
Interestingly, for all three objectives, there are clear parallels with financial stability- related communication (Born et al., 2012). At the same time, however, there are also important differences between these two types of communication. First, the opera- tional objective of monetary policy is generally much more precisely defined than that of financial stability, which often has multiple facets, instruments and, moreover, authorities that are in charge of it, thus contributing to the complexity of communication about financial stability.
Central banks have become much more transparent about their conduct of monetary policy over the last decades, along with an increasing importance given to communication. There is a debate on possible limits to central bank transparency (Morris and Shin, 2002; Mishkin, 2004; Svensson, 2006) but the arguments are much less contentious than in the case of financial stability-related communication. As demonstrated by Cukierman (2009), a clear case for limiting transparency can be made
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
704 TH E E CONOM I C J O U RN A L [ J U N E
when the central bank has private information about problems within segments of the financial system. Release of such information may potentially be harmful, e.g. by triggering a run on the financial system. This suggests that policy makers need to be even more careful when designing a communication strategy with regard to their financial stability objectives.
While the literature on central bank communication for monetary policy purposes has been growing rapidly over the recent decade, the communication on financial stability has received less attention. Oosterloo and de Haan (2004) found that there is often a lack of accountability requirements for central banks’ financial stability objectives. Svensson (2003) argues that through the publication of indicators of financial stability in FSRs, central banks can issue early warnings to economic agents, thereby ideally preventing financial instability from materialising, ensuring that financial stability concerns do not impose a constraint on monetary policy. Cihak (2006, 2007) provides a systematic overview of FSRs as the main communication channel that central banks use for this purpose. He documents, on the one hand, that the reports have become considerably more sophisticated over time, with substantial improvements in the underlying analytical tools and on the other hand, that there has been a large increase in the number of central banks that publish FSRs. The frontrunners are the Bank of England, the Swedish Riksbank and Norges Bank (Norway’s central bank), all of which started publication in 1996/7. It is probably not a coincidence that these three central banks are typically also listed in the group of the most transparent central banks with regard to monetary policy issues (Eijffinger and Geraats, 2006; Dincer and Eichengreen, 2010). In the meantime, around 50 central banks are now releasing FSRs.
A first empirical analysis of FSRs has been conducted by Oosterloo et al. (2007), with the aim of understanding who publishes FSRs, for what motives and with what content. Their results indicate that there are mainly three motives for publication, namely to increase transparency, to contribute to financial stability and to strengthen cooper- ation between different authorities with financial stability tasks. They also find that the occurrence of a systemic banking crisis in the past is positively related to the likelihood that an FSR is published.
Even less work has been done with regard to the effects of financial stability-related communication. To our knowledge, the only exception is Allen et al. (2004), who conducted an external evaluation of the Riksbank’s work on financial stability issues and came up with a number of recommendations, such as making the objective of the Riksbank’s FSRs explicit, providing the underlying data, or expanding the scope of the FSR to, e.g., the insurance sector. The present article aims to fill this gap and analyses how central bank communications about financial stability are received in financial markets.
2. Measuring Communication and the Effects on Financial Markets
This Section introduces the dataset that we develop to study the effects of financial stability-related communication. We start by explaining the choice of data frequency, the sample of countries and time that we use and the choice of the financial sector stock market indices as our measure for financial markets. Subsequently, we describe
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 705
the process for identifying the relevant communications, how their content is coded, and the econometric methodology.
2.1. Choice of Data Frequency, Data Sample and the Relevant Financial Markets
We are interested in the effects of financial stability-related communication on financial markets. A first choice that is required relates to the frequency of the analysis. Given the speed of reactions in financial markets, it is necessary to identify the timing of the events as precisely as possible. Identification of a precise time stamp will allow for an analysis in a very tight time window around the event, thereby ensuring that the market reaction is not distorted by other news. We opted for a daily frequency for two practical reasons. First, given the aim of providing a cross-country study over a relatively long horizon, financial market data are not consistently available at higher frequencies. Second, the identification of the precise days of the release of central bank communications has already not been trivial in many cases, whereas the identification of the exact time of the release within a day is largely impossible. While a higher frequency might have been desirable, it is important to note that daily frequency is commonly employed in the announcements effect literature – for instance, two classic references with regard to the effect of monetary policy on stock markets, Rigobon and Sack (2004) and Bernanke and Kuttner (2005) both use daily data.
The sample of countries and the time period of the study have been determined on the basis of the release of FSRs. We tried to identify the release dates of the FSRs or relevant speeches or interviews by central bank governors for all those central banks listed in Cihak (2006, 2007), i.e. for all central banks which release FSRs. We succeeded in identifying such release dates for 35 countries, 24 of which are advanced economies according to the IMF’s country classification. Additionally, we included the euro area, as well as the US as the only country that does not release an FSR, restricting ourselves to studying the effect of speeches and interviews in this case. In total, our sample therefore covers 37 central banks (see Table 1). Our sample starts in 1996, i.e. the year when the first FSR was released by the Bank of England. The data were extracted in October 2009, such that the sample ends on September 30, 2009. Importantly, this sample allows differentiating the effects during the global financial crisis from those before, as we would expect the central bank influence to differ substantially across these two periods.
As to the selection of a financial market that shall be subject of this study, we opted for stock market indices relating to the financial sector, as we expect that empirical effects of financial stability communication should be most easily detectable for this sector. Such data are available from Datastream back to 1996, i.e. to the start of our sample period, for all the countries in our sample. This choice is partially due to the large cross-country dimension and the need to get historical data for nearly one and a half decades, which limited the availability of less traditional market measures, such as implied volatilities or expected default frequencies (EDFs). While the link of these measures to financial stability would have been relatively direct, we hope that the financial sector stock indices (using Morgan Stanley Capital International (MSCI) indices) provide a measure that is reasonably closely related to financial stability issues, too. All stock indices are expressed in local currency, given that we are interested in the
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
706 TH E E CONOM I C J O U RN A L [ J U N E
Table 1
Summary Statistics for FSRs and Speeches and Interviews
FSRs Speeches and interviews
By country Argentina 12 13 Australia 11 25 Austria 17 11 Belgium 7 3 Brazil 14 9 Canada 14 22 Chile 11 15 China 5 28 Czech Republic 5 11 Denmark 11 2 Euro Area 10 48 Finland 23 12 France 13 31 Germany 5 58 Greece 1 26 Hong Kong 12 44 Hungary 17 17 Indonesia 6 Ireland 4 2 Israel 6 7 Japan 8 32 Netherlands 8 17 New Zealand 10 18 Norway 20 3 Philippines 50 Poland 10 13 Portugal 5 8 Singapore 7 1 South Africa 11 20 South Korea 9 14 Spain 14 10 Sri Lanka 3 2 Sweden 24 18 Switzerland 7 16 Turkey 8 22 United Kingdom 25 23 United States 111
By year 1996 1 14 1997 3 39 1998 5 118 1999 7 56 2000 10 37 2001 14 17 2002 18 33 2003 25 32 2004 40 26 2005 53 17 2006 51 17 2007 54 68 2008 51 179 2009 35 115
Overall 367 768
Note: The Table shows the number of FSRs and speeches and interviews that are contained in the database, by country and by year.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 707
response of national financial markets to national communication. We will further- more show that our results are robust to using the overall stock market indices, rather than focusing on the financial sector stocks alone.
2.2. Choice and Identification of Communication Events
At the core of this article is a measure of communication events that quantifies their content. We focus on the two most important channels of communication about financial stability issues, namely FSRs and speeches and interviews. FSRs are typically relatively comprehensive documents that discuss various aspects of financial stability. They normally begin with an overall assessment of financial stability in the respective country, often including an international perspective. They usually contain an evaluation of current macroeconomic and financial market developments and the assessment of risks to banks and systemically relevant non-banking financial institutions. Cihak (2006) calls these sections the ‘core’ part of an FSR and differentiates them from the ‘non-core’ part that includes research articles on special issues, often written by outside experts. The weights attributed to these two parts vary considerably across central banks. The spectrum ranges from FSRs that only cover the core part (e.g. Norway) to FSRs which only consist of articles covering a special topic (e.g. France). Most central banks lie somewhere in between this range and are usually closer to the first type. Typically, FSRs are published twice a year, i.e. are relatively infrequent communications.
A second important channel for central banks to communicate about financial stability is to give speeches and interviews. By their very nature, these are much more flexible than FSRs. Their timing can often be chosen flexibly (Ehrmann and Fratzscher (2007) have shown this for monetary policy-related speeches), and their content can be much more focused. Of course, this is also due to the fact that they are much shorter than FSRs.
As we are interested in testing the response of financial markets to central bank communication, we need to identify the release dates as a first step (recall that we will conduct the analysis at a daily frequency, hence there is no need to identify the timing within a given day – as long as the release takes place before markets close). As to FSRs, we carefully ensured a proper identification of their release dates, mainly based on information provided on central banks’ websites and by central bank press offices and complemented with information from news reports about the release of FSRs as recorded in Factiva, a database that contains newspaper articles and newswire reports from 14,000 sources. As shown in Table 1, the dataset contains information on 367 FSRs. The increasing tendency of central banks to publish FSRs is reflected in this database. Starting from less than 10 FSRs per annum in the 1990s, we could identify around 50 FSRs each year in the mid-2000s (note that the drop in numbers in 2009 is entirely due to the fact that the sample ends in September, i.e. covers only three quarters of the year). As to the country coverage, the early publishers are obviously represented more frequently, with 20 and more reports, whereas ‘late movers’ have far fewer observations, down to one for the case of the Bank of Greece, which published its first FSR in June 2009 (for Indonesia and the Philippines, we could not identify the release dates; note that dropping these two countries from the sample does not affect our results in any substantive way).
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
708 TH E E CONOM I C J O U RN A L [ J U N E
Identifying speeches and interviews is more difficult. Our objective is to extract all relevant public statements that relate to financial stability. For tractability, we restricted our search to speeches by the central bank governor – even in cases where a central bank has a member of its governing body who has an explicit assignment regarding financial stability. We used Factiva and extracted all database entries containing the name of the policy maker together with some keywords that appear regularly in the editorials of the FSRs.2 From all hits obtained, we extracted those containing statements by the relevant policy maker with a reference to financial stability issues. Since newswire reports typically record the precise time stamp, we were in a position to allocate the speeches and interviews to the appropriate trading days. Communications during weekends were allocated to the subsequent Monday, communications in the evening – such as dinner speeches – to the subsequent trading day. Furthermore, we very carefully chose only the first report about a given statement, which typically originated from a newswire service. This choice has the advantage that the reporting is very timely, usually coming within minutes of each statement, and that it is mostly descriptive without providing much analysis or interpretation. To avoid double counting, we discarded all subsequent reports or analysis of the same statement.
Several issues are worth noting about this data extraction exercise. First, the search was conducted only for English language items. We might therefore not have discovered all statements, if these were made and reported upon exclusively in other languages. However, this issue should not be very problematic as Factiva also contains newswire reports and the coverage of this topic by newswires is extensive.
Second, one can easily think of other keywords to use in the database search. We have experimented with larger sets, e.g. including also the terms ‘volatile’, ‘volatility’, ‘risk’, ‘adverse’ or ‘pressures’. However, the additional hits typically related to monetary policy communications (such as central bank governors talking about inflationary ‘pressures’, ‘risks’ to price stability etc.), such that the resulting dataset on financial stability communications was basically unaltered.
Third, the news sources might be selective in their reporting, thus possibly not covering all relevant statements. However, given the sensitivity of the topic and the importance that it has for financial markets, we are confident that the coverage is close to complete. Furthermore, as we are interested in testing the market response to communication, it makes sense to focus only on those statements that actually reach market participants and this is best achieved by looking at prominent newswire services.
Fourth, our news sources may wrongly report or misinterpret a statement by policy makers. Again, our objective is to assess communication from the perspective of financial markets and therefore we analyse the information market participants actually receive.
The resulting dataset contains 768 communications. The breakdown by year in Table 1 reveals large time variations, with a massive increase in speeches in 1998,
2 To be precise, we used the following search terms: ‘financial stability or systemic or systemically or crisis or instability or instabilities or unstable or fragile or fragility or fragilities or banking system or disruptive or imbalances or vulnerable or strains’.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 709
i.e. during the Asian and the Russian crisis, as well as during the financial crisis of 2007–9. This suggests that the occurrence of speeches and interviews is responsive to the prevailing circumstances, which is in stark contrast to FSRs, which are typically released at pre-specified dates. Speeches and interviews do therefore provide the central bank with a very flexible instrument for communicating financial stability concerns, as their timing can often be as the bank wishes.
2.3. Measuring the Content of Communications
Once we have identified the communication events, it is necessary to measure their content in order to make the data amenable to econometric analysis. In other words, we want to capture those dimensions and elements of FSRs and speeches/interviews that are relevant for financial market participants and thus will be reflected in asset prices.
A discussion of the various possibilities of achieving this is provided in Blinder et al. (2008). The simplest option consists of assigning a dummy variable that is equal to one on event days and to zero otherwise. While easily done, this approach limits the analysis severely, namely to a study whether communication affects volatility or absolute returns. If we are interested in the effect of the content of communication, a method for quantification of such content is required. The approach adopted in some parts of the literature on monetary policy-related communication, namely, reading the communications and coding them on various scales, was not feasible for our purposes, given the amount of text that needed to be quantified. We have therefore opted for an automated approach for the current article.3
We used the computerised textual-analysis software DICTION 5.0, which searches text for different semantic features by using a corpus of several thousand words and scores the text along an optimism dimension. This dimension may be important as it provides agents with information about the current state and the prospects of the financial system and underlying risks. The respective scores are computed by adding the standardised word frequencies of various subcategories labelled as optimistic and by subtracting the corresponding frequencies of pessimistic subcategories. In broad terms, optimism refers to ‘language endorsing some person, group, concept or event, or highlighting their positive entailments’.4
This software has been used extensively in communication sciences and in political sciences, e.g. for analysing speeches of politicians (Hart and Jarvis, 1997; Hart, 2000), but has also been applied in the context of central banks (Bligh and Hess, 2007; Armesto et al., 2009). Furthermore, Davis et al. (2012) have used it to measure the
3 An alternative approach is used by Lucca and Trebbi (2009), where FOMC statements are cut down into small segments of text, the semantic orientation of which is then calculated by checking how often these text segments appear in conjunction with the words dovish or hawkish in a large body of text.
4 The scores are computed using scores from six subcategories, by adding the standardised word frequencies of the subcategories labelled as optimism increasing by DICTION (praise, satisfaction and inspiration), while our pessimism score is computed by adding the standardised word frequencies of the subcategories labelled optimism decreasing by DICTION (blame, hardship and denial). The Praise score, for example, includes words that isolate social qualities (witty), physical qualities (strong), intellectual qualities (reasonable), entrepreneurial qualities (successful) and moral qualities (good). For more details, see the DICTION 5.0 manual http://www.dictionsoftware.com/files/dictionmanual.pdf.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
710 TH E E CONOM I C J O U RN A L [ J U N E
reaction of financial markets to earnings announcements and find a significant incremental market response to optimistic and pessimistic language usage in earnings press releases.
There are several advantages of this approach over human coding of the text. First, the software creates a coding that is more mechanical and thus objective, compared to human coding which tends to be more judgmental. While some subjectivity could arise due to the choice of the content of the dictionaries against which a text is assessed, it is important to note that the corpus has been defined based on linguistic theory and without active participation by the authors of this article. Another advantage is the replicability of the coding, which is in stark contrast to human coding and also allows more text to be added without distorting the scoring process. Third, the automated approach allows a consistent coding of long passages of text and across a large number of communications. Human coding of long texts is rather difficult, as no part should in principle be given a larger weight in the assessment. Given the breadth of FSRs, this issue is particularly severe in the current application. At the same time, a drawback of the automated approach is that it does not consider the context of the text and thus cannot generate a ‘tailor-made’ coding for financial stability-related communication.
Based on this computerised textual-analysis software, we computed a score for each individual speech or interview (note that, effectively, we are coding the content of the related news reports, rather than the original source text) and for the overview part of each FSR.5 Subsequently, we transformed the resulting scores into a discrete variable, which takes the value of �1 for the lowest third of the distribution, a value of 0 for the middle part of the distribution, and the value of +1 for the upper third of the distribution. That is, a value of +1 denotes a relatively optimistic text, while a value of �1 corresponds to a relatively pessimistic statement. The discretisation of scores is required for the subsequent analysis, where we are interested in the market effects of optimistic versus pessimistic communications, rather than the effect of an incremental change in tone. This transformation was applied for the speeches as well as for the FSRs. Note that we test for robustness using a very different measurement approach, which also attempts to capture the surprise component contained in the respective communications, as well as (for the parts of the subsequent analysis where discretisation is not required) using the raw optimism scores given by the software.
It is important to note that this implies a relative coding, i.e. a given communication is scored in a comparative fashion against the other texts in the sample. However, due to the large sample, both across countries and along the time dimension, our communications cover periods of relative stability and tranquillity, as well as periods of financial market crises or turbulence. Accordingly, the overall sample of text should be relatively balanced, such that text which is coded with plus or minus one should indeed represent a corresponding opinion. We denote the resulting indicators by I
optimism;FSR it
and I optimism;speech it , respectively, where i denotes a given country and t stands for time.
5 While this overview carries different names across central banks, e.g. editorial, introductory chapter, executive summary, etc., it is rather similar in nature for all FSRs.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 711
To get back to our initial example of Alan Greenspan’s irrational exuberance speech, both the speech as such and the corresponding Reuters article6 would be classified as pessimistic. In the Appendix, we provide more examples of speeches and interviews and of how they were coded.
2.4. The Event Study Methodology
What are the effects of FSRs and speeches/interviews on financial markets? The natural econometric approach to testing our hypotheses of interest is the event study methodology. We use this methodology because we are interested not only in the contemporaneous effect of financial stability statements, but we also want to know how persistent the effect is over time. We can define the release of an FSR, or the delivery of a speech or an interview as an event. The question we want to address is whether the event affects stock markets in a causal fashion. For that purpose, it is essential that we can compare the stock market evolution following the event to the counterfactual, i.e. a predicted value that we believe would have occurred had the event not happened. A crucial issue in any event study is therefore to find a benchmark model to calculate expected returns, which in turn allows calculation of abnormal returns.7 Most event studies look at the effect of events, such as earnings announcements or stock splits, on individual stocks and use some variant of a factor model, such as the Fama and French (1993) three-factor model or the Carhart (1997) four-factor model, which extends the previous model by a momentum factor.
Given that we are interested in the evolution of national stock market indices rather than of individual stocks, the book-to-market ratio and the size factor of the Fama– French model are not applicable. Following Edmans et al. (2007) and Pojarliev and Levich (2008), we start by defining normal returns as:
Rit ¼ c0i þ c1iRit�1 þ c2iRmt�1 þ c3iRmt þ c4iRmtþ1
þ c5iDt þ c6iTit�1 þ c7iSit�1 þ c8iMit�1 þ eit ; ð1Þ where Rit is the daily local currency return on the financial sector stock market index for country i on day t, Rmt is the daily US dollar return on Datastream’s global financial sector stock market index and Dt denotes dummy variables for Monday to Thursday. Tit�1 stands for the trend in stock markets over the 20 days prior to the event, Sit�1 for the standard deviation of daily stock market returns over the 20 days prior to the event and Mit�1 for the ‘misalignment’ of stock indices on the day preceding the event, measured as the percentage deviation of the stock indices from their national average over the entire sample period.
The first five factors follow Edmans et al. (2007). The lagged index return controls for possible first-order serial correlation. The global stock market index is meant to capture the effects of international stock market integration and since some indices
6 ‘Stock markets can become too exuberant – Greenspan. WASHINGTON, Dec 5 (Reuters) – The Federal Reserve must be wary when “irrational exuberance” infects stock and other asset markets because that could end up doing damage to the economy, Fed Chairman Alan Greenspan said on Thursday. […]’.
7 For overviews of the event study literature see, for example, MacKinlay (1997) or Kothari and Warner (2007).
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
712 TH E E CONOM I C J O U RN A L [ J U N E
might be lagging or leading the world index, Edmans et al. (2007) not only include the contemporaneous global returns, but furthermore a lead and a lag. By doing so, we assume exogeneity of Rm – if outcomes of central bank communications are correlated across countries, this biases the results against us. We test for the presence of international spillovers and find these to be very limited, such that the assumption of exogeneity might not be particularly stringent. The last three terms are due to earlier event studies on exchange rates such as Pojarliev and Levich (2008) or Fratzscher (2009). The trend factor attempts to allow for persistence in stock market movements and is therefore closely related to the momentum factor in the Carhart four-factor model. The inclusion of the standard deviation is an attempt to capture the effect of market volatility. Finally, the misalignment factor is based on the idea that there might be booms or busts in stock markets and that over a sufficiently long sample, there could be some mean reversion (albeit possibly allowing for a drift). We test for robustness to the exclusion of these last three terms, given that they are derived from the exchange rate literature rather than the stock market event studies and find our results to be qualitatively unaltered.
Model (1) is estimated country by country, only including days that were neither preceding nor preceded by communication events for 60 days (in each direction). Based on the estimated parameters (denoted by hats), it is then possible to calculate abnormal returns on event days as
êit ¼Rit � ðĉ0i þ ĉ1iRit�1 þ ĉ2iRmt�1 þ ĉ3iRmt þ ĉ4iRmtþ1
þ ĉ5iDt þ ĉ6iTit�1 þ ĉ7iSit�1 þ ĉ8iMit�1Þ: ð2Þ The hypothesis to be tested is whether communication leads to abnormal returns in
the expected direction, i.e. whether
êit [ 0 if I optimism;c it ¼ 1 or êit\0 if I
optimism;c it ¼ �1; ð3Þ
where the superscript c stands for the two communication types, FSR and speeches or interviews. A more complex approach is required if we want to calculate the longer- term effects of communication beyond the event day. While we assume that world markets are exogenous to a communication in an individual country also over extended time windows, this is obviously not the case for the own lag, the recent trend, standard deviation and misalignment: as of the second day, it is necessary to calculate predicted returns for the preceding day and to substitute these into (2), thus yielding
êitþk ¼Ritþk � ðĉ0i þ ĉ1i R̂itþk�1 þ ĉ2iRmtþk�1 þ ĉ3iRmtþk þ ĉ4iRmtþkþ1
þ ĉ5iDtþk þ ĉ6i T̂itþk�1 þ ĉ7i Ŝitþk�1 þ ĉ8iM̂itþk�1Þ: ð4Þ Note that compared to (2), Rit�1, Tit�1, Sit�1 and Mit�1 have all been replaced by their predicted value in the absence of a communication event. For k ¼ 0, the two coincide, whereas for all days k > 0, it is important to calculate the appropriate predicted values. Tests for the effects of communication over longer time horizons with a time window of K days then amount to asking whether
XK
k¼0
êitþk [ 0 if I optimism;c it ¼ 1 or
XK
k¼0
êitþk\0 if I optimism;c it ¼ �1: ð5Þ
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 713
Following common practice in the event study literature, we employ two types of tests for the effects of communications (both described in detail in MacKinlay, 1997). First, we apply a non-parametric sign test to study whether the above conditions hold in more than 50% of all cases. The underlying idea is that by construction – if the factor model is correct – abnormal returns and cumulated abnormal returns are on average zero and that it is equally probable that they are positive or negative. If the events systematically move stock markets in the expected direction, we should find that the abnormal returns are non-zero and of the expected sign, in significantly more than 50% of cases. The second (parametric) test checks the average size of the (cumulated) abnormal returns and tests these against the null hypothesis that they are zero.
In a similar vein, to test whether communications affect stock market volatility, we furthermore study whether
rêi;t=tþk \rêi;t�1=t�1�k
if Dc it ¼ 1; ð6Þ
where rêi;t=tþk is the standard deviation of daily abnormal returns in country i from time t
to t + k, rêi;t�1=t�1�k , their standard deviation over the k days prior to the event and Dc
it a dummy variable that is equal to one on the days when a communication of type c is released in country i.8 Also here, we apply the non-parametric sign test whether the above conditions hold in more than 50% of all cases and the test whether the difference of the two standard deviations is equal to zero.
3. The Effects of Financial Stability-related Communication
This Section starts by identifying and testing for the effects of FSRs on financial markets (subsection 3.1). It proceeds by presenting a number of sample splits and robustness tests that also shed further light on the channels through which FSRs affect markets (subsection 3.2), before it turns to discussing the effect of speeches and interviews in subsection 3.3.
3.1. Effects of FSRs
We now turn to the question of the extent to which FSRs were affecting financial markets. A first test is provided in Figure 1, which compares the actual evolution of stock markets following FSRs to the predicted evolution on the basis of the benchmark model (1). The solid line plots the average actual cumulated returns over 60 days following the communication events. The dashed line, in contrast, shows the expected cumulated returns that would result from the benchmark model in the absence of a communication event. To combine pessimistic as well as optimistic FSRs in one chart, the cumulated returns are multiplied by �1 for pessimistic communications, whereas they are left unchanged for optimistic communications. Accordingly, we would expect the actual returns to lie above the predicted returns after FSRs if the markets follow the point of view expressed by the central bank (i.e. we observe negative abnormal returns
8 Excluding the daily abnormal returns on day t from calculating the post-event standard deviations does not alter our results. This implies that the results are not driven by the initial market reaction on the day of the announcement.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
714 TH E E CONOM I C J O U RN A L [ J U N E
in response to pessimistic statements and positive ones in the case of optimistic communications).
Figure 1 provides a compelling picture of the effects of central bank communication. It shows that markets move in the direction of the central bank view, since the actual returns are substantially larger than the predicted returns. Moreover, the effect is quite sizeable economically: for several time windows, FSR releases move equity markets on average by more than 1% in the direction indicated by the FSRs.
Figure 1 also suggests that central bank communications are potentially affecting financial markets even at very long horizons, given that the gap between predicted and actual cumulated returns is present for the entire horizon of time windows we look at and begins to narrow only towards the end of the horizon.
The formal test results for the effects of FSRs are provided in Table 2. The first set of results relates to (5), i.e. tests whether optimistic statements yield positive abnormal returns, and pessimistic ones lead to negative abnormal returns. The first column shows the share of cases in which the condition was met, as well as the results of the non-parametric sign test. Shares above 0.5 would suggest that stock markets move in the direction of the content of communications. The statistical significance is assessed by asterisks (*** for 1%, ** for 5% and * for 10% significance) – whereas numbers that are significantly smaller than 0.5 would be characterised by apostrophes (‴ for 1%, ″ for 5% and ′ for 10% significance).
There is clear evidence that the views represented in FSRs get reflected in financial markets, in significantly more than 50% of all cases. In terms of magnitudes, which
–0.5
0
0.5
1.0
1.5
2.0
2.5
1 2 3 4 5 10 15 20 25 30 35 40 45 50 55 60
Predicted Actual
Fig. 1. Predicted Versus Actual Evolution of Stock Markets After FSRs Notes. The Figure compares the actual evolution of cumulated stock market returns (in %) following FSRs to the predicted evolution on the basis of the benchmark model (1). The solid line plots the average actual cumulated returns starting from day 1 after the communication event and up to day 60. The dashed line shows the expected cumulated returns that would result from the benchmark model in the absence of a communication event. The cumulated returns are multiplied by �1 for pessimistic communications, whereas they are left unchanged for optimistic communications.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 715
T ab
le 2
E ff ec ts of
FS R s
N o o f d ay s
Jo in t m o d el
P es si m is ti c F SR
s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
1 0. 54
0. 27
** *
– –
0. 44
′ �0
.3 3
– –
2 0. 54
0. 33
** –
– 0. 46
�0 .5 4
– –
3 0. 58
** 0. 46
** *
– –
0. 40
″ �0
.7 5
– –
4 0. 57
** 0. 54
** *
0. 51
�0 .0 8*
0. 39
‴ �0
.7 3
0. 50
�0 .1 0
5 0. 53
0. 44
** 0. 53
�0 .0 7*
0. 47
�0 .4 9
0. 51
�0 .0 5
10 0. 53
0. 63
** 0. 55
** �0
.0 8*
* 0. 50
�0 .4 0
0. 48
�0 .0 4
15 0. 57
** 0. 64
** 0. 52
�0 .0 6*
0. 44
′ �0
.2 9
0. 51
�0 .0 8
20 0. 56
** 0. 92
** 0. 55
** �0
.0 5*
0. 50
�0 .0 1
0. 56
* �0
.0 7
25 0. 57
** 1. 27
** *
0. 56
** *
�0 .0 7*
* 0. 48
�0 .2 8
0. 60
** �0
.1 3*
* 30
0. 58
** *
1. 39
** *
0. 56
** *
�0 .0 5*
0. 49
�0 .3 6
0. 57
** �0
.1 1*
35 0. 57
** 1. 27
** *
0. 56
** *
�0 .0 5*
0. 51
0. 11
0. 56
* �0
.1 0*
40 0. 53
1. 21
** 0. 55
** �0
.0 4
0. 54
0. 33
0. 52
�0 .0 7
45 0. 56
** 1. 41
** *
0. 55
** �0
.0 5*
0. 52
0. 17
0. 56
* �0
.1 2*
* 50
0. 56
** 1. 60
** *
0. 56
** *
�0 .0 6*
* 0. 51
�0 .1 1
0. 56
* �0
.1 2*
* 55
0. 56
** 1. 47
** 0. 56
** �0
.0 5*
0. 55
0. 31
0. 58
** �0
.1 1*
* 60
0. 55
* 1. 21
** 0. 55
** �0
.0 5*
0. 54
0. 62
0. 58
** �0
.1 1*
*
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
716 TH E E CONOM I C J O U RN A L [ J U N E
N o o f d ay s
N eu
tr al
F SR
s O p ti m is ti c F SR
s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
1 0. 49
�0 .0 9
– –
0. 53
0. 20
* –
– 2
0. 52
0. 10
– –
0. 55
0. 14
– –
3 0. 53
0. 18
– –
0. 55
0. 20
– –
4 0. 51
0. 12
0. 52
�0 .1 3*
* 0. 54
0. 37
** 0. 51
�0 .0 2
5 0. 55
0. 21
0. 55
�0 .1 1*
0. 54
0. 39
* 0. 53
�0 .0 5
10 0. 48
�0 .3 8
0. 55
�0 .0 8
0. 56
* 0. 84
** *
0. 61
** *
�0 .1 1*
* 15
0. 51
0. 02
0. 50
�0 .0 2
0. 58
** 0. 95
** *
0. 55
* �0
.0 8*
* 20
0. 61
** 0. 51
0. 55
�0 .0 2
0. 61
** *
1. 75
** *
0. 55
�0 .0 6*
* 25
0. 59
** 0. 69
0. 55
�0 .0 4
0. 62
** *
2. 18
** *
0. 54
�0 .0 5
30 0. 59
** 0. 90
* 0. 57
* �0
.0 2
0. 65
** *
2. 33
** *
0. 55
�0 .0 3
35 0. 62
** *
1. 50
** 0. 55
�0 .0 4
0. 64
** *
2. 53
** *
0. 58
** �0
.0 1
40 0. 61
** 1. 47
** 0. 57
* �0
.0 3
0. 59
** 2. 63
** *
0. 56
* �0
.0 1
45 0. 57
* 1. 45
** 0. 58
** �0
.0 4
0. 63
** *
2. 86
** *
0. 52
�0 .0 1
50 0. 59
** 1. 46
* 0. 58
** �0
.0 5
0. 63
** *
2. 97
** *
0. 55
�0 .0 1
55 0. 61
** 2. 12
** 0. 57
* �0
.0 4
0. 66
** *
3. 09
** *
0. 52
0. 00
60 0. 61
** 2. 66
** *
0. 55
�0 .0 5
0. 63
** *
2. 87
** *
0. 52
0. 00
N ot es : T h e
T ab
le sh o w s re su lt s o f th e
te st
fo r co
m m u n ic at io n
ef fe ct s.
T h e
fi rs t se t o f re su lt s (r et u rn s,
n o n -p ar am
et ri c)
te st s th e
sh ar e
o f ca se s in
w h ic h
P K k¼
0 ê i tþ
k [
0 if
Io p ti m is m ;F SR
it ¼
1 o r P
K k¼ 0 ê i tþ
k \ 0 if
Io p ti m is m ;F SR
it ¼
�1 , fo r d if fe re n t ti m e w in d o w s K
in th e ro w s o f th e T ab
le . T h e se co
n d
co lu m n
(r et u rn s,
p ar am
et ri c)
sh o w s th e av er ag e si ze
o f th e cu
m u la te d ab
n o rm
al re tu rn s ð1 = N ÞP
N n ¼1
P K k¼
0 Io
p ti m is m ;F SR
n t
ê n tþ
k an
d te st s w h et h er
th es e ar e d if fe re n t fr o m
ze ro . T h e
co lu m n s fo r ‘S D ’ sh o w th e sh ar e o f ca se s in
w h ic h th e st an
d ar d d ev ia ti o n o f ab
n o rm
al re tu rn s o ve r k d ay s af te r th e re le as e o f an
F SR
is sm
al le r th an
th e st an
d ar d
d ev ia ti o n
d u ri n g th e k d ay s p ri o r to
th e re le as e,
i. e.
r ê i; t= tþ
k \ r ê
i; t�
1 = t�
1� k if
D c it ¼
1 (n
o n -p ar am
et ri c) , an
d th ei r av er ag e d if fe re n ce
(p ar am
et ri c)
an d
te st s th es e
ag ai n st 0. 5 an
d 0,
re sp ec ti ve ly .T
h e p an
el s ‘P es si m is ti c F SR
s’ ,‘ N eu
tr al
F SR
s’ an
d ‘O
p ti m is ti c F SR
s’ o f th e T ab
le re p ea ts th e ex
er ci se
fo r F SR
s th at
h av e b ee
n co
d ed
as
Io p ti m is m ;F SR
it ¼
�1 , Io
p ti m is m ;F SR
it ¼
0 an
d Io
p ti m is m ;F SR
it ¼
1, re sp ec ti ve ly . St an
d ar d
d ev ia ti o n s ar e o n ly
ca lc u la te d
fo r ti m e w in d o w s ex
ce ed
in g th re e b u si n es s d ay s.
** *,
** , an
d *i n d ic at e st at is ti ca l si gn
ifi ca n ce
ag ai n st th e n u ll h yp o th es is at
th e 1%
, 5%
an d 10
% le ve ls , re sp ec ti ve ly . ‴,
″, an
d ′i n d ic at e st at is ti ca l si gn
ifi ca n ce
ag ai n st
th e al te rn at iv e h yp o th es is at
th e 1%
, 5%
an d 10
% le ve ls , re sp ec ti ve ly .
T ab
le 2
(C on ti n u ed )
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 717
are reported in the second column, FSRs generate abnormal returns on the day of the release of 0.27% on average and cumulated abnormal returns up to 1.6% in the longer run, with the largest effects found after 25–50 trading days, i.e. after 5–10 weeks. Such an effect is indeed sizeable and economically meaningful, in particular when considering that FSRs are generally released twice a year per country.
Are these effects equally generated by optimistic, neutral or pessimistic communi- cations? The existing announcement literature does not help in generating clear-cut hypotheses, because the various studies of differential responses of asset prices to positive and negative news have come to conflicting results: whereas Andersen et al. (2003), for instance, show that asset prices respond more strongly to negative news, Entorf et al. (2012) come to the opposite conclusion.
Table 2 provides a breakdown of our results according to the type of the FSR and reveals that optimistic FSRs affect financial markets particularly strongly. They typically generate positive abnormal returns, which are furthermore large in magnitude, thus leading to statistically significant estimates. The cumulated abnormal returns are largest after 55 days, amounting to more than 3%. This suggests that an optimistic assessment provided in FSRs leads to an improvement in stock market sentiment over a fairly long horizon, in a way that is not matched by pessimistic FSRs leading to a deterioration in sentiment. Importantly, this implies that FSRs tend not to be successful in conveying warning signals about systemic risks, which is of course one of their main intentions.
As we show, this result depends crucially on the market environment – prior to the global financial crisis, optimistic FSRs lead to increasing stock returns, whereas this effect disappears during the financial crisis. One possibility for rationalising these findings is therefore the existence of a confirmation bias (Lord et al., 1979; Hirshleifer, 2001), whereby stock market participants react more strongly to news that confirms their previous beliefs, which in this case boils down to a stronger reaction to positive news.
Table 2 also provides the results for tests of whether the release of FSRs lowers stock market volatility, i.e. tests of whether condition (6) holds, again using both the non- parametric sign test and the parametric test. There is compelling evidence that FSRs do indeed lead to a significant reduction in market volatility.
To summarise, these findings suggest, first, that communication about financial stability has the potential to affect financial markets. The views expressed in FSRs get reflected in stock market returns in a long-lasting fashion, in particular if the FSR contains an optimistic assessment of the risks to financial stability. FSRs also manage to reduce market volatility somewhat.
3.2. Sample Splits and Robustness
We have subjected our benchmark results to a number of sample splits and robustness tests and studied various extensions, which we now describe. All results are provided in Table 3. Given the large number of tests, we only show results for a time window of 25 business days.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
718 TH E E CONOM I C J O U RN A L [ J U N E
T ab
le 3
E ff ec ts of
FS R s – Sa m pl e Sp li ts an
d R ob u st n es s
Jo in t m o d el
P es si m is ti c F SR
s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
(A ) B en ch m ar k
0. 57
** 1. 27
** *
0. 56
** *
�0 .0 7*
* 0. 48
�0 .2 8
0. 60
** �0
.1 3*
* (B
) Sa m pl e sp li ts
1. C o u n tr y gr o u p
A d va n ce d ec o n o m ie s
0. 56
* 0. 91
** 0. 59
** *
�0 .1 1*
** 0. 46
�0 .3 1
0. 62
** *
�0 .2 0*
** E m er gi n g ec o n o m ie s
0. 62
** 2. 27
** 0. 48
0. 05
0. 55
�0 .1 4
0. 50
0. 20
2. C ri si s ve rs u s p re -c ri si s
P re -c ri si s
0. 63
** *
2. 10
** *
0. 55
* �0
.0 5*
* 0. 39
″ �1
.0 6
0. 61
** �0
.0 8*
F in an
ci al
cr is is 20
07 –1
0 0. 45
�0 .5 5
0. 60
** �0
.1 3*
0. 58
0. 65
0. 58
* �0
.2 0
3. Su
p er vi so ry
ro le
C B is su p er vi so r
0. 56
1. 47
** 0. 55
* �0
.0 9*
0. 61
0. 96
0. 39
�0 .1 5
C B is n o t su p er vi so r
0. 58
** 1. 17
** 0. 57
** �0
.0 6*
0. 44
�0 .6 7
0. 66
** *
�0 .1 3*
(C ) R ob u st n es s
A ll st o ck s
0. 58
** *
1. 16
** *
0. 55
** �0
.0 2
0. 50
�0 .3 1
0. 58
** �0
.0 5
R is k- fr ee
ra te
as p re d ic to r
0. 58
** *
1. 42
** *
0. 60
** *
�0 .1 0*
* 0. 55
0. 03
0. 65
** *
�0 .2 3*
** A lt er n at iv e co
d in g
0. 53
0. 72
* 0. 56
** *
�0 .0 7*
* 0. 52
0. 43
0. 52
0. 01
R aw
D ic ti o n sc o re s
– 0. 50
** *
– –
– –
– –
(D ) T es ti n g fo r th e si gn al li n g ch an
n el
Sh o rt -t er m
in te re st
ra te s
0. 54
* 0. 05
0. 55
** 0. 01
0. 50
�0 .0 6
0. 59
** 0. 00
L o n g- te rm
in te re st
ra te s
0. 53
0. 02
0. 52
0. 00
0. 45
�0 .0 3
0. 51
0. 00
(E ) In te rn at io n al
sp il lo ve rs
U K F SR
s 0. 47
�1 .0 1″
0. 54
** 0. 03
0. 53
1. 43
** 0. 57
** *
0. 00
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 719
N eu
tr al
F SR
s O p ti m is ti c F SR
s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
(A ) B en ch m ar k
0. 59
** 0. 69
0. 55
�0 .0 4
0. 62
** *
2. 18
** *
0. 54
�0 .0 5
(B ) Sa m pl e sp li ts
(1 ) C o u n tr y gr o u p
A d va n ce d ec o n o m ie s
0. 57
* 1. 10
** 0. 59
** �0
.0 9*
* 0. 58
* 1. 62
** *
0. 55
�0 .0 3
E m er gi n g ec o n o m ie s
0. 64
�0 .8 3
0. 40
0. 17
0. 69
** *
3. 21
** *
0. 51
�0 .0 7
(2 )C
ri si s ve rs u s p re -c ri si s
P re -c ri si s
0. 61
** 0. 77
* 0. 55
�0 .0 1
0. 64
** *
2. 73
** *
0. 51
�0 .0 5*
F in an
ci al
cr is is 20
07 –1
0 0. 52
0. 47
0. 58
�0 .1 1
0. 52
�0 .3 2
0. 65
** �0
.0 1
(3 )S u p er vi so ry
ro le
C B is su p er vi so r
0. 63
* 0. 50
0. 56
�0 .0 5
0. 64
** 2. 63
** *
0. 63
** �0
.1 0*
C B is n o t su p er vi so r
0. 57
0. 79
0. 55
�0 .0 4
0. 59
* 1. 80
** *
0. 46
0. 00
(C ) R ob u st n es s
A ll st o ck s
0. 54
0. 00
0. 48
0. 04
0. 66
** *
1. 96
** *
0. 60
** *
�0 .0 6*
* R is k- fr ee
ra te
as p re d ic to r
0. 63
** *
1. 11
0. 53
�0 .0 2
0. 70
** *
2. 76
** *
0. 61
** *
�0 .0 5
A lt er n at iv e co
d in g
0. 57
* 0. 40
0. 63
** *
�0 .1 2*
* 0. 59
** 1. 86
** *
0. 54
�0 .1 0*
* R aw
D ic ti o n sc o re s
– –
– –
– –
– –
(D ) T es ti n g fo r th e si gn al li n g ch an
n el
Sh o rt -t er m
in te re st
ra te s
0. 58
* 0. 12
* 0. 51
0. 01
‴ 0. 58
** 0. 04
0. 55
0. 00
L o n g- te rm
in te re st
ra te s
0. 57
* 0. 03
0. 50
0. 00
0. 52
0. 00
0. 56
* 0. 00
(E ) In te rn at io n al
sp il lo ve rs
U K F SR
s 0. 59
** *
0. 65
* 0. 50
0. 06
′ 0. 48
0. 81
0. 70
** *
�0 .0 9
N ot es : Se
e n o te s to
T ab
le 2.
A ll re su lt s re la te
to th e ef fe ct
o f F SR
s at
a ti m e w in d o w o f 25
d ay s. R o w 1 re p o rt s th e b en
ch m ar k re su lt s, ea ch
su b se q u en
t ro w re p o rt s
re su lt s o f a sp ec ifi c sa m p le
sp li t o r ro b u st n es s te st . Sa m p le
sp li ts fo r ad
va n ce d / em
er gi n g ec o n o m ie s, p re -c ri si s/ fi n an
ci al
cr is is , C B as
su p er vi so r o r n o t. R o b u st n es s
te st s re la te
to u si n g o ve ra ll st o ck
in d ic es
ra th er
th an
fi n an
ci al
se ct o r st o ck s in d ic es ,t o re p la ci n g th e m o d el
o f p re d ic te d re tu rn s in
(1 ) b y th e ri sk
fr ee
ra te ,a
s w el l as
to u si n g an
al te rn at iv e co
d in g o f th e co
n te n t o f th e co
m m u n ic at io n s, o r u si n g th e ra w D ic ti o n o p ti m is m
sc o re s d ir ec tl y, ra th er
th an
th ei r d is cr et is ed
ve rs io n s. P an
el (D
) sh o w s th e ef fe ct s o n sh o rt
an d lo n g- te rm
in te re st
ra te s, an
d p an
el (E ) te st s fo r th e in te rn at io n al
ef fe ct s o f U K F SR
s.
T ab
le 3
(C on ti n u ed )
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
720 TH E E CONOM I C J O U RN A L [ J U N E
3.2.1. Sample splits The first set of results relates to various sample splits. Given the large number of countries and the long time sample, it might be the case that there is substantial heterogeneity across countries or over time that we do not capture in the full sample. The first such split addresses possible cross-country heterogeneity, by re-running the estimation separately for all advanced and all emerging market economies (following the IMF’s country classification). Results are overall robust. The interesting insight, though, is that there is a reduction in volatility following FSRs by central banks in advanced countries, which is not found in emerging market economies.
The existing literature does not provide much guidance how to interpret these differences. Two studies explicitly compare central bank-related announcements and the corresponding market response in advanced and emerging market economies, namely Kuttner and Posen (2010) and Moser and Dreher (2010), both of which analyse the effect of appointments of new central bank governors. Moser and Dreher find for emerging market economies that such appointments can lead to a loss of credibility, whereas no such effect is found by Kuttner and Posen in advanced economies. If we apply these results to our case, a lower level of credibility could indeed also imply that FSRs are less capable of calming markets in emerging market economies, thus explaining the absence of a reduction in volatility in their case.
Also the second split along the time dimension reveals interesting patterns. Separate tests for the period prior to the financial crisis 2007–9 (defining the starting date in September 2007, i.e. with Northern Rock; defining the start of the crisis with Lehman does not affect our results) and the time of the crisis shows that FSRs have exerted no systematic effect on stock markets during the crisis.
The third sample split intends to identify whether the role of the central bank in financial supervision matters, by testing once for the effects of communication by central banks that do have a formal role in financial supervision and once for those central banks without such a task. The classification is based on the Central Bank as Financial Authority (CBFA) index developed in Masciandaro and Quintyn (2009).9
This differentiation does not seem to play an important role, given that the results are robust, and no major differences between the two groups emerge.
Further sample splits could be interesting to study. For instance, the style of FSRs has changed over time and differs across countries, such that a split into relatively more and less informative FSRs could be interesting. We leave this for future research.
3.2.2. Robustness The rows of panel (C) in Table 3 present several robustness tests. First, replacing the financial sector stock indices with the broad national stock market index, we can test whether our results apply more broadly, or are confined to the financial sector. In this
9 This index takes the value 1 if the central bank is not assigned the main responsibility for banking supervision; 2 if the central bank has the main (or sole) responsibility for banking supervision; 3 if the central bank has furthermore responsibility for either insurances or the securities markets; 4 if the central bank has responsibility in all three sectors. We allocate central banks to the group with supervisory functions if their index value is larger than one.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 721
case, the prediction model of (1) is re-estimated, with the global financial sector stock market index replaced by the overall global stock market index as provided by MSCI. The results are remarkably robust.
A second robustness test replaces the prediction model of (1) by a very simple alternative, namely the risk-free rate (proxied by the three-month US t-bill rate). Here also, results are robust qualitatively, and even strengthen quantitatively.10
Furthermore, results are also not sensitive to the precise way we had split the communications into optimistic and pessimistic content. To test for this, we take two routes: first, by defining an alternative approach to discretising the codes that attempts to control for the expected component contained in the communication and to construct a surprise measure instead. We do so by means of the following auxiliary regression:
C optimism;c it ¼ a0i þ a1q þ a2Tit�1 þ a3Sit�1 þ a4Mit�1 þ lit ; ð7Þ
where C optimism;c it denotes the raw Diction coding of a given communication of type c
along the optimism dimension and a0i and a1q are country-fixed effects and time- fixed effects for each quarter of the sample, respectively. The country-fixed effects allow for the possibility that there is a different style in the reporting, thus leading to a different mean coding for each country. Such differences should be well known to observers and therefore not be a surprise. The time-fixed effects control for a common evolution across countries, given that often developments in financial markets are internationally determined. Such common time patterns should also not come as a surprise to financial markets. The last three explanatory factors are as described in benchmark model (1), i.e. they control for the trend, for stock market volatility and for a possible stock market misalignment. We retrieve the residuals l̂it from these regressions and define a communication to be optimistic if l̂it is above the 66th percentile in the distribution, as pessimistic if it is below the 33rd percentile and as neutral otherwise. Even though this classification is very different from the original, unconditional, one, it turns out that the results are remarkably robust.
Our second test for the role of our discretisation method reverts to the original, raw, scores generated by Diction. Higher scores denote more optimistic communications, such that we would expect stock returns to increase correspondingly. This is indeed what we find, consistently with our earlier results. With this measure, we are of course not able to separate out optimistic and pessimistic communications, such that we are neither able to conduct the non-parametric test, nor to fill the table where we break down the results by the content of the communication.
3.2.3. The importance of the signalling channel The next point we address here is through which channel communication affects financial markets. Is it that communication affects markets because it contains relevant information and thus coordinates markets and functions as a focal
10 We have also tested whether replacing the global financial sector stock index in (1) by the overall global stock index changes results and found this not to be the case (results available upon request).
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
722 TH E E CONOM I C J O U RN A L [ J U N E
point – akin to what is known as a coordination channel (Sarno and Taylor, 2001; Fratzscher, 2008)? Or is it that market participants believe that financial stability communication has a bearing on monetary policy decisions by central banks – or what is referred to as a signalling channel? The evidence discussed so far, in particular the persistence of the effects of communication, strongly points towards the coordination channel being at work (Sarno and Taylor, 2001). Yet a more direct test of these two channels is to ask whether financial market participants perceive that financial stability communication by central banks could be followed by monetary policy decisions, which should imply that market interest rates are reactive to such communications. As can be seen in panel (D) of Table 3, it is clear that there is no systematic reaction of short (three-month) or long (five to ten-year) interest rates. Thus, this is further evidence suggesting that there is very little role for a signalling channel and that it is rather the coordination channel that is at work – which is not too surprising, after all: most central banks in our sample have a clearly defined price stability objective and use their monetary policy tools to pursue this objective, such that there should be little scope for a signalling channel.
3.2.4. International spillovers Finally, we study spillover effects, i.e. whether the publication of an FSR in one country has effects on other countries. To implement this test, we took the country with the largest number of FSRs in our sample (namely the UK) and entered these communications into the model. Whenever a national communication has taken place 10 days before or after a given British FSR, we do not use these events. We find that the British FSRs tend to reduce volatility internationally but do not exert systematic effects on returns. The evidence points thus to some, albeit limited, international spillovers.
To summarise, the findings suggest that the effects of communication are not universal. Market conditions seem to matter, with different effects during the financial crisis. The origin of the communication also is important, with central banks in advanced economies exerting different effects from those in emerging economies. Finally, the evidence here further supports the conclusion that it is mainly a coordination channel that is at work – i.e. that communication provides relevant information about financial stability itself, rather than giving a signal about monetary policy, thereby affecting financial markets.
3.3. Effects of Speeches and Interviews
Having studied in great detail how financial markets react to central banks’ FSRs, we now turn to the effect of speeches and interviews by central bank governors. A first important difference between these two types of communications relates to the fact that FSRs typically have a predefined release schedule, whereas speeches and interviews are often much more flexible with regard to their timing. Figure 2 illustrates this flexibility by plotting the total number of speeches and interviews in all countries in a given quarter on the right-hand axis and the standard deviation of daily returns of the global financial stock index in each quarter on the left-hand
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 723
axis. The evolution of the two lines is extremely close, clearly suggesting that communication intensifies in times of financial market turbulence.11
In the light of these findings, one might ask whether speeches and their content are predictable, such that financial markets might have priced in the effects already prior to the communication event. In such a case, our event study methodology would not be appropriate. Speeches tend to be given in times when stock market volatility is high and when stock markets are declining. While this is true, the mere fact that stock market volatility is high and/or stock markets are declining is not sufficient to predict the occurrence of a speech. There are many days when these conditions hold, yet no speeches are given. This might partially be due to the fact that often the date of a speech and its topic are decided upon long before delivery. Accordingly, their occurrence might not be as easy to predict based on market conditions. We will try to address the potential endogeneity in the robustness tests.
Turning to the effects on financial markets, Figure 3 repeats the analysis of Figure 1, by comparing the actual and the predicted evolution of stock markets after speeches and interviews. The findings are remarkably different from those shown in Figure 1.
0
20
40
60
80
100
0
1
2
3
4
1995q1 2000q1 2005q1 2010q1
Volatility (Left-axis) Speeches & Interviews (Right-axis)
Fig. 2. Stock Market Volatility and the Occurrence of Speeches and Interviews Notes. The Figure shows the total number of speeches and interviews in all countries in a given quarter on the right-hand axis (solid line) and the standard deviation of daily returns of the global financial stock index in each quarter on the left-hand axis (dashed line).
11 Results of a more formal test (available upon request) show that on days before an event, the standard deviation of daily stock market returns is substantially higher than on non-event days, which is in contrast to results for FSRs. Furthermore, speeches and interviews intensify during periods of stock market declines: whereas the average stock return prior to non-event days is typically positive, it is on average negative prior to speeches and interviews. No such pattern is visible for FSRs.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
724 TH E E CONOM I C J O U RN A L [ J U N E
Speeches and interviews typically follow stock market declines and the model clearly predicts further declines subsequently (the dashed line in the Figure). As a matter of fact, actual returns do on average decline after a speech or an interview; however, comparing the expected with the actual evolution, it is also apparent that the stock markets decline by less than expected in the presence of central bank communications.
The results of the econometric tests are provided in Table 4, which is built in analogy to Table 2. The picture that emerges differs from the one for FSRs in several ways. With regard to the effects of speeches and interviews on returns, both the parametric and the non-parametric tests show results over longer horizons that are similar to those for FSRs – at the same time, for shorter horizons, we cannot detect any statistically significant effects. Furthermore, speeches do not lower stock market volatility – if anything, there is some tendency, especially of optimistic speeches, to somewhat increase it.
Of course, we have also subjected these results to the same sample splits and robustness tests as before. Again, given the large number of tests, we only show results for a time window of 25 business days in Table 5.
An interesting difference compared to FSRs arises when splitting the sample into the pre-crisis and the crisis period – whereas FSRs were found not to exert systematic effects on stock markets during the crisis, it is precisely then when speeches and interviews had their effects, underlining that speeches and interviews may be much more influential during periods of financial stress. Another difference is obtained for the robustness test using the risk-free rate as the prediction model. In contrast to the findings for FSRs, here, results weaken. This is not surprising, given that our prediction
–3.5
–30
–2.5
–2.0
–1.5
–1.0
–0.5
0
0.5
1 2 3 4 5 10 15 20 25 30 35 40 45 50 55 60
Predicted Actual
Fig. 3. Predicted Versus Actual Evolution of Stock Markets After Speeches and Interviews Notes. The Figure compares the actual evolution of cumulated stock market returns (in %) following speeches/interviews to the predicted evolution on the basis of the benchmark model (1). The solid line plots the average actual cumulated returns starting from day 1 after the communication event and up to day 60. The dashed line shows the expected cumulated returns that would result from the benchmark model in the absence of a communication event. The cumulated returns are multiplied by �1 for pessimistic communications, whereas they are left unchanged for optimistic communications.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 725
T ab
le 4
E ff ec ts of
Sp ee ch es
an d In te rv ie w s
N o o f d ay s
Jo in t m o d el
P es si m is ti c sp ee ch
es an
d in te rv ie w s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
1 0. 45
″ �0
.0 9
– –
0. 54
0. 12
– –
2 0. 48
�0 .1 0
– –
0. 57
** 0. 38
** –
– 3
0. 49
�0 .1 0
– –
0. 52
0. 28
– –
4 0. 51
0. 11
0. 47
′ 0. 02
0. 51
0. 07
0. 50
�0 .1 8*
5 0. 53
* 0. 26
0. 48
0. 01
0. 49
�0 .0 2
0. 53
�0 .1 9*
* 10
0. 55
** *
0. 55
** 0. 49
0. 00
0. 46
0. 05
0. 51
�0 .0 4
15 0. 54
* 0. 74
** 0. 48
0. 04
0. 47
0. 06
0. 49
0. 05
20 0. 52
0. 73
** 0. 49
0. 06
″ 0. 50
0. 17
0. 46
0. 06
25 0. 55
** 1. 04
** 0. 50
0. 06
″ 0. 47
0. 04
0. 48
0. 04
30 0. 54
** 1. 04
** 0. 51
0. 06
″ 0. 50
0. 63
0. 50
0. 04
35 0. 56
** *
1. 06
** 0. 50
0. 06
′ 0. 48
0. 70
0. 51
0. 04
40 0. 54
* 1. 01
* 0. 50
0. 05
′ 0. 51
0. 93
0. 52
0. 03
45 0. 52
0. 95
* 0. 50
0. 05
′ 0. 52
1. 13
0. 53
0. 01
50 0. 55
** *
1. 24
** 0. 51
0. 04
′ 0. 49
1. 06
0. 54
* 0. 00
55 0. 55
** 1. 58
** 0. 52
0. 04
′ 0. 50
0. 58
0. 53
0. 01
60 0. 55
** 1. 63
** 0. 50
0. 03
0. 50
0. 38
0. 51
0. 00
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
726 TH E E CONOM I C J O U RN A L [ J U N E
N o o f d ay s
N eu
tr al
sp ee
ch es
an d in te rv ie w s
O p ti m is ti c sp ee
ch es
an d in te rv ie w s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
1 0. 48
�0 .0 6
– –
0. 45
″ �0
.0 5
– –
2 0. 44
″ �0
.2 7
– –
0. 52
0. 17
– –
3 0. 46
�0 .5 9
– –
0. 50
0. 07
– –
4 0. 45
′ �0
.4 9
0. 46
0. 08
0. 53
0. 28
0. 46
0. 15
′ 5
0. 45
″ �0
.4 7
0. 45
′ 0. 10
0. 55
* 0. 48
* 0. 46
0. 11
10 0. 48
�0 .2 9
0. 49
0. 04
0. 57
** 1. 11
** *
0. 46
0. 01
15 0. 49
�0 .4 9
0. 50
0. 04
0. 54
1. 47
** *
0. 45
′ 0. 03
20 0. 45
′ �0
.7 0
0. 53
0. 05
0. 54
1. 54
** *
0. 47
0. 07
25 0. 45
′ �0
.4 2
0. 56
** 0. 02
0. 56
** 2. 02
** *
0. 48
0. 12
″ 30
0. 46
�0 .3 9
0. 53
0. 02
0. 57
** *
2. 55
** *
0. 49
0. 12
″ 35
0. 49
�0 .2 9
0. 52
0. 01
0. 60
** *
2. 67
** *
0. 49
0. 12
″ 40
0. 49
0. 07
0. 51
0. 01
0. 57
** *
2. 78
** *
0. 47
0. 11
″ 45
0. 50
0. 05
0. 51
0. 01
0. 56
** 2. 83
** *
0. 47
0. 12
″ 50
0. 50
0. 20
0. 51
0. 02
0. 59
** *
3. 33
** *
0. 48
0. 11
″ 55
0. 50
0. 06
0. 53
0. 01
0. 59
** *
3. 55
** *
0. 50
0. 10
″ 60
0. 49
0. 25
0. 52
0. 00
0. 60
** *
3. 45
** *
0. 48
0. 09
′
N ot es : Se
e n o te s to
T ab
le 2 b u t al l re su lt s re la te
to sp ee
ch es
an d in te rv ie w s ra th er
th an
F SR
s.
T ab
le 4
(C on ti n u ed )
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 727
T ab
le 5
E ff ec ts of
Sp ee ch es
an d In te rv ie w s – Sa m pl e Sp li ts an
d R ob u st n es s
Jo in t m o d el
P es si m is ti c sp ee
ch es
an d in te rv ie w s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
(A ) B en ch m ar k
0. 55
** 1. 04
** 0. 50
0. 06
″ 0. 47
0. 04
0. 48
0. 04
(B ) Sa m pl e sp li ts
(1 ) C o u n tr y gr o u p
A d va n ce d ec o n o m ie s
0. 54
* 1. 02
** 0. 48
0. 07
″ 0. 50
0. 56
0. 47
0. 04
E m er gi n g ec o n o m ie s
0. 57
* 1. 10
0. 56
** 0. 04
0. 36
″ �1
.7 4
0. 50
0. 03
(2 ) C ri si s ve rs u s p re -c ri si s
P re -c ri si s
0. 52
0. 48
0. 51
0. 02
0. 48
0. 47
0. 06
F in an
ci al
cr is is 20
07 -1 0
0. 59
** *
1. 87
** 0. 49
0. 11
″ 0. 45
0. 29
0. 48
0. 02
(3 ) Su
p er vi so ry
ro le
C B is su p er vi so r
0. 54
0. 75
0. 50
0. 06
0. 50
0. 61
0. 49
�0 .0 3
C B is n o t su p er vi so r
0. 56
** 1. 37
** 0. 51
0. 06
0. 44
′ �0
.4 9
0. 47
0. 10
′ (4 ) C lu st er in g
Sp ee
ch es
as p ar t o f cl u st er
0. 56
** 1. 41
** 0. 52
0. 02
0. 47
0. 18
0. 52
�0 .0 2
Sp ee
ch es
o u ts id e cl u st er
0. 53
0. 46
0. 47
0. 12
″ 0. 47
�0 .1 9
0. 41
′ 0. 15
′ (C
) R ob u st n es s
A ll st o ck s
0. 51
0. 87
** *
0. 48
0. 06
″ 0. 51
�0 .2 2
0. 48
0. 05
R is k- fr ee
ra te
as p re d ic to r
0. 50
�0 .4 3
0. 53
** 0. 00
0. 51
�0 .1 4
0. 51
0. 01
H ec km
an se le ct io n m o d el
0. 55
** 0. 97
** 0. 51
0. 06
′ 0. 47
0. 04
0. 48
0. 04
A lt er n at iv e co
d in g
0. 53
0. 75
* 0. 50
0. 06
″ 0. 48
0. 37
0. 47
0. 09
′ R aw
D ic ti o n sc o re s
– 0. 13
** –
– –
– –
– (D
) T es ti n g fo r th e si gn al li n g ch an
n el
Sh o rt -t er m
in te re st
ra te s
0. 49
�0 .0 7
0. 53
* �0
.0 3*
0. 44
″ �0
.0 1
0. 53
�0 .0 2
L o n g- te rm
in te re st
ra te s
0. 49
�0 .0 6
0. 50
0. 00
0. 48
�0 .0 2
0. 50
�0 .0 1*
* (E ) In te rn at io n al
sp il lo ve rs
U S sp ee
ch es
an d in te rv ie w s
0. 49
0. 09
0. 51
0. 00
0. 54
** *
0. 20
0. 50
0. 01
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
728 TH E E CONOM I C J O U RN A L [ J U N E
N eu
tr al
sp ee
ch es
an d in te rv ie w s
O p ti m is ti c sp ee
ch es
an d in te rv ie w s
R et u rn s
SD R et u rn s
SD
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
N o n -p ar am
et ri c
P ar am
et ri c
(A ) B en ch m ar k
0. 45
′ �0
.4 2
0. 56
** 0. 02
0. 56
** 2. 02
** *
0. 48
0. 12
″ (B ) Sa m pl e sp li ts
(1 ) C o u n tr y gr o u p
A d va n ce d ec o n o m ie s
0. 45
′ �0
.3 4
0. 52
0. 02
0. 58
** 2. 53
** *
0. 46
0. 15
″ E m er gi n g ec o n o m ie s
0. 47
�0 .6 3
0. 65
** *
0. 03
0. 52
0. 60
0. 52
0. 06
(2 ) C ri si s ve rs u s p re -c ri si s
P re -c ri si s
0. 48
�0 .8 8
0. 55
* 0. 01
0. 52
0. 84
0. 52
0. 01
F in an
ci al
cr is is 20
07 -1 0
0. 43
″ 0. 10
0. 56
* 0. 03
0. 62
** *
3. 58
** *
0. 43
′ 0. 27
″ (3 ) Su
p er vi so ry
ro le
C B is su p er vi so r
0. 48
0. 11
0. 50
0. 08
0. 56
* 1. 81
** 0. 50
0. 12
C B is n o t su p er vi so r
0. 43
′ �0
.9 7
0. 62
** *
�0 .0 4
0. 57
* 2. 30
** *
0. 45
0. 12
(4 ) C lu st er in g
Sp ee
ch es
as p ar t o f cl u st er
0. 47
�0 .0 8
0. 57
** 0. 00
0. 59
** 2. 92
** *
0. 48
0. 09
Sp ee
ch es
o u ts id e cl u st er
0. 43
′ �0
.9 6
0. 53
0. 05
0. 53
0. 69
0. 48
0. 16
′ (C ) R ob u st n es s
A ll st o ck s
0. 48
�0 .2 8
0. 53
0. 01
0. 52
1. 46
** *
0. 44
″ 0. 13
‴ R is k- fr ee
ra te
as p re d ic to r
0. 51
�0 .0 7
0. 60
** *
�0 .0 8
0. 50
�0 .9 3
0. 49
0. 08
H ec km
an se le ct io n m o d el
0. 45
′ �0
.4 2
0. 56
** 0. 02
0. 57
** 1. 95
** *
0. 49
0. 13
″ A lt er n at iv e co
d in g
0. 48
�0 .5 1
0. 53
0. 01
0. 53
1. 84
** *
0. 51
0. 08
R aw
D ic ti o n sc o re s
– –
– –
– –
– –
(D ) T es ti n g fo r th e si gn al li n g ch an
n el
Sh o rt -t er m
in te re st
ra te s
0. 41
‴ �0
.1 7
0. 54
* �0
.0 2
0. 42
‴ �0
.1 5
0. 52
�0 .0 4*
L o n g- te rm
in te re st
ra te s
0. 42
‴ �0
.0 4
0. 53
0. 00
0. 47
�0 .1 2
0. 48
0. 01
(E ) In te rn at io n al
sp il lo ve rs
U S sp ee
ch es
an d in te rv ie w s
0. 55
** *
0. 50
* 0. 53
** �0
.0 2
0. 53
* 0. 41
0. 49
0. 01
N ot es : Se
e n o te s to
T ab
le 3 b u t al l re su lt s re la te
to sp ee
ch es
an d in te rv ie w s ra th er
th an
F SR
s. T h e T ab
le al so
co n ta in s te st re su lt s fo r sp ee
ch es
an d in te rv ie w s th at
ar e
p ar t o f a cl u st er
o r n o t, an
d re su lt s fo r u n ex
p ec te d sp ee
ch es
b as ed
o n es ti m at io n o f a H ec km
an se le ct io n m o d el .
T ab
le 5
(C on ti n u ed )
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 729
model contained a momentum factor which was very important for speeches and interviews – they occur in times of declining stock markets and the prediction model predicts further declines. The choice of the prediction model is therefore important for speeches and interviews but less so for FSRs. The evidence regarding international spillovers reinforces our previous conclusion. In this estimation, we included the speeches and interviews by the chairman of the FOMC, as this provides us with the largest number of communications. We had seen before that FSRs had not generated large spillovers and we find here that speeches by the chairman of the FOMC generally do not affect international stock markets at all. This confirms that international spillovers tend to be rather limited.
Table 5 provides two additional results that we had not reported for FSRs. The first is an additional sample split and tests whether speeches and interviews exert different effects if they are clustered. We define a communication event to be part of a cluster if other speeches or interviews occur within 60 days after the event, or have occurred within 60 days before the event. As a matter of fact, these types exert very different effects. Speeches that are part of a cluster are influencing the market view, whereas isolated speeches do not and tend to increase market volatility.
The second additional test tries to tackle a possible endogeneity of speeches and interviews. We estimated a two-stage Heckman selection model, where the central bank in the first stage decides to deliver a speech or not and in the second stage, conditional on delivering a speech, decides whether the content is optimistic or pessimistic. We include the absolute value of the market trend, market volatility and the absolute value of our misalignment measure as predictors for the first stage and the market trend, market volatility and our misalignment measure as predictors for the second stage. Despite the fact that the model performs rather poorly, there are indeed a few speeches that get predicted by this model. Taking only the surprise component of our speeches, we can generate an alternative set of results, which is provided in Table 5. Results are remarkably robust to this variation. While this is due, in part, to the poor predictability of speeches, it also implies that the previous results were not driven by large effects of the predictable speeches.
To summarise, these findings suggest that speeches and interviews respond to economic conditions and are as such a rather flexible communication tool for central banks. Importantly, a sequence of speeches has a larger influence on financial markets than an isolated communication by the central bank governor. In contrast to FSRs, however, these communications affect markets only modestly in the short term and leave market volatility mainly unaffected; however, during the financial crisis they became more influential. An assessment of the effects of these tools therefore needs to clearly distinguish between the two.
4. Conclusions
This article has provided an empirical assessment of the effects of central bank communication about financial stability, a topic that has remained almost entirely unexplored in the literature to date. The article has studied the impact of central bank statements on financial markets, arguably one of the most important target groups of this type of communication. In more detail, it has constructed a unique dataset
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
730 TH E E CONOM I C J O U RN A L [ J U N E
covering over 1,000 communication events (a third of which being FSRs and two-thirds being speeches and interviews by central bank governors) by 37 central banks over a time period from 1996 to 2009, i.e. spanning nearly one and a half decades and analysed the reaction of financial sector stocks to these events.
The article’s findings suggest that communication about financial stability has important repercussions on financial sector stock prices. However, there are clear differences between FSRs on the one hand and speeches and interviews on the other. FSRs clearly create news in the sense that the views expressed in FSRs get reflected in stock market returns. These effects are furthermore long lasting. They also reduce market volatility. These effects are particularly strong if FSRs contain optimistic assessments of the risks to financial stability. Speeches and interviews, in contrast, move financial markets far less on average. In particular, while having only modest effects on stock market returns, they do not reduce market volatility. However, speeches and interviews were affecting market returns significantly more during the 2007–10 global financial crisis, indicating the potential importance of this communication tool during periods of financial stress. Importantly, our results generalise to overall stock market indices, i.e. are not confined to financial sector stocks and are robust to potential endogeneity of communications.
The mechanism by which the central bank affects financial markets seems to be related to the notion of a coordination channel, whereby communication by the central bank works as a coordination device, thereby reducing heterogeneity in expectations and information and thus inducing asset prices to more closely reflect the underlying fundamentals.
The article has also demonstrated how flexibly speeches and interviews can be used as a communication tool, with a higher frequency in times of heightened financial market volatility. In contrast to FSRs with their pre-defined release schedules, the mere occurrence of a speech or an interview can constitute news to financial markets in itself, a fundamental difference that might explain why the two communication channels have so different effects on market volatility. The findings of the article therefore underline that communication by monetary authorities on financial stability can influence financial market developments but that it needs to be employed with utmost care, stressing the difficulty of designing a successful communication strategy on financial stability.
Appendix A: Examples of Speeches and Interviews and Their Coding
5 March 1996: ‘Brazil Central Bk President Denies Bank Sector Instability’ ‘Central bank President Gustavo Loyola Tuesday denied rumors of instability in Brazil’s banking sector and said increasing bank investigations and encouragement for bank mergers have quelched any possibility of a crisis […]’. Source: Dow Jones International News Coded: Optimism = 1
27 October 1997: ‘China c. Banker Sees More Small Bank Bankruptcies’ ‘Some smaller Chinese banks and credit cooperatives could sink into bankruptcy due to bad loans, although a banking crisis was unlikely, central bank governor Dai Xianglong has said’. Source: Reuters News Coded: Optimism = �1
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 731
28 January 1998: ‘UK BOE’s George Confident Asia Contagion Can Be Avoided’ ‘Governor of the Bank of England Eddie George said Wednesday he was ‘reasonably confident’ wider financial contagion from the Asia crisis could be avoided’. Source: Dow Jones International News Coded: Optimism = 1
9 November 2000: ‘Korea Markets Unstable as Worries Linger – c. Bank’ ‘South Korea’s financial markets continue to show signs of instability as the second phase of financial restructuring progresses, the governor of the central Bank of Korea said on Thursday’. Source: Reuters News Coded: Optimism = �1
19 September 2002: ‘Mboweni Confident of Financial Stability’ ‘SA’s financial regulators are highly optimistic about the stability of the country’s financial system, Tito Mboweni, the SA Reserve Bank governor, said yesterday […]’. Source: All Africa Coded: Optimism = 1
10 April 2003: ‘Fukui Says Should Consider Preemptive Move on Banks’ ‘Bank of Japan Governor Toshihiko Fukui said on Thursday that Japan should consider ways to provide ailing banks with capital as a preemptive measure before any financial crisis occurred’. Source: Reuters News Coded: Optimism = 0
24 September 2003: ‘Argentina’s Central Bank Downplays Big Bank Restructuring’ ‘Plans to restructure the Argentine financial sector in the wake of last year’s financial crisis do not entail a widespread shakeup of the country’s banks, top Argentine Central Bank officials said Tuesday’. Source: Dow Jones International News Coded: Optimism = 0
17 March 2004: ‘Greenspan says U.S. Banking System Healthy’ ‘Federal Reserve Chairman Alan Greenspan said on Wednesday the US banking system weathered the 2001 recession well, and was in good shape to help finance the economic recovery’. Source: Reuters News Coded: Optimism = 1
11 September 2007: ‘CREDIT WRAPUP 5-Trichet Sure Major Banks Sound, Bernanke Silent’ ‘Europe’s banks are sound despite the confidence blow from a US subprime crisis, said the head of the European Central Bank on Tuesday, while the […]’. Source: Dow Jones International News Coded: Optimism = 1
5 February 2008: ‘ECB’s Noyer: Global Fincl System In Crisis For More Than A Year’ ‘The global financial system has been in a crisis situation for over a year and the crisis isn’t over, Bank of France Governor Christian Noyer said Tuesday’. Source: Dow Jones International News Coded: Optimism = �1
24 September 2008: ‘Swedish c. Bank Head Repeats Financial System Stable’ ‘Swedish Riksbank Governor Stefan Ingves said on Wednesday Sweden was now feeling the effects of the recent market turmoil more strongly, but repeated reassurances that the financial system was stable’. Source: Reuters News Coded: Optimism = 1
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
732 TH E E CONOM I C J O U RN A L [ J U N E
3 October 2008: ‘Bernanke: Fed To Do All It Can To Combat Crisis’ ‘Federal Reserve Chairman Ben Bernanke said on Friday the U.S. central bank will do whatever it can to combat the credit crisis and help the economy’. Source: Reuters News Coded: Optimism = 0
6 October 2008: ‘Turkish Banks Face Narrower Credit Channels – c. Bank’ ‘Central Bank Governor Durmus Yilmaz said on Monday Turkish banks were facing narrower credit channels due to the global credit crisis, but said they faced no difficulty in renewing external loans’. Source: Reuters News Coded: Optimism = 0
University of Mannheim European Central Bank Deutsches Institut f€ur Wirtschaftsforschung
Submitted: 30 August 2011 Accepted: 7 November 2012
References Allen, F., Francke, L. and Swinburne, M.W. (2004). ‘Assessment of the Riksbank’s work on financial stability
issues’, Sveriges Riksbank Economic Review, vol. 2004(3), pp. 5–26. Andersen, T., Bollerslev, T., Diebold, F. and Vega, C. (2003). ‘Micro effects of macro announcements: real-
time price discovery in foreign exchange’, American Economic Review, vol. 39(1), pp. 38–62. Armesto, M., Hernandez-Murillo, R., Owyang, M. and Piger, J. (2009). ‘Measuring the information content of
the Beige Book: a mixed data sampling approach’, Journal of Money, Credit and Banking, vol. 41(1), pp. 35– 55.
Bernanke, B. and Kuttner, K. (2005). ‘What explains the stock market’s reaction to Federal Reserve policy’, Journal of Finance, vol. 60(3), pp. 1221–57.
Bligh, M. and Hess, G.D. (2007). ‘The power of leading subtly: Alan Greenspan, rhetorical leadership, and monetary policy’, Leadership Quarterly, vol. 18(2), pp. 87–104.
Blinder, A., Ehrmann, M., Fratzscher, M., de Haan, J. and Jansen, D.-J. (2008). ‘Central bank communication and monetary policy: a survey of theory and evidence’, Journal of Economic Literature, vol. 46(4), pp. 910– 45.
Born, B., Ehrmann, M. and Fratzscher, M. (2011). ‘How should central banks deal with a financial stability objective? The evolving role of communication as a policy instrument’, in (S. Eijffinger and D. Masciandaro, eds.), Handbook of Central Banking, Financial Regulation and Supervision after the Financial Crisis, pp. 245–68, Cheltenham: Edward Elgar.
Born, B., Ehrmann, M. and Fratzscher, M. (2012). ‘Communicating about macroprudential supervision – a new challenge for central banks’, International Finance, vol. 15(2), pp. 179–203.
Brunnermeier, M., Crockett, A., Goodhart, C., Persaud, A. and Shin, H.S. (2009). The Fundamental Principles of Financial Regulation. London: Centre for Economic Policy Research.
Carhart, M. (1997). ‘On persistence in mutual fund performance’, Journal of Finance, vol. 52(1), pp. 57–82. Cihak, M. (2006). How do central banks write on financial stability?, Working Paper, International Monetary
Fund. Cihak, M. (2007). ‘Central banks and financial stability: a survey’, http://ssrn.com/abstract=998335 (last
accessed: 4 February 2013). Cukierman, A. (2009). ‘The limits of transparency’, Economic Notes, vol. 38(1–2), pp. 1–37. Davis, A.K., Piger, J. and Sedor, L.M. (2012). ‘Beyond the numbers: measuring the information content of
earnings press release language’, Contemporary Accounting Research, vol. 29(3), pp. 845–68. Dincer, N. and Eichengreen, B. (2010). ‘Central bank transparency: causes, consequences and updates’,
Theoretical Inquiries in Law, vol. 11(1), pp. 75–123. Edmans, A., Garcia, D. and Norli, O. (2007). ‘Sports sentiment and stock returns’, Journal of Finance, vol. 62
(4), pp. 1967–98. Ehrmann, M. and Fratzscher, M. (2007). ‘The timing of central bank communication’, European Journal of
Political Economy, vol. 23(1), pp. 124–45.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
2014] F I N A N C I A L S T A B I L I T Y C OMMUN I C A T I O N 733
Eijffinger, S. and Geraats, P. (2006). ‘How transparent are central banks?’, European Journal of Political Economy, vol. 22(1), pp. 1–21.
Entorf, H., Gross, A. and Steiner, C. (2012). ‘Business cycle forecasts and their implications for high frequency stock market returns’, Journal of Forecasting, vol. 31(1), pp. 1–14.
European Central Bank (2010). Financial Stability Review, December 2010, Frankfurt. Fama, E.F. and French, K.R. (1993). ‘Common risk factors in the returns on stocks and bonds’, Journal of
Financial Economics, vol. 33(1), pp. 3–56. Fratzscher, M. (2008). ‘Oral interventions versus actual interventions in FX markets – an event-study
approach’, ECONOMIC JOURNAL, vol. 118(530), pp. 1079–106. Fratzscher, M. (2009). ‘How successful is the G7 in managing exchange rates?’, Journal of International
Economics, vol. 79(1), pp. 78–88. Hart, R.P. (2000). Campaign Talk: Why Elections Are Good For Us, Princeton, NJ: Princeton University Press. Hart, R.P. and Jarvis, S. (1997). ‘Political debate: forms, styles, and media’, American Behavioral Scientist,
vol. 40(8), pp. 1095–1122. Hirshleifer, D. (2001). ‘Investor psychology and asset pricing’, Journal of Finance, vol. 56(4), pp. 1533–97. Kaminsky, G. and Lewis, K. (1996). ‘Does foreign exchange intervention signal future monetary policy?’,
Journal of Monetary Economics, vol. 37(2), pp. 285–312. Kothari, S.P. and Warner, J.B. (2007). ‘Econometrics of event studies’, in (B.E. Eckbo, ed.), Handbook of
Corporate Finance: Empirical Corporate Finance, vol. 1, pp. 3–36, Amsterdam: Elsevier. Kuttner, K. and Posen, A. (2010). ‘Do markets care who chairs the central bank?’, Journal of Money, Credit and
Banking, vol. 42(2–3), pp. 347–71. Lord, C.G., Ross, L. and Lepper, M.R. (1979). ‘Biased assimilation and attitude polarization: the effects of
prior theories on subsequently considered evidence’, Journal of Personality and Social Psychology, vol. 37 (11), pp. 2098–2109.
Lucca, D.O. and Trebbi, F. (2009). ‘Measuring central bank communication: an automated approach with application to FOMC statements’, Working Paper, National Bureau of Economic Research.
MacKinlay, A.C. (1997). ‘Event studies in economics and finance’, Journal of Economic Literature, vol. 35(1), pp. 13–39.
Masciandaro, D. and Quintyn, M. (2009). ‘After the Big Bang and before the next one? Reforming the financial supervision architecture and the role of the central bank. A review of worldwide trends, causes and effects (1998–2008)’, Working Paper, Paolo Baffi Centre.
Mishkin, F.S. (2004). ‘Can central bank transparency go too far?’, in (C. Kent and S. Guttmann, eds.), The Future of Inflation Targeting, pp. 48–65, Sydney: Reserve Bank of Australia.
Morris, S. and Shin, H.S. (2002). ‘Social value of public information’, American Economic Review, vol. 92(5), pp. 1521–34.
Moser, C. and Dreher, A. (2010). ‘Do markets care about central bank governor changes? Evidence from emerging markets’, Journal of Money, Credit and Banking, vol. 42(8), pp. 1589–612.
Oosterloo, S. and de Haan, J. (2004). ‘Central banks and financial stability: a survey’, Journal of Financial Stability, vol. 1(2), pp. 257–73.
Oosterloo, S., de Haan, J. and Jong-A-Pin, R. (2007). ‘Financial stability reviews: a first empirical analysis’, Journal of Financial Stability, vol. 2(4), pp. 337–55.
Pojarliev, M. and Levich, R.M. (2008). ‘Do professional currency managers beat the benchmark?’, Financial Analysts Journal, vol. 64(5), pp. 18–32.
Rigobon, R. and Sack, B. (2004). ‘The impact of monetary policy on asset prices’, Journal of Monetary Economics, vol. 51(8), pp. 1553–75.
Sarno, L. and Taylor, M. (2001). ‘Official intervention in the foreign exchange market: is it effective, and if so, how does it work?’, Journal of Economic Literature, vol. 39(3), pp. 839–68.
Svensson, L.E.O. (2003). ‘Monetary policy and real stabilization’, Working Paper, National Bureau of Economic Research.
Svensson, L.E.O. (2006). ‘Social value of public information: Morris and Shin (2002) is actually pro transparency, not con’, American Economic Review, vol. 96(1), pp. 448–51.
© 2013 The Author(s). The Economic Journal © 2013 Royal Economic Society.
734 TH E E CONOM I C J O U RN A L [ J U N E 2014]
Competition and Financial Stability Different Author 2004.pdf
Competition and Financial Stability Author(s): Franklin Allen and Douglas Gale Source: Journal of Money, Credit and Banking, Vol. 36, No. 3, Part 2: Bank Concentration and Competition: An Evolution in the Making A Conference Sponsored by the Federal Reserve Bank of Cleveland May 21-23, 2003 (Jun., 2004), pp. 453-480 Published by: Ohio State University Press Stable URL: http://www.jstor.org/stable/3838946 .
Accessed: 06/02/2015 10:50
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Ohio State University Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of Money, Credit and Banking.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN
DOUGLAS GALE
Competition and Financial Stability
Competition policy in the banking sector is complicated by the necessity of maintaining financial stability. Greater competition may be good for (static) efficiency, but bad for financial stability. From the point of view of welfare economics, the relevant question is: what are the efficient levels of competi- tion and financial stability? We use a variety of models to address this question and find that different models provide different answers. The relationship between competition and stability is complex: sometimes com- petition increases stability. In addition, in a second-best world, concentration may be socially preferable to perfect competition and perfect stability may be socially undesirable.
JEL codes: D4, D5, D6, G2 Keywords: crises, banking concentration, dynamic, spatial,
and Schumpeterian competition.
IN THE BANKING SECTOR, unlike other sectors of the
economy, competition policy must take account of the interaction between competi- tion and financial stability. Greater competition may be good for (static) efficiency, but bad for financial stability.1 In this paper, we shall argue that the relationship between competition and financial stability is considerably more complex than this
simple "trade-off" suggests, but understanding why increasing competition might reduce economic stability is a good starting point.
In an important paper, Keeley (1990) provided a theoretical framework and
empirical evidence that deregulation of the banking sector in the U.S. in the
1. See Canoy et al. (2001) and Carletti and Hartmann (2003) for excellent surveys of the literature on financial stability and competition.
Prepared for the World Bank and Federal Reserve Bank of Cleveland project on Bank Concentration. Presented at the April 3-4, 2003 conference at the World Bank and the May 21-23, 2003 conference at the Federal Reserve Bank of Cleveland. We are grateful to an anonymous referee, to our discussants Stephen Haber and Charles Kahn, and to Elena Carletti, Qian Liu, Lemma Senbet, Andrew Winton, and participants in the conferences.
FRANKLIN ALLEN is a professor offinance in the Department of Finance, Wharton School, University of Pennsylvania. E-mail: [email protected] DOUGLAS GALE is a professor in the Department of Economics, New York University. E-mail: [email protected] Received May 29, 2003; and accepted in revised form May 29, 2003.
Journal of Money, Credit, and Banking, Vol. 36, No. 3 (June 2004, Part 2) Published in 2004 by The Ohio State University Press.
I
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
454 : MONEY, CREDIT, AND BANKING
1970s and 1980s had increased competition and led to a reduction in monopoly rents. This reduction in "charter value" magnified the agency problem between bank owners and the government deposit insurance fund. The bank owners or managers acting on their behalf had an increased incentive to take on extra risk, given the guaranteed funds available to them because of deposit insurance. As in the agency problem identified by Jensen and Meckling (1976), if the gamble was successful the equity owners would obtain the rewards while if it was unsuccessful the cost would be born by the deposit insurance fund. The extra risk that banks took on as a result of this agency problem caused a dramatic increase in bank failures during the 1980s. The US is not the only country where there appears to be an empirical relationship between increased competition and financial instability. Beck, Demir- guc-Kunt, and Levine (2003) find using data from 79 countries that crises are less likely in more concentrated banking systems.
Various empirical studies have found that the cost of financial instability is high. For example, Hoggarth and Saporta (2001) find that the average fiscal costs of banking resolution across countries are 16% of GDP. For emerging countries the figure is 17.5% and for developed countries it is 12%. As Table 1 shows, the costs of banking crises alone are estimated at 4.5% of GDP. Although these costs are substantial, they are much lower than the costs, estimated at 23% of GDP, of banking and currency crises occurring together. A proportion of the fiscal costs are transferred, so these figures do not represent the deadweight economic costs. A number of studies measure the cumulative output loss resulting from a financial crisis by using the deviation from trend output. Table 2 gives estimates for these costs. The average cumulative output loss for all crises is 16.9% of GDP. Here the costs of twin crises are again higher; the loss caused by twin banking and currency crises is 29.9% of GDP versus 5.6% for banking crises alone. However, in contrast to fiscal costs,
TABLE 1
AVERAGE CUMULATIVE FISCAL COSTS OF BANKING CRISES IN 24 CRISES, 1977-2000
Non-performing loans Fiscal costs of banking resolution Number of crises (percentage of total loans) (percentage of GDP)
All countries 24 22 16 Emerging market countries 17 28 17.5 Developed countries 7 13.5 12 Banking crisis alone 9 18 4.5 Banking and currency crises of which 15 26 23 Emerging market countries 11 30 25 Developed countries 4 18 16 Banking and currency crises with 11 26 27.5
previous fixed exchange rate of which
Emerging market countries 8 30 32 Developed countries 3 18 16
NOTE: Source: Hoggarth and Saporta (2001, p. 150).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 455
TABLE 2
OUTPUT LOSSES ASSOCIATED WITH BANKING CRISES, 1977-98
Average crisis length Average cumulative output losses Number of crises (years) (percentage of GDP)
All 43 3.7 16.9 Single banking crises 23 3.3 5.6 Twin banking and currency crises 20 4.2 29.9 Developed countries 13 4.6 23.8 Emerging market countries 30 3.3 13.9
NOTE: Source: Hoggarth and Saporta (2001, p. 155).
developed countries have a greater loss, 23.8% of GDP, than emerging countries, 13.9% of GDP.
The large literature on the efficiency of the banking industry (for a survey see, e.g., Berger and Humphrey, 1997) is mostly concerned with the cost- and profit- efficiency of retail banking. For example, Canoy et al. (2001) summarize the evidence as suggesting that the average bank operates at a cost level that is 10% or 20% above the best-practices level. This is just one (probably small) part of the total costs of deviations from perfect competition. Unfortunately, the total costs of a deviation from perfect competition have not been documented as carefully as the costs of financial instability.
Given the large and visible costs of financial instability, it is natural for policymak- ers to make the avoidance of financial crises a high priority. By contrast, the difficulty of measuring the efficiency costs of concentration may suggest that competition policy warrants a lower priority. In fact, the uncertainty about the costs of concentra- tion together with the perceived (negative) trade-off between competition and finan- cial stability may actually encourage policymakers to favor concentration at the expense of competition policy. This subordination of competition policy to financial stability may be unwise for a number of reasons, however. In the first place, the extent to which there is a negative trade-off between competition and financial stability may be questioned. The costs of financial crises are undoubtedly high, but it does not follow that it is necessary to reduce competition to avoid those costs. Secondly, the wide range of estimates of the efficiency costs from concentration is at least consistent with a high efficiency gain from greater competition. Thirdly, the costs of financial crises occur infrequently, perhaps every decade or few decades, whereas the inefficiency cost concentrations are born continuously.
The proper balance between competition and financial stability presupposes a framework in which we can identify the welfare costs and benefits of different levels of competition and financial stability. Our objective in this paper is to review a number of theoretical models as a prelude to the development of a theoretical framework in which the optimal policy can be identified. From the point of view of welfare economics, the relevant question is: What are the efficient levels of competition and financial stability? We use a variety of models to address this
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
456 : MONEY, CREDIT, AND BANKING
question and find that different models provide different answers. This should not be surprising. In a second-best world, concentration may be preferred to perfect competition (Schumpeter 1950) and perfect stability may be socially undesirable (Allen and Gale 1998). When we consider the relationship between competition and stability, we find that the idea of a simple negative trade-off is, again, too simple: sometimes competition decreases stability and sometimes perfect competition is compatible with the socially optimal level of stability.
We begin in Section 1 by describing the general equilibrium model of financial intermediaries and markets from Allen and Gale (2003a). They provide analogues of the classical theorems of welfare economics for a model of intermediation with asymmetric information. If financial markets are complete and contracts between intermediaries and their customers are complete, the perfectly competitive equilib- rium allocation is incentive-efficient. In this sense, perfect competition is socially optimal. There is no financial instability because contracts are completely contingent and hence there is no need to default. Similarly, if contracts are incomplete, the perfectly competitive equilibrium allocation is constrained-efficient, but now finan- cial instability is necessary for efficiency. If the banks cannot meet the fixed payments that they have promised, there is a financial crisis. A deviation from competition may increase financial stability, but cannot increase and is likely to reduce welfare. This result illustrates that, in general, there should be no presumption that reducing competition in order to increase financial stability is socially desirable.
In simple partial-equilibrium models, it is possible to generate a negative trade-off between competition and financial stability. However, even in this case, the nature of the trade-off between competition and stability is more complicated than was first thought. For example, Allen and Gale (2000a, chap. 8), Boyd and De Nicolo (2002), and Perotti and Suarez (2003) have identified a number of different effects of increased competition on financial stability. In some circumstances, increased competition can actually increase financial stability. These models are discussed in Section 2.
Introducing other kinds of frictions produces further complications to our picture of competition. On the one hand, as mentioned above, Allen and Gale (2000a, chap. 8) show that when search costs are introduced, competition among a large number of unitary banks may result in the monopoly price being charged. On the other hand, a system with two banks with branches at each location may result in the perfectly competitive price. A Hotelling-type model of spatial competition introduces a rich variety of effects concerning regional diversification and risk sharing. The profitability of banks is shown to be extremely sensitive to the precise form of local interactions.
Section 4 describes a model of Schumpeterian competition, in which firms compete by developing new products. The firm that makes the best innovation manages to capture the whole market. The equilibrium price equals the difference between the value of the successful firm's product and the value of the second-best product. So the successful firm's profit is equal to the social value of its innovation. This provides the firms with the right incentives to innovate efficiently. In this context, perfect competition is again desirable and can lead to efficiency. Clearly, this kind
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 457
of innovation process is not consistent with financial stability. The successful innova- tor will survive while the unsuccessful will fail. Again, as in the benchmark model, efficiency requires a combination of perfect competition and financial instability. If the government is concerned with financial stability it may ensure that banks survive by regulating entry in different submarkets. We consider a setting where banks are assured of a monopoly in their region and consider the incentives to innovate. It is shown that this enforced stability leads to a welfare loss as might be expected. Less obvious is the result that there may be too little or too much investment in innovation. There can be too little because each bank only obtains profits from its own region. There can be too much because the bank is assured of a return no matter what happens.
Contagion is another important source of financial instability. It occurs when some shock, possibly small, spreads throughout the financial system and causes a systemic problem. Allen and Gale (2000b) developed a model of contagion with a perfectly competitive banking sector. It was shown that a shock that was arbitrarily small relative to the economy as a whole could cause all the banks in the financial system to go bankrupt. The contagion spreads through the interbank market. Section 5 extends the Allen and Gale (2000b) model of contagion to allow for imperfect competition in the banking sector. It is shown that in this case the economy is not as susceptible to contagion as it is with perfect competition. Each oligopolistic bank realizes that its actions affect the price of liquidity. By providing sufficient liquidity to the market they can ensure that contagion and their own bankruptcy are avoided. In this case there is a trade-off between competition and stability.
Concluding remarks are contained in Section 6.
1. COMPETITION AND CRISES
In the Arrow-Debreu model of general equilibrium, the fundamental theorems of welfare economics show that perfect competition is a necessary condition for efficiency. Allen and Gale (2003a) show that analogous results hold for a model of financial crises with complete markets. In this setting, perfect competition is compati- ble with the efficient level of financial stability. In this sense, there is no "trade- off" between competition and stability. We begin by describing the model of perfect competition in a financial system consisting of financial intermediaries and markets and then summarize the theoretical results in Allen and Gale (2003a).
There are three dates t = 0,1,2 and a single good at each date. The good is used for consumption and investment.
The economy is subject to two kinds of uncertainty. First, individual agents are subject to idiosyncratic preference shocks, which affect their demand for liquidity (these will be described later). Second, the entire economy is subject to aggregate shocks that affect asset returns and the cross-sectional distribution of preferences. The aggregate shocks are represented by a finite number of states of nature. All agents have a common prior probability density over the states of nature. All
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
458 : MONEY, CREDIT, AND BANKING
uncertainty is resolved at the beginning of date 1, when the aggregate state is revealed and each agent discovers his/her individual preference shock.
Each agent has an endowment of one unit of the good at date 0 and no endowment at dates 1 and 2. So, in order to provide consumption at dates 1 and 2, they need to invest.
There are two assets distinguished by their returns and liquidity structure. One is a short-term asset (the short asset), and the other is a long-term asset (the long asset). The short asset is represented by a storage technology: one unit invested in the short asset at date t = 0,1 yields a return of one unit at date t + 1. The long asset yields a return after two periods. One unit of the good invested in the long asset at date 0 yields a random return of more than one unit of the good that depends on the aggregate state at date 2.
Investors' preferences are distinguished ex ante and ex post. At date 0 there is a finite number n of types of investors, indexed by i = 1,...,n. We call i an investor's ex ante type. An investor's ex ante type is common knowledge and hence contractible.
While investors of a given ex ante type are identical at date 0, they receive a
private, idiosyncratic, preference shock at the beginning of date 1. The date 1
preference shock is denoted by Oi C Oi, where 0i is a finite set. We call 0i the investor's ex post type. Because 0i is private information, contracts cannot be explicitly contingent on Oi.
Investors only value consumption at dates 1 and 2. An investor's preferences are
represented by a von Neumann-Morgenster utility function, ui(cl, c2; 0i), where ct denotes consumption at date t = 1,2. The utility function ui(.; Oi) is assumed to be concave, increasing, and continuous for every type Oi. Diamond and Dybvig (1983) assumed that consumers were one of two ex post types, either early diers who valued consumption at date 1 or late diers who valued consumption at date 2. This is a special case of the preference shock 0i. The present framework allows for much more general preference uncertainty.
Allen and Gale (2003a) consider two different versions of the model, depending on the kind of contracts financial institutions offer to their customers. In the first version, contracts are completely contingent, subject only to incentive-compatibility constraints. More precisely, contracts are required to be incentive-compatible and are allowed to be contingent on the aggregate states qr and individuals' reports of their ex post types. Each intermediary offers a single contract and each ex ante type is attracted to a different intermediary.
One can, of course, imagine a world in which a single "universal" intermediary offers contracts to all ex ante type of investors. A universal intermediary could act as a central planner and implement the incentive-efficient allocation of risk. There would be no reason to resort to markets at all. Our world view is based on the assumption that transaction costs preclude this kind of centralized solution and that decentralized intermediaries are restricted in the number of different contracts they can offer. This assumption provides a role for financial markets in which financial intermediaries can share risk and obtain liquidity.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 459
At the same time, financial markets alone will not suffice to achieve optimal risk sharing. Because individual economic agents have private information, markets for individual risks are incomplete. The markets that are available will not achieve an incentive-efficient allocation of risk. Intermediaries, by contrast, can offer individuals incentive-compatible contracts and improve on the risk sharing provided by the market.
In the Diamond and Dybvig (1983) model, all investors are ex ante identical. Consequently, a single representative bank can provide complete risk sharing and there is no need for markets to provide cross-sectional risk sharing across banks. Allen and Gale (1994) showed that differences in risk and liquidity preferences can be crucial in explaining asset prices. This is another reason for allowing for ex ante heterogeneity.
In the context of intermediaries with complete markets and complete contingent incentive-compatible contracts, Allen and Gale (2003a) prove the following result.
PROPOSITION 1: Under the maintained assumptions, the equilibrium allocation of the model with complete markets and incomplete contracts is constrained-efficient.
Proposition 1 assumes that intermediaries use complete, incentive-compatible contracts. In reality, we do not observe such complex contracts, for reasons that are well documented in the literature, including transaction costs, asymmetric infor- mation, and the nature of the legal system. These frictions can justify the use of debt and many other kinds of incomplete contracts that intermediaries use in practice. The second version of the model presented by Allen and Gale (2003a) assumes that intermediaries are restricted to using a set of incompletely contingent contracts. This framework allows for many special cases, including at one extreme the earlier model with completely contingent contracts and completely non-contingent debt contracts. Note that this framework allows for a wide variety of assumptions about what is feasible, but takes the set of feasible contracts as given. To endogenize the set of feasible contracts one would have to appeal to factors such as transaction costs, non-verifiable information, and so on.
When contracts are complete, there is no incentive for intermediaries to enter into commitments that they cannot carry out. When contracts are constrained to be incomplete, it may be (ex ante) optimal for the intermediary to plan to default in some states. In the event of default, it is assumed that the intermediary's assets, including the Arrow securities it holds, are liquidated and the proceeds distrib- uted among the intermediary's investors. For markets to be complete, which is an assumption we maintain here, the Arrow securities that the bank issues must be default free. Hence, we assume that these securities are collateralized and their holders have priority. Anything that is left after the Arrow security holders have been paid off is paid out pro rata to the depositors. Allen and Gale (2003a) demonstrate the following result for the case when contracts are incomplete and take the form of deposit contracts.
PROPOSITION 2: Under maintained assumptions, the equilibrium allocation of the model with complete markets and incomplete contracts is constrained-efficient.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
460 : MONEY, CREDIT, AND BANKING
This is an important result. It shows that, in the presence of complete markets and perfect competition, the incidence of default is optimal in a laisser-faire equilib- rium. There is no scope for welfare-improving government intervention to prevent financial crises. In fact competition and financial instability are both necessary for constrained efficiency.
This result demonstrates that in a standard framework achieving optimality does not require trading off competition and financial stability. As we will see below this result extends to a number of other circumstances.
It is important to stress that the results in this section are simply benchmarks to illustrate what may happen. The only costs modeled are the losses to consumers from inefficient risk sharing. Many features that may be important in practice, such as unemployment and bankruptcy costs to firms are excluded. When these are taken into account, there may be a role for government intervention to reduce the incidence of financial crises. What the results do show is that the operation of the financial system and the occurrence of crises when there are complete markets are not the problem. There must be some form of market failure for financial crises to be undesirable.
2. AGENCY COSTS
Keeley (1990) developed a simple model of risk taking by banks with two dates and two states when there is deposit insurance. He showed that as competition increased risk taking by banks also increased. In fact, deposit insurance is not necessary for this effect to be present although it does exacerbate it. Allen and Gale (2000a, chap. 8) developed a simple model of competition and risk taking to illustrate the agency problem.
When firms are debt-financed, managers acting in the shareholders' interests have an incentive to take excessive risks, because the debtholders bear the downside risk while the shareholders benefit from the upside potential. This well known problem of risk shifting is particularly acute in the banking sector where a large proportion of the liabilities are in the form of debt (deposits). The risk-shifting problem is exacerbated by competition. Other things being equal, greater competition reduces the profits or quasi-rents available to managers and/or shareholders. As a result, the gains from taking excessive risks become relatively more attractive and this increases the incentive to exploit the non-convexity in the payoff function. Any analysis of the costs and benefits of competition has to weigh this effect against the supposed efficiency gains of greater competition.
To illustrate these ideas, consider the problem faced by a banking regulator who controls entry into the banking industry by granting charters to a limited number of banks. We use a model of Cournot competition, in which banks choose the volume of deposits they want, subject to an upward sloping supply of funds schedule. Having more banks will tend to raise the equilibrium deposit rate and increase the tendency to shift risks. What is the optimal number of charters?
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 461
Should the regulator restrict competition by granting only a few charters or encourage competition by granting many?
2.1 A Static Model
Suppose that the regulator has chartered n banks, indexed i = l,...,n. Each bank chooses a portfolio consisting of perfectly correlated risks. This as-
sumption is equivalent to assuming that the risk of each investment can be decom- posed into a common component and a purely idiosyncratic component. If there is a very large number of investments, the purely idiosyncratic components can be pooled perfectly. Then the idiosyncratic risks disappear from the analysis and we are left with a common component representing the systematic risks.
A portfolio is characterized by its size and rate of return. The bank's investments have a two-point return structure: for each dollar invested, bank i will receive a return Y, with probability p(yi); with probability (1 - p(yi)) they pay a return 0. The bank chooses the riskiness of its portfolio by choosing the target return Yi on its investments. The function P(Yi) is assumed to be twice continuously differentiable and satisfies
p(O) = 1, p(y) = 0, and p'(Yi) < 0, p"(Yi) < 0, VO < yi < .
The higher the target return, the lower the probability of success and the more rapidly the probability of success falls. Because the investments have perfectly correlated returns, the portfolio return has the same distribution as the returns to the individual investments.
Let di > 0 denote the total deposits of bank i, which is by definition the total number of dollars the bank has to invest. (For the moment, we ignore bank capital.) There is an upward sloping supply-of-funds curve. If the total demand for deposits is D = ,idi, then the opportunity cost of funds is R(D), where R(D) is assumed to be a differentiable function satisfying
R'(D) > 0, R"(D) > 0, R(0) = 0 and R(oo) = oo.
We assume that all deposits are insured, so the supply of funds is independent of the riskiness of the banks' portfolios. In the sequel, we consider the case where banks bear the cost of deposit insurance.
The payoff to bank i is a function of the riskiness of its own portfolio and the demand for deposits of all the banks
rri(y, d) = p(yi)[ydi - R(D)di] ,
where d = (dl,...,dn) and y = (l,... ,yn). Note that we have ignored the cost of deposit insurance to the bank in calculating its net return.
Since a bank can always ensure non-negative profits by choosing di = 0, it will always earn a non-negative expected return in equilibrium, that is, yidi - R(D)di > 0. There is no need to introduce a separate limited-liability constraint.
In a Nash-Courot equilibrium, each bank i chooses an ordered pair (Yi, di) that is a best response to the strategies of all the other banks. Consider an equilibrium
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
462 MONEY, CREDIT, AND BANKING
(y, d) in which each bank i chooses a strictly positive pair (Yi, di) >> O. As a necessary condition for a best response, this pair must satisfy the following first-order conditions
p(Yi)[i - R(D) - R'(D)di] = O,
'(yi)[yi - R(D)]di + p(yi)di = 0.
Assuming that the equilibrium is symmetric, that is, (yi, di) = (y, d) for every i, the first-order conditions reduce to
y - R(nd) - R'(nd)d = 0,
p'(y)[y - R(nd)] + p(y) = 0,
and this implies that
_P - y - R(nd) = R'(nd)d. p'(y)
Given our assumptions on p(y), an increase in y reduces - p(y)/p'(y). Suppose that there are two symmetric equilibria, (y, d) and (y', d'). Then y > y' implies that R'(nd)d < R'(nd')d', which, given our assumptions on R(D), implies that d < d' and R(nd) < R(nd'). Then clearly y - R(nd) > y' - R(nd'), contradicting the first equa- tion. So there is at most one solution to this set of equations, which determines both the size and the riskiness of the banks' portfolio in a symmetric equilibrium.
PROPOSITION 3: Under the maintained assumptions, there is at most one symmetric equilibrium (y*, d*)>> 0 which is completely characterized by the conditions
p(y) -P,) = y - R(nd) = R'(nd)d .
What can we now say about the effect of competition on risk taking? Suppose that we identify the degree of competitiveness of the banking sector
with the degree of concentration. In other words, the larger the number of banks, the more competitive the banking sector is. So a first attempt at answering the question would involve increasing n ceteris paribus and observing how the riskiness of the banks' behavior changes.
With a fixed supply-of-funds schedule R(-) it is most likely that the volume of deposits will remain bounded as n increases. More precisely, if we assume that R(D) -- oo as D -- oo then it is clear that D -- oo is inconsistent with equilibrium. Then the equilibrium value of D is bounded above (uniformly in n) and this implies that d Din -> 0 as n -- oo. This in turn implies that R'(nd)d - 0 from which it
immediately follows that y - R(nd) - 0 and p(y) - 0, or in other words, that y and R(nd) both converge to y.
PROPOSITION 4: If R(D) -- oo as D - oo then in any symmetric equilibrium, y - R(nd) -> 0 and y -> y as n -> oo.
The effect of increasing competition is to make each bank much smaller relative to the market for funds and this in turn reduces the importance of the price effect
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 463
(the R'(nd)d term) in the bank's decision. As a result, banks behave more like perfect competitors and will increase their business as long as profits are positive. Equilibrium then requires that profits converge to zero, and this in turn implies that banks have extreme incentives for risk taking. In the limit as n -> oo, they will choose the riskiest investments possible in an attempt to earn a positive profit.
The effect of replicating the market. This exercise is enlightening but somewhat artificial since it assumes that we are dealing with a market of fixed size and increasing the number of banks without bound in order to achieve competition. Normally, one thinks of perfect competition as arising in the limit as the number of banks and consumers grows without bound. One way to do this is to replicate the market by shifting the supply-of-funds function as we increase the number of banks. Precisely, suppose that the rate of return on deposits is a function of the deposits per bank
R = R(D/n) .
In effect, we are assuming that, as the number of banks is increased, the number of depositors is increased proportionately, so that the supply of funds in relation to a particular bank is unchanged.
The effect of this change in the model is to make it more like the traditional model of a market in which, as the number of firms increases, the effect of any firm supply on the price of the product becomes vanishingly small. Here the effect of any bank's demand for deposits on the equilibrium deposit rate becomes vanishingly small in the limit as the number of banks becomes unboundedly large. To see this, note that the first-order conditions become
y - R(d) - R'(d)dn-~ = 0
and
p'(y)[y - R(d)] + p(y) = 0.
As before, we can ensure that d remains bounded as n -> oo by assuming that R(d) -> oo as n -> oo. Then the last term on the left hand side of the first equation will vanish as n -> o , leaving a limiting value of (y, d) that satisfies y = R(d). Substituting this in the second equation tells us that p(y) = 0. In other words, as the number of banks increases, the profit margins fall to zero, with the result that banks choose riskier and riskier investments.
PROPOSITION 5: If R(D) -> oo as D -> oo then in any symmetric equilibrium, y - R(d) -> 0 and y -> y as n -> oo.
This is a highly stylized model, so the results have to be taken with a grain of salt; nonetheless, they illustrate clearly the operative principle, which is that competi- tion, by reducing profits, encourages risk taking.
In this particular case, we have constant returns to scale in banking, so that in the limit, when there is a large number of individually insignificant banks, profits must converge to zero. In other words, banks will expand the volume of their
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
464 : MONEY, CREDIT, AND BANKING
deposits and loans until the deposit rate approaches the expected return on invest- ments. But this gives them an extreme incentive to shift risks to the depositors or the deposit insurance agency, since it is only by doing so that they can get positive profits at all.
With constant returns to scale, zero profit is always a necessary condition of equilibrium in a competitive industry. However, there are other ways of ensuring the same outcome even if constant returns to scale is not assumed. We replicated the market by increasing the number of banks and potential depositors in the same proportion. This is an interesting thought experiment, but it is not the same as the comparative static exercise the regulator is undertaking. Presumably, the regulator has to choose n optimally, taking as given the supply-of-funds schedule. Suppose that m is the number of depositors and n the number of banks. Then the market supply-of-funds schedule can be written as R(D/m) if R(.) is the individual supply- of-funds schedule. When m is very large, the supply of funds is elastic, other things being equal, so the banks will take the marginal cost of funds as being equal to the average cost R(D/m). However, increasing the number of banks n in relation to m will force profits down. If d remains bounded away from zero, the average cost of funds must increase to oo and if d goes to zero, profits will also go to zero. In this way, the regulator can achieve the effects of free entry, but there is no need to do this in order to ensure competition. Competition, in the sense of price-taking behavior, follows from having a large market, that is, a large value of m, independent of whether n is large or not. Clearly, the regulator does not want to drive profits to zero if it can be helped, because of the incentives for risk taking that that creates.
Cost of deposit insurance. The preceding analysis assumes that all deposits are insured and that the costs are not born by the banks. This is clearly unrealistic, so it makes sense to consider explicitly the cost of deposit insurance. We assume that the premium for deposit insurance is set before the banks choose their strategies and that it is the same for each bank, independently of the strategy chosen. In equilibrium, the premium accurately reflects the cost of deposit insurance provided by a risk neutral insurer.
Let x denote the premium per dollar of deposits. Then the objective function of bank i is p(yi)(yi - R(D) - p)di and the first-order conditions in a symmetric equilib- rium in which banks choose the strategy (y, d) will be
y - R(d) - - R'(d)dn-1 = 0,
p'(y)[y - R(d) - I] + p(y) = 0.
In equilibrium, the premium must be set so that the expected return on deposits is equal to the return demanded by depositors
R(d) = p(y)(R(d) + ) .
Substituting Xr = [(1 - p(y))/p(y)]R(d) into the first-order conditions yields
y - R(d)/p(y) - R'(d)dn- = 0,
p'(y)[y - R(d)/p(y)] + p(y) = 0.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 465
Let (yn, dn) be a symmetric equilibrium when there are n banks and suppose that (yn, dn) -> (y0, d0) as n -> oo. Then the first-order conditions imply that
lim yn - R(dn)/p(yn) = 0, n->oo
which is only possible if yn > y and p(yn) -> O, as before. Efficiency. Let us leave distributional questions on one side for the moment,
although historically they have been at the center of the arguments for competition in banking, and suppose that the regulator is only interested in maximizing surplus. Reverting to the constant-returs-to-scale case, two necessary conditions for Pareto optimality are that the average cost of funds be equal to the expected return on investments, and that the expected return on investments should be a maximum
R(D/m) = p(y)y ,
p(y)y > p(y)y', Vy'E[O, y] .
Neither of these conditions will hold in equilibrium when m and n are very large. The first condition requires that the volume of deposits expand until the cost of funds equals the expected value of investments. However, when the market is highly competitive, we have y - R(D/m) = O, so that p(y)y - R(D/m) < 0. As we also saw, the second condition cannot be satisfied in equilibrium, since as the market grows large (m, n -> oo), we have y -> y < oo and p(y) -> 0, so p(y)y -> O. This is not only sub-optimal but the worst possible outcome because it minimizes the total surplus.
Suppose instead that we hold the value of m fixed and adjust n to maximize total surplus, taking the equilibrium values (y(n), d(n)) as given functions of n determined by the equilibrium conditions
y - R(nd/m) - R'(nd/m)dm-1 = 0
and
p'(y)[y - R(nd/m)] + p(y) = O .
A "small change" in n will increase the expected revenue by
{p'(y(n))y(n) + p(y(n))}ny'(n) + p(y(n))y(n),
and the cost by
R(nd(n)/m){nd'(n) + d(n)},
so a necessary condition for an (interior) optimum is
{p'(y(n))y(n) + p(y(n))}ny'(n) + p(y(n))y(n) = R(nd(n)/m){nd'(n) + d(n)} .
This can be rewritten as
p(y(n))y(n) - R(nd(n)/m)d(n) = - {p'(y(n))y(n) + p(y(n))}ny'(n) + R(nd(n)/m)nd'(n),
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
466 : MONEY, CREDIT, AND BANKING
where the left hand side is the expected surplus generated by a single bank and the right hand side is the change in expected revenue per bank as n increases plus the change in the cost per bank of the funds borrowed. Since the left hand side is positive, the right hand side must be positive, too. But we know that the second term on the right must be negative since adding more banks reduces the volume of business each bank does, so the first term on the right is positive. We know that y'(n) is negative-increased competition leads to increased risk taking-so the term in braces must be positive. Assuming that p(y)y is concave in y, this tells us that n will be chosen so that y(n) is less than the value that maximizes expected revenue.
A fortiori, it will not, as we have seen, be as great as the value under free entry, since it is never optimal to let n -> oo. It may in fact, be optimal to let the number of banks remain quite small.
2.2 Loan Market Competition The model in Allen and Gale (2000a) analyzes competition in the deposit market.
The bank is assumed to invest deposits directly in a portfolio of assets with given risk characteristics. The bank directly determines the riskiness of its portfolio. As profits decline the bank's preference for risk increases, so increasing competition leads to increasing risk and a decrease in stability. Boyd and De Nicolo (2002) point out that the assumption that banks invest directly in assets is crucial for the result. To show this, they extend the Allen and Gale model to include entrepreneurs. The entrepreneurs obtain loans from the banks and invest the money in risky ventures. Each entrepreneur chooses the riskiness of the venture he invests in. The entrepre- neurs, like the banks in the Allen and Gale model, have a greater incentive to take risk when profits are lower. However, the effect of competition among banks here is the opposite of what we observed in the Allen and Gale (2000a) model. Greater competition among banks reduces the interest rates that borrowers pay, increases the profitability of their ventures, and hence reduces the incentive to take risk. Thus, increased competition among banks leads to increased financial stability. The effect of competition in the deposit market is the same as before but Boyd and De Nicolo are able to show that the loan market effect dominates. The trade-off between competition and stability presented in the Allen and Gale model is reversed in the Boyd and De Nicolo model. As competition between banks increases the risks taken by borrowers is unambiguously reduced and financial stability is improved.
2.3 Dynamic Competition The results in Section 2.1 demonstrate how the limited liability of managers and
shareholders in a moder banking corporation can produce a convex objective function which in turn leads to risk-shifting behavior. This kind of behavior is most likely to occur when the bank is "close to the water line," that is, when the risk of bankruptcy is imminent. For banks which are not in immediate danger of bankruptcy, the risk-shifting argument may be less relevant. However, even if a bank is not close to the water line, there may be other reasons for thinking that its objective
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 467
function is convex. Consider, for example, the winner-takes-all nature of competition. When banks compete for market share, the bank that ends up with the largest share may be able to exploit its market power to increase profitability. In this case, the profit function may be convex in market share, that is, doubling market share may more than double profits. Another reason is the presence of increasing returns to scale. If larger banks have lower average costs, then profits will be a convex function of the size of the bank. Either of these possibilities will give the bank an incentive to take riskier actions, even when the bank is not in immediate danger of bankruptcy.
These incentives for risk-taking behavior are naturally studied in a dynamic context. Suppose that a group of banks are competing over time. Their activities are constrained by the minimum capital ratio, so the only way to expand is to acquire more capital. Because of agency costs or the adverse signaling effects, it may be expensive for banks to raise capital from external sources, so they try to accumu- late capital by retaining earnings. The relative size of the bank matters, because it gives the bank a competitive edge over other banks. Because their reduced-form profit functions are convex and they are constrained by their capital, the game is a race to see who can accumulate capital or market share fastest.
When banks compete to capture greater market share or to reap economies of scale, they consider the effect of their actions not only on immediate profits but also on their future position in the market. How the bank's current actions will affect its future position in the market depends on the nature of the risks involved and on the behavior of the other banks. Even if the profit function is only convex when the bank is close to the bankruptcy point, the bank's objective function may be convex over a much wider region because the bank's objective function incorporates or discounts future possibilities which are still far away. This may influence the shape of the bank's objective function globally, through backward induction.
Allen and Gale (2000a, chap. 8) consider a variety of different models and show that risk taking can either be increased or decreased by competition in a dynamic setting. The first case analyzed involves a pair of duopolists who compete for market share. They play repeatedly for many periods. In each round market shares can go up or down a small amount. By taking a risky action instead of a safe action in any period they increase the variance of the change in market share. A crucial assumption is that there are "reflecting barriers" at the extreme values of market share. When a bank's market share hits zero, it does not go out of business; at worst it will remain at zero for some period before bouncing back. This non-convexity is like having increasing returns in the neighborhood of zero. Similarly, when a bank's market share hits 100% it must eventually bounce back, and this is like having locally decreasing returns. A simple numerical example is used to illustrate the effect of non-convexity on the bank's behavior. When a bank's market share is low, its objective function is convex and it has an incentive to take risk; when its market share is high, its objective function is concave and it has an incentive to avoid risk.
The second case analyzed assumes that there are "absorbing barriers." Now when market share hits zero or one it stays the same with probability one. In this case it is shown that the incentive to take risks is eliminated. The reason is the assumption
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
468 : MONEY, CREDIT, AND BANKING
that the bank's position can only change a little bit at a time. If the period length and the step size are made very small, the binomial process considered would approximate Brownian motion, which has continuous sample paths with probability one. It is the continuity of the movement of the bank's market share over time which eliminates the usual incentive for risk taking. The bank becomes "bankrupt" as soon as its market share hits zero. It cannot go below the line and so it cannot shift risk to depositors or other creditors. Whether incentives to take risks are greater or less with reflecting barriers is ambiguous. While absorbing barriers eliminate the positive incentive for risk taking with reflecting barriers, they also eliminate the in- centive to avoid risk when market share is high that was found in the same model.
Perotti and Suarez (2002) consider the effect of dynamic competition in a model where there can be a banking duopoly or monopoly. If a bank fails when there is a duopoly the market structure switches temporarily to a monopoly. Banks can lend prudently in which case their portfolio of loans has a safe payoff. The alternative is to lend speculatively in which case there is some probability of a high payoff but the average payoff is low. Limited liability and deposit insurance mean lending speculatively can shift risks and be advantageous in the short run. However, a bank can be hit by a random solvency shock. If it lent prudently it will always survive this shock but if it lent speculatively it will fail unless loan returns are high. It is shown that this effect introduces an incentive for banks to lend prudently. A prudent bank will be less exposed to the risk of being driven out of business and will emerge as a monopolist if the other duopolist lent speculatively and is hit by a solvency shock. In their model this "last one standing effect" makes duopolistic banks unam- biguously more prudent and encourages stability.
The range of results obtained with these dynamic models illustrate how crucial the particular details of the model are in determining whether or not competition leads to more or less financial stability.
3. SPATIAL COMPETITION
So far, we have only considered competition in markets for homogeneous com- modities or services. Product differentiation occurs in the financial sector, just as it does in non-financial sectors, and it is important to consider the effect of product differentiation on the competitive process. Models of spatial competition are used to represent competition among firms with differentiated products and the same can be done for the banking sector. We can interpret the spatial dimension literally as representing banks with different locations or we can interpret it metaphorically as representing some other qualitative difference in the services provided. In either case, we find that the effects of concentration on competition in spatial models can be quite different from the results obtained for markets with undifferentiated products. In this section we consider two models of bank competition. The first model, from Allen and Gale (2000a), compares branch banking with unitary banking and shows that competition among a small number of banks (with many branches) may be
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 469
more aggressive than competition among a large number of unitary banks. We then consider a Hotelling-type model of spatial competition and again compare competi- tion among a small number of banks (with many branches) with competition among a large number of unitary banks. The results concerning the trade-off between competition and diversification are very sensitive to the spatial arrangement of the branches.2
3.1 Unitary Banking versus Branch Banking The model presented below exploits two types of imperfections arising from
asymmetric information. The first is the presence of "lock-in" effects. Information is costly for both banks and their customers, whether borrowers or depositors, and once relationship-specific investments in information have been made, the parties may find themselves "locked-in" to the relationship. For example, a borrower having incurred a fixed cost of revealing its type to a bank will suffer a loss if it switches to another bank. In addition there is the well known "lemons effect," that arises if the borrower leaves the bank with which it has been doing business for many years. These lock-in effects will be modeled by simply assuming that there is a fixed cost of switching banks. As is well known (see, e.g., Diamond 1971), switching costs give the bank a degree of monopoly power, even if the bank is not "large."
There are other reasons why banks are monopolistic competitors. For locational reasons their services are not perfect substitutes. Differences in size and products and specialized knowledge also make them imperfect substitutes. Here, we focus on the lock-in effect.
The second essential imperfection arises from the fact that a bank's customers have incomplete information about the services offered by a bank and the prices at which these services are offered, at the time when the relationship has begun. In fact, the smaller the bank is, the less likely it is that the bank's reputation will be an adequate source of information about the quality and prices of the bank's products.
A third important feature of this model is the fact that banks offer a variety of services. The simplest example of this is the case of a bank with a large number of branches. Since customers have different preferences over branch location, branches at different locations are offering different services. This means that a bank with many branches is offering a bundle of different services to their customers.
Location is not the only dimension along which banks differ, of course. They will offer different menus of accounts, or concentrate on different types of lending business; they will attract a different mix of retail or wholesale funds; they may diversify into non-bank products such as insurance or mutual funds. Location is a convenient metaphor for these different dimensions.
By exploiting these three features of the model, lock-in effects, limited information, and product diversity, we can reverse the usual presumption that greater concentra- tion leads to more efficient outcomes.
2. For other welfare analyses involving spatial competition and risk see Besanko and Thakor (1992) and Matutes and Vives (1996, 2000).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
470 : MONEY, CREDIT, AND BANKING
The model. There is a finite set of locations indexed by I = 1,...,L and at each location there are two banking offices j = 1,2. Time is divided into an infinite number of discrete periods t = 1,2,.... There is a large number of individuals allo- cated exogenously to the different locations. Each period, these individuals have a demand for a unit of banking services which provides them with a surplus v. The value of banking services to individuals is distributed according to the distribution function F(v), that is, F(v) is the fraction of the population with valuation less than or equal to v.
At each date consumers are randomly assigned to a new location. They have an equal probability of arriving at any location and we assume that the number of locations is so large that the probability of returning to the same location is negligible and can be ignored. This is an extreme assumption, to be sure, but it serves to eliminate inconvenient and apparently unimportant complications. Each location is also assumed to receive a representative sample of the different types of individuals so, whatever the number of individuals at a given location, the distribution of types is F(v).
For simplicity, banks are assumed to have a zero marginal cost of providing banking services. Profit is thus identical to revenue.
At each date, the market is assumed to clear as follows. First, the individuals who have gathered at a particular location choose which bank to patronize. They do this before they know the price that the bank will charge. Next, the bank sets the price for its product (the interest rate on loans or deposit accounts, or the fees for other bank services). Finally, the consumers make one of three choices: to purchase the current bank's services at the quoted price, to switch to the other bank, whose price by now has been fixed, or to do without the services of a bank. There is a fixed cost c > 0 of switching from one banking office to the other.
We consider two limiting cases of bank organization. In the first, which we call unitary banking, each bank has a single branch. In other words, each banking office represents an independent bank. In the second case, which we call branch banking, there are only two banks, each owning one branch in each location. That is, all the banking offices are organized into two large networks. Regardless of the form of organization, we use the term banking office to denote the smallest unit of the bank, whether it constitutes the entire bank or a branch of a larger bank network.
Individuals are assumed to observe only what happens at their own bank in each period and, since they move to a different location in each period, they have no knowledge of the previous behavior of the bank they are patronizing in the current period. The banks themselves are assumed to condition their behavior in each location on their experience at the same location. This maintains an informational symmetry between the unitary- and branch-banking forms of industrial organization, i.e., unitary banking and branch banking, since in each case only local information is being used to condition the (local) pricing decision.
Unitary banking. In this form of industrial organization each bank consists of a single office. A bank sets the price of its product in each period to maximize the present value of profits. In the static version of this model, it is well known that
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 471
the unique equilibrium involves each banking office choosing the monopoly price. More precisely, suppose that there is a unique price PM such that
PM(1 - F(pM)) > p(l - F(p)), Vp.
A bank will clearly never want to charge more than PM. If the two banks at some location happen to charge prices p < p' < pM, where p < PM, the bank charging p can always raise its price by e without losing any customers, because of the fixed cost of switching. This will clearly increase its profits, so the only equilibrium is for both banking offices to charge PM.
In a dynamic context things are generally more complicated, because of the possibility of supporting a different equilibrium by means of punishment strategies. Under the maintained assumptions, however, there is no possibility of using such strategies to increase the set of equilibria. Because individuals only observe what happens at their own locations and never return to the same location, nothing that a banking office does in the current period will be observed by individuals who will visit that location in the future. Further, since banks at other locations do not condition their behavior on what happens at this location, there is no possibility of the bank's future customers being indirectly informed of a deviation through another bank's reaction to this bank's current deviation. Our informational assumptions have effectively severed any possible feedback from a current change in price to a future change in demand, so the argument used in the static model continues to apply. We conclude, then, that the unique, subgame perfect equilibrium of the unitary banking game consists of each bank, in each location, charging the monopoly price PM in every period. Consumers whose valuation v is greater than PM will purchase banking services; those whose valuation is less than PM will not. The equilibrium is inefficient for the usual reason: the monopoly price is too high and the monopoly quantity is too low.
Branch banking. Now suppose that banking offices are formed into two large networks. Each bank has one branch in each of the locations. Although consumers move from location to location they can stay with the same bank if they wish. The possibility of staying with the same bank generates a plethora of other equilibria. We describe one such equilibrium to illustrate the possibilities.
In each period, at each location, half the customers patronize each of the banks. Along the equilibrium path, the banks charge a price p = ? > 0 in every period. If, in any period, one of the banking offices has deviated from the equilibrium strategy, all the customers will leave the bank that last deviated and henceforth patronize branches belonging to the other bank. The banks continue to charge the same price. If no bank deviated, but some of the customers deviated in the past, the banks continue to charge the same price and the customers continue to patronize the two banks equally. The profits (for a single banking office) from deviating last one period and are less than or equal to (PM - e)(1 - F(PM)); on the other hand, it loses the profits from these customers in each future period until they all disappear from the game. Customers only last a finite number of periods, but if the number of locations is large, they will be around on average for a very long time. Each period, a fraction
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
472 : MONEY, CREDIT, AND BANKING
- 1 of the customers dies and is replaced. So the equilibrium profits lost by a single banking office's deviation are equal to [E/(1 - ?-1)(1 - 8)](1 - F(E)). For 8 suffi- ciently close to 1 and I sufficiently large, the profits from deviating are less than the profits of the equilibrium strategy. This shows that under branch banking, it is possible to support equilibria which are "more efficient" than the unique equilibrium in the case of unitary banking, where "more efficient" means that the sum of consumers' and producers' surplus is greater.
How should we interpret these results? The lock-in effects in the banking sector may be substantially greater than in most service industries and may be one of its distinguishing features. Small banks with a limited range of services and a limited geographical presence may have a greater incentive to exploit the lock-in effect than a large bank, because the large bank is always competing for the customer's future business, in another product line or another location.
Empirical evidence. There is some empirical evidence in support of the view of competition presented above.3 Bordo, Rockoff, and Redish (1994) compare the Canadian and US banking systems from 1920 to 1980. During this period Canada had a few branch banks while the US had unit banking in many parts. It is found that the Canadian system outperformed the US system in a number of respects. First, in Canada the interest rates paid on deposits were generally higher and the income received by security holders was generally slightly higher than in the US. Second, the interest rates charged on loans were generally quite similar in the two countries. Finally, the returns on equity were generally higher in Canada. Taken together this evidence is consistent with the Canadian branch banking system being more competitive than the US unitary banking one.
Carlson and Mitchener (2003) also find evidence that branch banking is more competitive than unitary banking. Using data on national banks from the 1920s and 1930s, they compare states which just have unit banks with ones that have both branch banking and unitary banking. In the former states, many banks that charge high rates of interest on loans and pay low rates on deposits survive. However, in states with both types of banks this kind of local monopoly is eliminated. Both branch banks and unit banks price in exactly the same way.
3.2 Spatial Competition and Diversification We next consider another model of spatial competition in the tradition of Hotelling.
It is shown that competition can be consistent with diversification and hence stability. However, the precise way in which this works depends crucially on the particular assumptions made.
Let locations be denoted by n = 0, ?1, +2,... and assume that there is a single bank at each location. Identical individuals are uniformly distributed on the real line. Each has one unit of a good that can be invested by the bank in a risky asset with return Rn. For simplicity assume that the returns Rn are i.i.d. with distribution
3. We are grateful to Stephen Haber for bringing this evidence to our attention.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 473
R -JR w.pr. n n- 0 w.pr. 1 - 7.
In a symmetric equilibrium a bank will draw clients from the interval [n - 1/2, n + 1/2] and offer them a risk sharing contract that promises a payment D if the project is successful and nothing otherwise (the banker is risk neutral but has limited liability). The expected utility of the contract is nU(D) + (1 - n)U(O) = ntU(D) if U(O) = 0.
The bank chooses D to maximize profits taking as given the expected utility a offered by the adjacent banks. Assuming linear transportation costs of one utility per unit distance, the marginal agent m < n who goes to bank n is determined by the condition that
rU(D) - (n - m) = - (m - n + 1)
or
2(n - m) = nU(D) - a + 1 .
The profit per agent is n(R - D) so total profits are
n(R - D)2(n - m) = n(R - D)(7cU(D) - u + 1).
The first-order condition is
7U'(D) = R R-D
If we allow a bank to occupy several locations, it can pool several independent risks, which allows it to offer better risk sharing to its customers. However, it may face less competition. Whether it does depends on precisely which locations a bank is allowed to occupy. For example, if two banks occupy alternate locations, we get improved risk sharing with no loss of competition. Here there is no trade-off between competition and stability. If a bank occupies a sequence of adjacent loca- tions, it can extract a large amount of surplus from the consumers located in the middle. Here there is a trade-off between competition and stability.
4. SCHUMPETERIAN COMPETlIION
Technological innovation is one of the major sources of growth in welfare. As Schumpeter (1950) famously pointed out, perfect competition undermines the incen- tive to innovate (when intellectual property rights are weak) and in that sense imperfect competition may be more "efficient" than perfect competition. Similar ideas apply in the financial sector. If banks innovate they may be able to capture the market and drive other banks out of business. Thus Schumpeterian competition may be associated with financial instability (creative destruction). We start by considering
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
474 : MONEY, CREDIT, AND BANKING
a "winner-takes-all" model of competition based on the work of Allen and Gale (2000c) which has this feature.
4.1 Winner-Takes-All Competition There is a finite number of locations i = 1,..., n with a bank at each location.
There are two dates, t = 0,1. At date 0, each bank i invests ki > 0 in the development of a product. The banks' opportunity cost of capital is R. The value of the product developed by bank i is given by Vi(ki, o). There is symmetric information all agents know the function Vi(.) and the investment ki and have the same continuous probability distribution F(-) over the states of nature ca We assume that V(0, co) = 0 for all o so capital is essential to the development of a useful product. In general, the more the capital that is provided the greater is the probability that the value of the product is high. We assume that once the new product is developed it can be produced at constant marginal cost and, without essential loss of generality, we set the marginal cost equal to zero.
At date 1 identical consumers are uniformly distributed on the line interval [0, n]. Each consumer wants to consume at most one unit of a new product.
4.2 Equilibrium At date 0 the banks jointly choose their investment strategies k = (kl,..., kn). At
the beginning of date 1, the state of nature o is realized and the banks observe the quality of the product they have developed
V(k, co ) = Vl(kl, o ),... V(kn, )).
Then the banks engage in Bertrand competition. Since the qualities are assumed to be continuously distributed (for ki > 0) the probability of ties can be ignored. Then Bertrand competition will lead to an outcome in which the best product captures the entire market and the price charged for this product is equal to the difference between the value of the first- and second-best products. The price for every other product is zero. At this price, consumers are indifferent between the first- and second- best products, but they will demand only the first-best product in equilibrium (if a positive fraction of consumers were expected to choose the second-best product, the firm with the first-best product would have chosen a slightly lower price to capture the entire market). Formally, for any bank i let
V-i(k-i, co) = (Vl(kl, o ) ...,Vi_1(ki_I, Co ),Vi+l(ki+l, CO), ... Vn(kn, ))
denote the vector of the qualities of products j X i; let
k-i = (kl- ...ki-1, k-i+ , ..., kn)
denote the allocation of investment in products j ? i; and let
V*_i(k_i, w) = maxjti{Vj(kj, o )}
denote the highest value in the vector V_i(k_i, o). Then, in the second-period equilib- rium, the price charged for the ith product is denoted by pi(k, co) and satisfies
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE 475
pi(k, co) = max{Vi(k, o) - V*i(k_i, co),0}, Vi.
Since the demand is equal to one for the best product and zero for the rest, the revenue of bank i is also equal to pi(k, o).
At the first date, we look for a Nash equilibrium in the investment levels. The ith bank chooses ki to maximize E[pi(ki, k_i, o )] - Rki, taking as given the invest- ment levels of the other banks, k_i. So a Nash equilibrium is a vector k* such that
kiEarg maxki>o{E[pi(ki, k_i, co)] - Rki}
for each i.
4.3 Optimum Since the cost of production at date 1 is zero, the surplus generated by consuming
the ith product is Vi(ki, o). Surplus is maximized by having all consumers consume the best product, so the total surplus at date 1 is
V*(k, co) _ maxi= ...{Vi(ki, co ) .
Assuming that the consumers are also risk neutral and that lump sum transfers are possible, the first-best efficient allocation is found by maximizing net surplus, that is, by solving the planner's problem
n
maxkV*(k) - Rki,
i=1
where V*(k) = E[V*(k, o )] is the expected value of V*(k, co). Define V*_i(k_i) = E[V*-i(k_i, o )]. Then the objective function V*(k) can be written
equivalently as
V*(k) = E[max{Vi((ki, o )- V*i(k-i, co), 0} + V*i(k-i, o )] = E[max{Vi(ki, co) = V*-i(k-i, o ),0}] + V*i(k-i)
and the planner's problem can be rewritten equivalently as n
maxk>oE[max{Vi(ki, co) - V*i(k, o), O} + V*i(ki)- REki i=1
Suppose that k* is a solution to the planner's problem above. A necessary condition is that ki maximizes
E[max{Vi(ki, co) -- Vi*(k*i, o ), 0}] - Rki = E[pi(ki, k* )] -Rki.
But this means that ki* satisfies the equilibrium condition for the ith bank's choice of ki. Hence, we have the following result.
PROPOSITION 6: If k* is a solution to the planner's problem, then k* is a Nash equilibrium of the banks' investment "game."
In this case the allocation produced by the winner-takes-all competition is efficient. The innovating banks have exactly the right incentives to invest. However, only
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
476 : MONEY, CREDIT, AND BANKING
one bank survives in each period. In other words there is considerable financial instability. In order to prevent this instability the government may wish to restrict competition and give each bank a monopoly in a particular region. We turn next to this restricted competition.
4.4 Restricted Competition Next assume that the bank located at n has a legal monopoly of the region [i - 1/2,
i + 1/2]. Then it maximizes E[V(ki,o) - Rki]. Clearly, surplus must be lower, both because we are not providing the best product to all consumers but also because we are not using the right investment level.
Will investment be too high or too low? Examples can be constructed to show that the answer is ambiguous. On the one hand losers are protected under the "competitive" arrangement. On the other hand winners suffer because they cannot capture the entire market.
To see this consider the following simple example. There is a continuum of identical consumers with measure one and two banks. Each bank can invest 0 or 0 < I < 1/2. If a bank invests zero the value of its good is zero; if it invests I the value of its good is 1. If the market is unified, there are three equilibria, two asymmetric equilibria in which only one bank invests and a mixed strategy equilib- rium in which the probability of investment is X = 1 - I. Now suppose we divide the market between the two banks, each getting 1/2 of the market. Then each bank will invest for sure, since I < 1/2.
Comparing equilibria, we see that total investment is greater in the divided market. Also, comparing the unique pure strategy equilibrium of the divided market with the mixed strategy of the unified market, we can see that (the probability of) innovation is strictly higher in the divided market. In the mixed strategy equilibria of the divided market the probability of innovation is the same.
For the case I > 1/2 there will be no investment, and hence no innovation, in the segmented market. In this case investment and innovation are unambiguously lower than in winner-takes-all case.
These examples are enough to indicate that even in simple models it is hard to obtain robust comparative static results. As with the other models we have examined, the relationship between competition and stability is complex and nuanced.
5. CONTAGION
One source of instability in financial systems is the possibility of contagion, in which a small shock that initially affects one region or sector or perhaps even a few institutions, spreads from bank to bank throughout the rest of the system, and then affects the entire economy. There are a number of different types of contagion that have been suggested in the literature. The first is contagion through interlinkages between banks and financial institutions (see, e.g., Rochet and Tirole, 1996a, 1996b, Freixas and Parigi, 1998, Freixas, Parigi, and Rochet, 2000, and Allen and Gale,
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 477
2000b, for theoretical analyses and Van Rijckeghem and Weder, 2000, for empirical evidence). The second is contagion of currency crises (see, e.g., Masson, 1999, Eichengreen, Rose, and Wyplocz, 1996, and Glick and Rose, 1999). The third is contagion through financial markets (see, e.g., King and Wadwhani, 1990, Kyle and Xiong, 2001, and Kodres and Pritsker, 2002).
The notion of financial fragility is closely related to that of contagion. When a financial system is fragile a small shock can have a big effect. A financial crisis may rage out of control and bring down the entire economic edifice (see, e.g., Kiyotaki and Moore, 1997, Chari and Kehoe, 2000, Lagunoff and Schreft, 2001, and Allen and Gale, 2003b).
In this section we are interested in the relationship between contagion and financial fragility and competition. Allen and Gale (2000b) develop a model of financial contagion through the interbank market. They assumed perfect competition. In that case a small aggregate shock in liquidity demand in a particular region can lead to systemic risk. Although the shock is small it may cause a bank to go bankrupt and liquidate its assets. This in turn causes other banks which have deposits in it to also go bankrupt and so on. Eventually all banks are forced to liquidate their assets at a considerable loss. Perfect competition in the interbank market plays an important role in this contagion. Since each bank is small, acts as a price taker and assumes its actions have no effect on the equilibrium, no bank has an incentive to provide liquidity to the troubled bank.
Saez and Shi (2004) have argued that if banks are limited in number they may have an incentive to act strategically and provide liquidity to the bank that had the original problem. This will prevent the contagion and make the banks providing funds in this way better off. The formal model of this phenomenon that captures the role of imperfect competition is similar to the model in Bagnoli and Lipman (1989). They consider the problem of the provision of public goods through private contributions. There is a critical level of resources required to provide the public good. Each person becomes pivotal and in this case the public good can be provided. Similarly, here each bank would be pivotal in the provision of liquidity to pre- vent contagion.
Suppose there are two banks, each holding A units of an illiquid asset and M units of money. One unit of the asset is worth one unit if liquidated today and R units if liquidated tomorrow. Money is storable, so one unit of money today is worth one unit tomorrow. Each bank owes one unit of money to a depositor who withdraws today and the liquidation value of the portfolio to a depositor who withdraws tomorrow. In addition, bank 1 owes bank 2 B units today. Suppose that bank 1 has K early consumers and 1 - X late consumers. If k + B > M + A - ((1 - X)/R), then bank 1 cannot meet its commitments on its own. If it fails, however, bank 2 can claim only a fraction (B/(1 + B))(M + A) < B. Bank 2 has several actions it can take. It can forgive part of its debt. It can wait for payment at date 2. It can make a cash transfer to bank 1 at date 1. If R is big enough, this may be better for bank 2 than
enforcing its debt now.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
478 : MONEY, CREDIT, AND BANKING
The same argument obviously applies to any number of banks, though the coordi- nation problem will become more severe as the numbers increase. If there is a continuum of banks or asymmetric information, it may be impossible to sustain the cooperative equilibrium. Note that even in the case with a finite number of banks there may be a coordination failure; if every bank thinks the others will not contribute anything, it may be optimal not to contribute anything (this may depend on the extensive form, e.g., simultaneous moves rather than sequential moves).
Gradstein, Nitzan, and Slutsky (1993) show that Bagnoli and Lipman's result is not very robust to the introduction of uncertainty. It remains to be seen whether a model of contagion and imperfectly competitive banks would also not be robust.
A model of an imperfectly competitive interbank market may therefore be more stable than the case where there is perfect competition. As in the original agency model of Keeley (1990) there is again a trade-off between competition and finan- cial stability.
6. CONCLUDING REMARKS
In this paper we have considered a variety of different models of competition and financial stability. These include general equilibrium models of financial interme- diaries and markets, agency models, models of spatial competition, Schumpeterian competition, and contagion. There is a very wide range of possibilities concerning the relationship between competition and financial stability. In some situations there is a trade-off as is conventionally supposed but in others there is not. For example, with general equilibrium and Schumpeterian models efficiency requires the combination of perfect competition and financial instability.
There is a large policy literature based on the conventional view that there is a trade-off between competition and stability. Since competition is generally viewed as being desirable because it leads to allocational efficiency, this perceived trade-off lead to calls for increased regulation of the banking sector to ensure the coexistence of competition and financial stability. The most popular instrument for achieving this end was the imposition of minimum capital requirements on banks. If the owners of banks were forced to put up significant amounts of capital, they would be unwilling to take risks because they would again stand to loose large amounts of funds. The Basel Agreement of 1988 imposed capital controls on banks so that the incentive to take risks would be reduced and they could compete on equal terms. A large literature has developed analyzing the effect of capital controls. For example, Hellman, Mur- dock, and Stiglitz (2000) showed in the context of a simple model of moral hazard that capital controls were not sufficient. In addition to capital controls, deposit rate controls were also necessary to achieve Pareto efficiency.
Our analysis suggests that the issue of regulation and its effect on competition and financial stability is complex and multi-faceted. Careful consideration of all the factors at work both at a theoretical and empirical level is required for sound policy.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
FRANKLIN ALLEN AND DOUGLAS GALE : 479
LITERATURE CITED
Allen, F., and D. Gale (1994). "Limited Market Participation and Volatility of Asset Prices." American Economic Review 84, 933-955.
Allen, F. and D. Gale (1998). "Optimal Financial Crises." Journal of Finance 53, 1245-1284.
Allen, F., and D. Gale (2000a). Comparing Financial Systems. Cambridge, MA: MIT Press.
Allen, F., and D. Gale (2000b). "Financial Contagion." Journal of Political Economy 108, 1-33.
Allen, F., and D. Gale (2000c). "Corporate Governance and Competition." In Corporate Governance: Theoretical and Empirical Perspectives, edited by X. Vives, pp. 23-94. Cambridge, UK: Cambridge University Press.
Allen, F., and D. Gale (2003a). "Financial Intermediaries and Markets." Working Paper 00- 44-C, Wharton Financial Institutions Center. Econometrica, forthcoming.
Allen, F, and D. Gale (2003b). "Financial Fragility, Liquidity and Asset Prices." Working Paper 01-37-B, Wharton Financial Institutions Center.
Bagnoli, M., and B. Lipman (1989). "Provision of Public Goods: Fully Implementing the Core through Private Contributions." Review of Economic Studies 56, 583-601.
Beck, T., A. Demirguc-Kunt, and R. Levine (2003). "Bank Concentration and Crises." Working Paper, World Bank.
Berger, A., and D. Humphrey (1997). "Efficiency of Financial Institutions: International Survey and Directions for Future Research." European Journal of Operational Research 98, 175-212.
Besanko, D., and A. Thakor (1992). "Banking Regulation: Allocational Consequences of Relaxing Entry Barriers." Journal of Banking and Finance 16, 909-932.
Bordo, M., H. Rockoff, and A. Redish (1994). "The U.S. Banking System from a Northern Exposure: Stability versus Efficiency." Journal of Economic History 54, 325-341.
Boyd, J., and G. De Nicolo (2002). "Bank Risk-Taking and Competition Revisited." Working Paper, Carlson School of Management, University of Minnesota.
Canoy, M., M. van Dijk, J. Lemmen, R. de Mooij, and J. Weigand (2001). Competition and Stability in Banking. The Hague, Netherlands: CPB Netherlands Bureau for Economic Policy Analysis.
Carletti, E., and P. Hartmann (2003). "Competition and Financial Stability: What's Special about Banking?" In Monetary History, Exchange Rates and Financial Markets: Essays in Honour of Charles Goodhart, Vol. 2, edited by P. Mizen. Cheltenham, UK: Edward Elgar.
Carlson, M., and K. Mitchener (2003). "Branch Banking, Bank Competition and Financial Stability." Working Paper, Department of Economics, Santa Clara University.
Chari, V., and P. Kehoe (2000). "Financial Crises as Herds." Working Paper, Federal Reserve Bank of Minneapolis.
Diamond, P. (1971). "A Model of Price Adjustment." Journal of Economic Theory 3, 156- 168.
Diamond, D., and P. Dybvig (1983). "Bank Runs, Deposit Insurance, and Liquidity." Journal of Political Economy 91, 401-419.
Eichengreen, B., A. Rose, and C. Wyplocz (1996). "Contagious Currency Crises: First Tests." Scandinavian Journal of Economics 98, 463-484.
Freixas, X., and B. Parigi (1998). "Contagion and Efficiency in Gross and Net Interbank Payment Systems." Journal of Financial Intermediation 7, 3-31.
Freixas, X., B. Parigi, and J. Rochet (2000). "Systemic Risk, Interbank Relations and Liquidity Provision by the Central Bank." Journal of Money, Credit, and Banking 32, 611-638.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
480 : MONEY, CREDIT, AND BANKING
Glick, R., and A. Rose (1999). "Contagion and Trade: Why are Currency Crises Regional?" In The Asian Financial Crisis: Causes, Contagion and Consequences, edited by P. Agenor, M. Miller, D. Vines, and A. Weber, chap. 9. Cambridge, UK: Cambridge University Press.
Gradstein, M., S. Nitzan, and S. Slutsky (1993). "Private Provision of Public Goods under Price Uncertainty." Social Choice and Welfare 10, 371-382.
Hellmann, T., K. Murdock, and J. Stiglitz (2000). "Liberalization, Moral Hazard in Banking, and Prudential Regulation: Are Capital Requirements Enough?" American Economic Review 90, 147-165.
Hoggarth, G., and V. Saporta (2001). "Costs of Banking System Instability: Some Empirical Evidence." Financial Stability Review (June 2001), 148-165.
Jensen, M., and W. Meckling (1976). "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure." Journal of Financial Economics 3, 305-360.
Keeley, M. (1990). "Deposit Insurance, Risk and Market Power in Banking." American Economic Review 80, 1183-1200.
King, M., and S. Wadhwani (1990). "Transmission of Volatility between Stock Markets." Review of Financial Studies 3, 5-33.
Kiyotaki, N., and J. Moore (1997). "Credit Chains." Journal of Political Economy 105, 211-248.
Kodres L., and M. Pritsker (2002). "A Rational Expectations Model of Financial Contagion." Journal of Finance 57, 768-799.
Kyle, A., and W. Xiong (2001). "Contagion as a Wealth Effect." Journal of Finance 56, 1401-1440.
Lagunoff, R., and S. Schreft (2001). "A Model of Financial Fragility." Journal of Economic Theory 99, 220-264.
Masson, P. (1999). "Contagion: Monsoonal Effects, Spillovers and Jumps between Multiple Equilibria." In The Asian Financial Crisis: Causes, Contagion and Consequences, edited by P. Agenor, M. Miller, D. Vines, and A. Weber, chap. 8. Cambridge, UK: Cambridge University Press.
Matutes, C., and X. Vives (1996). "Competition for Deposits, Fragility and Insurance." Journal of Financial Intermediation 5, 184-216.
Matutes, C., and X. Vives (2000). "Imperfect Competition, Risk Taking, and Regulation in Banking." European Economic Review 44, 1-34.
Perotti, E., and J. Suarez (2003). "Last Bank Standing: What Do I Gain if You Fail?" European Economic Review 46, 1599-1622.
Rochet, J., and J. Tirole (1996a). "Interbank Lending and Systemic Risk." Journal of Money, Credit, and Banking 28, 733-762.
Rochet, J., and J. Tirole (1996b). "Controlling Risk in Payment Systems." Journal of Money, Credit, and Banking 28, 832-862.
Saez, L., and X. Shi (2004). "Liquidity Pools, Risk Sharing and Financial Contagion." Journal of Financial Services Research 25, 5-23.
Schumpeter, J. (1950). Capitalism, Socialism and Democracy, 3rd edition. New York: Harper & Row.
Van Rijckeghem, C., and B. Weder (2000). "Spillovers through Banking Centers: a Panel Data Analysis." IMF Working Paper WP/00/88, Washington, DC: International Monetary Fund.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:50:11 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. [453]
- p. 454
- p. 455
- p. 456
- p. 457
- p. 458
- p. 459
- p. 460
- p. 461
- p. 462
- p. 463
- p. 464
- p. 465
- p. 466
- p. 467
- p. 468
- p. 469
- p. 470
- p. 471
- p. 472
- p. 473
- p. 474
- p. 475
- p. 476
- p. 477
- p. 478
- p. 479
- p. 480
- Issue Table of Contents
- Journal of Money, Credit and Banking, Vol. 36, No. 3, Part 2: Bank Concentration and Competition: An Evolution in the Making A Conference Sponsored by the Federal Reserve Bank of Cleveland May 21-23, 2003 (Jun., 2004), pp. 433-654
- Front Matter
- Bank Concentration and Competition: An Evolution in the Making [pp. 433-451]
- Competition and Financial Stability [pp. 453-480]
- Comment on "Competition and Financial Stability" by Franklin Allen and Douglas Gale [pp. 481-486]
- Crises in Competitive versus Monopolistic Banking Systems [pp. 487-506]
- Comment on "Crises in Competitive versus Monopolistic Banking Systems" by John H. Boyd, Gianni De Nicoló, and Bruce D. Smith [pp. 507-509]
- How Foreign Participation and Market Concentration Impact Bank Spreads: Evidence from Latin America [pp. 511-537]
- Comment on "How Foreign Participation and Market Concentration Impact Bank Spreads: Evidence from Latin America" by Maria Soledad Martinez Peria and Ashoka Mody [pp. 539-542]
- Real Effects of Bank Competition [pp. 543-558]
- Comment on "Real Effects of Bank Competition" by Nicola Cetorelli [pp. 559-562]
- What Drives Bank Competition? Some International Evidence [pp. 563-583]
- Comment on "What Drives Bank Competition? Some International Evidence" by Stijn Claessens and Luc Laeven [pp. 585-592]
- Regulations, Market Structure, Institutions, and the Cost of Financial Intermediation [pp. 593-622]
- Comment on "Regulations, Market Structure, Institutions, and the Cost of Financial Intermediation" by Asli Demirgüç-Kunt, Luc Laeven, and Ross Levine [pp. 623-626]
- Bank Competition and Access to Finance: International Evidence [pp. 627-648]
- Comment on "Bank Competition and Access to Finance: International Evidence" by Thorsten Beck, Asli Demirgüç-Kunt, and Vojislav Maksimovic [pp. 649-654]
- Back Matter
Competition and Financial Stability.pdf
Comment on "Competition and Financial Stability" by Franklin Allen and Douglas Gale Author(s): Charles M. Kahn Source: Journal of Money, Credit and Banking, Vol. 36, No. 3, Part 2: Bank Concentration and Competition: An Evolution in the Making A Conference Sponsored by the Federal Reserve Bank of Cleveland May 21-23, 2003 (Jun., 2004), pp. 481-486 Published by: Ohio State University Press Stable URL: http://www.jstor.org/stable/3838947 .
Accessed: 13/02/2015 16:16
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Ohio State University Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of Money, Credit and Banking.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:16:11 PM All use subject to JSTOR Terms and Conditions
CHARLES M. KAHN
Comment on "Competition and Financial
Stability" by Franklin Allen and Douglas Gale JEL codes: G28, E44, L51
Keywords: financial contagion, banking regulation.
Allen and Gale (2004, this issue of JMCB) provide a thought-provoking tour d'horizon. I'd call it "ambitious," except that what they label a "prelude to the development of a theoretical framework" is too honest to be ambitious: they stead- fastly refuse to give a definitive answer to their fundamental question, "Does competi- tion encourage or discourage adequate financial stability?"-in other words, "Is there a tradeoff between competition and financial stability or are the two goals mutually reinforcing?"
In addressing this question, Allen and Gale employ several different interpretations of the two goals: In some sections, "competition" means price-taking behavior. In others, they interpret competition to be imperfect competition with some degree of market power. By "instability" Allen and Gale sometimes simply mean "the possibility of default." Often instability means the choice of overly risky or undiversi- fled investments. But in their most interesting case instability means a high degree of interdependence between bank defaults.
1. THE ARGUMENTS
I will also use this review as an opportunity to consider the often uncomfortable relationship between economic theory and policy making. Policy makers want, and need, definitive answers to questions like Allen and Gale's fundamental question. Theorists spin answers in every possible direction. When we are lucky, empiricists sort through those answers to determine their relative importance. Nonetheless this
CHARLES M. KAHN is a professor in the Departments of Economics and Finance at the University of Illinois Urbana-Champaign. E-mail: [email protected]
Received May 29, 2003; and accepted in revised form May 29, 2003.
Journal of Money, Credit, and Banking, Vol. 36, No. 3 (June 2004, Part 2) Published in 2004 by The Ohio State University Press.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:16:11 PM All use subject to JSTOR Terms and Conditions
482 : MONEY, CREDIT, AND BANKING
process takes time and often fails to reach any definite conclusions. In the meanwhile, the theoretical arguments are all we have, and like Harry Truman, current policy makers sigh for a "one-handed economist."
Allen and Gale give them no comfort. They provide, not two, but many hands:1 1. The invisible hand. Allen and Gale (2003), following Prescott and Townsend
(1984), show conditions under which the welfare theorems hold in economies with financial intermediaries. In terms of this work "competition" means Walrasian price taking and "financial instability" means violating a contract by declaring bankruptcy. In this sense, financial instability is only present in an incomplete market setting; in a complete contract setting, there is no sense to the notion of violating a contract because every possible reaction and consequence is included in the terms of the contract itself. Allen and Gale (2003) specify a class of models in which incom- plete contracts are optimally chosen, provided the market for underlying Arrow securities is complete. Crucial to the structure is the assumption of no ex ante differential information; contracts are written and Arrow securities traded by symmet- rically informed agents. In such a world, there is no reason for a government to intervene to stabilize a financial market.
Although this is a natural starting point for a discussion, the argument seems to me to be a straw man. Even though bankruptcy procedures are expensive, sometimes particular banks ought to fail; few regulators or policy makers would argue otherwise. The real costs of financial instability must be found elsewhere.
2. A neighborly hand. Geographic diversification reduces risk. It may also reduce competition, depending on whether the geographically diverse banks have local monopolies or end up competing region by region. Nonetheless, since it is hard to see how consolidation could increase competition, score one for the tradeoff camp.
3. A hand in the cookie jar. Limited liability and high leverage, possibly exacer- bated by deposit insurance, by a winner-take-all objective function, or by some, though not all, market-share objective functions (or by options in executives' com- pensation packages?) tempt banks into overly risky positions. The lower the bank's charter value (for example due to increased competition), the greater the temptation. Allen and Gale's model of this is attractive in its simplicity and illustrates one classic form of the tradeoff between risk and competition. And yet, as they point out, the result is reversed if the potential for excessive risk taking is in the hands of the borrowers rather than the banks, or if the objective includes being the "last one standing."
4. The dead hand of the past. Lock-in of demand, through fixed costs of switching banks, can cause otherwise competitive firms to have monopoly power. Suppose nationwide banks are more likely to have continuing relationships with customers than are small banks. Then it is at least possible that such continuing relationships can lead to sustainable low prices through the operation of threats and counter threats in an infinitely repeated game.
1. And since they are old hands at this material, the fingerprints from these hands are often their own; several of the models they describe are found in greater detail in Allen and Gale (2000).
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:16:11 PM All use subject to JSTOR Terms and Conditions
DISCUSSION : 483
There is some empirical evidence that branch banks price more competitively than do unit banks. But for the story to be taken seriously we need better evidence that, for example, mobility causes nationwide banks to have more long-lived relations with customers than do local banks.
5. The hand of Shiva. Under Schumpeterian creative destruction, efficiency (that is, innovation) comes hand-in-hand with instability (that is, failure of unsuccessful enterprises). Allen and Gale present a model in which efficient innovation is achieved in an environment in which only one firm survives. The tradeoff between dynamic efficiency and static efficiency means that restriction of the competition may be desirable, but the correct remedy and the consequences for innovation and for stability are complex.
6. Dirty hands. Contagion-the spread of financial malady from one institution to another-is Allen and Gale's final interpretation of instability. The development of theories of financial fragility is currently a growth industry (in addition to the papers they cite, there are noteworthy recent contributions by Chang and Velasco, 2001, Cooper and Corbae, 2002, and Diamond and Rajan, 2001). Allen and Gale point out that competition, meaning price-taking behavior, increases the likelihood of instability, because financial system stability is a public good, one subject to free riding.
2. IMPLICATIONS: POLICY AND RESEARCH
What are we to make of all of this? Despite the high-minded even-handedness of the piece, I detect a mildly anti-interventionist undertone. Many clever arguments for intervention have been thought up over the years. Each requires a precise calculation of type and degree of intervention. Offsetting one argument against another implies a preference for inaction.
The political economy of the relationship between economists and policy makers ought to reinforce this bias. There is a great temptation for policy advocates to use interventionist theories as cover for rent-seeking behavior: "We need to limit entry into these financial activities in order to increase the stability of the financial system; we need charter value to encourage us to keep to the straight-and-narrow." Since the benefits of these interventions are tightly focused, and the benefits of nonintervention are widespread and dispersed, caution dictates skepticism when facing advocates of intervention.
Note that most of the arguments Allen and Gale adapt are not particular to the financial industry. The problems of excessive risk taking apply to all limited liability corporations. Many of the models they employ were originally developed for examin- ing manufacturing firms and industrial organization (spatial competition and patent races, for example). Their message seems to be that these models should make no more of a case for intervention in the financial industry than in other industries of the economy.
The exception, the one argument for intervention which is distinctively focused on financial intermediation, is the fear of contagion. It is also by far the most
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:16:11 PM All use subject to JSTOR Terms and Conditions
484 : MONEY, CREDIT, AND BANKING
important argument in the arsenal, the justification that financial regulators consider most significant and the one that economists must take most seriously.
The issue is not new. "Financial contagion" is merely the latest version of a traditional theme in the literature on intermediation. Financial intermediaries pro- vide so many different, nebulous services that it is a recurring rhetorical device to say, "Banks are on the one hand focused on doing A, but meanwhile they also (possibly inadvertently) do B." In various incarnations of the argument, a variety of activities have played the role of "A" and "B." By providing credit, banks inadvertently expand the money supply. By cutting back on their lending activi- ties, banks inadvertently cause feedbacks throughout the macroeconomy. The most recent twists go through the payments system and liquidity channels: "Contagion" and "financial fragility" are just the latest buzzwords. (I too am guilty of packing old wine in new skins; see Kahn and Santos, 2003).
So what to do? As I said Allen and Gale are too wise and too honest to provide an easy answer. I am either less honest or more foolhardy. Is there a tradeoff between financial efficiency and financial stability? Perhaps. Is financial instability expensive? Probably, although not all of the things Allen and Gale call "instability" are equally expensive. But inefficiency is always with us, and instability is (by definition, I suppose) intermittent. Rather than sacrifice efficiency, it is better be prepared to mitigate crises when they do arise.
Indeed, it is hopeless to try to approach regulation any other way: many times the "benefits" from instability are in the form of discipline on agents in stable times. But such discipline is a myth if crises are handled badly. If in the midst of a crisis the regulator bails out the directors and shareholders as well as the depositors, then allowing the crisis is not efficient after all.
Try as it might, a regulator can no more commit not to intervene in an LTCM or September 11th crisis than it could commit to its own abolition. Crisis interventions can be carried out well or badly, but some sort of intervention is inevitable. The best that can be hoped for is that the regulator has plans and procedures in place to serve, in effect, as the template-the commitment device-for the techniques to be used in that intervention.
It is important to rethink several aspects of banking regulation in this light: deposit insurance, closure policy, and payment system regulations. Regulatory policies can have long-run effects on a financial system's efficiency and stability. But the effect of announced policies will inevitably be filtered through participants' perceptions of their credibility. To do so requires modeling the activities of a regulator with its own bureaucratic objectives. Relevant studies of financial regulation are Boot and Thakor (1993), Campbell, Chan, and Marino (1992), Goodhart et al. (1998), Kahn and Santos (2001), Kane (1990), Mailath and Mester (1994), and Repullo (2000); more generally, see Alesina and Tabellini (2003) and the literature on political business cycles.
This perspective also provides renewed significance for a once-common type of study within economics: the regulatory history. Examinations of regulators' re- sponses to particular crises can help us understand the effects and limitations of
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:16:11 PM All use subject to JSTOR Terms and Conditions
DISCUSSION : 485
having particular frameworks in place. (The classic example of this sort of study is of course Friedman and Schwartz, 1971. For other examples relevant to central
banking, see Fleming and Garbade, 2002, Lowenstein, 2001, McAndrews and Potter, 2002, Meltzer, 2002, and Shapiro, 1980).
3. CONCLUSION
While Allen and Gale have refused to take an explicit stand on whether competi- tion and financial stability are conflicting objectives, they have, by balancing their
arguments so finely, implicitly argued against competition-reducing interventions. But the most powerful argument in favor of intervention is the one they treat most
briefly: the fear that financial contagion will cause systemic crisis. If we wish to
prevent regulators from fighting financial crises by reducing competition in normal
times, we need to ensure that they have in place effective procedures to mitigate crises when they do arise.
LITERATURE CIT'ED
Alesina, Alberto, and Guido Tabellini (2003). "Bureaucrats or Politicians?" Working Paper No. 238, Innocenzo Gasparini Institute for Economic Research (June 2003).
Allen, Franklin, and Douglas Gale (2000). Comparing Financial Systems. Cambridge, MA: MIT Press.
Allen, Franklin, and Douglas Gale (2003). "Financial Intermediaries and Markets." Working Paper No. 00-44-C, Wharton Financial Institutions Center, Econometrica, forthcoming.
Allen, Franklin, and Douglas Gale (2004). "Competition and Financial Stability." Journal of Money, Credit, and Banking 36, 453-480. (this issue of JMCB)
Boot, Amoud W.A., and Anjan V. Thakor (1993). "Self-interested Bank Regulation."American Economic Review 83, 206-212.
Campbell, T.S., Y.S. Chan, and A.M. Marino (1992). "An Incentive-Based Theory of Bank Regulation." Journal of Financial Intermediation 2, 255-276.
Chang, Roberto, and Andres Velasco (2001). "A Model of Financial Crises in Emerging Markets." Quarterly Journal of Economics 116, 489-517.
Cooper, Russell, and Dean Corbae (2002). "Financial Collapse: A Lesson from the Great Depression." Journal of Economic Theory 107, 159-190.
Diamond, Douglas, and Raghuram Rajan (2001). "Liquidity Risk, Liquidity Creation and Financial Fragility: A Theory of Banking." Journal of Political Economy 109, 287-327.
Fleming, Michael J., and Kenneth D. Garbade (2002). "When the Back Office Moved to the Front Burner: Settlement Fails in the Treasury Market after 9/11." Federal Reserve Bank of New York Economic Policy Review 8, 35-57.
Friedman, Milton, and Anna Jacobson Schwartz (1971). Monetary History of the United States, 1867-1960. Princeton, NJ: Princeton University Press.
Goodhart, C., P. Hartmann, D. Llewellyn, L. Rojas-Suarez, and S. Weisbrod (1998). Financial Regulation: Why, How, and Where Now? New York: Routledge.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:16:11 PM All use subject to JSTOR Terms and Conditions
486 : MONEY, CREDIT, AND BANKING
Kahn, Charles M., and Joao A.C. Santos (2001). "Allocating Bank Regulatory Powers: Lender of Last Resort, Deposit Insurance and Supervision." Federal Reserve Bank of New York Working Paper (November 2001).
Kahn, Charles M., and Joao A.C. Santos (2003). "Endogenous Financial Systems and Prudential Regulation." Federal Reserve Bank of New York (June 2003), unpublished manuscript.
Kane, E.J. (1990). "Principal-Agent Problems in S&L Salvage." Journal of Finance 45, 755-764.
Lowenstein, Roger (2001). When Genius Failed: The Rise and Fall of Long-Term Capital Management. New York: Random House.
Mailath, G.J., and L.J. Mester (1994). "A Positive Analysis of Bank Closure." Journal of Financial Intermediation 3, 272-299.
McAndrews, James J., and Simon M. Potter (2002). "Liquidity Effects of the Events of September 11,2001." Federal Reserve Bank of New York Economic Policy Review 8, 59-79.
Meltzer, Allan H. (2002). A History of the Federal Reserve: 1913-1951, Vol. 1. Chicago: University of Chicago Press.
Prescott, E., and R. Townsend (1984). "Pareto Optima and Competitive Equilibria with Adverse Selection and Moral Hazard." Econometrica 52, 21-45.
Repullo, R. (2000). "Who Should Act as a Lender of Last Resort? An Incomplete Contracts Model." Journal of Money Credit and Banking 32, 580-605.
Shapiro, Max (1980). The Penniless Billionaires. New York: Times Books.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:16:11 PM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. [481]
- p. 482
- p. 483
- p. 484
- p. 485
- p. 486
- Issue Table of Contents
- Journal of Money, Credit and Banking, Vol. 36, No. 3, Part 2: Bank Concentration and Competition: An Evolution in the Making A Conference Sponsored by the Federal Reserve Bank of Cleveland May 21-23, 2003 (Jun., 2004), pp. 433-654
- Front Matter
- Bank Concentration and Competition: An Evolution in the Making [pp. 433-451]
- Competition and Financial Stability [pp. 453-480]
- Comment on "Competition and Financial Stability" by Franklin Allen and Douglas Gale [pp. 481-486]
- Crises in Competitive versus Monopolistic Banking Systems [pp. 487-506]
- Comment on "Crises in Competitive versus Monopolistic Banking Systems" by John H. Boyd, Gianni De Nicoló, and Bruce D. Smith [pp. 507-509]
- How Foreign Participation and Market Concentration Impact Bank Spreads: Evidence from Latin America [pp. 511-537]
- Comment on "How Foreign Participation and Market Concentration Impact Bank Spreads: Evidence from Latin America" by Maria Soledad Martinez Peria and Ashoka Mody [pp. 539-542]
- Real Effects of Bank Competition [pp. 543-558]
- Comment on "Real Effects of Bank Competition" by Nicola Cetorelli [pp. 559-562]
- What Drives Bank Competition? Some International Evidence [pp. 563-583]
- Comment on "What Drives Bank Competition? Some International Evidence" by Stijn Claessens and Luc Laeven [pp. 585-592]
- Regulations, Market Structure, Institutions, and the Cost of Financial Intermediation [pp. 593-622]
- Comment on "Regulations, Market Structure, Institutions, and the Cost of Financial Intermediation" by Asli Demirgüç-Kunt, Luc Laeven, and Ross Levine [pp. 623-626]
- Bank Competition and Access to Finance: International Evidence [pp. 627-648]
- Comment on "Bank Competition and Access to Finance: International Evidence" by Thorsten Beck, Asli Demirgüç-Kunt, and Vojislav Maksimovic [pp. 649-654]
- Back Matter
Defining Financial Stability.pdf
WP/04/187
Defining Financial Stability
Garry J. Schinasi
© 2004 International Monetary Fund WP/04/187
IMF Working Paper
International Capital Markets Department
Defining Financial Stability1
Prepared by Garry J. Schinasi
October 2004
Abstract
This Working Paper should not be reported as representing the views of the IMF. The views expressed in this Working Paper are those of the author(s) and do not necessarily represent those of the IMF or IMF policy. Working Papers describe research in progress by the author(s) and are published to elicit comments and to further debate.
The main objective of this paper is to propose a definition of financial stability that has some practical and operational relevance. Financial stability is defined in terms of its ability to facilitate and enhance economic processes, manage risks, and absorb shocks. Moreover, financial stability is considered a continuum: changeable over time and consistent with multiple combinations of the constituent elements of finance. The paper also discusses several practical implications of the definition that should be considered when using it for policy analysis or developing an analytical framework. JEL Classification Numbers: E60, G00, H00 Keywords: Finance, stability, fragility, crises Author(s) E-Mail Address: [email protected]
1 This paper was written while on sabbatical from the IMF and is part of a manuscript on financial stability issues. I gratefully acknowledge the IMF’s financial support under its Independent Study Leave Program. I also gratefully acknowledge the support and encouragement of De Nederlandsche Bank (DNB) and the European Central Bank (ECB) while visiting them in 2003 and 2004, especially Tommaso Padoa-Schioppa, Mauro Grande, and John Fell at the ECB and Henk Brouwer, Jan Brockmeijer, Aerdt Houben, and Jan Kakes at the DNB. I am grateful to Tommaso Padoa-Schioppa, John Fell, Mauro Grande, Aerdt Houben, Jan Kakes, and Jukka Vesala for extensive discussions on this topic and for comments on earlier drafts of this paper.
- 2 -
Table of Contents
Page
I. Introduction....................................................................................................................3
II. Prior Concepts of Finance and Finance’s Strengths and Weaknesses ...........................4
III. Key Principles for Defining Financial Stability.............................................................6
IV. Definition of Financial System Stability........................................................................8
V. Some Practical Implications of the Definition.............................................................11
Annex: Alternative Definitions of Financial Stability .............................................................13 References................................................................................................................................17
- 3 -
I. INTRODUCTION
Does financial stability require the soundness of institutions, the stability of markets, the absence of turbulence, low volatility, or something more fundamental? Can it be achieved and maintained through individual private actions and unfettered market forces alone? If not, what is the role of the public sector in fostering financial stability, as opposed to private- collective action: is it just to make way for the private sector to achieve an optimum on its own, or is a more proactive role necessary for achieving the full private and social benefits of finance? Is there a consensus on how to achieve and maintain financial stability?
The last three questions are not likely to have clear answers without a useful answer to the first question. Likewise, without a good working definition, the growing financial stability profession will continue to find it difficult to develop useful analytical frameworks for examining policy issues. Unfortunately, there is no single, widely accepted and used definition of financial stability. There have been recent attempts to define financial stability, but most of them seem to fit into a particular theme of a paper or speech. In addition, most authors prefer to define financial instability or systemic risk (see the attached Annex starting on page 13).
The approach taken here is to define financial stability rather than its absence, in part because this is likely to be the more useful “policy” objective. A policy objective of avoiding financial instability or crisis—or of managing systemic risk—could bias policy decisions, analyses, and analytical frameworks towards sacrificing both private and social benefits of finance. A more positive or constructive approach—such as the one proposed in this paper— may serve additional practical purposes, including leaving open the possibility of assessing whether the private and social benefits of finance can be increased further. This would be particularly useful in countries that have relatively undeveloped financial systems.
As anyone who has tried to define financial stability knows, there is as yet no widely accepted model or analytical framework for assessing financial system stability and for examining policies as there is for economic systems and in other disciplines.2 This is because the analysis of financial stability is still in its infant stage of development and practice, as compared with—for example—the analysis of monetary and/or macroeconomic stability. In the rare cases in which financial systems are expressed rigorously, they constitute one or two equations in a much larger macroeconomic model possessing most of the usual macro- equilibrium and macro-stability conditions. In addition, there are reasons to believe that a single target variable cannot be found for defining and achieving financial stability—as there is believed to be for defining and achieving monetary stability—although many doubt that a single target variable approach accurately represents actual practice in monetary policymaking.
2 See the paper by Houben, Kakes, and Schinasi (2004), which proposes a framework for financial stability (and also draws on concepts developed here). The IMF’s bilateral and multilateral financial market and system surveillance, and the IMF’s and World Bank’s Financial Sector Assessment Program are also making progress in this direction.
- 4 -
Lacking a framework, a set of models, or even a concept of equilibrium, it is difficult to envision a definition of financial stability akin to that which economists normally demand and use. Nevertheless, it would be useful to have one that allows for the development of policy frameworks and analytical tools. The definition proposed in this paper is one step in this direction, and is offered for wider debate.
The paper is organized as follows: Section II briefly presents some prior concepts of finance and its strengths (benefits) and weaknesses (fragilities), drawing on analysis in a companion study. These concepts serve as both practical and analytical focal points for developing a concept of financial stability in the absence of a widely accepted concept of equilibrium and analytical framework. For simplicity, Section III identifies five principles that a useful definition could encompass. It also makes a case for seeing financial stability as occurring along a changeable continuum or range of conditions of the constituent parts of the financial system, as opposed to a single configuration or state of these parts as is most often used in microeconomic and macroeconomic models. Section IV proposes a broad definition and discusses the meaning of some of its language. The final section identifies several practical implications of the definition that should be carried over into any policy or analytical framework that utilizes it.
II. PRIOR CONCEPTS OF FINANCE AND FINANCE’S STRENGTHS AND WEAKNESSES
Before developing a working definition of financial stability, it would be useful to consider the following understandings as prerequisites or as relevant concepts and ideas.3
First, a barter economy is less effective and efficient in allocating scarce resources than is an economy with the ability to use financial claims on future real resources. A discussion of financial stability must necessarily take place within the context of a monetary economy in which there exists a money (now usually fiat money) that is universally accepted as the economy’s unit of account and means of payment.
Second, money is not necessarily the most desirable store of value—except in the very short run or during episodes of financial distress and dysfunctions. Throughout recorded history, human ingenuity has driven an evolutionary process of finance to overcome this persistent deficiency. Modern finance provides substitutes for money that provide temporary and reversible intertemporal means-of-payment and store-of-value services. These substitutes are promises to pay money in the future and are designed in part to facilitate intertemporal resource allocations.
Third, many of the services provided by money and finance are both private and public goods. They are private goods in providing benefits to individuals in their private affairs, benefits that convey only to the counterparts engaged in specific transactions. They separately and jointly provide public goods as well, because they allow multilateral trade and exchange to be more efficient, in part by eliminating the need for Jevons’ “double coincidence of wants,” both sectorally at moments in time and intertemporally. In addition, 3 See Schinasi (2004) for a more detailed discussion and analysis of many of these points.
- 5 -
finance provides public goods beyond those of fiat money: by enhancing and distributing the public-good characteristics of fiat money, finance enlarges society’s opportunities for—and efficiency in—intertemporal economic processes such as trade, production, wealth accumulation, economic development and growth, and ultimately social prosperity. In sum, the universal acceptability of money and the existence of an effective process of finance together create an environment that provides collective benefits to all members of society.
Fourth, an alternative and useful way of seeing finance is to bring to the surface one of its defining characteristics. Unlike fiat money—which eliminates the element of human trust in trade and exchange—finance involves human promises to pay back specific amounts of fiat money in the future. In this way, finance existentially embodies uncertainty (about human trust). Modern financial systems have evolved to provide beneficial and necessarily imperfect ways of transforming this fundamental uncertainty into quantifiable and “priceable” risks, such as default risk, and, through social arrangements (both markets and financial institutions), also market risk, liquidity risk, and so on. In less traditional but no less appropriate terms, modern finance provides societies with effective, albeit imperfect, mechanisms for transforming, pricing, and allocating economic and financial uncertainties and risks.
Finally, because finance existentially embodies uncertainty, there are both potential benefits and costs associated with it.4 On the one hand, finance enhances the private and social benefits of fiat money: in part by enlarging the pool of liquidity available for production, consumption, and exchange; and in part by facilitating and enhancing the efficiency of intertemporal economic processes. In effect, the willingness to engage in finance (i.e., to take the leap of faith) and accept the uncertainty of trust has created social welfare gains far beyond what fiat money alone could provide.
On the other hand, trust is fragile: it can, and often enough does, become a source of potential financial instability, which in the wrong circumstances can affect both individual and social welfare. To the extent that doubts about human trust are transformed by the financial system into market and other financial risks, they too can become companion sources of instability—even more so if a society’s financial-market mechanisms are impaired and unable to effectively reallocate and price such doubts. How such doubts propagate through the financial system is an important determinant of whether they either self-correct and remain isolated and harmless or become widespread, harmful, and perhaps even systemic. Because finance supports and facilitates real economic processes, these potential instabilities may well extend to the real economy.
4 Diamond and Dybvig (1983) and Diamond and Rajan (2000) explore this in the context of bank intermediation.
- 6 -
III. KEY PRINCIPLES FOR DEFINING FINANCIAL STABILITY
While there is scope for being more comprehensive and inclusive, a small number of key principles can be identified for developing a working definition of financial stability.5 One that requires more elaboration than the others is that it is useful to consider financial stability as occurring along a continuum—rather than as a static condition.
The first principle is that financial stability is a broad concept, encompassing the different aspects of finance (and the financial system)—infrastructure, institutions, and markets. Both private and public persons participate in markets and in vital components of the financial infrastructure (including the legal system and official frameworks for financial regulation, supervision, and surveillance). Governments borrow in markets, hedge risks, operate through markets to conduct monetary policy and maintain monetary stability, and own and operate payments and settlement systems. Accordingly, the term “financial system” can be seen as encompassing both the monetary system with its official understandings, agreements, conventions, and institutions as well as the processes, institutions, and conventions of private financial activities.6 Given the tight interlinkages between all of these components of the financial system, (expectations of) disturbances in any of the individual components can undermine the overall stability, requiring a systemic perspective. At any given time, stability or instability could be the result of either private institutions and actions, or official institutions and actions, or both simultaneously and/or iteratively.
A second useful principle is that financial stability not only implies that finance adequately fulfills its role in allocating resources and risks, mobilizing savings, and facilitating wealth accumulation, development, and growth; it should also imply that the systems of payment throughout the economy function smoothly (across official and private, retail and wholesale, and formal and informal payments mechanisms). This requires that fiat (or central bank) money—and its close-substitute, derivative monies (such as demand deposits and other bank accounts)—can adequately fulfill its role as the universally accepted means of payment and unit of account and, when appropriate, as a (short-term) store of value. In other words, financial stability and what is usually regarded as a vital part of monetary stability overlap to a large extent.
A third principle is that the concept of financial stability relates not only to the absence of actual financial crises but also to the ability of the financial system to limit, contain, and deal with the emergence of imbalances before they constitute a threat to itself or economic processes. In a well-functioning and stable financial system, this occurs in part through self-corrective, market-disciplining mechanisms that create resilience and prevent
5 There are also many prerequisites for establishing a sound and stable financial system, such as: macroeconomic stability and a policy framework for maintaining it; an adequate—if not effective—framework for financial regulation, supervision, and surveillance (implicitly mentioned in the text as infrastructure; well- established codes, standards, and business practices), and more generally private incentive structures; and an enforceable legal system that supports productive private financial contracts.
6 This is adapted from the definition of “international financial system” in Truman (2003).
- 7 -
problems from festering and growing into system-wide risks. In this respect, there may be a policy-related trade-off entailing the choice between allowing market mechanisms to work to resolve potential difficulties and intervening quickly and effectively—through liquidity injections via markets, for example—to restore risk-taking and/or to restore stability. Thus, financial stability entails both preventive and remedial dimensions.
A fourth important principle is that financial stability be couched in terms of the potential consequences for the real economy. Disturbances in financial markets or at individual financial institutions need not be considered threats to financial stability if they are not expected to damage economic activity at large. In fact, the incidental closing of a financial institution, a rise in asset-price volatility, and sharp and even turbulent corrections in financial markets may be the result of competitive forces, the efficient incorporation of new information, and the economic system’s self-correcting and self-disciplining mechanisms. By implication, in the absence of contagion and the high likelihood of systemic effects, such developments may be viewed as welcome—if not healthy—from a financial stability perspective.
A fifth principle—consistent with those already discussed and the actual dynamism of finance—is that financial stability be thought of as occurring along a continuum. An example that is more transparent is the health of an organism, which also occurs along a continuum. A healthy organism can usually reach for a greater level of health and well being, and the range of what is normal is broad and multi-dimensional. In addition, not all states of un- health (or illness) are significant, systemic, or life threatening. And some illnesses, even temporarily serious ones, allow the organism to continue to function productively and can have a cleansing effect, leading to greater health. One implication of seeing financial stability in this way is that maintaining financial stability does not necessarily require that each part of the financial system operate persistently at peak performance; it is consistent with the financial system operating on a “spare tire” from time to time.7
The concept of a continuum is relevant because finance fundamentally involves uncertainty, is dynamic (meaning both inter-temporal and innovative), and is composed of many interlinked and evolutionary elements (infrastructure, institutions, markets). Accordingly, financial stability is expectations-based, dynamic, and dependent on many parts of the system working reasonably well. What might represent stability at one point in time, might be more stable or less stable at some other time, depending on other aspects of the economic system—such as technological, political, and social developments. Moreover, financial stability can been seen as being consistent with various combinations of the conditions of its constituent parts, such as the soundness of financial institutions, financial markets conditions, and effectiveness of the various components of the financial infrastructure.
7 See Greenspan (1999).
- 8 -
IV. DEFINITION OF FINANCIAL SYSTEM STABILITY
Broadly, financial stability can be thought of in terms of the financial system’s ability: (a) to facilitate both an efficient allocation of economic resources—both spatially and especially intertemporally—and the effectiveness of other economic processes (such as wealth accumulation, economic growth, and ultimately social prosperity); (b) to assess, price, allocate, and manage financial risks; and (c) to maintain its ability to perform these key functions—even when affected by external shocks or by a build up of imbalances—primarily through self-corrective mechanisms.
A definition consistent with this broad view is as follows:
A financial system is in a range of stability whenever it is capable of facilitating (rather than impeding) the performance of an economy, and of dissipating financial imbalances that arise endogenously or as a result of significant adverse and unanticipated events.
The meanings of several phrases need to be explained.
First, the concept of a range of stability represents the concept of a continuum as a key building block. The continuum for financial stability can be thought of as multidimensional and occurring across a multitude of observable and measurable variables. The set of variables should encompass a subset that tries to quantify, albeit imperfectly, how well finance is facilitating economic and financial processes such as savings and investment, lending and borrowing, liquidity creation and distribution, asset pricing, and ultimately wealth accumulation and growth.
As a continuum, financial stability can be seen practically as somewhat broader and less precise than the ability to return to a single and sustainable position or time path after a shock or perturbation, as with other (Newtonian) concepts of equilibrium and stability in some disciplines (including economics). The proposed definition is consistent with a financial system being in a perpetual state of flux and transformation while its ability to perform its key functions remains well within a set of tolerable boundaries—defined over a set of measurable variables—that are consistent with it successfully playing its important facilitative and efficiency-enhancing roles. Observable states approaching these boundaries would indicate that the financial system is losing some of its ability to perform; observations outside these boundaries would indicate that the system is no longer effectively facilitating economic processes, perhaps because aggregate production is substantially below its potential on account of funds not being channeled to profitable activities, risks not being managed, and shocks not being absorbed. In such cases, remedial action would be required, which in the extreme would mean crisis resolution and restoration.
To illustrate the multidimensional nature of the definition, consider a very simplistic two-dimensional example. In assessing the joint stability of financial markets and financial institutions, one might be able to identify combinations of interest rate spread volatility (as a possible market source of instability) and banking system capital (as an institutional source of shock-absorptive capacity) that are consistent with the financial system continuing
- 9 -
effectively to facilitate efficient resource allocation. Likewise, other combinations could be identified that would not be consistent with stability. The former would constitute the range of stability and the latter would fall outside this range.
A more comprehensive sets of factors could be envisioned for determining a grid over which a continuum is defined. Statistical tools could be utilized to select such factors by considering historical episodes of both stability and instability, in part by using forward- looking, market-determined expectations of future outcomes and matching them with actual outcomes. This methodology could, in principle, also help to establish estimates of boundaries or zones separating stability from potential instability.8
A second phrase that needs some explanation is facilitating (rather than impeding) the performance of an economy. This phrase means, among other things, that finance is contributing to (rather than impeding) the efficient allocation of real resources, the rate of growth of output, and the processes of saving, investment, and wealth creation—and may also entail and include other observable and measurable aspects of economic performance.
Third, the term, “dissipate financial imbalances,” means a movement along the continuum in the direction of stability (away from boundaries) and not instability, for example in asset prices and portfolio flows, implicitly through self-corrective mechanisms. Such adjustments would include the exit and entry of market participants (financial institutions or nonfinancial entities acting on behalf of others or individuals acting directly in the markets).
There are other aspects of the definition worth noting. The proposed definition leaves open the possibility that the financial system could become capable of impeding the performance of the economy endogenously, even in the absence of unanticipated events (shocks), for example through the accumulation of imbalances caused by asset mispricing and/or other market “imperfections.” This is consistent with the ample historical evidence that financial systems, particularly banking systems, are prone to the build up of imbalances (credit-risk concentrations or illiquidity, for example) and even instability. Banks internalize the fragilities associated with the properties of liquidity, and are therefore prone to instability themselves.9 Banks, other financial institutions, and even markets can be seen as social arrangements—or as clearing houses—for assessing, pricing, and trading human promises necessarily involving uncertainty and risk, including uncertainty about the fundamental element of trust in financial contracts. Social arrangements and institutional features of economic systems try to internalize the potential adverse consequences of negative externalities associated with the frailties of human trust. A tangible example of this is that banks internalize the potential adverse consequences of failures of trust by economizing on
8 In principle, this approach could be generalized and made amenable to theoretical and empirical model building. For example, one could define a set of n variables that encompass all relevant measures of aspects of financial stability. The range of stability could be defined as a subset of n-tuples bounded by n functions (most likely nonlinear) defining the limits of stability in terms of n variables.
9 See Diamond and Rajan (2000 and 2002).
- 10 -
information about large pools of debtors and their ability to pay future claims or promissory notes. In internalizing these elements of financial risk and uncertainty, financial institutions and markets themselves embody the potential for financial fragility, which ultimately finds its source in a failure of human trust in some meaningful way (for example, a default).
The definition also presupposes that there are aspects of finance that embody either negative or positive externalities. In this sense, improvements in the ability of finance to facilitate rather than impede economic processes—including providing greater financial stability—is welfare improving in terms of enhancing the efficiency of resource allocation (and pricing), especially inter-temporally.10 Some points along the continuum of financial stability are more welfare-improving (and efficiency-enhancing) than others, and some points along the continuum of instability are to be avoided, seemingly at all costs.11 Thus, in moving from a condition of stability to instability, the contribution of the financial system to aggregate economic welfare is being reduced.
A stable financial system is one that enhances economic performance in many dimensions, whereas an unstable financial system is one that detracts from economic performance. In this sense the definition is “normative.” Ultimately—and unlike physical instabilities such as earthquakes, floods, and sunspots—financial instability can be dealt with through massive intervention by authorities, including redefining the rules of the market place. But these measures would be “last resort” reforms to prevent the economic system from collapsing—as, for example, during the world-wide depression in the 1930s and more recently in Asia during 1997–98.
To illustrate the broad nature of this definition of financial stability, two corollary definitions are useful:
(i) A financial system is entering a range of instability whenever it is threatening to impede the performance of an economy.
(ii) A financial system is in a range of instability when it is impeding performance and threatening to continue to do so.
A more general definition that does not require the specification of what constitutes a “financial system” is:
Financial stability is a condition in which an economy’s mechanisms for pricing, allocating, and managing financial risks (credit, liquidity, counterparty, market, etc.) are functioning well enough to contribute to the performance of the economy (as defined above).
10 See Schinasi (2004).
11 There would seem to be a trade-off in financial systems between financial stability and efficiency, but this is difficult to analyze given that there are different concepts of both stability and efficiency. There is some work on this in the theoretical banking literature, but none could be found at the financial-system level.
- 11 -
V. SOME PRACTICAL IMPLICATIONS OF THE DEFINITION
Looking at financial stability in this way allows for the delineation of financial conditions and potential difficulties according to their intensity, scope, and potential threat to systemic stability. One could, for example, think of potential financial difficulties as falling into one of the following fairly broad categories:
• difficulties in a single institution or market not likely to have system-wide consequences for either the banking or financial system;
• difficulties that involve several relatively important institutions involved in market activities with some nontrivial probability of spillovers and contagion to other institutions and markets; and
• problems likely to spread to a significant number and types of financial institutions and across usually unrelated markets for managing liquidity needs, such as forward, interbank, and even equity markets.
Problems occurring within each of these categories would require different diagnostic tools and policy responses, ranging from doing nothing to intensifying supervision or surveillance of a specific institution or market, to liquidity injections into the markets to dissipate strains, to interventions into particular institutions.
The financial stability definition also involves several complexities that have practical significance in terms of assessing risks to the well-functioning of the financial system and the contribution public policy can make to ensuring financial stability.
• Developments in financial stability cannot be summarized in a single quantitative indicator. In contrast with price stability, for instance, there is as yet no unequivocal unit of measurement for financial stability.12 This reflects the multifaceted nature of financial stability as it relates to both the stability and resilience of financial institutions, and to the smooth functioning of financial markets and settlement systems. Moreover, diverse factors need to be weighed in terms of their potential ultimate influence on real economic activity.
• Developments in financial stability are inherently difficult to forecast. Assessing the state of financial stability should not only take stock of disturbances as they emerge, but also indicate the risks and vulnerabilities that could lead to such disturbances occurring in the future. A forward-looking approach is therefore needed in order to establish the build-up of risks and imbalances and to take account of the transmission lags in policy instruments. The challenge is that financial crises are inherently difficult to predict because of many factors, for example, contagion effects and nonlinearities in the relationships between the constituent parts of finance. In
12 See Andy Haldane (2004) for an attempt to set out how this might be done for financial stability.
- 12 -
addition, risks to financial stability often reflect the far-reaching consequences of unlikely events. This implies that the focus of attention should not be the mean, median, or mode of possible outcomes or states but the entire distribution of them, and particularly the left “tail.” Beyond this, the distribution of possible outcomes may be subject to greater fundamental uncertainty (in the sense of Knight, 1921) than traditional macroeconomic projections, reflecting lack of knowledge about the actual shape of the probability distribution. This would imply that forecasts of financial stability might be inherently less reliable than forecasts of monetary or macroeconomic stability, for which there are well-worked and more reliable models and more timely and useful data. Thus, in the large sets of financial indicators that are now being used by central banks and international financial institutions, the relationships between indicators and financial stability conditions may not be strong or robust enough to be reliable for assessments and prediction. Nevertheless, in looking at a broader array of indicators, in developing better frameworks, and in utilizing sophisticated statistical tools, there may be scope for improving the ability to monitor and assess financial stability in the future.
• Developments in financial stability are only partly controllable. The policy instruments that can be used to safeguard financial stability generally have other primary objectives, such as protecting the interests of deposit holders (in the case of prudential instruments), fostering price stability (in the case of monetary policy), or promoting a swift settlement of financial transactions (in the case of policies governing payment and settlement systems). Besides timing lags, the impact of these policy instruments on financial stability is thus often indirect; in some cases there may even be friction with the instrument’s initial objective. Moreover, developments in financial stability are highly susceptible to exogenous shocks—ranging from natural catastrophes to abrupt swings in market sentiment—further limiting their controllability.
• Policies aimed at financial stability often involve a trade-off between resilience and efficiency. Measures to enhance financial stability often involve weighing the pursuit of an efficient allocation of financial resources against the ability to exclude or absorb shocks to the financial system. This implies a risk/return judgment that is difficult to make in a fully objective manner. For instance, in the sphere of prudential policies, higher solvency requirements will reduce the risk of a bank not being able to absorb an adverse shock but will also imply capital costs and foregone investment opportunities. Similarly, exchange restrictions may reduce or exclude certain risks related to international capital flows but may also limit the efficiency of the domestic financial market.
• Policy requirements for financial stability may be time inconsistent. Since the use of some public policy instruments to safeguard financial stability circumvents market forces, the short-term stability gain may come at the cost of a longer-term stability loss. In particular, measures such as the provision of lender-of-last-resort finance or deposit guarantee may undermine market discipline, thereby creating moral hazard or adverse selection. This intertemporal trade off is a fundamental issue in financial system policymaking.
- 13 - ANNEX
ALTERNATIVE DEFINITIONS OF FINANCIAL STABILITY
This appendix provides an overview of definitions or descriptions of financial stability by a selected group of officials, central banks, and academics.13
John Chant and others (Bank of Canada)14
“Financial instability refers to conditions in financial markets that harm, or threaten to harm, an economy’s performance through their impact on the working of the financial system.... Such instability harms the working of the economy in various ways. It can impair the financial condition of non-financial units such as households, enterprises, and governments to the degree that the flow of finance to them becomes restricted. It can also disrupt the operations of particular financial institutions and markets so that they are less able to continue financing the rest of the economy.... It differs from time to time and from place to place according to its initiating impulse, the parts of the financial system affected, and its consequences. Threats to financial stability have come from such diverse sources as the default on the bonds of a distant government; the insolvency of a small, specialized, foreign exchange bank; computer breakdown at a major bank; and the lending activities of a little- known bank in the U.S. Midwest (pp. 3–4).
Andrew Crockett (Bank for International Settlements and Financial Stability Forum)15
“...define financial stability as an absence of instability...a situation in which economic performance is potentially impaired by fluctuations in the price of financial assets or by an inability of financial institutions to meet their contractual obligations. I would like to focus on four aspects of this definition.
“Firstly, there should be real economic costs.... Secondly, it is the potential for damage rather than actual damage which matters.... Thirdly, my definition refers...not just to banks but to nonbanks, and to markets as well as to institutions.... Fourth, my definition allows me to address the question of whether banks are special...all institutions that have large exposures—all institutions that are largely interconnected whether or not they are themselves directly involved in the payments system—have the capacity, if they fail, to cause much widespread damage in the system.”
13 Some authors choose not to define financial stability and instead use the concept of systemic risk. See Oosterloo and Haan (2003) for a discussion of this concept.
14 See Chant (2003).
15 See Crockett (1997).
- 14 - ANNEX
Deutsche Bundesbank16
“The term financial stability broadly describes a steady state in which the financial system efficiently performs its key economic functions, such as allocating resources and spreading risk as well as settling payments, and is able to do so even in the event of shocks, stress situations, and periods of profound structural change.”
Wim Duisenberg (European Central Bank)17
“...monetary stability is defined as stability in the general level of prices, or as an absence of inflation or deflation. Financial stability does not have as easy or universally accepted a definition. Nevertheless, there seems to be a broad consensus that financial stability refers to the smooth functioning of the key elements that make up the financial system.”
Roger Ferguson (Board of Governors of the U.S. Federal Reserve System)18
“It seems useful...to define financial stability...by defining its opposite: financial instability. In my view, the most useful concept of financial instability for central banks and other authorities involves some notion of market failure or externalities that can potentially impinge on real economic activity.
“Thus, for the purposes of this paper, I’ll define financial instability as a situation characterized by these three basic criteria: (i) some important set of financial asset prices seem to have diverged sharply from fundamentals; and/or (ii) market functioning and credit availability, domestically and perhaps internationally, have been significantly distorted; with the result that (iii) aggregate spending deviates (or is likely to deviate) significantly, either above or below, from the economy’s ability to produce.
Michael Foot (U.K. Financial Services Authority)19
“...we have financial stability where there is: (a) monetary stability; (b) employment levels close to the economy’s natural rate; (c) confidence in the operation of the generality of key financial institutions and markets in the economy; and (d) where there are no relative price movements of either real or financial assets within the economy that will undermine (a) or (b).
“The first three elements of this definition are, I hope, noncontentious. In respect of (a) and (b), it seems implausible to define financial stability as occurring in a period of rapid inflation, or in a mid-1930s style period of low inflation but high unemployment.
16 Deutsche Bundesbank (2003).
17 See Duisenberg (2001).
18 See Ferguson (2003).
19 See Foot (2003).
- 15 - ANNEX
“Similarly in respect of (c), it would be strange to argue that there was financial stability in a period when banks were failing, or when normal conduits for long-term savings and borrowing in either the personal or corporate sectors were seriously malfunctioning. Such circumstances would mean the participants had lost confidence in financial intermediaries. It would mean, almost certainly, that economic growth was being damaged by the unavailability or relatively high cost of financial intermediation.
“This leaves us with (d).... I would say that there are four main channels by which changes in asset prices might affect the real economy: by changing household wealth and thereby consumption…by a change in equity prices...by their impact on firms’ balance sheets which can then affect corporate spending...[and] by their impact on capital flows, with for example inflows of capital—as during the dot.com boom in the US—strengthening the domestic currency.”
Sir Andrew Large20
“In a broad sense.....think of financial stability in terms of maintaining confidence in the financial system. Threats to that stability can come from shocks of one sort or another. These can spread through contagion, so that liquidity or the honoring of contracts becomes questioned. And symptoms of financial instability can include volatile and unpredictable changes in prices. Preventing this from happening is the real challenge.”
Frederick Mishkin (Columbia University)21
...Financial instability “occurs when shocks to the financial system interfere with information flow so that the financial system can no longer do its job of channeling funds to those with productive investment opportunities.”
Norges Bank22
“Financial stability means that the financial system is robust to disturbances in the economy, so that it is able to mediate financing, carry out payments, and redistribute risk in a satisfactory manner.”
Tommaso Padoa-Schioppa (European Central Bank)23
“...[financial stability is] a condition where the financial system is able to withstand shocks without giving way to cumulative processes, which impair the allocation of savings to investment opportunities and the processing of payments in the economy. 20 See Large (2003).
21 See Mishkin (1999).
22 See Norwegian Central Bank (2003).
23 See Padoa-Schioppa (2003).
- 16 - ANNEX
“The definition immediately raises the related question of defining the financial system...[which] consists of all financial intermediaries, organized and informal markets, payments and settlement circuits, technical infrastructures supporting financial activity, legal and regulatory provisions, and supervisory agencies. This definition permits a complete view of the ways in which savings are channeled towards investment opportunities, information is disseminated and processed, risk is shared among economic agents, and payments are facilitated across the economy.”
Anna Schwartz (National Bureau of Economic Research)24
“A financial crisis is fueled by fears that the means of payment will be unobtainable at any price and, in a fractional reserve banking system, leads to a scramble for high-powered money. It is precipitated by actions of the public that suddenly squeeze the reserves of the banking system.... The essence of a financial crisis is that it is short-lived, ending with a slackening of the public’s demand for additional currency.”
Nout Wellink (De Nederlandsche Bank)25
“According to our own definition at the Nederlandsche Bank, a stable financial system is capable of efficiently allocating resources and absorbing shocks, preventing these from having a disruptive effect on the real economy or on other financial systems. Also, the system itself should not be a source of shocks. Our definition thus implies that that money can properly carry out its functions as a means of payment and as a unit of account, while the financial system as a whole can adequately perform its role of mobilizing savings, diversifying risks, and allocating resources. Financial stability is a vital condition for economic growth, as most transactions in the real economy are settled through the financial system. The importance of financial stability is perhaps most visible in situations of financial instability. For example, banks may be reluctant to finance profitable projects, asset prices may deviate excessively from their underlying intrinsic values, or payments may not be settled in time. In extreme cases, financial instability may even lead to bank runs, hyperinflation, or a stock market crash.”
24 Schwartz (1986).
25 See Wellink (2002).
References
Chant, John, 2003, “Financial Stability As a Policy Goal,” in Essays on Financial Stability, by John Chant, Alexandra Lai, Mark Illing, and Fred Daniel, Bank of Canada Technical Report No. 95 (Ottawa: Bank of Canada), September, pp. 3–4.
Crockett, Andrew, 1997, “The Theory and Practice of Financial Stability,” GEI Newsletter Issue No. 6 (United Kingdom: Gonville and Caius College Cambridge), 11–12 July.
Davis, Philip, 2002, “A Typology of Financial Instability,” Financial Stability Report 2 (Vienna: Österreichische Nationalbibliothek).
Deutsche Bundesbank (2003), “Report on the Stability of the German Financial System,” Monthly Report, Frankfurt, December.
Diamond, Douglas W., and Raghuram G. Rajan, 2001, “Liquidity Risk, Liquidity Creation, and Financial Fragility: A Theory of Banking,” Journal of Political Economy, Vol. 109, 2, pp. 287–327.
——— , 2001, “Banks, Short Term Debt, and Financial Crises: Theory, Policy Implications, and Applications,” Proceedings of Carnegie Rochester Series on Public Policy (Amsterdam: Elsevier Science),Vol. 54, No. 1, pp. 37–71.
Duisenberg, Wim F., 2001, “The Contribution of the Euro to Financial Stability,” in Globalization of Financial Markets and Financial Stability—Challenges for Europe (Baden-Baden: Nomos Verlagsgesellschaft), pp. 37–51.
Ferguson, Roger, 2002, “Should Financial Stability Be An Explicit Central Bank Objective?” (Washington: Federal Reserve Board).
Foot, Michael, 2003, “What Is ‘Financial Stability’ and How Do We Get It?” The Roy Bridge Memorial Lecture (United Kingdom: Financial Services Authority), April 3.
Greenspan, Alan, 1999, “Do Efficient Markets Mitigate Financial Crises?” speech delivered before the 1999 Financial Markets Conference of the Federal Reserve Bank of Atlanta.
Haldane, Andrew, 2004, “Defining Monetary and Financial Stability” (unpublished; London: Bank of England), February.
Houben, Aerdt C.F.J., Jan Kakes, and Garry Schinasi, 2004, “Toward a Framework for Safeguarding Financial Stability,” IMF Working Paper 04/101 (Washington: International Monetary Fund) and DNB Occasional Paper (Amsterdam, Forthcoming).
Large, Sir Andrew, 2003, “Financial Stability: Maintaining Confidence in a Complex World,” in Financial Stability Review (London: Bank of England), December, pp. 170–74.
- 18 -
Mishkin, Frederick, 1999, “Global Financial Instability: Framework, Events, Issues,” Journal of Economic Perspectives, Vol. 13 (Fall), pp. 3–20.
Norwegian Central Bank, 2003, Financial Stability Review, February.
Padoa-Schioppa, Tomasso, 2003, “Central Banks and Financial Stability: Exploring the Land In Between,” in The Transformation of the European Financial System, ed. by Vitor Gaspar and others (Frankfurt: European Central Bank), pp. 269–310.
Schinasi, Garry J., 2003, “Responsibility of Central Banks for Stability in Financial Markets,” in Current Developments in Monetary and Financial Law—Volume 2 (Washington: International Monetary Fund), Chapter 17.
——— , 2004, “Private Finance and Public Policy,” IMF Working Paper 04/120 (Washington: International Monetary Fund).
Schwartz, Anna J., 1986, “Real and Pseudo-Financial Crises” in Financial Crises and the World Banking System, ed. by Forrest Capie and Geoffrey E. Woods (New York: St. Martin’s Press).
Truman, Edwin, 2003, Inflation Targeting in the World Economy (Washington: Institute for International Economics).
Wellink, Nout, 2002, “Current Issues in Central Banking”, (Oranjestad: Central Bank of Aruba), November 14.
Financial Regulation, Monetary Policy, and Inflation in the Industrialized World.pdf
Financial Regulation, Monetary Policy, and Inflation in the Industrialized World Author(s): Mark S. Copelovitch and David Andrew Singer Source: The Journal of Politics, Vol. 70, No. 3 (July 2008), pp. 663-680 Published by: The University of Chicago Press on behalf of the Southern Political Science Association Stable URL: http://www.jstor.org/stable/10.1017/S0022381608080687 .
Accessed: 06/02/2015 10:59
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
The University of Chicago Press and Southern Political Science Association are collaborating with JSTOR to digitize, preserve and extend access to The Journal of Politics.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
Financial Regulation, Monetary Policy, and Inflation in the Industrialized World
Mark S. Copelovitch University of Wisconsin – Madison
David Andrew Singer Massachusetts Institute of Technology
This article argues that the institutional mandates of central banks have an important influence on inflation outcomes in the advanced industrialized countries. Central banks that are also responsible for bank regulation will be more sensitive to the profitability and stability of the banking sector and therefore less likely to alter interest rates solely on the basis of price stability objectives. When bank regulation is assigned to a separate agency, the central bank is more likely to enact tighter monetary policies geared solely toward maintaining price stability. An econometric analysis of inflation in 23 industrial countries from 1975 to 1999 reveals that inflation is significantly higher in those countries with central banks that are vested with bank regulatory responsibility, although this effect is conditional on the choice of exchange rate regime and the relative size of the banking sector. We also conduct a case study of the Bank of England, which lost its bank regulatory authority to a new agency in 1998. We find that the new Labour government under Tony Blair imposed the institutional change on the Bank of England in part to remove the bank stability bias from its monetary policymaking. These findings suggest that the mandates of central banks not only have important influences on macroeconomic outcomes, but may also be modified in the future by governments seeking to impose their own monetary policy preferences.
I n today’s world of global capital markets, political leaders face two primary challenges in regulating their economies. First, leaders must ensure the
stability of their country’s financial institutions. The frequency and magnitude of banking crises have reached levels not seen since the interwar period (Bordo et al. 2000). The United States, for example, experienced a dramatic bout of banking instability during the 1980s, with more banks collapsing in 1985–87 than in the prior 30 years (FDIC 1998). The United Kingdom faced a similar crisis among small banks during the 1970s, followed by the sudden and dramatic collapse of a prestigious bank in the mid- 1980s. Such episodes of financial instability have prompted governments to place ‘‘prudential regula- tion’’—rules and policies designed to ensure the solvency and soundness of banks and other financial institutions—at the top of their economic policy agendas (Eichengreen 1999; Llewellyn 1999). Second, politicians face the age-old challenge of fighting inflation—an economic scourge that has become even more virulent in today’s environment of floating exchange rates and full capital mobility (Mosley 2003). In nearly all industrialized countries, politi- cians have granted their central banks a high degree
of insulation from political pressures, leaving them ostensibly to focus on keeping prices stable.
The dual policy goals of financial stability and low inflation generate a challenge for monetary policy makers. The central bank’s main monetary policy instrument is the interest rate, which dampens infla- tionary pressures when raised. However, interest rate hikes are potentially harmful to banks’ profits and increase the probability of bank failures (Cukierman 1991; OECD 1992). A policy of financial stability may require a more gradual monetary tightening in the face of inflationary pressures, thereby allowing banks more time to adjust their balance sheets. On the other hand, a policy of strict price stability would require more aggressive interest rate adjustments. With one policy instrument and two potentially conflicting goals, a central bank must decide how much empha- sis to place on fighting inflation versus maintaining financial stability.
In this article, we focus on the institutional tension between the dual goals of financial stability and low inflation. Specifically, we argue that mone- tary policy makers have a less aggressive stance toward inflation when governments combine the bank regulatory- and monetary-policymaking functions
The Journal of Politics, Vol. 70, No. 3, July 2008, Pp. 663–680 doi:10.1017/S0022381608080687
� 2008 Southern Political Science Association ISSN 0022-3816
663
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
within the central bank. Such ‘‘regulatory central banks’’—which exist in more than one-third of industrialized countries—have incentives to be espe- cially sensitive to the profitability and stability of the banking sector when setting monetary policy. We therefore argue that the presence of bank regulatory authority in the central bank’s mandate generates a bias in its monetary policymaking. On the other hand, in countries where bank regulation is assigned to a separate agency, the central bank is less likely to be biased by bank stability concerns and more likely to enact tighter monetary policies geared toward maintaining low inflation. Policy makers’ success in fighting inflation therefore varies with the institu- tional locus of bank regulatory authority.
We further argue that the effect of a central bank’s mandate on inflation is conditional on the govern- ment’s choice of exchange rate regime. Assuming full capital mobility, countries that adopt floating ex- change rates maintain the ability to conduct autono- mous monetary policy. Under such conditions, a central bank’s regulatory responsibilities play a signifi- cant role in shaping its monetary policy choices. In contrast, a central bank operating under fixed ex- change rates will not have the ability to pursue an independent monetary policy, and therefore its institutional features will be of little importance (O’Mahony 2007). We also test an ancillary hypothesis regarding the effect of the central bank’s regulatory mandate conditional upon the size of the domestic banking sector. We hypothesize that when banks represent a larger share of the national economy, a regulatory central bank will be particularly sensitive to bank stability when setting monetary policy.
In focusing on the relationship between central banks’ regulatory mandates and monetary policymak- ing, this paper contributes to two theoretical litera- tures in political economy. The first literature, which examines the politics of central banking, has focused overwhelmingly on the relative degree of political independence as the driving influence on the central bank’s monetary policymaking (e.g., Broz 2002; Clark 2002; Cukierman 1992; Franzese 1999; Grilli, Mas- ciandaro, and Tabellini 1991; Henning 1994; Max- field 1997). This literature depends critically on the implicit assumption that all central banks have the same responsibilities; thus, the ‘‘institutional struc- ture’’ of the central bank can be equated solely with its degree of insulation from the political whims of politicians (Bernhard, Broz, and Clark 2002).1
Politicians, who view monetary policy from the perspective of political expediency, will encourage a dependent central bank to adjust interest rates for political purposes (e.g., to provide a temporary boost to the economy before an election). In the absence of political interference, central banks are assumed to be aggressive inflation fighters.2 While we do not dispute the importance of a central bank’s political inde- pendence on its monetary policymaking behavior, we argue that the institutional locus of regulatory au- thority also strongly influences how central bankers make decisions—even for central banks with high degrees of independence. This argument essentially turns the central banking literature on its head: rather than assuming a uniform political influence on monetary policymaking, we argue that central banks can be pulled in different directions as a result of their institutional mandates.
The second relevant literature pertains to the dynamics of principal-agent relationships, specifically those that arise when elected governments delegate important tasks to bureaucratic agents. There are many studies of the challenges inherent in the dele- gation of authority to an agent whose preferences may differ from those of the principal and of the strategies available to principals to ensure the obedience of their agents (e.g., McCubbins and Schwartz 1984; Weingast 1984). The paper does not challenge this literature; instead, it illustrates how politicians’ initial choice to delegate multiple tasks to a single bureaucratic agency can have significant consequences for policy out- comes (Dewatripont, Jewitt, and Tirole 2000). The ‘‘multitask’’ dilemma exists in many policy domains, including environmental and energy policy, financial regulation, and health and safety. Consider, for example, the consequences of delegating responsibility for energy efficiency and environmental conservation to the same agency. If policies designed to minimize environmental damage have the side effect of raising energy costs, then an agent could find itself compro- mising one objective in favor of the other. A similar tension can be found in the mandates of financial regulators, who are frequently held accountable for the competitiveness and stability of the regulated industry. We apply this reasoning to central banks and examine how their anti-inflation policies might vary as a function of their regulatory responsibilities.
In the case of central banking in the developed world, governments generally decided their delega- tion structures in the early twentieth century, long
1Central bank independence is not always treated as a single analytical concept; see, e.g., Hallerberg (2002).
2For a discussion of central bankers’ preferences based on their varied career backgrounds, see Adolph (2005).
664 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
before the emergence of prudential regulation as a politically prominent issue. As regulation gained political salience, the preferences of regulatory central banks began to change. We note below that the mandates of central banks have been largely static over the years, but the evolution in the preferences of regulatory central banks has led some governments to alter their longstanding delegation decisions—specifi- cally by removing regulatory authority from the central bank—to ensure that monetary policy out- comes remain in line with government preferences.
The remainder of the article proceeds as follows. In the next section, we discuss the significant variation in the regulatory responsibilities of central banks, emphasizing that this variation has, until very recently, been entirely cross-national rather than intertemporal. We then explore the theoretical reasons why this variation in the central bank’s institutional mandate might influence its monetary policymaking choices, after which we move on to empirical tests of our argument. Using data from 23 industrialized countries from 1975 to 1999, we show that inflation has been systematically higher in countries where the central bank is vested with bank regulatory authority; as expected, however, we find that this effect is condi- tional on both the choice of exchange rate regime and the size of the domestic banking sector.
An important policy implication of this finding is that the removal of bank regulatory responsibil- ity from the central bank is a potential strategy for countries seeking to enhance the inflation-fighting credibility of their central banks. We test this logic by conducting a case study of the Bank of England, which lost its bank regulatory authority to the newly formed Financial Services Authority in a major British govern- ment initiative in 1998. We find that the new Labour government under Tony Blair was eager to control inflation and imposed the institutional change on the Bank of England in part to remove the bank stability bias from its monetary policymaking. Finally, we conclude with a discussion of the broader implications of these findings for our understanding of the politics of bureaucratic decision making.
Regulatory Responsibilities of Central Banks
Across the world, central banks vary in the responsi- bilities delegated to them by politicians. Central banks are generally responsible for implementing a country’s monetary policy by controlling the money supply and setting interest rates based on current economic
conditions. In addition, some—but not all—central banks serve as bank regulators with responsibility for implementing rules and restrictions on banking activ- ity, supervising compliance with prudential regulation and applicable laws, and otherwise safeguarding the stability of the banking sector. This variation has been understudied in the literature on the political econ- omy of inflation.3 Some scholars have examined the central bank’s regulatory mandate in the context of measuring its overall degree of independence from political pressures (Banaian, Burdekin, and Willett 1995). Other scholars have grappled with the inter- play between monetary policy and financial stability more generally (Cukierman 1991, 1992). However, in empirical analyses of inflation, central bank inde- pendence reigns supreme, and the central bank’s regulatory responsibilities are generally ignored.
The regulatory responsibilities of central banks are rooted in history. Until the early 1800s, the focus of central banks was predominantly on wartime finance. For example, the British government created the Bank of England in 1694 to raise funds for war and conquest (North and Weingast 1989). Norway’s central bank, the Norges Bank, was created in the wake of the Napoleonic Wars to stabilize the currency after a period of highly inflationary wartime expen- ditures (Capie, Goodhart, and Schnadt 1994). The exigencies of wartime finance eventually gave way to the more general goal of maintaining the internal and external value of the currency, which often involved ensuring convertibility in commodity-based exchange rate regimes. For many central banks, maintaining the value of the currency was closely intertwined with the regulation of the banking system, since commer- cial banks play a major role in the implementation of monetary policy (Capie, Goodhart, and Schnadt 1994). Central banks in countries such as Italy and the Netherlands gradually assumed bank regulatory responsibility in the twentieth century. The Bank of England operated for decades as an informal but powerful bank regulator and gained statutory author- ity to regulate banks with the Banking Act of 1979. In the United States, Congress created the Office of the Comptroller of the Currency—the regulator of na- tionally chartered banks—in 1863, but later delegated the statutory authority to regulate state-chartered
3An exception is Di Noia and Di Giorgio (1999), which offers mostly descriptive statistics and no theoretical argument. Posen (1995) includes the central bank’s regulatory mandate as part of a broader measure of financial opposition to inflation. In addition, there are several analyses of the pros and cons of separating the monetary-policy and regulatory functions, including Abrams and Taylor (2000); Barth et al. (2002); Goodhart (2001); Goodhart and Schoenmaker (1995); and Peek, Rosengren, and Tootell (1999).
financial regulation, monetary policy, and inflation in the industrialized world 665
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
banks and bank holding companies to the U.S. Federal Reserve, founded in 1913.
In contrast, central banks and regulatory agencies emerged as completely separate entities in countries such as Canada and the Scandinavian countries. Canada, in fact, developed a bank regulator first—in 1925—and waited nine years before establishing its central bank in 1934. For countries such as Sweden and Norway, central banks have existed for centuries, yet bank regulatory authority lies with independent agencies subsequently established in the nineteenth and early twentieth centuries.4
Today there is considerable cross-national varia- tion in the institutional mandates of central banks. Of the 23 industrial countries listed in Table 1, 14 currently have separate central banks and bank regu- lators, while nine have unified systems. It is interesting to note that levels of central bank independence do not appear to be associated with regulatory structure. Countries with unified systems have central banks ranging from highly independent (Germany and the United States) to relatively dependent (Spain).
The regulatory responsibilities of central banks in industrialized countries have demonstrated very little variation over time. This path-dependent characteristic stands in marked contrast to other aspects of central banks, such as their degrees of political independence, which have experienced more frequent changes over time (see Bernhard 1998). Indeed, altering the institu- tional mandate of the central bank is costly, both politically and financially. To remove the regulatory responsibilities of the central bank, governments must enact highly detailed legislation to establish a separate bank regulatory agency with its own staff, budget, and bylaws. In the industrialized world, there are only three exceptions to the endurance of the central bank’s mandate: the United Kingdom and Australia, both of which transferred bank regulatory responsi- bilities to newly created separate agencies in 1998, and Iceland, which made a similar change in 1999.5 In the
overwhelming majority of cases, however, the regu- latory responsibilities of the central bank have re- mained steady through the years. This historical claim will prove important in our statistical analysis of inflation performance from 1975 to 1999, in which we assume that the central bank’s mandate is exoge- nously determined.
TABLE 1 Location of Bank Regulatory Authority, Industrial countries (2005)
Central bank Separate agency
France1 Australia2
Greece Austria Ireland Belgium Italy Canada Netherlands Denmark New Zealand Finland Portugal Germany3
Spain Iceland4
United States5 Japan Luxembourg Norway Sweden Switzerland United Kingdom6
1The Banking Commission (Commission Bancaire) is a compo- site body chaired by the Governor of the Banque de France, with representatives from the Treasury. While the World Bank data classify this as a separated regime, we follow Goodhart and Schoenmaker (1995) in treating this as unification. However, our results are not sensitive to the alternative specification. 2Prior to 1999, the Reserve Bank of Australia regulated banks. This authority now resides with the Australian Prudential Regulatory Authority. 3The Federal Financial Services Supervisory Authority (Bunde- sanstalt für Finanzdienstleistungsaufsicht, BaFIN), established in 2002, united the three previously independent supervisory agencies for banking, securities, and insurance. BaFIN is en- trusted with control over the ‘‘sovereign measures’’ (licensing, issuing regulations) whereas the Bundesbank only participates in the ‘‘operational tasks’’ of supervision (e.g., collecting and processing banks’ prudential returns). See Goodhart and Schoenmaker 1995, http://www.bundesbank.de/bankenaufsicht/ bankenaufsicht_bafin.en.php, and http://www.bafin.de/bafin/ aufgabenundziele_en.htm#n6. Goodhart and Schoenmaker clas- sify this arrangement as a separated regime, while the World Bank treats it as ‘‘unification.’’ We follow the former, but our results are not sensitive to the alternative specification. 4Prior to 1999, the Bank Inspectorate was part of the Central Bank of Iceland. It was merged with the Insurance Supervisory Authority into a separate entity, the Financial Supervisory Authority, on January 1, 1999 (http://www.invest.is/files/ 2075240179Financial_System.qxp.pdf). 5The Federal Reserve is not the only bank regulator in the U.S., but it has overall responsibility for banks within the Federal Reserve system, and also regulates all bank holding companies. Note that our definition of ‘‘unification’’ is that the central bank has or shares responsibility for regulation. 6Prior to 1999, the Bank of England regulated banks. This authority has since been transferred to the Financial Services Authority.
4The Riksbank was established in Sweden in 1668, while the Royal Inspectorate of Banks was founded in 1907, and continues to operate today as the Swedish Financial Supervisory Authority (Finansinspektionen). In Norway, the Norges Bank was founded in 1816, while the Banking Inspectorate was founded in 1825; it operates today as the Banking, Insurance, and Credit Commis- sion under the Ministry of Finance.
5The new agencies are called the Financial Services Authority in the United Kingdom, the Australian Prudential Regulatory Authority in Australia, and the Financial Supervisory Authority in Iceland. Korea also experienced an institutional change in 1998, but it was not considered an industrial country for much of the time period of our analysis, and thus is not included in our sample.
666 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
Central Banks, Regulation, and Monetary Policy Bias
Scholars of monetary policymaking have long argued that the central bank faces dueling pressures to keep prices stable and to maintain economic growth and full employment. The classic model by Barro and Gordon (1993) indicates a tension between maintain- ing inflation at a socially optimal level and respond- ing to political pressure to decrease unemployment. It is assumed that inflation aversion varies across countries. Consider the following quadratic loss function (adapted from Persson and Tabellini 2000):
L ¼ ½lðp � pÞ2 þ ðx � xÞ2�=2 ð1Þ
where p and x represent inflation and output (with overbars representing society’s preferred values), and l is the relative weight that the policy maker places on controlling inflation. The political economy liter- ature generally assumes that central bank independ- ence (CBI) is a useful proxy for l, with the expectation that politically insulated central banks will be more conservative, or inflation hawkish, than politically dependent ones.
While we do not dispute the importance of a central bank’s political independence on its monetary policymaking behavior, we argue that a sole focus on CBI ignores the fact that central banks also vary on another dimension: their regulatory mandates. Cen- tral banks responsible for bank regulation must respond appropriately to exogenous influences on the price level without triggering politically costly instability in the banking sector. When the central bank raises interest rates by decreasing the money supply, it deters investment and dampens aggregate demand, thereby keeping inflation in check. How- ever, interest rate changes motivated exclusively by expected changes in the price level have potentially adverse consequences for the profitability and sol- vency of banks (Cukierman 1991, 1992; Goodhart and Schoenmaker 1995).6 Banks are particularly vulnerable to changing financial market conditions because they must commit to loan terms in advance. More specifically, banks that issue fixed-rate loans, such as mortgages and certain corporate loans, will face declining profits as increasing interest rates force them to raise their own deposit rates. Bank customers might also withdraw their money in favor of higher-
yielding money market accounts or other invest- ments. The overall increase in the cost of funds, in turn, cuts into banks’ profits and increases the like- lihood of bank failures (OECD 1992). Increasing interest rates also leads to a greater risk of default by bank customers with flexible-rate loans, which also increases the potential for bank failures (OECD 1992; Tuya and Zamalloa 1994).7
The connection between monetary policy and bank stability does not inhere exclusively in the level and volatility of interest rates. A more general tension is that the cyclical effects of monetary policy and bank regulation push in opposite directions (Good- hart and Schoenmaker 1993). Monetary policy tends to move in countercyclical fashion: in the event of an economic slowdown, the central bank expands the money supply and provides more funds to speed up the economy’s recovery. However, the effects of bank regulation—especially prudential regulation, such as capital adequacy requirements—are procyclical, re- quiring a contraction of banking activity when the economy slows. For example, during a recession, a bank regulator might require an increase in loan-loss reserves and an improvement in the quality of banks’ lending portfolios. The resulting decrease in lending activity would result in tighter credit just as the monetary authorities were attempting to facilitate new lending to spur investment and consumption.
Examples of the tension between tight monetary policy and financial stability can be found across the developed and developing world. In New Zealand in 1985, a run on foreign exchange reserves prompted the central bank to increase interest rates dramati- cally, but concern over the soundness of the banking sector led to a reversal in policy and a prompt injection of liquidity (Healey 2001). In the United States, in 1979, Federal Reserve Chairman Paul Volcker announced that the Federal Open Market Committee (FOMC) would adopt a tight-money strategy called ‘‘nonborrowed reserve targeting,’’ in which changes in monetary aggregates such as M1 and M2—rather than interest rates or other eco- nomic targets—would serve as the target of monetary policy. When the LDC Debt Crisis began to unfold in 1982, policy makers and economists viewed the stringency of this approach as a contributing factor
6For a similar argument about the vulnerability of the financial sector to interest-rate defenses of fixed exchange rates, see Walter and Willett (2007).
7A drop in interest rates may also have adverse consequences for the banking system. If banks have loan portfolios that are more sensitive to interest rate fluctuations than their liabilities (e.g., treasury bonds and customer deposits), then a drop in interest rates—instigated by the central bank to spur economic growth or prevent disinflation—will lead to a deterioration in bank profit- ability in the short run. See Di Noia and Di Giorgio (1999).
financial regulation, monetary policy, and inflation in the industrialized world 667
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
to the growing debt problems of the LDCs and the resulting instability of overexposed U.S. banks (Goodhart and Schoenmaker 1995). By the autumn of 1982, Volcker abandoned the FOMC’s sole focus on money supply targets in favor of inflation targets and other macroeconomic measures, which proved more sensitive to the plight of the banking sector. Finally, there are countless examples in the develop- ing world of national banking systems actively resist- ing the tight money policies of central banks during financial crises (Walter and Willett 2007). The central bank of Indonesia, for example, raised interest rates sharply in November 1997 in the wake of the Asian financial crisis, but quickly injected new liquidity into the economy as banks began to collapse (Boorman et al. 2000).
Given the tension between monetary tightening and bank stability, how will the central bank make monetary policy? Our central claim, which emanates from the earlier work of Cukierman (1991, 1992) and others, is that the presence of regulatory responsi- bility in the central bank’s institutional mandate introduces an important bias into its monetary policymaking calculus. When the central bank has official responsibility for regulating the banking sector, it is held publicly accountable in the event of bank failures or a sharp decline in bank profit- ability.8 For example, after the sudden collapse of Johnson Matthey Bankers (JMB) in the United King- dom in 1984, the Bank of England—which at the time was also the bank regulator—faced a barrage of criticism from Parliament. Indeed, the reappoint- ment of a prominent deputy governor of the Bank to another five-year term was delayed because of the negative fallout from the JMB affair (Singer 2007). It should come as no surprise that politicians find it politically expedient to castigate bank regulators for bouts of financial instability; in extreme cases, elected leaders will force the heads of regulatory agencies to resign. Any central bank with regulatory responsibil- ity will therefore be especially sensitive to short-term bank stability when making policy decisions.
In contrast, central banks without regulatory authority will, all else equal, base their monetary policy decisions on the price level, with less emphasis on the impact of interest rates on bank solvency or profitability. If banks become unstable, policy makers will place blame on the agency with official respon- sibility for bank supervision, such as Canada’s Office of the Superintendent of Financial Institutions or the
Australian Prudential Regulatory Authority. Central bank policy might be the focus of legislative discus- sion, but it is unlikely that the central bank itself will be held ultimately responsible if banks should fail. The U.S. savings and loan (S&L) crisis is illustrative: the rapid interest rate increases during the early 1980s squeezed the profit margins of S&Ls that issued long- term fixed-rate mortgages. Those S&Ls that did not take proper steps to rebalance their portfolios found themselves in dire financial trouble, and many collapsed altogether. The Federal Reserve, however, was not held publicly accountable for the S&L crisis despite the fact that its aggressive monetary policies were a contributing factor.9 Indeed, banks do not fail because of interest rate changes per se; they fail because of improper risk management, inadequate capital, or other forms of malfeasance—all of which are part of the bank regulator’s jurisdiction.
Returning to equation (1), our argument can be stated more formally: lr , ls , where the subscripts r and s refer to regulatory and separate central banks, respectively. In other words, separate central banks should be more inflation hawkish than regulatory central banks, all else equal. The discrepancy between lr and ls arises as a result of the varying principal- agent relationships between elected leaders and the central bank. When politicians delegate regulatory authority to the central bank, banking instability be- comes personally costly to central bankers. We there- fore expect central bankers to adjust their monetary policies with careful attention to bank stability.
Although we expect this logic to hold across countries and over time, it is important to acknowl- edge three important caveats to our argument. First, as noted previously, the influence of the institutional mandate of the central bank on inflation outcomes is likely to be conditional upon the prevailing exchange rate regime. Whether a country fixes or floats its exchange rate will influence the degree of monetary policy autonomy of the central bank. In a fixed exchange rate regime, the central bank has limited or no autonomy to set an independent monetary policy, and thus its characteristics—including its regulatory responsibilities—will have little impact on inflation outcomes (Bearce 2003; Oatley 1999; O’Mahony 2007).10
8For an example of this argument applied to the U.S. Federal Reserve, see Cukierman (1991).
9In contrast, Congress quickly dismantled the S&L regulator—the Federal Home Loan Bank Board—and fired its chairman.
10Central banks can employ sterilization as a monetary policy strategy in the short term, but this strategy is generally not viable in the medium- or long-term given the continual need to spend down foreign exchange reserves.
668 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
Second, it is plausible that the influence of the central bank’s regulatory mandate on inflation out- comes is conditional on the size of the domestic banking sector. In particular, the central bank’s regulatory responsibilities might be more likely to influence the magnitude of l (the relative weighting given to inflation fighting) in countries where the banking sector is large relative to the overall size of the economy. The logic is straightforward: the size of the banking sector is an excellent proxy for the potential magnitude of the career costs associated with bank instability. In contrast, the central bank’s institutional mandate should have less of an impact on monetary policymaking when the banking sector is relatively small. As Figure 1 illustrates, there is significant variation in the size of the domestic banking sector across the advanced industrialized countries. While we believe the central bank’s regu- latory responsibilities are influential in all countries, we acknowledge the possibility that banking sector size may be an important determinant of the magni- tude of a regulatory central bank’s bias.
Finally, it is important to emphasize that no central bank ignores financial stability entirely. To be sure, the central bank is always concerned about the stability of the banking system, and bank stability factors into monetary policy regardless of the central bank’s additional responsibilities (Briault 1999). Cen- tral banks rely on the banking system for the smooth functioning of open market operations, and for the satisfactory supply of credit to the economy. Bank of England Governor Eddie George makes the case more
bluntly: ‘‘It is inconceivable that the monetary au- thorities could quietly pursue their [price] stability- oriented monetary policy objectives if the financial system through which policy is carried on . . . were collapsing around their ears’’ (George 1994). On this point we do not disagree; certainly no central bank would ever preside insouciantly over a faltering fi- nancial system. Rather, as depicted in equation (1), our argument is one of degree. Central banks without regulatory authority will place less emphasis on the negative externalities of bank stability in their policy choices, and more emphasis on fighting inflation. Thomas Cargill notes that ‘‘[t]he channels of conflict are fundamentally related to the need to have an economy-wide perspective for the conduct of mon- etary policy as opposed to an industry-wide per- spective for the conduct of financial regulation’’ (Cargill 1989, 60). Since monetary policy influences the objectives of financial regulation—namely bank stability—a central bank with regulatory power is expected to be more sympathetic to the perspective of industry than a central bank without regulatory authority.
Nevertheless, it is also important to note that the link between institutional structure and inflation could obtain even if central bankers make monetary policy decisions without actively taking their varying regulatory responsibilities into account. If market actors have expectations that the central bank might be influenced by the presence or absence of regu- latory responsibility, then they will react accordin- gly—setting wage contracts and locking in their
FIGURE 1 Average Banking Sector Size, Industrialized Countries, 1975–99
financial regulation, monetary policy, and inflation in the industrialized world 669
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
expected changes in the price level (Di Noia and Di Giorgio 1999; Goodhart and Schoenmaker 1995).
Our expectations of different policies as a result of central banks’ varying institutional structures are consistent with a variety of analytical approaches to regulation. Scholars such as Stigler (1971) argue that regulators are ‘‘captured’’ by the regulated industry, and enact policies that reflect the industry’s short- term needs. Central banks with regulatory authority (or ‘‘captured central banks’’) would therefore enact monetary policies more favorable to banks than central banks without regulatory authority. As Frieden (1991) notes, banks are generally considered to be a strong constituency in favor of price stability. How- ever, in the event of exogenous shocks, banks would prefer that the central bank take an ‘‘interest rate smoothing’’ approach—in which interest rates are increased in relatively small and steady increments— even if a more aggressive strategy would be more effective in controlling inflation (Cukierman 1991; see also Walter and Willett 2007). Other scholars argue that regulators safeguard their decision-making discretion and avoid policy choices that increase the chances of political intervention (Ferejohn and Shipan 1990; Singer 2004, 2007; Weingast and Moran 1983; Woolley 1984). Under this assumption, a central bank with regulatory responsibility would be especially sensitive to the threat of bank instability, since the legislature would likely react to bank failures with heightened monitoring of the central bank’s policies.11 In addition, our argument is consistent with the emerging literature on multitask incentive prob- lems in principal-agent theory (Dewatripont, Jewitt, and Tirole 2000). Agents with multiple responsibil- ities will behave differently than single-task agents, especially when there are potential conflicts between the agents’ tasks.
Empirical Analysis
Thus far, we have argued that a central bank’s regulatory mandate is an important yet underappre- ciated factor influencing its monetary policymaking behavior. In this section, we test this theory by exploring the relationship between the institutional locus of financial regulation—that is, whether the central bank has regulatory responsibility or not—
and inflation outcomes for a panel of 23 industrial- ized countries for the 1975–99 period. The empirical analysis follows the econometric specifications used in recent political economy work on inflation out- comes (Keefer and Stasavage 2002, Franzese 1999, Hall and Franzese 1998, Cukierman, Webb, and Neyapti 1992). The dependent variable is the (logged) five-year average inflation rate. Period averages are appropriate because many of the institutional deter- minants of inflation rates in our model, including the central bank’s mandate, change infrequently.12
The main explanatory variable, regulatory sepa- ration (hereafter Separate), is a binary variable that takes the value of 0 if the central bank holds or shares bank regulatory authority (unification) and a value of 1 if the task of bank regulation has been delegated to an agency separate from the central bank (separa- tion).13 Based on the theory explicated above, we expect this variable to be negatively associated with inflation outcomes—that is, countries that vest reg- ulatory authority in a separate agency should have lower inflation rates than countries that combine the two functions in the central bank. Data on Separate are compiled from a recently available World Bank dataset, Bank Regulation and Supervision, which assembles the results of a cross-national survey on the structure of financial markets and financial regulation (Barth, Caprio, and Levine 2003).14 We supplement the World Bank source with data from Goodhart and Schoenmaker (1995) and national monetary policy and regulatory policy authorities.15
In the first model, we include a standard battery of control variables, including central bank inde- pendence (CBI), exchange rate regime, trade open- ness, capital account openness, and the log of GDP and GDP per capita. Data on CBI are based on Cukierman’s (1992) methodology for calculating legal independence and are compiled from the Com- parative Political Dataset for the period 1975–96 (Armingeon et al. 2005) and Polillo and Guillen
11The policy literature on the institutional structure of central banks includes ‘‘increased politicization’’ as one of the disadvan- tages of vesting regulatory authority with the central bank. See, e.g., Barth et al. (2002).
12Inflation data are taken from the World Bank’s World Develop- ment Indicators.
13Note that the dependent variable does not take into account whether a single regulator oversees banking, securities, and insurance, or whether multiple agencies participate in bank regulation. Separate simply classifies the central bank’s partic- ipation in bank regulation. Data are taken from question 12.1 of the World Bank survey utilized by the authors to compile the dataset: ‘‘What body or agency supervises banks?’’
14For complete background on the dataset, as well as the data itself, see http://www.worldbank.org/research/projects/bank_ regulation.htm.
15See Table 1 for further details.
670 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
(2003) for the remaining years.16 The exchange rate regime variable takes the value of 0 for floating or managed floating regimes and 1 for all varieties of ‘‘hard’’ fixed exchange rates (currency boards, monetary unions, hard pegs).17 Data are based on the International Monetary Fund’s classification of each country’s choice of exchange rate regime as published in the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions. Each obser- vation in the dataset is the average of the annual observations over the relevant five-year period.
Trade openness, measured as imports plus ex- ports as a percentage of GDP, is taken from the World Development Indicators database, along with GDP and GDP per capita. We also include the Chinn-Ito index of capital account openness (‘‘KAOPEN’’), in which higher values indicate greater degrees of openness (Chinn and Ito 2006). KAOPEN measures the extent of legal restrictions on cross-border financial transactions. It is based on the binary coding of restrictions in the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions and focuses on four dimensions of restrictions: the existence of multiple exchange rates, restrictions on the current and capital accounts (where the latter are measured as the proportion of the last five years without controls), and requirements to surrender export proceeds.18 The index has a mean of zero and ranges from -2.66 (full capital controls) to 2.66 (complete liberalization). We also include GDP to control for the significant disparity in size between the smallest industrialized countries (Luxembourg, Iceland, New Zealand) and the largest (United States, Japan, Germany). Likewise, including GDP per capita alle- viates our concerns that some of our institutional variables (e.g., CBI, KAOPEN) are merely proxies for different levels of development across the developed countries (Keefer and Stasavage 2002).
We include several additional controls. First, we add a variable that captures whether there was a currency crisis in the country during the four-year period. We also add a similar variable for the presence of a banking crisis. Data for both variables is taken from Glick and Hutchinson (1999), with missing data for Australia and the United States filled in using Caprio and Klingebiel (2003). Each of these variables is measured for individual years over the 1975–99 period, taking a value of 1 if there was a crisis and 0 otherwise. For each five-year period in the sample, each variable represents the sum of these individual year observations divided by five, thereby indicating the portion of time within each five-year period in which a crisis was occurring. All else equal, we expect inflation to be higher during periods that were more crisis-prone. Second, we include a dummy variable to capture whether a country has implemented explicit deposit insurance. Our expect- ation is that all central banks—regardless of their institutional mandates—have greater leeway to pur- sue price stability-oriented monetary policy when an explicit deposit insurance scheme protects depos- itors from losses in the case of bank failures. Data are taken from the World Bank’s Deposit Insurance Around the World database (Demirgücx-Kunt, Kar- acaovali, and Laeven 2005). Third, we include the size of the domestic banking sector, measured as domestic credit provided by the banking sector as a percentage of GDP. Data are taken from the World Bank’s World Development Indicators. Finally, we include a time trend variable that ranges from 1 to 5 in accordance with the five periods in the dataset. This variable controls for unobserved characteristics across each five-year period, including the factors that may have contributed to the secular decline in inflation rates across the developed world over the last 25 years.19
Models and Results
We first specify a basic linear regression (OLS) model with panel-corrected standard errors. This specifica- tion accounts for heteroskedasticity and spatial error correlation, common problems in time-series cross- sectional data that are evident in our sample (Beck
16Alternative CBI measures are available for only a subset of our sample; we therefore rely on Cukierman’s index. For a discussion of the challenges of measuring CBI, see Banaian, Burdekin, and Willett (1995).
17Alternative specifications using a 3-point ordered scale (float- ing, managed floating/intermediate, fixed) produce similar re- sults. Since we are primarily concerned with the distinction between ‘‘hard’’ pegs and other regimes, we utilize the simpler binary scale in the analysis that follows. The authors thank Nancy Brune for sharing her dataset, which draws on Bernhard and Leblang (1999) and compiles and codes the raw IMF data.
18For a detailed description of this measure, see Chinn and Ito (2006). 1975–79 data are missing for Switzerland and 1975–80 data are missing for the Netherlands; we use the 1980 and 1981 values, respectively, for these periods.
19In 1975–79, average inflation in the 23 industrialized countries was 12.02% per year. While average inflation rose slightly in 1980–84 (12.19%), it subsequently declined in each period (6.06% in 1985–89; 4.35% in 1990–94; 1.97% in 1995–99).
financial regulation, monetary policy, and inflation in the industrialized world 671
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
2004; Beck and Katz 1995). The basic model is as follows:
Log Inflationit¼ aþ b ðIndependent VariablesÞit þ e
In our second model, we more directly assess the argument that the influence of the institutional struc- ture of the central bank (Separate) on inflation is conditional upon the exchange rate regime. When the exchange rate is fixed, the central bank will have limited or no autonomy to conduct an independent monetary policy, and thus the central bank’s institutional struc- ture should have little effect on inflation outcomes. We therefore include a multiplicative interaction term that captures the influence of Separate conditional upon the exchange rate regime. All else equal, we expect that a country which separates the bank regulatory authority from the central bank (Separate51) will realize lower inflation rates under floating exchange rates (Exchange Rate Regime50), since the central bank’s monetary policy autonomy increases and it will be free to pursue strict price stability. Under fixed exchange rates, how- ever, we expect Separate to have no significant effect on inflation outcomes, since the commitment to a fixed exchange rate commits all central banks—regardless of their regulatory mandates—to a price stability-oriented monetary policy. We therefore expect the interaction term to be positively associated with inflation outcomes.
Finally, in model 3 we test the argument that the influence of Separate on inflation is conditional not only on the exchange rate regime, but also on the size of the domestic banking sector. In this model, we therefore include a three-way multiplicative interac- tion, which captures the influence of Separate on inflation conditional on both the exchange rate regime and the banking sector size.
Table 2 presents the regression results for all three models, which are robust to alternative specifica- tions.20 In the models, several of the controls have the expected effects on inflation outcomes, including capital account openness and currency crises. The time trend variable is also significant and negative. Likewise, inflation rates are significantly lower in countries that choose fixed exchange rate regimes than those that pursue floating rates. Country size, as measured by GDP, is also associated with lower
inflation outcomes in model 3 but is not statistically significant in models 1 and 2. The remaining varia- bles appear to have had less impact on industrialized countries’ inflation rates during the 1975–99 period: trade openness, GDP per capita, deposit insurance, and banking crises are insignificant in all three models. Despite the extensive literature arguing that independence is the key institutional factor affecting
TABLE 2 Regression Results
Dependent variable: ln(inflation) 1 2 3
GDP (log) 20.039 20.041 20.087** (0.050) (0.046) (0.042)+
GDP Per Capita (log) 20.067 20.068 20.006 (0.119) (0.106) (0.086)
Trade Openness 20.002 20.002 20.003 (0.002) (0.002) (0.002)
Capital Account Openness (KAOPEN)
20.271*** (0.043)
20.257*** (0.044)
20.250*** (0.044)
Deposit Insurance (1 5 yes)
0.077 (0.094)
0.11 (0.093)
0.12 (0.098)
Currency Crisis (1 5 yes)
0.720** (0.304)
0.689** (0.280)
0.597** (0.253)
Banking Crisis (1 5 yes)
0.103 (0.135)
0.057 (0.136)
0.034 (0.126)
Time 20.288*** 20.298*** 20.323*** (0.058) (0.058) (0.058)
CBI 0.361 0.493 0.246 (0.309) (0.308) (0.323)
Exchange Rate Regime (1 5 fixed)
20.281*** (0.087)
20.410*** (0.099)
20.751*** (0.227)
Banking Sector Size
20.004* (0.002)
20.003 (0.002)
20.001 (0.002)
Separate Central Bank
20.151** (0.071)
20.382*** (0.114)
20.037 (0.226)
Exchange Rate Regime * Separate
0.383*** (0.139)
0.131 (0.129)
Exchange Rate Regime * Banking Sector Size
0.003 (0.003)
Banking Sector Size * Separate
20.004** (0.002)
ER Regime * Banking Sector Size * Separate
0.003** (0.001)
Constant 4.826*** 4.849*** 5.553*** (1.631) (1.508) (1.428)
Observations 115 115 115 Number of countries 23 23 23 Log-likelihood 257.167 254.153 249.709 R2 0.809 0.819 0.832 Adjusted R2 0.786 0.795 0.805
Panel corrected standard errors in parentheses. *p , 5.10; **p , 5.05; ***p , 5.01.
20The results do not change substantively when including the ‘‘lagged dependent variable’’ (e.g., the previous five-year period’s inflation average) or when including period effects. Additional explanatory variables, such as union density, centralization and coordination of wage bargaining, veto players, the partisan composition of government, unemployment rates, GDP growth, and a dummy for an election year, also do not affect the basic results.
672 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
central banks’ monetary policy behavior, CBI is also not significantly associated with lower inflation in our models. This finding could be attributable to measurement issues with CBI (e.g., Banaian, Burde- kin, and Willett 1995) or to theoretical problems with the link between CBI and inflation (e.g, Posen 1995).
In contrast, the results provide strong support for our argument that central banks’ regulatory man- dates are a key determinant of monetary policy- making and inflationary outcomes in the developed world. In the basic model, Separate is negative and highly significant. Since the significance of Separate in the interactive models is conditional on the exchange rate regime and banking sector size, we cannot directly interpret the regression coefficients in models 2 and 3 for these variables. However, F-tests of the joint significance of the interaction terms and their components in both models 2 and 3 are significant at the 1% level. We further analyze and interpret these interactive results below.
Given that the dependent variable in our model is the natural log of inflation, interpreting the magni- tude of these regression coefficients even in the ‘‘base’’ model is not straightforward. Therefore, in Table 3, we present first differences that illustrate the predicted effect of a one standard deviation (or equiv- alent relevant change) increase in each significant independent variable on average inflation, holding all other variables constant at their means.21 These re- sults further clarify the importance of a central bank’s regulatory mandate as a key determinant of inflation outcomes. All else equal, countries in which the central bank does not regulate banks (Separate51) have inflation rates that are 0.81 % lower than in ‘‘unified’’ (Separate50) countries. This effect is quite large in relation to both the average inflation rate in the industrialized countries from 1975 to 1999
(7.32%) and the predicted rate with all variables held at their means (5.38%).
Table 3 also illustrates the relative effect of the significant control variables from model 1 on infla- tion. A one standard deviation increase in capital account openness (KAOPEN), roughly equivalent to a shift from the mean level to its maximum, reduces inflation by 1.64%, a decrease of 30.5% from the predicted rate of inflation when all variables are set at their means (5.38%). Similarly, a shift in the time trend variable from period 2 (1980–84) to period 4 (1990–94) reduces inflation by 3.13%. This finding reinforces the observation that inflation throughout the developed world has declined sharply since the late 1970s. As expected, fixed exchange rates also reduce inflation: a shift from floating to fixed rates reduces average inflation by 1.57%. Currency crises, on the other hand, increase average inflation by 5.42%.
Although these results clearly support our basic hypothesis, they do not test our conditional argu- ment that the effect of Separate should depend on the choice of exchange rate regime. Thus, in Table 4, we present the relevant quantities of interest for the two-way interactive model (model 2). Here, too, the results strongly support our argument. Under float- ing exchange rates, separation of monetary policy- making and bank regulation significantly reduces inflation. When the central bank regulates banks under floating rates, predicted average inflation is 7.16%. However, when responsibility for bank regulation is assigned to a separate agency, average predicted inflation declines sharply to 4.28%. As the table illustrates, this is a statistically significant differ- ence at the 95% confidence level. In contrast, ‘‘sep- aration’’ has no significant effect on inflation outcomes under fixed exchange rates. Indeed, pre- dicted average inflation under fixed rates is almost identical in both ‘‘separated’’ and ‘‘unified’’ systems (4.29% to 4.33%, respectively). This result confirms
TABLE 3 First Differences, Model 1* Predicted average inflation rate, all variables at mean: 5.38%
Variable Predicted change in
average inflation Interpretation of one standard deviation
(or equivalent) change in X
KAOPEN 21.64% 1.25 to 2.59 (maximum openness 5 2.65) Currency Crisis 5.42% 0 to 1 time 23.13% Time 5 2 (1980-84) to Time 5 4 (1990-94) Exchange Rate Regime 21.57% 0 to 1 Separate 20.81% 0 to 1
*All other variables held constant at their means
21Calculations done using the Stata post-estimation software CLARIFY (Tomz, King, and Wittenberg 2003).
financial regulation, monetary policy, and inflation in the industrialized world 673
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
our expectation that the institutional structure of the central bank has clear and significant effects on monetary policymaking and inflation outcomes, but that this effect is conditional on the choice of exchange rate regime.
Finally, in Figures 2 and 3, we analyze and interpret the three-way interactive model (model 3), in which Separate is conditional on both the choice of exchange rate regime and the size of the domestic banking sector. In Figure 2, we graph the marginal effect of Separate under both fixed and floating exchange rates as banking sector size varies across the range of observed values in our dataset.22 As is evident from the chart, separation has no significant effect on inflation under fixed exchange rates, regard- less of the size of the banking sector (dashed line). In contrast, separation has a clear and significant neg- ative effect on inflation under floating rates, but only at mid- to high levels of banking sector size.23 This result further confirms our expectation that the regulatory mandate of the central bank has clear and significant effects on monetary policymaking and inflation outcomes, but it also further clarifies the conditional nature of this expectation. Figure 3 builds on this finding by graphing the predicted inflation level under separation (SEPARATE- 5 1) and float- ing exchange rates as banking sector size varies across the range of observed values in our sample. Once again, the large and significant negative impact of separation on inflation outcomes under this unique combination of conditions is clearly evident. As these results illustrate, the institutional structure of the
central bank has the most significant impact on monetary policymaking and inflation when two conditions are met: (1) A country is operating under floating exchange rates; (2) The banking sector represents a sizeable portion of the domestic macroeconomy.24
As a robustness check, we consider the possibility that selection into the ‘‘separate’’ category may not be exogenous to the explanatory variables in our data—or in other words, the same variables that explain variation in inflation outcomes also may explain a country’s initial decision to separate bank regulatory authority from the monetary policy author- ity. To address this concern, we use propensity score matching to ‘‘preprocess’’ our data (Ho et al. 2007; Leuven and Sianesi 2003; Simmons and Hopkins 2005). The critical idea behind propensity score matching is to match each ‘‘treated’’ observation (in this case, each country-year observation of ‘‘separation’’) with a ‘‘control’’ observation (i.e., a country-year observation of nonseparation) for which all the values of the explanatory variables are as close to identical as possible. The matching estimation results demonstrate that the paper’s main findings still hold after controlling for possible selection bias.25
A Case of Institutional Change: The Bank of England
As discussed earlier, governments in the industrial- ized world decided the regulatory responsibilities of their central banks many years ago, and those institutional configurations have been largely static over time. However, three industrialized coun- tries—the United Kingdom, Australia, and Ice- land—modified the mandates of their respective central banks in the late 1990s by transferring bank regulatory responsibility to newly created regulatory agencies. The argument presented above suggests that
TABLE 4 Predicted Inflation by SEPARATE, Conditional on Exchange Rate Regime*
Exchange rate Regime
Does the CB regulate banks?
Predicted average inflation rate (%)
Confidence interval (95%)
Floating Yes (Separate 5 0) 7.16 6.18–8.28 No (Separate 5 1) 4.28 3.27–5.60
Fixed Yes (Separate 5 0) 4.29 3.49–5.28 No (Separate 5 1) 4.33 3.71–5.04
*All other variables held constant at their means
22Table 6 was generated using the Stata code for analyzing interactive terms developed by Brambor, Clark, and Golder 2006 (http://homepages.nyu.edu/~mrg217/interaction.html).
23The mean value of banking sector size in the sample is 90.4%, with a minimum of 27.0% and a maximum of 239.4%.
24The results for the size of the banking sector should be viewed with caution; while its marginal effect on inflation (given a separate central bank and a floating exchange rate) is statistically significant, the confidence intervals are fairly wide as a result of the small sample size. 25The results are available upon request by the authors.
674 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
a government’s price-stability bias could lead it to alter the mandate of the central bank. In the follow- ing brief case study, we explore the applicability of our argument to the institutional evolution of the Bank of England in the United Kingdom. We find that price stability was a primary concern behind the modification of the Bank of England’s mandate, but that exogenous factors—namely the rise of financial conglomerates with multiple financial functions— were also an important influence behind the transfer of regulatory responsibilities to new all-encompassing ‘‘super regulator.’’ The case study demonstrates the possibility that governments in the future could be swayed to shift regulatory responsibility away from the central bank for price-stability concerns. At the same time, it also underscores the point that the central bank’s mandate is subject to exogenous (i.e., noninflation related) influences.26
Inflation, Regulation, and the Bank of England
In 1997, the incoming Labour government in the United Kingdom made an abrupt change in the mandate of the Bank of England. The new Chancellor of the Exchequer, Gordon Brown, announced that the Bank’s regulatory responsibilities would be transferred to a newly created financial regulator, the Financial Services Authority (FSA). In a separate announcement, Chancellor Brown also transferred full monetary policy authority to the Bank, thus
changing its status from an agent of the Treasury to a relatively independent central bank, similar to its counterparts in the industrialized world. These two policy initiatives were not unrelated; indeed, the preceding discussion about the conflicting mandates of central banks and bank regulators helps us to understand the rationale behind Brown’s actions.
The Bank of England faced a two-pronged challenge in the 1980s and 1990s. The first challenge was to maintain price stability in the face of increas- ing economic globalization and a turbulent foreign exchange market. The United Kingdom faced average annual inflation rates of approximately 10% between 1980 and 1984, and in excess of 5% through the mid- 1990s. In contrast, the United States kept average inflation below 8% between 1980 and 1984 and well below 4% through the mid-1990s.27 The United Kingdom’s struggle with inflation can be traced to its history of adopting and then abandoning a series of exchange rate regimes during the period after the fall of Bretton Woods (Bernanke et al. 1999). The most dramatic episode in Britain’s recent monetary history came in 1992, when currency speculators attacked the pound, leading to an abrupt devaluation and Britain’s hasty exit from the Exchange Rate Mechanism (ERM). In the absence of an exchange rate target, policy makers grew concerned about the credibility of monetary policy. Norman Lamont, then the Chancellor of the Exchequer, promptly an- nounced a temporary inflation target to guide the Treasury’s monetary policy decisions through the end of the current Parliament, with the understanding
FIGURE 2 Marginal Effect of SEPARATE (0 to 1), Three-way interaction
FIGURE 3 Predicted Values of Inflation by Banking Sector Size (with Separate Central Bank and Floating Exchange Rates)
26This case study draws upon a series of interviews conducted by the authors with regulators and financial consultants at the Bank of England, the Financial Services Authority, and private financial institutions in London during the summer of 2005. 27Inflation data taken from the World Development Indicators.
financial regulation, monetary policy, and inflation in the industrialized world 675
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
that the next government—whether Conservative or Labour—would have to make a decision about the future course of monetary policy after the 1997 election (Bernanke et al. 1999).
The second challenge faced by the Bank of England pertained to bank regulation—and in particular, to the increasing prominence of bank instability. The Bank’s regulatory responsibilities evolved quickly in the 1970 and 1980s. During the 1970s, the Bank regulated informally, using ‘‘moral suasion’’ to keep commercial banks in line with its wishes (Penn 1989). However, the ‘‘secondary banking crisis’’ of 1973–75, in which a large number of community banks collapsed as a result of imprudent lending decisions and changing macro- economic conditions, prompted Parliament to grant the Bank formal regulatory responsibility in 1979. The Banking Act of 1979 established an authorization process whereby deposit-taking institutions were re- quired to secure a license from the Bank and undergo mandatory examinations by regulators. Just five years later in 1984, Johnson Matthey Bankers (JMB), a prestigious bank heavily involved in gold trading and real estate investment, became insolvent after a series of lending mishaps (Grady and Weale 1986). The col- lapse of JMB caused great embarrassment for the Bank, and led to another legislative initiative by Parliament: the Banking Act of 1987, which among other things created a bank supervisory board within the Bank and raised the profile of the Bank’s regulatory functions (Hall 1999).
In the wake of these episodes of financial insta- bility, the Bank quickly became a worldwide leader in the development of prudential regulations for com- mercial banks. Bank Governor Eddie George was instrumental in the creation of the 1988 Basel Accord, a capital adequacy standard for internationally active banks in the G-10 industrialized countries. At the same time, however, the large banks in the United Kingdom were finding new ways to take on risk, as manifested by the spectacular collapse of Barings in 1995. The country’s numerous smaller banks—in- cluding some 60 ‘‘building societies’’ which catered to local communities—were also struggling with the economic recession of the early 1990s (Logan 2001).
When the Labour Government came to power in 1997, it was clear that the Bank was being pulled in two different directions. On the one hand, the Treasury was adamant about reining in inflation through tighter monetary policy and increased credibility; on the other, the Bank itself was being held increasingly accountable for instability in the banking sector. Chancellor Gordon Brown’s strategy simultaneously addressed both challenges. In transferring operational
responsibility for monetary policy from the Treasury to the Bank, Brown enhanced the Bank’s independ- ence, bolstered the credibility of its inflation-fighting mandate, and signaled to the public a strong commit- ment to price stability. At the same time, by trans- ferring bank regulatory responsibility from the central bank to the newly created Financial Services Authority (FSA), the Government removed the potential bias that could compromise the Bank’s inflation-fighting mandate. In terms of fighting inflation, the combina- tion of increased independence and the removal of regulatory responsibility appears to have been suc- cessful: from 1999 to 2005, the United Kingdom’s average annual inflation rate was 2.38% and did not exceed 3% in any individual year.28
An alternative but complementary rationale for the removal of regulatory responsibility from the Bank was the increasing prominence of financial conglom- erates in London, and the corresponding challenges of orchestrating multiple functional regulators—including the Bank of England, the Securities and Investments Board (SIB), the Personal Investment Authority, and many others. Before the creation of the FSA, a large financial institution with banking, securities, and insurance arms could find itself subject to regulation and supervision from a welter of regulators, some of them with conflicting objectives. The new FSA, however, is a ‘‘super regulator’’ in that all forms of financial regulation—from insurance supervision to the prevention of securities fraud—fall within its jurisdiction. Proponents of the new agency argue that it leads to increased efficiency through economies of scale and that financial institutions are generally relieved to work with only one regulator with a unified set of rules and procedures (Briault 1999). According to this line of reasoning, the Government removed the Bank’s regulatory responsibilities in order to create a new, coherent financial regulator to manage London’s new financial conglomerates. It would have been impractical to grant the Bank the full respon- sibility for all of these regulatory functions. Some regulators also note that the Bank would be deemed far too powerful in the eyes of the Treasury if it was granted both political independence and a host of new regulatory responsibilities.29
While this alternative explanation for the change in the Bank’s mandate is attractive for its simplicity,
28Of course, it is impossible to determine whether the drop in inflation is attributable to the increase in independence, the removal of regulatory authority, or some other exogenous variable.
29We thank Andrew Bailey at the Bank of England for this point.
676 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
it cannot explain the timing of the institutional change. Unlike the United States, in which the Graham-Leach-Bliley Act of 1999 relaxed a long-held prohibition against financial conglomerates, the United Kingdom’s experience with financial con- glomerates began in earnest in the 1970s and 1980s (Maycock 1986). Concerns about the inefficiency of multiple regulators and the blurring of lines between different sectors of the financial services industry therefore would have dictated an institutional change long before 1997.
Ultimately, the Bank of England case demon- strates the potential conflict between bank regulation and monetary policy. In 1997, the incoming Labour Government was eager to establish its inflation-fight- ing credentials after more than a decade of poor macroeconomic performance. The establishment of an independent central bank was an important first step, but the United Kingdom’s history of bank instability made it clear that the Bank’s responsibility for bank regulation could compromise its efforts to fight inflation. The transfer of bank regulatory authority to the FSA proved to be the capstone of the Government’s reengineering of monetary policy- making in the 1990s.
Conclusion
In this paper, we argue that a central bank’s institu- tional mandate—specifically, whether or not it also regulates banks—is an important determinant of its monetary policy decisions. The institutional locus of regulatory authority influences how the central bank resolves the tension between price stability and bank stability. All else equal, we argue that a central bank without regulatory responsibility is more likely to enact tighter monetary policies geared solely toward maintaining price stability. Analyzing data from 23 industrialized countries from 1975 to 1999, we find strong support for our argument: inflation rates have been significantly lower, on average, in countries where the central bank and the bank regulator are separate agencies. However, this effect of a central bank’s mandate on inflationary outcomes is condi- tional on both the choice of exchange rate regime and the size of the domestic banking sector. Under float- ing rates, the central bank’s regulatory responsibilities play a significant role in shaping its monetary policy choices, but only when the domestic banking sector is sufficiently large. In contrast, a central bank operat- ing under fixed exchange rates will pursue price
stability-oriented policies, regardless of its regulatory mandate or the size of the domestic financial sector. Thus, a central bank’s regulatory mandate influences monetary policy outcomes, but only under certain institutional and economic conditions.
These findings have several key policy and re- search implications. Above all, they suggest that our understanding of the political economy of central banks and monetary policymaking has been limited by an exclusive focus on independence as the primary institutional characteristic affecting central banks’ behavior. Indeed, as our results indicate, other facets of a central bank’s mandate—in particular, the scope of its regulatory responsibilities—also play a signifi- cant role in shaping its monetary policymaking decisions. This finding suggests that politicians should be cautious in assuming that granting the central bank operational independence will automati- cally enhance the credibility of its commitment to price stability, thereby resulting in lower average inflation. In many cases, central bank independence may indeed provide an ‘‘institutional fix’’ for the problem of high inflation. Whether or not this is the case, however, depends critically on the broader institutional mandate of the central bank, as well as on the structure of other monetary institutions (e.g., the choice of exchange rate regime) and the charac- teristics of the domestic financial system.
In addition, our findings about the influence of a central bank’s mandate on its policy choices shed light on a broader issue concerning the delegation of authority to bureaucratic agencies. In particular, they suggest that the behavior of agents in delegation situations depends critically not only on the type of tasks that the principal assigns them, but also on the number of tasks that they have been delegated (Dewatripont, Jewitt, and Tirole 2000). Specifically, they illustrate how politicians’ choice to delegate multiple tasks to a single bureaucratic agency can have significant consequences for policy outcomes. This is especially likely in cases where these policy goals (e.g., price stability and bank stability) are potentially conflicting, and bureaucrats have a pri- mary ‘‘tool’’ (e.g., interest rate manipulation) by which to achieve them. Thus, while it may be politically and administratively easier for politicians to assign multiple tasks to an existing bureaucracy rather than incurring the costs of creating a new agency, the bundling of multiple tasks within a single agency may produce unintended policy outcomes.
Finally, this paper also suggests an additional question for future research: what are the effects of institutional design on regulatory policymaking? Our
financial regulation, monetary policy, and inflation in the industrialized world 677
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
argument has focused on the monetary policy im- plications of the institutional design of central banks, but there may be substantive differences in bank regulation as well. Are central banks more stringent bank regulators than stand-alone regulatory agencies? This question is beyond the scope of the literature on the political economy of monetary policy, but none- theless important for the expanding literature on comparative financial regulation (e.g., Rosenbluth and Schaap 2003). Exploring the relationship be- tween the mandates of regulatory agencies and key policy outcomes (e.g., financial stability, bank profit- ability) would further enhance our understanding of the ways in which institutions shape economic policy outcomes.
Acknowledgments
We thank Rob Franzese, Jeff Frieden, Alexandra Guisinger, Steven Hall, Lisa Martin, Bumba Mukher- jee, David Nickerson, Will Phelan, Naunihal Singh, and three anonymous reviewers for helpful com- ments and suggestions. An earlier version of this paper was presented at the 2007 annual meeting of the Midwest Political Science Association.
Manuscript submitted 17 April 2007 Manuscript accepted for publication 11 August 2007
References
Abrams, Richard K., and Michael W. Taylor. 2000. ‘‘Issues in the Unification of Financial Sector Supervision.’’ IMF Working Paper 213, December.
Adolph, Chris. 2005. The Dilemma of Discretion: Career Ambi- tions and the Politics of Central Banking. Ph.D. Dissertation, Harvard University.
Armingeon, Klaus, Philipp Leimgruber, Michelle Beyeler, and Sarah Menegale. 2005. Comparative Political Data Set, 1960– 2003. Institute of Political Science, University of Berne.
Banaian, King, Richard Burdekin, and Thomas D. Willett. 1995. ‘‘On the Political Economy of Central Bank Independence.’’ In Monetarism and the Methodology of Economics: Essays in Honor of Thomas Mayer, ed. Kevin D. Hoover, and Steven Sheffrin. Aldershot, England: Edward Elgar, 178–97.
Barro, Robert, and David B. Gordon. 1983. ‘‘Rules, Discretion, and Reputation in a Model of Monetary Policy.’’ Journal of Monetary Economics 12 (1): 101–21.
Barth, James R., Gerard Caprio Jr., and Ross Levine. 2003. Bank Regulation and Supervision database. The World Bank. http:// www.worldbank.org/research/projects/bank_regulation.htm (February 25, 2008).
Barth, James. R., Daniel E. Nolle, Triphon Phumiwasana, and Glenn Yago. 2002. ‘‘A Cross-Country Analysis of the Bank Supervisory Framework and Bank Performance.’’ Economic
Policy Analysis Working Paper 2002-2, U.S. Comptroller of the Currency.
Basel Committee. 1997. ‘‘Core Principles for Effective Banking Supervision.’’ Bank for International Settlements. September.
Bearce, David H. 2003. ‘‘Societal Preferences, Partisan Agents, and Monetary Policy Outcomes.’’ International Organization 57 (2): 373–410.
Beck, Nathaniel. 2004. ‘‘Longitudinal (Panel and Time Series Cross-Section) Data.’’ Unpublished manuscript. http://www. nyu.edu/gsas/dept/politics/faculty/beck/beck_home.html (February 25, 2008).
Beck, Nathaniel, and Jonathan N. Katz 1995. ‘‘What to do (and not to do) with Time-Series Cross-Section Data.’’ American Political Science Review 89 (3): 634–47.
Beck, Thorsten, Asli Demirgücx-Kunt, and Ross Levine. 2000. ‘‘A New Database on Financial Development and Structure.’’ World Bank Economic Review 14 : 597–605.
Bernanke, Ben S., Thomas Laubach, Frederic Mishkin, and Adam Posen. 1999. Inflation Targeting: Lessons from the International Experience. Princeton: Princeton University Press.
Bernhard, William. 1998. ‘‘A Political Explanation of Variations in Central Bank Independence.’’ American Political Science Review 92 (2): 311–27.
Bernhard, William, and David Leblang, 1999, ‘‘Democratic Institutions and Exchange Rate Commitments.’’ International Organization 53 (1): 71–97
Bernhard, William, J. Lawrence Broz, and William Roberts Clark. 2002. ‘‘The Political Economy of Monetary Institutions.’’ International Organization 56 (4): 693–723.
Boorman, Jack, Timothy Lane, Marianne Schulze-Ghattas, Ales Bulir, Atish R Ghosh, Javier Hamann, Alexandros Mourmou- ras, and Steven Phillips. 2000. ‘‘Managing Financial Crises: The Experience in East Asia.’’ IMF Working Paper 00/107.
Bordo, Michael, Barry Eichengreen, Daniela Klingebiel, and Maria Soledad Martinez-Peria. 2000. ‘‘Is the Crisis Problem Growing More Severe?’’ Unpublished manuscript. Rutgers University. http://econweb.rutgers.edu/bordo/Crisis_Problem_ text.pdf (February 25, 2008).
Brambor, Thomas, William Roberts Clark, and Matt Golder. 2006. ‘‘Understanding Multiplicative Interaction Models: Improving Empirical Analyses.’’ Political Analysis 14 (1): 63–82.
Briault, Clive. 1999. ‘‘The Rationale for a Single National Financial Services Regulator.’’ Occasional Paper, Financial Services Authority (U.K.).
Broz, J. Lawrence. 2002. ‘‘Political System Transparency and Monetary Commitment Regimes.’’ International Organization 56 (4): 861–87.
Capie, Forrest, Charles Goodhart, and Norbert Schnadt. 1994. ‘‘The Development of Central Banking.’’ In The Future of Central Banking, ed. Forrest Capie, Charles Goodhart, Stanley Fischer, and Norbert Schnadt. Cambridge: Cambridge Uni- versity Press, 1–231.
Caprio, Gerard, and Daniela Klingebiel. 2003. ‘‘Episodes of Systemic and Borderline Financial Crises.’’ World Bank Re- search dataset. http://econ.worldbank.org/WBSITE/EXTERNAL/ EXTDEC/EXTRESEARCH/0,contentMDK:20699588~pagePK: 64214825~piPK: 64214943~theSitePK: 469382,00.html (February 25, 2008).
Cargill, Thomas F. 1989. ‘‘Central Bank Independence and Regulatory Responsibilities: The Bank of Japan and the Federal Reserve.’’Monograph 1989–2, Salomon Brothers
678 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
Center for the Study of Financial Institutions, Leonard N. Stern School of Business, New York University.
Chinn, Menzie, and Hiro Ito. 2006. ‘‘What Matters for Financial Development: Capital Controls, Institutions, and Interac- tions.’’ Journal of Development Economics 81 (1): 163–92.
Clark, William Roberts. 2002. ‘‘Partisan and Electoral Motivations and the Choice of Monetary Institutions Under Fully Mobile Capital.’’ International Organization 56 (4): 725–49.
Cukierman, Alex. 1991. ‘‘Why Does the Fed Smooth Interest Rates?’’ In Monetary Policy on the 75th Anniversary of the Federal Reserve System, ed. Michael T. Belongia. Boston: Kluwer Academic Publishers, 111–47.
Cukierman, Alex. 1992. Central Bank Strategy, Credibility, and Independence. Cambridge: MIT Press.
Cukierman, Alex, Steven Webb, and Bilin Neyapti. 1992. ‘‘Meas- uring the Independence of Central Banks and Its Effects on Policy Outcomes.’’ The World Bank Review 6 (3): 353–98.
Demirgücx-Kunt, Asli, Baybars Karacaovali, and Luc Laeven. 2005. ‘‘Deposit Insurance around the World: A Comprehensive Database.’’ Policy Research Working Paper 3628, World Bank, Washington.
Dewatripont, Mathias, Ian Jewitt, and Jean Tirole. 2000. ‘‘Multi- task Agency Problems: Focus and Task Clustering.’’ European Economic Review 44 (4–6): 869–77.
Di Noia, Carmine, and Giorgio Di Giorgio. 1999. ‘‘Should Banking Supervision and Monetary Policy Tasks Be Given to Different Agencies?’’ International Finance 2 (3): 361–78.
Eichengreen, Barry. 1999. Toward a New International Financial Architecture: A Practical Post-Asia Agenda. Washington: In- stitute for International Economics.
Federal Deposit Insurance Corporation (FDIC). 1998. ‘‘A Brief History of Deposit Insurance in the United States.’’ Interna- tional Conference on Deposit Insurance, Washington.
Ferejohn, John, and Charles R. Shipan 1990. ‘‘Congressional Influence on Bureaucracy.’’ Journal of Law, Economics, and Organization 6: 1–21.
Franzese, Rob 1999. ‘‘Partially Independent Central Banks, Politically Responsive Governments, and Inflation.’’ American Journal of Political Science 43 (3): 681–706.
Frieden, Jeffry. 1991. ‘‘Invested Interests: The Politics of National Economic Policies in a World of Global Finance.’’ Interna- tional Organization 45 (4): 425–51.
George, Edward. 1994. ‘‘The Bank of England—Objectives and Activities.’’ Speech given to the Capital Market Research Group, Frankfurt University, 5 December.
Glick, Reuven, and Michael Hutchison, 1999. ‘‘Banking and Currency Crises: How Common Are Twins?’’ Pacific Basin Working Paper Series 99–07, Federal Reserve Bank of San Francisco.
Goodhart, Charles. 2001. ‘‘The Organizational Structure of Bank- ing Supervision.’’ In Financial Stability and Central Banks, eds. Richard Brealey, Alastair Clark, Charles Goodhart, Juliette Healey, Glenn Hoggarth, David T. Llewellyn, Chang Shu, Peter Sinclair, and Farouk Soussa. London: Routledge, 79–106.
Goodhart, Charles, and Dirk Schoenmaker. 1993. ‘‘Institutional Separation between Supervisory and Monetary Agencies.’’ Special Paper No. 52, LSE Financial Markets Group, April.
Goodhart, Charles, and Dirk Schoenmaker. 1995. ‘‘Should the Functions of Monetary Policy and Banking Supervision Be Separated?’’ Oxford Economic Papers 47 (4): 539–60.
Grady, John, and Martin Weale. 1986. British Banking, 1960–85. London: Macmillan.
Grilli, Vittorio, Donato Masciandaro, and Guido Tabellini. 1991. ‘‘Political and Monetary Institutions and Public Financial Policies in the Industrial Countries.’’ Economic Policy 6 (2): 341–92.
Hall, Maximilian J.B. 1999. Handbook of Banking Regulation and Supervision in the United Kingdom. Cheltenham, UK: Edward Elgar.
Hall, Peter A., and Rob Franzese. 1998. ‘‘Mixed Signals: Central Bank Independence, Coordinated Wage Bargaining, and European Monetary Union.’’ International Organization 52 (3): 505–36.
Hallerberg, Mark. 2002. ‘‘Veto Players and the Choice of Monetary Institutions.’’ International Organization 56 (4): 775–802.
Healey, Juliette. 2001. ‘‘Financial Stability and the Central Bank: International Evidence.’’ In Financial Stability and Central Banks, eds. Richard Brealey, et al . London: Routledge, 19–78.
Henning, C. Randall. 1994. Currencies and Politics in the United States, Germany, and Japan. Washington: Institute for Inter- national Economics.
Ho, Daniel, Kosuke Imai, Gary King, and Elizabeth Stuart. 2007. ‘‘Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference.’’ Political Analysis 15 (3): 199–236.
Kapstein, Ethan B. 1989. ‘‘Resolving the Regulator’s Dilemma: International Coordination of Banking Regulations.’’ Interna- tional Organization 43 (2): 323–47.
Keefer, Philip, and David Stasavage. 2002. ‘‘Checks and Balances, Private Information, and the Credibility of Monetary Com- mitments.’’ International Organization 56 (4): 751–74.
Llewellyn, David. 1999. ‘‘The Economic Rationale for Financial Regulation.’’ Occasional Paper No. 1. London: Financial Services Authority. http://www.fsa.gov.uk/pubs/occpapers/ OP01.pdf. (February 25, 2008).
Leuven, Edwin, and Barbara Sianesi. 2003. ‘‘PSMATCH2: Stata Module to Perform Full Mahalanobis and Propensity Score Matching, Common Support Graphing, And Covariate Im- balance Testing.’’ http://ideas.repec.org/c/boc/bocode/s432001. html. (February 25, 2008).
Logan, Andrew. 2001. ‘‘The United Kingdom’s Small Banks Crisis of the Early 1990s: What Were the Leading Indicators of Failure?’’ Bank of England Working Paper 139.
Maxfield, Sylvia. 1997. Gatekeepers of Growth: The International Political Economy of Central Banking in Developing Countries. Princeton: Princeton University Press.
Maycock, James. 1986. Financial Conglomerates: The New Phenomenon. Aldershot: Gower Publishing.
McCubbins, Mathew D., and Thomas Schwartz. 1984. ‘‘Con- gressional Oversight Overlooked: Police Patrols versus Fire Alarms.’’ American Journal of Political Science 28 (1): 165–79.
Mosley, Layna. 2003. Global Capital and National Governments. New York: Cambridge University Press.
North, Douglass C., and Barry R. Weingast 1989. ‘‘Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth Century England.’’ Journal of Economic History 49 (4): 803–32.
Oatley, Thomas. 1999. ‘‘How Constraining is Capital Mobility? The Partisan Hypothesis in an Open Economy.’’ American Journal of Political Science 43 (4): 1003–27.
financial regulation, monetary policy, and inflation in the industrialized world 679
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
O’Mahony, Angela. 2007. ‘‘Escaping the Ties that Bind: Exchange Rate Choice Under Central Bank Independence.’’ Compara- tive Political Studies 40 (7): 808–31.
Organization for Economic Cooperation and Development. 1992. Banks under Stress. Paris: OECD.
Peek, Joe, Eric S Rosengren, and Geoffrey M.B. Tootell 1999. ‘‘Is Bank Supervision Central to Central Banking?’’ Quarterly Journal of Economics 114 (2): 629–53.
Persson, Torsten, and Guido Tabellini. 2000. Political Economics: Explaining Economic Policy. Cambridge: MIT Press.
Polillo, Simone, and Mauro F. Guillén, 2005. ‘‘Globalization Pressures and the State: The Global Spread of Central Bank Independence.’’ American Journal of Sociology 110 (6): 1764– 1802.
Posen, Adam. 1995. ‘‘Declarations Are Not Enough: Financial Sector Sources of Central Bank Independence.’’ NBER Macro- economics Annual. Cambridge: MIT Press.
Rosenbluth, Frances M., and Ross Schaap. 2003. ‘‘The Domestic Politics of Banking Regulation.’’ International Organization 57 (2): 307–36.
Simmons, Beth A., and Daniel J. Hopkins 2005. ‘‘The Constrain- ing Power of International Treaties: Theory and Methods.’’ American Political Science Review 99 (4): 623–31.
Singer, David Andrew. 2004. ‘‘Capital Rules: The Domestic Politics of International Regulatory Harmonization.’’ Interna- tional Organization 58 (3): 531–65.
Singer, David Andrew. 2007. Regulating Capital: Setting Standards for the International Financial System. Ithaca: Cornell Uni- versity Press.
Stigler, George J. 1971. ‘‘The Theory of Economic Regulation.’’ Bell Journal of Economics and Management Science 2: 3–21.
Tomz, Michael, Jason Wittenberg, and Gary King. 2003. CLAR- IFY: Software for Interpreting and Presenting Statistical Results.
Version 2.1. Stanford University, University of Wisconsin, and Harvard University. http://gking.harvard.edu. (February 25, 2008).
Tuya, Jose, and Lorena Zamalloa. 1994. ‘‘Issues on Placing Banking Supervision in the Central Bank.’’ In Frameworks for Monetary Stability, eds. Tomas J.T. Balinoand Carlo Cottarelli. Washington: International Monetary Fund, 663–90.
Walter, Stefanie, and Thomas Willett. 2007. ‘‘Delaying the Inevitable: A Political Economy Model of Currency Defenses and Capitulation.’’ Presented at the annual meeting of the Midwest Political Science Association\.
Weingast, Barry. 1984. ‘‘The Congressional-Bureaucratic System: A Principal Agent Perspective (With Applications to the SEC).’’ Public Choice 44 (1): 147–91.
Weingast, Barry, and Mark Moran. 1983. ‘‘Bureaucratic Discre- tion or Congressional Control? Regulatory Policymaking by the Federal Trade Commission.’’ Journal of Political Economy 91 (5): 765–800.
Woolley, John T. 1984. Monetary Politics: The Federal Reserve and the Politics of Monetary Policy. Cambridge: Cambridge Uni- versity Press.
World Bank. World Development Indicators. Electronic Database.
Mark S. Copelovitch is assistant professor of political science and public affairs, University of Wisconsin-Madison, Madison, WI 53706. David Andrew Singer is assistant professor of political science, Massachusetts Institute of Technology, Cam- bridge, MA 02139.
680 mark s. copelovitch and david andrew singer
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:09 AM All use subject to JSTOR Terms and Conditions
Financial Stability and Central Banks.pdf
Financial Stability and Central Banks by Juliette Healey; Peter Sinclair Review by: Philip Arestis Eastern Economic Journal, Vol. 31, No. 1 (Winter, 2005), pp. 141-143 Published by: Palgrave Macmillan Journals Stable URL: http://www.jstor.org/stable/40326327 .
Accessed: 06/02/2015 10:57
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve and extend access to Eastern Economic Journal.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:57:02 AM All use subject to JSTOR Terms and Conditions
BOOK REVIEWS 141
BOOK REVIEWS
Financial Stability and Central Banks. Edited by Juliette Healey and Peter Sinclair. London and New York: Routledge 2001. pp. 254. $100.00, ISBN 0-415-25775-1 (hard- back); $36.95, ISBN 0-415-25776-X (paperback).
Philip Arestis Jerome Levy Economics Institute
Recent developments in monetary policy emphasise two complementary aspects: the pursuit of price stability and the maintenance of financial stability. This volume concentrates on the second while recognizing that the two aspects "go hand in hand. You cannot really expect to have one without the other" as Bank of England Governor Sir Edward George reminds us in his Forward to this book.
The papers in the volume were all presented at the 7th Central Bank Governors Symposium hosted by the Bank of England. Authors were asked to investigate issues and challenges common to all economies (industrialized, developing and transition). There are eight chapters in all, including a concluding chapter.
Chapter 1, "Financial Stability and Central Banks: An Introduction" by Peter Sinclair, sets the scene and summarizes the papers in the collection. It also poses twelve key questions discussed throughout the collection. Among other issues, it cov- ers financial crises and bank failures, competition and safety in financial markets, and the links between financial stability and monetary policy. These important issues demonstrate that maintaining financial stability is a core responsibility of modern central banking. Indeed, it is no less important than open market operations and the conduct of monetary policy.
Chapter 2, "Financial Stability and the Central Bank: International Evidence" by Juliette Healey, presents survey results of 37 central banks from industrialised, develop- ing and transition countries, and compares the responsibilities and tools for achieving "financial stability" in different countries. The survey results indicate that central banks in smaller and poorer countries exercise a larger range of functions. Healy also explores the question of whether a central bank should carry out regulatory and supervisory functions in addition to monetary policy. The evidence here suggests an inverse relationship between central bank independence and regulatory or supervisory functions.
Chapter 3, "The Organisational Structure of Banking Supervision" by Charles Goodhart, deals with bank supervision and its organizational structure. Goodhart argues that industrialized countries should avoid giving independent central banks excessive power and that supervisory functions should reside outside the central bank. Close contact between the central bank and the supervisory body is paramount, how- ever. The main reason for this is that an independent central bank with supervisory powers might be a too powerful institution within a democratic context. In contrast, for developing and transition countries, supervision and regulation should stay within central banks. This, Goodhart argues, is likely to result in more independent, better funded, and therefore more reliable and expert central banks.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:57:02 AM All use subject to JSTOR Terms and Conditions
142 EASTERN ECONOMIC JOURNAL
Chapter 4, "Alternative Approaches to Regulation and Corporate Governance in Financial Firms" by David Llewellyn, concerns the structure of bank supervision. What matters, according to Llewellyn, is not so much regulation per se, but the regu- lation regime. Emphasis should be placed not on just one element of regulation, but on all of them and especially on their interaction. The optimum combination of the key elements changes over time and across firms. This limits what regulation and supervision of banks and financial institutions can achieve in the real world.
Chapter 5, "Bank Capital Requirements and the Control of Bank Failure" by Richard Brealey, discusses the role of capital requirements in reducing bank failures in the UK and the U.S. Brealey argues that the Basel system of weighting credit risk con- tains several weaknesses. Diversification of the loan portfolio is ignored, and classifi- cation of loans is treated in a broad-brush manner. Revisions may tackle the second weakness, but the first may be less tractable. Ultimately, banks with deposit insur- ance need to remain strongly regulated institutions.
Chapter 6, "Crisis Management, Lender of Last Resort and the Changing Nature of the Banking Industry" by Glen Hoggarth and Farouk Soussa, concentrates on the involvement of central banks in crisis management and on their role as lenders of last resort. It argues that more collaboration and coordination between banks and other financial institutions within and across countries has emerged as a key element in this respect. It describes how the changing nature of the banking industry (especially the blurring distinction between banks and non-banks) requires adapting regulations. This is especially true of the lender-of-last-resort function. Current Basel proposals recognize this, and emphasize the need for market discipline and supervision of bank management systems.
Chapter 7, "International Capital Movements and the International Dimension of Financial Crises" by Peter Sinclair and Chang Shu, concerns international crises. It begins with a sensible summary of the benefits and drawbacks of international capital movements. Capital flows can increase national income, provide opportunities for smoothing consumption, help diversify risk and intertemporal trade, as well as accommo- date current account imbalances. Capital controls may be helpful temporarily but there are undeniable disadvantages associated with them. Surveying the literature on capital controls produces a number of interesting insights: there is no effective single capital control measure; capital controls are effective, albeit only temporarily; to be effective, capital controls should be high and comprehensive, and accompanied by suitable macroeconomic policy; ultimately, it is argued, there are no alternatives to prudential regulation and prompt action.
In his concluding comments Alastair Clark highlights key policy issues and looks at the future role of central banks. Five issues are at the center of current discussions on financial stability: transparency, the use of information, the tension between glo- bal business and national jurisdictions, surveillance, and crisis management. Preserv- ing financial stability will be of vital importance in the future.
This set of papers is worth reading. It explores the rapidly changing monetary and financial environment, and discusses a range of policy responses that might be appro- priate during times of financial crises. This collection is particularly valuable because it recognizes that global financial developments have created problems, because it
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:57:02 AM All use subject to JSTOR Terms and Conditions
BOOK REVIEWS 143
carefully analyzes these problems, because it sets forth possible solutions that are not always apparent or easy. Overall, it is a rich and stimulating collection. However, while it purports to cover a wide range of countries, the focus tends to be on industrial- ized countries. A more balanced approach would have improved this otherwise interest- ing and informative collection.
Evolution and Path Dependence in Economic Ideas: Past and Present Edited by Pierre Garrouste and Stavros Ioannides. Cheltenham: Edward Elgar, 2001. pp. 256. $90.00. ISBN 1840640812.
Mark Setterfield Trinity College, Connecticut
Mainstream economics is timeless, analyzing economic outcomes as end states that are reached independently of the paths taken towards them. For Garrouste and Ioannides, this explains the mainstream's disdain for the history of economic thought and its focus on the genesis and development of economic ideas. The essential theme of this book, which compiles papers presented at the 1997 meetings of the European Association for Evolutionary Political Economy, is that both economic outcomes and economic ideas must be understood as arising from path- dependent processes.
Although most contributions in the volume focus on the history of economic thought, two chapters directly address the question of how path dependence and eco- nomic evolution should be conceived and applied. Paul David reconsiders the econom- ics of path dependence and its public policy implications in light of some recent criti- cisms, primarily by Liebowitz and Margolis [1995] . David shows that his critics errone- ously focus on an outcome that may be associated with path dependence (the ineffi- ciency of economic outcomes) rather than on the intrinsic nature and properties of path dependence itself. His critics seem to be motivated by a normative advocacy of unfettered private behavior rather than a desire better to understand economic dynam- ics, having decided that the real agenda of those interested in path dependence is increased government intervention. Ultimately, they succeed only in diverting atten- tion away from the importance of studying economic dynamics.
John Foster addresses the association between economic competition and biologi- cal natural selection, arguing that frequently the former is ill-defined and the latter ill-applied. By tracing changes in the concept of economic evolution over time, Foster identifies the essential difference between Smithian and subsequent thinking: the former is concerned with acts of creation rather than selection between given alterna- tives. He argues that Smithian thinking has resurfaced in recent neo-Schumpeterian work, and shows how this can be used as a springboard for developing a concept of economic evolution that is suited to the properties of social systems.
Some of the history of thought contributions focus on the path dependence of economic thinking, while others address path dependence in the history of economic thought. Among the former, Francisco Lougá examines the correspondence between
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:57:02 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 141
- p. 142
- p. 143
- Issue Table of Contents
- Eastern Economic Journal, Vol. 31, No. 1 (Winter, 2005), pp. 1-158
- Front Matter
- Is the Human Development Index Redundant? [pp. 1-5]
- The Determinants of Shirking: Analysis and Evidence on Job Loser Unemployment [pp. 7-21]
- Expense Preference and Student Achievement in School Districts [pp. 23-44]
- Symposium: Coordination Failure in Macroeconomics
- Coordination Failure in Macroeconomics: An Overview [pp. 46-54]
- Do Sunspots Reflect Consumer Confidence? An Empirical Investigation [pp. 55-73]
- The Index of Leading Economic Indicators as a Source of Expectational Shocks [pp. 75-95]
- Coordination, Fragility, High-Powered Money, and the Liquidity Trap: A "Tobinesque" Parable [pp. 97-106]
- Estimation and Identification of Structural Parameters in the Presence of Multiple Equilibria [pp. 107-130]
- Economics Forum
- The New Classical Counter-Revolution: A False Path for Macroeconomics [pp. 131-140]
- Book Reviews
- Review: untitled [pp. 141-143]
- Review: untitled [pp. 143-146]
- Review: untitled [pp. 146-148]
- Review: untitled [pp. 148-150]
- Review: untitled [pp. 151-153]
- Review: untitled [pp. 153-155]
- Review: untitled [pp. 155-158]
- Back Matter
FINANCIAL STABILITY AND MONETARY POLICY INTERNATIONAL SURVEILANCE.pdf
FINANCIAL STABILITYAND MONETARY
POLICY: NEED FOR INTERNATIONAL
SURVEILLANCE
Gary Hufbauer* and Daniel Danxia Xie**
ABSTRACT
In this article, we propose a new monetary framework that defines a broader
set of assets, De Facto Money (DFM), as the benchmark for improving
financial stability. DFM is defined as traditional monetary aggregates plus
other liquid assets such as stocks and bonds. Empirical evidence for the
USA, other Organisation for Economic Co-operation and Development
countries, and a few emerging countries lends strong support for the con-
nection between exceptionally fast growth of DFM and subsequent financial
instability. We recommend several potential policy instruments to implement
the new monetary framework. Due to cross-country spillovers from national
financial crises, we suggest that international surveillance will be necessary to
monitor DFM and thus the underlying conditions for financial stability in
major countries. We argue that the International Monetary Fund is the ideal
institution to carry out this task in terms of its reputation and expertise.
I. INTRODUCTION
In the wake of the Great Crisis of 2008, regulatory proposals have empha-
sized the micro-prudential space: greater bank liquidity and more bank
capital, strict supervision of too-big-to-fail financial firms, ‘living wills’,
new resolution mechanisms, compensation that tracks risk. Proposed
reforms also address troublesome problems created by cross-border financial
institutions and international financial networks.
An important ingredient is missing from this menu. Micro-prudential
reforms that focus on financial firms are essential, but it is equally important
* Reginald Jones Senior Fellow, Peterson Institute for International Economics, Washington,
DC. E-mail: [email protected].
** Research Analyst, Peterson Institute for International Economics, Washington, DC when this
article was written. E-mail: [email protected].
The views expressed are the opinions of the authors, not necessarily the views of the Institute.
Journal of International Economic Law 13(3), 939–953 doi:10.1093/jiel/jgq035.
Journal of International Economic Law Vol. 13 No. 3 � Oxford University Press 2010, all rights reserved
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
to reform the macro-economic behavior of central banks. New evidence
shows a strong link between loose monetary policy and greater risk-taking
by banks and other financial institutions. Regulatory failures in the USA
and Europe may have been the proximate cause of the crisis, but massive
monetary policy errors set the stage.
In this article, we suggest that the monetary policy framework should be
expanded to take explicit account of financial stability as a policy objective
for central banks—a policy objective in addition to inflation and output.
To fulfill this objective, we commend a new measure of monetary aggregates,
De Facto Money (DFM). DFM attempts to measure the quantity of a broad
set of liquid assets, as a useful indicator of systematic financial risk in the
economy. Empirical evidence for the USA indicates that DFM can be useful
for identifying asset booms and moderating asset busts.
We also present cross-country evidence for the 2000s that shows the
connection between DFM and the international background of the Great
Crisis. For that exercise, a panel data set of 18 countries explores correl-
ations between the DFM dimension of monetary policy and current account
balances. The panel regressions indicate that loose monetary policies—
defined in terms of their wealth effects—are robustly related to current
account deficits. In their summit meetings, G-20 countries have stressed
that global imbalances should be curtailed to ward off the next crisis.
It follows that greater emphasis on financial stability, especially in the
boom phase of the economic cycle, should be explored as an instrument
for achieving this policy objective.
Monetary policy has historically belonged squarely in the realm of national
sovereignty, and has not been a subject of international surveillance. As US
Treasury Secretary John Connally famously said, when the Bretton Woods
system of fixed exchange rates was collapsing: ‘‘The dollar is our money
and your problem.’’1 As a consequence of the Great Crisis, however, the
downside risks of financial globalization now challenge the conventional
assignment of monetary policy strictly to national authorities. Cross-country
spillover effects call for surveillance of national monetary policies, especially
for ‘systemically important countries’.
II. BOOMING WEALTH AND PRIVATE RISK-TAKING
We are certainly not the first to critique the role of loose monetary policy,
during the period 2002–05, for setting the stage of the Great Crisis. John
B. Taylor was an early and prominent critic of the Federal Reserve. Taylor
1 John Connally made this comment to European finance ministers in 1971.
940 Journal of International Economic Law (JIEL) 13(3)
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
used a simulation exercise to indicate that tighter monetary policy could have
slowed the huge housing boom in the USA.2
In fact, well before the bubble, Bernanke and Gertler,3 and then
Bernanke, Gertler, and Gilchrist proposed a ‘financial accelerator’ model
in which monetary policy pumps up the financial sector.4 These research
papers illustrate how frictions within the credit markets can amplify shocks to
the macroeconomy. In good times, private actors underestimate risk, load up
on assets, and enlarge their debt burdens. In bad times, risk premiums soar,
and deteriorating credit market conditions worsen the economic downturn.
Borio and Zhu emphasized the explicit ‘risk-taking channel’ of monetary
policy—easier money by the central banks, higher risk by private lenders.5
In a similar spirit, Rajan argued that loose monetary policy can induce asset
managers to ‘search for yield’, bending constraints such as contract struc-
tures and other institutional features.6 Cukierman even mentioned Madoff as
an example of outright fraud facilitated by easy access to credit.7
Seminal research by Jiménez et al. showed the impact of low interest rates
on the appetite of Spanish banks for credit risk. Under expansive monetary
policy, Spanish banks relaxed their lending standards and extended loans to
borrowers with weak credit histories.8
2 John B. Taylor, ‘Housing and Monetary Policy’, http://www.stanford.edu/�johntayl/Housing%
20and%20Monetary%20Policy–Taylor–Jackson%20Hole%202007.pdf (visited 5 August
2010). 3 Ben Bernanke and Mark Gertler, ‘Agency Costs, Net Worth, and Business Fluctuations’,
79 American Economic Review 14 (1989). 4 Ben Bernanke, Mark Gertler and Simon Gilchrist, ‘The Financial Accelerator in a Quantitative
Business Cycle Framework’, in John B. Taylor and Michael Woodford (eds), Handbook of
Macroeconomics (Amsterdam: Elsevier, 1999) 1341–93. 5 Claudio Borio and Haibin Zhu, ‘Capital Regulation, Risk-Taking and Monetary Policy: A
Missing Link in the Transmission Mechanism?’, Bank for International Settlements Working
Papers No 268, December 2008, http://www.bis.org/publ/work268.htm (visited 5 August
2010). 6 Raghuram Rajan, ‘Has Financial Development Made the World Riskier?’, http://www.imf.org/
external/np/speeches/2005/082705.htm (visited 5 August 2010). Adrian and Shin carry the
analysis further and explore connections between monetary policy and the cyclical behavior
of the balance sheets of financial institutions. Tobias Adrian and Hyun Song Shin, ‘Money,
Liquidity, and Monetary Policy’, 99 American Economic Review 600 (2009). Tobias Adrian
and Hyun Song Shin, ‘Financial Intermediaries and Monetary Economics’, in Benjamin
Friedman and Michael Woodford (eds), Handbook of Monetary Economics (North Holland,
2010). 7 Alex Cukierman, ‘Reflections on the Crisis and on its Lessons for Regulatory Reform and for
Central Bank Policies’, Journal of Financial Stability (2010), http://www.tau.ac.il/�alexcuk/pdf/
Bocconi-Revised%20&%20Expanded-12-09.pdf (visited 4 November 2010). 8 Gabriel Jiménez et al., ‘Hazardous Times for Monetary Policy: What do Twenty-Three Million
Bank Loans Say About the Effects of Monetary Policy on Credit Risk-Taking?’, Bank of Spain
(Banco de España) Working Papers No 0833, http://www.bde.es/webbde/Secciones/
Publicaciones/PublicacionesSeriadas/DocumentosTrabajo/08/Fic/dt0833e.pdf (visited 5 August
2010). Smaller Spanish banks were found to be more affected by loose monetary policy than
larger ones, at 8, http://www.bde.es/webbde/Secciones/Publicaciones/PublicacionesSeriadas/
DocumentosTrabajo/08/Fic/dt0833e.pdf
Financial Stability and Monetary Policy 941
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
The Bank for International Settlements (BIS) further confirmed the
risk-taking channel of monetary policy, using a large new cross-country
data set.9 Drawing on this data set, Gambacorta found robust evidence
that the default probability for banks increased in countries where interest
rates remained low for a lengthy period prior to the crisis.10 In their latest
research, Altunbas, Gambacorta, and Marques-Ibanez explored a detailed
database of European and American banks, and confirmed that enduring
low interest rates contributed to higher risk-taking behavior.11 To round
out the story, Schularick and Taylor provide long-run historical evidence
of the linkage between fast credit growth and financial crises.12
To summarize, an academic consensus is forming around the idea that
prolonged easy money can set the stage for financial crisis. Of course, central
bankers are quick to proclaim ‘not on my watch’. Bernanke is a prominent
example.13 But these self-serving declamations are not turning the tide of
academic analysis.
III. THE MAINSTREAM MONETARY REGIME
The great inflation of the 1970s and early 1980s turned mainstream econo-
mists and central bankers alike into inflation hawks. New Zealand pioneered
the adoption of inflation targeting (IT) as its monetary policy framework in
1989. Since then, IT has been popularly adopted in both developed and
developing countries. In fact, IT has become the dominant monetary
regime, with various degrees of formality. Developed countries with explicit
targets often identify 2–3% as the desirable range of inflation: the UK uses
2.5%, Korea uses 2.5–3.5%, Sweden uses 2% (�1%) and Spain uses 2%.
With or without explicit targets, central bankers settled on the view that,
9 ‘International Banking and Financial Market Developments’, BIS Quarterly Review,
December 2009, http://www.bis.org/publ/qtrpdf/r_qt0912.htm (visited 5 August 2010). 10 Leonardo Gambacorta, ‘Monetary Policy and the Risk-taking Channel’, BIS Quarterly
Review, Special Features, December 2009, at 43, http://www.bis.org/publ/qtrpdf/
r_qt0912.htm (visited 5 August 2010). 11 Yener Altunbas, Leonardo Gambacorta and David Marques-Ibanez, ‘Does Monetary Policy
Affect Bank Risk-taking?’, BIS Working Papers No 298, March 2010, http://www.bis.org/publ/
work298.htm (visited 5 August 2010). 12 Moritz Schularick and Alan M. Taylor, ‘Credit Booms Gone Bust: Monetary Policy,
Leverage Cycles and Financial Crises, 1870–2008’, National Bureau of Economic Research
Working Paper No 15512, November 2009, http://www.jfki.fu-berlin.de/faculty/economics/
team/persons/schularick/Schularick__Taylor_Credit_Booms_Gone_Bust.pdf (visited 5 August
2010). 13 Speech of Ben Bernanke, ‘Monetary Policy and the Housing Bubble’, Annual Meeting of the
American Economic Association, Atlanta, Georgia, 3 January 2010, http://www.federalreserve
.gov/newsevents/speech/bernanke20100103a.htm (visited 5 August 2010).
942 Journal of International Economic Law (JIEL) 13(3)
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
when inflation was acceptably low, all was well in the economic kingdom. As
Blanchard, Giovanni dell’Ariccia, and Paolo Mauro put the consensus:
we thought of monetary policy as having one target, inflation, and one instrument, the policy rate. So long as inflation was stable, the output gap was likely to be small and stable and monetary policy did its job . . . . There was an increasing consensus that inflation should not only be stable, but very low. . ..14
Bernanke et al. published the first systematic study of IT and found that
IT countries typically enjoyed lower inflation, lower inflation expectations,
and lower nominal interest rates.15 Moreover, temporary shocks to the price
level had a smaller ‘pass-through’ effect on inflation. Based on their findings,
Bernanke et al. list the following principles for an IT framework:
� Explicit central bank commitment to low and stable inflation as the
overriding objective
� Public disclosure of the official inflation target over a defined time
horizon
� Mechanisms that compel the central bank to comply with its commitment
A later study by Truman likewise reported that, after adopting this
‘benevolent’ monetary regime, IT countries generally achieved lower infla-
tion levels and higher gross domestic product (GDP) growth rates.16 Before
the Great Crisis, nearly all empirical studies confirmed the fine performance
of IT.
This happy consensus was shattered by the events of 2008 and 2009.
Buiter for one argues that the narrow focus of monetary policy on IT dis-
tracted the authorities from the equally important goal of financial stability.17
The former Chief Economist of the Central Bank of Iceland, Gudmundsson
confessed: ‘I was a great fan of IT. However, experience has brought
with it a better appreciation of the challenges that come with it. Iceland
was the first country that I am aware of to suspend IT because of a financial
crisis.’18
14 Olivier Blanchard, Giovanni dell’Ariccia and Paolo Mauro, ‘Rethinking Macroeconomic
Policy’, IMF Staff Position Note, SPN/10/03, 12 February 2010, at 3–4, http://www.imf
.org/external/pubs/ft/spn/2010/spn1003.pdf (visited 5 August 2010). 15 Ben Bernanke et al., Inflation Targeting: Lessons from the International Experience (Princeton,
NJ: Princeton University Press, 1999). 16 Section 3 of Edwin Truman, Inflation Targeting in the World Economy? (Washington, DC:
Peterson Institute for International Economics, 2003). 17 Willem Buiter, ‘The Unfortunate Uselessness of Most ‘‘State of the Art’’ Academic Monetary
Economics’, 6 March 2009, http://www.voxeu.org/index.php?q=node/3210 (visited 5 August
2010). 18 Már Gudmundsson, ‘Challenges to Inflation Targeting: Raising Some Issues’, BIS Papers No
51, http://www.bis.org/publ/bppdf/bispap51b.pdf (visited 5 August 2010).
Financial Stability and Monetary Policy 943
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
IV. THE POLITICAL ECONOMY OF FINANCIAL CRISIS
Conservative economists sometimes argue that financial crises are an
efficient way of wiping out weak companies, leading to a post-crisis boost
in productivity. This view does not sit well with a public that must bear the
cost of high unemployment and lost retirement wealth. In democracies, and
even in some autocracies, political pressures to arrest financial distress are
considerable. Central banks will, over time, reflect public preferences. In the
early 1980s, inflation was the major concern, and IT eventually became the
dominant monetary policy. In the early 2010s, financial stability looks like a
bigger concern and this may set the stage for a new approach to monetary
policy. As Feldstein points out:
The Fed has been subject to substantial, and I believe justified, criticism
for its failure to prevent the behavior in those institutions that contributed
directly to the recent financial crisis . . . . In the years before the meltdown
that began in 2006 and 2007, Fed officials frequently indicated that bank
capital was quite adequate and not a cause for concern.19
While central bank independence has come to be regarded as the Holy Grail
of monetary policy, the taxpaying public, unemployed workers, and impover-
ished retirees will all be heard when central bank errors contribute to financial
crises.20 This suggests that Financial Stability—capital F, capital S—will be
added to IT as an explicit policy objective.
V. RETHINKING THE MONETARY POLICY FRAMEWORK
In an unpublished paper, Xie proposed that the new monetary policy frame-
work should reflect a broad definition of aggregate assets to track DFM.21
DFM measures the quantity of all liquid assets in the economy, expressed in
the national currency unit, such as dollars, euros, yen, or yuan. The goal of
Financial Stability can then be defined in terms of moderating fluctuations in
DFM. In comparison with two or three decades ago, most assets are more
tradable and more liquid than they were owing to the creation of new finan-
cial products, advances in information technology, much lower transaction
costs, and a high degree of global financial integration.
19 Martin Feldstein, ‘What Powers for the Federal Reserve?’, 48(1) Journal of Economic
Literature 134 (2010). Feldstein goes on to recite a familiar list of remedies for Wall Street
excess (tighter supervision of the large bank holding companies, etc.). 20 Alberto Alesina and Andrea Stella, ‘The Politics of Monetary Policy’, National Bureau of
Economic Research Working Paper No 15856, April 2010, http://www.nber.org/papers/
w15856.pdf?new_window=1 (visited 5 August 2010). 21 Daniel Danxia Xie, ‘Beyond Inflation Targeting in the Post Crisis World: Towards a New
Monetary Regime’ (Unpublished mimeo, Peterson Institute for International Economics,
Washington, DC, 2009, on file with the authors).
944 Journal of International Economic Law (JIEL) 13(3)
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
Thanks to these innovations, a growing proportion of assets have come to
share the attributes of traditional money (M1, M2, and M3).22 Moreover,
the shadow banking system has grown so large that it rivals the traditional
banking system which is safeguarded by deposit insurance in many countries.
In a financial crisis, the absence of deposit insurance for a shadow bank
(think Northern Rock, Bear Stearns, and AIG) hardly means that the insti-
tution can be allowed to fail, wiping out creditors as well as shareholders.
Our proposal emphasizes Xie’s augmented measure, DFM, in order to
take account of the liquidity characteristics of a broad spectrum of assets,
including traditional bank assets, bonds and shares, real estate, and new
financial instruments (like mortgage-backed securities). Since different cate-
gories of assets have different degrees of liquidity, ideally we would like to
sum up the total of liquidity-weighted assets in the economy. Equation (1)
provides a simple definition of DFM:
DFM ¼ Xn
i¼1
Liquidityi � Vi ð1Þ
In equation (1), Vi denotes the total value of asset category i, and Liquidityi
measures the degree of liquidity of this type of asset, on a scale from zero to
one. For example, bank deposits, certificates of deposit and money market
funds might have a liquidity index of 1.0, bonds might have 0.9, stocks 0.7,
and residential real estate 0.5.
To illustrate the calculation of DFM in the USA, we make the strong
assumption that all assets have the same degree of liquidity. This assumption
gives too much weight to real estate relative to stocks and bonds (even
though real estate is much more liquid now than it was two decades ago).
With this assumption we can borrow the wealth account data constructed by
Jorgenson, and Landefeld and Jorgenson.23 The wealth account is the sum of
the reproducible and tangible assets held by households and non-profit or-
ganizations (NPOs), and the government. It includes an adjustment for the
net international investment position of the USA. Averaged over the decade
from 1990 to 2000, data shows that 39% of total US wealth was in equity,
bonds, and mutual funds, while 22% was in residential housing. Another
22 There are small variations in the definition of monetary aggregates for different countries, but
the definitions are very similar. For the USA, M1: The total of currency in circulation,
checking accounts, and currency held as bank reserves; M2: In addition to M1, bank savings
accounts, money market accounts, retail money market mutual funds, and certificates of
deposit under $100,000; M3: In addition to M2, all other Certificate of Deposit (CD)s,
time deposits over $100,000; institutional money market mutual funds; deposits of
Eurodollars, and repurchase agreements (repos). 23 Dale W. Jorgenson, ‘A New Architecture for the US National Accounts’, 55 (1) Review of
Income and Wealth 1 (2009); Dale W. Jorgenson and J. Steven Landefeld, ‘Blueprint
for Expanded and Integrated U.S. Accounts: Review, Assessment, and Next Steps’, in
Dale W. Jorgenson, J. Steven Landefeld, and William D. Nordhaus (eds), A New Architecture
for the U.S. National Accounts (Chicago: University of Chicago Press, 2006) 13–112.
Financial Stability and Monetary Policy 945
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
20% of US wealth was held by the government. For our purposes, wealth
held by the government is excluded.24
Figure 1 plots the GDP growth rate, the inflation rate (measured by the
Consumer Price Index, or CPI), the M2 growth rate, and the growth rate of
private US wealth for the USA from 1973 to 2009.25 We see no large
anomalies in the GDP growth rate or inflation rate between 1993 and
2006. However, the growth rates of M2 and wealth accelerated after 1993.26
Period averages are summarized in Table 1. From Figure 1 and Table 1,
we can see that during the 2003–06 periods, GDP, CPI, M1 and M2
generally grew at slower speeds than during the 1996–99 period, and the
speeds were comparable to the 1992–95 averages. Note, however, the
unusual feature of the 2003–06 period: wealth growth reached 12.2%
annually, much faster than inflation or GDP. Column (6) shows the
difference between wealth growth rates and inflation rates; column (7)
shows the difference between wealth growth rates and GDP growth rates.
The period 2003–06 stands out among the past three decades. The
large gap between wealth growth and either inflation or GDP growth
indicates conditions leading to a crisis. Meticulous calculations by
Taylor likewise imply a deviation from the eponym Taylor Rule27 from
-24 -22 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8
10 12 14 16 18
19 73
19 74
19 75
19 76
19 77
19 78
19 79
19 80
19 81
19 82
19 83
19 84
19 85
19 86
19 87
19 88
19 89
19 90
19 91
19 92
19 93
19 94
19 95
19 96
19 97
19 98
19 99
20 00
20 01
20 02
20 03
20 04
20 05
20 06
20 07
20 08
20 09
A nn
ua l P
er ce
nt C
ha ng
e
Real GDP CPI M2 Household & NPO Net Wealth
Figure 1. US growth rates of GDP, CPI, M2 and wealth, 1973–2009.
24 Government wealth, such as military bases, national parks and office buildings is highly
illiquid and plays little role in financial volatility. 25 Wealth figures from 2002 to 2009 are based on our own calculations. The source of the
figures for 1973-2001 is the Federal Reserve Board. 26 The growth of M3 also accelerated after 1993, but the Federal Reserve discontinued this M3
series in 2006. 27 The Taylor rule stipulates how much the nominal interest rate should be changed in response
to weighted divergences between: (i) the actual inflation rate and the target inflation rate; and
(ii) the actual level of GDP and the potential level of GDP.
946 Journal of International Economic Law (JIEL) 13(3)
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
2003 to 2006.28 However, excessive monetary ease shows up more clearly in
the gap between wealth growth and the sum of inflation and GDP growth
rates for this period, as shown in column (8).
VI. CROSS-COUNTRY COMPARISONS
Global imbalances are often cited as a factor contributing to the Great Crisis
(e.g. Bernanke), Obstfeld and Rogoff, and Bergsten and Subramanian).29
Obstfeld and Rogoff offer the nuanced view that current account imbalances
‘magnified the ultimate causal factors behind the recent financial crisis’.30
Table 1. US Growth Rates of GDP, CPI, M1, M2, and Wealth, 1977–2006
Periods GDP CPI M1 M2 Wealth Wealth�CPI Wealth�GDP Wealth�
(CPI + GDP)
(1) (2) (3) (4) (5) (6) (7) (8)
1977–81 3.1 9.8 7.4 8.8 12.2 2.3 9.1 �0.8
1982–91 3.0 4.1 7.6 6.8 7.5 3.4 4.5 0.4
1992–95 3.2 2.9 6.1 1.9 6.0 3.1 2.7 �0.1
1996–99 4.4 2.3 �0.1 6.2 11.3 9.0 6.9 4.7
2003–06 3.0 2.9 2.9 5.1 12.2 9.3 9.2 6.4
Source: U.S. Bureau of Economic Analysis, The Federal Reserve Board, and Jorgenson and
Landefeld (2006, 2009).
Note: For the purpose of this table, wealth is defined as the net worth of households and NPO
with an adjustment for the net international investment position.
28 See Taylor, above n 2; John B. Taylor, ‘The Financial Crisis and the Policy Responses: an
Empirical Analysis of What Went Wrong’, National Bureau of Economic Research Working
Paper No 14631, January 2009, at 5, http://www.nber.org/papers/w14631.pdf?new_window=1
(visited 5 August 2010). 29 Speech of Ben Bernanke, ‘Financial Reform to Address Systemic Risk’, The Council on
Foreign Relations, Washington, DC, 10 March 2009, http://www.federalreserve.gov/
newsevents/speech/bernanke20090310a.htm (visited 5 August 2010); Maurice Obstfeld and
Kenneth Rogoff, ‘Global Imbalances and the Financial Crisis: Products of Common Causes’,
http://elsa.berkeley.edu/�obstfeld/santabarbara.pdf (visited 5 August 2010); Fred Bergsten
and Arvind Subramanian, ‘America Cannot Resolve Global Imbalances on Its Own’,
Financial Times, 19 August 2009, http://www.iie.com/publications/opeds/oped.cfm?
ResearchID=1283 (visited 5 August 2010).
For a deeper debate about global imbalances, contrast the views represented by Obstfeld and
Rogoff, and Bergsten, who all think that large current account imbalances are not sustainable
in the long run, with the views supported by Dooley, Folkerts-Landau, and Garber and
Caballero, Farhi, and Gourinchas, as well as Cooper, who all argue that current account
imbalances can be a global equilibrium solution to differing savings and investment prefer-
ences in different countries, and can be sustained for long periods. Maurice Obstfeld and
Kenneth Rogoff, ‘Global Current Account Imbalances and Exchange Rate Adjustments’,
Brookings Papers on Economic Activity 1 (2005) 67–146. Michael Dooley, David
Folkerts-Landau, and Peter Garber, International Financial Stability: Asia, Interest Rates, and
the Dollar (New York: Deutsche Bank Global Research, 2005). Ricardo Caballero, Emmanuel
Farhi, and Pierre-Olivier Gourinchas, ‘An Equilibrium Model of ‘‘Global Imbalances’’ and
Low Interest Rates’ 98 American Economic Review 1 (2008) 358–93. Richard N. Cooper,
‘Living with Global Imbalances’, Brookings Papers on Economic Activity 2 (2007) 91–107. 30 See Obstfeld and Rogoff (2005), ibid, at 10.
Financial Stability and Monetary Policy 947
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
With this view in mind, it is worth exploring the relationship between
global imbalances and monetary policy. We start with a cross-region
comparison for the years before and during the Great Crisis. Due to the
lack of comparable data across countries, we simply use the total value of
bonds, equities, and bank assets (compiled from IMF Global Financial
Stability Reports) as an indicator of changes in DFM. Figure 2 illustrates
the experience in five emerging market areas. Emerging Europe registered
both the highest growth of DFM before the crisis and the deepest recession
afterwards.31 Asset growth in emerging Europe was fueled by the bank lend-
ing channel, which empirically shows the highest correlation to an imminent
financial crisis.
Figure 3 shows the DFM experience of several developed countries from
2002 to 2008. Generally, the developed countries exhibit lower DFM vola-
tility and growth rates than emerging markets. While Greece was the cham-
pion in terms of DFM volatility and growth, Ireland, Spain, and the UK all
experienced significant DFM growth before the crisis.
Our data set for 18 Organisation for Economic Co-operation and
Development (OECD) countries includes as one crucial variable DFM, for
which our proxy measure is the total value of bonds, equities, and bank
assets (compiled from IMF Global Financial Stability Reports).32 We use
-80
-60
-40
-20
0
20
40
60
80
2002 2003 2004 2005 2006 2007 2008
G ro
w th
o f
pe rc
en t o
f G
D P
Emerging Asia Latin America Middle East Africa Emerging Europe
Note: Financial assets are defined as bonds, shares and bank assets.
Figure 2. Annual growth of financial assets in emerging markets, as percent of GDP.
31 See Morris Goldstein and Daniel Xie, ‘The Impact of the Financial Crisis on Emerging Asia’,
presented at the conference on Asia and the Global Financial Crisis, Santa Barbara, 18–20
October 2009, http://www.frbsf.org/economics/conferences/aepc/2009/09_Goldstein.pdf (vis-
ited 5 August 2010). 32 The database also includes current account data for the corresponding countries (from the
IMF WEO database).
948 Journal of International Economic Law (JIEL) 13(3)
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
panel regression with so-called ‘fixed effects’ to control both omitted
variables and country-specific characteristics.33
Column (1) of Table 2 shows the panel regression results where the cur-
rent account balance is the dependent variable, and the corresponding DFM
growth rate and its lagged value are independent variables. These empirical
results reveal a robust negative correlation between past DFM growth and
current account positions. Faster DFM growth in the lagged year is strongly
correlated with a larger current account deficit (or a smaller current account
surplus) in the present year.
Exchange rate movements may explain a good part of changes in current
account balances. To control for this effect, we single out the Euro area
countries and rerun the regressions. Since the Euro area countries have a
single currency, this experiment implicitly removes exchange rate changes.
-50
-40
-30
-20
-10
0
10
20
30
40
50
2002 2003 2004 2005 2006 2007 2008
G ro
w th
a s
pe rc
en t o
f G
D P US
Japan
Germany
Greece
Ireland
Spain
U.K.
Note: Financial assets are defined as bonds, shares and bank assets.
Figure 3. Annual growth of financial assets in developed countries, as percent of GDP.
Table 2. Current Account Balance and DFM Growth Panel Regression 2000–08
Independent variables (1) (2)
18 OECD countries 12 Euro countries
Current year DFM growth rate 0.005 (�0.009) 0.008 (�0.012)
Lagged year DFM growth rate �0.024** (�0.01) �0.029** (�0.013)
Number of observations 108 72
Number of countries 18 12
Note: Standard errors in parentheses; **P< 0.05; P : Probability.
33 For an explanation of the ‘fixed effects’ regression model, see Jeffrey M. Wooldridge,
Econometric Analysis of Cross Section and Panel Data (Cambridge: MIT Press, 2001).
Financial Stability and Monetary Policy 949
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
Column (2) of Table 2 shows the new results: the negative effect of
prior-year DFM growth persists.
How can these results be explained? Domestic monetary expansion pushes
up asset prices and attracts more capital from overseas. A current account
deficit then needs to emerge to offset the capital inflow, as Harry Johnson
explained long ago.34 An obvious channel for a rising current account deficit
is appreciation of the exchange rate—for those countries on flexible exchange
rate regimes. However, even for countries with fixed rate regimes—those in
the Euro area, for example—the negative relationship between DFM growth
and the current account balance persists. This implies that a second channel
is general economic expansion fueled by capital inflows, which in turn
enlarges the current account deficit.35
An interesting piece of evidence to support the negative relation between
the current account position and lagged DFM growth is that China may
witness a smaller current account deficit in 2010. One explanation for this
surprising phenomenon could be the recent strength of housing prices due to
super-loose monetary policy in 2009.
VII. ALTERNATIVE MONETARY TARGETS
A few scholars and policymakers have begun to reconsider the received
tenets of monetary policy, in light of the most serious economic crisis
since the Great Depression of the 1930s. Blanchard dell’Ariccia, and
Mauro, for example, suggest the possibility of increasing the desired inflation
targets, aiming to make more room for monetary easing36 in containing the
next economic crisis.37
Other researchers have proposed revisions to the way in which inflation is
measured in the present IT framework. For example, they would include
selected asset prices in the price index to warn of the next bubble.
However, unlike consumption goods, the role of asset expansion cannot be
entirely captured by a price index. For example, new financial products may
be created that have little influence on the average level of asset prices, but
still greatly expand the gross value of tradable and liquid assets. Moreover,
financial innovation keeps creating new products that might be overlooked
by financial supervisors in the early stages and not integrated into the price
34 Harry G. Johnson, ‘The Monetary Approach to Balance of Payments Theory’, 7 Journal of
Financial and Quantitative Analysis 1555 (1972). 35 A caveat is that our panel data set only covers developed countries, which means that major
developing countries like China, India, and Brazil are not included due to the difficulties of
data collection. The experience of these developing countries may differ from the results
reported here. 36 To stimulate economic activity, central bank might reduce interest rates and boost money
supply. 37 See Blanchard, dell’Ariccia, and Mauro, above n 14, at 5.
950 Journal of International Economic Law (JIEL) 13(3)
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
index until a crisis finally happens. Collateralized debt obligations (CDOs)
would be an example.
To achieve the policy objective of Financial Stability, in addition to IT,
additional policy targets and instruments are needed, not just an adjustment
of existing targets and instruments. As for targets, we have sketched DFM
and the way that the concept is summarized in equation (1) above is cer-
tainly measured better than we have done in this article.
Turning to instruments, various micro-economic controls are available,
in addition to the standard central bank suite of short-term policy interest
rates and M1 levels. For example, more stringent liquidity and capital
requirements for banks, beyond Basel II, are widely discussed, and these
can be varied in response to changes in financial stability conditions. An
obvious instrument for the housing sector is the mortgage down-payment
requirement. Margin requirements can be adjusted to moderate stock and
bond market booms. For derivatives, the authorities can insist on central
clearing houses, which in turn can enforce initial margin and variation
margin requirements. Another instrument is to regulate the extent of
credit card debt by imposing maximum interest rates, which in turn will
induce banks to raise their standards for issuing credit cards.38 All these
tools are clearly within the legal powers of central banks and other financial
supervisors.
VIII. INTERNATIONAL SURVEILLANCE
If the new monetary policy framework we suggest is so good, and if instru-
ments for DFM targeting are readily available, why don’t countries by their
own initiative seek to moderate financial fluctuations by better control of
national DFM levels? What is the rationale for international surveillance?
Our short answer is that the combination of rising asset values and low
inflation is a central banker’s dream. Under the legendary Alan
Greenspan, years of loose monetary policy created a vast real-estate boom
that ended in tears. During the good times, Greenspan was crowned ‘the
greatest central banker in history’ by no less a figure than Alan Blinder.39
The inevitable tears were confined neither to Wall Street nor the boundaries
of the US economy. Owing to global financial integration, damage was
38 Quite often, the credit card default rate is around 10 percent annually, which implies that
credit card interest rates must be at least 15 percent for the issuing firms to make money on
outstanding balances. On 30 April 2010, Senator Sheldon Whitehouse Democrat-Rhode
Island (D-RI) offered an amendment to the financial regulatory bill to force national card
issuers to comply with anti-usury laws in each state where their customers reside. If enacted,
this measure would dramatically curtail the issue of high-risk credit cards. In preference to
this approach, we favor granting the Federal Reserve the power to cap credit card interest
rates. 39 Alan S. Blinder, The Quiet Revolution: Central Banking Goes Modern (New Haven, CT:
Yale University Press, 2004).
Financial Stability and Monetary Policy 951
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
spread far and wide. In earlier but smaller episodes—the Latin American
debt crisis of the 1980s, and the Asian and Russian crises of the 1990s—
damage was also global.
National central banks are neither omniscient nor immune to the pleasures
of rising asset values. Moreover, national monetary authorities are unlikely to
give sufficient weight to the global consequences of ‘feel-good’ policies at
home. Even if central bankers are assigned explicit objectives—a combin-
ation of low inflation, full employment, and financial stability—they might
put insufficient weight on financial stability. From time to time, central
bankers need forceful warnings from their foreign peers.
For purposes of global financial stability, country size matters. A few large
countries bear nearly all the responsibility for stabilizing global asset values.
Any system of international financial surveillance needs to worry about the
USA, the EU, Japan, China, India, Brazil, and a very few others. This group
is the core of the G-20, the latest global steering group which came together
in the midst of the Great Crisis.40
The empirical results derived in the last section suggest that international
trade and monetary surveillance are complementary actions. Obstfeld and
Rogoff, for example, have proposed that current account balances should be
a consideration in the formation of monetary policy.41 According to our
empirical analysis, if excess monetary expansion is avoided, by adopting a
DFM target reinforced by international monetary surveillance, large and
persistent current account deficits could be mitigated.
At the G-20 Summit, held in London in April 2009, the Financial
Stability Forum of the BIS was enlarged to include all G-20 countries,
and renamed the Financial Stability Board (FSB), with the implication of
stronger powers and broader representation. However, so far the FSB has
only played a consultative role, with a focus on micro-prudential issues.
Despite its grand name, the FSB is not the right forum for international
surveillance of Financial Stability, in the sense that we conceive this
objective.
Instead the International Monetary Fund (IMF) is the prime candidate for
the task of macro-financial surveillance. The IMF already has extensive
experience in designing ‘early warning’ indicators. So far, however, the
IMF has been singularly ineffective in warding off looming financial crises.
Rather, the IMF’s role has historically been to discipline smaller countries
after a crisis erupts. We are suggesting that the IMF should widen its
40 The G-20 held its inaugural meeting in Washington in November 2008, with two meetings
since then – London in April 2009 and Pittsburgh in September 2009. The G-20 has essen-
tially eclipsed the G8 as the steering forum for the world economy. The next meetings are
scheduled for Toronto in July 2010 and Seoul in November 2010. 41 See Obstfeld and Rogoff (2005), above n 29, at 10.
952 Journal of International Economic Law (JIEL) 13(3)
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
ambitions to play a ‘crisis avoidance’ role for ‘systemically important coun-
tries’, namely the big G-20 nations.
The DFM framework that we have proposed, with suitable refinements,
should be used by the IMF to monitor anomalous financial imbalances in
important countries and then blow the whistle. As a first step this would
require enhancement of the two IMF data reporting systems: the SDDS
(Special Data Dissemination Standard) and the GDDS (General Data
Dissemination System). Wealth accounts and flow of funds accounts would
be needed for the important IMF members. A benchmark threshold can
then be set between the DFM growth rate and a combination of GDP
and inflation growth rates, drawing on empirical evidence from previous
boom and bust episodes. When the DFM growth rate for an important
country exceeds the benchmark threshold, the IMF should sound the alarm.
IX. CONCLUSIONS
The Great Crisis has motivated both scholars and officials to think about
tools for containing the next global financial crisis. Loose monetary policy
was among the causes of the Great Crisis. At present, the mainstream mon-
etary policy regime, which focuses on price stability, is not well equipped to
detect the beginning of an asset bubble, which in turn sets the stage for a
financial crisis. The huge social cost of a financial crisis should, however,
compel policymakers to include Financial Stability as a major policy object-
ive. We propose the DFM framework, which quantifies a broad set of liquid
assets, as the basis for measuring Financial Stability. Empirical evidence
shows that DFM can anticipate conditions that presage a financial crisis.
Cross-country comparisons illustrate that future fragility is correlated to con-
temporary rapid expansion of DFM. Empirical evidence thus lends strong
support to including DFM as an additional target for central banks to moni-
tor, and thereby promote financial stability. To implement this new monetary
framework, central banks need to invoke more instruments than traditional
short-term policy interest rates and M1 controls. The new instruments must
include control over bank capital and liquidity requirements, mortgage down
payments, and maximum credit card interest rates.
Spillover of financial crisis is an inevitable side effect of globalization.
International surveillance is needed because national central banks typically
enjoy the upside of an asset cycle, and typically discount the negative spill-
over effects from a domestic crisis. The IMF is the ideal candidate to carry
out the surveillance task.
Financial Stability and Monetary Policy 953
at B angor U
niversity on February 22, 2015 http://jiel.oxfordjournals.org/
D ow
nloaded from
Financial Stability in Asian Economies.pdf
Financial Stability in Asian Economies Author(s): Subbaiah Singala and Mukul G. Asher Source: Economic and Political Weekly, Vol. 43, No. 10 (Mar. 8 - 14, 2008), pp. 65-71 Published by: Economic and Political Weekly Stable URL: http://www.jstor.org/stable/40277232 .
Accessed: 06/02/2015 10:49
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to Economic and Political Weekly.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
Financial Stability in Asian Economies
SUBBAIAH SINGALA, MUKUL G ASHER
Financial innovations, exemplified by the increasing complexity and variety of financial instruments, and the
emergence of new financial players such as sovereign wealth funds have led to an unprecedented increase in
global financial assets and flows. While these changes have brought about improved liquidity, reduced transaction costs and more risk management options,
they have created major challenges for macroeconomic
policymakers. As incidents of financial disruption and
volatility increase and as their economic costs become
significant, ensuring financial stability has become a
major preoccupation of central banks.
The authors would like to thank Charles Adams for useful comments. The usual caveat applies. Views expressed in the article are personal.
Subbaiah Singala {[email protected]) is with the Reserve Bank of ìndia and Mukul G Asher ([email protected]) is with the National University of Singapore.
globalisation process, particularly since the 1980s,, has resulted in a radical transformation and expansion of
global financial markets. The world's financial stock
(government and corporate bonds, equity and bank deposits) to GDP has also increased from around 100 per cent in 1980 to over 325 per cent in 2004 [Mckinsey 2005].
The forces of globalisation, viz, deregulation, liberalisation, advances in information and computer technologies, etc, have eased the speed and lowered the cost with which capital can be moved
globally and financial transactions can be carried out. This has been
accompanied by the development of sophisticated new financial instruments for risk management;1 and by the emergence of new
players such as sovereign wealth funds (swfs).2 These changes are beneficiai overall but they have also been
accompanied by frequent financial disruptions (for example, bond market turbulence in the United States in 1996, Asian currency crisis in 1997, Argentine crises in 2001-02 and sub-prime mortgage crisis in 2007). There have been more than 10 incidents of market turbulence and financial crisis since 1992 [Schinasi 2005a].
These financial crises had significant macroeconomic costs. Goldstein et al (2000) found that there have been more than 65 developing country episodes where the banking system's capital was completely or nearly exhausted during 1980-95. In more than a dozen of these banking crises, the public-sector resolution costs amounted to 10 per cent or more of the country's gdp. The cost of bank recapitalisation in the countries affected by the Asian financial crisis is estimated to be around 58 per cent of gdp for
Indonesia, 30 per cent for Thailand, 16 per cent for south Korea and 10 per cent for Malaysia.
Crockett (1997), therefore, has appropriately argued that
securing stability should be the concern of public authorities as
instability has been associated with lower levels of savings and
investments, fiscal costs, and adverse impact on growth. Moreover, as Martin Wolf (2008) has argued "a financial sector that generates vast rewards for insiders and repeated crises for hundreds of millions of innocent bystanders is... politically unacceptable in the long run".
The recent sub-prime mortgage crisis and its impact on global financial institutions and markets have induced re-pricing of risk
by market participants and change in the monetary stance of the
major central banks. The Asian economies depend on exports to the developed world and the current signs of possible recession in the us and uk economies might adversely affect their growth rates. These countries are already facing the problem of manag- ing relatively large capital inflows, and the recent rate cuts by the Federal Reserve and the Bank of England could only exacerbate the problem. Recently the International Monetary Fund (imf) has
Economic &Political weekly ÛE253 march 8. 2008
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
SPECIAL ARTICLE = - - =
cautioned that unstable capital flows could pose a serious challenge to financial stability in Asia in future [imf 2007a]. The Global Financial Stability Forum has also set up a working group to explore the consequences of capital flows for emerging market economies, both the macroeconomic effects and the impact on the domestic financial systems [rbi 2007].
Safeguarding financial stability, therefore, has become an increasingly dominant objective in economic policymaking as it has all the essential characteristics of a public good3 [Schinasi 2005b]. Nearly 50 central banks and other international agencies such as the imf now publish regular reports on financial stability. Currently, asset bubbles and excessive leverage are engaging policy attention as the possible destabilising forces underlying the ongoing correction in global markets.
It is in the above context that this paper examines financial stability issues confronting Asian economies. It also suggests pubic policy measures, which could help enhance financial stability. The rest of the paper is structured as follows. Section 1 provides a brief overview of the approaches to safeguard finan- cial stability. Section 2 reviews the various macroeconomic and analytical models to safeguard financial stability. Section 3 presents a policy cycle concept for addressing financial stability challenges. Section 4 briefly discusses financial stability challenges for India. The concluding section suggests public policy measures to enhance financial stability.
1 Definition and Approaches There is no consensus on what constitutes financial stability. In general, financial stability refers to the smooth functioning of financial markets and institutions without serious disruptions [rbi 2006b]. More formally, "financial stability is a condition in which an economy's mechanisms for pricing, allocating, and managing financial risks (credit, liquidity, counterparty, market and so forth) are functioning well enough to contribute to the performance of the economy" [Schinasi 2005a].
There are three important aspects involved in comprehensive assessment of financial stability.4 These are: (i) identifying the plausible and systemically important sources of risks and vulner- abilities that could pose challenges to financial stability in future; (ii) an appraisal of the potential costs, that is, the ability of the financial system to cope, should some combination of these identified risks and vulnerabilities materialise; and (iii) forming a judgment about the individual and collective strengths and robustness of the constituent parts of the financial system - insti- tutions, markets and infrastructures.
In addition, the following aspects are also relevant: (i) taking steps to counteract risks and vulnerabilities and/or strengthening the financial system and/or its constituent parts if there is a chance that certain risks and vulnerabilities could overwhelm the financial system and the resulting costs are considered to be too great;5 and (ii) crisis management and resolution arrangements.
The existing approaches to safeguard financial stability have limitations. For example, Goodhart (2005) provides insights into the limitations of existing measures of supervisory capital requirements in the globalised world. He argues that Basel 11 regulations and the new accounting standards for financial 66
instruments (ias 39) are likely to be of limited help in fostering financial stability. Stliglitz (2006) argues that the current global (dollar) reserve system has enormous costs for developing countries, exerts a deflationary pressure on the global economy and contributes to global instability. Financial Times (2006) highlighted some of the limitations of using value-at-risk (var) models and emphasised that market liquidity risk cannot be measured by var during a market panic, as large-scale selling in the fear of further price decline is common during a crisis, while some hedging programmes require sales in the falling markets.
Several techniques and approaches have been suggested to cope with financial stability issues. The use of early warning signals to safeguard financial stability in emerging economies was studied by Goldstein et al (2000). The use of key standards and codes as a useful way of benchmarking progress in the reform of a financial system was advocated by Schneider (2003) based on the evidence that the adoption of standards and codes does lead to improved country credit rating, especially when standard assessments are allowed to be published. Both these approaches are quite popular and useful in understanding the possible sources of threats to financial stability.
There is, however, a constant need to refine these approaches as the sources of threats are constantly changing thanks to innova- tions in financial instruments and risk transfer mechanisms, the derivative structures (collateralised debt obligations, popularly cdos) involved in the sub-prime crisis being one such recent example. A similar argument was made by Brouwer (2003), who argued that the risks associated with hedge funds need to be regulated carefully and that constraining an excessive leverage of hedge funds effectively is particularly important from a public policy perspective. Similar sentiments were expressed by Asher (2008) for sovereign wealth funds, imf (2007b) cautions that Asia remains too dependent on exports as an engine of growth and the financial sector needs to be developed further to cope with capital flows.
2 Analytical Models The challenges to safeguarding financial stability are becoming increasingly complex and hence, need to be constantly monitored, measured and managed. However, issues such as increasing global imbalances, asset bubbles, complex risk-transfer instru- ments, increasing size of financial conglomerates, etc, could pose serious challenges to domestic financial stability. The use of statistical models in this process cannot be relied upon exclu- sively, given the behaviour of market participants. To counter these challenges, each country should put in place a rigorous analytical framework for safeguarding financial stability. However, the challenges in building an analytical framework are more complex than compared to developing a monetary policy framework. While the mandate of monetary policy is clear in terms of inflation levels, the financial stability challenge is a lot more complex, mainly because of the focus of financial stability policies on extreme (tail) events. A few analytical frameworks have been advanced recently, which aim to identify vulnerabili- ties to financial stability and means of preventing or controlling resulting risks. The Financial Sector Assessment Programme (fsap) of the imf and World Bank, the Bank of England framework and
march 8, 2008 EEE9 Economica Political weekly
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
the analytical framework developed by Garry J Schinasi are the most comprehensive of them.
2.1 IMF-World Bank Framework In practice, analytical frameworks to monitor financial stability are centred on the surveillance of macroeconomic conditions and financial market developments, macro-prudential surveillance and analysis of macro financial linkages which are mutually supportive and reinforcing [World Bank and imf 2005]. At the international level, the imf and World Bank have launched the fsap, which examines selected countries' financial soundness and assesses their compliance with financial system standards and codes.
2.2 Bank of England Framework The 2005 analytical framework of the Bank of England (boe) for financial stability is a suit of models for better gauging the fragilities and frictions in the financial system. The boe framework recognises that the vulnerabilities of the financial system to developments (such as credit risk transfers by new holders of credit risk like
hedge funds, growth of derivative instruments and advent of a
range of new asset classes) have added to the risks of financial instability arising through leverage, volatility, and opacity. The most
important elements of the financial stability function at the boe are to assess threats to financial stability, oversight of payment systems, provision of liquidity and preparation for a financial crisis.
2.3 Garry J Schinasi's Framework A key aspect of the framework developed by Schinasi (2005b) is that it brings together macroeconomic, monetary, financial market, supervisory, and. regulatory inputs. The purpose of this framework is to (i) foster early identification of potential risks and vulnerabilities; (ii) promote preventive and timely remedial
policies to avoid financial stability; and (iii) resolve instabilities when preventive and remedial measures fail.
The framework has the objective of preventing problems from
occurring or resolving difficulties if prevention fails. Factors
affecting the financial system performance are identified under
endogenous and exogenous factors. The focus is on identifying and dealing with the build-up of vulnerabilities prior to downward corrections in the market, prior to problems within institutions or prior to failures in financial infrastructure.
2.4 Review of Models The review of models suggest a few policy considerations. These are (a) financial stability assessments carry a higher degree of
uncertainty than that ordinarily associated with forecasts based on macro-econometric models; (b) normal distribution and
log-linear relationships that models follow may not be relevant in assessment of financial sector stability as crises can have
unpredictable non-linearity; (c) analytical frameworks could be useful in constantly monitoring risks and vulnerabilities and initi-
ating suitable pro-active or remedial actions; (d) undergoing the fsap of the IMF-Word Bank at regular intervals could point out the areas requiring focus from a financial stability angle; (e) the sources of threats to financial stability cannot be captured by any single model or framework and require constant vigil and
ongoing surveillance for early identification and suitable preven- tive or remedial action; (f) policy communication is a power tool to manage market perceptions of the macroeconomic, market, institutional and infrastructure factors affecting finan- cial stability; and (g) an eclectic framework, appropriate in the context of a specific country is needed requiring substantial capacity and expertise.
3 Financial Stability Cycle Framework Given the nature of financial stability issues, limitations of mathematical models and the emerging role of Asia in the global order, financial stability can be safeguarded by putting in place a "financial stability cycle" framework. The cycle gives a "helicopter" approach to financial stability and could foster inter- departmental coordination within the central banks and inter-
agency coordination in a country. The components of the cycle are already in place in most of the countries and having them organised in a cycle form and reviewing it, say every six months, could bring about an organised and systematic way of safeguard- ing financial stability. Setting up a dedicated group with the
support of top management to work on financial stability cycle issues would serve the function well. Though this could put strain on scarce human resources, the trade-offs are worth the invest- ment. The broad components of this framework are listed below:
3.1 Policy Actors Economic policy falls under "incremental approach" with the
policy framework evolving over a period [Dye 2005]. Financial
stability can also be treated as an incrementally developed approach. The policy actors in preserving financial stability can be grouped under the following categories:
International System: Financial stability is fostered through international cooperation and is contingent upon global order.
Any imbalances in the global economic order might pose a substantial risk to financial stability. The actors of the inter- national system that have a bearing on the financial system are the major economies and their central banks, financial market
players like foreign institutional investor, hedge funds, sovereign wealth funds, and financial institutions and international organi- sations such as IMF/World Bank, etc. International think tanks and press are becoming increasingly important.
Domestic State System: The capacity and autonomy of individual countries and their central banks to pursue sound macro- economic, prudent supervisory and pragmatic exchange rate
policies influence the financial stability of individual countries. The most important actor here is the central bank of the country. Other players include the government, particularly the finance
ministry, regulators of the financial system and market players. Business media plays an important role in this process by highlighting the instances of "wrong signals" to financial stabil-
ity and keeps the regulators on guard. Think tanks and research
organisations also play a role in finding enduring solutions to the
problems to financial stability. A robust database on financial market players and financial instruments is essential. As such a
Economic apolitical weekly GEE3 march 8, 2008 Ö7
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
SPECIAL ARTICLE ~ ' ^^^
database is a public good, it should be the responsibility of the central bank of the country.
3.2 Policy Instruments
Normally several instruments are used to preserve financial stability. Organisation based tools of the central bank such as prudential supervision, maintaining macroeconomic and price stability, market surveillance and monitoring are commonly used. Self-regulation by market players and advisory groups is also a strong policy instrument to promote financial stability. One of the voluntary tools, namely, market creation is also a powerful tool for maintaining financial stability, as it facilitates competitive pressure and orderly functioning of markets. Infor- mation is also used as a powerful tool to promote financial stabil- ity as it communicates micro and macro level issues to the public at large. Market based tools such as volatility in share prices, changes in bond spreads, implied volatility based on option prices, ratings are now commonly used [Martin 2006].
3.3 Policy Cycle The policy cycle (see the chart) may be characterised as having the following stages:
Policy Formulation: Policy formulation for financial stability follows a "policy renewal" style as there is a constant emergence of new problems and new players in the financial markets. The first two stages of the cycle focus on monitoring and analysing the factors that affect financial stability and a constant assess- ment of their possible impact. These could be endogenous factors such as institutional factors (financial, legal, reputation risks), market-based factors (run on markets, contagion, counter party risk) and infrastructure factors (regulatory, supervisory, domino effects) and exogenous factors such as macroeconomic distur- bances (Asian crisis) or events (outbreak of bird flu, saars) [Schinasi 2005a]. A robust macroeconomic surveillance function focusing on tail events (high risk-low probability), focus on
potential crisis. These stages may run parallel due to multiple problems or a single problem passing through these stages if not managed properly in the initial stages. Prevention of problems and identifying potential problems such as identifying irregular practices in a particular market, lack of internal controls and systems in financial institutions and weaknesses in the infra- structure such as payment systems, etc, is a crucial proactive way of addressing threats to financial stability while they are building up. Taking remedial action for potential high value problems such as rapid credit booms, inflation, asset price bubbles and capital flows is the most critical stage as it poses challenges for the identification of the right mix of solutions in determining the appropriate action. An overdose of policy response could stifle growth and the use of inappropriate instruments could bring about unintended consequences. The resolution of potential crises such as payments not settled in time, market fears follow- ing a major negative such as war and operational breakdowns is tricky as it has moral hazard issues. If the resolution is in the interest of systemic risk, then central banks need to provide liquidity and other support to resolve the crises.
Policy Evaluation and Communication: The policy evaluation process of financial stability is normally undertaken by the state actors. However, with the increasing realisation of the value of involving market players and social actors and to manage thoir perceptions, policy communication effort has received increasing attention. This presents an opportune time for policy actors to make an assessment of the period under review, the challenges, if any, posed to safeguarding financial stability and the efficacy of policy actions. This also gives an opportunity to present to the market players the likely situation of financial stability and shape market perceptions through financial stability reports, speeches and briefs.
Managing market perceptions is crucial for creating a stabilis- ing convention in the markets. Internationally, the trend is to publish a separate communication on financial stability. The
systemically important players through risk based supervision, use of tools like stress tests and liquidity tests are important areas to focus on at the policy formulation stage. In countries with or planning to set up financial centres, strengthen- ing the surveillance function to monitor exogenous shocks could be useful.
Policy Implementation: The policy implementation cycle of financial stability involves stages of prevention of problems and identification of potential problems, remedial action for potential systemic problems and resolution of
publication of such a report is less common in Asia, in spite of its potential contribution to credibility, education and anti- cyclicality.
The policy cycle framework could help enhance policy coherence among the various actors such as the central bank, ministry of finance, other regulators, etc, and would lend credibility to the actions and communication of state actors. This helps in managing market perceptions well. This could be an area for improvement in most Asian countries, particularly those with democratic govern- ments, and robust non-govern- ment controlled media.
Chart: Policy Cycle for Financial Stability
Source: Adopted from Schinasi (2005b).
68 march 8, 2008 B3E2 Economic &Political weekly
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
- SPECIAL ARTICLE
The cycle can have a renewal period of six months for under- taking a system health check for stability and communicate the status to markets.
4 Financial Stability Challenges for India Since adopting the open economy, open society paradigm in 1991, India is increasingly becoming integrated with the rest of the world in terms of international trade, investment and financial markets. However, the risks associated with globalisation have also sensitised domestic policymakers to the need for stability of the financial system. Financial stability has, therefore, become a dominant objective of economic policy in India [rbi 2006a]. Perseverance of domestic financial stability in the wake of several global financial crises in the 1990s and the early part of this decade bears testimony to the institutional strength and capacity of the government and rbi. However, the stock market volatility in May 2006 and the continued risks of financial markets turbu- lences are a constant reminder of the challenges that lie ahead in safeguarding financial stability in India.
India, like most Asian economies, is traditionally a bank dominated financial system, and hence regulation and super- vision of banks is at the core of financial stability policies. Pruden- tial regulation using risk management principles, development of sound payment and settlement systems, sound governance and disclosures, monitoring the macro prudential indicators and self-
regulation have been the key instruments of financial stability policies in India. The regulatory structure is that of multiple regulators and necessary institutional arrangements have been
.put in place to ensure that risks, particularly those emanating from financial conglomerates, are better managed through inter-
agency regulatory coordination. To safeguard financial stability, the rbi adopted a multi-pronged strategy based on international best practices with suitable adaptations to promote stability of institutions, markets and the financial infrastructure. Monetary policy also fostered a financial stability function by lowering inflation and stabilising inflation expectations [rbi 2005]. Thus, the rbi has pursued measures aimed at both micro and macro
prudential controls. The key to financial stability policies has been strengthening
financial institutions like commercial and cooperative banks, financial and non-financial companies, mutual funds, insurance
companies, etc. Of these, the banking sector, being the most dominant class of players in the financial system, has become
strong, healthy and resilient. With an asset quality, profitability and capital base that are comparable to international standards, the banking sector has become a source of stability. With the
present policies of consolidation, cost reduction, improved govern- ance and technological innovations, they are even aspiring to
compete at the international level. Strong prudential regulation, development of financial markets, deregulation of banking indus-
try and rationalisation of statutory requirements have been the most important factors in strengthening the banking sector.
The rbi places an equal emphasis on measurement, manage- ment and containment of risks of financial instruments, players and markets. Regulatory policies have focused on strengthening risk management systems, practices and management of market
players and the appropriate regulatory monitoring mechanism of the system's aggregate risk. Broader issues like anti-money laundering and combating the financing of terrorism have also been addressed through proactive regulation. Risk based super- vision, off-site monitoring and surveillance, encouraging banks to strengthen their internal controls, mandating consolidated accounts of banks and institutionalising a scheme of prompt corrective action as a structured early intervention system and involving external auditors have been the supervisory measures for fostering a stable banking system. Other segments of the financial system have also been strengthened in a calibrated manner, even though the outcomes, especially in case of urban cooperative banks are not as encouraging.
Financial markets development has been at the core of finan- cial stability policies in India, since they have a direct impact on the health of financial institutions through portfolio holdings and risk management. The money, forex and debt markets have
developed with a vigour and this has enabled better management of liquidity, transmission of monetary policy objectives, pricing of assets and risk management. At the same time, the Securities
Exchange Board of India has made efforts to develop the equity market. A closely related aspect is the thrust given to transpar- ency, data dissemination and disclosures - both by market players and the central bank. The involvement of market participants through committees and groups in advising the rbi has enabled market development and consolidation. The growth of financial markets is partly reflected in the phenomenal growth of deriva- tives markets - equity, forex, interest rate and commodity, and the availability of reasonable market liquidity, even though at times it could be a one-way market reflecting the shallowness of these markets.
A mention of the strength of the payment and settlement
systems and its role in preserving financial stability should be made. Several measures such as the delivery versus payment system and guaranteed settlement have already been put in place and recently a board for supervision and regulation of payment and settlement systems has been constituted and real time gross settlement in large-value funds transfer systems has been opera- tionalised. Implementation of international standards and codes is another positive aspect.
Thus, the financial stability function has been rather robust in India. More recently, the government of India in consultation with the rbi constituted a Committee on Financial Sector Assessment under the chairmanship of Rakesh Mohan to undertake a self- assessment of financial sector stability and development, using the IMF-World Bank fsap framework. While this will definitely contribute to the strengthening of domestic financial stability, the financial stability policy cycle framework suggested in the paper could play a complementary role. Also, in the light of the risks
posed by the international movements of capital, risk transfer, proxy pricing and other factors, the following suggestions may complement the ongoing efforts of the policymakers to strengthen financial stability function in India in future, (i) Organisational set up: Dedicated organisational division to monitor and coordinate the financial stability function may be set up.
Economic &Political weekly H23 march 8, 2008 Ö9
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
SPECIAL ARTICLE EEE _ =
(ii) Financial stability reports: Financial stability reports could be published at regular intervals. (iii) Policy cycle: A financial stability cycle framework suggested in the paper could be put in place, which can be constantly renewed to ensure an organised and systemic focus on preserv- ing financial stability function. (iv) Tools: To measure the impact of shocks to the system and the ability to withstand the same, tools such as stress tests, liquidity tests and contagion tests can be further developed, (v) Surveillance: There is a need to enhance the scope of both the macroeconomic surveillance and the micro/prudential surveil- lance, to focus on India's key economic partners, (vi) Prudential regulation: Prudential regulation involving both micro and macro controls could be strengthened to ensure that the risks are properly measured and reported by banks and financial institutions, more particularly by the financial conglomerates. Executive compensation in the financial sector will need to be monitored, without affecting the flexibility and competitiveness of the sector. (vii) The tendency of the swfs, particularly from Singapore to target India's financial sector, also require continuous monitoring and heightened scrutiny. (viii) Encouraging domestic investors: Prudential encouragement to domestic provident and pension fund organisations to be more active in financial and capital markets, and the development of more robust annuities market may assist in promoting financial stability, (iv) Education: A strategy may be put in place to educate the public at large as well as the political class and media on financial stability. The central bank may impress research and academic institutions and think tanks to undertake rigorous research on financial stability issues affecting India. Public perception of risks
and rewards in the financial sector will also need to be managed. These educational initiatives would supplement the efforts of the central bank in bolstering domestic financial stability.
5 Concluding Remarks
Strengthening the banking sector, through prudential regulation and risk based supervision should be at the core of financial stability policies. In the Asian context, this is all the more the case due to the dominance of banks in the financial sector and the underdevelopment of the corporate debt market. Effective supervision of the banking and financial system, therefore, is the most important aspect of preserving financial stability. Increas- ingly and in the context of financial globalisation, behaviour of markets that could lead to a one-way movement of asset prices, is engaging financial stability policies. Maintaining market confi- dence in the financial system, developing resilience of payment and settlement systems and providing liquidity to market players with genuine difficulty during crises have been the three aspects of financial stability oversight the world over during speculative attacks or threats to financial stability.
The correlation among international markets seem to be on a rise as does the homogeneity of views and risk management practices. The risk appetite of market players has also been rising, though the sub-prime crisis appears to have altered the pricing of risk. These have increased systemic risks and threats to financial stability besides contributing to under pricing of risks. From a financial stability perspective, therefore, it is important to encourage market players with different invest- ment objectives, balance sheet composition and capital struc- ture. This will reduce the homogeneity of views and massive uni-directional actions in domestic financial markets. A related
AFRICAN STUDIES ASSOCIATION OF INDIA (ASA) FELLOWSHIP ANNOUNCEMENT
ASA, India, invites applications for Ford Foundation Doctoral Fellowship from Indian Scholars registered for Ph.D in Indian universities working on South African issues.
Applications can be submitted along with research proposal, registration certificate of the university, detailed C.V and attested copies of previous degrees. Last date for submission of complete application form is 31 March 2008.
Please send your application to:
LProf. Ajay Dubey, 2. Prof. Aparajita Biswas, African Studies Association of India (ASA) Centre for African Studies, 351 - School of International Studies, Kalina Campus, Jawaharlal Nehru University, Vidya Nagari, University of Mumbai, New Delhi - 110067 Mumbai- 400 098 Email: [email protected] Email: [email protected]
march 8, 2008 BEE3 Economics Political weekly
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
- - - SPECIAL ARTICLE
approach could be to develop the domestic institutions by Asian economies and ensure that they provide stability during episodes of instability or crises, which are characterised by capital outflows and asset price depressions.
Many Asian countries are aspiring to set up international financial centres (ifcs). Financial surveillance in such a scenario analysis should cover not only the complex structure of the key financial institutions operating in an ifc but also the operations in which those institutions are engaged so as to understand the sources of risk (which are often outside the host jurisdiction) and the transfer of risks within and from the ifc. Cross-border movements of capital in ifcs might put a strain on payment and settlement systems, encourage risk taking behaviour in domestic markets by foreigners or by domestic entities in foreign countries and may contribute to higher volatility of growth. These risks could be managed though developing strong institutional capacity to supervise ifcs through an institutional arrangement of informa- tion sharing and policy coordination among national supervisors.
Globalisation necessitates global rules, institutions and
regulations and hence, a formal and legally binding framework for international finance needs to be developed. A financial
stability forum is seen as a step in this direction but it is infor- mal and does not have the mandate or power to adopt and
implement a regulatory policy of its own. A formal institution is therefore, an urgent need. A regional financial stability forum could be set up to coordinate the financial stability policies among Asian countries.
However, there are two broad issues, which may challenge the efforts to safeguard financial stability. First, the issue of
communicating financial stability issues could be a double edged sword. Financial stability communication has, so far taken place under benign conditions and not been tested for its efficacy during a crisis or adverse market conditions. Thus, communicat- ing financial stability issues without precipitating a crisis could be a challenge going forward. Second, the quality of indicators of financial stability in Asian markets needs to be improved. This calls for rigorous research on financial stability issues in these countries so as to develop indicators and measurement tools for financial stability and techniques to manage market perceptions and behavioural finance aspects.
As Asia emerges as the future engine of global economic growth, led by China and India, the opportunities for Asian countries to grow and ameliorate the living standards of their
people are immense. However, the forces of globalisation may also pose constant threats to domestic financial stability with the potential to derail the growth process. They missed the
opportunity to capitalise on the growth momentum achieved in the 1990s and can ill afford another crisis. Therefore, safeguard- ing financial stability should be pursued with right earnest.
The financial stability function is rather robust in India and there are ongoing efforts by the government of India and the rbi to further strengthen domestic financial stability. Putting in place the financial stability cycle framework discussed in this paper and other measures such as strengthening prudential regulation of financial conglomerates, encouraging a diverse domestic inves- tor base, developing financial stability tools and education of the
public at large on the need to preserve financial stability would
complement the efforts of policymakers and regulators.
NOTES
1 For example, the over the counter derivatives market has grown five times from $ 2.5 trillion to $ 10 trillion between 1998 and 2006 [BIS 2007].
2 The sovereign wealth funds (SWFs) are one of the pools of assets (primarily but not exclusively international) for achieving the government's economic, financial, and strategic objectives. In June 2007, 16 of the largest SWFs had assets of $ 2.1 trillion, and the corresponding countries had foreign exchange reserves of $ 4.1 trillion [Asher 2008]. Morgan Stanley estimates that by 2015, SWF's assets alone will be $ 12 trillion, equivalent to the GDP of US (ibid).
3 These are: (1) non-excludability of supply, i e, the producer of the good is unable .to control who benefits from consuming the good; and (2) non-rivalry in consumption, i e, consumption of the good by one person does not affect the benefits received in consuming the good by others. Pure public goods convey benefits that are both non-excludable in supply and non-rival in consumption.
4 European Central Bank, http://www.ecb.int/ press/pr/date/2oo5/html/pro5i2o8.en.html (last accessed on December 31, 2007).
5 Based on email correspondence with John Palmer, former deputy managing director, Monetary Authority of Singapore dated March 10, 2006.
REFERENCES Asher, Mukul G (2008): 'Sovereign Wealth Funds',
Pragati, January, pp 10-12, available online at http://pragati.nationalinterest.in
Bank of England (2005): Financial Stability Review June 2005, London.
BIS (2007): 'Semiannual OTC Derivatives Statistics at End-December 2006', available online at http:// www.bis.org/statistics/derstats.htm, accessed on June 12, 2007.
Brouwer, Gordon De (2003): Hedge Funds in Emerging Markets, Cambridge University Press, UK.
Crockett, Andrew (1997): 'Why Financial Stability a Goal of Public Policy', available online at http.V/www. kc.frb.org/publicat/sympos/1997/pdf/s97crock.pdf, accessed on June 12, 2007.
Dey, Thomas (2005): Understanding Public Policy, Pearson Prentice Hall, 11th ed, US.
Financial Times (2006): 'The unVaRnished Truth', LEX column on July 14, 2006, London, available on- line at http://search.ft.com/searchArticle7query Text=VaR++stress+test+&y=8&javascriptEnabl ed=true&id=o6o7i4ooo832&x=n, accessed on November 18, 2006.
Goldstein, Morris et al (2000): Assessing Financial Stability -An Early Warning System for Emerging Markets, Institute for International Economics, Washington DC.
Goodhart, Charles A E (2005): 'Some Reflections on Financial Stability in Globalisation of Capital Mar- kets and Monetary Policy', Jens Holscher et al (ed), Palgrave Macmillan, Houndmills, Basingstoke, Hampshire.
IMF (2007a): 'Asia Ten Years After', Finance and Devel- opment, Vol 44, No 2.
- (2007b): World Economic and binanciai burveys, Regional Economic Outlook, Asia and Pacific, October, Washington DC, available online at http://www.imf.org/external/pubs/ft/re0/2007/ APD/ENG/areoioo7.pdf
Martin, Cihak (2006): 'How Do Central Banks Write on Financial Stability', IMF Working Paper, Washington DC.
Mckinsey (2005): '$118 Trillion and Running: Taking Stock of the World's Capital Markets', available online at http^/mckinsey.com/mgi/publications/ gcm/index.asp, accessed on November 18, 2006.
Reserve Bank of India (2005): Report on Trend and Progress of Banking in India 2004-05, Mumbai.
- (2006a) : Speech delivered by the prime minister of India, Manmohan Singh, on March 18, 2006, available onlineathttp://www.rbi.org.in/scripts/BS_Speeches: View.aspx?Id=27i, accessed on November 16, 2006.
- (2006b): Speech delivered by the governor, Y V Reddy on March 10, 2006, 'Financial Sector Reform and Financial Stability', available online at http://www.rbi.org.in/scripts/BS_SpeechesView. aspx?Id=27o, accessed on November 16, 2006.
- (2007): press release, available online at http://rbi. org.in/scripts/BS_PressReleaseDisplay.aspx?prid= 16746, accessed on June 16, 2007.
Schinasi, Garry J (2005a): Safeguarding Financial Stability: Theory and Practice, International Monetary Fund, Washington DC.
- (2005b): 'Preserving Financial Stability', Eco- nomic Issues, 36, International Monetary Fund, Washington DC.
Schneider, Benu (2003): The Road to International Financial Stability - Are Key Financial Standards and Codes the Answer, Palgrave Macmillan, Houndmills, Basingstoke, Hampshire.
Stliglitz, Joseph (2006): in Inge Kaul and Pedro Conceicao (eds), The New Public Finance: Re- sponding to Global Challenges, Oxford University Press, New York.
Wolf, Martin (2008): 'Why It Is So Hard to Keep the Financial Sector Caged', Financial Times, Februarys.
The World Bank and IMF (2005): Financial Sector Assessment, Washington DC.
Economic & Political weekly 01353 march 8, 2008 71
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:49:29 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 65
- p. 66
- p. 67
- p. 68
- p. 69
- p. 70
- p. 71
- Issue Table of Contents
- Economic and Political Weekly, Vol. 43, No. 10 (Mar. 8 - 14, 2008), pp. 1-84
- Front Matter
- Letters
- Arguing with Sen on Nationalism [pp. 4-4]
- Editorials
- Is the Union Budget a Federal Budget? [pp. 5-6]
- Manufacturing a WTO Agreement [pp. 6-6]
- Australian Apology [pp. 7-7]
- From 50 Years Ago [pp. 7-7]
- Commentary
- Revising Estimates of Poverty [pp. 8-10]
- Amaresh Bagchi: Public Finance Economist Par Excellence [pp. 10-13]
- Cricket, Excesses and Market Mania [pp. 13-15]
- Myth of Judicial Overreach [pp. 15-18]
- Is Rajasthan Heading Towards Caste War? [pp. 19-21]
- Integrative Reconciliation: Mothers in the Naga Movement [pp. 21-23]
- Book Reviews
- Globalisation of Urban India [pp. 26-31]
- The Constitution and Censorship of Plays [pp. 31-33]
- Making Growth Inclusive [pp. 33-35]
- Prespective
- Does Land Still Matter? [pp. 37-42]
- Special Articles
- Ownership Holdings of Land in Rural India: Putting the Record Straight [pp. 43-47]
- Improving Land Access for India's Rural Poor [pp. 49-56]
- History, Memory and Localised Constructions of Insecurity [pp. 57-64]
- Financial Stability in Asian Economies [pp. 65-71]
- Notes
- Doing a Rashomon on the Hindutva Cases [pp. 72-77]
- Discussion
- More on Futures Trading and Commodity Prices [pp. 78-79]
- Current Statistics [pp. 80-81]
- Letters
- Nationalism – Boon or Curse? [pp. 82-82]
- Back Matter
financialstabilityreview201411.en.pdf
F inanc ial STaB i l i TY REV iEW noVEmBER 2014
FINANCIAL STABILITY REVIEW NOVEMBER 2014
In 2014 all ECB publications feature
a motif taken from the €20 banknote.
© European Central Bank, 2014
Address Kaiserstrasse 29 60311 Frankfurt am Main Germany
Postal address Postfach 16 03 19 60066 Frankfurt am Main Germany
Telephone +49 69 1344 0
Website http://www.ecb.europa.eu
All rights reserved. Reproduction for educational and non-commercial purposes is permitted provided that the source is acknowledged.
Unless otherwise stated, this document uses data available as at 14 November 2014.
ISSN 1830-2025 (epub) ISSN 1830-2025 (online) EU catalogue number QB-XU-14-002-EN-E (epub) EU catalogue number QB-XU-14-002-EN-N (online)
3 ECB
Financial Stability Review November 2014
CONTENTS FOREWORd 5
OVERVIEW 7
1 MACRO-FINANCIAL ANd CREdIT ENVIRONMENT 15 1.1 Ongoing moderate recovery, but downside risks on the rise 15
Box 1 Does the growing importance of emerging market banks pose a systemic risk? 22 1.2 Structural reform and fiscal consolidation needs remain high, despite contained
sovereign stress 24 1.3 Gradually improving financing conditions in the non-financial private sector,
but vulnerabilities remain 28
2 FINANCIAL MARkETS 37 2.1 Interbank activity in euro area money markets continues to normalise,
but fragmentation remains 37 2.2 Yields at record lows amid a slight increase in credit risk premia 41
Box 2 Structural and systemic risk features of euro area investment funds 43 Box 3 Financial market volatility and banking sector leverage 51
3 EuRO AREA FINANCIAL INSTITuTIONS 57 3.1 Balance sheet repair continues, but weak profitability persists in the euro area
banking sector 57 Box 4 The ECB’s comprehensive assessment exercise 64 Box 5 Do contingent convertible capital instruments affect the risk perceptions of senior
debt holders? 72 3.2 The euro area insurance sector: resilience amid continued headwinds 76 3.3 Macro-prudential policy measures announced in several countries 82 3.4 Reshaping the regulatory framework for financial institutions, markets
and infrastructures 87 Box 6 Regulatory initiatives to enhance overall loss-absorption capacity 90
SpECIAL FEATuRES 99 A Fire-sale externalities in the euro area banking sector 99 B Capturing the financial cycle in euro area countries 109 C Initial considerations regarding a macro-prudential instrument based on the net
stable funding ratio 118
STATISTICAL ANNEX S1
4 ECB Financial Stability Review November 20144
COuNTRIES BE Belgium LU Luxembourg BG Bulgaria HU Hungary CZ Czech Republic MT Malta DK Denmark NL Netherlands DE Germany AT Austria EE Estonia PL Poland IE Ireland PT Portugal GR Greece RO Romania ES Spain SI Slovenia FR France SK Slovakia HR Croatia FI Finland IT Italy SE Sweden CY Cyprus UK United Kingdom LV Latvia JP Japan LT Lithuania US United States
ABBREVIATIONS
5 ECB
Financial Stability Review November 2014
The Financial Stability Review (FSR) reviews developments relevant for financial stability, in addition to identifying and prioritising main risks and vulnerabilities for the euro area financial sector. It does so to promote awareness of these risks among policy-makers, the financial industry and the public at large, with the ultimate goal of promoting financial stability. The ECB defines financial stability as a condition in which the financial system – intermediaries, markets and market infrastructures – can withstand shocks without major disruption in financial intermediation and in the effective allocation of savings to productive investment.
The FSR also plays an important role in the ECB’s new macro- and micro-prudential tasks. With the establishment of the Single Supervisory Mechanism (SSM), the ECB was entrusted with the macro- prudential tasks and tools provided for under EU law. The FSR, by providing a financial system-wide assessment of risks and vulnerabilities, provides key input to the ECB’s macro-prudential policy analysis. Such a euro area system-wide dimension is an important complement to micro-prudential banking supervision which is more focused on the soundness of individual institutions. At the same time, whereas the ECB’s new roles in the macro- and micro-prudential realms rely primarily on banking sector instruments, the FSR continues to focus on risks and vulnerabilities of the financial system at large, including – in addition to banks – shadow banking activities including non-bank financial intermediaries, financial markets and market infrastructures.
This Review includes several special features that are aimed at deepening the ECB’s financial stability analysis and supporting the work underlying the ECB’s new macro-prudential function. The first analyses asset fire sales as a potential conduit of systemic stress in the banking system. The second discusses the measurement of the financial cycle in euro area countries, which can provide relevant information for the counter-cyclical objective of macro-prudential policies. The third presents some first considerations regarding the potential use of a macro-prudential instrument based on the net stable funding ratio, a new structural liquidity metric developed by the Basel Committee.
The Review has been prepared with the involvement of the ESCB/SSM Financial Stability Committee. This committee assists the decision-making bodies of the ECB, including the Supervisory Board, in the fulfilment of their tasks.
Vítor Constâncio Vice-President of the European Central Bank
FOREWORd
7 ECB
Financial Stability Review November 2014
OVERVIEW Despite intermittent financial market turbulence, euro area systemic stress has remained at low levels. Indicators of stress among euro area banks and sovereigns have declined further to levels last seen before the outbreak of the global financial crisis in 2007. Stress across the broader financial system has also remained contained (see Chart 1).
This belies a delicate situation, in which generally benign financial market sentiment has contrasted with a weak, fragile and uneven economic recovery. In fact, the environment of low nominal growth and high unemployment is the major underlying factor driving the challenges to financial stability. The resulting apparent disconnect between real economic and financial cycles has had implications for credit provision. On the one hand, bank-intermediated credit remains scarce, given a combination of weak demand and credit terms that may discourage borrowing and investment, which hinders the economic recovery. ECB monetary policy action – including the asset-backed securities purchase programme, the new covered bond purchase programme, and the targeted longer-term refinancing operations (TLTROs) – is providing key support, in particular to specific market segments that play a fundamental role in the financing of the economy. On the other hand, market-intermediated credit is rather abundant and available at conditions that resemble pre-crisis standards, underpinned by a global search for yield.
Amid these economic and financial developments, progress has continued in addressing legacy issues from the euro area crisis. A strengthening of euro area bank balance sheets continues, supported by the ECB’s comprehensive assessment. Capital positions have been strengthened further, amid increased transparency and balance sheet repair. Notwithstanding this progress in enhancing balance sheet resilience, many euro area – as well as global – banks are still confronted with profitability challenges amid a cyclical recovery proceeding at different speeds around the globe. At the same time, progress in repairing balance sheets in the non-financial sector also continues apace. In particular, euro area sovereigns have focused on repairing fiscal fundamentals alongside structural reforms, although at an uneven pace across countries. Public debt sustainability challenges nonetheless remain, implying that the work of restoring the soundness of public finances is unfinished, while structural reform efforts are needed to enhance macroeconomic growth prospects.
A combination of these legacy issues as well as emerging risks yields three key risks to euro area financial stability over the next year and a half (see Table 1) that have the potential to be mutually reinforcing if triggered. Underlying all of these key risks is the uncertainty surrounding the weak, fragile and uneven economic recovery and the current period of very low inflation, which has the potential to aggravate and trigger the existing vulnerabilities should the current situation continue for longer than expected or conditions deteriorate further.
Euro area stress has remained moderate…
… but vulnerabilities remain…
… despite progress in addressing banking and sovereign vulnerabilities
Three key risks to euro area financial stability
Chart 1 Measures of financial market, banking sector and sovereign stress in the euro area (Jan. 2011 – 14 Nov. 2014)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0
4
8
12
16
20
24
28 May FSR
Jan. July Jan. July Jan. July Jan. July
probability of default of two or more LCBGs (percentage probability; left-hand scale) composite indicator of systemic stress in financial markets (right-hand scale) composite indicator of systemic stress in sovereign bond markets (right-hand scale)
2011 2012 2013 2014
Sources: Bloomberg and ECB calculations. Notes: “Probability of default of two or more LCBGs” refers to the probability of simultaneous defaults in the sample of 15 large and complex banking groups (LCBGs) over a one-year horizon. For more information on composite indicators of systemic stress, see the notes to Chart 1.11.
8 ECB Financial Stability Review November 201488
Key risk 1: Abrupt reversal of the global search for yield, amplified by pockets of illiquidity, with signs of a growing use of leverage in the non-bank financial sector
Despite bouts of volatility – linked to rising geopolitical tensions and weak economic data – a search for yield has persisted across global financial markets. The price of risk has remained low in most market segments, supported by historically low risk-free rates and measures of market volatility. This has been associated with an increased correlation within and across euro area bond, equity and money markets reminiscent of the years before the onset of the global financial crisis.
In Europe, this global strong demand for riskier assets has been most prominently seen in corporate and sovereign bond markets (see Charts 2 and 3), but also in valuations of other assets such as
Continued global search for yield…
Table 1 key risks to euro area financial stability
Current level (colour) and recent change (arrow)*
1. Abrupt reversal of the global search for yield, amplified by pockets of illiquidity, with signs of a growing use of leverage in the non-bank financial sector
2. Persistent weak bank profitability in a weak, fragile and uneven macroeconomic recovery
3. Re-emergence of sovereign debt sustainability concerns, amid low nominal growth and wavering policy determination for fiscal and structural reforms
pronounced systemic risk *The colour indicates the current level of the risk which is a combination of the probability of materialisation and an estimate of the likely systemic impact of the identified risk over the next year and a half, based on the judgement of the ECB’s staff. The arrows indicate whether this risk has intensified since the previous FSR.
medium-level systemic risk
potential systemic risk
Chart 2 Cumulative changes in bond yields since May 2013
(2 May 2013 – 14 Nov. 2014; cumulative change in basis points; ten-year sovereign bond yields)
-240 -200 -160 -120 -80 -40
0 40 80
120 160 200
-240 -200 -160 -120 -80 -40 0 40 80 120 160 200
May Aug. Nov. Feb. May Aug. Nov.
May FSR
2013 2014
emerging market sovereign bonds US sovereign bonds German sovereign bonds euro area high-yield corporate bonds average for Irish, Italian, Portuguese and Spanish sovereign bonds
Sources: Bloomberg and JPMorgan Chase & Co.
Chart 3 Selected bond yields and expected euro area equity returns
(Jan. 1999 – Oct. 2014; percentages)
0
5
10
15
20
25
0
5
10
15
20
25
1999 2001 2003 2005 2007 2009 2011 2013
euro area expected equity returns emerging market external sovereign debt yield euro area non-financial corporate high-yield bond yield euro area corporate investment-grade bond yield
Sources: Bloomberg, Bank of America/Merrill Lynch indices, R. Shiller (Yale University), ECB and ECB calculations. Note: The euro area expected equity return is the inverted Shiller cyclically adjusted price/earnings ratio.
9 ECB
Financial Stability Review November 2014 9
OVERVIEW
9
equities and the prime segment of commercial property (i.e. modern office and retail space in capital cities). During the bouts of volatility in recent months linked to weak economic data releases and geopolitical tensions, investors showed some signs of increased credit risk aversion, especially towards high-yield corporates and, more recently, vulnerable sovereigns (see Chart 2), but appeared willing to continue to seek yield by increasing duration exposures to higher-rated issuers.
The resilience of strong investor demand for lower-rated bonds, equities and other higher- yielding asset classes depends on continued strong risk appetite. Indeed, outflows from high-yield bonds and bouts of increased financial market volatility in recent months (see Chart 4) highlight investor uncertainty regarding valuations and the potential for sharp adjustments in the future. Global investor sentiment remains sensitive to changes in the economic outlook, geopolitical tensions and emerging market risks, notably related to larger economies such as China. In addition, while monetary policy settings in major economies, including the euro area, provide an anchor for expectations regarding short-term interest rates, yields on longer-dated bonds remain vulnerable to an increase in US term premia.
At the same time, financial stability risks may arise from investor complacency especially during periods of weak returns on financial assets when investors hunt for yield. Such periods have the potential to breed systemic risks, if they lead to an excessive build-up of leverage or maturity extension and mismatches. While leverage in the banking sector has remained in check, signs of increasing leverage have started to emerge in securities markets and among shadow banking entities, albeit from relatively low levels. In addition, continued low yields may place additional pressure on investors to improve returns by taking on higher duration risk exposures.
Along with signs of some increase in leverage and duration, concerns remain that the impact of a possible reversal of hunt-for-yield flows could be amplified by low market liquidity in some segments. For instance, although primary bond markets have seen continued strong investor demand, secondary market liquidity, in particular in corporate bond markets, has deteriorated in recent years. This has included lower daily trading volumes – in an environment of significantly higher amounts outstanding – and a reduction in the number of market-makers. Markets that are important for the functioning of bond markets, such as repo markets, have seen reduced activity since the outbreak of the financial crisis as well.
Although the euro area banking sector remains exposed to the risk of a repricing of market risk, the steady increase in the euro area shadow banking sector in recent years, amid a gradual shift from bank to market-based funding in the economy, suggests that vulnerabilities are likely to have been
… dependent on sustained risk appetite
Vulnerabilities have been growing in the shadow banking sector…
Chart 4 Implied market volatilities
(Jan. 1999 – 14 Nov. 2014; ten-day moving average; index: average since 1999 = 100)
50
100
150
200
250
300
350
50
100
150
200
250
300
350
50
60
70
80
90
100
110
50
60
70
80
90
100
110
1999 2005 2011 May
VIX VSTOXX global FX volatility implied bond market volatility (one-month Treasury options)
Sep. 2014
Source: Bloomberg.
10 ECB Financial Stability Review November 20141010
growing more in this segment (see Chart 5). Any problem confronting investment funds could, however, propagate quickly to the banking sector and the real economy since they are highly interconnected with euro area credit institutions and an important source of funding for euro area banks, non-financial corporates and governments. The euro area investment fund sector has doubled in size since 2009, with assets reaching €8.9 trillion in the third quarter of 2014. Almost all of these funds are open-ended and the share of liquid assets as a percentage of shares/ units issued has declined from 40% in 2009 to 33% in the third quarter of 2014. This raises stability concerns as demandable equity in these funds can have the same fire-sale properties as short-term debt funding. In addition, some segments of the shadow banking sector appear to have become more concentrated, for instance with the largest global asset managers accounting for an increasing share of assets under management.
With monetary policies aimed at preserving price stability, prudential policies are needed to address vulnerabilities from financial excesses. As the potential for adjustment in financial markets remains, micro- and macro-prudential policies need to be considered to ensure that financial intermediaries have sufficient buffers to withstand a reversal of risk premia. It also calls for further initiatives to monitor and assess vulnerabilities in the growing shadow banking sector, and for continued efforts to improve the oversight and the tools available for mitigating action as currently available tools have limited scope to deal with risks from shadow banking activities.
Key risk 2: Persistent weak bank profitability in a weak, fragile and uneven macroeconomic recovery
A confluence of cyclical and structural factors has led to a low profitability or loss-making environment for euro area banks. Clearly, the emergence from crisis and recession in the euro area has had a significant impact – with one-fifth of euro area significant banking groups1 reporting losses in the first half of 2014, albeit down considerably from more than half of the banks reporting losses in the second half of 2013. Sluggish bank profitability has, however, not only been a challenge specific to euro area banks. Their aggregate financial performance closely resembles that of non-euro area European banks and – once correcting for provisioning – also that of their US peers (see Chart 6).
Persistent weak bank profitability could become a systemic concern if it limits banks’ ability to improve their shock-absorbing capacity via retained earnings and provisioning. This could prevent
1 “Significant banking groups” (SBGs) refers to around 90 euro area banking groups (depending on data availability) and is the consolidated group level analogue of the significant banks that fall under direct ECB supervision. Alongside this group of banks, the FSR also contains analysis of a sub-set of 18 euro area “large and complex banking groups” (LCBGs) – which is a sub-set of the SBGs – and 22 global LCBGs which are the largest, least substitutable and most interconnected banks. For further details, see “A new bank sample for the ECB’s Financial Stability Review”, Financial Stability Review, ECB, November 2013.
… which calls for implementation of prudential policies
Bank profitability remains weak…
Chart 5 Assets of selected euro area financial sectors
(Q1 2009 – Q3 2014; index: Q1 2009 = 100)
8080
100100
120120
140140
160160
180180
200200
220220
240240
8080
100100
120120
140140
160160
180180
200200
220220
240240
20092009 20102010 20112011 20122012 20132013 20142014
shadow banks: overallshadow banks: overall shadow banks: investment funds (excluding money market funds) shadow banks: investment funds (excluding money market funds) shadow banks: hedge fundsshadow banks: hedge funds banksbanks bank loans to households and non-financial corporationsbank loans to households and non-financial corporations
Sources: ECB and ECB calculations.
11 ECB
Financial Stability Review November 2014 11
OVERVIEW
11
banks around the world from engaging in new profitable lending activities and lead to more structural business model-related concerns in a low growth environment. In such circumstances, banks might be tempted to take on more risk to improve profitability, which in turn could make them more vulnerable to future shocks.
Cyclical headwinds affecting profitability are expected to dissipate as the economic environment improves, and signs of a levelling-off in the pace of non-performing loan formation have emerged in some countries. That said, the turning point does not appear to have been reached yet in some countries and the fragile and uneven economic recovery points to continued downside risks to the credit quality of banks’ borrowers. At the same time, large one-off costs stemming from past conduct irregularities weigh on banks and could lead to market volatility, in particular for some large euro area banks that are active in capital market businesses.
Although cyclical factors mainly related to high loan loss provisioning needs continue to weigh heavily on euro area banks’ financial performance, the fact that for many banks their return on equity has fallen below their cost of equity – shareholders’ expected rate of return – also points to a structural need for further balance sheet adjustment in parts of the banking system. Current lower levels of profitability are also a result of – the much needed – de-risking of bank balance sheets, including a stark reduction in leverage (see Chart 7). The challenge for banks is therefore to improve profitability without unduly taking on risk.
While the challenges confronting banks can to a large extent be linked to legacy issues stemming from the financial crisis, including the weak economic environment, in some countries signs of new potential risks are emerging. In particular, property market developments in both the residential and (prime) commercial segments in some countries have been frothy, leaving property markets
… due to both cyclical and structural factors…
Chart 6 pre- and post-provision return on equity of euro area and global large and complex banking groups (LCBgs) (H1 2007 – H1 2014; percentages; medians; two-period moving average)
Pre-provision return on equity Return on equity
5
10
15
20
25
30
35
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
35
0 2007 2009 2011 2013 2007 2009 2011 2013
euro area LCBGs non-euro area European LCBGs US LCBGs
Sources: SNL Financial and ECB calculations. Note: “Non-euro area European LCBGs” include banks from the United Kingdom, Switzerland, Sweden and Denmark.
Chart 7 Return on equity and leverage of euro area significant banking groups
(Q1 2004 – 2015; medians)
15
16
17
18
19
20
21
0
4
8
12
16
20
24
2004 2006 2008 2010 2012 2014
leverage (ratio; right-hand scale) return on equity (percentage; left-hand scale) return on equity, analyst forecast for 2015 (percentage; left-hand scale)
Sources: Bloomberg and ECB calculations.
12 ECB Financial Stability Review November 20141212
vulnerable to correction should investor sentiment deteriorate. The numerous property- related instruments in the newly acquired macro- prudential toolkit may contribute to attenuating financial cycles, while also increasing the resilience of banks and their borrowers.
Amid current legacy and new challenges confronting banks, efforts to clean up balance sheets, bolster capital positions and adjust business models continue. Bank balance sheets have been strengthened further, with a clear shift towards capital increases in 2014 – related to the comprehensive assessment carried out by the ECB – from deleveraging and de-risking in previous years. This has happened amid a significant reduction in the size of bank balance sheets since mid-2012. However, signs have emerged that the asset reduction process might have come to an end (see Chart 8), which, together with continuously improving capital buffers, can be seen as a positive development as it suggests that the trough in the bank performance cycle might have been reached. Aggregate data, however, conceal notable differences across banks and countries and further adjustments are needed in parts of the banking sector.
While the comprehensive assessment ensured that significant banks in the euro area have sufficient capital levels, progress needs to continue in parts of the banking system to address remaining fragilities and uncertainties. Further measures in this respect need to be taken mainly by banks themselves and needed action is likely to differ across banks or national banking sectors depending on whether banks are, for instance, faced with possible overcapacity in parts of the banking sector, high costs, or limited diversification of their income sources. In addition, continued prudent asset valuation enforcement, as well as timely and accurate risk controls by banks, should encourage banks to develop appropriate systems to deal with credit risk and enhance their capacity to deal with distressed borrowers. At the same time, further official sector policies can also provide support – in particular, legal frameworks should be sought that facilitate a timely and low-cost resolution of non-performing loans, thereby enabling a smooth interaction between banks and their distressed borrowers and freeing up additional lending capacity.
Key risk 3: Re-emergence of sovereign debt sustainability concerns, amid low nominal growth and wavering policy determination for fiscal and structural reforms
Sovereign stress has remained contained in the euro area since the publication of the May FSR, albeit with increasing challenges from a deterioration in the economic growth outlook. Building on the improved sovereign debt market conditions following the announcement of Outright Monetary Transactions in 2012 and more recent ECB policy action, market sentiment – especially towards more vulnerable euro area countries – has remained relatively favourable in recent months. Some gradual strengthening in cyclical economic conditions and the ongoing adjustment of fiscal fundamentals underpinned this development. The aggregate euro area fiscal deficit is expected to continue to fall and stay below the 3% Maastricht threshold this year as consolidation efforts and
… although efforts to build a stronger
banking sector have been significant
Sovereign stress has remained
contained…
Chart 8 Evolution of total assets of euro area monetary financial institutions
(May 2012 – Sep. 2014; EUR trillions)
30.0
30.5
31.0
31.5
32.0
32.5
33.0
33.5
34.0
34.5
35.0
30.0
30.5
31.0
31.5
32.0
32.5
33.0
33.5
34.0
34.5
35.0
€34.9 trillion
€31.2 trillion
May MaySep. Sep.Jan. May Sep.Jan. 2012 2013 2014
Source: ECB. Note: The red bars signal a decline and the green bars an increase in the given month.
13 ECB
Financial Stability Review November 2014 13
OVERVIEW
13
various reform and administrative measures have started to bear fruit and tax revenues have grown more strongly than initially expected in some countries. In addition, the unwinding of financial sector support is expected to contribute positively to the improvement of fiscal positions in 2014 and beyond in many countries (see Chart 9).
Sentiment towards sovereigns has also been supported by continued progress towards weakening the links between sovereigns and banks. Most notably, banking union preparations have continued with its first pillar, the Single Supervisory Mechanism (SSM), in place since 4 November. Regulatory initiatives – such as new bail-in rules – have also helped to weaken links between euro area banks and sovereigns, although the continued significant correlation in euro area banks’ and sovereigns’ borrowing costs highlights the need for continued progress (see Chart 10).
Despite the relatively benign sentiment towards euro area sovereigns, public debt sustainability challenges persist in the context of continued high debt levels in many countries, heightened
… but public debt sustainability challenges persist…
Chart 9 general government debt and deficits in the euro area
(percentage of GDP)
Austria
Belgium
Cyprus
Germany
Spain
Estonia
Finland
France
Greece
Ireland
Italy
Luxembourg Latvia
Malta Netherlands
Portugal
Slovakia
Slovenia
euro area
0
20
40
60
80
100
120
140
160
180
0
20
40
60
80
100
120
140
160
180 High debt and high deficits
High deficits
High debt
-1 0 1 2 3 4 5 6
2015 2014
x-axis: public deficits y-axis: general government debt
Source: European Commission.
Chart 10 Sovereign and bank credit default swap spreads
(July 2011 – 14 Nov. 2014; basis points)
0
50
100
150
200
250
300
350
400
450
0
50
100
150
200
250
300
350
400
450
0 50 100 150 200 250 300 350 400 450
y-axis: bank CDS spreads x-axis: sovereign CDS spreads
euro area July 2011 – May 2014 euro area May – Nov. 2014 global July 2011 – May 2014 global May – Nov. 2014
Sources: Bloomberg and ECB calculations. Note: Average CDS spread for euro area and global LCBGs versus the average sovereign CDS spread where the LCBGs are headquartered (France, Germany, Italy, Spain and the Netherlands for euro area LCBGs and the United States, the United Kingdom, Switzerland, Denmark, Sweden and Japan for global LCBGs).
14 ECB Financial Stability Review November 20141414
downside risks to the economic outlook and a low inflation environment. Uncertainties relating to sovereign debt sustainability are likely to remain over the medium term as government debt-to-GDP ratios are projected to stay at levels well above 100% in several euro area countries. This highlights the need for further adjustment of fiscal and economic fundamentals relevant for debt sustainability.
Debt sustainability concerns in the medium term remain susceptible to potential setbacks related to the aforementioned necessary further adjustment as well as the weak nominal growth outlook. The needed adjustment could run the risk of being delayed due to the recent relative calm in euro area financial markets, which has the potential to breed complacency in terms of fiscal consolidation and structural reforms. Reinforced rules at the European level should help to mitigate such risks, but reform fatigue or complacency at the national level could lead to a reassessment of sentiment towards euro area sovereigns. Debt sustainability concerns could also resurface during a prolonged period of very low inflation or if the economic outlook deteriorates, which would limit governments’ room for manoeuvre for further fiscal adjustment.
Clearly, risks to the sovereign outlook are also closely linked to the risks stemming from the ongoing global search for yield or further stress in the banking sector. A generalised abrupt reversal of the global search for yield could lead to renewed increases in sovereign bond yields, in particular in lower-rated euro area countries, and could also translate into losses for banks on their sovereign debt holdings. In addition, while new bail-in rules will help shield public balance sheets and taxpayers from future national costs of bank recapitalisation, the potential for renewed adverse feedback loops between banks and sovereigns also remains. Continued efforts to swiftly and effectively implement all pillars of the banking union as well as the Bank Recovery and Resolution Directive are therefore needed.
MACRO-pRudENTIAL pOLICY ACTION ANd REguLATORY INITIATIVES Progress towards a safer financial system continues, with macro-prudential policy action and regulatory advancements at both the European and global levels.
In line with newly acquired macro-prudential policy mandates, a number of euro area countries have already announced and also implemented macro-prudential measures. These include systemic risk measures aimed at mitigating vulnerabilities stemming from the significant size, high concentration and interconnectedness of banking sectors. Different types of residential property measures have been adopted as well, with the aim of addressing unfavourable developments in property markets.
In the regulatory field, progress in strengthening banking sector resilience has continued, including weakening the links between sovereigns and banks. Notably, significant achievements have been made since the publication of the last issue of the FSR in the areas identified as central elements of an integrated financial framework in Europe, particularly in the euro area, namely the establishment of a Single Supervisory Mechanism, a common resolution framework and a Single Resolution Mechanism, along with more harmonised deposit insurance scheme parameters. Measures have not only been taken in the banking domain, however, but also across other financial institutions, as well as market infrastructures. Perhaps most importantly, as shadow banking grows in breadth around the world, regulation of this segment has also gathered pace.
… which are also linked to search for
yield and banking sector concerns
Policy action and regulatory
advancements have continued
15 ECB
Financial Stability Review November 2014
1 MACRO-FINANCIAL ANd CREdIT ENVIRONMENT Macro-financial conditions remain fragile in the euro area, with a very modest economic recovery contrasting with generally robust financial market sentiment. At the country level, fragmentation of the real economy continues to weigh on the underlying growth momentum despite further progress in euro area rebalancing. Risks surrounding the fragile, low nominal growth environment appear to have risen. In particular, geopolitical tensions – despite a limited global impact to date – have the potential to reignite risk aversion in financial markets and potentially also trigger a broad-based adjustment in global capital flows. Ultimately, uncertainties regarding the pace and sustainability of economic recovery in both emerging and advanced economies within and outside the euro area remain – amid continued macro-financial vulnerabilities and structural reform needs along the path to normalisation of macroeconomic policies in some major advanced economies.
Euro area sovereign stress has remained contained amid further improving market sentiment towards more vulnerable euro area economies, as well as some gradual strengthening in cyclical economic conditions and the ongoing adjustment of fiscal fundamentals. Risks nonetheless have increased in the current fragile growth environment, with related challenges for several countries in durably restoring the sustainability of public finances in the context of a prolonged period of low inflation and heightened downside risks to the economic outlook.
The weak economic recovery to date has also entailed challenges for the non-financial private sector, given muted developments in income and earnings. At the same time, household and corporate indebtedness remain high in several euro area countries. Financing conditions for euro area households and firms continued to ease, while recently introduced unconventional measures by the Eurosystem will help further reduce persistent fragmentation across countries and firm sizes. With time, a strengthening macroeconomic recovery should gradually translate into improved income and earnings prospects for households and non-financial corporations, which – together with the favourable interest rate environment – should help support the ongoing process of balance sheet repair.
In this environment, overall developments in euro area property markets have shown incipient signs of recovery, in particular driven by a turnaround in some countries and market segments with considerable post-crisis adjustments. Nonetheless, fragmentation across countries and different property types remains pronounced, albeit declining in terms of both price developments and valuations. The ongoing hunt for yield in prime commercial property, a weaker than expected economic recovery as well as possible corrections in some jurisdictions and regions with signs of overvaluation still represent risks to financial stability going forward.
1.1 ONgOINg MOdERATE RECOVERY, BuT dOWNSIdE RISkS ON ThE RISE
The economic recovery has continued in the euro area in 2014, but has lost some of its momentum amid a softening in some major euro area countries towards the middle of the year. Aggregate euro area economic growth continued to be supported by domestic demand, which benefited from favourable real income developments and financing conditions. The economic recovery has been also buttressed by further reduced macroeconomic uncertainty, with all the different types of uncertainty now below their long-run average despite some pick-up in financial market uncertainty more recently (see Chart 1.1).
The recent slowdown in momentum notwithstanding, the economic recovery in the euro area is expected to continue on a moderate upward path in the medium term. Support stems from an accommodative monetary policy stance (notably including the standard and non-standard
A loss of economic momentum…
… but the recovery continues at a modest pace…
16 ECB Financial Stability Review November 20141616
Eurosystem measures introduced in June and September 2014), favourable financial market conditions, recovering world trade, a depreciation of the euro, gradual improvements in the labour market as well as continued fiscal consolidation and structural reforms in several euro area countries. The September 2014 ECB staff macroeconomic projections for the euro area indicate annual real GDP growth of 0.9% for 2014 – somewhat lower than the 1.1% projected back in June and corresponding to a similar downward revision to the outlook of professional forecasters.
The economic recovery remains fragile going forward in the light of increasing downside risks to the economic outlook (see Chart 1.2). Over the short to medium term, several factors could weigh significantly on the underlying euro area growth momentum, including the heightened geopolitical tensions across the globe, the still ongoing process of balance sheet repair in the financial and non-financial private sectors and the continued need for further fiscal
… amid increasing downside risks
Chart 1.1 Economic, political and financial market uncertainty in the euro area
(Q1 2006 – Q3 2014; standard deviations from average over 1996-2014)
-2
-1
0
1
2
3
4
-2
-1
0
1
2
3
4
2006 2007 2008 2009 2010 2011 2012 2013 2014
financial market uncertainty economic uncertainty political uncertainty
Sources: Consensus Economics, Eurostat, Baker, Bloom and Davis (2013), European Commission and ECB calculations. Notes: Based on households’ and firms’ perceived uncertainty about the future economic situation taken from surveys (economic), various financial market indicators (financial market) and economic policies (political). For further details on the methodology, see “How has macroeconomic uncertainty in the euro area evolved recently?”, Monthly Bulletin, ECB, October 2013.
Chart 1.2 distribution of the 2015 real gdp growth forecasts for the euro area and the united States (probability density)
Jan. 2014 forecast for 2015 May 2014 forecast for 2015 Nov. 2014 forecast for 2015
x-axis: real GDP growth rate
a) euro area b) United States
0.0
0.5
1.0
1.5
2.0
2.5
0.0
0.5
1.0
1.5
2.0
2.5
0.50.0 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4. 5 0.0
0.5
1.0
1.5
2.0
2.5
0.0
0.5
1.0
1.5
2.0
2.5
Sources: Consensus Economics and ECB calculations.
17 ECB
Financial Stability Review November 2014 17
I� Macro-F I�nancI�al and credI�t
envI�ronMent
17
consolidation in some countries. That said, the economic outlook for the euro area remains subdued in comparison to the growth prospects of other major advanced and emerging market economies (see Chart 1.3).
At the euro area country level, real fragmentation – albeit significantly lower than during the euro area sovereign debt crisis – remains a cause for concern, with more recently some signs of a renewed widening in the cross-country dispersion of growth rates. Moreover, labour market conditions are still very divergent within the euro area, as high unemployment in more vulnerable countries contrasts with relatively benign labour market conditions in other euro area economies (see Chart 1.4). This heterogeneity continues to highlight the need for employment-enhancing structural reforms with a view to fostering an inclusive economic recovery.
Efforts to restore competitiveness are ongoing in a number of euro area countries. The overall competitiveness of more vulnerable euro area countries has improved considerably since the onset of the crisis, as indicated by major current account corrections. A large part of the underlying adjustment has been of a non-cyclical nature and is therefore likely to be sustained (see Chart 1.5). In the context of the ongoing rebalancing, structural reforms have proven decisive for competitiveness gains in several vulnerable euro area countries, as shown by notable improvements in global competitiveness rankings, for example in Greece and Portugal (see Chart 1.6). Still, structural reforms need to continue in order to help further reduce the real and financial fragmentation across the euro area, to enhance the euro area’s medium-term growth potential and to further narrow the still sizeable, albeit diminishing, negative output gaps, particularly in vulnerable euro area economies.
Real fragmentation remains a cause for concern…
… despite the ongoing rebalancing in the euro area
Chart 1.3 Evolution of forecasts for real gdp growth in selected advanced and emerging economies for 2015 (Jan. 2014 – Nov. 2014; percentage change per annum)
0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
6
7
8
euro area
United States
United Kingdom
China Brazil
Russia
India
Japan
Jan. Feb. Mar. Apr. May June July Aug. Sep. Oct. Nov. 2014
Sources: Consensus Economics and ECB. Note: The chart shows the minimum, maximum, median and interquartile distribution across the 11 euro area countries surveyed by Consensus Economics (Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, the Netherlands, Portugal and Spain).
Chart 1.4 developments in the number of unemployed across the euro area
(Jan. 2008 – Sep. 2014; index: Q1 2008 = 100)
80
100
120
140
160
180
200
220
240
260
80
100
120
140
160
180
200
220
240
260
2008 2009 2010 2011 2012 2013 2014
euro area recession
euro area average vulnerable euro area countries
other euro area countries
Sources: Eurostat and ECB. Note: Vulnerable euro area countries include Cyprus, Greece, Ireland, Italy, Portugal, Slovenia and Spain.
18 ECB Financial Stability Review November 20141818
A gradual recovery in global economic activity has continued. Economic momentum in advanced economies strengthened further, albeit at an uneven pace across regions, while growth in emerging markets has also rebounded after a temporary dip in 2013. The accommodative monetary policy stance in advanced economies – though showing signs of divergence – has continued to provide vital support to a fragile economic recovery. Notwithstanding the recent rise, overall volatility appears to have remained subdued in global financial markets (see Chart 1.7). In particular, emerging markets have witnessed a drop in financial market pressures (see Chart 1.8), as improved global risk sentiment has encouraged capital flows back to emerging markets following intermittent turbulences since mid-2013. While global growth is expected to pick up gradually, risks to the global outlook remain tilted to the downside. Heightened geopolitical risks, persistent macroeconomic and/or financial imbalances, as well as a sharp repricing of risk with ensuing corrections in asset prices and a potential disorderly unwinding of capital flows, could have negative repercussions for the global economy.
Zooming in on the main global economic regions, economic momentum in many advanced economies outside the euro area is firming slowly, despite short-term volatility. Recent trends indicate a continued recovery ahead, but the pace of progress varies across countries, as still weak labour market conditions, continued balance sheet adjustments in the financial and non-financial
Global recovery continues along a
gradual but uneven growth path
Economic momentum in
advanced economies is firming slowly…
Chart 1.6 Changes in competitiveness across the euro area
(2010, 2014; ranks)
0 5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
average rank (2010) average rank (2014) range of ranks (2014)
FI DE NL IE AT PT SI ES LV SK IT GRBE FR EE LU CY
Sources: World Bank, International Finance Corporation, World Economic Forum, Transparency International and ECB calculations. Notes: The average rank represents the simple average of the country’s rank in Transparency International’s Corruption Perceptions Index (2013, 2010), the World Economic Forum’s Global Competitiveness Report (2014-15, 2010-11) and the World Bank/International Finance Corporation’s Ease of Doing Business (2015, 2011) rankings.
Chart 1.5 Current account rebalancing across the euro area
(2008 – 2013; percentage of GDP)
-17.5
-15.0
-12.5
-10.0
-7.5
-5.0
-2.5
0.0
2.5
5.0
7.5
10.0
12.5
non-cyclical change cyclical change current account balance 2008 current account balance 2013
CYGR LV PT ES EE SK IE SI MT IT FR BE EA FI NL AT LU DE
Sources: ECB and ECB calculations. Notes: The estimates of cyclical and non-cyclical changes are based on a current account model in the vein of the IMF’s External Balance Assessment. For further details see External Balance Assessment Methodology: Technical Background, Research Department, IMF, June 2013.
19 ECB
Financial Stability Review November 2014 19
I� Macro-F I�nancI�al and credI�t
envI�ronMent
19
private sectors and a still incomplete process of fiscal consolidation continue to weigh on near-term growth prospects in several countries. Beyond the support of accommodative monetary policies, economic growth in advanced economies will increasingly benefit from waning private sector deleveraging and fiscal drag, enhanced confidence and falling unemployment.
In the United States, the economic recovery has gained traction after a weather-related weak start to 2014, supported by favourable housing and labour market developments. Notwithstanding some temporary increase in volatility in October, financial conditions have eased further, with indicators of financial stress at or near all-time lows. In the context of generally improving economic prospects, the Federal Reserve concluded its asset purchase programme in October, while preserving a highly accommodative monetary stance, as reflected by the low target range for the federal funds rate and the expected maintenance of its longer-term securities holdings at sizeable levels. Looking ahead, economic activity should become more sustained due to the ongoing recovery in labour and housing markets, accommodative monetary and financial conditions, as well as fading headwinds from fiscal policy and household deleveraging, with the household debt-to-income ratio now having arguably returned to levels closer to equilibrium.1
In Japan, the economy contracted strongly in the second quarter of 2014, as demand rebalanced after the VAT hike in April and the frontloaded spending in the first quarter. Growth is expected to resume towards the end of 2014 supported by accommodative monetary policy, including a newly adopted set of measures taken at the end of October. Despite the consumption tax rise, fiscal challenges remain and fiscal consolidation over the medium term remains a necessity to ensure long-term debt sustainability. In addition, banks’ sovereign exposure, albeit declining, remains a concern for the profitability and solvency of the 1 For further details, see Albuquerque, B., Baumann, U. and Krustev, G., “Has US household deleveraging ended? A model-based estimate
of equilibrium debt”, Working Paper Series, No 1643, ECB, March 2014.
… but risks continue to be tilted to the downside
Chart 1.7 Financial market volatility and economic policy uncertainty
(2003 – 2014; number of standard deviations)
Ec on
om ic
p ol
ic y
un ce
rta in
ty
Fi na
nc ia
l m ar
ke t
vo la
til ity
S&P 500
Europe
China
India
pre-crisis financial crisis 2014
-2
-1
0
1
2 US
UK
EUR/USDEME FX index
US interest rate swaps
Sources: Haver Analytics, Bloomberg, ECB and Baker, Bloom and Davis (2013). Notes: The chart shows differences in the number of standard deviations between the values of each indicator over three different periods and the average of 2003-14. The pre-crisis period refers to January 2003 – August 2007. The financial crisis period refers to August 2007 – December 2009, while 2014 refers to available observations this year. Series above the horizontal line are measures of economic policy uncertainty from Baker, S., Bloom, N. and Davis, S., “Measuring Economic Policy Uncertainty”, Chicago Booth Research Paper No 13/02, January 2013. Below the line, implied volatilities on three-month USD/EUR options, the JPMorgan emerging market FX index, the S&P 500 and US interest rate swaps are shown.
Chart 1.8 Financial conditions in selected advanced and emerging market regions
(Jan. 2006 – Nov. 2014; number of standard deviations)
-14 -12 -10 -8 -6 -4 -2 0 2 4 6 8
10
-14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10
2006 2007 2008 2009 2010 2011 2012 2013 2014
United States euro area Asia (excluding Japan)
ea sin
g fin
an ci
al co
nd iti
on s
tig ht
en in
g fin
an ci
al co
nd iti
on s
Source: Bloomberg.
20 ECB Financial Stability Review November 20142020
Japanese banking sector, in case of a repricing of risk in financial markets and the related potential increase in government bond yields.
The United Kingdom maintained its strong growth momentum in the first half of 2014, buttressed by rising household confidence, improving labour market conditions and a buoyant housing market. Leading indicators point to a growth moderation in the short run, while structural factors, such as the need for further balance sheet repair in the private and public sectors, will weigh on economic activity over the medium term. Headwinds also relate to ongoing geopolitical risks and the sharp housing market recovery that may provide some relief for highly indebted households, but may also render them more vulnerable to potential corrections in property markets.
Emerging markets have benefited from easing financial market pressures in recent months, as reflected by lower sovereign bond spreads, stabilising equity prices and renewed capital inflows into some of those countries that were more affected by the bouts of volatility in 2013 and early 2014 (see Chart 1.9). This was underpinned by progress in reducing macroeconomic imbalances in some of the more vulnerable economies, also bolstering investor confidence. In tandem with reduced financial stress, economic activity rebounded in several countries, even though remaining rather subdued relative to past years’ experience. Going forward, growth in some emerging economies is likely to be restrained by structural factors, such as infrastructure bottlenecks and capacity constraints, while in other countries that were highly dependent on capital inflows, activity is likely to be dampened as economies rebalance and adjust to tighter financial conditions and the expected adjustment of US monetary policy. In the latter context, depleted foreign exchange reserves after the 2013-14 tensions may render some emerging economies with larger external imbalances more vulnerable.
The economic recovery continued in most emerging European economies, notably the EU countries in central and eastern Europe, supported by strong exports and domestic demand. The impact of the Ukraine-Russia crisis on the region has remained contained to date, given rather limited direct trade linkages and contained financial market spillovers. Still, the main downside risk to the region’s economic recovery is related to a further escalation of this conflict, which could also lead to a deepening of sanctions between the EU and Russia. Given strong trade and financial linkages, economic activity in the region is expected to benefit from the ongoing gradual euro area recovery, but also from a further strengthening of domestic demand. However, the outlook for domestic demand in several countries continues to be constrained by a still incomplete process of balance sheet adjustment in the private and public sectors, which in some countries is further complicated by existing currency mismatches. In spite of improved economic activity, credit growth remains subdued in most countries amid a still elevated level of non-performing loans and the ongoing
Emerging markets are recovering as
financial tensions subside
Growth outlook in emerging Europe
benefits from gradual euro area
recovery
Chart 1.9 Equity and bond flows to advanced and emerging market economies
(Jan. 2008 – Nov. 2014; index: Jan. 2008 = 100)
50
75
100
125
150
175
200
225
50
75
100
125
150
175
200
225
2008 2010 2012 2014
eastern Europe, the Middle East and Africa Latin America vulnerable euro area countries Asia other advanced economies other euro area countries
50
75
100
125
150
175
200
225
50
75
100
125
150
175
200
225
2008 2010 2012 2014
Bonds Equities
Source: EPFR. Note: Bonds include both sovereign and corporate bonds.
21 ECB
Financial Stability Review November 2014 21
I� Macro-F I�nancI�al and credI�t
envI�ronMent
21
deleveraging by foreign banks. At the same time, foreign banks have continued to adjust towards a more self-sustained and domestically funded business model that should help mitigate risks to financial stability in the region.
Following a period of weak growth, economic momentum has strengthened in emerging Asia, particularly in India and some other emerging Asian economies, where growth prospects have started to improve following corrections in external imbalances and structural reforms, although growth in China has weakened lately. Looking ahead, a gradual moderation in regional growth dynamics is expected, mainly in China, where high credit growth and leverage, as well as a strongly expanding shadow banking sector, require close monitoring. A slowdown in China would have knock-on effects for other Asian economies with close trade and financial links, but downside risks continue to relate to larger than expected spillovers from capital outflows linked to Federal Reserve tapering. Economic activity in Latin America has lost some traction in 2014 and growth has become more uneven across economies. In Brazil, the economy dipped into outright recession following three years of weak growth, while a deeper recession is underway in Argentina, where the outlook has deteriorated further after the debt default in July. The region is expected to undergo a period of subdued growth before gradually benefiting from improved external demand. Risks to the outlook remain tilted to the downside. The main concerns relate to a further tightening of external financing conditions, a more pronounced decline in commodity prices and the risk of a prolonged period of weakness in economic activity in Brazil.
Overall, a moderate but uneven global recovery across countries and regions is expected, with inherent fragilities being somewhat masked by continued benign financial market sentiment. Risks remain tilted to the downside as long-standing and newly emerging underlying vulnerabilities continue to pose a threat to recovery across the globe. Alongside persistent real and financial global imbalances, which remain high in a historical context despite having narrowed markedly since the onset of the global crisis, the recent intensification of geopolitical tensions represents an increasing cause for concern – this not only in the context of the still ongoing Ukraine-Russia crisis, but also related to other recent incidents in the Middle East. These tensions have the capacity to trigger a spike in commodity prices that may endanger the global recovery and also contribute to preserving global imbalances. This said, intensified geopolitical risks have had a largely muted effect on commodity markets to date (see Chart 1.10), with oil and non-oil commodity prices generally declining over the past months driven by short-run supply- side (e.g. limited disruptions to oil production, emergence of alternative production techniques) and demand-side (e.g. moderate global growth, buoyant risk appetite) fundamentals, which have offset upward pressures related to heightened
Economic activity has rebounded in Asia, but lost momentum in Latin America
Rising geopolitical tensions represent an increasing cause for concern…
Chart 1.10 Selected commodity price developments
(Jan. 2006 – Nov. 2014; index: Dec. 2005 = 100)
0
100
200
300
400
500
600
0
100
200
300
400
500
600
2006 2007 2008 2009 2010 2011 2012 2013 2014
oil gold platinum wheat silver copper
Source: Bloomberg.
22 ECB Financial Stability Review November 20142222
geopolitical risks. Lastly, the risk of a disorderly and broad-based unwinding of global search- for-yield flows as a result of a faster than expected exit from unconventional monetary policies by some major central banks in advanced economies remains a cause for concern.
In sum, important macro-financial risks to euro area financial stability stem from global factors, including rising geopolitical tensions as well as uncertainties regarding the pace and sustainability of the economic recovery in emerging and advanced economies. Most notably, the risk of possible renewed tensions in global financial markets coupled with a potential unwinding of search-for- yield flows continues to represent a cause for concern. At the same time, macro-financial risks also continue to originate from within the euro area in a fragile, low nominal growth environment. In particular, the still ongoing process of balance sheet adjustment in both the financial and non-financial sectors in several countries and continued (albeit diminishing) real and financial fragmentation still weigh on euro area growth momentum.
... with related risks to euro area financial stability
Box 1
dOES ThE gROWINg IMpORTANCE OF EMERgINg MARkET BANkS pOSE A SYSTEMIC RISk?
One side effect of the global financial crisis has been strong growth in the weight of emerging market banks in the global financial system. Indeed, financial deepening in emerging markets has accelerated in recent years as the financial crisis has triggered both increased capital flows to these economies, as well as deleveraging of banks in advanced economies. By the end of 2013, 28 of the 100 largest banks globally were headquartered in emerging markets, compared with 17 only five years earlier (see Chart A). As the resulting geographical structure of the global financial system has evolved, the monitoring of risks clearly also needs to be adapted.
Tracking the main regions exhibiting a rapid expansion of financial sector size, banks from six emerging market economies (EMEs) are represented in the set of the 100 largest banking groups worldwide – i.e. China (15), Brazil (4), South Korea (4), Singapore (2), Russia (2) and India (1). Also, the market capitalisation of emerging market banks has almost quadrupled since the peak of the financial crisis and accounted for 35% of global bank market value just before the onset of the “taper tantrum” in May 2013 (see Chart B). Against this background, the purpose of this box is to provide empirical evidence about whether or not, in line with the share of the emerging market financial sector in world markets, their systemic importance for the global financial system has increased over the recent past.
Chart A Number of emerging market banks in the world’s 100 largest banks by total assets (number of institutions)
0
5
10
15
20
25
30
0
5
10
15
20
25
30
2008 2013
South Africa India Singapore Russia South Korea Brazil China
Source: relbanks.com
23 ECB
Financial Stability Review November 2014 23
I� Macro-F I�nancI�al and credI�t
envI�ronMent
23
To gauge the systemic importance of emerging market banks, two popular measures of conditional risk (co-risk) can be employed: the conditional value at risk (CoVaR) and the conditional expected shortfall (CoES).1 These measures capture tail dependence between equity price return distributions of individual institutions and the financial system as a whole. In this application, the two metrics represent, respectively, the value at risk (VaR) and the expected shortfall (ES) of the global banking system conditional on a particular emerging market bank being in distress.2
The model estimates suggest that, despite rapid growth in emerging market banks, there has not been a meaningful increase in the systemic importance of emerging market banks for the global banking system.3 In fact, the two co-risk measures indicate that, at times when emerging market banks were at risk, the global banking sector experienced a median loss in the range of one to two times of the daily standard deviation prevailing in the respective calendar year (see Charts C and D). The evolution of the two co-risk measures over time does not exhibit a downward-sloping trend, i.e. more negative returns for the global banking sector during periods of financial stress among emerging market banks. If anything, the co-risk measures have, in recent years, moderated towards lower conditional losses in global banking sector prices, whereas they peaked in periods of global or euro area market turbulence in 2008, 2009 and 2011.4
Overall, the empirical evidence confirms earlier findings in the literature suggesting that tail dependence measures, like standard correlation coefficients, tend to increase globally in periods of global market turbulence. At the same time, the above findings are consistent with recent studies on emerging market banks which find that the global footprint of emerging market banks has remained regionally confined so far.5 Notwithstanding this finding, a changing geographical importance of global financial institutions requires close monitoring given the prospect that market prices underlying these empirical measures may adapt in ways that cause past empirical
1 See Brunnermeier, M. K. and Adrian, T., “CoVaR”, Federal Reserve Bank of New York Staff Reports, No 348, September 2008 (revised in September 2011).
2 CoVaR/CoES are measures of the excess loss of the euro area banking system at the tail of a bank i’s return distribution, implied by the bank’s individual VaR/ES at the qth percentile, relative to its median.
3 The sample of emerging market banks is composed of the three largest, non-foreign-owned, listed banks (in terms of total assets) from six EMEs which have systemically relevant financial sectors according to the IMF as well as three advanced Asian economies that exhibit a high degree of integration with the banking sector in emerging Asia.
4 A major caveat of this CoVaR/CoES approach is that any interdependence of price movements between emerging market banks and the global financial system may also stem from global factors. At the same time, the presented set-up largely rules out the possibility of reverse causality (i.e. that shocks to the global banking sector determine price movements of emerging market banks).
5 See Van Horen, N., “Branching Out: The Rise of Emerging Market Banks”, in Reuttner, I. (ed.), The Financial Development Report 2012, World Economic Forum, New York, 2012; and BIS, “EME banking systems and regional financial integration”, CGFS Publications, No 51, Committee on the Global Financial System, March 2014.
Chart B Emerging market banks’ market capitalisation and share in global bank market value (Jan. 2004 – Sep. 2014; USD billions; percentages)
0.0
0.5
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0
250
500
750
1,000
1,250
1,500
1,750
2,000
2,250
2004 2006 2008 2010 2012 2014
emerging market banks’ share in global bank market value (right-hand scale) emerging market banks’ market value (left-hand scale)
Source: Datastream.
24 ECB Financial Stability Review November 20142424
1.2 STRuCTuRAL REFORM ANd FISCAL CONSOLIdATION NEEdS REMAIN hIgh, dESpITE CONTAINEd SOVEREIgN STRESS
Sovereign stress in the euro area has remained contained, with the composite indicator of systemic stress in sovereign bond markets being close to levels last seen before the financial crisis despite a small uptick more recently (see Chart 1.11). While fiscal positions are generally on a more solid footing than at the height of the sovereign debt crisis on account of consolidation efforts, gradually strengthening economic growth and favourable financing conditions, further reform progress over the past six months has been uneven across euro area countries. In terms of fiscal adjustment, in some countries (e.g. Cyprus, Ireland and Spain) various reform and administrative measures have started to bear fruit and tax revenues have grown more strongly than initially expected. At the same time, most recent incoming macroeconomic data have shown a loss of economic momentum amid uncertainties surrounding the reform process in some euro area countries.
Sovereign stress in the euro area
has remained contained…
regularities to break down. Moreover, while a mainly regional footprint may limit the prospect of systemic risk at the global level, regional aspects may nonetheless be relevant for euro area financial stability. Emerging market banks located in EU neighbouring countries have recently intensified their financial linkages with the euro area/EU, for instance by setting up offices in the EU and by participating actively in deposit gathering and loan operations in the region. Given that financial stress among emerging market banks can be transmitted to the euro area via both direct and indirect exposures, significant emerging market banks in general can have financial stability repercussions on the euro area financial sector.
Chart C daily value at risk of the global financial system conditional on EME banks at risk (ΔCoVaR1%
system|i) (2006 – 2013; percentage of daily standard deviation)
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
-4.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
2006 2007 2008 2009 2010 2011 2012 2013
Sources: Bloomberg and ECB calculations. Notes: The charts depict the distribution and the median CoVaR/CoES estimates based on eight non-overlapping annual samples of daily observations from 2006 to 2013. The black line represents the median of the 26 EME banks’ daily ΔCoVaRq
system|i/ ΔCoESq system|I in per cent of the daily standard
deviation of the global banking sector’s return distribution. A negative (positive) value represents a conditional loss (gain). The blue box represents the 25% to 75% quantile of banks. The blue vertical lines represent the minimum and the maximum estimates.
Chart d daily expected shortfall of the global financial system conditional on EME banks in distress (ΔCoES1%
system|I) (2006 – 2013; percentage of daily standard deviation)
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
-3.5
-3.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
2006 2007 2008 2009 2010 2011 2012 2013
Sources: Bloomberg and ECB calculations. Notes: The charts depict the distribution and the median CoVaR/CoES estimates based on eight non-overlapping annual samples of daily observations from 2006 to 2013. The black line represents the median of the 26 EME banks’ daily ΔCoVaRq system|i/ ΔCoESq system|I in per cent of the daily standard deviation of the global banking sector’s return distribution. A negative (positive) value represents a conditional loss (gain). The blue box represents the 25% to 75% quantile of banks. The blue vertical lines represent the minimum and the maximum estimates.
25 ECB
Financial Stability Review November 2014 25
I� Macro-F I�nancI�al and credI�t
envI�ronMent
25
Given the progress made with correcting fiscal imbalances, the focus has increasingly moved towards a more growth-friendly composition of consolidation. In this respect, several governments have recently announced or approved income tax cuts (e.g. Spain and the Netherlands), while planning to stay within their nominal fiscal targets. At the same time, other countries will most likely miss their 2014 fiscal targets mainly on account of weaker than expected macroeconomic developments.
Despite the progress made to date in reducing fiscal and macroeconomic imbalances, sovereign risks remain elevated. First, room for fiscal manoeuvre tends to be limited to a small number of euro area countries, as government debt levels continue to be high and still rising in many countries. This limits considerably the scope for fiscal stimulus through cuts in taxes without corresponding compensatory measures on the spending side. Any delay in debt stabilisation can affect countries’ creditworthiness, as recently stressed by major rating agencies. Moreover, despite the progress achieved in the past years, many euro area countries are still far away from their medium-term objective of a close-to-balanced structural budget (see Chart 1.12). For the euro area as a whole, the improvement in the structural balance is expected to fall considerably short of the Stability and Growth Pact’s requirements, with Germany being the only euro area country that is expected to over- achieve the requirements under the Pact in 2014 and 2015 (see Chart 1.12). Second, sizeable reform
… despite continued vulnerabilities
Chart 1.11 Composite indicator of systemic stress in euro area sovereign bond markets (SovCISS) (Jan. 2005 – Oct. 2014)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
minimum-maximum range vulnerable euro area countries euro area average other euro area countries
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Sources: ECB and ECB calculations. Notes: Aggregation of country indicators capturing several stress features in the corresponding government bond markets (changing default risk expectations, risk aversion, liquidity risk and uncertainty) for vulnerable (Greece, Ireland, Italy, Portugal and Spain) and other (Austria, Belgium, Germany, Finland, France and the Netherlands) countries. The range reflects the maximum and minimum across the entire set of above-mentioned countries. For further details on the CISS methodology, see Hollo, D., Kremer, M. and Lo Duca, M., “CISS – a composite indicator of systemic stress in the financial system”, Working Paper Series, No 1426, ECB, March 2012.
Chart 1.12 Structural balances and medium-term fiscal objectives across the euro area (2014, 2015; percentage of GDP)
-4 IE FR MT BE SI ES SK LV PT AT FI EA IT CY EE NL DE LU
-3
-2
-1
0
1
2
2014 structural balance 2015 structural balance medium-term objective
Sources: European Commission autumn 2014 economic forecasts and 2014 national stability programmes. Notes: Greece does not have an updated medium-term fiscal objective and is not shown in the chart (its structural surplus is estimated in the European Commission’s forecast at 2.0% in 2014 and 1.6% in 2015).
26 ECB Financial Stability Review November 20142626
commitments remain to be implemented, as highlighted in the past European Semester given only minor advances over the last six months. In line with the country-specific recommendations adopted by the ECOFIN Council in July 2014, several governments have cut the high tax wedge on income to promote employment and long-term growth, but deeper structural reforms, particularly those in the labour and product markets or pension systems, would bring long-term benefits without endangering fiscal solvency. At the same time, even in countries with limited fiscal space, fiscal policy can still support economic recovery by altering the composition of the budget – in particular by simultaneously cutting distortionary taxes and unproductive expenditure. Third, while alleviating fiscal costs, the currently low sovereign yields on the outstanding debt in many euro area countries (see Chart 1.13) may expose some countries to sudden flow reversals, especially if macroeconomic developments or reform efforts turn out to be less favourable than currently envisaged.
Against this background, under current government plans, the aggregate euro area fiscal deficit would continue to fall and stay below the 3% Maastricht threshold. According to the Commission’s autumn 2014 forecast, the budget deficit for the euro area (18-country aggregate) will fall from 2.9% of GDP in 2013 (following a positive revision of 0.1 percentage point implied by the transition to the European System of Accounts 2010) to 2.6% in 2014 and 2.4% in 2015. After the incorporation of fiscal measures underlying governments’ 2015 draft budgetary plans, structural balances are projected to deteriorate in half of the euro area countries and to remain flat for the euro area aggregate. However, cyclical developments and temporary factors are seen to be supportive to fiscal positions in 2015 so that headline balances follow a more favourable path in most euro area countries. Compared with the Commission’s spring 2014 forecasts, the short-term fiscal outlook deteriorated marginally for the euro area aggregate, triggered by larger deteriorations in France, Italy, Portugal and Finland.
The unwinding of financial sector support is expected to contribute to the improvement of fiscal balances in 2014 and beyond in many countries. In Greece and Slovenia, the bank recapitalisation costs of 2013 were a one-off. In Portugal, the cash reserves earmarked for potential support to the financial sector were used in mid-2014 as a loan to the Portuguese Resolution Fund for use in the isolated bail-in case of Banco Espirito Santo. Going forward, bail-in and bank resolution arrangements based on the provisions of the Bank Recovery and Resolution Directive and the Single Resolution Mechanism, as well as, more generally, steps taken at the European level towards a banking union, might imply a new paradigm relative to the last years, notably with regard to the sovereign-bank nexus. The explicit and transparent framework for sharing resolution costs with
Fiscal deficit is forecast to drop
further in 2014 and 2015…
… as support to the financial sector
weighs less on public finances
Chart 1.13 Average nominal yields on debt securities issued by euro area governments
(Sep. 2014; percentage per annum)
2.0
2.5
3.0
3.5
4.0
4.5
5.0
2.0
2.5
3.0
3.5
4.0
4.5
5.0
BE
DE
IE
GR
ES
FR
IT
CY MT
NL
AT
PT
SI
SK
FI
LV
EA
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
x-axis: issued between October 2013 and September 2014 y-axis: outstanding at end-September 2014
Source: ECB. Notes: The connecting dotted lines map the evolution over the past year. Yields are averages, weighted with the amount of outstanding, and respectively, newly issued securities. Issuances over the past year reflect, inter alia, an extension of maturities in several countries with longer-dated securities bearing higher yields.
27 ECB
Financial Stability Review November 2014 27
I� Macro-F I�nancI�al and credI�t
envI�ronMent
27
bank creditors along with a Single Resolution Fund clearly has the potential to reduce prospective contingent liabilities of any given country vis-à-vis its banking sector.
Despite progress in fiscal adjustment, the public debt-to-GDP ratio for the euro area (18-country aggregrate) is still rising, but is projected in the Commission’s autumn 2014 forecast to peak in 2015 at 95% of GDP. This is mainly attributable to adverse interest rate-growth differentials and deficit-debt adjustments, which are expected to exceed the primary surplus projected as of 2014. As these two inhibiting factors are expected to wane, the public debt ratio for the euro area as a whole is projected to decline as of 2016 for the first time since 2008. At the country level, public debt ratios remain on an increasing path in the majority of euro area countries (see Chart 1.14).
Regarding debt sustainability, the most important risks across the euro area relate to the potential complacency in terms of fiscal adjustment and structural reforms, a slowdown in economic growth dynamics and a prolonged period of low inflation.2 Such developments would impede the debt-servicing abilities of sovereigns, in particular of those which currently face heightened market optimism and downward rigidities in fiscal positions. Simulation results suggest that a combined lasting shock of lower growth, higher yields and worsened structural balances, which could emerge from a lack or reversal of structural reforms and fiscal consolidation efforts, would put debt sustainability at risk (see Chart 1.15). In general, the higher the debt levels and the deeper the economic and institutional rigidities, the less resilient countries are to adverse shocks.
The euro area sovereign debt crisis has illustrated that alongside perceived credit risks liquidity strains in the public sector may also pose a risk to financial stability. In fact, sovereign 2 For more details on the financial stability challenges posed by very low rates of consumer price inflation, see Box 1 in Financial Stability
Review, ECB, May 2014.
Public debt is expected to peak in 2015 and decline gradually thereafter…
… but uncertainties relating to sovereign debt sustainability persist
Financing needs remain sizeable in several countries in 2015…
Chart 1.14 Changes in public debt levels across the euro area between 2013 and 2015 (2013 – 2015; percentage points of GDP)
-15.0
-12.5
-10.0
-7.5
-5.0
-2.5
0.0
2.5
5.0
7.5
10.0
12.5
15.0
CY SI ES IT FR FI AT BE EA NLMTLU SK EE LV PT DEGR IE
interest rate-growth differential primary deficit deficit-debt adjustment change in debt 2013-2015
Source: European Commission autumn 2014 economic forecast.
Chart 1.15 Reaction of the public debt ratio to standardised macro and fiscal shocks
(percentage points of GDP)
0 5
10 15 20 25 30 35 40 45 50 55
0 5 10 15 20 25 30 35 40 45 50 55
euro area average maximum country value minimum country value
real growth
potential growth
marginal interest
rate
structural primary balance
GDP deflator
combined shock
Source: ECB. Notes: The chart shows the reaction of the debt ratio for the euro area average and the individual countries’ range as of 2024 to standardised (1 percentage point) adverse shocks to real growth (for three years), potential growth, the marginal interest rate, the fiscal position (structural primary balance), the GDP deflator, and a combined shock of the above. The shocks are permanent (the three-year real growth shock translates into partial potential shock deterioration) and are applied as of 2015. The deterministic debt simulations are conducted in a partial equilibrium framework, which takes into account feedback effects between fiscal, macro and financial variables.
28 ECB Financial Stability Review November 20142828
financing needs for 2015 remain significant in many euro area countries (see Chart 1.16), according to securities redemption data up to September 2014. Maturing sovereign debt in the near-to-medium term remains high in the euro area too, albeit with major cross- country differences. As at end-September 2014, securities with a residual maturity of up to one year accounted for about 20% of total outstanding debt securities in the euro area or 15.5% of GDP. The average residual maturity of outstanding euro area government securities was 6.3 years, with the residual maturities ranging from 3.2 years in Cyprus to 12.0 years in Ireland.
Sovereign financing needs may – to some extent – be alleviated by resorting to existing financial assets. The consolidated financial assets held by euro area general governments averaged some 36.7% of GDP at the end of the first quarter of 2014, with some variation across countries. At the same time, the market value of consolidated general government liabilities in the euro area was 104.3% of GDP, yielding net financial liabilities of 67.6% of GDP.
1.3 gRAduALLY IMpROVINg FINANCINg CONdITIONS IN ThE NON-FINANCIAL pRIVATE SECTOR, BuT VuLNERABILITIES REMAIN
While recovering somewhat amid moderately improving macroeconomic conditions, income and earnings for the euro area non-financial private sector have remained sluggish. The income situation of households appears to have stabilised further, but disposable income dynamics have remained muted and households’ financial situation expectations have become somewhat less optimistic as the economic recovery has shown signs of losing some of its momentum. While there are tentative signs of improvements in labour market conditions at the aggregate euro area level (see Chart 1.17), the situation continued to be particularly weak in vulnerable euro area countries, thereby further weighing on households’ income prospects. As signalled by a distance-to-distress indicator capturing household balance sheet risks, overall credit risks from household balance sheets in the euro area have increased somewhat in the last quarters, but are still much less pronounced than during the stressed conditions of the euro area sovereign debt crisis (see Chart 1.18).
Similar to households, the earnings-generating capacity of euro area non-financial corporations has improved somewhat driven by the gradual economic recovery to date, yet corporate profitability has remained muted. Gross operating income has picked up slightly, amid lower negative earnings growth per share and expected default frequencies for listed firms close to pre-crisis lows. Being a function of overall macroeconomic developments, corporate earnings in the euro area are expected to rise as the
… but available financial assets may
mitigate financing needs
Gradual economic recovery alleviates
income and earnings risks somewhat
Chart 1.16 Maturing government debt securities and projected deficit financing needs of euro area governments in 2015 (percentage of GDP)
0
3
6
9
12
15
18
21
24
27
FR IT ES BE EA PT CY NL GR DE MT SI FI SK AT IE LV EE LU
maturing government securities general government deficit
Sources: European Commission autumn 2014 economic forecast, ECB and ECB calculations. Notes: Gross financing needs are estimates of government debt securities maturing in 2015, based on ECB data as at end-September 2014, and the Commission’s government deficit projections for 2015. The estimates are subject to the following caveats. First, they only account for redemptions of debt securities, while maturing loans are not included. Second, estimates disregard that some maturing government securities are held within the government sector. Finally, refinancing needs corresponding to short-term debt issued after September 2014 are assumed to be the same as in the fourth quarter of 2014, which may imply an overestimation for some countries.
29 ECB
Financial Stability Review November 2014 29
I� Macro-F I�nancI�al and credI�t
envI�ronMent
29
economic recovery gathers pace, though there is a risk that firms’ capacity to retain earnings may remain weak until this materialises.
Despite the projected gradual improvement in income and earnings prospects, legacy balance sheet issues continue to weigh on the aggregate euro area non-financial private sector. On average, euro area households’ indebtedness amounted to some 64% of GDP, while for non-financial corporations the number is more elevated, at 104% of GDP (or some 90% of GDP on a consolidated basis). However, a gradual balance sheet adjustment is underway, even if the adjustment to date may seem rather modest at the aggregate euro area level (see Chart 1.19). Indeed, a much more nuanced picture emerges at the level of individual countries or sectors of economic activity. When tracking private sector (in particular corporate) deleveraging at the country level, the pace of adjustment differed considerably across the euro area, with
Private sector indebtedness remains elevated amid continued heterogeneity at the country and sector levels
Chart 1.17 Expectations about households’ financial situation and changes in the number of unemployed in the euro area (Jan. 2005 – Oct. 2014; number in thousands, seasonally adjusted; percentages; percentage balances; three-month moving averages)
0
2
4
6
8
10
12
14
16
-200
-100
0
100
200
300
400
500
600
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
monthly change in the number of unemployed (left-hand scale) unemployment rate (right-hand scale) expectations about households’ financial situation over the next 12 months (right-hand scale)
Sources: European Commission Consumer Survey and Eurostat. Note: Expectations about households’ financial situation are presented using an inverted scale, i.e. an increase (decrease) of this indicator corresponds to less (more) optimistic expectations.
Chart 1.18 households’ distance to distress in the euro area
(Q1 2005 – Q2 2014; number of standard deviations from mean)
0
5
10
15
20
25
30
0
5
10
15
20
25
30
2005 2006 2007 2008 2009 2010 2011 2012 2013
Sources: ECB, Bloomberg, Thomson Reuters Datastream and ECB calculations. Notes: A lower reading for distance to distress indicates higher credit risk. The chart shows the median, minimum, maximum and interquartile distribution across 11 euro area countries for which historical time series cover more than one business cycle. For details of the indicator, see Box 7 in Financial Stability Review, ECB, December 2009.
Chart 1.19 Indebtedness of the non-financial corporate sector in the euro area
(Q1 2006 – Q1 2014; percentage of GDP; unconsolidated)
70
80
90
100
110
120
130
140
150
160
170
180
70
80
90
100
110
120
130
140
150
160
170
180
2006 2007 2008 2009 2010 2011 2012 2013
euro area average percentile range
Sources: ECB and ECB calculations. Notes: Based on ESA 95 standards. Debt includes loans, debt securities and pension fund reserves. The chart shows the average non-financial corporate indebtedness in the euro area and the interquartile distribution (25th and 75th percentile) across individual euro area countries.
30 ECB Financial Stability Review November 20143030
deleveraging being more pronounced in countries which had accumulated large amounts of debt in the run-up to the crisis. The same pattern emerges at the sector level, whereby overindebted sectors, such as the construction and real estate services sector, continue to deleverage more strongly than less indebted ones such as industry or wholesale and retail trade.
In the current environment of low interest rates and a low cost of market-based funding, households’ and non-financial firms’ interest payment burden has remained at record lows (see Chart 1.20). Borrowers in countries with ongoing relative price adjustments, however, have seen some rise in their real debt burden amid recent low inflation outturns. In terms of risks, the ongoing process of balance sheet repair should help offset the challenges related to an eventual normalisation of interest rates and the ensuing rise in the debt servicing burden. Such challenges might be greatest for those countries where loans with floating rates or rates with rather short fixation periods predominate. That said, a higher debt service burden for borrowers in a rising interest rate environment is likely to be partly offset by the positive impact of an economic recovery on households’ and firms’ income and earnings situation.
Bank lending flows to the non-financial private sector have remained muted, partly reflecting the ongoing balance sheet repair in both the financial and non-financial sectors. On average, bank lending to euro area households has remained subdued, mirroring sluggish dynamics of household income, high levels of unemployment and housing market weakness in some countries. However, rather heterogeneous developments at the country level form the basis of this relatively weak aggregate picture (see Chart 1.21). Looking at the components of bank lending by purpose, modest annual growth in loans for house purchase has been offset by a continued drop in consumer loans and other types of lending. Nonetheless, in line with the gradual economic recovery, the October 2014
Favourable interest rate environment
facilitates debt servicing
Lending to the non-financial private sector
remains muted
Chart 1.20 Interest payment burden of the euro area non-financial private sector (Q1 2006 – Q1 2014; four-quarter moving sums; percentages)
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
2006 2007 2008 2009 2010 2011 2012 2013
net interest payments-to-gross operating surplus ratio of non-financial corporations households’ interest payment burden as a percentage of gross disposable income
Sources: ECB and Eurostat. Note: Based on ESA 95 standards.
Chart 1.21 MFI lending to euro area households
(Jan. 2006 – Sep. 2014; percentage change per annum)
-10
-5
0
5
10
15
20
25
30
35
-10
-5
0
5
10
15
20
25
30
35
2006 2007 2008 2009 2010 2011 2012 2013 2014
interquartile range minimum-maximum range euro area average
Source: ECB. Note: Data have been adjusted for securitisation.
31 ECB
Financial Stability Review November 2014 31
I� Macro-F I�nancI�al and credI�t
envI�ronMent
31
euro area bank lending survey suggests further improvements in households’ financing conditions, as reflected by the continued easing of credit standards on loans to households and the further net increase in demand for such loans.
Cross-country disparities in supply conditions fell overall for loans to households, thus pointing to a decrease in financial market fragmentation. Supply-side constraints appear to be easing particularly for consumer loans and other lending to households, and to a lesser extent also for housing loans. Improving supply-side conditions reflect lower pressures from cost of funds and balance sheet constraints, but competition has also contributed to the net easing of credit standards, mainly for loans to households for house purchase. By contrast, a re-emergence of risk concerns had a slightly restrictive impact on credit standards for both housing and consumer loans. At the same time, improving housing market prospects and consumer confidence have translated into a continued net increase in demand for housing loans and consumer credit.
The net external financing of euro area non-financial corporations continued to fall, albeit at a slower pace than in recent quarters (see Chart 1.22). Corporate disintermediation continued, but the issuance of market-based debt still fell short of offsetting the decline in new MFI loans to non-financial corporations. However, funding substitution has remained limited to larger corporations and predominantly those which are domiciled in countries with more developed corporate bond markets (e.g. Germany and France), while small and medium-sized enterprises (SMEs) and large firms located in more vulnerable countries remained more dependent on bank funding. That said, the results of the latest euro area bank lending survey suggest that underwriting terms for corporate loans have continued to improve, as reflected by easing credit standards, in particular for large firms. Similarly to household loans, supply- side conditions for corporate loans point to decreasing fragmentation across countries. Demand for corporate loans in the euro area continued to rise, although cross-country heterogeneity has remained considerable. Increased demand largely reflects higher financing needs, mainly for mergers and acquisitions and debt restructuring, while financing needs related to fixed investment dampened demand for loans to euro area enterprises. Firms’ internal financing capacity and the issuance of debt securities by non- financial corporations contributed negatively to loan demand. Alongside improving supply and demand-side conditions, targeted Eurosystem measures to revive lending, i.e. the targeted longer-term refinancing operations or the asset- backed securities and covered bond purchase programmes, should promote the recovery of credit going forward, while at the same time contributing to a further decrease in funding costs for non-financial firms in the euro area.
A drop in bank lending to non-financial corporations is partly offset by the issuance of market-based debt…
Chart 1.22 External financing of euro area non-financial corporations
(Q1 2006 – Q3 2014; EUR billions; net annual flows)
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
11.0
11.5
-200
-100
0
100
200
300
400
500
600
700
800
2006 2007 2008 2009 2010 2011 2012 2013 2014
bonds loans net share of bonds in total corporate debt (right-hand scale)
Sources: ECB and ECB calculations.
32 ECB Financial Stability Review November 20143232
Corporate liquidity has remained at record highs in several euro area countries, suggesting that non-financial firms can also rely on internal funds as a financing source in addition to loans and debt securities. Firms’ liquidity holdings reached almost 30% of GDP in early 2014, but amid a large degree of cross-country variation across the euro area (see Chart 1.23). These high liquidity buffers may reflect a lack of investment opportunities, precautionary motives (i.e. mitigating the risk of limited access to external financing in the future) in the context of a low opportunity cost of holding liquid assets and continued credit supply constraints in some countries.
Nominal funding costs of the euro area non- financial private sector have continued to decline across most business lines, maturities and funding sources. Nominal financing costs for euro area households reached their lowest levels since the start of the reporting of harmonised euro area bank lending rates in 2003 for all categories of lending except consumer credit, while real funding costs have remained broadly unchanged since early 2014 (see Chart 1.24). Likewise,
… amid high corporate liquidity
Funding costs have touched record lows
on average…
Chart 1.23 Liquidity position of non-financial corporations in selected euro area countries
(Q1 2006 – Q1 2014; percentage of GDP)
15
20
25
30
35
40
45
50
15
20
25
30
35
40
45
50
2006 2007 2008 2009 2010 2011 2012 2013
euro area Italy Spain Germany France Netherlands
Sources: ECB and ECB calculations. Notes: Based on ESA 95 standards. Liquidity is defined as the sum of currency and deposits, short-term securities and mutual fund shares.
Chart 1.24 Euro area bank lending rates on new loans to households in nominal and real terms (Jan. 2006 – Sep. 2014; percentages)
consumer lending lending for house purchase other lending
a) nominal b) real
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
2006 2008 2010 2012 2014 0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
2006 2008 2010 2012 2014
Source: ECB. Note: Real bank lending rates are calculated by deflating nominal lending rates with the Harmonised Index of Consumer Prices.
33 ECB
Financial Stability Review November 2014 33
I� Macro-F I�nancI�al and credI�t
envI�ronMent
33
non-financial corporations’ overall financing costs have continued to fall across most external financing sources (see Chart 1.25), supported by a low interest rate environment and favourable financial market conditions. Bank lending rates have declined further across the maturity spectrum, though the latest easing in monetary policy rates remains yet to be fully passed through (see Chart 1.26). At the same time, the cost of equity has increased since early 2014 amid ebullient equity markets and rising equity risk premia in many countries – a development which contrasts with a continued fall in the cost of market-based debt.
Fragmentation in both nominal and real lending conditions persists, despite having decreased since the height of the euro area sovereign debt crisis. The cross-country heterogeneity in the euro area, as measured by the range between the lowest and highest interest rate charged on loans to households, has remained at elevated levels, reflecting different country-specific risk constellations and persisting fragmentation afflicting some euro area countries. The same holds true for firms, where lending rates continue to vary widely across the euro area. At the same time, developments in firms’ financial conditions continue to vary also in terms of firm size. The strong difference between the loan pricing conditions for small and large firms, which primarily results from the divergence in firm-specific risks, highlights the still less favourable conditions faced by small firms, particularly in more vulnerable countries. In addition, according to the ECB’s latest survey on access to finance of enterprises in the euro area, banks’ willingness to grant a loan continues to be higher for large firms (see Chart 1.27). This is also corroborated by the fact that the success of large firms when applying for a bank loan was higher than for SMEs, indicating overall better access to finance of large firms compared with SMEs. Finally, collateral requirements also appear to be less strict for large firms than for SMEs.
… but fragmentation in lending conditions persists across countries and firm sizes
Chart 1.25 Nominal cost of external financing of euro area non-financial corporations (Jan. 2006 – Oct. 2014; percentages)
0
2
4
6
8
10
12
0
2
4
6
8
10
12
2006 2007 2008 2009 2010 2011 2012 2013 2014
cost of quoted equity
cost of market-based debt
overall cost of financing
short-term MFI lending rates
long-term MFI lending rates
Sources: ECB, Merrill Lynch, Thomson Reuters Datastream and ECB calculations. Note: The overall cost of financing for non-financial corporations is calculated as a weighted average of the cost of bank lending, the cost of market-based debt and the cost of equity, based on their respective amounts outstanding derived from the euro area accounts.
Chart 1.26 The ECB policy rate and the composite cost-of-borrowing indicator for non-financial corporations (Sep. 2011 – Sep. 2014; cumulative percentage point changes)
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
-2.5
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
minimum bid rate in main refinancing operations minimum and maximum change
2014 Sep. Mar. Sep. Mar. Sep. Sep.Mar.
2011 2012 2013
Sources: ECB and ECB calculations. Notes: For methodological details on the construction of the cost-of-borrowing indicator, see “Assessing the retail bank interest rate pass-through in the euro area at times of financial fragmentation”, Monthly Bulletin, ECB, August 2013.
34 ECB Financial Stability Review November 20143434
Mirroring overall macroeconomic trends, the overall development of euro area property markets remained subdued in the first half of 2014, but with signs of a recovery in some countries. Residential property prices have stabilised on an annual basis at the aggregate euro area level, following a sharp turnaround in some euro area countries that experienced significant price corrections in recent years. Similarly, euro area commercial property markets have shown further signs of stabilisation, but the underlying price dynamics in the prime and non-prime segments continued to diverge strongly (see Chart 1.28).
Prime commercial property (i.e. modern retail and office buildings in metropolitan areas) continued on its ebullient course in the context of the current low yield environment and the related ongoing search for yield. Accordingly, investment activity in commercial property markets has remained buoyant in recent quarters, with underlying transaction volumes reaching multi-year highs (see Chart 1.29). Activity has been increasingly driven by domestic investors, but foreign – in particular non-European – investors have remained active as well. Increased investor interest went hand in hand with a broad-based decline in yields on prime commercial property. Perhaps most noteworthy, the significant pick-up in demand
Euro area property markets show signs
of an incipient recovery…
… amid a continued ebullience in prime
commercial property markets
Chart 1.27 Financing conditions of euro area SMEs in comparison with large firms
(H1 2009 – H1 2014; net percentages of respondents; changes over the past six months)
-60
-50
-40
-30
-20
-10
0
10
20
30
40
50
-60
-50
-40
-30
-20
-10
0
10
20
30
40
50
2009 2010 2011 2012 2013 2014 2009 2010 2011 2012 2013 2014 2009 2010 2011 2012 2013 2014 Willingness of banks to provide credit Level of interest rates Collateral requirements
large firms small and medium-sized enterprises
Improving financing conditions
Deteriorating financing conditions
Source: ECB calculations based on the survey on access to finance of enterprises (SAFE). Note: The level of interest rates and collateral requirements are presented using an inverted scale.
Chart 1.28 Euro area commercial and residential property values and the economic cycle (Q1 2004 – Q3 2014; percentage change per annum)
-15
-10
-5
0
5
10
15
20
25
-15
-10
-5
0
5
10
15
20
25
2004 2006 2008 2010 2012 2014
GDP growth
prime commercial property prices commercial property prices
residential property prices
Sources: Eurostat, ECB, experimental ECB estimates based on IPD and national data, and Jones Lang LaSalle.
35 ECB
Financial Stability Review November 2014 35
I� Macro-F I�nancI�al and credI�t
envI�ronMent
35
for commercial property in countries that had previously witnessed pronounced price declines, such as Ireland and Spain, has also contributed to narrowing yield dispersion across the euro area.
In terms of property price dynamics, fragmentation at the country level has been declining, particularly in the prime commercial segment where most recently almost all euro area countries have seen an increase in prices. By contrast, residential property prices continued to drop – to varying degrees – in countries such as Cyprus, Greece, Italy and Slovenia. This illustrates the high degree of cyclicality of commercial property prices which tend to be more volatile and track the economic cycle with greater amplitude than residential property prices. That said, after a major multi-year adjustment, country-level data suggest a sharp rebound in residential and commercial property markets in some countries, notably Ireland. At the same time, country-level developments often mask underlying regional disparities, with strong house price growth in metropolitan areas and comparably subdued price movements in remaining regions (e.g. Austria, Germany and Ireland), highlighting the risk that strong house price growth could potentially ripple out to surrounding areas. So far there are no signs of the ongoing recovery or the regional buoyancy of euro area residential property markets translating into buoyant housing loan growth (see Chart 1.30), suggesting some transitory phenomena such as pent-up demand from cash buyers and the presence of foreign buyers in certain (mainly high-priced) market segments, especially in some large cities.
In terms of valuations, for the euro area as a whole, residential property prices are broadly in line with fundamentals, but valuation estimates for prime commercial property are still somewhat above their long- term average. However, property markets are inherently local, so that such aggregates belie heterogeneous developments at both the country and regional level. Residential and prime commercial property valuations
Fragmentation at the country and regional levels persists, although diminishing
Overvaluation is a concern in some countries…
Chart 1.29 Commercial property price changes and investment volumes in the euro area (Q1 2009 – Q3 2014; average of price changes in Austria, France, Germany, Ireland, the Netherlands and Spain)
0
2
4
6
8
10
12
14
16
18
-25
-20
-15
-10
-5
0
5
10
15
20
2009 2010 2011 2012 2013 2014
transaction volumes – overall market (EUR billions; right-hand scale) prices – prime property (percentage change per annum; left-hand scale) prices – overall market (percentage change per annum; left-hand scale)
Sources: DTZ Research, ECB, experimental ECB estimates based on IPD and national data, and Jones Lang LaSalle. Note: Four-quarter moving average of investment volumes.
Chart 1.30 Residential property price and housing loan growth across the euro area
(H1 2014; percentage change per annum)
-10
-8
-6
-4
-2
0
2
4
6
8
10
12
14
-10
-8
-6
-4
-2
0
2
4
6
8
10
12
14
Germany
Italy
France
Spain
NetherlandsPortugal
Belgium
Estonia
SloveniaCyprus
Malta
Luxembourg
Ireland
Finland
Slovakia
Greece
Austria
euro area
Latvia
-7.5 -5.0 -2.5 0.0 2.5 5.0 7.5 10.0 12.5 15.0
x-axis: housing loan growth y-axis: residential property price growth
Sources: ECB and ECB calculations. Notes: Bank lending data are not adjusted for securitisation. Securitisation may play an important role in some countries, for example Belgium, where the time series adjusted for securitisation would result in annual housing loan growth of 3% for the first half of 2014.
36 ECB Financial Stability Review November 20143636
have come down considerably from previous peaks in several countries (e.g. Ireland and Spain) as the unwinding of pre-crisis excesses brought prices down to or below the level suggested by underlying values. By contrast, estimated overvaluation has remained high in both market segments in Belgium, Finland and France (see Chart 1.31). Similar disparities may emerge at the regional level, as reflected by the estimated significant overvaluation of residential property in some large cities in Germany and Austria. It is worth emphasising though that valuation estimates are surrounded by a high degree of uncertainty as they do not capture country-level specificities, such as fiscal treatment or various structural property market characteristics.
A key downside risk to euro area property markets relates to a weak or stalling economic recovery, given the high cyclicality of many property market segments. Indeed, a negative economic shock could create at least three challenges: first, to those commercial property investors who are already confronted with difficulties (e.g. those in negative equity positions due to prices being below previous years’ peaks); second, as a trigger for house price corrections in countries with signs of overvaluation (or it could reverse the ongoing recovery in others); and third, for debt servicing in countries with a highly indebted household sector. From a financial perspective, a potential increase in global risk aversion and the related rise in long-term interest rates could affect the debt servicing capacity of both households and commercial property investors via the more limited availability and higher cost of funding, thereby contributing to rising rollover risks and aggravating the interest payment burden. The numerous property-related instruments in the newly acquired macro-prudential toolkit may help alleviate any future cyclical challenges, while also contributing to increasing the resilience of banks and their borrowers.
… while risks remain tilted to the
downside
Chart 1.31 Estimated over/undervaluation of residential and prime commercial property prices in selected euro area countries (Q2 2014; percentages)
-30
-20
-10
0
10
20
30
40
50
60
-30
-20
-10
0
10
20
30
40
50
60
Austria
Belgium Germany
euro area
Spain
Finland
France
Italy
Netherlands
Portugal
Ireland
-15 -10 -5 0 5 10 15 20 25
y-axis: commercial property under/overvaulation x-axis: residential property under/overvaluation
Sources: Jones Lang LaSalle, European Commission, ECB and ECB calculations. Notes: The size of the bubble reflects the projected change in real GDP growth in 2015. Estimates for residential property prices refer to Q1 2014 for Finland, Germany, Ireland, the Netherlands and Portugal and are based on four different valuation methods: price-to-rent ratio, price-to-income ratio and two model-based methods. For details of the methodology, see Box 3 in ECB, Financial Stability Review, June 2011. For further details on valuation estimates for prime commercial property, see Box 6 in ECB, Financial Stability Review, December 2011.
37 ECB
Financial Stability Review November 2014
2 FINANCIAL MARkETS Supported by historically low risk-free rates and subdued market volatility, a search for yield continues in global financial markets. Despite bouts of volatility – linked to rising geopolitical tensions and weak economic data for the euro area – the price of risk remains low across global market segments and duration exposures have increased.
Within the euro area money market segment, low and even negative rates have encouraged an increase in interbank activity and a move into slightly longer maturities. In bond markets, yields have generally fallen further despite bouts of volatility and some outflows of foreign investment from lower-rated euro area markets. Likewise, credit spreads remain at relatively low levels, though risk premia in credit markets have not been immune to strong outflows from the high- yield segment amid rising global risk aversion and concerns about overheating. Indeed, investors concentrated yield-seeking behaviour on the investment-grade segment of the bond market, which experienced a further increase in duration. Equity market rallies have only seen brief interruptions and valuations remain elevated, particularly for US markets.
As a broad-based search for yield continues, vulnerabilities are building up in global capital markets. While estimates of prospective asset overvaluations in any individual market segment differ, it is clear that asset price movements are becoming increasingly correlated across segments. In addition, current high valuations are being sustained by historically low levels of risk-free rates and subdued levels of market volatility, which could be tested by a withdrawal of accommodative global monetary policy. At the same time, investor appetite for riskier euro area assets depends on a fragile economic recovery with significant downside risks.
A combination of three amplifying factors could disrupt financial stability should the search for yield exhibit a sustained reversal. First, bouts of market volatility have shown that secondary market liquidity in fixed income markets is low. Second, while banking sector leverage continues to decline, use of leverage in securities markets is increasing. Moreover, similar to leverage risk, redemption risk for investment funds embeds the possibility of forced selling leading to prospective fire-sale spirals. Finally, duration risk exposure is elevated, which would also magnify future price corrections.
2.1 INTERBANk ACTIVITY IN EuRO AREA MONEY MARkETS CONTINuES TO NORMALISE, BuT FRAgMENTATION REMAINS
Conditions in euro area money markets continue to improve, though fragmentation remains a concern. Recent developments include a further decline in market-based measures of stress, a broad- based increase in interbank activity and improved access for banks from vulnerable countries to the secured segment (see Charts 2.1 and 2.2). The decisions of the ECB’s Governing Council to lower the deposit facility rate to a negative level in June and cut it further in September have clearly had an impact on money market rates and have contributed to an increase in interbank turnover. However, the rate cuts have had a limited impact on fragmentation. Increased activity has been concentrated largely on transactions involving highly rated counterparties and/or collateral. However, positive rating actions on sovereigns have eased fragmentation by improving access to secured markets for banks from vulnerable countries. In addition, the preliminary results of the latest Euro Money Market Survey indicate that credit policies are no longer exerting a strong contractionary impact on bank lending and banks expect an expansionary impact going forward. This survey also reports an improvement, from low levels, in market functioning across all segments, both in terms of liquidity and efficiency. However, increased activity in certain segments, for example the overnight index swap market, may not reflect improved market functioning but rather an increased need to hedge against falling interest rates.
Conditions in euro area money markets continue to normalise
38 ECB Financial Stability Review November 20143838
Following a seven-year decline, interbank activity in the unsecured money market segment is showing signs of a tentative recovery, although access remains challenging for lower-rated banks. The latest Euro Money Market Survey signals a slight increase in unsecured activity in the second quarter of 2014 which was, according to EONIA volumes, sustained in the months following ECB rate decisions. Unsecured money market interest rates have declined and become negative for a maturity of up two weeks, but increased activity in the segment remains concentrated among higher-rated entities.1 Meanwhile, market access for banks from vulnerable euro area countries remains limited to small amounts at overnight maturities.
The repayment of three-year longer-term refinancing operations (LTROs), positive rating actions and the increased use of repos
1 According to the October 2014 Euro Money Market Survey, five institutions account for almost 90% of activity in the unsecured segment.
Unsecured segment shows signs of a
tentative recovery…
… but activity remains
concentrated in the secured segment
Chart 2.1 Turnover in selected euro area money market segments
(Q2 2003 – Q2 2014; EUR trillions)
0
5
10
15
20
25
30
35
2003 2005 2007 2009 2011 2013
unsecured
0
5
10
15
20
25
30
35
secured overnight interest rate swaps
Source: ECB Euro Money Market Survey.
Chart 2.2 Spreads between unsecured interbank lending and overnight index swap rates
(Jan. 2007 – Nov. 2014; basis points; three-month maturities)
euro area
UK
US
0
50
100
150
200
250
300
350
400
2007 2008 2009 2010 2011 2012 2013
GBPUSDEUR
Finalisation of the May 2014 FSR (16 May)
0
10
20
30
0
10
20
30
16 May 16 July 16 Sep. 0
50
100
150
200
250
300
350
400
2014
Sources: Bloomberg and ECB calculations. Notes: Red indicates rising, yellow moderating and green falling pressure in the respective money markets. For more details, see Box 4 entitled “Assessing stress in interbank money markets and the role of unconventional monetary policy measures” in Financial Stability Review, ECB, June 2012.
39 ECB
Financial Stability Review November 2014 39
2� F InancIal Markets
39
by banks’ treasuries for liquidity management purposes have contributed to increased activity and less fragmentation in the secured money market segment. The rating upgrade/stabilisation of vulnerable euro area sovereigns has resulted in improved access for their banks to repo markets. However, fragmentation and local bias among banks as regards counterparties and collateral persist. Similar to the unsecured segment, increased activity following the introduction of negative policy rates appears concentrated on high credit quality. Repo rates in non-vulnerable countries have fallen and remained at negative levels, while those in vulnerable countries have oscillated around zero. A strong preference for credit quality is evident in repo trading volumes, where transactions backed by high-quality collateral have experienced a steady increase since June 2014 and a sharp increase following the second rate cut in September, while transactions backed by lower-rated collateral are currently close to May 2014 levels.2 Banks are also continuing to move away from bilateral trading towards the use of central clearing counterparties (CCPs). The latest Euro Money Market Survey shows that the share of transactions conducted via CCPs remained stable at around 73% of bilateral turnover compared with 74% in 2013.
The interest rate environment has proven challenging for euro area money market funds (MMFs). Outflows from euro area MMFs continued in the second quarter of 2014, taking assets under management for the industry 36% (€488 billion) below their pre-crisis level (see Chart 2.4). While the average large MMF has some room to absorb the impact of recent rate declines (given a gross average yield of 35 basis points and a net average yield of 18 basis points at end-May 2014), the pressure of negative money market rates has resulted in some fund managers activating reverse distribution mechanisms (to maintain value at par) and temporary “soft closures”, while others
2 Average daily turnover for the Eurex GC Pooling ECB Basket (which includes assets rated A-/A3 and above) has been rising since June and increased markedly following the September rate cut (from €14 billion to €20 billion), while turnover for the ECB Extended Basket (where assets are rated according to ECB eligibility criteria, currently BBB-/Baa3) remains close to May levels.
The current interest rate environment is challenging for money market funds
Chart 2.3 daily turnover in the Eurex gC pooling ECB and ECB Extended Baskets
(Jan. 2011 – Nov. 2014; EUR billions)
0
5
10
15
20
25
0
5
10
15
20
25
2011 2012 2013 2014 Jan. July Jan. July Jan. July Jan. July
core extended
Sources: Bloomberg and ECB calculations. Notes: The core ECB basket only includes assets rated A-/A3 and above. The extended basket includes assets based on ECB eligibility criteria, currently BBB-/Baa3.
Chart 2.3 daily turnover in the Eurex gC pooling ECB and ECB Extended Baskets
(Jan. 2011 – Nov. 2014; EUR billions)
0
5
10
15
20
25
0
5
10
15
20
25
2011 2012 2013 2014 Jan. July Jan. July Jan. July Jan. July
core extended
Sources: Bloomberg and ECB calculations. Notes: The core ECB basket only includes assets rated A-/A3 and above. The extended basket includes assets based on ECB eligibility criteria, currently BBB-/Baa3.
Chart 2.4 Assets of euro area money market funds and the ECB deposit rate
(Q1 2006 – Q3 2014; EUR billions; index of notional stocks; percentages)
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
100
105
110
115
120
125
130
135
140
2006 2008 2010 2012 2014
index of notional stocks for euro area money market funds (left-hand scale) ECB deposit facility rate (right-hand scale)
Sources: ECB and ECB calculations.
40 ECB Financial Stability Review November 20144040
asked for early repayments of commercial paper by issuers, in order to roll over into longer maturities before rates become negative.3
Money market investors have responded to falling euro area money market rates by rebalancing portfolios towards longer-dated funds and non-euro area instruments. The aforementioned decline in the assets of MMFs has coincided with a significant expansion of euro area bond funds.4 Following the introduction of negative policy rates in June, MMFs reported a further rebalancing by investors away from short-term MMFs (with a weighted average maturity of 120 days) towards longer-dated MMFs (with a weighted average maturity of up to one year), euro area bond funds and bank deposits. At the same time, the widening of the spread between euro area money market rates and those of foreign markets may be contributing to a rebalancing away from euro area instruments (see Chart 2.5). Over the past year euro area investors have switched from being net sellers to net purchasers of foreign money market instruments, while foreign investors have become net sellers of euro area money market instruments.
As MMF assets have declined, so too have their holdings of euro area bank debt securities. At the same time, Basel III regulation encourages banks to lengthen their funding maturity structure. From their peak in March 2009, the value of MMF holdings of euro area banks’ debt securities has fallen by €125 billion, a figure equivalent to 15% of all short-term (with an original maturity of less than two years) bank debt securities outstanding at that time, while loans to banks have declined by €76 billion.5 However, over this period banks have, in response to regulatory and market pressures, reduced their reliance on market debt funding and lengthened the maturity of debt funding: the outstanding amount of short-term bank debt securities has fallen by a third, while that of longer-term debt securities has fallen by 5%. Moreover, some investment outflows from MMFs may have been diverted directly (via increased bank deposits) or indirectly (via euro area bond funds) to banks.6
Changes in the regulation of US MMFs will more closely align the structures of the US and European MMF industries and could have important implications for short-term US dollar funding for large euro area banks. The US Securities and Exchange Commission is requiring prime funds
3 Large money market fund refers to the 29 large funds rated by S&P. Reverse distribution mechanisms allow fund managers to reduce the number of outstanding shares in proportion to the reduction in value of the fund over a day in which returns were negative. Soft-closing a fund to new investors avoids the returns of existing investors being heavily diluted by a need to buy paper with a zero or even negative yield.
4 Assets of euro area bond funds have increased by 55% (€1.6 trillion) since June 2008. 5 Money market funds may also purchase securities with a short-term remaining maturity. The €125 billion figure is equivalent to 3% of all
bank debt securities at that time. 6 Euro area bond funds have increased their holdings of bank debt securities by €25 billion over the crisis period. Meanwhile, the ECB’s
Money Market Contact Group reports some disintermediation from MMFs towards bank deposits.
Investors are rebalancing away
from euro area money market funds
and instruments…
… which has implications for short-term bank
funding…
… while changes in US regulation have implications for US
dollar funding for some large euro area
banks
Chart 2.5 One-year forward overnight index swap rates in one year in the euro area and the united States (May 2013 – Nov. 2014; percentages)
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
May July Sep. Nov. Jan. Mar. May July Sep. Nov. 2013 2014
EUR OIS one-year in one year forward USD OIS one-year in one year forward
Source: Bloomberg.
41 ECB
Financial Stability Review November 2014 41
2� F InancIal Markets
41
(invested in non-government securities) and municipal funds (invested in securities issued by local authorities) held by institutional investors to convert from constant net asset value (CNAV) to variable net asset value (VNAV). Around 40% of the US industry will be affected by the mandatory conversion which will result in a closer alignment of the US and European industries.7 Potential outflows from US MMFs as a result of regulatory changes might represent a challenge for some large euro area banks. The five euro area banks most active in US commercial paper have around USD 200 billion in outstanding issues that are subscribed by US MMFs which are likely to experience outflows following the change in regulation.
2.2 YIELdS AT RECORd LOWS AMId A SLIghT INCREASE IN CREdIT RISk pREMIA
Global credit markets have been affected by bouts of volatility and an increase in risk aversion amid rising geopolitical tensions and concerns regarding the global growth outlook. Similar to events last summer, high-yield corporate bond and equity markets were hit hardest during bouts of market tensions.8 In contrast to last year, adjustments in euro area equity and certain sovereign bond markets have been larger than those observed in other regions, a reflection of diverging economic cycles. Although short-lived, these gyrations highlighted four key vulnerabilities in global financial markets. First, there is a growing correlation in global asset price movements. Second, current high valuations are supported by low risk-free rates and subdued market volatility, both of which are sensitive to negative economic news and changing expectations regarding the future path of global monetary policy. Third, concerns regarding stretched valuations for lower-rated corporate bonds make this market segment particularly vulnerable to changing risk sentiment. Finally, low levels of secondary market liquidity in fixed income markets will amplify the price impact of future outflows.
In many ways, current conditions in financial markets echo those of the pre-crisis era: low yields, high correlations across markets and compressed credit spreads sustained by relatively low levels of market volatility and expected default frequencies for corporates (see Chart 2.6). However, while these conditions were conducive to a significant build-up of financial sector leverage during the pre-crisis era, the post-crisis environment has been characterised by an ongoing process of bank deleveraging (see Box 2). At the same time, however, investment funds, which embed leverage- like redemption risk, have been growing in size and their role in financial markets and credit intermediation has increased considerably. In addition, use of leverage in securities markets has been increasing, particularly in the US, where growth in leveraged financing, collateralised loan obligations (CLOs), collateralised debt obligations (CDOs) and the use of margin financing has been quite strong.9
The euro area investment fund sector has doubled in size since 2009, with assets over €10 trillion in September 2014 (see Overview Chart 5 and Box 2). In terms of assets, over 99% of funds are open-ended, while a declining proportion of their assets are liquid (see Chart 2.7). This raises stability concerns as demandable equity in these funds can have the same fire-sale properties as
7 European MMFs are about 55% invested in VNAV and 45% in CNAV. 8 In May and June 2013, a sharp change in market expectations regarding the Federal Reserve’s asset purchase programme resulted in
market tensions. 9 Leveraged financing has been increasing and is expected to reach USD 925 billion globally in 2014. Within Europe, issuance looks set
to reach a post-crisis peak of €150 billion this year, almost treble the level it was two years ago, if 25% below its 2007 peak. Issuance of CLOs and CDOs has also been increasing. While record levels of CLO issuance in the United States are dominating global developments, signs of a recovery in this market segment are also evident in the euro area. See Securities Markets Risk Outlook 2014-15, International Organization of Securities Commissions, October 2014.
Bouts of volatility hint at vulnerabilities in financial markets
While banking sector leverage continues to fall…
… stability concerns arise from growing leverage-like risks in the non-bank financial sector…
42 ECB Financial Stability Review November 20144242
short-term debt funding. Periods of market volatility have shown that investors in these funds, in particular high-yield bond funds and exchange-traded funds, are quite sensitive to price developments and changing expectations regarding the growth outlook or the future path of monetary policy. In addition, certain asset managers report that relatively low cash buffers are being compensated for with credit lines to the banking sector. Funds resident in the euro area are highly interconnected with euro area credit institutions as well as an important and growing source of credit for non-financial corporates and governments (see Chart 2.8). These funds hold 9% of outstanding debt securities issued by euro area credit institutions and provide €370 billion in loans to euro area banks. In addition, they hold a quarter of debt securities issued by euro area non-financial corporates. Therefore, difficulties in the sector can propagate quickly to the banking sector and real economy.
The substantial expansion of fixed income markets has coincided with a decline in secondary market liquidity. Changes in
… and a decline in secondary market
liquidity.
Chart 2.7 Liquid assets as a percentage of shares/units issued by euro area bond funds (Q4 2009 – Q3 2014; percentages; four-quarter moving average)
30
32
34
36
38
40
42
30
32
34
36
38
40
42
2010 2011 2012 2013 2014
Sources: ECB and ECB calculations. Note: Liquid assets include all euro area government debt securities, debt securities issued by euro area residents with an original maturity of up to one year, debt securities issued by non-euro area residents with an original maturity of up to one year and equities issued within the European Union, the United States and Japan.
Chart 2.8 percentage of debt securities issued by euro area non-financial corporations, MFIs (excluding the Eurosystem) and governments held by euro area investment funds (Q4 2008 – Q3 2014; percentages)
0
5
10
15
20
25
0
5
10
15
20
25
MFI debt securities
Government debt securities
Non-financial corporate debt
securities
Q4 2008 Q3 2014
Sources: ECB and ECB calculations. Note: MFIs refer to monetary financial institutions (excluding the Eurosystem) which comprise credit institutions and money market funds.
Chart 2.6 Average cross-correlations between CISS sub-indices and yields on high-yield euro area corporate bonds (Jan. 1999 – Nov. 2014; cross-correlation; percentages)
0
5
10
15
20
25
30
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
1999 2002 2005 2008 2011 2014
periods of rising correlation and falling yields CISS correlation contribution (left-hand scale) yields on euro area high-yield bonds (right-hand scale)
Sources: Bloomberg, Bank of America Merrill Lynch and ECB calculations. Note: For further details, see Hollo, D., Kremer, M. and Lo Duca, M., “CISS – a composite indicator of systemic stress in the financial system”, Working Paper Series, No 1426, ECB, March 2012.
43 ECB
Financial Stability Review November 2014 43
2� F InancIal Markets
43
secondary markets following the outbreak of the financial crisis have profoundly altered the supply and demand of market liquidity. Post-crisis regulation and the substantial expansion of the bond market have reduced the ability and willingness of some market participants to provide sufficient liquidity. Credit disintermediation has seen the outstanding stock of euro area non- financial corporate (NFC) debt securities double to reach €1.2 trillion in 2014, while the supply of market-making services by traditional market-makers, in particular banks, has declined.10 While it cannot be excluded that other market participants may fill the void over time, there is a risk of a shortage in market-making services in the short run. Recent bouts of volatility have highlighted that liquidity problems are not confined to the corporate segment but are broad based across fixed income markets. In addition, other markets that contribute to a smooth functioning of secondary fixed income markets have also declined during the post-crisis era.11 At the same time, structural changes in the asset management industry – for example, a proliferation of passive trading strategies and liquidity transformation – may have increased the pro-cyclicality of demand for market liquidity during stressed times.
10 While the outstanding stock of NFC debt securities has doubled, euro area banks’ holdings of these securities have fallen from €250 billion (over 40% of debt securities outstanding) to €150 billion (less than 13% of debt securities outstanding).
11 For example, since the outbreak of the financial crisis repo volumes have fallen considerably in the euro area and other advanced economy markets.
Box 2
STRuCTuRAL ANd SYSTEMIC RISk FEATuRES OF EuRO AREA INVESTMENT FuNdS
In addition to remarkable growth in the euro area shadow banking sector over the last years, its structure has also been evolving.1 By mid-2014, investment funds domiciled in the euro area had grown to a large size – with money market funds (MMFs) and non-MMF investment funds (IFs) representing almost half of the €19.6 trillion euro area shadow banking sector. Clearly, these structural changes require an adaptation of financial stability monitoring, to understand the role of the investment fund sector and its prospective role in originating or transmitting systemic risk. To this end, this box uses granular data for a sub-sample of all euro area investment funds to further characterise the euro area investment fund universe (including MMFs and IFs but excluding hedge funds).2 This sample excludes hedge funds and covers roughly half of the euro area investment fund population. Within the aggregated assets under management (AuM) of the analysed sample, equity funds represent the largest share of this total (33.1%) followed by bond (29.8%), money market (17.6%) and mixed (14.7%) funds (see Chart A).
The analysis in this box provides evidence of concentration of investment funds managed by individual asset management companies at both the asset class and the aggregate portfolio
1 This approximation follows the Financial Stability Board’s broad measure adding together data on the assets of MMFs and other financial intermediaries (OFIs). The ECB’s 2014 Banking Structures Report reviews in detail the different components of the euro area non-bank financial sector (including the shadow banking sector) at the aggregate level.
2 The box uses end-June 2014 data from Lipper for Investment Management (LIM) covering 26,392 domiciled investment funds in the euro area and managing approximately €5.4 trillion of assets. By comparison, ECB statistics indicate that IFs (including hedge funds) managed almost €10 trillion of assets as at the second quarter of 2014 (see http://www.ecb.europa.eu/stats/money/mfi/html/index.en.html).
44 ECB Financial Stability Review November 20144444
levels. This, combined with significant cross- border retail flows, calls for a close financial stability monitoring, not least given the open- ended nature of much of this sector and its associated vulnerability to run risk.
Euro area investment funds are open-ended funds commonly subject to early redemption claims…
Investment funds invest in assets – equities or debt instruments with predominantly medium to longer-term maturity – while being financed by liabilities (commonly shares/units issued) redeemable at short notice. In a scenario of systemic stress, the structural aspects related to this redeemable-on-demand feature, the use of leverage and knowledge of the ultimate risk bearer are particularly relevant. Within the analysed sample, 69% of funds and 87% of AuM are regulated by the UCITS (Undertakings for Collective Investment in Transferable Securities) Directive.3 The UCITS label is only applicable to (and hence a proxy for the predominance of) open-ended structures. It implies a primarily EU investor base not necessarily corresponding to the fund domicile. Due to their intra-day tradability and specific liquidity features, the early redemption risks of exchange-traded funds (ETFs) are considered even higher. Within the analysed sample, euro area-domiciled ETFs – 95% of which are regulated as UCITS – account for 5% of funds and 6% of AuM. They predominantly invest in less liquid assets as reflected in a preponderance of structures with an investment policy linked to commodities, other assets and equities. For the analysed sample of euro area investment funds, only 1.4% of AuM and 2.5% of funds are potentially leveraged, a reflection of the high proportion of UCITS funds which face restrictions as regards their use of leverage and the exclusion of hedge funds from the sample.4
… and are predominantly owned by retail investors not necessarily residing in the fund domicile jurisdiction
From a financial stability perspective, information on the investor base is important to identify the ultimate risk bearer and to assess the likelihood of contagion to other parts of the financial system under stressed conditions. It also provides a gauge for the likely reaction speed of the investor base to market developments. For example, the experience from the period surrounding
3 Directive 2014/91/EU of the European Parliament and of the Council of 23 July 2014. 4 LIM allocates a leverage flag to investment funds foreseeing as part of their investment mandates to borrow money or to invest based
on anticipated future returns.
Chart A Size and number of funds in the euro area investment fund universe by investment policy (Q2 2014; EUR trillions; number of funds by underlying regulatory framework)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.07,2487,248 1,903 4,580
1,818
871 207 4,237
2,141
603 1,125
37 229 535
685 121 52
1 2 3 4 5 6 7 8
not UCITS UCITS
5 Alternatives 6 Real estate 7 Other 8 Commodity
1 Equity 2 Bond 3 Money market 4 Mixed assets
Sources: LIM and ECB calculations. Note: The UCITS label proxies the predominance of open-ended fund structures within an investment policy category.
45 ECB
Financial Stability Review November 2014 45
2� F InancIal Markets
45
and including the money fund crisis of September 2008 indicates that, for MMFs, institutional investors tended to react more quickly to deteriorating market conditions and prospects of perceived liquidity shortfalls than retail investors did.5 Within the analysed sample of euro area investment funds, 80% of assets on average are held by retail investors, compared with 13.8% by institutional investors and 6.2% by other investor types.6 Only in the MMF category do institutional investors own a relatively higher share of assets (41.5%) compared with retail investors (53.1%) and other investors (5.4%).
Large fund size variation with big players in each asset class…
While large investment funds can be economically efficient, their size naturally determines the market impact of any investment decisions they take. The distribution of euro area-domiciled fund sizes points for each investment policy to a concentration of assets managed in a number of bigger funds (see Chart B). This feature is particularly noteworthy for MMFs, where the average size is 8.4 times the median fund size, compared with 3.9 and 4.1 times for bond and equity funds respectively.
… and funds managed by a small number of large management companies shape market developments
The concentration at individual fund level is further augmented by the concentration of assets managed (across investment policies) at the individual management company level. The combination of size, range of funds managed and consequently importance in different market segments leads these institutions – through investment, portfolio allocation or rebalancing decisions – to define or to drive market developments in normal and in stressed conditions. A Lorenz curve representation illustrates the dominance of a limited number of asset management companies (see Chart C). This concentration has potential consequences: (i) developments at an individual fund could have an adverse impact on the reputation of a specific management company as a whole; or (ii) it could drive market developments or spread market shocks in the financial system. The footprint of a small set of large asset management companies in the euro area investment fund sector (representing 40% of AuM and 21% of funds) is particularly noteworthy in this context (see Chart D).
5 Schmidt, L., Timmermann, A. and Wermers, R., “Runs on Money Market Mutual Funds”, working paper, 2 January 2013. 6 LIM defines institutional funds as funds targeting institutional investors and likely to require a large minimum investment. Other funds
are defined as insurance funds (i.e. an insurance product) plus private funds (i.e. a fund with less than 50 investors). Retail funds are approximated by subtracting institutional and other funds from the total number of funds.
Chart B Investment fund size distribution by investment policy
(Q2 2014; EUR millions)
0
100
200
300
400
500
600
700
800
900
1,000
0
100
200
300
400
500
600
700
800
900
1,000
1 2 3 4 5 6 7 8 5 Alternatives 6 Real estate 7 Other 8 Commodity
1 Equity 2 Bond 3 Money market 4 Mixed assets
interquartile range median average
Sources: LIM and ECB calculations. Notes: Only interquartile ranges, medians and averages are represented. High average figures indicate the presence of very large funds.
46 ECB Financial Stability Review November 20144646
gOVERNMENT dEBT MARkETS Yields on global government bonds for advanced regions with safe-haven status have fallen further to historically low levels (see Chart 2.9). Safe-haven assets attracted strong demand during the summer amid rising political tensions and concerns regarding growth and low inflation, particularly in the euro area. As a result, yields on higher-rated government bonds fell to new troughs. The decline in yields on German government bonds amplified the decreases in the yields on other safe-haven assets outside the euro area owing to further monetary policy easing and market expectations of the introduction of further non-standard measures by the ECB. For the first time on record, the yield on the two-year German government bond fell into negative territory and the yield on the Bund declined markedly below 1%. On the other side of the Atlantic, strong economic data and the phasing-out of quantitative easing by the Federal Reserve offset somewhat the
Yields on higher- rated government
bonds are at historical lows
Chart 2.9 Nominal yields on selected ten-year government bonds compared with historical levels (Jan. 1914 – Nov. 2014; percentages; interquartile range)
0
2
4
6
8
10
12
14
16
0
2
4
6
8
10
12
14
16
United States Japan Germany United Kingdom
median interquartile range
November 2014
Source: Global Financial Data.
Chart C Lorenz curve for the distribution of assets by management company parent
(Q2 2014; x-axis: percentage of fund management company parent; y-axis: percentage of assets managed; Gini coefficient (percentage))
0
20
40
60
80
100
0
20
40
60
80
100
total (Gini coefficient: 90.5) equity (87.8) bond (84.7) money market (87.8) mixed assets (86.1) alternatives (75.4) real estate (85.5) other (81.9) commodity (67.5)
0 20 40 60 80 100
Sources: LIM and ECB calculations.
Chart d Assets and number of euro area funds managed of the top-15 management company parents (Q2 2014; EUR trillions; number of funds)
0.0
0.1
0.2
0.3
0.4
0.0
0.1
0.2
0.3
0.4
156451467727336255944825548096
172 722
856
406
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 BlackRock 2 Amundi Group 3 Deutsche Bank 4 JPMorgan Chase & Co 5 Franklin Templeton 6 Eurizon Financial Group 7 Union Gruppe 8 UBS
9 UniCredit 10 DekaBank 11 PIMCO 12 Goldman Sachs 13 Groupe BPCE 14 BNP Paribas 15 FIL International
Sources: LIM and ECB calculations.
47 ECB
Financial Stability Review November 2014 47
2� F InancIal Markets
47
Chart 2.10 Yield on the ten-year german government bond and spreads between it and selected euro area government bonds (Jan. 2009 – Nov. 2014; percentage points)
0
5
10
15
20
25
30
35
40
0
2
4
6
8
10
12
14
16
Germany Italy Spain
Portugal Ireland Greece (right-hand scale)
0
2
4
6
8
10
0
1
2
3
4
5
2009 2010 2011 2012 2013 2014 16 May 16 Sep.
Sources: Bloomberg and ECB calculations. Note: Long-term average refers to the period from 1965 to 2014.
compression of yield spreads on US Treasuries resulting from safe-haven flows. As a result, the spread between the US and the German ten-year government bond yields widened to over 160 basis points, its highest level since the beginning of the single monetary policy in the euro area.
The broad-based rally within euro area government bond markets was briefly interrupted by bouts of market volatility owing to concerns about euro area growth and fiscal debt sustainability for certain countries. Investors appear to be increasingly discriminating among euro area sovereign bonds based on the evolution of fiscal fundamentals. Within the higher-rated segment, yields on ten-year Belgian bonds fell below those of France. These countries stand in contrast as regards fiscal developments this year (see Section 1.2). Within the lower-rated segment, the gap between yields on Spanish and Italian government bonds has widened further (see also Section 1.2). Meanwhile, Greek government bond yields rose sharply amid public debt sustainability concerns.
Intra-euro area spreads hit new post-crisis troughs and remain at low levels for most countries despite bouts of market tensions (see Chart 2.10). Yields on lower-rated euro area sovereign bonds have benefited from sovereign rating upgrades and proved resilient to rising geopolitical tensions but vulnerable to negative economic data and concerns regarding fiscal sustainability in one country. Worryingly, market gyrations in October hinted at low levels of secondary market liquidity in certain segments and highlighted the ability for difficulties in one market to quickly propagate to another. Riskier sovereign markets did experience a withdrawal of foreign investment that was offset by demand from euro area investors. Although part of the euro area support appears to have been domestic bank-based, non-domestic institutional investors played an important role as well. While lower-rated sovereigns have taken advantage of benign conditions to improve their fiscal outlook by frontloading issuance, smoothing repayment schedules and lengthening maturities, market conditions are vulnerable to any further signs of weakness in the euro area recovery.
The broad-based rally in euro area sovereign bond markets continues…
… and intra-euro area spreads have fallen further
48 ECB Financial Stability Review November 20144848
While nominal yields on government bonds are touching record lows, real yields on government bonds are less extreme. Real yields on higher-rated government bonds (United States, Germany, Japan and the United Kingdom) are above record lows, but do fall within the lowest quartile of observations over the last century (see Chart 2.11). Meanwhile, real yields on lower-rated euro area bonds (such as those in Italy and Spain) are close to their century medians. The current compressed level of real yields on higher-rated bonds reflects strengthened demand (owing to regulatory considerations) for a reduced pool (owing to rating downgrades) of high-quality liquid assets and – in the case of the United States, the United Kingdom and Japan – large acquisitions by central banks.12 Indeed, while interest rates on such sovereign paper continue to touch historically low levels, government debt-to-GDP ratios remain elevated and there is a risk of potential sharp adjustments as central banks exit from quantitative easing programmes. Moreover, recent market gyrations indicated that such adjustments could be amplified by lower levels of secondary market liquidity post crisis.
One factor that could underpin the current low level of nominal and real yields is that markets are pricing in the potential for a protracted period of low growth, low inflation and therefore accommodative global monetary policy. If borne out, a protracted period of low growth could hamper debt sustainability. The level of public (and private sector) debt-to-GDP ratios is historically high across most regions (see Section 1). If, on the other hand, the recovery in the United States and the United Kingdom endures, monetary tightening could be implemented sooner than expected by markets and, despite ample warnings, substantial corrections could be triggered. Under such a scenario, a sharp adjustment in US term premia is likely. While weaker than expected euro area growth remains the most significant threat to the euro area government bond markets, a sharp increase in US term premia is also a cause for concern. While forward guidance has been successful in containing spillovers from rising US money market rates, the extent to which the long end of the euro area bond yield curve might react to a significant repricing of US term premia is still a worry.
CORpORATE CREdIT MARkETS A search for yield continues in corporate credit markets. While rising geopolitical tensions and concerns regarding stretched valuations in the high-yield segment temporarily affected investor appetite for credit risk, investors were willing to increase duration exposure (see Charts 2.12 and 2.13). At the same time, credit spreads for both the investment-grade and high-yield segments remain at relatively low levels and the market continues to absorb record levels of corporate bond issuance. In addition, investor demand for higher-yielding complex products – such as corporate hybrids – remains strong. 12 The Bank of England, the Bank of Japan and the Federal Reserve hold roughly 27%, 24% and 15% of domestic government bonds
respectively. The ECB holds less than 3% of euro area government debt securities.
Real yields are less extreme but still
low, despite elevated government debt-to-
GDP levels…
... and vulnerable to changing market
expectations regarding the growth outlook and the path
of global monetary policy.
The search for yield continues in
corporate credit markets
Chart 2.11 Real yields on selected ten-year government bonds compared with historical levels (Jan. 1914 – Oct. 2014; percentages; current, median and interquartile range)
-3
-2
-1
0
1
2
3
4
5
6
-3
-2
-1
0
1
2
3
4
5
6
United States Japan Germany United Kingdom
median November 2014
Source: Global Financial Data. Note: Yields are deflated using the consumer price index measure of inflation.
49 ECB
Financial Stability Review November 2014 49
2� F InancIal Markets
49
Prices and average durations for euro area investment-grade corporate bonds maintained their steady rise during the summer, although geopolitical tensions temporarily weighed on issuance. Having taken advantage of attractive funding costs in the first half of the year, issuers did not appear willing to test the market over the summer amid increased global risk aversion. As a result, issuance was weak but rebounded in the autumn as geopolitical tensions subsided somewhat.13 At the same time, credit spreads for investment- grade bonds reached a new post-crisis trough, while average duration rose above pre-crisis levels.
Low risk-free rates have sustained high-yield corporate bond yields at historical lows despite a widening of credit spreads amid investor outflows from lower-rated bond funds (see Charts 2.12 and 2.14). Weak returns during the year and concerns regarding stretched valuations, particularly in the US market, made the corporate segment quite vulnerable to the sudden change in market sentiment that occurred during the summer. In the euro area, concerns regarding Banco Espirito Santo and Portugal Telecom temporarily added to negative market sentiment. Weekly outflows from US and European lower-rated bond funds reached a magnitude that surpassed levels observed last summer during the so-called “taper tantrum”.
13 It was the strongest September for euro investment-grade fixed rate issuance since 2012.
Investors have increased duration exposure to investment-grade issuers…
… and withdrawn from the high-yield segment
Chart 2.12 Spreads on investment-grade and high-yield corporate bonds
(Jan. 1998 – Nov. 2014; basis points)
0
500
1,000
1,500
2,000
2,500
0
200
400
600
800
1,000
0
200
400
600
0
100
200
300
US investment-grade US high-yield (right-hand scale)
euro investment-grade euro high-yield (right-hand scale)
1998 2001 2004 2007 2010 2013 15 May 15 Sep.
Sources: Bloomberg and ECB calculations.
Chart 2.13 Modified duration of long-term investment-grade euro area corporate bonds by rating category (Jan. 2000 – Nov. 2014; 30-day moving averages; years)
5.5
6.0
6.5
7.0
7.5
5.5
6.0
6.5
7.0
7.5
A-rated AA-rated BBB-rated
Finalisation of the May 2014 FSR
2000 2002 2004 2006 2008 2010 2012 2014
Sources: Bank of America Merrill Lynch. Note: Long-term bonds refer to bonds with maturities of between seven and ten years.
50 ECB Financial Stability Review November 20145050
While noteworthy, the recent outflows and increase in credit spreads need to be placed in the context of substantial inflows over the past three years which have pushed credit spreads close to pre-crisis lows, while issuance has reached record levels (see Chart 2.15). High-yield credit spreads have fallen almost 20 percentage points from crisis peaks to within 150 basis points of pre-crisis troughs. While euro area corporate bond issuance slowed in the third quarter of this year owing to weakened demand, it was still the strongest third quarter for deal volumes on record. Moreover, underwriting standards of high-yield issuances continue to weaken, as evidenced by increased growth in covenant-lite loans and payment-in-kind bonds.
The speed and magnitude of the declines in corporate credit spreads (for both investment- grade and high-yield bonds) from sovereign crisis peaks mirror developments during the pre-crisis era (see Chart 2.16). Similar to that period, current low levels of market volatility and expected default frequencies provide some justification for the compressed level of credit spreads. However, levels of corporate indebtedness are much higher now (see Section 1.3). In addition, increases in average maturity and durations raise concerns over whether investors are adequately compensated for the default rates and market
Nonetheless, credit spreads remain at relatively low
levels…
… raising some concerns that investors may
not be adequately compensated for risk
Chart 2.14 Net weekly flows of retail and institutional investors to/from high-yield euro area bond funds (Jan. 2011– Nov. 2014; USD millions)
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
2011 2012 2013 2014
institutional retail
Sources: EPFR and ECB calculations. Note: Data capture funds located in Austria, Belgium, Cyprus, Estonia, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal and Spain.
Chart 2.15 quarterly issuance of euro area high-yield and investment-grade corporate bonds (Q1 2000 – Q3 2014; EUR billions)
0
20
40
60
80
100
120
0
20
40
60
80
100
120
2000 2002 2004 2006 2008 2010 2012 2014
investment-grade high-yield
Sources: Dealogic and ECB calculations.
Chart 2.16 developments in credit spreads on BBB and CCC-rated euro bonds since 2011 compared with 2002 (Oct. 2002 – Nov. 2014; basis points)
0
50
100
150
200
250
300
350
400
450
500
0
50
100
150
200
250
300
350
400
450
500
Oct. 02 Dec. 11
Oct. 04 Dec. 13
Oct. 06 0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000BBB-rated CCC-rated
May 2014 FSR cut-off
Oct. 02 Dec. 11
Oct. 04 Dec. 13
Oct. 06
period from October 2002 period from December 2011
May 2014 FSR cut-off
Sources: Bloomberg, Bank of America Merrill Lynch and ECB calculations.
51 ECB
Financial Stability Review November 2014 51
2� F InancIal Markets
51
volatility they could expect over the entire life of the bond.14 Moreover, past experience teaches us that pervasive low levels of volatility are rare and tend to be short-lived (see Box 3). In addition, current low levels of default are dependent on the endurance of: (i) low market volatility; (ii) the recovery in euro area growth; and perhaps also (iii) low interest rates. Furthermore, the strong correlation between corporate and sovereign bonds (particularly within vulnerable euro area countries) suggests that risk factors affecting sovereign bond markets, mainly a worsening of the still fragile economic recovery and a disorderly repricing in global markets, could propagate quickly to corporate bond sectors.
Euro area corporate hybrid bonds exhibited some temporary price and issuance volatility in recent months, owing to geopolitical tensions and a one-off shock to the banking sector. There was a hiatus in bank Additional Tier 1 contingent convertible bond issuance during the summer, as banks were unwilling to test the market following the bail-in of the subordinated bonds of Banco Espirito Santo. However, the impact of the banking sector shock proved short-lived and issuance and prices rebounded strongly in autumn. During this period bank issuance offset a slowdown in NFC hybrid issuance as firms started to fulfil their targeted programme amounts and the large-scale mergers and acquisitions that would warrant hybrid issuance to protect ratings did not materialise.
Demand for complex high-yielding products is evident in a resurgence of CLOs, particularly in the United States, and the emergence of capital relief trades (CRTs). While global issuance of securitised products remains flat, issuance of CLOs has grown significantly, surpassing pre-crisis peaks in the United States.15 CLO issuance in the euro area has been growing, but remains subdued relative to pre-crisis peaks, a reflection perhaps of post-crisis risk-retention rules. However, a rebound in the issuance of other securitised products in the euro area, in particular asset-backed securities (ABSs), is expected over the coming year following the ECB’s announcement that it would engage in purchases of senior ABS tranches and mezzanine tranches provided that they are guaranteed.16 A number of sophisticated CRTs, whereby a bank pays a third party to take on some risk associated with its asset exposures, have been reported over the past year.17
14 The average maturity of a euro area corporate bond issued in the third quarter of 2014 was six years. 15 In 2014, the issuance of securitised products is expected to reach USD 691 billion globally, still well below its peak. 16 Issuance increased noticeably in September following the ECB announcement. At the same time, a Bloomberg survey among market
participants found that they expect the euro area ABS market to grow significantly in the coming year. Nearly 60% of respondents to a Bloomberg survey think that structured finance issuance will increase over the next 12 months, compared with 33% in the previous survey. The Q3 2014 reading is the highest in the survey history and is higher than for any other asset class.
17 These include the sale of shipping loans by Citigroup to Blackstone, the sale of multiple loan portfolios by Unicredit to Barclays and the sale of trade finance loans by Standard Chartered.
Demand for complex high-yielding products remains strong at the euro area…
… and global level
Box 3
FINANCIAL MARkET VOLATILITY ANd BANkINg SECTOR LEVERAgE
Global asset market volatility remained persistently at historical lows across financial asset classes and economic regions from the third quarter of 2013 up until early October 2014.1 Low financial market volatility may in many ways reflect fundamentals, including low uncertainty regarding policies, limited surprises in economic releases and the stabilising influence of more
1 In October 2014, a deterioration of the economic outlook in major advanced and emerging economies, including the United States and China, triggered an episode of market volatility in several asset markets.
52 ECB Financial Stability Review November 20145252
stringent post-crisis regulation of the financial sector. At the same time, financial stability risks may arise from investor complacency especially during periods of weak returns on financial assets when investors hunt for yield. Such periods have the potential to embed systemic risk, if they lead to an excessive build-up in leverage or maturity extension.
The broad-based nature of this current period of record low volatility is particularly noteworthy. Option-implied stock market volatility (as measured, for instance, by the VIX) and derived measures of uncertainty and risk aversion have approached record low levels.2 At the same time, realised market volatility has remained at extremely low levels for the past five consecutive quarters (up until the end of the third quarter of 2014) in thirteen major asset markets (G3 equity, government bond, corporate bond and FX markets, as well as two major commodity markets; see Chart A). Indeed, the average annualised daily market volatility of these markets has fallen to a range of 6.3% to 9.5% – even lower than daily volatility of 7.9%-12.8% for global bond and equity markets on the eve of the global financial crisis. Moreover, volatility is touching record lows across a much broader range of asset categories than it did during the pre-crisis era and is proving more persistent (see Chart A). The former may reflect the growing correlation of global asset markets in the post-2008 period.
According to the volatility paradox hypothesis3, an environment of low yields and volatility could invite excessive risk-taking by financial investors. First, risk aversion tends to decline during prolonged periods of low volatility as suggested by estimates of the volatility risk premium (see Section 2.2). A lower premium amounts to investors demanding less compensation for holding risky assets. Such a fall in the price of risk changes the relative price of assets with a given risk/return trade-off and may lead to portfolio rebalancing in favour of riskier assets. Second, low volatility mechanically compresses backward-looking risk measures, such as the value at risk (VaR), which shape investors’ risk management decisions. In fact, the unit VaR – calculated as the VaR per unit of assets – of a sample of large euro area banks lags a measure 2 For further details, see Financial Stability Review, ECB, May 2014, pp. 55-56 and BIS Quarterly Review, September 2014, pp.10-11. 3 Adrian, T. and Shin, H., “Procyclical Leverage and Value-at-Risk”, NBER Working Paper Series, No 18943, 2013; Adrian, T. and
Boyarchenko, N., “Intermediary Leverage Cycles and Financial Stability”, Federal Reserve Bank of New York Staff Report No 567, 2013; and Brunnermeier, M. and Sannikov, Y., “A Macroeconomic Model with a Financial Sector”, American Economic Review, Vol. 104(2), pp. 379-421, 2014.
Chart A heat map of levels of volatility across major asset markets
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 USEquity EMU JP USGovernment
bonds EMU JP USCorporate
bonds EMU JP USD/EURFX JPY/USD Gold (WTI)Commodities Oil (Brent)
Sources: Thomson Reuters Datastream and ECB calculations. Notes: Volatility estimates are derived from non-overlapping quarterly samples of daily price data. The colour code is based on the ranking of these quarterly estimates in the respective asset market. A red, yellow and green colour code indicates, respectively, a high, medium and low volatility estimate compared with other periods. Equity markets are represented by the respective MSCI price index at the country or region level. Bond markets are represented by the respective JPMorgan government bond index at the country or region level (local currency/all maturities). The last observation is for 30 September 2014. White indicates non-availability of data.
53 ECB
Financial Stability Review November 2014 53
2� F InancIal Markets
53
of stock market volatility in the euro area by about a year. This pro-cyclical behaviour of the VaR allows investors to increase their exposure to assets which are prone to bursts of volatility for a given risk threshold. Finally, cheap funding and subdued risk measures allow investors to increase their leverage, thereby reinforcing the vulnerability of the financial sector at large.
The period of low volatility leading up to the global financial crisis commencing in 2007 is illustrative of such risks via leverage. In that episode, the build-up of banking sector leverage was certainly a side-effect of low market volatility. From 2002 to 2007 banking sector leverage in the United States and the euro area rose considerably (see Chart B). During this period, financial market volatility as measured by the VIX, which is often also interpreted as a yardstick of global risk aversion, was at very low levels. By contrast, the decline in market volatility since mid-2009 has so far not been associated with a renewed increase in banking sector leverage (see Chart B).
There are a number of reasons why the mechanical link between market volatility, risk appetite and banking sector leverage observed ahead of the last crisis does not hold for current developments. Between 2002 and 2007 the pro-cyclical nature of the leverage cycle appeared to follow an empirical regularity whereby in periods of low volatility and low measured market risk, lower risk weights for banks to meet capital adequacy requirements enabled them to build up leverage. Since mid-2009, this mechanism has not yet started to operate for two reasons. First, capital and liquidity requirements for regulated banks have been tightened in the context of more stringent regulatory requirements. Second, the legacy of the crisis has led to a prolonged period of low economic growth. As a result, low credit growth has partly been driven by subdued demand for loans. Finally, the reasons for low market volatility during the leverage cycle between 2002 and 2007 might have been different from those in recent years.
Chart B Banking sector leverage and financial market volatility in the united States and the euro area (percentages)
a) United States b) euro area
40
60
80
100
120
140
160
180
200
0
5
10
15
20
25
30
35
40
VIX – Q4 2002 to Q2 2007 (left-hand scale) VIX – Q2 2009 to Q4 2013 (left-hand scale) banking sector leverage – Q4 2002 to Q2 2007 (right-hand scale) banking sector leverage – Q2 2009 to Q4 2013 (right-hand scale)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 0
25
50
75
100
125
150
175
200
225
250
0
5
10
15
20
25
30
35
40
45
50
VSTOXX – Q4 2002 to Q2 2007 (left-hand scale) VSTOXX – Q2 2009 to Q4 2013 (left-hand scale) banking sector leverage – Q4 2002 to Q2 2007 (right-hand scale) banking sector leverage – Q2 2009 to Q4 2013 (right-hand scale)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Sources: Bloomberg, ECB and ECB calculations. Notes: Data on banking sector leverage (debt/equity) in the United States and the euro area are based on partly consolidated data for comparability purposes. Banking sector debt includes total loans given to banks by non-banks, money deposited by non-banks, total debt securities issued and money market fund shares.
54 ECB Financial Stability Review November 20145454
EquITY MARkETS The broad-based rally in global stock markets was interrupted by bouts of volatility amid growing global risk aversion owing to concerns regarding rising geopolitical tensions and the global growth outlook (see Chart 2.17). Price corrections in euro area stock markets amplified those in the US market for three key reasons. First and foremost, weaker than expected economic data releases for the euro area weighed on earnings expectations. Second, geopolitical tensions weighed more heavily on euro area stocks as the macro-financial consequences of the Ukraine-Russia conflict were considered more severe for the euro area (see Section 1). Finally, certain euro area financial stocks were affected by one-off country and sector-specific shocks.
While there are no clear signs of overvaluation in aggregate euro area stock price indices, price/ earnings ratios for some national markets are significantly above their long-run averages. Although the recovery in euro area stock markets has been remarkable in recent years, the EURO STOXX index still remains 27% below its level in 2007 (see Chart 2.17). Moreover, metrics such as the
Weak economic growth and rising
geopolitical tensions temporarily impacted global stock markets
There are no clear signs of
overvaluation in euro area stock
markets…
Nevertheless, the current period of low volatility may be contributing to rising leverage outside the regulated banking sector (see Section 2.2).
Ultimately, the elusive and time-varying nature of many of these explanatory factors implies a need for monitoring persistently low financial market volatility for financial stability risks. Indeed, given the profound impact of the global financial crisis on both the financial system and the economy, the nature of systemic risks may too be evolving – requiring a broad-based monitoring of low volatility with various measures of leverage including leverage outside the regulated banking sector (for example, embedded in financial market transactions of certain market segments such as derivatives, securities financing or repo markets) as well as any prospect of broad-based liquidity or maturity mismatch that could cause system-wide stress.
Chart 2.17 developments in uS and euro area stock markets
(Jan. 2007 – Nov. 2014; index: Jan. 2007 = 100)
0
50
100
150
0
50
100
150
0
50
100
150
0
50
100
150 S&P 500 (144)
German DAX (107)
French CAC (85) EURO STOXX 500 (73) Spanish IBEX (72)
Italian MIB (52)
Greek ASE (15)
Finalisation of the May 2014 FSR
2007 2009 2011 2013 15 May 15 Oct.
Sources: Bloomberg and ECB calculations.
55 ECB
Financial Stability Review November 2014 55
2� F InancIal Markets
55
cyclically adjusted price/earnings (CAPE) ratio and Tobin’s Q (the ratio of a firm’s market value to its replacement costs) suggest valuations are still in check, with both measures close to long-run averages. However, recent price adjustments, in particular for German and French stock markets, have shown that current valuations depend on a fragile economic recovery with increasing downside risks. Indeed, current valuations are supported by expectations of robust (double-digit) earnings growth for euro area firms over the next year. Moreover, the rally in euro area markets and strong earnings expectations seem to contrast with the growing share of loss-making firms within the region (up from 15% in 2011 to 22% in 2014).18 In addition, at the national level, the trailing price/earnings ratios for stocks in Belgium, Ireland, Spain, the Netherlands and France now deviate substantially from their long-run means and lie outside their interquartile ranges (see Chart 2.18).
As US stock prices enter their fourth year of increase, commonly used metrics of overvaluation signal that valuations are becoming stretched. Both the CAPE ratio for the S&P 500 index and Tobin’s Q for US firms are well above their long-run averages (see Chart 2.19). The CAPE for the S&P 500 is 60% above its long-run average, having reached a level that has only been surpassed on three other occasions in its 188-year history: 1929, 1999 and 2007 (years which preceded significant stock price collapses). Meanwhile Tobin’s Q has risen above 1 for US non-financial firms for the second time in its 69-year history, the only other occasion being the period ahead of the dot-com collapse. As these valuations have grown, the use of leverage also appears to be on the rise. Data on margin financing indicate a large increase in the use of leverage to fund US securities purchases. The rally in the S&P 500 has coincided with a sharp increase in margin financing, which has grown by 350% in the past year to reach record levels in real terms.
18 See Société Générale Cross Asset Research factsheet.
… but some signs of stretched valuations are evident in US markets
Chart 2.18 price/earnings ratios for selected Eu countries and the united States
(Jan. 1980 – Oct. 2014; ratios; maximum, minimum and interquartile range)
0
5
10
15
20
25
0
5
10
15
20
25
BE NL IE FR GR EA ES AT IT FI DE PT
interquartile range mean post-2007 minimum October 2014
Sources: Thomson Reuters Datastream, ECB and ECB calculations. Notes: Unbalanced panel with series starting between 1980 and 1990.
Chart 2.19 Cyclically adjusted price/earnings ratio and Tobin’s q for the uS stock market
(ratios)
0
5
10
15
20
25
30
35
40
45
50
0
5
10
15
20
25
30
35
40
45
50
18811911 1941 1971 2001
Shiller CAPE ratio Tobin’s Q
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
19451960197519902005
US US long-run average
Sources: R. Shiller (Yale University), Federal Reserve Board and ECB calculations.
57 ECB
Financial Stability Review November 2014
3 EuRO AREA FINANCIAL INSTITuTIONS Euro area financial institutions have continued to make steady progress in tackling legacy issues from the financial crisis, while adapting to an evolving regulatory and prudential environment. Bank balance sheets have been strengthened further, with a clear shift towards capital increases in 2014, from deleveraging and de-risking in previous years. While some asset quality concerns remain, the pace of deterioration has slowed considerably. The comprehensive assessment has brought much-needed transparency and confirms that a large majority of the most significant euro area banks is well equipped to withstand a severe economic downturn.1
Notwithstanding these efforts to strengthen balance sheets, a combination of cyclical and structural headwinds has implied weak profitability in many parts of the euro area banking sector. In particular, elevated loan loss provisions and subdued revenues remain a drag on profits in an environment of low growth and flat yield curves. While cyclical headwinds should abate as economic conditions improve, there is a clear need to continue to adapt bank strategies and business models so as to sustainably improve profitability in a post-crisis environment, notably to foster internal capital generation. In this context, bank lending activity remains subdued – with loans to non-financial corporations developing particularly sluggishly, mainly on account of anaemic credit demand and persistent fragmentation of credit conditions. Over time, further progress in removing impediments to the supply of bank credit – also including disposals of non-performing loans – should help improve credit conditions, as should, in particular, the ECB’s targeted measures to improve access to finance essential for economic growth.
Not only banks, but also insurers, for whom a prolonged period of low yields remains a key concern, have been adapting their business models to the prevailing macro-financial environment. While low yields have placed pressure on earnings in the latter sector, the financial performance and capital positions of large euro area insurers have remained sound.
On the policy front, progress continues apace in the regulatory and prudential domains. In the regulatory field, further advances have been made, in particular, in weakening the links between sovereigns and banks, and in building a more resilient banking sector. Since the publication of the last issue of the Financial Stability Review (FSR), much has been achieved to put in place central elements of an integrated financial framework in Europe, especially the euro area, namely (i) the Single Supervisory Mechanism, (ii) a common resolution framework, (iii) a Single Resolution Mechanism and (iv) harmonised deposit insurance. In line with a new and reinforced prudential mandate, a number of euro area Member States have announced specific macro-prudential measures. These include systemic risk measures in order to mitigate systemic risks originating from the significant size, high concentration and interconnectedness in their banking sectors. Different types of property-related measures have been adopted as well, with the aim of addressing unfavourable developments in the property market (see Section 3.3 for a description of measures taken).
3.1 BALANCE ShEET REpAIR CONTINuES, BuT WEAk pROFITABILITY pERSISTS IN ThE EuRO AREA BANkINg SECTOR
FINANCIAL CONdITION OF EuRO AREA BANkS Euro area banks’ profitability remained weak in the first three quarters of 2014, given a confluence of both cyclical and structural factors. In the third quarter of 2014, the median return on equity (ROE) of significant banking groups (SBGs) in the euro area remained broadly unchanged from three months earlier, at around 4%, and showed only a slight improvement on a year-on-year basis (see Chart 3.1). Elevated loan loss provisions remained the most important cyclical drag on bank 1 Given the broad nature of the comprehensive assessment, including a bottom-up stress-test exercise, and the forthcoming stress test by
EIOPA on insurers, sensitivity analyses for financial institutions are not presented in this issue of the FSR.
Bank profitability remains under pressure…
58 ECB Financial Stability Review November 20145858
performance, even if these provisions have fallen somewhat over the last half year. Furthermore, banks are also struggling to boost revenues in an environment of low growth and flat yield curves. In addition to cyclical factors, one-off factors also affected some banks, mainly in the form of large non-recurring expenses related to litigation charges or goodwill write-downs that depressed profits.
At the same time, the de-risking and deleveraging of bank balance sheets (see Chart 3.2) as well as some structural factors – such as strong domestic competition or remaining cost inefficiencies in some parts of the euro area banking sector – have also contributed to lower profitability. This combination of both cyclical and structural headwinds has pushed banks’ ROE well below their cost of equity in the past few years. As the impact of cyclical factors eventually fades away, any weak structural profitability remaining could limit banks’ internal capital generation and provide incentives for banks to take on more risks. Moreover, for some banks, persistently weak profitability also raises questions about the viability of their business models. In this respect, while a number of euro area banks have made progress in restructuring their operations since the start of the crisis, driven by continued pressure to contain costs and reduce non-core activities, the advances have been uneven across different parts of the banking sector. Therefore, further measures need to be taken in parts of the banking sector to adapt business models to new realities, for instance, by refocusing activities on profitable core business, diversifying income sources or further improving cost efficiency.
Low profitability remains a concern for most euro area banks, although the main drivers have differed somewhat across banks and countries in recent years. In countries that experienced a recession in the last few years and where economic recovery remains weak, low or negative bank profitability has been driven primarily by high loan loss provisions (see Chart 3.3). More generally, over the past few years, pre-impairment operating profits remained rather subdued, or showed a decline, on account of a combination of narrowing net interest margins and weak loan volume
… given a combination
of cyclical and structural headwinds…
… due to elevated credit risk costs and compressed interest
margins…
Chart 3.1 Euro area banks’ return on equity
(2007 – Q3 2014; percentages; 10th and 90th percentiles and interquartile range distribution across SBGs)
-35 -30 -25 -20 -15 -10 -5 0 5
10 15 20 25
-35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25
-47% -83%
2007 2009 2011 2013 2013 2014 Q3 Q4 Q1 Q2 Q3
median for SBGs median for LCBGs
Source: SNL Financial. Note: Based on publicly available data on SBGs that report annual financial statements and on data on a sub-set of those banks that report on a quarterly basis.
Chart 3.2 Return on equity and leverage for large euro area banks
(Q1 2004 – Q2 2014; median values for SBGs)
15
16
17
18
19
20
21
0
4
8
12
16
20
24
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
return on equity (percentage; left-hand scale) leverage (multiple; right-hand scale)
Sources: Bloomberg and ECB calculations. Note: Based on publicly available data on a sub-sample of listed SBGs that report quarterly financial statements.
Chart 3.2 Return on equity and leverage for large euro area banks
(Q1 2004 – Q2 2014; median values for SBGs)
15
16
17
18
19
20
21
0
4
8
12
16
20
24
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
return on equity (percentage; left-hand scale) leverage (multiple; right-hand scale)
Sources: Bloomberg and ECB calculations. Note: Based on publicly available data on a sub-sample of listed SBGs that report quarterly financial statements.
59 ECB
Financial Stability Review November 2014 59
3� Euro arEa F inancial
inst itutions
59
growth. For banks in vulnerable countries, net interest income had been negatively affected by higher funding costs as a consequence of the sovereign crisis, while interest margins in some core countries (notably in Germany) have been structurally low for a long time, mainly on account of intense bank competition, a situation that has recently also been exacerbated by low interest rates.
More recently, however, euro area banks’ operating performance showed signs of a moderate improvement – with median pre- impairment profits for SBGs increasing somewhat in the first half of 2014 (see Chart 3.4). This mainly reflected a modest overall increase in net interest income as average funding costs declined more than asset yields (see Chart 3.5), albeit with significant cross-country heterogeneity. In particular, many banks from vulnerable countries recorded an improvement – contrasting with flat or even declining patterns for a number of banks in other countries.
Chart 3.3 Euro area banks’ return on assets, pre-impairment profits and impairment costs in vulnerable and other countries (2007 – H1 2014; percentage of total assets; median values for SBGs)
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2007 2008 2009 2010 2011 2012 2013 H1 2014
H1 2014
2007 2008 2009 2010 2011 2012 2013
pre-impairment profit impairment costs ROA
Vulnerable countries Other countries
Source: SNL Financial. Notes: Based on publicly available data on SBGs that report on a semi-annual basis. Two-period averages for the first half of 2014. “Vulnerable countries” refer to Cyprus, Greece, Ireland, Italy, Portugal, Slovenia and Spain.
Chart 3.4 Euro area banks’ pre-impairment profits and their main components
(2007 – H1 2014; percentage of total assets; median values for SBGs)
-1.6
-1.2
-0.8
-0.4
0.0
0.4
0.8
1.2
1.6
2.0
2.4
-1.6
-1.2
-0.8
-0.4
0.0
0.4
0.8
1.2
1.6
2.0
2.4
2007 2008 2009 2010 2011 2012 2013 H1 2014
operating costs net interest income net fee and commission income trading income pre-impairment profits
Source: SNL Financial. Notes: Based on publicly available data on SBGs that report on a semi-annual basis. Two-period averages for the first half of 2014.
60 ECB Financial Stability Review November 20146060
These cross-country differences in funding costs mainly reflect the marked fall in sovereign yields in vulnerable countries. In these countries, a median decline of 21% in interest costs in the first half of 2014 – resulting from a spillover of lower sovereign yields to both deposit and wholesale funding costs – contrasted with a more moderate decrease in interest costs for banks in other countries (median decline of 9%). Mirroring these patterns, banks in vulnerable countries registered a median increase of 4% in net interest income in the first half of 2014, as compared with a year earlier, compared with a median increase of 2% for banks in other countries.
At the same time, non-interest income decreased slightly in the first half of 2014 due to lower trading income, while fee and commission income remained stable. In the same period operating costs, expressed as a percentage of total assets, decreased somewhat on average reflecting banks’ continued efforts to cut costs (see Chart 3.4). That said, the progress in improving cost efficiency remains uneven across banks with more than one-fifth of SBGs maintaining cost-to-income ratios above 70%, suggesting that for several banks there is scope for further cost containment.
Despite some easing of cyclical headwinds, banks’ financial results have continued to be heavily affected by high impairment costs, albeit to a lesser extent than six months earlier. Stark differences in impairment costs across banks persisted, with smaller banks from vulnerable countries bearing much of the negative impact on results. In the first half of 2014, the median value of loan loss provisions (the bulk of impairment costs) for SBGs in vulnerable countries was still above the average over the five years preceding the sovereign debt crisis (2007-11). By contrast, average loan loss provisions for banks in other countries remained at moderate levels (see Chart 3.6). Furthermore, additional provisioning needs identified by the asset quality review (AQR) are likely to be recognised mostly in banks’ fourth-quarter or full-year 2014 results.
… with high impairment costs affecting mainly
banks in vulnerable countries…
Chart 3.5 Interest spread and its components for significant banking groups in the euro area (2007 – H1 2014; percentages; median values for SBGs)
0.4
0.6
0.8
1.0
1.2
1.4
1.6
-6
-4
-2
0
2
4
6
2007 2008 2009 2010 2011 2012 2013 H1 2014
asset yield (left-hand scale) cost of funds (left-hand scale) interest spread (right-hand scale)
Source: SNL Financial. Note: Based on publicly available data on SBGs that report on a semi-annual basis.
Chart 3.6 Loan loss provisions of banks in vulnerable and other euro area countries
(2007 – H1 2014; percentage of total loans; median values)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2007 2008 2009 2010 2011 2012 2013 H1 2014
2007-11 averages
vulnerable countries other countries
Source: SNL Financial. Notes: Based on publicly available data on SBGs that report on a semi-annual basis. “Vulnerable countries” refer to Cyprus, Greece, Ireland, Italy, Portugal, Slovenia and Spain.
61 ECB
Financial Stability Review November 2014 61
3� Euro arEa F inancial
inst itutions
61
Divergent reported asset quality trends across banks continued into the first half of 2014 (see Chart 3.7), with banks in vulnerable countries experiencing a further deterioration, albeit at a slowing rate. This development was mainly linked to weak macroeconomic conditions in these countries, although some of the increase in non-performing loan (NPL) ratios may also have been related to a reclassification of restructured loans in anticipation of the future implementation of harmonised European Banking Authority (EBA) standards for NPLs.
Moreover, for the 130 banks subject to the comprehensive assessment, the AQR resulted in an increase of €136 billion, or 18%, in non-performing exposures (NPEs) with respect to figures reported for end-2013 (see also Box 4). By asset class, AQR-related increases in NPEs in absolute terms were largest for property-related and large corporate exposures, followed by large SMEs (see Chart 3.8).
Looking ahead, banks with a large stock of NPLs on their balance sheets still face the challenge of dealing with their problem assets, even if banks in some vulnerable countries have made some progress in writing off or disposing of bad loans (over and above the transfer of assets to bad banks/ asset management companies). Further significant progress in this area is all the more important as a slow resolution of NPLs could limit banks’ potential for new (profitable) lending.
Despite higher provisioning by a number of banks, coverage of impaired (non-performing) loans by reserves remained broadly stable in the first half of 2014, with the median coverage ratio for SBGs standing at 54% at end-June (see Chart 3.9). Loan loss reserves of large and complex banking groups (LCBGs) remained considerably higher than those of smaller SBGs, with the median value for the largest banks reaching 61% in mid-2014.
… on account of a further increase in non-performing loans…
… while coverage ratios remained broadly stable
Chart 3.7 Impaired loan ratios of euro area banks
(2007 – H1 2014; percentages; 10th and 90th percentiles and interquartile range distribution across SBGs)
0
4
8
12
16
20
24
28
0
4
8
12
16
20
24
28
2007 2008 2009 2010 2011 2012 2013 H1 2014
median for SBGs median for LCBGs
Source: SNL Financial. Note: Based on publicly available data on SBGs that report semi- annual financial statements.
Chart 3.8 Impact of the AqR on non-performing exposures by asset class
(EUR billions)
0
25
50
75
100
125
150
175
200
225
250
0
25
50
75
100
125
150
175
200
225
250
pre-AQR post-AQR
1 2 3 4 5 6 7 98 1 Property-related 2 Large SMEs 3 Large corporates 4 Residential property 5 Retail SMEs
6 Other retail 7 Shipping 8 Other non-retail 9 Project finance
Source: ECB. Note: The AQR-related changes reflect the impact of the application of the EBA’s simplified NPE approach and the credit file review.
62 ECB Financial Stability Review November 20146262
Overall, following the ECB’s comprehensive assessment exercise, long-lingering concerns about the asset and collateral valuation of significant banks in the euro area, NPL recognition as well as provisioning practices have largely dissipated. While the asset quality review, in the case of some banks, has led to higher provisions and reported NPLs in the short term, it should help strengthen confidence in the sector.
While banks’ subdued earnings performance continued to limit internal capital generation, a steady across-the-board increase in euro area banks’ risk-weighted capital ratios continued in the first half of 2014. Core Tier 1 (CT1) capital ratios increased only slightly in comparison with the levels at end-2013, and even decreased for LCBGs, given the one-off increase in risk- weighted assets following the implementation of the Capital Requirements Directive IV (CRD IV) (see left-hand panel of Chart 3.10). This affected both credit and counterparty risk-related and market risk-related risk-weighted assets due to, among other things, the new calculation of risk-weighted assets for the credit valuation adjustment (CVA) and the inclusion of former capital deduction items for higher risk securitisation positions.
Banks improved risk-weighted capital
ratios further…
Chart 3.10 Core Tier 1 (CT1)/common equity Tier 1 (CET1) capital ratios of euro area banks (2008 – H1 2014; percentages; 10th and 90th percentiles and interquartile range distribution across SBGs)
4
6
8
10
12
14
16
18
4
6
8
10
12
14
16
18
2008 2009 2010 2011 2012 2013 H1 2014
2013 H1 2014
CRD IVCT1/phased-in CET1 ratios
Fully loaded CET1 ratios
median for SBGs median for LCBGs
Source: SNL Financial. Note: Based on publicly available data on SBGs that report annual financial statements and on data on a sub-set of those banks that report on a semi-annual basis.
Chart 3.11 decomposition of changes in euro area banks’ aggregate Core Tier 1 capital ratio (2011 – H1 2014; percentages and percentage points)
6
7
8
9
10
11
12
13
14
6
7
8
9
10
11
12
13
14
9.6 0.6 0.1
0.6 10.9 0.4
1.1 -0.312.1
0.8 -0.3 -0.6
12.0
CT1 ratio 2011
CT1 ratio 2012
CT1 ratio 2013
CT1 ratio
H1 2014
change in capital change in total assets change in average risk weight
Sources: SNL Financial and ECB calculations. Notes: Based on publicly available data for a sample of 65 SBGs that report at least on a semi-annual basis. The increase in the average risk weight in the first half of 2014 was mostly due to the implementation of CRD IV.
Chart 3.9 Coverage ratios of euro area banks
(2008 – H1 2014; loan loss reserves as a percentage of impaired loans; 10th and 90th percentiles and interquartile range distribution across SBGs)
30
40
50
60
70
80
90
100
110
30
40
50
60
70
80
90
100
110
2008 2009 2010 2011 2012 2013 H1 2014
median for SBGs median for LCBGs
Source: SNL Financial. Note: Based on publicly available data on SBGs that report annual financial statements and on data on a sub-set of those banks that report at least on a semi-annual basis.
63 ECB
Financial Stability Review November 2014 63
3� Euro arEa F inancial
inst itutions
63
Based on a fully loaded common equity Tier 1 (CET1) definition, the median CET 1 ratio for banks participating in the comprehensive assessment exercise was 11.1% at 1 January 2014 (pre-AQR). Public disclosures by a sub-sample of SBGs suggest that fully loaded CET1 ratios may have improved further in the first six months of this year, with the median ratio for 45 reporting SBGs rising by nearly 80 basis points (see right-hand panel of Chart 3.10).
A decomposition of changes in banks’ aggregate risk-weighted capital ratio over the last two and a half years shows a shift towards capital increases in the first half of 2014 (see Chart 3.11). Recent increases in CET1 capital have mainly resulted from a further expansion of equity capital, which has amounted to over €50 billion for SBGs since end-2013. Furthermore, some banks completed or announced capital increases in the third quarter of 2014, partly in preparation for the comprehensive assessment to address capital shortfalls. By contrast, increasing risk-weighted assets contributed to lower capital ratios on account of both increasing average risk weights (due mainly to the implementation of CRD IV) and the reversal of asset deleveraging for a number of banks.
Thanks to a significant pick-up in banks’ equity issuance, euro area SBGs also continued to improve their balance sheet-based leverage ratios, with the median ratio of tangible common equity to tangible assets rising to 5.1% in mid-2014, from 4.5% at end-2013 (see Chart 3.12). However, the improvement of leverage ratios was more muted for LCBGs, with some of the largest banks remaining in the lowest quartile of the SBG distribution. In fact, despite recent improvements, large euro area banks continue to lag behind their global peers in terms of their leverage ratios when measured by adjusted tangible equity over adjusted tangible assets on a comparable basis (see Chart 3.13).
… mainly through capital increases…
… while large banks lag behind their global peers in improving leverage ratios
Chart 3.12 Euro area banks’ leverage ratios (tangible common equity to tangible assets)
(2007 – H1 2014; percentages; 10th and 90th percentiles and interquartile range distribution across SBGs)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
median for SBGs median for LCBGs
2007 2008 2009 2010 2011 2012 2013 H1 2014
Source: SNL Financial. Note: Based on publicly available data on SBGs, including LCBGs, that report annual financial statements and on data on a sub-set of those banks that report on a semi-annual basis.
Chart 3.13 CET1 ratio and adjusted leverage ratio for large banks in the euro area, other European countries and the united States (H1 2014; percentages)
2
3
4
5
6
7
8
9
2
3
4
5
6
7
8
9
8 9 10 11 12 13 14 15 16
euro area other European countries United States
x-axis: CET1 ratio y-axis: adjusted leverage ratio
Sources: Federal Deposit Insurance Corporation and SNL Financial. Notes: The adjusted leverage ratio is calculated as adjusted tangible common equity to adjusted tangible assets. Horizontal and vertical lines show median values.
64 ECB Financial Stability Review November 20146464
Box 4
ThE ECB’S COMpREhENSIVE ASSESSMENT EXERCISE
The results of the ECB’s comprehensive assessment, a thorough and unprecedented examination of 130 euro area banks, were published on 26 October 2014. This box presents the scope, main findings and conclusions of the comprehensive assessment exercise.
Scope of the comprehensive assessment
The exercise was undertaken as part of the preparations for the ECB’s assumption of supervisory responsibilities on 4 November 2014. The 130 banks participating in the exercise had total assets of €22 trillion at the end of 2013, accounting for more than 80% of total assets of the euro area banking system.
The comprehensive assessment exercise had two components:
• An asset quality review (AQR) of the assets held by banks at end-2013, in the course of which banks’ accounting models, policies and practices were checked on the basis of a common methodology1 and harmonised definitions across all participating countries.
• A constrained bottom-up stress test, in the course of which banks were requested to project the impact of hypothetical baseline and adverse macro-financial scenarios on their balance sheets and income statements.
The results of both components were joined together using a methodology that adjusted the stress-test results to reflect the findings of the AQR,2 a unique feature of the comprehensive assessment in comparison with similar stress-testing exercises.
Both components of the comprehensive assessment exercise were subject to a rigorous quality assurance process, comprising banks, national supervisors and the ECB, in order to ensure the appropriate degree of conservatism and a level playing field for all participating banks. The adverse macro-financial scenario for the stress test was designed by the European Systemic Risk Board. It captured the most relevant threats to the stability of the EU banking system that were identified in the spring of 2014, including an increase in global bond yields, a deterioration in credit quality, stalling policy reforms that lead to a re-emergence of sovereign risk and a lack of the balance sheet repair necessary to sustain market funding at affordable rates. Overall, these risks still remain relevant to date. The comprehensive assessment was a prudential exercise. By design, its scope did not include some of the macro-prudential risks related to, for example, the interconnectedness of participating banks or second-round effects arising from banks’ endogenous response to macro-financial stress.
Main findings
The comprehensive assessment concluded that most of the euro area banks would be resilient under the adverse macro-financial scenario in spite of a significant depletion of their capital.
1 See Asset quality review – Phase 2 Manual, ECB, March 2014. 2 See Comprehensive assessment stress test manual, ECB, August 2014.
65 ECB
Financial Stability Review November 2014 65
3� Euro arEa F inancial
inst itutions
65
The Common Equity Tier 1 (CET1) capital of the participating euro area banks would be reduced by €216 billion (see Chart A), €34 billion of which is due to the adjustment made in the course of the AQR, and €182 billion to the losses projected in the adverse scenario of the stress test.3 In addition, the minimum capital requirements would rise by €47 billion as a result of the increase in risk-weighted assets.
It was found that 25 euro area banks did not have sufficient capital to meet the CET1 capital ratio requirements specified for the comprehensive assessment exercise of 8% for the baseline and 5.5% for the adverse scenario. The total capital shortfall amounts to €24.6 billion prior to mitigating actions taken after the end-2013 reference date.
The AQR concluded that, under the common methodology and harmonised definitions, the non- performing exposures (NPEs) of participating banks should increase by €136 billion, or 18%, with respect to the stock of NPEs reported at the end of 2013. The review of impairment provisions related to both NPEs and other assets found that banks would mark down their assets by a further €43 billion on a pre-tax basis.
The baseline scenario of the stress test entailed an only slight increase in the CET1 capital ratio, reflecting the subdued operating profitability of participating banks. Under the adverse scenario, loan losses would nearly double with respect to the baseline case, and net interest income would contract by about 10%. A somewhat less material contribution to aggregate losses came from
3 See Aggregate report on the comprehensive assessment, ECB, October 2014.
Chart A Total impact of the adverse scenario of the comprehensive assessment on capital (EUR billions)
0
50
100
150
200
250
300
0
50
100
150
200
250
300
AQR after-tax impact
Stress-test capital impact
RWA effect on capital
requirements
Total CA impact
34
182
47
263
Source: ECB.
Chart B distribution of the CET1 capital ratios of banks participating in the comprehensive assessment (x-axis: percentile of the distribution, y-axis: CET1 capital as a percentage of risk-weighted assets)
0
5
10
15
20
25
30
0
5
10
15
20
25
30
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
end-2013 before CA
end-2013 AQR-adjusted end-2016 baseline
end-2016 adverse
Source: ECB. Note: Distribution censored at 5th and 95th percentile to remove outliers.
66 ECB Financial Stability Review November 20146666
BANkINg SECTOR OuTLOOk ANd RISkS
Outlook for the banking sector on the basis of market indicators Market-based indicators suggest an unchanged outlook for euro area banks over the last few months. In particular, the improving trend in euro area LCBGs’ price-to-book ratios that started around mid-2013 appears to have come to a halt in the second quarter of 2014 (see Chart 3.14). On the one hand, this mirrors similar developments for other global banks, including US LCBGs. On the other hand, the latest reading of this ratio suggests a weaker outlook for euro area banks compared with US peers, possibly reflecting concerns about the profit-generating capacity of euro area banks in an environment of low nominal growth.
Indeed, market expectations suggest a weak earnings outlook for euro area banks, with many banks expected to achieve returns below their cost of equity. In fact, while the latest earnings forecasts for euro area banks signal an improvement for 2015, market expectations of profitability remain at rather moderate levels (see Chart 7 of the Overview). Similarly, a frequently cited market-based measure of systemic banking sector stress suggests that, following the significant decline since mid-2013, systemic risk within euro area banks has stabilised at a low level (see Chart 3.15).
Market-based indicators point to
a stabilisation of banks’ outlook
a downward revaluation of trading assets and sovereign bonds, as well as from non-interest income. Overall, the capital ratio of the median bank would be reduced by around 4 percentage points, to about 8.3% (see Chart B).
Conclusions
The comprehensive assessment has caused euro area banks to take extensive action that has raised capital and reduced risk to mitigate potential capital shortfalls. In addition to capital measures taken prior to the end-2013 cut-off date of the comprehensive assessment exercise, banks continued to strengthen their balance sheets in 2014 (see Section 3.1 for more details). Twelve of the banks that were found to have a capital shortfall had already covered these shortfalls prior to the end of the exercise. The remaining 13 banks, with a combined capital shortfall of €9.5 billion, are implementing capital plans and are expected to reinforce their capital buffers. The capital actions should be completed within six months of the end of the assessment4 if shortfalls result from the AQR or the baseline scenario, or within nine months in case of shortfalls resulting from the adverse scenario.
From a forward-looking perspective, the results of the comprehensive assessment represent a major step towards balance sheet repair and strengthening the euro area banking sector, which in turn is key to enable the sector to support the economic recovery in the euro area. The results have shown that the vast majority of significant euro area banks are able to withstand a major adverse macro-financial shock without breaching the 5.5% CET1 ratio threshold. The findings of the ECB’s latest bank lending survey, which indicate that banks have begun to ease their lending standards, corroborate the conclusion reached in the comprehensive assessment that the importance of supply-side constraints in euro area credit markets has diminished.
4 These results include two banks which are implementing restructuring plans agreed with the European Commission, under which one bank would have a zero shortfall and one bank would have a small shortfall.
67 ECB
Financial Stability Review November 2014 67
3� Euro arEa F inancial
inst itutions
67
Credit risks emanating from banks’ loan books The level of credit risk in the loan book of the euro area banking sector remains elevated against the background of a tenuous economic recovery and legacy balance sheet issues that still represent a challenge in several countries. Bank lending has remained weak, particularly lending to the corporate sector, while lending to households has declined only slightly (see Chart 3.16). Although the effects of this are mitigated or offset by financial disintermediation in the case of larger firms with access to international bond markets, small and medium-sized firms that are reliant on bank-based finance continue to bear the negative consequences.
This challenge for the euro area banking sector is, however, part of a broader phenomenon of non- financial sector deleveraging in many advanced economies. Indeed, credit conditions across OECD economies have remained relatively weak by historical standards, with the global credit gap for OECD countries remaining well below its early warning threshold for costly asset price booms, despite some further improvement up to the first quarter of 2014 (see Chart 3.17).
These aggregate developments, however, conceal major differences in lending conditions across regions and countries as economic recoveries proceed at different speeds. Within the euro area, credit developments differed significantly across countries (see Chart S.1.14), with continued sharp declines in lending to non-financial corporations in more vulnerable countries contrasting with flat lending volumes in core countries, thereby raising concerns regarding a credit-less recovery.
Bank lending survey information suggests that much of the observed weakness in credit flows over the past year or so has been more closely linked to anaemic credit demand, with credit supply constraints playing a diminished role. In this vein, the results of the October 2014 euro area bank lending survey reveal some signs of easing credit standards for loans to both non-financial
Credit risk remains elevated…
… while credit standards show some signs of easing…
Chart 3.14 price-to-book ratios of large and complex banking groups in the euro area and the united States (Jan. 2007 – Nov. 2014; ratio)
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
2007 2008 2009 2010 2011 2012 2013 2014
difference between US and euro area LCBGs euro area LCBGs US LCBGs
Sources: Bloomberg, SNL and ECB calculations Note: Median values for LCBGs in the United States and the euro area.
Chart 3.15 Measure of euro area banking sector stress
(Jan. 2010 – Nov. 2014; probability; percentages)
0
5
10
15
20
25
30
0
5
10
15
20
25
30 Finalisation of the May 2014 FSR
2010 2011 2012 2013 2014
Sources: Bloomberg and ECB calculations. Notes: The measure contains the credit default swap implied probability of two or more of a sample of 15 banks defaulting simultaneously over a one-year horizon. See Box 8 in Financial Stability Review, ECB, June 2012, for further details.
68 ECB Financial Stability Review November 20146868
corporations (NFCs) and households. They also point to a recovery in credit demand not only by households, irrespective of the purpose of the loan, but also by NFCs, regardless of the firm size (see Chart 3.18).
While these signs could indicate a turning point in credit flows, they are closely tied to the pace of economic expansion and its impact on income and earnings risks for households and NFCs in a context of ongoing challenging balance sheet adjustment.
Notwithstanding the importance of demand conditions, legacy asset quality problems in vulnerable countries also weigh on new lending. At the country level, a continued expansion of NPLs is particularly visible in the most vulnerable euro area countries, although there are some tentative signs of a slowdown in new NPLs in some countries, or even of a reversal of worsening asset quality trends, most notably in Spain.
While a further expansion of NPLs is likely in countries with weak macroeconomic conditions in the coming quarters, there are some tentative
Chart 3.16 gdp growth and growth in credit to households and non-financial corporations in the euro area (Q1 2002 – Q2 2014; percentage change per annum)
-6
-4
-2
0
2
4
6
-10
-5
0
5
10
15
20
loans to households loans to non-financial corporations loans to the non-financial private sector real GDP growth
2002 2004 2006 2008 2010 2012
Sources: ECB and Eurostat. Note: The Q3 2014 GDP growth figure is based on the flash estimate by Eurostat.
Chart 3.17 global credit gap and optimal early warning threshold
(Q1 1980 – Q2 2014; percentages)
-6
-4
-2
0
2
4
6
-6
-4
-2
0
2
4
6
1980 1984 1988 1992 1996 2000 2004 2008 2012
Sources: ECB and ECB calculations. Note: Index for 18 OECD countries – see Alessi, L. and Detken, C., “Quasi real time early warning indicators for costly asset price boom/bust cycles: A role for global liquidity”, European Journal of Political Economy, Vol. 27(3), September 2011.
Chart 3.18 Credit standards and demand conditions in the non-financial corporation sector (Q1 2006 – Q4 2014; weighted net percentages)
-40
-20
0
20
40
60
80
-40
-20
0
20
40
60
80
2006 2007 2008 2009 2010 2011 2012 2013
large firms small and medium-sized enterprises
Source: ECB. Notes: The solid lines denote credit standards, while the dotted lines represent credit demand. Credit standards refer to the net percentage of banks contributing to a tightening of credit standards, while credit demand indicates the net percentage of banks reporting a positive contribution to demand.
69 ECB
Financial Stability Review November 2014 69
3� Euro arEa F inancial
inst itutions
69
signs that the pace of credit quality deterioration could slow in an increasing number of countries as the economic recovery gains momentum. In fact, the combined quarterly change in corporate NPLs in three of the vulnerable countries where sectoral NPL data are available (Spain, Italy and Portugal) shows a decline in the first two quarters of 2014, although it was driven mainly by developments in Spain (see Chart 3.19). At the same time, there is little sign of a pick-up in loan write-offs, suggesting that banks in these countries still need to make further progress in resolving the issue of NPLs.
The comprehensive assessment exercise accelerated the process of bank balance sheet repair, ensuring prudent asset valuation and stricter loan loss recognition, as well as providing more transparency on asset quality. Complementing this, the cleaning-up of bank balance sheets should be fostered at the national level by removing legal and judicial obstacles to timely NPL resolution (see Chart 3.20).
Finally, for some euro area banks, credit risks also emanate from their significant cross-border exposures. Indeed, some SBGs remain highly exposed to emerging market economies (EMEs), based on the ratios of their exposure at default (EAD) to common equity, in particular to countries in “developing Europe”.2 A few banks with exposures to the most vulnerable EMEs (including Russia and Ukraine) have incurred higher credit losses in the first half of 2014, and face the risk of asset quality deterioration in the event of geopolitical tensions persisting for longer and/or the macroeconomic environment in some EMEs deteriorating further. The SBGs exposed most to those EMEs could face higher loan losses on these portfolios in the period ahead.
Funding liquidity risk Market-based bank funding conditions remained very favourable, with average spreads on bank debt stabilising below the levels seen in early 2010, i.e. before the start of the sovereign debt crisis. Spreads on different debt instruments have diverged somewhat since mid-2014, with a further
2 See Financial Stability Review, ECB, May 2014.
…with further progress needed in the disposal of NPLs
Funding conditions remained very favourable…
Chart 3.19 quarterly change in non-performing loans and loan write-offs in Spain, Italy and portugal (Q1 2010 – Q2 2014; EUR billions)
-15
-10
-5
0
5
10
15
20
25
30
35
40 Corporates Households
-15
-10
-5
0
5
10
15
20
25
30
35
40
2010 2011 2012 2013 2014 2010 2011 2012 2013 2014
write-offs NPLs
Source: National central banks.
Chart 3.20 Length and cost of contract enforcement and non-performing loan ratios across the euro area (2014)
0
5
10
15
20
25
30
35
40
0
5
10
15
20
25
30
35
40
y-axis: Cost of contract enforcement (percentage of claim) x-axis: Length of contract enforcement (days)
AT ES
FR
IE
LV NL
PT
SK
GR
DE
IT
BE
SI
EE
CY
LU
FI
MT long & high cost short & high cost
long & low costshort & low cost
200 400 600 800 1,000 1,200 1,400
Sources: World Bank Doing Business 2014 and ECB. Note: The size of the bubble represents the NPL ratio at year- end 2013.
70 ECB Financial Stability Review November 20147070
tightening of those on covered bonds and, to a lesser extent, senior unsecured debt contrasting with some widening of spreads on subordinated debt (see Chart 3.21). Fragmentation in the pricing of bank debt declined further, as reflected, for instance, in the narrowing differential between spreads on covered bonds issued by banks in vulnerable and other countries, which recently also benefited from the ECB’s announcement of a third covered bond purchase programme (CBPP3) (see Chart 3.22). Market-based funding remained widely available, although debt issuance by euro area banks in recent months was below last year’s levels, including for banks in vulnerable countries, on the back of increased volatility in credit markets.
Debt issuance patterns reflected banks’ efforts to adapt their debt and capital structures to new regulatory requirements, as well as continued strong investor demand for higher-yielding bank debt. As a result, subordinated debt issuance has seen the most significant increase in the year to date, including both additional Tier 1 and Tier 2 instruments (see Chart 3.23), as banks continued to build up their subordinated debt buffers in preparation of meeting the CRR/CRD IV total capital/Tier 1 capital ratio, as well as minimum bail-in requirements. Despite a recent slowdown, issuance of junior
Chart 3.21 Spreads on banks’ senior debt, subordinated debt and covered bonds
(Jan. 2010 – Nov. 2014; basis points)
0
100
200
300
400
500
600
700
0
50
100
150
200
250
300
350
Previous cut-off date
2010 2011 2012 2013 2014
iBoxx EUR banks senior (left-hand scale) iBoxx EUR covered (left-hand scale) iBoxx EUR non-financial senior (left-hand scale) iBoxx EUR banks subordinated (right-hand scale)
Sources: ECB and Markit.
Chart 3.22 Covered bond spreads in vulnerable and other euro area countries
(Jan. 2010 – Nov. 2014; basis points)
-100
0
100
200
300
400
500
600
700
800
900
1,000
-100
0
100
200
300
400
500
600
700
800
900
1,000
Previous cut-off date
CBPP3 announcement
France Germany Spain Italy Portugal
2010 2011 2012 2013 2014
Sources: ECB and Markit.
Chart 3.23 Issuance of subordinated debt by euro area banks
(Q1 2009 – Q3 2014; EUR billions)
0
2
4
6
8
10
12
14
16
18
20
22
0
2
4
6
8
10
12
14
16
18
20
22
2009 2010 2011 2012 2013 2014
Tier 1 Tier 2
Source: Dealogic. Note: Excludes retained deals and government-guaranteed issuance.
71 ECB
Financial Stability Review November 2014 71
3� Euro arEa F inancial
inst itutions
71
debt by euro area banks in the first nine months of 2014 more than tripled in comparison with a year earlier. Issuance activity in the senior unsecured debt market has slowed since mid-2014, partly also reflecting reduced funding needs following robust issuance in the first half of 2014 (see Chart 3.24). Meanwhile, covered bond issuance up to October remained slightly below last year’s level, although it started to show some signs of a pick-up in November, also thanks to the implementation of the ECB’s CBPP3.
At the same time, issuance of asset-backed securities (ABSs) by euro area banks remains moderate. In fact, in 2014 thus far, euro area banks have placed less than €30 billion of ABSs with investors, around 30% less than a year earlier. Going forward, however, the ABS market and euro area banks’ off-balance-sheet financing are likely to benefit from the ECB’s ABS purchase programme.
Turning to structural changes in bank funding, deposit flows slowed in the first nine months of 2014, with further negative net flows of wholesale funding – consistent with continued deleveraging – while the share of customer deposits increased further (see Chart 3.25). As a result, the median ratio of customer deposits to total liabilities for SBGs reached 53% in mid-2014, up from 46% at the end of 2012 (see Chart 3.26). Providing yet another sign of declining euro area fragmentation, banks in both vulnerable and other countries benefited from a shift towards deposit funding (as a share of total funding), even if this was due more to shrinking reliance on other funding sources such as wholesale and Eurosystem funding than to deposit growth.
Similarly, banks’ loan-to-deposit ratios (a proxy of their reliance on wholesale funding) continued to decline gradually in the first half of 2014, with the median ratio for SBGs reaching 115% at the end of June, representing a significant fall from its pre-crisis peak of 143% in 2007. Nevertheless, the dispersion of loan-to-deposit ratios remains wide, and some institutions continue to be dependent
… and the shift towards deposit funding continued
Chart 3.24 Cumulative yearly issuance of senior unsecured debt and covered bonds by euro area banks (Jan. 2011 – Nov. 2014; EUR billions)
0 20 40 60 80
100 120 140 160 180 200 220 240 260
0 20 40 60 80 100 120 140 160 180 200 220 240 260
senior unsecured covered bonds
Jan. Mar. May July Sep. Nov. Jan. Mar. May July Sep. Nov.
2011 2012 2013 2014
Source: Dealogic. Notes: Excludes retained deals and government-guaranteed issuance. November 2014 includes data up to the middle of the month.
Chart 3.25 Twelve-month flows in the main liabilities of the euro area banking sector
(Jan. 2010 – Sep. 2014; 12-month flows; EUR billions)
-2,000
-1,500
-1,000
-500
500
0
1,000
1,500
2,000
-2,000
-1,500
-1,000
-500
500
0
1,000
1,500
2,000
2010 2011 2012 2013 2014
private sector deposits (euro area)
wholesale funding (euro area)
capital and reserves
foreign deposits
Eurosystem funding
total assets (excl. remaining assets)
Source: ECB. Notes: Total assets are adjusted for remaining assets, which consist largely of derivatives. Wholesale funding comprises interbank liabilities and debt securities.
72 ECB Financial Stability Review November 20147272
on wholesale funding. These banks need to make further adjustments in their funding profiles, with some business models (e.g. those of some German Landesbanken) facing particular challenges in this regard.
Looking at funding challenges beyond the short term, banks’ changing debt/capital structures – characterised by the rising share of loss- absorbing and bail-inable instruments – should contribute to a safer system and more efficient resolution mechanisms. However, these changes also create challenges of their own. The fast-growing market for contingent convertible capital instruments (CoCos) remains untested, with no investor loss event (trigger or coupon deferral) having occurred thus far, creating some uncertainty as to whether such an event would be seen as idiosyncratic or could affect the asset class more profoundly. This highlights the need for investors to gain a better understanding of how different features of CoCos impact on the risk profile of these investments (see Box 5).
Regarding potential implications of bail-ins, the subordinated debt market remained resilient to recent bail-ins (Banco Espirito Santo and Hypo Alpe Adria), although this may also reflect the relatively small size of the bailed-in debt involved. Looking ahead, however, as some countries are planning to bring forward senior debt bail-in rules as of 2015, rating agencies have indicated that they would review ratings on the basis of how the bail-in legislation is expected to affect government support. This could cause rating agencies to reduce or eliminate systemic support in the ratings, which would put pressure on senior debt ratings, in particular for those banks that currently enjoy a multi-notch uplift through implied government support.
Chart 3.26 Share of customer deposits in total liabilities for euro area banks
(2007 – H1 2014; percentage of total liabilities; 10th and 90th percentiles and interquartile range distribution across SBGs)
0
10
20
30
40
50
60
70
80
0
10
20
30
40
50
60
70
80
2007 2008 2009 2010 2011 2012 2013 H1 2014
Source: SNL Financial. Note: Based on publicly available data on SBGs that report annual financial statements and on data on a sub-set of those banks that report on a semi-annual basis.
Box 5
dO CONTINgENT CONVERTIBLE CApITAL INSTRuMENTS AFFECT ThE RISk pERCEpTIONS OF SENIOR dEBT hOLdERS?
Contingent convertible capital instruments or bonds (CoCos) are hybrid instruments that are automatically transformed into equity or are written off in the event of a capital shortfall. CoCos thus contain built-in mechanisms for absorbing losses when trigger points are reached. CoCos are flexible instruments that are able to boost regulatory CET1 capital ratios when necessary, while preserving the respective debt status if the pre-specified trigger level is not reached. They have grown in popularity in recent years, not least on account of their state-contingent nature, their distinct accounting treatment and the fact that they combine elements of debt and equity.
73 ECB
Financial Stability Review November 2014 73
3� Euro arEa F inancial
inst itutions
73
The attractive features of CoCo instruments for issuers and investors have led to marked growth in this market. But as the importance of this nascent market for the structure of banks’ liabilities increases, the risks involved may rise as well. The market has experienced dramatic growth over the last few years, with an increasing share of write-down instruments.1 The supply of such hybrids appears closely related to a need of banks to increase their capital ratios in line with the new Basel III standards. On the demand side, the higher coupons paid to investors in CoCos in comparison with those of many other financial assets have proven to be very attractive in the current low-yield environment (see Chart A). The market is quite important in Europe, which has seen greater use of CoCos than the rest of the world (see Chart B).
One factor obfuscating an aggregate view of risk related to the growing market for these instruments is that contingent convertible bonds are complex in structure and, as a result, no two such hybrid instruments are identical. That said, the underlying loss-absorption mechanism is a key channel through which risk may arise, as this conduit for risk-taking incentives for holders of equity can create externalities.2 The theoretical literature on hybrid debt is closely related to whether such instruments contain “write-down” or “conversion” clauses. Since write-down instruments imply that losses at the trigger point are first borne by CoCo investors, this could increase the risk-taking incentives for bank owners. By contrast, instruments with a conversion- to-equity clause imply that, if triggered, current equity holders suffer from the dilution of their shares. This aligns the interests of CoCo investors and shareholders, incentivising the latter to limit risk-taking in order to avoid triggering the CoCos. Hilscher and Raviv analyse the stabilising effect of CoCos on the issuing bank, conditional on the features of the instrument, concluding that a high conversion ratio significantly reduces the risk-taking incentives of stock-holders.3 Berg and Kaserer show that a significant reliance on CoCos can lead to more
1 See also Box 9 in Financial Stability Review, ECB, May 2014. 2 It should be noted that shareholders may be reluctant to allow capital levels to reach the trigger point as that could lead to restrictions
on dividend payments. 3 See Hilscher, J. and Raviv, A., “Bank stability and market discipline: The effect of contingent capital on risk taking and default
probability”, Journal of Corporate Finance, 2014.
Chart A Contingent convertible bond issuance: write-down versus conversion
(July 2009 – Aug. 2014; EUR billions)
5
0
10
15
20
25
30
35
40
45
0
5
10
15
20
25
30
35
40
45
2009 2010 2011 2012 2013 2014
conversion write-down
Sources: Dealogic, Bloomberg and ECB calculations.
Chart B Cumulated amounts of contingent convertible bonds issued, broken down by region (Aug. 2014; EUR billions)
0
10
20
30
40
50
60
70
0
10
20
30
40
50
60
70
Euro area Non-euro area Europe
Rest of the world
conversion write-down
Sources: Dealogic, Bloomberg and ECB calculations.
74 ECB Financial Stability Review November 20147474
risk-taking, especially when capital ratios approach the trigger level.4 Such behaviour could be amplified further by write-down clauses, as they imply only losses for holders when the trigger is reached. A significant level of dilution can hence help align the incentives of shareholders and those of the bondholders and reduce endogenous risk. These considerations raise the question as to whether different CoCo features create incentives for risk-taking by issuing banks.
An analysis of the effect of CoCo issuance on the pricing of senior unsecured debt (five-year credit default swap (CDS) spreads) suggests that the risk perception of senior bond holders depends crucially on the risk-taking incentives that CoCos may create for equity holders. The sample covers quarterly panel data for the period from the third quarter of 2009 to the first quarter of 2014 and for 60 banks (20 CoCo issuers and 40 non-issuers) from 19 countries.5 First, the analysis aims at disentangling the effect of conversion/write-down CoCo dummies on CDS spreads. In a second step, the explanatory power of the quantity of CoCos as a percentage of equity is analysed. Since the control group is represented by non-issuers, the coefficients in the second column of the table below represent the effect of adding one more percentage point of CoCos relative to equity.
The point estimates in the first column of the table below show that the effect of the write- down dummy is positive and significant. Hence, a bank with write-down CoCos is perceived by senior bond holders to be riskier when compared with non-issuers, and this is reflected in a significantly larger increase in CDS spreads. Moving to the second column of the table of results, the effect of write-down instruments as a proportion of total equity is also positive. This implies that higher costs for protection against default are associated with a stronger reliance on write-down instruments in the capital structure. These results are quite illustrative, as empirical work on CoCo instruments and their impact on risk perceptions and incentives has remained limited, despite the recent surge in theoretical research.
Such results are consistent with the notion that issuing CoCos with a write-down clause appears to increase the perceived risk of a bank. On the other hand, the results suggest that holding instruments that are converted to equity if triggered has a negative impact on the change in bank CDS spreads, although that impact is insignificant in terms of quantities. As the prevalence of these instruments increases, a better understanding of their characteristics and behavioural implications in stressed market conditions is crucial for understanding their prospective impact on financial stability.
4 See Berg, T. and Kaserer, C., forthcoming. 5 For further details on the empirical analysis, see Bicu, A., Stolz, S. and Wedow, M., “Layer cake: Risk incentive effects of CoCos”.
Impact of contingent convertible bonds on the change in banks’ CdS spreads
Variables ΔCDS ΔCDS
Conversion dummy -31.62* Write-down dummy 28.21*** Conversion quantity in total equity -2.97 Write-down quantity in total equity 2.83** R2 0.471 0.470
Notes: The analysis is performed using a panel fixed effects estimator, with bank individual effects, quarter dummies and bank-clustered standard errors. The regressions are augmented with bank balance sheet variables (bank balance sheet and regulatory indicators, size) and country risk (sovereign CDS spread), but their effect is not shown. ***, **, * indicates significance at the 1, 5 and 10% levels.
75 ECB
Financial Stability Review November 2014 75
3� Euro arEa F inancial
inst itutions
75
Market-related risks Banks’ interest rate risk has remained material against the background of both still high sovereign exposures in some parts of the euro area and the continued flattening of the euro area yield curve, which has adverse implications for the profits banks garner from maturity transformation activities (see above). Since the finalisation of the May 2014 FSR, there has been a further substantial decline in sovereign yields, particularly at the long end of the yield curve (see Chart 3.27), with continued yield compression also extending to bonds of lower-rated sovereigns. Against this backdrop, euro area banks remain vulnerable to a potential reassessment of risk premia in global markets, in particular through their direct exposures to higher-yielding debt instruments, via possible valuation losses on their sovereign bond exposures, depending on the duration of these portfolios and on the extent to which their positions are hedged.
In this regard, data on the holdings of government debt by monetary financial institutions (MFIs) in the euro area show a continuation of home bias in sovereign debt holdings for banks in most euro area countries (see Chart 3.28). Despite recent declines, sovereign bond holdings as a percentage of total assets remain well above pre-crisis levels in some countries. Furthermore, some banks attempted to offset declining yields by extending the duration of their bond portfolios. As confirmed by bank-level data from the comprehensive assessment exercise, mid-sized SBGs have higher exposures, on average, to lower-rated sovereigns in their respective countries, leaving them more vulnerable than larger banks to adverse yield movements.
Interest rate risk remains material…
... with some banks still exposed to lower-rated sovereign debt…
Chart 3.27 developments in the euro area yield curve
(percentages)
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
-0.2
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
1 month
1.5 years
3 years
4.5 years
6 years
7.5 years
10 years
June 2012 FSR
December 2012 FSR November 2013 FSR
May 2013 FSR May 2014 FSR
14 November 2014
Sources: ECB and Thomson Reuters.
Chart 3.28 MFIs’ holdings of sovereign debt, broken down by country
(Sep. 2013 – Sep. 2014; percentage of total assets; annual growth rate)
-20
-15
-10
-5
0
5
10
15
20
25
-7.5
-5.0
-2.5
0.0
2.5
5.0
7.5
10.0
12.5
holdings of other Member States’ government debt as a share of total assets holdings of domestic government debt as a share of total assets annual growth rate of holdings of euro area government debt (right-hand scale)
4 Italy 5 Other vulnerable countries 6 Other euro area countries
1 Germany 2 France 3 Spain
1 2 3 4 5 6
Source: ECB. Note: “Other vulnerable countries” refer to Cyprus, Greece, Ireland, Portugal and Slovenia.
76 ECB Financial Stability Review November 20147676
With respect to other fixed-income exposures, euro area MFIs’ holdings of euro area non- financial corporate debt were stable in the first two quarters of 2014, with the share of these securities in banks’ balance sheets remaining limited at around 0.5%. This suggests that the direct impact of a sharp adjustment of risk premia on euro area corporate bonds would be contained at the aggregate level. However, some banks with material exposures to high-yield or EME corporate bonds could be more negatively affected in such a scenario.
Finally, euro area banks’ exposure to equity markets remained, on average, broadly unchanged in the first half of 2014, but with significant heterogeneity across banks of different sizes (see Chart 3.29). In particular, LCBGs have increased their exposure to this asset class since end-2012. This could be related in part to the fact that low equity market volatility tends to compress backward- looking risk measures, such as the value at risk (VaR), thereby inducing some banks to increase their exposure.
3.2 ThE EuRO AREA INSuRANCE SECTOR: RESILIENCE AMId CONTINuEd hEAdWINdS
FINANCIAL CONdITION OF LARgE INSuRERS 3
The performance of large euro area insurers remained stable, despite headwinds from a low interest rate environment and only moderate economic growth. Overall, the sector exhibited modest growth in premiums written during the second and third quarters of 2014 (see Chart S.3.22 in the Statistical Annex), although median growth was relatively muted in the life insurance sub-sector during the first half of the year (see Chart 3.30). Life insurers appear to be particularly affected by the low interest rate environment – especially those offering guaranteed products. Nevertheless, this segment appears to be weathering the headwinds, given continued significant cost savings and an optimised product mix. Overall, combined ratios (i.e. incurred losses and expenses as a proportion of premiums earned) were somewhat
3 The analysis is based on a varying sample of 21 listed insurers and reinsurers with total combined assets of about €4.9 trillion in 2013, which represent around 80% of the assets in the euro area insurance sector. Quarterly data were only available for a sub-sample of these insurers.
… while corporate bond exposures remain limited
Insurers resilient so far…
Chart 3.29 Euro area banks’ holdings of equity instruments
(2007 – H1 2014; percentage of total assets; 10th and 90th percentiles and interquartile range distribution across SBGs)
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
2007 2008 2009 2010 2011 2012 2013 H1 2014
median for SBGs median for LCBGs
Source: SNL Financial.
Chart 3.30 gross-premium-written growth for a sample of large euro area insurers
(2011 – Q3 2014; percentages; 10th and 90th percentiles, interquartile distribution and median)
-25
-20
-15
-10
-5
0
5
10
15
20
25
-25
-20
-15
-10
-5
0
5
10
15
20
25
-25
-20
-15
-10
-5
0
5
10
15
20
25
-25
-20
-15
-10
-5
0
5
10
15
20
25
2011 2013 2014 Q1 Q2 Q3
2011 2013 2014 Q1 Q2 Q3
Life insurance Non-life insurance
Sources: Bloomberg, individual institutions’ financial reports and ECB calculations.
77 ECB
Financial Stability Review November 2014 77
3� Euro arEa F inancial
inst itutions
77
higher in the second quarter of 2014, impacted by higher loss ratios (see Chart S.3.23). Still solid investment income and the absence of any major global natural catastrophe have both been crucial factors underpinning the stable profitability of large euro area insurers (see Chart S.3.21). Moreover, the heterogeneity of investment income performance, which previously had exhibited a strong cross-country dimension, seems to have subsided considerably, mainly on account of a convergence of the yields on benchmark euro area government bonds.
The capital base of large euro area insurers remained stable at comfortable levels (see Chart 3.31), supported by falling yields on government bonds, which form the bulk of insurers’ assets. While this signals an average underlying resilience of these large insurers, regulatory factors may be playing a role as well, since fair value accounting of assets, but not of liabilities, as is applied in most jurisdictions, implies accounting benefits from the decline in most sovereign yields.4
INSuRANCE SECTOR OuTLOOk: MARkET INdICATORS ANd ANALYSTS’ VIEWS Market-based indicators suggest a relatively stable outlook for the euro area insurance sector next year. The share prices of the most important euro area insurance companies showed some volatility in the summer and, most notably, in September when, following a change in management at PIMCO, turbulence relating to the share price of Allianz created some volatility in fixed-income markets, in which insurers are very active players (see Chart S.3.30). In addition, the downward trend in credit default swap (CDS) spreads across large insurers stabilised somewhat at relatively low levels in the last months (see Chart S.3.28).
Analysts also expect euro area insurance earnings to remain relatively stable in 2014 and 2015, although subdued economic growth may pose additional challenges to profitability (see Chart 3.32). Given historically
4 Upon the implementation of Solvency II in 2016, valuation of assets and liabilities will shift to a market-based approach.
… despite low yields in all euro area jurisdictions
Market indicators show some volatility
Analysts expect stable earnings
Chart 3.31 Capital positions of large euro area insurers
(2005 – H1 2014; percentage of total assets; 10th and 90th percentiles, interquartile distribution and median)
5
10
15
20
25
30
35
40
5
10
15
20
25
30
35
40
2005 2007 2009 2011 2013 H1
20142013 H2 H1
Sources: Bloomberg, individual institutions’ financial reports and ECB calculations. Note: Capital is the sum of borrowings, preferred equity, minority interests, policyholders’ equity and total common equity.
Chart 3.32 Earnings per share of selected large euro area insurers and real gdp growth
(Q1 2002 – 2015)
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
-6
-5
-4
-3
-2
-1
0
1
2
3
4
5
2002 2004 2006 2008 2010 2012 2014
actual earnings per share (EUR) real GDP growth (percentage change per annum) earnings per share forecast for 2014 and 2015 (EUR) real GDP growth forecast for 2014 and 2015 (percentage change per annum)
Sources: ECB, Thomson Reuters and ECB calculations.
78 ECB Financial Stability Review November 20147878
low interest rates in all jurisdictions, there is considerable pressure on insurance companies to seek higher returns on their investments. At the same time, analysts generally expect most euro area insurers to be able to meet their guarantees for a prolonged period, even in the case of low investment returns, as other sources of income from new business should be supported by product innovations and a temporary revival of demand for traditional life insurance products in core markets. In addition, cost-cutting appears to be a common trend throughout the industry.
Analysts have also noted an increase, at an industry level, in risk appetite in terms of longer duration and increasing demand for corporate debt within fixed-income portfolios. Thus far, this appears to be still relatively contained for large euro area insurers, as aggregate volumes of high-yield bonds and other more risky investments remain stable. High levels of capitalisation, in particular in the reinsurance sector, have increased expectations of higher dividends.
Despite the generally stable outlook for the euro area insurance sector, challenges persist in the months ahead. In the reinsurance sub-sector, an abundant supply and stagnant demand are expected to fuel further declines in prices in 2015, making it challenging for reinsurers to earn their cost of capital. In addition, analysts expect the low-yield environment to have a negative impact on investment income, hampering profitability throughout the insurance sector in the euro area and testing the long-run viability of some life insurers’ business models. Finally, individual insurers in some euro area jurisdictions may be confronted with higher than expected litigation costs.
INVESTMENT RISk Investment activity remains highly concentrated on traditional fixed-income segments, such as government and corporate bond markets (see Chart S.3.25). However, given the crucial role that investment income plays in insurers’ business models and the expected persistence of currently low yields in fixed-income markets, insurance companies have been seeking higher returns in alternative investments. Some signs of portfolio adjustments were visible in some large euro area insurers, with investment in equities increasing since 2013 and investment in structured credit and commercial property declining slightly over the same period. In addition, although fixed-income portfolios are clearly dominated by highly rated bonds, there was a very slight increase in the proportion of higher-yield bonds (see Chart 3.33).
In terms of geographical orientation, long- term investors have further increased their exposure towards emerging economies’ bond markets. Emerging market debt accounts for an increasing share of the return-seeking portfolios of both life and non-life insurers. Although the proportion of emerging market bonds in the fixed-income portfolios of most euro area large insurers is currently relatively low, sizeable future increases would create concerns about currency risk on their books. On the one hand,
Sector faces multiple challenges
Emerging market debt increasing
Chart 3.33 Bond investments of selected large euro area insurers split by rating categories (percentage of total bond investments; weighted averages)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
35
2012 2013 H1 2014
AAA AA A BBB Non- investment
grade
Unrated
Sources: Company reports, JPMorgan Cazenove and ECB calculations. Note: Based on the available data for 12 large euro area insurers.
79 ECB
Financial Stability Review November 2014 79
3� Euro arEa F inancial
inst itutions
79
with Solvency II, 25% of capital will be required to be held against assets held in any currency other than that used to prepare the insurer’s financial statements. On the other hand, hedging currency risk – for instance, by means of a deliverable forward contract – is also expensive, which might act as a pecuniary deterrent for insurance companies.
An investment uncertainty map signals stress in several markets (see Chart 3.34). With government bond yields reaching historical lows in almost all jurisdictions during the summer and investors expecting rates to remain low, challenges to economic solvency and investment income persist. If sustained, this environment – together with weak economic growth – could potentially impact profitability further, eroding capital positions, in particular of small and medium-sized life insurers in jurisdictions where fixed guarantees are offered to policyholders. Naturally, given the weight of fixed-income securities in insurers’ assets, a major concern remains the potential for a sudden rise of risk-free rates.5 On the one hand, in the medium and long term, the impact of a rise is deemed to be mainly positive in terms of higher investment income, economic solvency and embedded value. Life insurers would benefit most, given the longer duration of their liabilities relative to assets. On the other hand, in the short term, the impact thereof on stated equity and price-to-book ratios may also be a concern, leading to an abrupt temporary increase in market volatility with a potential short-term risk to share prices. This could affect insurers with short-duration assets, particularly
5 The impact of rising or falling interest rates is only relevant if there is a duration mismatch between assets and liabilities. If an insurer is short duration (i.e. lower asset duration than the liability duration), a rising interest rate is beneficial as the fall in asset value is lower than the fall in liability value, i.e. the capital position improves. This is normally the case for the majority of the life insurers. Very rarely, insurers are long duration (i.e. asset duration higher than the liability duration) although technically non-life insurers could be so.
Widespread low yields are a real threat…
… although the industry is prepared for sudden-rise scenarios
Chart 3.34 Investment uncertainty map for the euro area
(Jan. 1999 – Oct. 2014)
Government bond markets
Corporate bond markets
Stock markets
Structured credit Commercial property markets
2011 20132005 2007 20091999 2001 2003
greater than 0.9
between 0.7 and 0.8 between 0.8 and 0.9
below 0.7 data not available
Sources: ECB, Bloomberg, JPMorgan Chase & Co., Moody’s, Jones Lang LaSalle and ECB calculations. Notes: Each indicator is compared with its “worst” level since January 1999. “Government bond markets” represent the euro area ten-year government bond yield and the option- implied volatility of German ten-year government bond yields, “Corporate bond markets” A-rated corporate bond spreads and speculative-grade corporate default rates, “Stock markets” the level and the price/earnings ratio of the Dow Jones EURO STOXX 50 index, “Structured credit” the spreads of residential and commercial mortgage-backed securities, and “Commercial property markets” commercial property values and value-to-rent ratios.
Chart 3.35 portfolio transactions of euro area insurance companies
(H1 2008 – H1 2014; EUR billions)
-20
0
20
40
60
80
100
-20
0
20
40
60
80
100
H1 H1 H1 H1 H1 H1 H1H2 H2 H2 H2 H2 H2 2008 2009 2010 2011 2012 2013 2014
debt securities issued by governments
debt securities issued by MFIs debt securities issued by NFCs
debt securities issued by OFIs
Source: ECB. Notes: Data availability varies across countries and investment categories. “MFIs” refer to monetary financial institutions, “NFCs” to non-financial corporations and “OFIs” to other financial intermediaries. Counterparties reside in the euro area.
80 ECB Financial Stability Review November 20148080
if they offer attractive dividends. In addition, non-life insurers might be tempted to use higher investment incomes to cut prices and reduce underwriting margins. At a global level, the desire to remain flexible in the face of possibly rising interest rates is inducing more insurance companies to consider absolute-return investment approaches, ahead of other approaches, such as book yield, relative return and liability matching.6
Euro area insurers have increased their holdings of government bonds in almost all jurisdictions (see Chart 3.35) – in some cases with a high domestic sovereign focus – according to transactional data, which exclude valuation changes. Holdings of debt issued by euro area corporates appears to also be on the rise. At the same time, insurers in the euro area have decreased their holdings of debt issued by euro area monetary financial institutions, although some analysts expect this trend to reverse in the near future.
While the insurance sector is increasing its non-traditional activities in an endeavour to boost income, their use remains limited thus far, on aggregate. Although evidence of such activities (mainly sales of credit risk protection and direct lending to counterparties) exists, levels at an aggregate euro area level remain low, and even declined slightly within the euro area in the first half of 2014.
The use of captives7 by insurance companies raises concerns about capital arbitrage and financial soundness. The sharp increase in captive insurance entities (in particular, in the United States) and their weak disclosure obligations have recently gained the attention of the international financial stability community. Most concerns come from the use of captive life reinsurers for life insurance reserve financing and the use of inter-company loans, activities sometimes called “shadow insurance”. Although currently only limited signs of such activities exist in the euro area, an expected increase in formations of captives in Europe (which currently accounts for an estimated 28% of all captives worldwide) warrants close monitoring.
uNdERWRITINg RISk Expectations of depressed top-line growth in life insurance markets in the future and a continued softening of reinsurance pricing pose challenges to the reinsurance and life insurance business models. Both life and non-life companies have further increased their amounts of premiums written in emerging markets. Such expansion brings diversification benefits in markets that are highly profitable and relatively underpenetrated at the moment. However, new challenges emerge in terms of risk management, currency risk, new product developments and group supervision.
The reinsurance industry recorded manageable and below-average natural catastrophe losses in 2014 (see Chart 3.36). However, Europe was the only region to have above-average insured losses in the first half of the year. Severe thunderstorms and hail in early June caused significant damage in France, Germany and the Netherlands, with total insured losses estimated at USD 2.5 billion. In addition, aviation disasters in 2014 thus far could cost the insurance industry as much as USD 1.5 billion.
6 A relative-return approach rates the performance of the fixed-income portfolio relative to that of a public benchmark. An absolute or total- return approach considers performance relative to zero-risk assets. Relative return gives asset managers a yield target above the market average, but this may not be enough to provide the cash-flow matching and yield that insurers are seeking.
7 “Captives” are insurance companies established with the objective of financing specific risks borne by their respective owner, affiliated businesses or a designated set of companies. In the case of non-financial companies, use of captives is motivated by sound risk management and a cost-efficient pooling of risks. However, the use of captives by insurance companies might be driven by the ability to effectively move assets (and their associated liabilities) off the balance sheet in order to reduce regulatory capital requirements.
Slow portfolio adjustment…
… with limited non- traditional activities
thus far…
… despite the increase in captive
activities by insurance companies
Increased exposure to emerging markets
Manageable insured catastrophe losses…
81 ECB
Financial Stability Review November 2014 81
3� Euro arEa F inancial
inst itutions
81
Over the past two years, the reinsurance industry has seen an inflow of approximately USD 20 billion of new capital from an ever-broadening investor base,8 precipitating the most marked change to the sector’s capital structure in recent times. Capital has entered the market through investment in insurance-linked securities (mainly catastrophe bonds), funds and “sidecars”, as well as through the formation of hedge fund-related reinsurance companies and collateralised reinsurance vehicles.
These investors have been drawn to (re)insurance on account of the advantages the sector offers in terms of being a non-correlating asset class, as well as the absence of attractive investments given the current level of interest rates. Indeed, the performance of catastrophe bonds relative to other traditional asset classes through different financial market cycles demonstrates the value of this asset class and its non-correlative basis (see Chart 3.36). Consequently, the first half of 2014 saw the highest issuance of catastrophe bonds in any six-month period, with a record high of USD 5.7 billion. These trends are expected to continue for the full year 2014 (see Chart 3.37).
This excess of capital and capacity, combined with benign developments in natural catastrophe insured losses since 2013, has been reflected in a significant decline in prices of reinsurance policies (see Chart 3.36). In addition, new premiums written are continuing to decline as ceding companies use less reinsurance (increasing retention ratios via consolidation) as a means of stabilising profitability levels. Given these developments – weakening fundamentals and a challenging market environment – the European reinsurance sector was given a negative outlook by all rating agencies in the course of 2014. In an attempt to change the dynamics of the market, some European reinsurers have been
8 Including hedge funds, pension funds, endowments, sovereign wealth funds and asset managers.
… combined with an excess of capital and supply…
… driven by the good non-correlative performance of insurance-linked securities…
… with stagnant demand in a soft market…
Chart 3.36 Cumulative return profiles, broken down by market asset class and reinsurance pricing (Q1 2002 – Q3 2014; index: Q1 2002 =100)
100
110
120
130
140
150
160
0
50
100
150
200
250
300
350
2002 2004 2006 2008 2010 2012 2014
pricing (right-hand scale) catastrophe bonds (left-hand scale) US stocks (left-hand scale) hedge funds (left-hand scale) commodities (left-hand scale) euro area stocks (left-hand scale)
Sources: Bloomberg, Guy Carpenter and ECB calculations. Notes: S&P 500 and EURO STOXX are used as benchmark indices for US and euro area stocks respectively. The Guy Carpenter World Property Catastrophe RoL Index tracks changes in property catastrophe reinsurance premium rates on a worldwide basis.
Chart 3.37 Insured catastrophe losses and catastrophe bond issuance
(1997 – H1 2014; USD billions)
0
1
2
3
4
5
6
7
8
0
20
40
60
80
100
120
140
1997 1999 2001 2003 2005 2007 2009 2011 2013
insured losses (left-hand scale) issuance of catastrophe bonds (right-hand scale)
Sources: EQECAT, Munich Re, Swiss Re, Guy Carpenter and ECB calculations. Note: Data for 2014 refer to the first six months of the year.
Chart 3.37 Insured catastrophe losses and catastrophe bond issuance
(1997 – H1 2014; USD billions)
0
1
2
3
4
5
6
7
8
0
20
40
60
80
100
120
140
1997 1999 2001 2003 2005 2007 2009 2011 2013
insured losses (left-hand scale) issuance of catastrophe bonds (right-hand scale)
Sources: EQECAT, Munich Re, Swiss Re, Guy Carpenter and ECB calculations. Note: Data for 2014 refer to the first six months of the year.
82 ECB Financial Stability Review November 20148282
releasing their excess capital, via share buy-backs, higher than expected dividend payments or capital injections into direct insurance business lines, placing increased pressure on primary insurance pricing. Further cost-cutting and some consolidation are expected in the sector. As positive trends, selected lines (aviation) and countries (Germany) may enjoy slight pricing gains due to recent loss developments. Product innovation, such as protection against cyber risks, has also been pursued by some reinsurers. However, cyber risk has been poorly defined in reinsurance coverage thus far, and the market is at an incipient stage, with rather customised policies dominated by a few large providers.
The investment guarantees that life insurers can offer new customers are driven by the yields on the bonds they can invest in. Low yields reduce the level (or increase the price) of guarantees that insurers can offer, making guaranteed savings products unattractive to customers, hampering volumes of new premiums written and potentially making the business unviable for small, not well- diversified institutions that were unable or unwilling to mitigate the risk in advance through a close matching of cash flows or hedging activities.
3.3 MACRO-pRudENTIAL pOLICY MEASuRES ANNOuNCEd IN SEVERAL COuNTRIES
This section considers the macro-prudential measures that have been implemented, or proposed, in a number of euro area countries since November 2013. It draws on a quarterly update provided by Member States. The measures introduced by the countries concerned can be grouped into two categories, depending on the risks being addressed: real estate measures and systemic risk measures. They are summarised in Table 3.1.
SYSTEMIC RISk MEASuRES A number of member countries recently introduced measures to mitigate systemic risks originating from the significant size, high concentration and interconnectedness of their banking sectors. The measures ranged from the instruments provided for in the Capital Requirements Regulation/Capital Requirements Directive IV (CRR/CRD IV) to country-specific measures. For instance, Estonia put a systemic risk buffer (SRB) in place, while the Netherlands decided to introduce both an SRB and a buffer for other systemically important institutions (O-SII buffer), with phase-in arrangements. Belgium and Slovenia introduced ad hoc measures to address country-specific aspects of systemic risk, namely excessive trading activities of banks (Belgium) and funding liquidity (Slovenia).
In December 2013, Belgium decided to apply targeted Pillar 2 capital surcharges to banks’ trading activities above a certain threshold. Prior to the recent crisis, a number of Belgian banks’ trading activities were undesirably high. Although banks have since reduced their trading activities, the purpose of the surcharge is to deter banks from engaging in an undesirable level of trading activity, such as that observed prior to the crisis, and to ensure that trading activities do not become a significant obstacle to banks’ solvency. The surcharge is to be applied if a bank exceeds the threshold set for either of two indicators, a volume-based indicator and a risk-based indicator. The volume-based indicator consists of all held-for-trading assets that are not used for hedging the banks’ own positions. If the volume-based indicator exceeds the mark of 15% of the bank’s total assets, a capital surcharge equal to the amount by which the indicator exceeds the threshold will be applied. The risk-based indicator consists of the regulatory capital requirements for market risk (excluding foreign exchange risk). A capital surcharge will be applied if the “adjusted” market risk capital requirement exceeds 10% of total regulatory capital requirements, and the surcharge will equal three times the amount by which market risk capital requirements exceed the threshold. The thresholds of the indicators were determined on the basis of banks’ trading activities in the pre-crisis period. The measure is not subject to any predefined time limit.
… create a challenging outlook for the reinsurance
sector
Life business model tested by low yields
Systemic risk measures were
introduced in a number of countries…
… including Belgium…
83 ECB
Financial Stability Review November 2014 83
3� Euro arEa F inancial
inst itutions
83
The Netherlands decided in April 2014 to require an O-SII buffer of 1-2% for the most systemically important banks in the country, and an SRB of 3% for all Dutch banks with a balance sheet size (on and off-balance-sheet items) equal to at least 50% of the country’s annual gross domestic product (GDP), with the higher of the two requirements applying to each of the credit institutions concerned. As a result, a capital buffer of 3% of the respective risk-weighted assets (CET1 capital) was imposed for ING Bank, Rabobank and ABN AMRO, while one of 1% was required of SNS Bank. Banks are able to phase in these buffers between 2016 and 2019. This will raise future CET1 capital levels required of the three major banks to at least 10% of their risk-weighted assets, and that required of SNS Bank to 8%. The reasons for the imposition of these requirements are to be found in the relatively large size of the Dutch banking sector, in terms of GDP, and its level of concentration. To determine which banks are systemically important, De Nederlandsche Bank (DNB) assessed banks against a number of criteria such as the size of a bank relative to Dutch GDP, a bank’s interconnectedness with other financial institutions and the substitutability of certain crucial functions performed by a bank. On the basis of these criteria, DNB determined that ING Bank, Rabobank and ABN AMRO are the systemically most important banks. The size of the balance sheet of each individual major bank is in excess of 50% of Dutch GDP – in the case of ING Bank and Rabobank, the size actually exceeds 100% of GDP. The three major banks are also strongly interconnected, and are interwoven with other Dutch and international financial institutions. Finally, taken together, they are responsible for most lending to Dutch households (85%) and companies (60%). Although SNS Bank is far smaller and has a smaller share in the services provided to the real economy, it is likewise systemically important: it holds a relatively large proportion of Dutch consumers’ savings, and part of these savings is guaranteed under the deposit guarantee scheme. In addition, SNS Bank is an important player in the domestic mortgage loan market.
Slovenia decided in April 2014 to introduce minimum requirements on changes in loans to the non-banking sector relative to changes in non-banking sector deposits. The ratio is calculated on changes in stocks before considering impairments (gross loan-to-deposit flows). The measure was introduced to counter the observed acceleration of the decline in banks’ loan-to-deposit ratios in recent years (from a peak of 162% in 2008 to 130% in 2012, and further to 109% at the end of 2013), which was in turn accompanied by a decline in commercial wholesale funding and the contraction of the banking system’s total assets. By way of this measure, Banka Slovenije aims to stabilise the funding structure of the banking system and mitigate system-wide funding liquidity risk, as well as to restrict negative feedback between the condition of banks, real sector activity, system-wide liquidity and loan quality. Banka Slovenije expects the measure to reduce the migration of, and competition for, deposits. The calibration of gross loan-to-deposit flows was based on historical experience and simulations for individual banks. The minimum requirements set the floor for the measure as follows: 0% in the first year, and 40% in the second year. The instrument is being introduced on a temporary basis, until the banks’ funding structure has been stabilised successfully, and until system-wide funding liquidity risk has been reduced. Since the measure is to apply solely to banks in Slovenia, scope for cross-border spillover effects is very limited.
Estonia decided in May 2014 to set up a systemic risk buffer requirement of 2%, starting on 1 August 2014. The systemic risk buffer applies to all credit institutions licensed in Estonia. In Eesti Pank’s assessment, the main reasons for introducing the systemic risk buffer were the structural vulnerabilities of both the Estonian economy and its financial sector. The former stems primarily from the small size and from the openness of the Estonian economy. The ongoing convergence and build-up of a capital stock make the development of the economy more volatile than that of most other EU countries. Moreover, in Eesti Pank’s view, the financial buffers of the real economy,
… the Netherlands…
… Slovenia…
… and Estonia
84 ECB Financial Stability Review November 20148484
although growing, are still relatively small and provide only limited protection against sudden shocks, particularly external shocks. The structural vulnerabilities of the financial sector include the high concentration of the banking sector and the exposures of institutions to the same set of economies and economic sectors, which include exposures via other subsidiaries of parent banking groups. Although the direct exposures of credit institutions in Estonia to one another may be considered to be fairly limited, the structure of their credit portfolios indicate either that they have significant direct exposures to the domestic real sector or that they are likely to be significantly affected through second-round effects if a bank with a significant market share should fail to provide services. As the total capital requirement in Estonia was set at 10% from 1997 to 2013, and as all banks there fulfilled the requirement with a sufficient excess at the end of 2013, the introduction of the measure is expected to have an only limited impact both on the capitalisation of banks and on the financing conditions of the real economy.
REAL ESTATE MEASuRES Different types of real estate measures have been adopted, with the aim of addressing unfavourable developments in property markets. Real estate typically represents a large proportion of banks’ credit exposures, and of households’ assets, thus making imbalances in this sector particularly important in terms of financial stability. In this regard, Belgium, Slovakia, Ireland and Estonia decided to introduce national measures to address specific risks in the property markets.
In November 2013, Belgium decided to increase banks’ risk weights for certain exposures through a modification of the Belgian Own Funds Regulation. This decision was a result of an analysis both of the risks to the Belgian banking sector as a result of Belgian residential mortgages and of the adequacy of the capital requirements applicable to Belgian credit institutions (in Belgium, mortgage lending is undertaken primarily by Belgian credit institutions). The analysis was motivated by the significant increase in residential mortgage lending, as well as by the potential risk of an overvaluation of real estate in Belgium in recent years. Before the change, the capital requirements applicable to residential mortgages were relatively low for credit institutions relying on internal risk models (i.e. those using internal ratings-based (IRB) approaches) in Belgium (on average, 9.6% of the respective asset value), and were (and continue to be) significantly lower, on average, than those applied under the Basel II framework (35%). This is due to the fact that internal risk models are calibrated on historical credit loss data, and to the absence of a major crisis in the Belgian housing market in the past. Considering the findings of the analysis, the Nationale Bank van België/Banque Nationale de Belgique (NBB/BNB) increased the capital requirements applicable to exposures secured by mortgages on residential property in Belgium through the Basel II Pillar I framework. For IRB banks, the increase was 5 percentage points, while nothing changed for banks using the standardised approach. Once this macro-prudential measure has been implemented, the average risk weight for domestic mortgage loans for Belgian IRB banks will increase to around 14.6%, which is closer to the average risk weight observed in other core European countries. The NBB/BNB decided in March 2014 to uphold the increase in the capital requirements.
In October 2014, Slovakia decided to issue a non-binding recommendation on risks related to market developments in retail lending. The measure is to be introduced to counter the rapid pace of credit growth, the significant proportion of loans with high loan-to-value ratios (LTVs) and the high proportion of housing loans used to refinance other loans which do not involve any verification of the borrower’s income and which are not subject to any interest rate stress tests. The aim of the recommendation is to keep the parameters of new retail housing loans at sustainable levels, avoiding any underestimation of risks due to a higher level of competition. It provides for the share of high LTV loans (currently between 90% and 100%) to be limited to 25% in June 2015, to 20%
Real estate measures
introduced in…
… Belgium…
… Slovakia…
85 ECB
Financial Stability Review November 2014 85
3� Euro arEa F inancial
inst itutions
85
in March 2016, to 15% in December 2016 and to 10% in 2017. In addition, it stipulates that no new loans with LTV ratios of more than 100% should be extended. Slovakia moreover recommends that the banks impose own limits on their debt-to-income ratios and that they verify the income generated. Banks are also asked to implement interest rate testing when granting individual loans, as well as to perform portfolio stress testing for increases in interest rates and unemployment. Lending at long maturities, with progressive or deferred repayment, is not significant, but Národná banka Slovenska advises that such lending be avoided altogether. It recommends that banks take a prudential approach to loan refinancing and lending through intermediaries. The recommendation is considered a preventive step. Slovakia believes a non-binding measure to be proportionate to the current situation. Binding measures are regarded as unnecessary since the level of risks is not high. The need for additional measures will be assessed via regular follow-up procedures and reporting. The recommendation will enter into force in November 2014 (in case of LTV limits) and March 2015 (for other issues).
In October 2014, Ireland proposed that regulations placing ceilings on the share of mortgage lending at both high loan-to-value (LTV) ratios and high loan-to-income (LTI) ratios be introduced. The reason for the proposed regulation is the need to increase the resilience of Irish households and banks to residential property, in the context of high exposure of these sectors to property, and given the fact that a significant share of new lending is taking place at high LTV ratios and there have been sharp movements in house prices. Moreover, property lending tends to be subject to cyclical fluctuations which are amplified if lending standards are eased. The preceding crisis has shown the need for a policy overlay that would restrict imprudent lending throughout the credit cycle. The Central Bank of Ireland has acknowledged that loans at higher LTV and LTI rates can be appropriate in certain circumstances. For this reason, instead of imposing absolute limits, Ireland has proposed proportionate limits. The proposed measures will require banks to restrict lending for principal dwelling houses (PDHs) at rates above 80% LTV to no more than 15% of the value of all new PDH loans, and to restrict lending for PDHs at rates above 3.5 times LTI to no more than 20% of that aggregate value. Furthermore, the proposed regulation provides for a lower threshold for buy-to-let (BTL) property, requiring banks to limit BTL housing loans at rates above 70% LTV to 10% of all BTL housing loans. The rationale behind adopting limits on LTV and LTI together is to be found in the fact that both measures complement each other, with the LTI addressing the borrower’s loan affordability and the LTV lender’s losses in the event of default. Such thresholds are aimed at ensuring a greater degree of safety around the mortgage business. The objectives of the proposed regulations are to increase the resilience of the banking and household sectors with respect to the property market and to dampen the risk of self-reinforcing dynamics between property lending and house prices. While the regulations are not yet in place, regulated lenders have been instructed to take account of the probable introduction of such a regime and to already start adapting their lending practices in anticipation of its introduction.
In October 2014, Estonia announced plans to set limits on the granting of housing loans as from 2015. Eesti Pank plans to introduce three requirements targeted at the housing market: LTV ratios, debt service-to-income ratio (DSTI) limits and a maximum maturity. The LTV will be limited to 85% (90% in the case of housing loans guaranteed by the state foundation KredEx). The DSTI limit will restrict the total amount of monthly loan, lease principal and interest payments to below 50% of the borrower’s net monthly income. Finally, the maximum maturity of housing loans will be set at 30 years. The requirements are to be introduced as a precautionary measure to address the potential risk of an overvaluation of the property market. Eesti Pank does not expect the new limits to tighten prevailing lending conditions. The measures will affect all banks operating in Estonia, including branches of foreign banks.
… Ireland…
… and Estonia
86 ECB Financial Stability Review November 20148686
Table 3.1 Overview of macro-prudential policy measures implemented and proposed in euro area countries since November 2013
Country Measure Summary description Date of entry into force
Reasons for implementation
Systemic risk measures
Belgium Capital surcharge for excessive trading activities
The capital surcharge will be applied as a Pillar 2 add-on if a bank exceeds the threshold set for either of two indicators, a volume-based indicator or a risk-based indicator.
December 2013 Prevent a build-up of systemic risk
Netherlands Systemic risk buffer (SRB) and buffer for other systemically important institutions (O-SII buffer)
An O-SII buffer of 1-2% for the most systemically important banks and an SRB set at 3% of the total risk exposure (consolidated basis) for all Dutch banks with a balance sheet size equal to at least 50% of Dutch GDP. For each credit institution, the higher of the two requirements applies.
Phased in from January 2016 to January 2019
Mitigate the long-term non-cyclical systemic risk emanating from the large and concentrated banking sector in the Netherlands
Slovenia Liquidity requirements The measure is based on the gross loan-to-deposit flows ratio. The ratio required has been set at 0% in the first year and at 40% in the second year. It is a temporary measure.
June 2014 Prevent and mitigate systemic risk emanating from an excessive maturity mismatch and from funding illiquidity
Estonia Systemic risk buffer An SRB set at 2% of the total risk exposure (consolidated and individual basis).
August 2014 Structural vulnerabilities of the economy and financial sector
Real estate instruments
Belgium Risk weights Five percentage point add-on to risk weights of Belgian residential mortgage loans calculated by banks that use an internal ratings-based approach.
November 2013 Increase in residential mortgage lending
Slovakia Recommendation on lending criteria
Non-binding recommendation related to the risks in the housing lending market.
First phase November 2014
Excessive credit growth and significant proportion of loans with high LTV
Ireland LTV and LTI limits Proposal for proportionate limits on LTI and LTV ratios currently under consultation: new principal dwelling house (PDH) loans at rates above 80% LTV may not exceed 15% of the total value of all new PDH lending, and new PDH loans at rates above 3.5 times LTI may not exceed 20% of the total value of new PDH lending. Buy-to-let (BTL) housing property loans at rates above 70% LTV restricted to 10% of the total value of all BTL housing property lending.
To be announced Increase resilience of households and banks to property, given large share of high LTV loans and sharp movements in house prices.
Estonia Requirements for housing loans
Introduction of requirements for housing market: LTV limits (85% and 90% in case of loans guaranteed by KredEx), debt service-to-income (DSTI) limits (50%), maximum maturity of housing loans (30 years)
Early 2015 Address the potential risk of an overvaluation of real estate market and protect financial system against the excessive risk-taking by banks in credit booms
87 ECB
Financial Stability Review November 2014 87
3� Euro arEa F inancial
inst itutions
87
3.4 REShApINg ThE REguLATORY FRAMEWORk FOR FINANCIAL INSTITuTIONS, MARkETS ANd INFRASTRuCTuRES
This section provides an overview of a number of regulatory initiatives in the banking, insurance and market spheres that are of primary importance for enhancing financial stability in the European Union (EU).
REguLATORY INITIATIVES FOR ThE BANkINg SECTOR The key elements of the regulatory requirements for financial institutions operating in the EU, as well as the framework for the supervisory review and evaluation process and the mechanism for coordinating the activities of national and EU authorities, are set out in the Capital Requirements Regulation/Capital Requirements Directive IV (CRR/CRD IV). This prudential framework is complemented by the Single Supervisory Mechanism Regulation (SSM Regulation) that provides the ECB with strong powers for the supervision of all banks in participating Member States, as well as with additional tasks and responsibilities in the area of macro-prudential policy. While many elements of the CRR/CRD IV package are already in force, some remaining elements are still subject to finalisation and calibration, including the liquidity regulation, the leverage ratio provisions and the securitisation rules.
The international framework for liquidity regulation includes two policy instruments, namely the liquidity coverage ratio (LCR) and the net stable funding ratio (NSFR). The LCR is aimed at promoting the short-term resilience of the liquidity risk profile of banks, while the NSFR is aimed at diminishing maturity mismatches between assets and liabilities, thereby reducing funding risks of banks. Since the publication of the final definition of the LCR by the Basel Committee in January 2013, the European Commission has made significant progress with respect to the implementation of this liquidity standard in the EU. A key element of this process was the publication of the final delegated act on the LCR in October 2014.9 The remaining work primarily concerns the scope of supervisory reporting before the LCR is phased in next year, with an initial minimum requirement of 60%.
As regards the NSFR, the Basel Committee published a consultative document on a revised calibration of the measure in January 2014, also aligning the NSFR with the LCR in terms of the treatment of high-quality liquid assets. In the European context, the European Banking Authority (EBA) has set up a team to assess the impact and appropriate calibration of the NSFR. A final report by the EBA is expected to be delivered to the European Commission by the end of 2015.
The ECB actively supports the ongoing work on the leverage ratio that is aimed at preventing the build-up of excessive leverage in the financial system. Following the endorsement of the revised definition of the leverage ratio by the governing body of the Basel Committee in January 2014, the Group of Central Bank Governors and Heads of Supervision (GHOS), the European Commission issued a delegated act that broadly aligns the CRR/CRD IV definition with the revised international standard. As regards the implementation of the leverage ratio as a supervisory tool, banks will be required to publicly disclose their leverage ratios as from January 2015.10
9 See the Commission Delegated Regulation of 10 October 2014 (available at: http://ec.europa.eu). 10 See Special Feature C in this issue of the FSR for further details on the NSFR.
Work on the finalisation and calibration of certain key elements of the CRR/CRD IV is continuing
88 ECB Financial Stability Review November 20148888
In the area of securitisation, significant work is underway at the global and EU levels. This stems from the policy objective of reviving securitisation markets in a sustainable manner, and reflects the positive effects that sound securitisation practices can have on the financing of the real economy.
At the international level, the Basel Committee on Banking Supervision (BCBS) and the International Organization of Securities Commissions (IOSCO) set up a Task Force on Securitisation Markets earlier this year, with the aim of (i) identifying factors that may be hindering the development of sustainable securitisation markets and the participation of certain types of investors, and (ii) defining criteria to identify and assist in the development of simple and transparent securitisation structures.11 These criteria could inform future regulatory actions, such as those of the BCBS, which pledged, at its meeting in September 2014, to consider in 2015 how to incorporate the BCBS-IOSCO criteria, once finalised, into the securitisation capital framework.12
In the EU in October, the European Commission adopted two delegated acts under the Solvency II Directive and – for the LCR – the CRR that establish a differentiated regulatory treatment of securitisations that meet certain criteria in terms of simplicity and transparency. In addition, following a call for advice from the European Commission on the appropriateness of the prudential requirements provided for in the CRR/CRD IV in relation to long-term financing and, in particular, securitisations, the EBA determined that certain simple, standard and transparent securitisations merit differentiated capital treatment, developed draft criteria to identify such securitisations and launched a public consultation that is to be closed in mid-January 2015.13
Notwithstanding the ongoing work on the above-mentioned prudential requirements, several policy tools are already available for also macro-prudential purposes. Subject to strict notification and coordination mechanisms between national and EU authorities, including the ECB under the SSM Regulation, the CRR/CRD IV defines a set of instruments that can be applied by macro-prudential authorities to address risks to financial stability.
As required by the CRR, the revision by the European Commission of the macro-prudential rules is an ongoing process. The revision is focusing on the assessment of the effectiveness, efficiency and transparency of the policy framework and on the adequacy of the coverage of, and possible overlap between, tools, as well as on the interaction between internationally agreed standards and the provisions of the CRR/CRD IV.
The European Systemic Risk Board (ESRB) and the EBA have already provided the Commission with their assessment of the adequacy of the macro-prudential policy framework and have set out a number of proposals with regard to possible ways of improving the framework. The ECB, too, is currently assessing the adequacy of the macro-prudential rules in the CRR/CRD IV, with a specific focus on identifying the main issues arising from the establishment of the SSM and on ensuring consistency between the SSM Regulation and the CRR/CRD IV.
With regard to recently passed legislation or ongoing regulatory initiatives, Tables 3.2 to 3.4 provide an update of the major strands of work in the EU, followed by a short overview of selected policy measures from the perspective of financial stability and macro-prudential policy.
11 See the BCBS-IOSCO press release of 3 July 2014 (available at: http://www.bis.org/press/p140703.htm). 12 See BCBS press release (available at: http://www.bis.org/press/p140925.htm). 13 See EBA, “EBA Discussion Paper on simple, standard and transparent securitisations”, 14 October 2014.
The concept of “simple and
transparent securitisation”
is spreading ever faster at both the
global and the EU level
89 ECB
Financial Stability Review November 2014 89
3� Euro arEa F inancial
inst itutions
89
Another key area of significant progress comprises steps taken towards a banking union in Europe, namely the establishment of (i) a single supervisory mechanism, (ii) a single resolution framework, (iii) a single resolution mechanism and (iv) harmonised deposit insurance. The first pillar of the banking union, the Single Supervisory Mechanism became operational on 4 November.
Important complementary elements of single supervisory arrangements are a common EU framework for bank recovery and resolution, as well as a single resolution mechanism. As of 1 January 2015, the Bank Recovery and Resolution Directive (BRRD) will be implemented by all Member States.14 The BRRD establishes common and efficient tools and powers for addressing a banking crisis pre-emptively, and for managing failures of credit institutions and investment firms in an orderly manner throughout the EU.
14 With the exception of the bail-in tool, which will follow by 1 January 2016 at the latest.
Significant progress made in the establishment of the banking union
The BRRD will provide common and efficient tools and powers for addressing a banking crisis
Table 3.2 Selected new legislation and proposals for legislative provisions on the banking sector in the Eu
Initiative Description Current status Single Supervisory Mechanism Regulation (SSM Regulation)
The SSM Regulation establishes a Single Supervisory Mechanism (SSM) with strong powers for the ECB (in cooperation with national competent authorities) for the supervision of all banks in participating Member States (euro area countries and non-euro area Member States which join the system).
The SSM came into force on 4 November 2014, and the ECB took up its new role of supervisor. The results of the comprehensive assessment of all banks that are under its direct supervision were published on 26 October 2014.
Bank Recovery and Resolution Directive (BRRD)
The BRRD sets out a framework for the resolution of credit institutions and investment firms, with harmonised tools and powers relating to prevention, early intervention and resolution for all EU Member States.
The BRRD entered into force on 2 July 2014. Member States have to transpose the BRRD into national legislation by 31 December 2014, and to apply it as from 1 January 2015. However, the bail-in provisions will only be applicable as of 1 January 2016, at the latest.
Deposit Guarantee Scheme Directive (DGS Directive)
The DGS Directive deals mainly with the harmonisation and simplification of rules and criteria applicable to deposit guarantees, a faster pay-out, and an improved financing of schemes for all EU Member States.
The DGS Directive entered into force on 2 July 2014. Member States will have to transpose most provisions into national legislation by 3 July 2015, and in full by 31 May 2016.
Single Resolution Mechanism Regulation (SRM Regulation)
The SRM Regulation establishes a single system, with a single resolution board and single resolution fund, for an efficient and harmonised resolution of banks within the SSM. The SRM would be governed by two main legal texts: the SRM Regulation, which covers the main aspects of the mechanism, and an Intergovernmental Agreement (IGA) relating to some specific aspects of the Single Resolution Fund (SRF).
The SRM Regulation entered into force on 19 August 2014. It will be partly applicable as of 1 January 2015, whereas most resolution functions (including the SRF) will apply as from 1 January 2016 (or when the IGA becomes applicable, if later). The IGA on the SRF was signed by all Member States (except the United Kingdom and Sweden) on 21 May 2014, and its ratification by national parliaments is now pending.
Regulation on structural measures The Regulation introduces restrictions on certain activities and sets out rules on structural separation, with the aim of improving the resilience of EU credit institutions.
The European Commission’s proposal was published on 29 January 2014. Preliminary discussions have started in the European Council. The ECB’s legal opinion on the proposal was published on 21 November 2014.
90 ECB Financial Stability Review November 20149090
The Single Resolution Mechanism (SRM) will establish a single system, with a Single Resolution Board (SRB) and a Single Resolution Fund (SRF) at its centre, for the resolution of banks in Member States participating in the SSM. The SRM is a necessary complement to the SSM in order to achieve a well-functioning banking union and to sever the link between banks and their sovereigns. Thus, the SRM will apply to all banks supervised within the scope of the SSM, and accordingly, any Member State outside the euro area which opts to join the SSM will automatically also fall under the SRM. The SRM will ensure that in the event of a bank failing, and if it is in the public interest to resolve it, its resolution can be managed efficiently, jointly and in the common interest. The SRM will be better placed to take due account of contagion and spillovers when making resolution decisions. It will also ensure a consistent application of resolution principles and tools throughout the banking union, also for banks with no cross-border activity.
The SRM will be governed by two main legal texts: (i) the SRM Regulation, which covers the main aspects of the mechanism and is based on the BRRD, and (ii) an Intergovernmental Agreement (IGA), which covers some specific aspects of the SRF. Whereas most of the provisions of the SRM Regulation will apply as from 1 January 2016, the SRB will become operational on 1 January 2015. This will allow the SRB to engage in recovery and resolution planning during 2015.15 The European Commission is responsible for the establishment of the SRB, and a dedicated Commission Task Force has been set up for this purpose.
The IGA on the transfer and mutualisation of contributions to the SRF was signed by 26 Member States.16 All signatories of the IGA are to complete its ratification according to their national procedures before 1 January 2016. This is expected to take place soon, given that the Commission has recently adopted a delegated act and a proposal for a Council implementing act on the risk-based bank contributions to national resolution funds and the SRF, as required by the BRRD and the SRM Regulation respectively.
15 This may include, for example, the examination of recovery plans received from the ECB or national competent authorities in order to identify any actions which may adversely impact the resolvability of the institutions, and the drafting and adoption of resolution plans, including the assessment of resolvability, the application of simplified obligations for certain institutions and the determination of the minimum requirements of eligible liabilities and own funds for bail-ins, for all covered institutions.
16 The IGA was signed by all Member States except the United Kingdom and Sweden.
The SRM will create a single
system for resolution
Box 6
REguLATORY INITIATIVES TO ENhANCE OVERALL LOSS-ABSORpTION CApACITY
One of the key objectives of the resolution frameworks introduced in response to the recent crisis, such as the Bank Recovery and Resolution Directive (BRRD) in the EU, is the shifting of the cost of bank failures from the taxpayer to, first and foremost, the shareholders and creditors of the failing bank. This is important for many reasons, not least that of solving the too-big-to-fail problem of large banks, which – unless there is a credible resolution option – often have to be bailed out by the public at huge cost. These banks have often been perceived by markets as having an implicit state guarantee, which creates not only a moral hazard problem, but also an uneven playing field among banks, in that large banks in fiscally strong countries can fund themselves far more cheaply than smaller banks or banks in countries with weaker public finances. Thus, the introduction of a credible resolution framework contributes to weakening the link between banks and their sovereigns, which proved to be both costly and destabilising in the recent crisis.
91 ECB
Financial Stability Review November 2014 91
3� Euro arEa F inancial
inst itutions
91
An important tool for attaining this objective is the bail-in tool, which enables the resolution authority to write down, or convert into equity, the claims of a broad range of creditors. However, some types of liabilities are excluded from the scope of a bail-in, such as secured liabilities and covered deposits. Furthermore, in exceptional circumstances, other liabilities may also have to be excluded on a case-by-case basis, either because it is not possible to bail them in quickly enough or because this is necessary in order to attain the resolution objectives. Consequently, in order to ensure that the bail-in tool will still be efficient in resolution, there is a need to make sure that there are sufficient own funds and liabilities in banks for bail-ins, when needed.
Under the BRRD, Member States are required to ensure that institutions meet a minimum requirement for own funds and eligible liabilities (MREL) for bail-ins.1 An adequate level of own funds and eligible liabilities will be key to ensure that there is sufficient loss-absorbing capacity within institutions when they fail, thereby underpinning the efficient application of the bail-in tool. It will also protect the resolution funds, including the Single Resolution Fund, as own funds and eligible liabilities, as defined by the MREL, and other bail-inable liabilities will be used before a resolution fund may contribute to the funding of any resolution.
Some technical details on the MREL remain to be finalised before it becomes operational along with the bail-in tool in 2016. In particular, the European Banking Authority will draft regulatory technical standards by July 2015 which will specify how the MREL is to be determined for each institution. By December 2016, the European Commission will submit a legislative proposal on the harmonised application of the MREL. Such a proposal may include the introduction of an appropriate number of different MRELs that take account of the different business models of institutions and groups, as well as possible adjustments to ensure consistency with any international standards that have been developed by international fora in this area.
Currently, an international standard is also under discussion within the G20 and the Financial Stability Board (FSB) so as to end the too-big-to-fail problem of the global systemically important banks (G-SIBs). The FSB, in consultation with the Basel Committee on Banking Supervision, has developed proposals on the adequacy of loss-absorbing capacity of G-SIBs in resolution, in response to a call by G20 leaders at the 2013 St Petersburg summit. The proposal is subject to public consultation and a quantitative impact study, before being finalised by the FSB in 2015. This proposal would be the international equivalent of the MREL in the BRRD, applicable to G-SIBs only. Although similar, the draft FSB proposal for G-SIBs’ total loss-absorbing capacity (TLAC) in resolution differs from the MREL in some key areas (see the table below).
1 Within the SRM, the SRB will be the authority, after consulting competent authorities, including the ECB, which determines the MREL for all entities under direct ECB supervision and for all cross-border groups.
key features of the MREL and the TLAC
MREL TLAC
Scope All banks in scope of the BRRD G-SIBs only
Set-up A minimum requirement in parallel to Basel III minimum capital requirements for banks, calculated as the amount of own funds (including buffers) and eligible liabilities.
A minimum requirement incorporating Basel III minimum capital requirements and excluding Basel III buffers for G-SIBs.
92 ECB Financial Stability Review November 20149292
key features of the MREL and the TLAC (cont’d)
MREL TLAC
Determination Determined on an individual basis for each institution.
A common minimum Pillar 1 requirement set within the range of 16-20% of RWAs and at least twice the Basel III Tier 1 leverage ratio requirement1 as a floor for all G-SIBs, with the possibility for authorities to top it up on an individual basis through a Pillar 2 component. Also sets out how TLAC is distributed among material institutions within a group when the whole group is resolved or when various sub-sets of the group are resolved together.
Eligible instruments Capital instruments can simultaneously satisfy both minimum regulatory capital requirements (including buffers) and the MREL. To be eligible, liabilities need to fall within the scope of bail-in. This will exclude e.g. covered deposits and, in principle, secured liabilities. Additionally, eligible liabilities must satisfy certain criteria, such as issued and fully paid up, not owed to, secured or guaranteed by the institution itself, not arise from a derivative or from a preferred deposit, and have a remaining maturity of at least one year.
Capital instruments can simultaneously satisfy both minimum regulatory capital requirements and TLAC, but only CET1 capital in excess of that required to satisfy these requirements may count towards the capital buffers. Certain liabilities are excluded from consideration for TLAC, e.g. liabilities arising from derivatives, insured deposits and liabilities which are preferred to normal senior unsecured creditors under the relevant insolvency law. Eligible external TLAC must be unsecured, must have a minimum remaining maturity of at least one year and must not be subject to set off or netting rights. Credible ex ante commitments by authorities to recapitalise a G-SIB, which may be required to contribute to resolution funding, may count towards a firm’s Pillar 1 minimum TLAC, subject to certain strict conditions (e.g. the commitments must be pre-funded by industry contributions).
Priority Priority is not a precondition in the BRRD. Eligible external TLAC must absorb losses prior to excluded liabilities in insolvency or in resolution without giving rise to material risk of successful legal challenge or compensation claims.
Regulation of investors Without prejudice to the existing large exposure regime Member States have to ensure that in order to provide for resolvability of institutions/groups, resolution authorities limit the extent to which other institutions hold liabilities eligible for the bail-in tool, save for liabilities that are held at entities that are part of the same group.
G-SIBs must deduct from their own TLAC or regulatory capital exposures to eligible external TLAC liabilities issued by other G-SIBs in a manner generally parallel to the existing provisions in Basel III that require a bank to deduct from its own regulatory capital certain investments in the regulatory capital of other banks. Further provisions, also for non G-SIBs, are envisaged.
1) The calibration is subject to a quantitative impact study and market survey which will be carried out in early 2015.
A final element of the banking union is the establishment, in the medium term, of a common deposit guarantee fund in the EU. A first step in this direction was the entry into force of the recast Deposit Guarantee Scheme Directive (DGS Directive) on 2 July 2014.17
17 By 3 July 2019, the Commission must submit a report and, if appropriate, a legislative proposal to the European Parliament and the Council, setting out how deposit guarantee schemes operating in the EU may cooperate through a European scheme so as to prevent risks from arising from cross-border activities and to protect deposits against such risks.
Improved depositor protection in the
EU
93 ECB
Financial Stability Review November 2014 93
3� Euro arEa F inancial
inst itutions
93
The DGS Directive will ensure that deposits in all Member States will continue to be guaranteed up to an amount of €100,000 per depositor and bank. It will also ensure faster pay-outs with specific repayment deadlines, which will gradually be reduced from 20 to 7 working days by 2024. It will also ensure a strengthened financing of deposit guarantee schemes, notably by requiring a significant level of ex ante funding (0.8% of covered deposits) which is to be met within ten years. At most 30% of the funding could be made up of payment commitments. In case of insufficient ex ante funds, the deposit guarantee scheme would collect immediate ex post contributions from the banking sector and, as a last resort, the scheme would have access to alternative funding arrangements, such as loans from public or private third parties. In addition, a voluntary mechanism for mutual borrowing between national deposit guarantee schemes in the EU is also provided for.
On 29 January 2014, the European Commission presented its proposal for a Regulation on structural measures for EU credit institutions. The proposal aims at improving the resilience of European banks by preventing contagion from banks’ trading activities to traditional banking activities. This would be done by prohibiting banks from carrying out proprietary trading, i.e. securities trading not related to client activity or hedging, and only for the purpose of making a profit for their own account. Furthermore, it is proposed that supervisors can require a bank to shift other trading activities to trading entities, which are legally, economically and operationally separated from the deposit-taking entity of the bank. Importantly, trading in government bonds issued by Member States will be exempted from the prohibition, as well as from the separation requirements. Likewise, the deposit-taking entity will still be able to use financial instruments aimed at hedging its own risks. The regulation will cover all global systemically important banks in the EU, as well as other banks with sufficiently large trading activities.
Another key objective of this proposal is to reduce banks’ incentives to take excessive risks on the back of the safety net (resolution funds, deposit insurance funds and, ultimately, governments), and to make banks less complex to resolve. In ensuring that, the proposal can complement the BRRD and may, at the same time, contribute to enhancing systemic stability in Europe. Also, by harmonising rules on structural regulation, the proposal seeks to create a level playing field for banks inside the EU.
The ECB supports this proposal in principle. It will contribute towards ensuring a harmonised EU framework that addresses concerns related to banks that are “too big to fail” and “too interconnected to fail”. Nevertheless, the ECB considers it important to sufficiently preserve the market-making activities of banks in order to maintain or increase asset and market liquidity, to moderate price volatility and to increase securities markets’ resilience to shocks.
REguLATORY INITIATIVES FOR FINANCIAL MARkETS ANd INFRASTRuCTuRES In addition to initiatives in the area of banking regulation, several steps have been taken to also strengthen the resilience of financial infrastructures.
The ECB Regulation on oversight requirements for systemically important payment systems came into force on 12 August 2014. The Regulation aims to ensure the efficient management of legal, credit, liquidity, operational, general business, custody, investment and other risks, as well as sound governance arrangements, objective and open access and the efficiency and effectiveness of systemically important payment systems (SIPSs). It implements the principles for financial market infrastructures (PFMIs) developed jointly by the Committee on Payments and Market Infrastructures and IOSCO in a legally binding way, and covers both large-value and retail payment systems of systemic importance, irrespective of whether they are operated by Eurosystem national central
The proposal for a Regulation on structural measures aims at improving the resilience of European banks
Adoption of an ECB Regulation on oversight requirements for systemically important payment systems
94 ECB Financial Stability Review November 20149494
banks or private entities. Four SIPSs have been identified: TARGET2 (operated by the Eurosystem), EURO1 and STEP2 (both operated by EBA Clearing), and CORE (FR) (operated by STET). The Eurosystem will review this list on the basis of updated statistical data each year. For consistency with international practices, and to take account of the increased integration of retail payment systems in the Single Euro Payments Area (SEPA), the Eurosystem has also undertaken a comprehensive review of the oversight standards for euro retail payment systems that are not SIPSs. As a result of this review, the ECB published the “Revised Oversight Framework for Retail Payment Systems” on 21 August 2014.
Table 3.3 Selected new legislation and legislative proposals for financial markets and infrastructures in the Eu
Initiative Description Current status
ECB Regulation on oversight requirements for systemically important payment systems
The Regulation aims at ensuring the efficient management of all types of risk that systemically important payment systems (SIPSs) face, together with sound governance arrangements, objective and open access, as well as the efficiency and effectiveness of SIPSs.
The Regulation entered into force on 12 August 2014.
European Market Infrastructure Regulation (EMIR)
The Regulation aims to bring more safety and transparency to the over-the-counter derivatives market and sets out rules for, inter alia, central counterparties and trade repositories.
The Regulation entered into force in August 2012. Implementation is in progress.
Regulation on improving the safety and efficiency of securities settlement in the EU and on central securities depositories (CSD Regulation)
The Regulation introduces an obligation of dematerialisation for most securities, harmonised settlement periods for most transactions in such securities, settlement discipline measures and common rules for central securities depositories.
The Regulation entered into force on 17 September 2014. Implementation is in progress.
Review of the Markets in Financial Instruments Directive and Regulation (MiFID II/MiFIR)
The legislation will apply to investment firms, market operators and services providing post-trade transparency information in the EU. It is set out in two pieces of legislation: a directly applicable regulation dealing, inter alia, with transparency and access to trading venues, and a directive governing authorisation and the organisation of trading venues and investor protection.
The Directive 2014/65/EU on markets in financial instruments (MiFID II) and the Regulation (EU) No 600/2014 on markets in financial instruments (MiFIR) were both published in the Official Journal of the EU on 12 June 2014.
Proposal for a Money Market Fund Regulation (MMF Regulation)
The proposal addresses the systemic risks posed by this type of investment entity by introducing new rules aimed at strengthening their liquidity profile and stability. It also sets out provisions that seek, inter alia, to enhance their management and transparency, as well as to standardise supervisory reporting obligations.
The European Commission’s proposal was published in September 2013 and has since been subject to discussions at the trialogue level by the European Parliament and, lately, by the European Council.
Proposal for a Regulation on reporting and transparency of securities financing transactions
The proposal contains measures aimed at increasing the transparency of securities lending and repurchase agreements through the obligation to report all transactions to a central database. This seeks to facilitate regular supervision and to improve transparency towards investors and on re-hypothecation arrangements.
The European Commission’s draft proposal was published in January 2014. The ECB expressed its support, in principle, of the proposal in its legal opinion of 24 June 2014.
95 ECB
Financial Stability Review November 2014 95
3� Euro arEa F inancial
inst itutions
95
Implementation of the European Market Infrastructure Regulation (EMIR) has continued to make progress. The Regulation seeks to bring more stability, transparency and efficiency to derivatives markets by requiring, inter alia, standard derivative contracts to be cleared through central counterparties (CCPs), and all European derivative transactions to be reported to trade repositories. CCPs that were previously authorised in a Member State had to apply for authorisation under EMIR by 15 September 2013. On 18 March 2014, the first EU CCP was authorised under EMIR. In the meantime, further EU CCPs that had filed an application have been authorised to offer services and conduct activities in the EU.18 The first authorisations of CCPs under EMIR have set in motion the process of determining the classes of derivatives subject to the mandatory clearing obligation. The European Securities and Markets Authority (ESMA) submitted final draft regulatory standards on the clearing obligation to the European Commission in October 2014, covering several classes of over-the-counter (OTC) interest rate derivatives. Mandatory clearing of these products will enter into force gradually as from 2015. The Eurosystem complements EMIR and uses the PFMIs as its oversight standards for CCPs.
The Regulation on improving securities settlement in the EU and on central securities depositories (the CSD Regulation) entered into force on 17 September 2014. The aim of the Regulation is to increase the safety and efficiency of securities settlement and settlement infrastructures (i.e. central securities depositories – CSDs) in the EU. It introduces, inter alia, an obligation of dematerialisation for most securities, harmonised settlement periods for most transactions in such securities, settlement discipline measures and common rules for CSDs. The CSD Regulation enhances the legal and operational conditions for cross-border settlement in the EU. It delegates to ESMA and the EBA the drafting, in close cooperation with the members of the ESCB, of technical standards within nine months of its entry into force (i.e. before end-June 2015). The PFMIs complement the provisions of the CSD Regulation with respect to the Eurosystem’s oversight standards.
In the field of shadow banking, the FSB carried on with the deliverables agreed at the G20 Summit in St Petersburg in 2013, with a view to presenting an updated roadmap in time for the Brisbane Summit on 15-16 November 2014. Milestones attained in the last six months include:19
(i) The publication in October of a revised regulatory framework on haircuts for non-centrally cleared short-term financing transactions to limit the build-up of excessive leverage outside the banking system and help reduce pro-cyclicality.20 The framework includes a consultative proposal on the application of numerical haircut floors to transactions between non-banks.
(ii) The review of standards and processes for global securities financing data collection and aggregation ahead of their planned public consultation.
(iii) The approval of a work plan to examine a possible harmonisation of regulatory approaches to re-hypothecation of client assets and possible financial stability issues related to collateral re-use.
18 An up-to-date list of authorised CCPs can be found on ESMA’s website at: http://www.esma.europa.eu/content/Registries-and-Databases. 19 See the FSB press release issued following the FSB Plenary Meeting on 17 and 18 September 2014 in Cairns, Australia (available at:
https://www.financialstabilityboard.org/press/pr_140918.htm). 20 See the FSB press release of 14 October 2014 (available at: https://www.financialstabilityboard.org/press/pr_141013.htm).
The FSB makes further progress with its shadow banking agenda
96 ECB Financial Stability Review November 20149696
The FSB intends, in 2015, to launch a peer review of the jurisdictional implementation of the high- level policy framework for strengthening oversight and regulation of shadow banking entities (other than MMFs).
REguLATORY INITIATIVES FOR ThE INSuRANCE SECTOR The Solvency II Directive will harmonise the different regulatory regimes for insurance corporations in the European Economic Area and will introduce risk-based capital requirements for the first time. After the adoption by the Council of the Omnibus II Directive, which amends the Solvency II Directive, the European Commission and the European Insurance and Occupational Pensions Authority (EIOPA) are working on rules and guidelines to specify more detailed requirements for individual undertakings, as well as for groups. In October, the Commission published the Solvency II Delegated Act, which covers the scope of the valuation of assets, capital requirements, governance, group supervision, third country equivalence, and reporting and public disclosure. The Delegated Act also sets out the details on the favourable treatment of long-term guarantee activities as agreed in the Omnibus II Directive, as well as details on the preferential regulatory treatment of high-quality securitisations. EIOPA is working on Implementing Technical Standards (ITSs) and Guidelines on Solvency II to ensure its uniform application. EIOPA has divided the ITSs and Guidelines into two sets. A first set of ITSs and Guidelines was submitted to the Commission on 31 October. A second set is scheduled to be published for consultation in December 2014, and is expected to be finalised by the middle of next year, before Solvency II is applied in 2016.
At the international level, the International Association of Insurance Supervisors (IAIS) has decided to identify, for 2014, the nine global systemically important insurers (G-SIIs) identified in 2013. A set of policy measures, such as higher loss absorbency (HLA), will apply to those insurers. As a basis for the HLA, the basic capital requirements (BCRs) for G-SIIs are currently being developed by the IAIS. The simple, factor-based BCRs will be replaced by a risk-sensitive global insurance capital standard (ICS) from 2019. The ICS will be applied not only to G-SIIs, but also to the wider group of internationally active insurance groups.
Progress made with the technical
implementation of the Solvency II
regime
Development of group-wide global insurance capital
standards
Table 3.4 Selected legislative proposals for the insurance sector in the Eu
Initiative Description Current status
Solvency II Directive/Omnibus II Directive
The Solvency II Directive is the framework directive that aims to harmonise the different regulatory regimes for insurance corporations in the European Economic Area. Solvency II includes capital requirements, supervision principles and disclosure requirements. The Omnibus II Directive aligns the Solvency II Directive with the legislative methods introduced by the Lisbon Treaty, incorporates new supervisory measures given to the European Insurance and Occupational Pensions Authority (EIOPA) and makes technical modifications.
The Solvency II Directive was adopted by the EU Council and the European Parliament in November 2009. It is now scheduled to come into effect on 1 January 2016. The European Commission has published the Delegated Act on Solvency II. EIOPA has submitted a first set of Implementing Technical Standards (ITSs) on approval processes and “Guidelines” relevant for approval processes, including Pillar 1 (quantitative basis) and internal models.
97 ECB
Financial Stability Review November 2014 97
3� Euro arEa F inancial
inst itutions
97
OThER INITIATIVES Finally, an issue closely related to financial regulation is the proposal published by the European Commission on 14 February 2013 for implementing a financial transaction tax (FTT) in 11 euro area Member States via enhanced cooperation. The European Parliament adopted a legislative resolution on the proposal, in which it supports the Commission’s proposal but calls for several amendments. The negotiations among Member States are continuing in the meantime. On 6 May 2014, new political impetus was given in a joint statement by ten ministers, issued in the context of the ECOFIN Council meeting. The statement envisages a staged approach (first equities and some derivatives, followed by other instruments at a later stage) and foresees that a first step of FTT implementation will enter into force in 2016.
Legislative proposals on tax policies do not fall within the fields of competence of the ECB. However, the ECB is monitoring the legislative process closely in view of the possible impact of the FTT on financial markets, financial market infrastructures, monetary policy implementation and financial stability.
99 ECB
Financial Stability Review November 2014
SpEC IAL FEATuRES A FIRE-SALE EXTERNALITIES IN ThE EuRO AREA BANkINg SECTOR1
This special feature studies the effects of fire-sale externalities in the euro area banking sector. Using individual bank balance sheet data and a framework developed by Greenwood et al. (forthcoming), an indicator is constructed to quantify the effects of fire-sale spillovers in terms of losses in equity capital in the banking system. For some countries, loans to monetary financial institutions are the most systemic assets, while for others loans to households can pose systemic risks. Thanks to the fine granularity of the background data and monthly updates, the index can be used as an early warning indicator and a measure of systemic risk.
INTROduCTION
The recent financial crisis has shown that a shock affecting a financial institution can propagate to other financial firms and jeopardise the stability of the whole financial system. One channel through which such contamination can spread is fire-sale spillovers.
The mechanics of such spillovers can be described as follows. As documented in a number of studies, financial firms often target leverage.2 When a bank experiences an adverse shock to its equity capital which increases its leverage, one way for the bank to return to the target leverage is to shed assets and pay off debt. At times when market liquidity is scarce or an asset is illiquid, a financial institution which is forced to liquidate that asset may depress its price. As a consequence, other financial institutions holding the same asset (or assets of the same asset class) will suffer a loss, even if they do not have direct links with the firms initiating the (fire) sale. Affected financial institutions may, in turn, sell other assets to bolster their balance sheets. Therefore, common asset exposures can result in contagion, even between seemingly unrelated assets and banks.
Fire sales and the ensuing liquidation spirals have received extensive attention in the literature and are believed to have contributed significantly to systemic risk in the financial system.3 The paper of Greenwood, Landier and Thesmar (Greenwood et al. (forthcoming)) proposes a framework to quantify such fire-sale externalities.4
By using individual bank balance sheet data, this special feature provides an aggregate vulnerability (AV) indicator for euro area banks which is based on the framework developed by Greenwood et al. This vulnerability indicator measures how much equity capital in the banking system is wiped out after a shock and when liquidation spirals occur.
The results of the analysis show that losses arising from asset fire sales can be large. The average value of fire-sale externalities after a 1% shock to assets throughout the sample is 37% of total euro area banking system equity. The AV index reaches its peak in autumn 2008, coinciding with the intensification of the financial crisis after the failure of Lehman Brothers. The outbreak of the euro area sovereign debt crisis in 2010 is also captured. Importantly, it is found that for some countries the
1 Prepared by Lorenzo Cappiello and Dominik Supera. 2 See, for instance, Adrian, T. and Shin, H., “Liquidity and Leverage”, Journal of Financial Intermediation, Vol. 19, No 3, July 2010, pp. 418-437. 3 See, for example, Shleifer, A. and Vishny, R., “Liquidation Values and Debt Capacity: A Market Equilibrium Approach”, Journal
of Finance, Vol. 47, No 4, 1992, pp. 1343-1366; Shleifer, A. and Vishny, R., “Fire Sales in Finance and Macroeconomics”, Journal of Economic Perspectives, Vol. 25, No 1, 2011, pp. 29-48; Brunnermeier, M. and Pedersen, L., “Market Liquidity and Funding Liquidity”, The Review of Financial Studies, Vol. 22, Issue 6, June 2009, pp. 2201-2238; and Allen, F., Babus, A. and Carletti, E., “Asset commonality, debt maturity and systemic risk”, Journal of Financial Economics, Vol. 104, Issue 3, 2012, pp. 519-534.
4 Greenwood, R., Landier, A. and Thesmar, D., “Vulnerable Banks”, Journal of Financial Economics, forthcoming.
The mechanics of fire sales
Losses arising from asset fire sales can be large
100 ECB Financial Stability Review November 2014100
most systemic assets in the banking system are loans to monetary financial institutions (MFIs), while for others loans to households can pose systemic risks. However, asset “systemicness” differs across countries. The framework applied in this study can also be used to analyse the systemicness of specific assets. For example, when assuming a 25% write-off on a given set of countries’ government bonds, the AV index increases well before the outbreak of the sovereign debt crisis.
The findings have important policy implications. First, the analysis sheds light on the importance of monitoring leverage as a complement to capital requirements. Second, it shows that systemic risk can build up when certain assets in the banking system keep on growing, even if leverage remains approximately constant.5 This suggests that in some cases a mere (relatively rapid) expansion of assets can pose risks to financial stability. Third, the study shows that banks in different countries can be vulnerable to different asset classes. This indicates that policy measures aimed at guaranteeing financial stability should be calibrated to the specific characteristics of different jurisdictions, a lesson which is very relevant for the euro area. Finally, since fire-sale spillovers can propagate across countries, it is essential that policy measures are coordinated internationally.
Greenwood et al. apply their framework to produce measures of the contribution of each bank to systemic risk, the interconnectedness between two banks and an AV indicator. In particular, using commercial bank exposures provided by the European Banking Authority’s July 2011 stress test, Greenwood et al. analyse the 2010-11 sovereign debt crisis and estimate the potential spillovers following the significant haircuts experienced by a set of European sovereigns. Furthermore, they evaluate the outcome of various policies aimed at reducing fire-sale spillovers during the crisis, i.e. forced mergers among the most exposed banks and equity injections.6
In a related work, Duarte and Eisenbach (2014) implement the Greenwood et al. framework7 to construct the time series of a systemic risk measure that quantifies vulnerability owing to fire- sale spillovers using the regulatory balance sheet data for US commercial banks. Not surprisingly, their measure reaches a peak in Q4 2007 and spikes again in Q3 2008 but, interestingly, it starts to increase already in 2004, showing its relevance as an early warning indicator.
5 In this case equity capital grows at the same pace as assets. 6 While forced mergers would not have substantially reduced systemic risk, equity injections can significantly decrease banking sector
vulnerability. 7 For more details on the methodology, see Greenwood et al., op. cit.; Duarte, F. and Eisenbach, T., “Fire-Sale Spillovers and Systemic
Risk”, Federal Reserve Bank of New York Staff Reports, No 645, 2014.
Policy implications
Box A.1
ThEORETICAL FRAMEWORk
To evaluate spillover losses, this special feature adopts the framework proposed by Greenwood et al. and assumes that banks are hit by a hypothetical shock, which either erodes their returns on assets or their equity capital. This will give rise to the liquidation spirals discussed in the Introduction above. In line with Greenwood et al., the framework is based on three hypotheses. First, it is assumed that banks target a given leverage and that, after a shock, they will sell assets in order to return to that target leverage. This leverage-targeting hypothesis is in line with
101 ECB
Financial Stability Review November 2014 101
SPECIAL FEATURE A
empirical evidence from, for example, Adrian and Shin (2010),1 who show that banks manage book leverage to offset shocks to asset values. Second, it is assumed that banks, after the initial shock, will sell assets proportionally to their existing holdings. The third assumption is that asset sales generate a price impact of 10 basis points per €10 billion worth of assets sold. This assumption is in line with Amihud (2002),2 who shows that this figure is close to the liquidity of a broad spectrum of stocks. Since most of the assets considered are less liquid than stocks, the price impact generated by the model is likely to be at a lower bound.
To understand the intuition of the model, it is useful to consider the following steps in the sequence of events occurring in a fire sale. The framework adopted quantifies each of those steps. The algebra is worked out in Greenwood et al. and Duarte and Eisenbach (2014).3
1) Initial stage (bank j)
A population of N banks and K assets is considered. For simplicity, it is assumed that N = 2 (indexed by j and h) and K = 3 (indexed by X, Y and Z). At time t = 0, bank j has total assets Aj,0, total liabilities (excluding capital) Lj,0 and total capital Ej,0. It is also assumed that at time t = 0, bank j’s asset holding is given by Xj,0, Yj,0, and Zj,0. Part 1 of the table below shows the balance sheet of bank j at time t = 0. For illustrative purposes, throughout the time periods of the exercise, we assume that bank h holds only assets Y and Z.
2) Initial shock and direct losses (bank j)
At time t = 1, a shock occurs that wipes out 50% of asset X value. As a result, bank j incurs direct losses since the value of asset X decreases from Xj,0 = €50 billion to Xj,1 = €25 billion. At the same time, its capital is eroded by the same amount from Ej,0 = €50 billion to Ej,1 = €25 billion. Part 2 of the table presents the balance sheet of bank j at time t = 1, after the shock. As a result of the haircut, the leverage of bank j increases from Levj,0 = Lj,0 / Ej,0 = 3 at time t = 0 to Levj,1 = 6 at time t = 1. To keep leverage constant at the level prevailing before the shock, bank j sells SOj,1 = Levj,0 * (Ej,0 – Ej,1) = €75 billion. Since it is assumed that bank h does not hold asset X, it will not be subject to the direct losses stemming from the initial shock.
3) Asset sales (bank j)
At time t = 2, bank j sells its assets proportionally to its holding at time t = 1:
– Asset X: SOj,1*Xj,1 / Aj,1 = €10.71 billion
– Asset Y: SOj,1*Yj,1 / Aj,1 = €21.42 billion
– Asset Z: SOj,1*Zj,1 / Aj,1 = €42.86 billion
Part 3 of the table reports the balance sheet of bank j at time t = 2.
1 Adrian and Shin (2010), op. cit. 2 Amihud, Y., “Illiquidity and stock returns: cross-section and time-series effects”, Journal of Financial Markets, Vol. 5, Issue 1,
2002, pp. 31-56. 3 Op. cit.
102 ECB Financial Stability Review November 2014102
4) Price impact (bank j)
Bank j’s asset sell-off affects the prices of assets at time t = 3. Assuming that the price impact is 10 basis points per €10 billion worth of assets sold, the liquidation of assets by bank j has the following price impact:
– Asset X: 14.29 basis points = 0.1429%
– Asset Y: 28.58 basis points = 0.2858%
– Asset Z: 57.14 basis points = 0.5714%
Therefore, bank j incurs additional losses stemming from the adverse price impact. The value of assets in the balance sheet of bank j decreases accordingly (as in step 2):
– Asset X: Xj,3 = Xj,2*(1 – 0.5714%) = €14.27 billion
– Asset Y: Yj,3 = Yj,2*(1 – 0.2858%) = €28.50 billion
– Asset Z: Zj,3 = Zj,2*(1 – 0.5714%) = €56.81 billion
Part 4 of the table reports bank j’s balance sheet at the end of time t = 3. The decrease in the value of assets – which takes into account the effects stemming from the price impact – triggers a second-round sell-off of assets (as in step 3).
Balance sheet of banks j and h throughout the sample
1. Initial stage, t = 0, bank j 2. Initial shock and direct losses, t = 1, bank j Assets: Liabilities: Assets: Liabilities:
Xj,0 = 50 Lj,0 = 150 Xj,1 = 25 Lj,1 = 150 Yj,0 = 50 Capital: Yj,1 = 50 Capital: Zj,0 = 100 Ej,0 = 50 Zj,1 = 100 Ej,1 = 25
3. Asset sales, t = 2, bank j 4. Price impact, t = 3, bank j Assets: Liabilities: Assets: Liabilities:
Xj,2 = 14.29 Lj,2 = 75 Xj,3 = 14.27 Lj,3 = 75 Yj,2 = 28.58 Capital: Yj,3 = 28.5 Capital: Zj,2 = 57.14 Ej,2 = 25 Zj,3 = 56.81 Ej,3 = 24.86
5. Initial stage, t = 2, bank h 6. Spillover losses, t = 3, bank h Assets: Liabilities: Assets: Liabilities:
Yh,2 = 75 Lh,2 = 100 Yh,3 = 74.79 Lh,2 = 100 Zh,2 = 50 Capital: Zh,3 = 49.71 Capital:
Eh,2 = 25 Eh,2 = 24.5
103 ECB
Financial Stability Review November 2014 103
SPECIAL FEATURE A
AppLICATION OF ThE MOdEL TO EuRO AREA BANkS
A framework based on that proposed by Greenwood et al. is implemented using granular balance sheet data for a large sample of euro area banks.8 Observations span from July 2007 until May 2014 at monthly frequency. The total AV index is computed at each point in time. Assuming an initial shock such that all assets decrease in value by 1%, the AV index is defined as the fraction of total banking system equity capital which would be wiped out owing to direct and second-round effects of fire-sale spillovers9 (see Chart A.1).
The index increases steadily from around 39% of banking system equity capital in July 2007 until it reaches its peak in September 2008 at 52.5%, at the time of the Lehman Brothers failure.
8 For the purpose of this special feature, we use confidential balance sheet panel data for the 177 largest euro area credit institutions. 9 The purpose of the exercise is not to identify the shock but to show the effects of the decrease in the value of the assets on equity.
The vulnerability index reaches its peak in September 2008 and decreases thereafter
5) Initial stage (bank h)
Since it is assumed that bank h holds only assets Y and Z, it will not be affected by the initial shock to asset X. However, bank j’s asset sell-off determines a price impact which affects bank h because of the decline in the value of assets Y and Z observed in step 4. In this example, the target leverage of bank h is assumed to be equal to 4. Therefore, the price impact on assets Y and Z triggers a sale of assets by bank h as well. Part 5 of the table shows bank h’s balance sheet at time t = 2.
6) Spillover losses (bank h)
The price impact determines the spillover losses to all banks holding assets of the same asset class as those sold. In this example, in order to keep leverage constant, bank h needs to sell a share of its assets (as in step 3). This action decreases the price of those assets which are sold off and triggers a liquidation spiral in the banking system. The result of the price impact through bank j’s sales is that the value of bank h’s assets will decrease as shown in the calculations for step 4:
– Asset Y: Yh,3 = Yh,2*(1 – 0.2858%) = €74.79 billion
– Asset Z: Zh,3 = Zh,2*(1 – 0.5714%) = €49.71 billion
Part 6 of the table shows bank h’s balance sheet after spillover losses. To keep leverage constant at the level prevailing before the shock (i.e. Levh,1 = 4), bank h needs to sell €2 billion worth of assets as described in step 3.
Chart A.1 Aggregate vulnerability index
(July 2007 – May 2014; fraction of total banking system equity)
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
2007 2008 2009 2010 2011 2012 2013 2014 0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
Sources: ECB and ECB calculations.
104 ECB Financial Stability Review November 2014104
The AV index then follows a downward sloping trend, with a spike in May 2010 capturing the outbreak of the sovereign debt crisis in the euro area. From September 2011 until May 2012 this trend comes to a halt and the index stabilises at around 35%, most likely reflecting the spread of the sovereign debt crisis within the euro area. Thereafter the index decreases almost continuously.
dECOMpOSITION OF ThE AggREgATE VuLNERABILITY INdICATOR
To understand the factors determining the extent of the spillover losses and how they vary over time, the AV index is decomposed into three components: the system assets (i.e. the total size of the assets in the banking system), the system leverage (i.e. the average leverage weighted by total liabilities), and the illiquidity concentration. The first factor is a relevant determinant of the AV index since the larger the size of the assets in the system, the larger the overall price effects. The system leverage contributes more than proportionately to the AV indicator because, for a given shock, the more highly leveraged the system, the larger the fire sales and, for a given fire sale, the larger the spillover losses in terms of equity capital. The illiquidity concentration denotes a modified Herfindahl-Hirschman index for asset classes. This factor indicates that if a given asset is widely held in banks’ balance sheets and has a large aggregate share, if it is illiquid and concentrated in banks which are large, relatively highly leveraged and exposed to the initial shock, then that asset will contribute significantly to the vulnerability of the system (see Duarte and Eisenbach (2014)). Charts A.2 to A.4 plot the evolution of total assets, system-wide leverage and illiquidity concentration against the AV index.
This set of charts suggests that it is mainly the increase in asset size and, to a lesser extent, the rise in system leverage that drive the increase of the AV index from July 2007 to September 2008. In particular, a hypothetical shock would have its largest effect on the AV index when assets grow very rapidly (at around 1.1% on average per month) between July 2007 and September 2008. The
Systemic risk can build up when assets
keep on growing, even if leverage remains
constant…
… but it decreases when banking system
leverage falls and assets grow at a slower
pace
Chart A.2 System assets and aggregate vulnerability index
(July 2007 – May 2014; EUR billions (left-hand scale); fraction of total banking system equity (right-hand scale))
0.22
0.27
0.32
0.37
0.42
0.47
0.52
10,000
10,500
11,000
11,500
12,000
12,500
13,000
13,500
system assets (left-hand scale) AV index (right-hand scale)
2007 2008 2009 2010 2011 2012 2013 2014
Sources: ECB and ECB calculations.
Chart A.3 Illiquidity concentration and aggregate vulnerability index
(July 2007 – May 2014; index: July 2007 = 100 (left-hand scale); fraction of total banking system equity (right-hand scale))
0.22
0.27
0.32
0.37
0.42
0.47
0.52
94
95
96
97
98
99
100
101
102
103
104
illiquidity concentration (left-hand scale) AV index (right-hand scale)
20082007 2009 2010 2011 2012 2013 2014
Sources: ECB and ECB calculations.
105 ECB
Financial Stability Review November 2014 105
SPECIAL FEATURE A
effect is smaller when assets grow at a slower pace (on average 0.27% per month) between October 2008 and March 2012. On the other hand, the substantial decrease in system leverage (from 16.5 in September 2008 to around 11.0 in May 2014) is largely responsible for the downward sloping trend of the AV index observed from September 2008. The illiquidity concentration is likely to contribute to the spike in the index observed in May 2010 and the interruption of the downward trend between September 2011 and May 2012, when the AV index tends to stabilise.
ASSET “SYSTEMICNESS”
The AV index is also decomposed according to the “systemicness” of each asset type – computing the contribution of an asset category to the aggregate vulnerability. Specifically, we consider the following question: how much equity capital would be lost owing to fire sales if a particular asset class were the only one that suffered a shock? Chart A.5 shows that the most systemic asset classes (with the average share in the index in parentheses) throughout our sample are loans to MFIs (31.3%), loans to households (18.5%) and loans to non-financial corporations with a maturity of over one year (13.3%). It should be pointed out that the contribution of each of these three asset classes to the index is different from their respective share in banks’ portfolios, namely 25.3%, 20.4% and 14.9%. This indicates that, besides the size of an asset class, it is its systemicness that plays an important role (i.e. the fact that the asset is held by systemic banks which are defined as those banks that are large and leveraged and, in turn, hold large proportions of other illiquid assets). It should be noted that the framework of fire-sale spillovers applies well to tradable assets, while less to loans. However, after an adverse shock to a given class of loan extended to a given sector, banks might reduce their lending to that sector. This increases the risk associated with that class of loan, which may trigger a further reduction in lending (or a tightening of lending standards, including an increase in loan interest rates). This can have a negative impact on the sector and backfire on the banks themselves in a self-reinforcing spiral. As a result, banks could further reduce lending and fire-sale (tradable) assets in their portfolio. Following this reasoning, even though loans are relatively illiquid assets, the framework of Greenwood et al. could still be applied to a bank’s entire balance sheet.10
Furthermore, the AV index is decomposed according to the “systemicness” of the banking sector of each euro area Member State. This enables the estimation of the contribution of each country’s banking sector to euro area banking sector fragility. As shown in Chart A.6, banks in group 1 countries – Austria, Belgium, Finland, France, Germany, Luxembourg, the Netherlands and Slovenia – contribute the most to the AV index (70.7%), while the contribution to the index of banks in group 2 countries – Greece, Ireland, Italy, Portugal and Spain – is on average smaller (20.3%).11
10 Indeed Greenwood et al. and Duarte and Eisenbach (2014) also apply the framework to loans. See also Ramcharan, R. and Rajan, R. “Financial Fire Sales: Evidence from Bank Failures”, Finance and Economics Discussion Series, Federal Reserve Board, June 2014.
11 Data for banks in Cyprus, Estonia, Malta and Slovakia were not available for the whole sample. Those banks are therefore excluded from the analysis.
Certain asset classes are more systemic than others
Group 1 countries make the largest contribution to the AV index
Chart A.4 System leverage and aggregate vulnerability index
(July 2007 – May 2014; fraction of total banking system equity)
0.22
0.27
0.32
0.37
0.42
0.47
0.52
11
10
12
13
14
15
16
17
18
system leverage (left-hand scale) AV index (right-hand scale)
2007 2008 2009 20112010 2012 2013 2014
Sources: ECB and ECB calculations.
106 ECB Financial Stability Review November 2014106
It is worth mentioning that the contribution of the banking sector of each country group is different from the share of that group’s banking sector, as the share is computed as a fraction of the total euro area banking sector assets. The share of the group 1 countries is 59.6%, while that of the group 2 countries is 31.2%. Thus, the group 1 countries’ banks are more systemic than banks in group 2 countries, not only because they have the largest share of assets but also because they hold a large proportion of illiquid assets.
The framework also enables vulnerability indices to be constructed grouping those countries’ banking sectors that share similar sources of fragility. As shown from the breakdown into asset classes of the AV index, the most systemic assets are loans to MFIs and households. The analysis shows that the countries mostly exposed to loans to MFIs are mainly group 1 countries, while the countries mostly exposed to loans to households are mainly group 2 countries. The share of loans to MFIs in the AV index for the first group of countries is on average equal to 39.3%, while the share of loans to households in the AV index for the second group is on average equal to 30%. The indices for both groups of countries are reported in Charts A.7 and A.8 and are characterised by a similar pattern.
The countries mostly exposed to loans to MFIs
are mainly group 1 countries, while the
countries mostly exposed to loans to households are
mainly group 2 countries
Chart A.5 Contribution of each asset to the aggregate vulnerability index
(July 2007 – May 2014; fraction of total banking system equity)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.0
0.1
0.2
0.3
0.4
0.5
0.6
shares sec_gov_dom sec_gov_ea sec_nonmfi sec_mfi loans_NFC_o1
loans_NFC_u1 loans_otherFI loans_HH loans_gov loans_mfi
2007 2008 2009 2010 2011 2012 2013
Sources: ECB and ECB calculations. Notes: loans_mfi: loans to MFIs; loans_HH: loans to households; loans_NFC_o1: loans to non-financial corporations with maturity over one year; loans_gov: loans to general government; loans_otherFI: loans to other financial institutions, pension funds and insurance corporations; loans_NFC_u1: loans to non-financial corporations with maturity up to one year; sec_mfi: securities of MFIs; sec_nonmfi: securities of non-MFIs (excluding general government); sec_gov_ea: securities of euro area governments (excluding the reference area); sec_gov_dom: securities of domestic government; shares: shares and other equities.
Chart A.6 Aggregate vulnerability index – country breakdown
(July 2007 – May 2014; fraction of total banking system equity)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.0
0.1
0.2
0.3
0.4
0.5
0.6
2007 2008 2009 2010 2011 2012 2013 2014
group 2 countries group 1 countries
Sources: ECB and ECB calculations. Notes: Group 1 countries refers to Austria, Belgium, Finland, France, Germany, Luxembourg, the Netherlands and Slovenia. Group 2 countries refers to Greece, Ireland, Italy, Portugal and Spain. Data for banks in Cyprus, Estonia, Malta and Slovakia were not available for the whole sample. Those banks are therefore excluded from the analysis.
107 ECB
Financial Stability Review November 2014 107
SPECIAL FEATURE A
The AV indices plotted in Charts A.7 and A.8 both reach a peak in autumn 2008 and then show a downward sloping trend. The difference between the indices for the two country groups mainly stems from the size of the fire-sale externalities. In the case of the first group (consisting mostly of group 1 countries), a 1% reduction in the value of all assets in the banking system would have wiped out around 62% of total equity capital at its peak in autumn 2008 and 46% on average throughout the sample. In the case of the second group (consisting mostly of group 2 countries), the direct effects and fire-sale externalities are of a lower magnitude – 39% at the peak and 26% on average.
EFFECTS OF AN AdVERSE ShOCk TO SOVEREIgN BONdS
Finally, the last experiment studies a bank’s susceptibility to the deleveraging cycle caused by a potential write-down of sovereign bonds. Echoing a similar exercise carried out by Greenwood et al., a 25% write-off in the value of Greek, Irish, Italian, Portuguese and Spanish government debt is considered.12 The data used provide information on banks’ exposure to the sovereign debt
12 The size of the assumed write-off is in line with the maximum drop in price observed for Spanish and Italian government bonds between August 2010 and November 2011, which was 21.8% and 29.5% respectively. By way of comparison, the price of Greek and Portuguese government bonds fell by 97.6% and 61.1% respectively between November 2009 and February 2012. The value of Irish government bonds decreased by 48.7% between November 2009 and June 2011.
Assuming a substantial write-off of group 2 countries’ government bonds, the AV index increases well before the outbreak of the sovereign debt crisis
Chart A.7 Aggregate vulnerability index specific to countries with loans to MFIs being the most systemic asset (July 2007 – May 2014; fraction of total banking system equity)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
2007 2008 2009 2010 2011 2012 2013
shares sec_gov_dom sec_gov_ea sec_nonmfi sec_mfi loans_NFC_o1
loans_NFC_u1 loans_otherFI loans_HH loans_gov loans_mfi
Sources: ECB and ECB calculations. Notes: loans_mfi: loans to MFIs; loans_HH: loans to households; loans_NFC_o1: loans to non-financial corporations with maturity over one year; loans_gov: loans to general government; loans_otherFI: loans to other financial institutions, pension funds and insurance corporations; loans_NFC_u1: loans to non-financial corporations with maturity up to one year; sec_mfi: securities of MFIs; sec_nonmfi: securities of non-MFIs (excluding general government); sec_gov_ea: securities of euro area governments (excluding the reference area); sec_gov_dom: securities of domestic government; shares: shares and other equities.
Chart A.8 Aggregate vulnerability index specific to countries with loans to households being the most systemic asset (July 2007 – May 2014; fraction of total banking system equity)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
shares sec_gov_dom sec_gov_ea sec_nonmfi sec_mfi loans_NFC_o1
loans_NFC_u1 loans_otherFI loans_HH loans_gov loans_mfi
2007 2008 2009 2010 2011 2012 2013
Sources: ECB and ECB calculations. Notes: loans_mfi: loans to MFIs; loans_HH: loans to households; loans_NFC_o1: loans to non-financial corporations with maturity over one year; loans_gov: loans to general government; loans_otherFI: loans to other financial institutions, pension funds and insurance corporations; loans_NFC_u1: loans to non-financial corporations with maturity up to one year; sec_mfi: securities of MFIs; sec_nonmfi: securities of non-MFIs (excluding general government); sec_gov_ea: securities of euro area governments (excluding the reference area); sec_gov_dom: securities of domestic government; shares: shares and other equities.
108 ECB Financial Stability Review November 2014108
of their own country of residence. However, the dataset only shows the banks’ holdings of aggregate foreign sovereign debt. For example, one cannot observe how much German or French sovereign debt is held by an Italian bank; only the total foreign euro area debt held by the Italian bank can be observed. To circumvent this data limitation, it is assumed that banks in the group 1 countries hold a share of group 2 countries’ government bonds equal to the share of outstanding public debt of group 2 countries in the total public debt of all of the euro area countries considered in this study.13 Chart A.9 plots the vulnerability indices after a 25% drop in the value of group 2 countries’ government bonds for i) the group 1 countries’ banks and ii) the group 2 countries’ banks only.14
The AV index for the banking system for group 1 countries remains stable at around 20-25% until May 2010. The index thereafter exhibits a downward sloping trend, which stabilises at about 10%. On the other hand, the AV index for the banks in group 2 countries increases well before the outbreak of the sovereign debt crisis – from 13% in July 2007 and reaching a peak of 31% in May 2010. It then decreases before rising again in April 2012 with the second wave of the sovereign debt crisis. The index has decreased since April 2013 as confidence in the group 2 countries improves.
CONCLudINg REMARkS
Using a simple framework and detailed balance sheet data for euro area banks, this special feature finds that spillover losses from fire sales can be large. The average value of fire-sale externalities throughout the sample from July 2007 until June 2014 is 37% of the total euro area banking system equity capital. Loans to MFIs, loans to households, and loans to firms with a maturity of over one year are the most systemic assets.
The AV index proposed can be used as a systemic risk measure and an early warning indicator. Its main advantage is that it is based on individual banks’ balance sheet data. The fine granularity offered by balance sheet data provides a detailed overview of the evolution, composition and determinants of fire-sale vulnerability in the euro area banking sector. Furthermore, since the dataset underlying the analysis can be updated at a monthly frequency, the AV index is well suited for timely monitoring.
13 It should be noted that this assumption can affect the results of the exercise. The conclusions should therefore be interpreted with caution. 14 The two AV indices are computed as the fraction of group 1 and group 2 countries’ banking system equity capital respectively.
Chart A.9 Vulnerability indices after a 25% drop in the value of group 2 countries’ government bonds (July 2007 – May 2014; fraction of total banking system equity)
0.08
0.12
0.16
0.20
0.24
0.28
0.32
0.08
0.12
0.16
0.20
0.24
0.28
0.32
2007 2008 2009 2010 2011 2012 2013 2014
vulnerability index for group 2 countries vulnerability index for group 1 countries
Sources: ECB and ECB calculations. Notes: Group 1 countries refers to Austria, Belgium, Finland, France, Germany, Luxembourg, the Netherlands and Slovenia. Group 2 countries refers to Greece, Ireland, Italy, Portugal and Spain. Data for banks in Cyprus, Estonia, Malta and Slovakia were not available for the whole sample. Those banks are therefore excluded from the analysis.
109 ECB
Financial Stability Review November 2014 109
SPECIAL FEATURE B
B CApTuRINg ThE FINANCIAL CYCLE IN EuRO AREA COuNTRIES1
This special feature discusses ways of measuring financial cycles for macro-prudential policy- making. It presents some estimates and empirical characteristics of financial cycles. Existing studies on financial cycle measurement remain quite nascent in comparison with the voluminous literature on business cycles. In this context, two approaches – turning point and spectral analysis – are used to capture financial and business cycles at the country level. The results of the empirical analysis suggest that financial cycles tend to be more volatile than business cycles in the euro area, albeit with strong cross-country heterogeneity. Both aspects underscore the relevance of robust financial cycle estimates for macro-prudential policy design in euro area countries.
INTROduCTION
Attenuating financial cycles is one of two fundamental goals of macro-prudential policy.2 Indeed, the recent global financial crisis provided a vivid illustration of the time series dimension of systemic risk – namely, that recessions associated with the build-up of financial sector disruptions exhibit much higher output losses than “normal” recessions.
Despite the prominence of this goal in macro-prudential policy, there is no generally agreed definition of the financial cycle3, and existing measurement methods yield only preliminary and incomplete results. Existing analysis on characterising financial cycles remains scarce and is in many ways not yet suitable for policy use in the euro area. Measurement limitations include the geographic coverage of the analysis (in that it tends to focus on a limited number of countries) and a lack of consensus on the mechanics of measurement, such as the choice of indicators and the method used to construct them. Ideally, a unique synthetic measure of the financial cycle would summarise the (co-)movements over time of a range of finanical sector variables, covering quantities and prices. In practice, however, over-reliance on a single composite measure is not advisable as each constituent variable contains relevant information for macro-prudential policy-making.
Measuring financial cycles for euro area countries has become more important in the context of ECB macro-prudential oversight and the launch of the Single Supervisory Mechanism (SSM). There is an urgent need to obtain a robust view on capturing financial cycles – balancing cross-country consistency with individual country relevance. This special feature presents the results of two different methodologies aimed at furthering the basis for country-specific macro-prudential policy- making in the euro area. One employs spectral methods for cycle extraction and characterisation for euro area countries, and the other characterises financial cycles on the basis of turning point analysis. Both approaches incorporate information from several macro-financial variables typically used in the growing body of literature to robustly capture the financial cycle across a diverse set of countries, and present relationships with business cycles extracted on a comparable methodological basis. In so doing, they provide information on differing properties of financial cycles across euro area countries, including their amplitude and persistence.
1 Prepared by Paul Hiebert, Benjamin Klaus, Tuomas Peltonen, Yves S. Schüler and Peter Welz. 2 The two commonly thought of goals are (i) attenuating the financial cycle and (ii) enhancing resilience of the financial system. See, for
instance, the Group of Thirty, Enhancing Financial Stability and Resilience: Macroprudential Policy, Tools, and Systems for the Future, Working Group on Macroprudential Policy, October 2010 (http://www.group30.org/images/PDF/Macroprudential_Report_Final.pdf).
3 One appealing characterisation of financial cycles relates to the pro-cyclicality of the financial system inherent in “self-reinforcing interactions between perceptions of value and risk, attitudes towards risk and financing constraints, which translate into booms followed by busts”. See Borio, C., “The financial cycle and macroeconomics: What have we learnt?”, Journal of Banking and Finance, Vol. 45, August 2014, pp. 182-198.
Need for measures of the financial cycle…
… to support macro- prudential policy- making
110 ECB Financial Stability Review November 2014110110
EXISTINg STudIES ON ThE FINANCIAL CYCLE: pOINT OF dEpARTuRE
Whereas the business cycle has been studied extensively, the comparable body of literature on financial cycles remains nascent. In general, cycles can be measured as classical cycles considering the level of the underlying time series, as growth cycles by removing a permanent component from the series under study, and as cycles in growth rates where the underlying time series is first transformed into growth rates.4 Of the studies which have been seminal in laying the ground work for a better understanding of financial cycles across major economies, two strands stand out as representative.
The first consists of turning point analysis applied to cycle extraction – examining key descriptive characteristics such as duration, amplitude and slope. An influential study in this respect is Claessens et al. (2012),5 which examines the phases of business and financial cycles and the resulting impact on macroeconomic performance for a broad range of advanced and emerging economies. Their findings suggest that, while financial variables tend to exhibit more variability than those related to the business cycle, this differs across financial assets such as equity (which has the longest upturn duration) and real estate variables (with housing exhibiting the longest downturn duration). Furthermore, their findings show that there is a close link between business and financial cycles.
The second strand focuses on frequency-based filters, in some cases complemented by turning point analysis. A widely cited study in this respect is Drehmann et al. (2012),6 which uses both frequency-based filters and turning point analysis to identify financial cycles for several advanced economies and compares them with business cycles. The study finds that financial cycles are considerably longer than business cycles and that financial cycle peaks tend to be associated with financial crises.7 Looking specifically at credit, Aikman et al. (forthcoming) apply a frequency- based filter to extract cyclical dynamics using very long time series and find that financial cycles tend to be longer than their economic analogue.8
While the available literature has greatly contributed to developing key methodologies, there is less consensus regarding which variables best help to capture the cycle and how to combine multiple variables. Moreover, systematic analysis has been limited for euro area countries.
WhICh VARIABLES COuLd CApTuRE ThE FINANCIAL CYCLE?
Absent a single summary measure for the state of the financial sector, a multivariate approach that relies on a range of macro-financial indicators seems to provide the best-suited method for obtaining a financial cycle estimate. Such an approach would map the methodology used in a large body of business cycle research that goes beyond using real GDP as a summary indicator and instead
4 For a classification of cycle measurement approaches, see Harding, D. and Pagan, A., “A suggested framework for classifying the modes of cycle research”, Journal of Applied Econometrics, Vol. 20, Issue 2, 2005, pp. 151-159.
5 See Claessens, S., Kose, M. and Terrones, M., “How do business and financial cycles interact?”, Journal of International Economics, Vol. 87, Issue 1, 2012, pp. 178-190.
6 See Drehmann, M., Borio, C. and Tsatsaronis, K., “Characterising the financial cycle: don’t lose sight of the medium term!”, BIS Working Paper, No 380, June 2012.
7 See Stremmel, H. and Zsámboki, B., “The relationship between structural and cyclical features of the EU financial sector”, Banking Structures Report, ECB, October 2014, where a frequency-based filter set up by Drehmann et al. (ibid.) is used to conduct a turning point analysis on the filtered series and map structural features of the banking sector on the financial cycle for EU Member States. Their findings suggest that structural banking sector characteristics influence the amplitude of the financial cycle.
8 See Aikman, D., Haldane, A. and Nelson, B., “Curbing the credit cycle”, The Economic Journal (forthcoming).
Nascent literature on financial cycles (in
contrast to business cycles)…
… including turning point analysis…
… and spectral analysis
No single indicator can capture the
financial cycle…
111 ECB
Financial Stability Review November 2014 111
SPECIAL FEATURE B
111
employs a broad set of indicators on economic activity.9 For instance, additional variables such as prices of goods and services (consumer price inflation) and the price of intertemporal substitution of consumption and investment (interest rates) should contain important information for a more accurate measurement of business cycles than reliance on any single variable such as GDP.
In conceptualising the determinants of the financial cycle, studies to date have focused predominantly on a combination of various measures of credit, e.g. transformations of total or bank-based credit, either in growth rates or credit to GDP ratios, as well as asset prices, especially real estate and equity prices. Measures of credit can give an impression of financial flows that form a conceptual analogue to flows of goods and services in business cycle research.
According to Schularick and Taylor (2012), the entire bank balance sheet and its decomposition may have macroeconomic implications.10 Notably, variation in leverage (proxied by credit) is related to asset price developments.11 With regard to the latter, while there are numerous possible proxies to capture asset price movements, a combination of residential property price indices, equity price indices and a measure of benchmark bond yields can provide a basis for capturing all main asset market segments.
Ultimately, it could be argued that a good starting point is an analysis of a parsimonious set of three or four variables for the financial cycle (credit, house prices, equity prices – as is standard in the literature to date – complemented by a country-specific benchmark interest rate as a proxy for bond market pricing) and three for the business cycle (GDP, consumer price inflation and the interest rate). Table B.1 summarises the key characteristics of these quarterly series in real terms for ten euro area countries since 1970.12
When looking at the characteristics of the series on a cross-country basis, it is clear that the indicators for the financial cycle tend to be more volatile than the indicators for the business cycle. This applies broadly to credit growth, house price growth and – in particular – equity price growth. By contrast, the mean growth rates of real GDP and inflation have been less volatile over approximately the last 40 years, albeit less stable in the early years for which data are available, mainly the 1970s.
Another important stylised fact relates to the considerable cross-country heterogeneity – arguably stronger for variables characterising the financial cycle (such as credit and asset prices) than those characterising the business cycle (such as GDP). Average total real credit growth, for instance, has been in the range of 4-5% per annum in most countries in the euro area since 1970, but rather low in Germany (at around 2.5%) and rather high in Ireland (in excess of 7%) over the same period. Similarly, while average real house price growth in these ten euro area countries has been around 1.5% since 1970, real house prices have actually fallen on average in Germany and nearly stagnated in Portugal. Likewise, average cross-country real equity price growth has been
9 For arguments in favour of using a variety of measures in the business cycle context, see, for example, Boehm, E., “A review of some methodological issues in identifying and analysing business cycles”, Melbourne Institute Working Paper, No 26/98, November 1998. See also Stock, J. and Watson, M.W., “Estimating Turning Points Using Large Data Sets”, Journal of Econometrics, Vol. 178, 2014, pp. 368-381.
10 See Schularick, M. and Taylor, A., “Credit Booms Gone Bust: Monetary Policy, Leverage Cycles, and Financial Crises, 1870–2008”, American Economic Review, Vol. 102, Issue 2, 2012, pp. 1029–1061.
11 Clearly, these variables represent a compromise to the ideal information set for financial cycle extraction. For example, it would be preferable to include key propagation mechanisms of systemic risk, such as actual leverage and maturity mismatch, but long time series at the country level for these series is unfortunately scarce for euro area countries.
12 Standard unit root tests suggest most variables in levels are integrated of order one in individual countries.
… but credit and asset prices are key ingredients for financial cycle estimation
Key variables for euro area countries…
… suggest financial cycle determinants are relatively volatile…
… as well as considerable cross-country heterogeneity
112 ECB Financial Stability Review November 2014112112
of a similar magnitude at 1.3%, but has declined in Spain, Italy and Portugal over the period. Real interest rate changes across the ten countries have been around zero on average since 1970, with less cross-country variation in averages, but a vast difference in terms of volatility or extremes. By contrast, real GDP growth and consumer price inflation rates have tended to be more homogeneous across the euro area countries over the past few decades.
METhOdS TO CApTuRE FINANCIAL CYCLES FOR SELECTEd EuRO AREA COuNTRIES
As financial cycles are not directly observable, they must be inferred. This gives rise to a potential for both data and model uncertainty and so the use of complementary analytical perspectives can enhance measurement and policy-making. Against this background, two commonly used types of methodology are presented below, which together provide a conceptually distinct but complementary means of cycle extraction.
Spectral analysis A widely applied method of extracting cycles is spectral analysis, which, simply put, means applying filters that exploit information on dominant frequencies in variables that capture respective cycles. To account for specification uncertainty, a rich set of alternatives is proposed in the literature, ranging from univariate analysis (of a single series) through to extracting cycles based on commonality across several variables (“cohesion”).
Results from two methods to infer financial cycles
Multivariate spectral analysis…
Table B.1 Summary of data for selected euro area countries
(annual percentage changes and percentage point changes)
Total credit House prices Equity prices Interest rates GDP Inflation
AT 4.9 (3.6) 2.5 (7.5) 1.6 (25.4) -0.1 (1.3) 2.4 (2.0) -0.0 (0.5) [-2.4,13.9] [-8.2,27.8] [-88.9,88.6] [-3.5,3.1] [-5.2,8.9] [-1.6,1.9]
BE 4.5 (4.3) 2.3 (5.5) 1.8 (20.5) -0.1 (1.9) 2.1 (2.0) -0.0 (0.7) [-10.7,12.2] [-14.3,11.5] [-71.8,44.8] [-7.7,6.0] [-4.4,7.0] [-2.0,2.8]
DE 2.6 (2.6) -0.2 (2.8) 2.3 (20.9) -0.1 (1.2) 2.0 (2.2) -0.0 (0.5) [-3.7,9.1] [-5.8,7.9] [-61.6,51.0] [-3.7,3.4] [-7.0,7.2] [-1.5,1.3]
ES 4.6 (6.2) 2.3 (9.7) -1.0 (26.4) 0.0 (2.7) 0.7 (5.0) -0.0 (1.0) [-9.1,18.3] [-14.0,31.0] [-71.5,74.6] [-9.7,10.2] [-23.7,10.5] [-3.4,5.1]
FI 4.3 (4.6) 1.3 (8.8) 4.5 (32.0) -0.1 (2.3) 2.4 (3.4) -0.0 (0.8) [-12.1,13.6] [-22.2,30.1] [-76.5,92.4] [-8.3,6.4] [-10.2,10.1] [-2.4,3.3]
FR 3.9 (3.2) 2.0 (5.2) 2.0 (22.7) -0.0 (1.4) 2.1 (1.7) -0.0 (0.6) [-2.1,11.9] [-9.3,13.0] [-56.3,51.3] [-4.7,3.9] [-4.0,5.6] [-2.4,2.8]
IE 7.5 (8.0) 2.0 (9.3) 1.8 (28.7) 0.1 (3.0) 1.7 (4.7) -0.1 (1.3) [-8.9,30.5] [-20.9,25.6] [-104.1,50.6] [-8.7,10.8] [-16.8,12.7] [-4.9,3.8]
IT 3.3 (4.6) 1.6 (10.1) -1.3 (29.0) -0.0 (3.0) 1.7 (2.6) -0.0 (1.0) [-6.9,11.9] [-20.2,48.4] [-66.6,89.2] [-8.7,12.9] [-7.2,9.4] [-3.9,3.5]
NL 5.5 (4.5) 1.9 (8.7) 1.7 (20.9) -0.1 (1.4) 2.2 (2.2) -0.0 (0.5) [-3.5,16.9] [-24.7,31.3] [-70.7,49.3] [-6.2,3.2] [-4.8,7.9] [-1.4,1.1]
PT 4.4 (6.2) 0.1 (3.0) -0.6 (46.9) 0.2 (6.0) 2.5 (3.4) -0.0 (2.4) [-13.1,18.3] [-5.6,8.2] [-198.2,127.6] [-34.7,34.3] [-6.5,11.3] [-11.8,14.2]
Avg. (Std.dev.) 4.6 (1.3) 1.6 (0.9) 1.3 (1.8) -0.0 (0.1) 2.0 (0.5) -0.0 (0.0)
Sources: Eurostat and ECB calculations. Notes: The table reports the mean, standard deviation (round brackets) and minimum as well as maximum value (square brackets). Quarterly variables are transformed to year-on-year changes and are in real terms, except for inflation. Real total credit, real house prices, real equity prices and real GDP are in annual percentage changes. Real interest rates and inflation reflect percentage point changes. Real interest rates represent deflated rates of long-term government bond yields. “Avg.” refers to the average of the country means and “Std.dev.” to the standard deviation of means across countries. The sample covers the period Q2 1972 – Q1 2014 (real total credit: Q2 1972 – Q4 2013). Exceptions are real house prices in the case of AT (starting Q3 1987), ES (starting Q1 1972), PT (starting Q1 1981), and BE/DE/DK/IT (ending Q4 2013). BE and FI (starting Q4 1971) and IE (starting Q2 1972) are the exceptions for real total credit. Real GDP ends in Q4 2013 for FI and IE. For ES and IE, industrial production (excluding construction) is used rather than GDP given the unit root properties of the latter series, likely associated with the services sector (starting Q1 1975 and Q1 1980 respectively).
113 ECB
Financial Stability Review November 2014 113
SPECIAL FEATURE B
113
The results from a multivariate spectral approach for characterising financial cycles at the country level are presented below. The cycles are extracted using a three-step procedure combining spectral analysis with principal component analysis.13 First, a cohesion measure capturing common movement regardless of the phase differences across variables is applied to capture the frequency range (or the length of cycles) with the highest co-movement across the set of indicators. Second, these country-specific frequency bands are used as an input into a band-pass filter to yield a continuous representation of country-specific cycles for each constituent indicator.14 Third, these cycles constructed for each individual indicator are aggregated into a common country-specific financial cycle through principal component analysis, with normalised indicators rebased to an average volatility across all series.
A key finding from the first step of this analysis is that, while providing further evidence to support the finding that financial cycles tend to be longer than business cycles, there is considerable heterogeneity across euro area countries. As indicated in Chart B.1, financial cycles measured on this basis indeed appear to be just under three times as long as the average business cycle of the ten euro area countries analysed – around 13 years as opposed to five years.15 But the dispersion of dominant country frequencies around this cross-country average is stark – with financial cycles lasting between seven and 17 years in contrast to business cycles lasting between three and eight years. Interestingly, the distribution of business cycle lengths seems to be skewed downwards compared with a relatively symmetric distribution for the financial cycle lengths, indicating a rather homogeneous cycle length for the business cycle. Clearly, there are some limitations to these data given structural changes over the last 40 years in many of the countries analysed, but the results are nonetheless illustrative.
Applying these dominant country frequencies can yield a composite estimate of a financial cycle for individual euro area countries, with a type of “concordance” across constituent explanatory variables around this principal signal. Specifically, filtering country credit, asset price (house and equity) and interest rate data using country-specific dominant frequencies yields a range for variables that are key underlying forces for the financial cycle – while the first principal component of these four variables yields a sort of “average financial cycle”, in the form of a linear combination of these individual cycles. Chart B.2 contains a representation of this output for an illustrative euro area country. The black line, representing the combination of cycles in individual variables at a given point in time, moves around a zero line representing deviations from a long-term historic average. The range of
13 For details, see Schüler, Y., Hiebert, P. and Peltonen, T., “Characterising financial cycles across Europe: one size does not fit all”, Working Paper Series, ECB (forthcoming).
14 See Christiano, L. and Fitzgerald, T., “The band-pass filter”, International Economic Review, Vol. 44, Issue 2, May 2003, pp. 435-465. 15 This broadly confirms the finding of Drehmann et al. (2012), who argue that the duration of the financial cycle is, on average, around 16
years, or 20 years when considering only cycles that peaked after 1998.
… as a basis for a three-step procedure to capture country financial cycles…
… suggesting relatively long (but heterogeneous) financial cycles…
… and enabling estimation of country-specific financial cycles…
Chart B.1 Estimated length of financial and business cycles for selected euro area economies (years; maximum; minimum; inter-quartile range and median)
0
2
4
6
8
10
12
14
16
18
0
2
4
6
8
10
12
14
16
18
Financial cycle Business cycle
Source: ECB calculations. Notes: Length of cycles refers to the point of maximum average cohesion across financial and real indicators along different frequencies. The sample includes data for ten euro area countries over Q1 1970 – Q1 2014, where available.
114 ECB Financial Stability Review November 2014114114
cycles around this dominant signal then provides information on the concordance, or agreement, of these individual cycles. Clearly, this tends to differ across time. From a policy perspective, such a signal can be important to understand both where one may stand in a financial cycle at a given point in time and the level of uncertainty attached to such a signal.
Mirroring heterogeneity in the source data across the euro area countries, the ability of a dominant principal component to capture movement across those variables meant to capture the financial cycle varies considerably across countries. Indeed, financial cycle determinants across countries appear to be more heterogeneous than modelled business cycle determinants. As shown in Chart B.3, credit and house prices appear to be dominant explanatory factors in many but not all countries, while equity prices and interest rates appear to be, in all but one case, less important. Clearly, country-specific lead/lag relationships across variables may affect the degree to which contemporaneous low frequency co-movement is present.
… with strong cross-country
differences
Chart B.3 Contributions to cyclical fluctuations across selected euro area countries
(percentage of total variance)
a) financial cycle b) business cycle
0 IE ES FI BE NL FR PT AT DE IT
10
20
30
40
50
60
0
10
20
30
40
50
60
total credit house prices equity prices interest rates
0
10
20
30
40
50
60
70
80
90
0
10
20
30
40
50
60
70
80
90
GDP inflation interest rates
FI IT IE AT FR PT BE ES DE NL
Source: ECB calculations. Notes: Figures reflect variance explained by a first principal component and the respective contributions of input variables. For the definitions and transformations of variables, see Table B.1. Variables have been standardised before obtaining the principal component.
Chart B.2 Illustrative financial cycle and cycles of constituent indicators
(quarterly data, standard deviation units from mean; rebased to series’ variance)
-35
-25
-15
-5
5
15
25
35
45
-35
-25
-15
-5
5
15
25
35
45
1987 1991 1995 1999 2003 2007 2011 2015
interest rates equity prices
house prices total credit financial cycle
Source: ECB calculations. Notes: The financial cycle is the first principal component of the four underlying smoothed and standardised indicators (coloured lines), namely interest rates, equity prices, residential property prices and total credit. Underlying indicators are smoothed using country-specific frequency windows as an input into a band-pass filter. The frequency window is determined via cohesion of indicators across cycle periods. The blue shaded area refers to banking crisis dates as specified in European Systemic Risk Board, “Operationalising the countercyclical capital buffer: indicator selection, threshold identification and calibration options”, Occasional Papers, No 5, June 2014. The yellow shaded area refers to forecasted cycles. Dates refer to the third quarter of the indicated year.
115 ECB
Financial Stability Review November 2014 115
SPECIAL FEATURE B
115
Turning point analysis The second type of methodology for cycle inference is turning point analysis. The approach followed in this special feature applies classical cycle measurement focusing on the level of the series as in Bry and Boschan (1971) and Harding and Pagan (2002).16 This type of turning point analysis has been mostly used to study the business cycle, but it has also been applied to financial series. Pagan and Sossounov (2003)17, for example, characterise the bull and bear market phases in equity prices; Claessens et al. (2012) analyse cycles in credit, house prices and equity prices; Drehmann et al. (2012) apply the algorithm to identify peaks and troughs in short-term and medium-term cycles of both GDP and financial series; and Bracke (2013) studies cycles in house prices.18
While a key advantage of turning point analysis is that it identifies the local minima and maxima in the levels of a series of interest, which is simple, transparent and robust to the inclusion of newly available data, a few parameters still need to be chosen. To ensure comparability with other studies, the analysis here uses for all series the common parameter settings as in Claessens et al. (2012). In particular, the initial turning points are searched within a window of two quarters and thereafter censoring rules of a minimum phase (complete cycle) length of two (five) quarters are applied.
The method is applied to investigate real credit, real residential property prices and real equity prices as the variables that may best capture information about the financial cycle from a macroeconomic perspective. In addition, turning points for the level of real GDP are determined to capture the business cycle. The sample includes the same ten euro area countries used in the spectral analysis.
Table B.2 summarises the results across the euro area countries of phase characteristics from turning point analysis, such as amplitude, duration and slope. The results reveal that, while the amplitude in the credit cycle is about twice as high as in the business cycle, they are roughly equally long. Property and equity price upturns tend to be shorter than those in credit and GDP. There is large country heterogeneity in real property price cycles, probably reflecting differing and complex
16 See Bry, G. and Boschan, C., Cyclical analysis of time series: Selected procedures and computer programs, National Bureau of Economic Research, Inc., 1971; and Harding, D. and Pagan, A., “Dissecting the cycle: a methodological investigation”, Journal of Monetary Economics, Vol. 49, Issue 2, 2002, pp. 365-381, where the former developed the methodology for monthly data and the latter adapted it to quarterly data.
17 See Pagan, A. and Sossounov, K., “A simple framework for analysing bull and bear markets”, Journal of Applied Econometrics, Vol. 18, Issue 1, 2003, pp. 23-46.
18 See Bracke, P., “How long do housing cycles last? A duration analysis for 19 OECD countries”, Journal of Housing Economics, Vol. 22, Issue 3, 2013, pp. 213-230.
Turning point analysis of financial variables in levels takes alternative view
Real credit, equity and housing prices are analysed separately and then compared
Table B.2 Summary of turning point analysis
(number of years; percentage changes; percentage changes per year)
Downturn Upturn Number
of phases Duration
(years) Amplitude Slope Number
of phases Duration
(years) Amplitude Slope
Real total credit 51 1.3 -4.0 -0.8 50 5.7 34.9 1.3 Real equity prices 106 1.6 -40.9 -6.8 103 1.8 48.3 7.0 Real residential property prices 65 2.0 -12.1 -1.4 66 3.3 22.2 1.8 Real GDP 64 1.0 -2.8 -0.6 56 5.6 18.1 0.7
Sources: BIS, OECD, Eurostat and ECB calculations. Notes: All statistics are computed over all countries included in the sample and separately for both cycle phases. Downturns (upturns) are defined as the phases between peak and trough (trough and peak). Duration measures the average length of a cycle phase in years. Amplitude refers to the average percentage change in a variable over upturns and downturns respectively. Slope refers to the ratio of amplitude to duration.
116 ECB Financial Stability Review November 2014116116
structural characteristics in regional housing markets, such as tax treatment of housing, macro- prudential and mortgage market features, and land and rental regulation.19
Based on the identified turning points, the extent to which cycles are synchronised, both within a given country and across countries, can also be determined. In particular, a useful measure is the concordance index proposed by Harding and Pagan (2002), which measures the fraction of quarters that two cycles are in the same phase.
Within countries, credit, business and housing cycles are the most strongly correlated, while equity cycles are much less synchronised with the other cycles (see Chart B.4), partly reflecting the close relationship observed between the business cycle and loans to the non-financial private sector. Specifically, this seems to be in line with the stylised fact observed for euro area aggregates that growth in loans to households, of which loans for house purchase constitute the largest fraction, is roughly coincident with growth in real GDP.20 As shown in Chart B.5, business, credit and equity cycles are more strongly correlated across countries than housing cycles, which also display a substantial variation in the degree of synchronisation, again owing to the diverse structure of housing markets across countries.
CONCLudINg REMARkS
One of the key goals of the new macro-prudential mandates around the world is to attenuate financial cycles. In the euro area, there is a need for country-level financial cycle estimates to provide a clear and consistent yardstick to guide forward-looking macro-prudential policy.
19 See, for example, “Institutional features and regulation of housing and mortgage markets” in European Commission, Quarterly report on the euro area, Vol. 13, Issue 2, June 2014.
20 See, for example, the box entitled “Stylised facts of money and credit over the business cycle” Monthly Bulletin, ECB, October 2013.
Strong correlation between credit,
housing and business cycles
within countries
Estimating financial cycles for euro area
countries…
Chart B.4 Synchronisation of cycles within countries
(share of quarters; maximum; minimum; inter-quartile range and median)
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Credit and
business
Housing and
business
Credit and
housing
Equity and
business
Equity and
housing
Credit and
equity
Sources: BIS, OECD, Eurostat and ECB calculations. Notes: The chart shows the degree of synchronisation of cycles within countries. Synchronisation is measured as the concordance between two series (e.g. real credit and real GDP) in each country.
Chart B.5 Synchronisation of cycles across countries
(share of quarters; maximum; minimum; inter-quartile range and median)
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Business Equity Credit Housing
Sources: BIS, OECD, Eurostat and ECB calculations. Notes: The chart shows the degree of synchronisation of cycles across countries. Synchronisation is measured as the concordance using a particular series (e.g. real credit) for two countries.
117 ECB
Financial Stability Review November 2014 117
SPECIAL FEATURE B
117
This special feature presented two methodologies to measure financial cycles for euro area countries and benchmarked these against business cycles obtained on a comparable basis. The methodologies are in many ways complementary – the turning point analysis considered in this special feature focuses on the levels of the underlying series, while spectral analysis looks at growth rates, thereby incorporating important information contained in stocks and flows.
Results suggest that the features of financial cycles tend to differ considerably from their business cycle counterparts. Both methodologies confirm the higher amplitude in the cycles of financial variables compared with the business cycle. The findings differ with regard to the length of the financial cycle, however, which can be attributed to the different definitions of cycles inherent to both methodologies. The relevance of measures of credit and asset prices in effectively capturing a synthetic financial cycle appears to vary at the country level, reflecting cross-country heterogeneity and idiosyncrasies in underlying driving forces.
… suggests financial and business cycles differ strongly
118 ECB Financial Stability Review November 2014118118
C INITIAL CONSIdERATIONS REgARdINg A MACRO-pRudENTIAL INSTRuMENT BASEd ON ThE NET STABLE FuNdINg RATIO1
The financial crisis led to a broad consensus among policy-makers and regulators that macro- prudential frameworks, in addition to micro-prudential policy, must be part of the solution to ensure the resilience of the financial system. The counter-cyclical capital buffer represents the first step in this direction taken by the Basel Committee on Banking Supervision. Regarding liquidity issues, two micro-prudential standards have been designed. The delegated act implementing the liquidity coverage ratio (LCR) at the European level has recently been adopted by the European Commission and the net stable funding ratio (NSFR) standard has just been finalised by the Basel Committee on Banking Supervision and was published on 31 October. After implementing these new standards, it will be necessary to monitor their impact on banks’ behaviour, market liquidity, monetary policy and financial stability before considering introducing any additional instruments. At this stage, the need for a liquidity-based macro-prudential tool is in the early stages of identification and discussion. Therefore, this special feature aims to provide some initial technical considerations regarding the macro-prudential use of the NSFR. The discussion considers two broad perspectives. The first is the need for a counter-cyclical NSFR to complement the counter-cyclical capital buffer. While capital and liquidity standards pursue different objectives, the two can also be used in conjunction depending on the specific risk to financial stability being targeted. The second perspective regards the use of the NSFR as a stand-alone macro-prudential tool, together with its potential trigger mechanism and its use in the current low yield environment.
INTROduCTION
The financial crisis highlighted the risks of unstable funding mixes and maturity mismatches on banks’ balance sheets. As a result, a series of micro-prudential standards have been developed, aimed at strengthening the resilience of banks confronted with liquidity shocks. One of the two instruments adopted by the Basel Committee on Banking Supervision in December 2010, together with the liquidity coverage ratio (LCR), is the net stable funding ratio (NSFR).2 The purpose of the NSFR is to ensure banks achieve a “stable funding profile” by limiting their excessive reliance on short- term wholesale funding relative to the liquidity risk characteristics of their assets and off-balance- sheet exposures (see Box C.1 for more detailed information on the composition of the NSFR). It supplements the LCR – which promotes banks’ short-term resilience to severe idiosyncratic and market-wide liquidity stress – by reducing the funding risk of institutions over a longer-term horizon.
Micro-prudential policy applies the same standards across banks, regardless of the impact of an institution’s failure on the financial system. Consequently, the micro-prudential approach assumes that the sources of risk are independent and exogenous to the collective behaviour of financial institutions. This shortcoming is addressed by the macro-prudential approach, which takes a systemic view rather than focusing on individual institutions. By considering both the systemic impact of financial institutions (the cross-sectional dimension) and the evolution of system-wide risk (the time dimension), the macro-prudential approach addresses the negative feedback loop that may emerge between the financial system and the real economy.
There has been significant progress in the design of macro-prudential tools, most notably the counter-cyclical capital buffer3 and the additional capital requirements for systemically important
1 Prepared by Andreea Bicu, Daniela Bunea and Michael Wedow. 2 Basel Committee on Banking Supervision, Basel III: the Net Stable Funding Ratio, 2014. 3 Basel Committee on Banking Supervision, Basel III: A global regulatory framework for more resilient banks and banking systems, 2011.
The NSFR is intended to increase
the resilience of banks confronted
with liquidity shocks…
… but as a micro-prudential
tool, it might not be sufficient
119 ECB
Financial Stability Review November 2014 119
SPECIAL FEATURE C
119
banks.4 Both tools require banks to hold greater amounts of capital, either in particular states of the economy (credit boom) or, in the case of systemically important institutions, at all times. However, in its discussions, the European Systemic Risk Board (ESRB)5 has highlighted that capital regulation may not be sufficient to limit systemic risk. Four sources of banking sector systemic risk have been identified: i) excessive credit growth and leverage; ii) excessive maturity mismatch and market illiquidity; iii) direct and indirect exposure concentrations; and iv) misaligned incentives with a view to reducing moral hazard. A combination of macro-prudential tools designed to address systemic risks posed by all of these four sources is hence needed.
Despite the significant progress made in understanding liquidity cycles, a framework that identifies systemic liquidity risks and guides the implementation of macro-prudential liquidity tools is still missing. Since the NSFR is by construction a micro-prudential tool, there is a debate regarding how the NSFR could be used as a macro-prudential instrument and, if necessary, how it should be modified for this purpose. The NSFR is a new liquidity metric and is yet to be implemented. Therefore, it must be kept in mind that a monitoring period for the ratio as well as more data and analysis are necessary in order to assess its practical use and shortcomings. While further adjustments to the NSFR may be premature at this stage, this special feature seeks to put forward some initial considerations regarding the potential use of the NSFR as a macro-prudential tool.
4 Basel Committee on Banking Supervision, Global systemically important banks: updated assessment methodology and the higher loss absorbency requirement, 2013; Financial Stability Board, 2013 update of group of global systemically important banks (G-SIBs), 2013.
5 European Systemic Risk Board, Flagship report on macro-prudential policy in the banking sector, 2014; European Systemic Risk Board, The ESRB handbook on operationalising macro-prudential policy in the banking sector, 2014.
Box C.1
WhAT IS ThE NET STABLE FuNdINg RATIO?
The purpose of the net stable funding ratio (NSFR), as a structural liquidity risk metric, is to reduce maturity mismatches between assets and liabilities over a one-year time horizon and, thereby, to reduce funding risk.
Under this standard, banks are required to hold a minimum amount of stable funding relative to the maturity/liquidity profile of their assets in order to limit their structural liquidity mismatch. It complements the liquidity coverage ratio (LCR) and is intended to limit the proportion of banks’ less liquid assets, such as long-term loans with maturities of over one year that are funded by short-term funding of less than one year, or funding sources considered less reliable and stable. In addition, the NSFR is intended to encourage a better assessment of funding risk across all on- and off-balance-sheet items and, overall, to promote funding stability.
The NSFR measures the ratio between the available amount of stable funding (ASF) and the required amount of stable funding (RSF). The ASF consists of weighted liabilities reflecting their contractual maturity or expected behavioural stability. The RSF consists of assets weighted by factors to reflect their contractual maturity or their expected market liquidity. The weights for assets and liabilities range from 100% to 0%. The ASF is the portion of a bank’s funding structure that is a reliable source of funding over a one-year time horizon, while the RSF is the portion of a bank’s assets and off-balance-sheet exposures viewed as illiquid over a one-year horizon and should thus be backed by stable funding sources.
120 ECB Financial Stability Review November 2014120120
LIquIdITY AS A COMpLEMENT TO OThER pRudENTIAL MEASuRES
The primary objective of micro-prudential regulation is “the promotion of safety and soundness of banks and the banking system”.6 The main regulatory standards which aim to fulfil this goal are based on capital and liquidity requirements. It is thus important to better understand the different objectives of and interactions between capital and liquidity requirements.
Capital and liquidity holdings are both important for increasing the resilience of banks. However, the nature of the shocks that capital regulation helps mitigate is different from the types of shock that liquidity regulation helps mitigate. The purpose of capital regulation is to limit the risk of insolvency, given the loss-absorbing capacity of this form of funding. By contrast, liquidity rules are intended to limit the maturity mismatch between liabilities and assets and, as a result, minimise funding liquidity risk (i.e. the inability to settle payment obligations) and market liquidity risk (i.e. the inability to sell or use assets without a significant impact on prices). Insufficient balance sheet liquidity can also lead to cash-flow insolvency7, even if a bank is still considered solvent from a capital perspective.8 Liquidity and solvency are closely interrelated. On the one hand, higher capital holdings reduce the need for liquidity buffers, all else being equal. Banks, however, still need to maintain adequate liquidity regardless of their capital levels since the two cannot perfectly substitute for one another. Therefore, strengthening capital buffers is not sufficient by itself to address liquidity risks affecting both sides of the balance sheet.9 Moreover, even a highly rated bank can have difficulties accessing private sources of funding, as the recent financial crisis has shown.10 Conversely, liquidity buffers can compensate to some extent for low capital levels and protect the bank when faced with a confidence shock. The importance of maintaining adequate capital and liquidity levels supports the need for liquidity standards to complement capital regulation.
According to the Bank of England,11 there are a number of channels through which the newly introduced liquidity standards interact with a bank’s capital position and vice versa. For instance, higher levels of capital give confidence to depositors and investors to provide or roll over funding to banks. Alternatively, increasing the NSFR/LCR by replacing illiquid loans with liquid assets leads to an improvement in capital ratios by decreasing risk-weighted assets. In addition, building capital and NSFR buffers is likely to be less costly for the bank when done in parallel, since an improvement in the NSFR will be accompanied by an increase in the capital ratio and vice
6 See Basel Committee on Banking Supervision, Core principles for effective banking supervision, 2012. 7 Cash flow insolvency is defined as the inability of a bank to repay its debts when they become due. 8 Farag, M., Harland, D. and Nixon, D., “Bank capital and liquidity”, Quarterly Bulletin, Bank of England, 2013. 9 European Systemic Risk Board, The ESRB handbook on operationalising macro-prudential policy in the banking sector, op. cit. 10 See van Rixtel, A. and Gasperini, G., “Financial crises and bank funding: recent experience in the euro area”, BIS Working Paper, No 406, 2013. 11 Farag, M., Harland, D. and Nixon, D., op. cit.
Adequate buffers for both capital
and liquidity are necessary
Interactions between liquidity and other
regulatory measures should be taken into
account
The design of the NSFR underwent some changes in January 2014 compared with its initial design proposed in December 2010. These changes included greater granular differentiation in terms of maturity and sought to reflect that the NSFR is a structural liquidity risk metric rather than a ratio calculated for stress scenarios. Overall, the revisions have made the tool more suited to detecting outlier banks with excessive maturity mismatches and thus fragile funding structures, as well as brought it more into line with the LCR in terms of the treatment of high- quality liquid assets. The final calibration of the NSFR was published in October 2014 and its implementation is foreseen for 2018.
121 ECB
Financial Stability Review November 2014 121
SPECIAL FEATURE C
121
versa.12 Moreover, the cost of increasing the NSFR gradually declines when more capital is raised, highlighting the synergies between the two standards.
In sum, prudential regulation should ensure that banks have sufficient capital and liquidity in order to avoid disrupting their financial intermediation function. The optimal combination should minimise the probability of distress, while balancing the benefits and costs of holding liquidity and capital.13
WhAT hAppENS TO ThE NSFR WhEN ThE COuNTER-CYCLICAL CApITAL BuFFER IS BuILT up?
Against the background of the link between the liquidity and capital standards, this section explores the relationship between the counter-cyclical capital buffer and the NSFR. It is important to understand how the two standards interact when the counter-cyclical capital buffer is activated. The starting point of the analysis is a stylised bank balance sheet with an initial NSFR close to the weighted average of the banks assessed under the Basel Committee’s Quantitative Impact Study (NSFR = 115%). Moreover, under all scenarios, the bank fulfils the minimum Basel III requirements for the risk-based capital and leverage ratios. The effect of implementing the full counter-cyclical capital buffer (2.5% of risk-weighted assets) on the NSFR for different starting bank capital ratios (8%, 9% and 10%) is considered. Under a first scenario, the bank maintains its entire existing capital buffer, even if it is above the minimum requirement. However, if the bank already has a capital buffer above the minimum requirement before the counter-cyclical capital buffer is built up, it could also choose to reduce this buffer to limit the potential impact on income and costs. Hence, the second scenario considers the case where the bank meets the higher minimum requirement by relying on the existing capital buffer. These two scenarios define a range for banks’ decisions when capital ratios need to be adjusted.
In order to estimate the effect on the NSFR of the build-up of capital, two broad benchmark cases are assessed, as illustrated in Chart C.1: (1) portfolio rebalancing via a shift towards assets with lower risk weights; and (2) balance sheet expansion resulting from an increase in capital.
Case 1 – portfolio rebalancing As an alternative to raising new equity, the bank may choose to decrease its risk-weighted assets while keeping the total size of the balance sheet unchanged. Under this scenario, replacing riskier assets by less risky assets is also likely to improve the NSFR, given that less risky assets are typically also more liquid and may thus also result in a lower required amount of stable funding (RSF).
Case 2 – Balance sheet expansion Under this scenario, the bank raises its capital ratio by issuing capital and/or retaining earnings, leading to an expansion of the balance sheet. Moreover, it is assumed that the bank invests the proceeds in assets requiring less regulatory capital. With regard to the NSFR, on the liability side, the increase in capital will lead to an improvement in the available amount of stable funding (ASF) of the same magnitude (100% factor). On the asset side, the investment will lead to a relatively smaller increase in the RSF for the majority of asset categories. As a consequence, the bank will see an improvement in its NSFR. The overall impact on the NSFR will be maximised by investing in assets with the lowest RSF, such as cash and sovereign bonds.
12 See King, M.R., “Mapping capital and liquidity requirements to bank lending spreads”, BIS Working Paper, No 324, 2010, and Basel Committee on Banking Supervision, An assessment of the long-term economic impact of stronger capital and liquidity requirements, 2010.
13 There are potentially also further interactions between the NSFR and the possible requirements for “bail-inable” debt for resolution purposes. These interactions are not considered in this special feature given that the work on resolution requirements is still ongoing.
122 ECB Financial Stability Review November 2014122122
NSFR levels following the balance sheet adjustments described in the two cases above are computed for different initial levels of capital. Chart C.2 shows the results obtained from implementing the two strategies to different degrees in order to visualise the range of possible NSFR changes. The horizontal line represents the starting NSFR level and is included as a benchmark. The most significant improvement in the NSFR is obtained from a reduction in risk-weighted assets following a rebalancing of the portfolio (Case 1). Balance sheet expansions (Case 2) have a weaker effect on the NSFR. Note, however, that the bank may also experience a slight decline in its NSFR
Implementing the counter-cyclical
capital buffer can also improve NSFR
levels…
Chart C.1 possible strategies for meeting the counter-cyclical capital buffer requirement
Case 1 Case 2
Assets Liabilities Assets Liabilities
Cash
Capital
High-quality liquid assets
Risky A
Low-risk assets
High-risk assets
Short-term funding
(<1 year)
Long-term funding
(>1 year)
Extra assets
Cash
High-quality liquid assets
Risky A
Short-term funding
(<1 year)
Long-term funding
(>1 year)
Capital
Extra capital
Note: The charts are used for illustrative purposes and are not based on actual balance sheet data used in the simulations.
Chart C.2 Impact of the counter-cyclical capital buffer on the NSFR: maintaining a constant capital buffer (left) and including the existing capital buffer in the counter-cyclical capital buffer (right) (percentages)
90
100
110
120
130
140
150
160
90
100
110
120
130
140
150
160 Case 1 Case 2 Case 1 Case 2 Case 1 Case 2
K=8 % K=9 % K=10 % 90
100
110
120
130
140
150
160
90
100
110
120
130
140
150
160 Case 1 Case 2 Case 1 Case 2 Case 1 Case 2
K=8 % K=9 % K=10 %
Source: ECB calculations. Notes: The vertical bars represent the NSFR following a range of balance sheet adjustments belonging to Case 1 (blue bars) or Case 2 (reddish orange bars) respectively. The capital level below each set of bars (labelled “K”) indicates the starting risk-based capital ratio prior to activating the counter-cyclical capital buffer.
123 ECB
Financial Stability Review November 2014 123
SPECIAL FEATURE C
123
if it increases its holdings of assets with a very high RSF factor.14 Overall, the simulations suggest an improvement in the NSFR as a result of implementing the counter-cyclical capital buffer. The rise in the NSFR is particularly pronounced for banks with low initial capital ratios which pursue adjustment strategies on the asset side and is largely muted for better capitalised banks.
Some caveats of our analysis should be noted. The mechanical scenarios do not take into account the potential offsetting behaviour of banks. Typically, a bank that follows one of the scenarios will try to offset the higher cost or the reduced income. Given that a bank cannot raise revenue by investing in riskier assets because of the impact on its risk-based capital ratio, it could compensate the increase in costs by shortening the term of its funding sources subject to any leeway obtained under the NSFR. Naturally, both cases are artificial in nature and banks typically use a combination of adjustments on both the asset and the liability sides. Moreover, given that the counter-cyclical capital buffer is likely to be implemented in buoyant times, raising capital appears the more likely scenario.
The analysis above has highlighted a positive relationship between the capital ratio and the NSFR, i.e. an increase in capital is also likely to increase the NSFR. This endogenous interaction can be desirable when there is a simultaneous need to build up resilience in terms of capital and the NSFR during a boom in the credit cycle. Under this assumption, macro-prudential policy could take this interaction between capital and the NSFR into account and, possibly, require a simultaneous build- up of an NSFR buffer. If, however, this is deemed unnecessary, banks should be allowed to flexibly use the additional stable funding resources. The subsequent section further discusses the potential use of the NSFR as a stand-alone macro-prudential instrument.
LIquIdITY AS AN INdEpENdENT MACRO-pRudENTIAL MEASuRE
In addition to micro-prudential rules, systemic liquidity risks need to be addressed by appropriately designed macro-prudential regulation. Systemic liquidity stress is defined by the ESRB as the failure of banks’ normal funding channels, leading to the central bank intervening as the lender of last resort.15 The recent crisis has highlighted that solvency regulation alone cannot fully address these risks and that macro-prudential liquidity instruments are necessary. The ESRB has identified the prevention of excessive maturity mismatch and market illiquidity as an intermediate macro-prudential objective.16 Considering that the aim of the NSFR is to prevent such mismatches, a well-designed and targeted (possibly time-varying) ratio could therefore help mitigate systemic liquidity risks.17
Acharya et al.18 discuss the counter-cyclical behaviour of liquidity in banks’ asset holdings, i.e. it tends to be inefficiently low during the business cycle upturn and excessively high during downturns. During boom periods, this behaviour is supported by the ease of obtaining funding owing to banks’ profitability as well as by a benign view on asset quality and liquidity, as reflected in the pledgeability of assets and low collateral haircuts. During downturns, by contrast, banks tend to have higher liquidity holdings as this acts as a form of insurance when facing uncertain liquidity withdrawals. Another reason is that they can then take advantage of fire sales if financial
14 The upper and lower bounds for the NSFR are obtained following very extreme balance sheet rebalancing and expansion strategies. The resulting interactions are hence relatively unlikely.
15 European Systemic Risk Board, The ESRB handbook on operationalising macro-prudential policy in the banking sector, op. cit. 16 ibid. 17 The LCR supplements the NSFR by promoting the short-term resilience of banks to severe liquidity shocks. Owing to the NSFR’s
structural nature, the longer horizon it targets and the intermediate systemic risk objectives it addresses, this special feature focuses on the macro-prudential use of the NSFR. The potential use of the LCR as a macro-prudential tool is not discussed in this special feature.
18 Acharya, V., Shin, H.S. and Yorulmazer, T., “Crisis resolution and bank liquidity”, The Review of Financial Studies, Vol. 24, No 6, 2011, pp. 2166-2205.
… however, banks may further adjust their balance sheets
NSFR adjustments may be necessary…
… considering the dynamics of banks’ (asset) liquidity
124 ECB Financial Stability Review November 2014124124
distress intensifies. From a financial stability perspective, this pattern raises a series of concerns. First, this counter-cyclical behaviour could support excessive credit growth during a boom and aggravate the economic downturn if banks hoard excessive liquidity during a bust. Second, the simultaneous large-scale sale of assets when financial distress intensifies leads to a vicious cycle of declining asset prices and losses on banks’ balance sheets, possibly precipitating further sales. Since the magnitude of fire sales is directly related to the balance sheet liquidity of the overall system, the counter-cyclical behaviour of liquidity across many market participants reinforces systemic stress during downturns. In addition, banks with insufficient cash and cash-like holdings may want to avoid selling other (less liquid) assets at discounted prices when financial stress is escalating, and rather increase their demand for additional funding. A system-wide increase in the demand for liquidity can precipitate funding liquidity stress, leading to spikes in funding costs and a breakdown in markets. This market failure may subsequently make central bank liquidity interventions necessary. As vividly demonstrated during the financial crisis, if banks fail to adequately manage liquidity and funding risk, this creates significant systemic vulnerabilities and threatens financial stability. The recent crisis has highlighted that capital regulation alone cannot fully address such vulnerabilities and that both micro-prudential and macro-prudential liquidity standards and instruments are necessary.19
The liquidity dynamics highlighted above are likely to be muted by the implementation of the new minimum standards for liquidity. However, this counter-cyclical behaviour could potentially persist even after the introduction of the liquidity standard. This would, in turn, be reflected in the NSFR, leading to relatively low NSFRs during booms and rising NSFRs during stress periods. As highlighted in this special feature, building up the capital buffer may already help increase the level of the NSFR during a boom. Nevertheless, liquidity and funding risks fluctuate over time and may not be sufficiently reflected in the NSFR given the static factors applied in its calculation. If the NSFR and the counter- cyclical capital buffer prove to be insufficient for limiting these risks, there will be some grounds for considering an additional liquidity macro-prudential tool to help address pro- cyclical risk-taking behaviour and to increase the resilience of banks.
As regards real NSFR figures, EU banks have experienced a continuous improvement in their NSFR since 2011, mainly owing to readjustments in their balance sheets and changes in the calibration of the NSFR.20
Chart C.3 illustrates the dynamics of the average NSFR for Group 1 and Group 2 banks21 during the six quarters covered by the Basel III monitoring exercise. At this point in time, it is still premature to assess the existence and
19 European Systemic Risk Board, The ESRB handbook on operationalising macro-prudential policy in the banking sector, op. cit. 20 European Banking Authority, Basel III monitoring exercise, September 2014. 21 The banks covered by the Basel III monitoring exercise are divided into two groups, with Group 1 made up of internationally active banks
with Tier 1 capital of more than €3 billion and Group 2 representing all other banks.
Chart C.3 NSFR levels for group 1 and group 2 Eu banks
(June 2011 – Dec. 2013; percentages, weighted averages)
80
85
90
95
100
105
110
115
80
85
90
95
100
105
110
115
June
Group 1 Group 2
Dec. Dec. Dec.June June 2011 2012 2013
Source: EBA Basel III monitoring exercise.
125 ECB
Financial Stability Review November 2014 125
SPECIAL FEATURE C
125
magnitude of any cyclical behaviour in the NSFR. In December 2013, which is the latest date for which public figures are available, the weighted average NSFR was above 100% for both groups of EU banks, at 109% and 102% respectively.
Despite these relatively comfortable NSFR levels, some current risks to financial stability have been highlighted. More specifically, the search for yield has contributed to asset price misalignments, as highlighted in the Overview and Section 2 of this issue of the FSR. In the current environment of high funding liquidity but subdued credit and economic growth, the counter-cyclical capital buffer may not be fully adequate to mitigate this risk to financial stability. Given that the NSFR explicitly incorporates securities at market prices, analysis needs to be carried out to establish whether these fluctuations in the NSFR are beneficial from a macro-prudential perspective. Depending on the conclusions of this analysis, an exploration of the scope for using an additional liquidity tool to address this risk may thus be appropriate. An understanding of the elements most likely to affect the NSFR could help in designing the counter-cyclical features of this instrument. As a macro- prudential tool aimed at preventing the build-up of systemic risk, a well-designed buffer could impose prudency in activities where financial stress would create significant negative effects. This seems particularly warranted when the financial cycle and the liquidity cycle are disconnected and may help to overcome the “inaction bias”.
WhAT FORM COuLd A COuNTER-CYCLICAL NSFR TAkE?
Imposing a higher minimum threshold for the NSFR when appropriate conditions are met would represent the most direct solution from an operational point of view. Similar to the counter-cyclical buffer, this would require the implementation of a trigger mechanism to signal when the NSFR minimum requirement is to be raised. While this is intuitively the most straightforward approach, it could have unintended consequences. For example, in the current environment, if banks increase their NSFR through even higher holdings of high-quality liquid assets, this could further aggravate asset price misalignments. A more targeted approach might therefore be warranted. In its current form, the NSFR relies on static RSF and ASF factors for assets and liabilities. Adjusting factors for particular asset classes, funding sources and/or sectors might therefore be preferable to imposing an overall higher NSFR requirement. Such an approach may, however, raise further complications in terms of implementation. Any deviation from internationally agreed standards should be subject to coordination and disclosure mechanisms across jurisdictions. Harmonisation is needed in order to ensure comparability and legal certainty within the Single Market.
With regard to assets, the RSF factors have been calibrated to reflect the need for stable funding sources. A possible avenue to address the risk of asset price misalignments could be to adjust the RSF factors upwards for those assets most affected, reflecting future risks of downward price adjustments, while leaving the overall minimum requirement of 100% unchanged. It should be noted that a rise in securities’ prices would, ceteris paribus, already lead to a decline in the NSFR. Therefore, any change in the RSF factors would further dis-incentivise demand and reduce the upward pressure on prices. This may be warranted during times when easy access to funding more than compensates for any inflationary effect on the RSF. It should be noted that the factors currently applied, particularly for high-quality liquid assets, have already been set at relatively conservative levels. Securities and certain equities that have been included as high-quality liquid assets in the LCR also have lower funding requirements under the NSFR, given the view that they can be used quickly to obtain stable funding either by outright sales or by using them in secured operations.
High NSFR levels may mask the build- up of risk
Banks could be required to maintain higher NSFR levels…
… or individual ASF/ RSF factors could be adjusted
126 ECB Financial Stability Review November 2014126126
Considering the overlaps, any adjustment in the RSF factors within the NSFR may thus also require further adjustments of the targeted assets included in the LCR. More generally, consistency across these two ratios may also be required in the broader context if either of the two ratios is used as a macro-prudential tool.
With regard to liabilities, a counter-cyclical NSFR could also be implemented by reducing ASF factors to reflect the (time-varying) stability of different funding sources. Revisions to ASF factors could be triggered by behavioural changes among depositors, by changes in the functioning of markets or if excessive reliance on certain funding sources emerges.
In the light of the discussion above, a set of trigger variables for the aggregate NSFR or for components of the ratio may be useful in the design of a counter-cyclical NSFR. This set of trigger variables could be based on volume and price-based indicators for liquidity risk. Cross-checking and combining information from multiple indicators may further improve accuracy when a warning signal is detected,22 but may also further complicate the trigger mechanism.
In addition to the LCR and NSFR, the Basel Committee on Banking Supervision also proposes that banks should report a series of additional liquidity monitoring metrics.23 These monitoring metrics may be particularly useful for identifying a systemic build-up of excessive funding risks. For example, the maturity ladder incorporates a broader set of maturity buckets going beyond the one-year horizon of the NSFR. Therefore, a counter-cyclical buffer could be activated when there is a build-up of maturing debt beyond the one-year horizon. This may be desirable when the maturity ladder across a wider part of the banking system indicates a future refinancing glut that could create strains in funding markets. By looking beyond a one-year horizon, mismatches could signal possible imbalances not yet captured by current NSFR levels. Regarding the monitoring metrics for the concentration of funding by counterparty/product, the NSFR could also target risk by reducing banks’ over-reliance on specific liquidity providers and instruments rather than simply raising the minimum requirement. In this context, however, the NSFR would pursue structural rather than cyclical policy objectives. Moreover, other tools may be more effective at addressing some of these issues, such as the large exposure requirements. A number of important sectors could be monitored and, if a build-up of risk in a specific sector (e.g. mortgages) is detected, the factors assigned to assets or liabilities related to these sectors could be adjusted. Finally, there is also scope to apply the NSFR as a tool for detecting excessive mismatches in the currency composition of assets and liabilities. This could be implemented by setting currency-specific NSFR requirements. The build- up of currency mismatches between assets and liabilities captured by the liquidity monitoring metrics could therefore be addressed by currency-specific NSFR requirements.24
According to the ESRB,25 simpler structural liquidity ratios such as the loan-to-deposit ratio and the core funding ratio are promising both in their role as indicators and as instruments addressing maturity mismatches and market illiquidity. The International Monetary Fund26 finds that higher
22 European Systemic Risk Board, The ESRB handbook on operationalising macro-prudential policy in the banking sector, op. cit. 23 These metrics are: a maturity ladder, the concentration of funding by counterparty, the concentration of funding by product type, the
concentration of counterbalancing capacity by issuer/counterparty, prices for various lengths of funding and the rollover of funding. See http://www.bis.org/publ/bcbs238.pdf for details.
24 Such an application would require taking into account banks’ currency risk management, e.g. whether they hedge these risks with appropriate financial instruments.
25 European Systemic Risk Board, The ESRB handbook on operationalising macro-prudential policy in the banking sector, op. cit. 26 See International Monetary Fund, Global financial stability report, October 2013.
Possible trigger variables can be
considered
127 ECB
Financial Stability Review November 2014 127
SPECIAL FEATURE C
127
loan-to-deposit ratios are associated with greater bank distress,27 both in advanced and emerging economies. Bologna28 also investigates the predictive power of the loan-to-deposit ratio for bank failures and finds that high loan-to-deposit levels increase the likelihood of a bank failure occurring two to three years later. The level of loan-to-deposit ratios one year prior to a failure is, however, not statistically significant, a pattern also highlighted by Marino and Bennett,29 who attribute this effect to a change in deposit composition and portfolio rebalancing at incipient signs of distress. Empirical research therefore suggests that the loan-to-deposit ratio is able to detect a build-up of risk with a substantial lead and it may thus be useful to include it in the design of the counter-cyclical NSFR as an early warning indicator. Moreover, considering the challenges associated with the operationalisation of a counter-cyclical NSFR, a time-varying loan-to-deposit ratio or core funding ratio may be easier to calibrate and implement.
The liquidity mismatch index proposed by Brunnermeier et al.30 represents an alternative measure of mismatch between bank assets and liabilities.31 It mirrors to some extent the NSFR design by assigning weights to balance sheet elements according to their ease of being sold (positive weights) as well as to the stability of funds and ease of rolling over debt (negative weights). Bai et al.32 implement the liquidity mismatch index and connect the liquidity premium on issuing liabilities and, hence, the time-varying stability and ease of obtaining funding to the spread between overnight index swaps and Treasury bills. More negative weights, indicating an increase in the volatility of funding sources, are assigned across all maturities during periods when there is a significant widening in the spread. The rationale behind this is that if the liquidity stress episode is severe and, hence, possibly long lasting, the stability of funding, even with a term beyond one year, becomes uncertain. When compared with its static design, the liquidity mismatch index calculated using time-varying weights was thus better able to capture the build-up of mismatches before 2008 when applied to a large sample of US bank holding companies. While an aggregate liquidity mismatch index has potential as a monitoring tool and could be used in the design of a counter-cyclical NSFR, its appropriateness has not yet been explored in the context of the European banking sector.
As highlighted by Bai et al.,33 the spread between overnight index swaps and Treasury bills contains important information regarding the stability of funding over the cycle. Moreover, the time-varying liability component is shown to be the main driving factor for liquidity mismatch dynamics. When spreads are compressed during boom periods, easing the access to funding, banks could be required to build up buffers since obtaining funding by issuing capital or liabilities can be achieved more easily and at lower cost. In a similar vein, Bloor et al.34 consider long-term funding costs as a natural trigger for the counter-cyclical NSFR buffer. The NSFR incentivises banks’ reliance on longer-term funding. Since the cost of accessing higher volumes of liquidity increases more steeply in long-term (less liquid) markets, meeting the requirement creates costs. This non-linear price-quantity relationship is further amplified in a crisis owing to high risk aversion and the drying-up of liquidity, especially at longer maturities. As a result, greater exposure to longer-term markets can lead to more adverse macroeconomic outcomes in the event of systemic market stress
27 A distressed bank is characterised by a low z-score, a low price-to-book ratio and a “sell” recommendation rating by bank equity analysts. 28 Bologna, P., “Structural funding and bank failures: Does Basel 3 net stable funding ratio target the right problem?”, Journal of Financial
Services Research, September 2013. 29 Marino, J.A. and Bennett, R.L., “The consequences of national depositor preference”, FDIC Banking Review, Vol. 12, No 2, 1999, pp.19-38. 30 Brunnermeier, M., Gorton, G. and Krishnamurthy, A., “Liquidity mismatch measurement”, in Brunnermeier, M. and Krishnamurthy, A.
(eds.), Risk Topography: Systemic Risk and Macro Modelling, NBER Books, 2014. 31 LMIω = Σi λωi Ai −Σj λωj Li , where λωi / λωj are weights applied to each asset and liability class i/j and are indexed by the state of the world ω. The
lower the index value, the higher the liquidity risk. 32 Bai, J., Krishnamurthy, A. and Weymuller, C.H., Measuring liquidity mismatch in the banking sector, 2013. 33 ibid. 34 Bloor, C., Craugie, R. and Munro, A., “The macroeconomic effects of a stable funding requirement”, Discussion Paper Series, DP2012/05,
Reserve Bank of New Zealand, 2012.
128 ECB Financial Stability Review November 2014128128
and the pro-cyclical effect of funding spreads is amplified.35 In periods of high funding liquidity and low long-term funding spreads, banks should thus be required to build up a buffer of long- term funding. Additionally, the buffer could be released during periods of stress to dampen adverse macroeconomic outcomes.
BROAdER CONSIdERATIONS REgARdINg ThE MACRO-pRudENTIAL uSE OF LIquIdITY STANdARdS
The design of a counter-cyclical NSFR needs to take into account the possibility that the buffer, similarly to the LCR, can be used during periods of stress. There are two possible complementary options for the implementation of the counter-cyclical NSFR in this respect: requiring a positive add-on for the NSFR while keeping the 100% as a lower, binding constraint, or allowing the NSFR to drop below 100% when liquidity conditions deteriorate. To the extent that the market allows banks to fall below the minimum requirement of 100%, the added flexibility should be reflected in the build-up phase of a counter-cyclical NSFR. If a level below 100% is indeed tolerated by the market and a jump in the risk perception of the bank is not a constraining factor, a lower required add-on for the NSFR could be designed, limiting the negative effects of too stringent an upper bound. On the other hand, dropping below 100% could still be perceived negatively by the market, limiting access to funding and sharply increasing borrowing costs. The relationship between the demand for long-term funding and the associated costs could thus be reinforced when the market perceives a NSFR below 100% as a negative signal, especially during times of financial stress. Building an additional buffer during boom periods might therefore minimise the risk of liquidity shortages and reduce uncertainty. Moreover, considering that one aim of a higher requirement is to “lean against the wind” during a cyclical upswing, a high add-on might still be preferable, independently of the possibility to go below 100% during times of crisis.
Well-designed macro-prudential tools should achieve maximum benefits with minimum costs. Additional changes to existing rules should be considered very carefully, taking into account that regulation that is too stringent might benefit other, less regulated parts of the financial sector and shift activity further towards the shadow banking sector. This would simply push the risks into these less regulated parts of the financial system and could even lead to an increase in systemic risk. On the other hand, the NSFR could contribute to the resilience of the financial system by dis-incentivising interlinkages between banks and non-bank financial institutions.
CONCLudINg REMARkS
This special feature highlights some initial considerations on the design and use of a counter- cyclical NSFR. Interactions with the counter-cyclical capital buffer show a positive relationship between the two. The counter-cyclical behaviour of bank liquidity indicates that an increase in the NSFR during a boom would be beneficial from a financial stability perspective. Therefore, the special feature highlights that the improvement in the NSFR arising from this interaction could be preserved and, possibly, further built on. Additional analysis should thus be carried out to determine whether the new standards designed by the Basel Committee on Banking Supervision are sufficient to address the counter-cyclical behaviour of banks.
35 The funding spread is the difference between long-term funding costs and the rollover of short-term funding. In good times, these spreads are compressed, while they increase in periods of stress. Costs of long-term funding may be further pushed upwards if demand is very high.
Liquidity regulation that is too stringent
may have unintended consequences
129 ECB
Financial Stability Review November 2014 129
SPECIAL FEATURE C
129
In designing the counter-cyclical features of the NSFR, various possible options have been highlighted. A higher minimum threshold would offer flexibility to banks in adjusting their balance sheet, while also being operationally easier to implement. On the other hand, a more targeted approach involving adjustments to individual ASF and RSF factors may be more appropriate if a build-up of risk in specific sectors or over different maturity horizons is detected.
Further work will be required to quantify the impact of the new Basel ratios after they are introduced. Second, the need for an additional instrument and its potential benefits and drawbacks have to be carefully assessed. Third, further work needs to be carried out on suitable trigger variables as well as on identifying appropriate buffer levels to be built up during upturns and released during downturns. Finally, it also remains to be analysed and discussed how other available macro-prudential instruments, such as the systemic risk buffer, might interact with a potential counter-cyclical NSFR, given the possible overlaps in the risk that these instruments address. The benefits of any mix of macro-prudential tools need to be assessed against the specific costs of implementation, including any distortions to the financial system or potential leakages.
STAT IST ICAL ANNEX
1 MACRO-FINANCIAL AND CREDIT ENVIRONMENT
S 1 ECB
Financial Stability Review November 2014
S.1.1 Actual and forecast real GDP growth S.1.2 Actual and forecast unemployment rates
(Q1 2004 - Q3 2014; annual percentage changes) (Jan. 2004 - Sep. 2014; percentage of the labour force)
-20 -18 -16 -14 -12 -10
-8 -6 -4 -2 0 2 4 6 8
10 12 14
- -
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14
euro area euro area - forecast for 2015
United States United States - forecast for 2015
2
4
6
8
10
12
14
16
18
20
22
24
26
28
-
- 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
2
4
6
8
10
12
14
16
18
20
22
24
26
28
euro area euro area - forecast for 2015
United States United States - forecast for 2015
Sources: Eurostat and European Commission (AMECO, Autumn 2014 forecast). Note: Data for GR, ES, CY and SK according to ESA95. The hatched area indicates the minimum-maximum range across euro area countries.
Sources: Eurostat and European Commission (AMECO, Autumn 2014 forecast). Note: The hatched area indicates the minimum-maximum range across euro area countries.
S.1.3 Citigroup Economic Surprise Index S.1.4 Exchange rates
(1 Jan. 2008 - 14 Nov 2014) (1 Jan. 2007 - 14 Nov 2014; units of national currency per euro)
euro area United Kingdom United States Japan
-200
-150
-100
-50
0
50
100
150
200
2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
-200
-150
-100
-50
0
50
100
150
200
USD GBP JPY (divided by 100) CHF
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0.60
0.80
1.00
1.20
1.40
1.60
1.80
Source: Bloomberg. Note: A positive reading of the index suggests that economic releases have, on balance, been more positive than consensus expectations.
Sources: Bloomberg and ECB calculations.
STAT I ST ICAL ANNEX
S 2 ECB Financial Stability Review November 2014
S.1.5 Quarterly changes in gross external debt
S.1.6 Current account balances in selected external
surplus and deficit economies (2014 Q1; percentage of GDP) (1997 - 2019; USD billions)
-40 -35 -30 -25 -20 -15 -10
-5 0 5
10 15 20 25 30 35
BE EE GR FR CY MT AT SI FI DE IE ES IT LV NL PT SK
0
120
240
360
480
600
720
840
960
1080
1200
general government (left-hand scale) MFIs (left-hand scale) other sectors 1 (left-hand scale) direct investment/inter-company lending (left-hand scale) gross external debt 2 (right-hand scale)+
-1000
-500
0
500
1000
1500
1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 -1000
-500
0
500
1000
1500
euro area United States Japan
China oil exporters
Source: ECB. Notes: For Luxembourg, quarterly changes were 0.1% for general government, -11.8% for MFIs, 13.7% for other sectors and 121.1% for direct investment/inter-company lending. Gross external debt was 5,482% of GDP. Comparable data for Ireland for 2014 Q1 is not available. 1) Non-MFIs, non-financial corporations and households. 2) Gross external debt as a percentage of GDP.
Source: IMF World Economic Outlook. Notes: Oil exporters refers to the OPEC countries, Indonesia, Norway and Russia. Figures for 2014 to 2019 are forecasts.
S.1.7 Current account balances (in absolute amounts) in
selected external surplus and deficit economies
S.1.8 Foreign exchange reserve holdings
(1997 - 2019; percentage of world GDP) (Aug. 2009 - Aug. 2014; percentage of 2009 GDP)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
all large surplus/deficit economies United States China
0
25
50
75
100
125
150
175
2010 2011 2012 2013 2014 0
25
50
75
100
125
150
175
advanced economies emerging Asia
CEE/CIS Latin America
Source: IMF World Economic Outlook. Notes: All large surplus/deficit economies refers to oil exporters, the EU countries, the United States, China and Japan. Figures for 2014 to 2019 are forecasts.
Sources: Bloomberg, IMF World Economic Outlook and IMF International Financial Statistics. Note: CEE/CIS stands for central and eastern Europe and the Commonwealth of Independent States.
STAT IST ICAL ANNEX
S 3 ECB
Financial Stability Review November 2014
S.1.9 General government deficit/surplus (+/-)
S.1.10 General government gross debt
(percentage of GDP) (percentage of GDP, end of period)
-2
0
2
4
6
8
10
12
14
16
18
- - - -
- -
- - - -
- - - - - - - -
EA BE EE GR FR CY MT AT SI FI DE IE ES IT LU NL PT SK
-2
0
2
4
6
8
10
12
14
16
18
four-quarter moving sum in Q2 2014
- European Commission forecast for 2015 European Commission forecast for 2016
0
20
40
60
80
100
120
140
160
180
- - -
-
-
-
- - -
-
-
- - -
-
- - -
EA BE EE GR FR CY MT AT SI FI DE IE ES IT LU NL PT SK
0
20
40
60
80
100
120
140
160
180
of which held by non-residents gross debt at end-Q1 2014
- European Commision forecast for 2015 European Commision forecast for 2016
Sources: National data, European Commission (AMECO, Autumn 2014 forecast) and ECB calculations. Notes: Data on four quarter moving sum refer to accumulated deficit/surplus in the relevant quarter and the three previous quarters expressed as a percentage of GDP.
Sources: National data, European Commission (AMECO, Autumn 2014 forecast) and ECB calculations based on ESA95 data. Notes: Government debt data for Q1 2014 are not available for Ireland and the Netherlands.
S.1.11 Household debt-to-gross disposable income ratio
S.1.12 Household debt-to-total financial assets ratio
(percentage of disposable income) (Q1 2009- Q2 2014; percentages)
-50
-25
0
25
50
75
100
125
150
175
200
225
250
275
EA US DE IE ES IT LV NL PT SK UK JP BE EE GR FR CY LU AT SI FI
-50
-25
0
25
50
75
100
125
150
175
200
225
250
275
2007 debt change in debt between 2007 and 2013
0
20
40
60
80
100
2009 2010 2011 2012 2013 0
20
40
60
80
100
euro area United Kingdom
United States Japan
Sources: ECB, Eurostat, US Bureau of Economic Analysis and Bank of Japan. Notes: Gross disposable income adjusted for the change in net equity of households in pension fund reserves. For Luxembourg initial debt data refer to 2008, change in debt refers to 2008 and 2012. For Japan, Estonia, Greece, Cyprus, Latvia and Slovakia change in debt refers to 2007 and 2012. Data for Malta are not available. The figures are based on both ESA2010 and ESA95 methodology.
Sources: ECB and ECB calculations, Eurostat, US Bureau of Economic Analysis and Bank of Japan. Note: The hatched/shaded areas indicate the minimum-maximum and interquartile ranges across euro area countries. The figures are based on on both ESA2010 and ESA95 methodology.
S 4 ECB Financial Stability Review November 2014
S.1.13 Corporate debt-to-GDP and leverage ratios
S.1.14 Annual growth of MFI credit to the private sector in
the euro area (percentages) (Jan. 2006 - Sep. 2014; percentage change per annum)
-50
0
50
100
150
200
250
300
350
400
450
500
- -
- -
- - - - - -
- - - - -
- - - - - -
-
EA US DE IE ES IT LV MT AT SI FI UK JP BE EE GR FR CY LU NL PT SK
-10
0
10
20
30
40
50
60
70
80
90
100
2007 debt (left-hand scale) change in debt between 2007 and 2013 (left-hand scale)
- 2013 leverage (right-hand scale)
-20
-10
0
10
20
30
40
2006 2007 2008 2009 2010 2011 2012 2013 2014 -20
-10
0
10
20
30
40
euro area
Sources: ECB, Eurostat, US Bureau of Economic Analysis and Bank of Japan. Note: The figures for Japan and Great Britain are based on ESA95 methodology. For Germany, Estonia and Latvia initial debt data refer to 2012, change in debt refers to 2012 and 2013. For Malta initial debt data refer to 2009, change in debt refers to 2009 and 2013.
Sources: ECB and ECB calculations. Notes: MFI sector excluding the Eurosystem. Credit to the private sector includes loans to, and holdings of securities other than shares of, non-MFI residents excluding general government; MFI holdings of shares, which are part of the definition of credit used for monetary analysis purposes, are excluded. The hatched/shaded areas indicate the minimum-maximum and interquartile ranges across euro area countries.
S.1.15 Changes in credit standards for residential
mortgage loans
S.1.16 Changes in credit standards for loans to large
enterprises (Q1 2003 - Q4 2014; percentages) (Q1 2003 - Q4 2014; percentages)
-50
-25
0
25
50
75
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 -50
-25
0
25
50
75
euro area (loans to households for house purchase) United States (all residential mortgage loans) United States (prime residential mortgage loans) United States (non-traditional residential mortgage loans) United Kingdom (secured credit to households)
-50
-25
0
25
50
75
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 -50
-25
0
25
50
75
euro area (loans to large enterprises) United States (commercial and industrial loans to large and medium-sized enterprises) United Kingdom (large and medium-sized enterprises)
Sources: ECB, Federal Reserve System and Bank of England. Notes: Weighted net percentage of banks contributing to the tightening of standards over the past three months. Data for the United Kingdom refer to the net percentage balances on secured credit availability to households and are weighted according to the market share of the participating lenders. Data are only available from the second quarter of 2007 and have been inverted for the purpose of this chart. For the United States, the data series for all residential mortgage loans was discontinued owing to a split into the prime, non-traditional and sub-prime market segments from the April 2007 survey onwards.
Sources: ECB, Federal Reserve System and Bank of England. Notes: Weighted net percentage of banks contributing to the tightening of standards over the past three months. Data for the United Kingdom refer to the net percentage balances on corporate credit availability and are weighted according to the market share of the participating lenders. Data are only available from the second quarter of 2007 and have been inverted for the purpose of this chart.
STAT IST ICAL ANNEX
S 5 ECB
Financial Stability Review November 2014
S.1.17 Changes in residential property prices
S.1.18 Changes in commercial property prices
(Q1 1999 - Q2 2014; annual percentage changes) (Q1 2006 - Q2 2014; capital value; annual percentage changes)
-60
-40
-20
0
20
40
60
2000 2002 2004 2006 2008 2010 2012 -60
-40
-20
0
20
40
60 euro area United States
-60
-50
-40
-30
-20
-10
0
10
20
30
40
2006 2007 2008 2009 2010 2011 2012 2013 -60
-50
-40
-30
-20
-10
0
10
20
30
40 euro area
Sources: National data and ECB calculations. Notes: The target definition for residential property prices is total dwellings (whole country), but there are national differences. The hatched/shaded areas indicate the minimum-maximum and interquartile ranges across euro area countries.
Sources: experimental ECB estimates based on IPD data and national data for Germany and Italy. Note: The hatched/shaded areas indicate the max.-min./interquartile range across EA countries, except DE, EE, GR, CY, LV, LU, MT, SI, SK and FI.
2 FINANCIAL MARKETS
S 6 ECB Financial Stability Review November 2014
S.2.1 Global risk aversion indicator
S.2.2 Financial market liquidity indicator for the euro
area and its components (3 Jan. 2000 - 14 Nov 2014) (4 Jan. 1999 - 14 Nov 2014)
-4
-2
0
2
4
6
8
10
12
2000 2002 2004 2006 2008 2010 2012 2014 Aug Sep Oct 2014
-4
-2
0
2
4
6
8
10
12
composite indicator foreign exchange, equity and bond markets money market
-6
-5
-4
-3
-2
-1
0
1
2
3
2000 2002 2004 2006 2008 2010 2012 2014 Q3 Q4 2014
-6
-5
-4
-3
-2
-1
0
1
2
3
Sources: Bloomberg, Bank of America Merrill Lynch, UBS, Commerzbank and ECB calculations. Notes: The indicator is constructed as the first principal component of five currently available risk aversion indicators. A rise in the indicator denotes an increase of risk aversion. For further details about the methodology used, see ECB, ’’Measuring investors’ risk appetite’’, Financial Stability Review, June 2007.
Sources: ECB, Bank of England, Bloomberg, JPMorgan Chase & Co., Moody’s KMV and ECB calculations. Notes: The composite indicator comprises unweighted averages of individual liquidity measures, normalised from 1999 to 2006 for non-money market components and over the period 2000 to 2006 for money market components. The data shown have been exponentially smoothed. For more details, see Box 9 in ECB, Financial Stability Review, June 2007.
S.2.3 Spreads between interbank rates and repo rates
S.2.4 Spreads between interbank rates and overnight
indexed swap rates (3 Jan. 2003 - 14 Nov 2014; basis points; 1-month maturity; 20-day moving average) (1 Jan. 2007 - 14 Nov 2014; basis points: 3-month maturity)
EUR GBP
USD JPY
-50
0
50
100
150
200
250
300
2004 2006 2008 2010 2012 2014 Aug Sep Oct 2014
-50
0
50
100
150
200
250
300
EUR GBP
USD JPY
-50
0
50
100
150
200
250
300
350
400
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
-50
0
50
100
150
200
250
300
350
400
Sources: Thomson Reuters, Bloomberg and ECB calculations. Notes: Due to the lack of contributors, the series for GBP stopped in October 2013.
Sources: Thomson Reuters , Bloomberg and ECB calculations.
STAT IST ICAL ANNEX
S 7 ECB
Financial Stability Review November 2014
S.2.5 Slope of government bond yield curves
S.2.6 Sovereign credit default swap spreads for
euro area countries (2 Jan. 2006 - 14 Nov 2014; basis points) (1 Jan. 2007 - 14 Nov 2014; basis points; senior debt; five-year maturity)
euro area (AAA-rated bonds) euro area (all bonds) United Kingdom United States
-100
0
100
200
300
400
2006 2008 2010 2012 2014 Aug Sep Oct 2014
-100
0
100
200
300
400
median
0
200
400
600
800
1000
1200
1400
1600
1800
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0
200
400
600
800
1000
1200
1400
1600
1800
Sources: European Central Bank, Bank for International Settlements, Bank of England and Federal Reserve System. Notes: The slope is defined as the difference between ten-year and one-year yields. For the euro area and the United States, yield curves are modelled using the Svensson model; a variable roughness penalty model is used to model the yield curve for the United Kingdom.
Sources: Thomson Reuters and ECB calculations. Notes: The hatched/shaded areas indicate the minimum-maixmum and interquartile ranges across national sovereign CDS spreads in the euro area. Following the decision by the International Swaps Derivatives Association that a credit event had occurred, Greek sovereign CDS were not traded between 9 March 2012 and 11 April 2012. Due to lack of contributors, Greek sovereign CDS spread is not available between 1st of March and 21 May 2013. For presentational reasons, this chart has been truncated.
S.2.7 iTraxx Europe five-year credit default swap
indices
S.2.8 Spreads over LIBOR of selected European AAA-rated
asset-backed securities (1 Jan. 2007 - 14 Nov 2014; basis points) (26 Jan. 2007 - 14 Nov 2014; basis points)
iTraxx Europe iTraxx Europe High Volatility iTraxx Europe Crossover Index (sub-investment-grade reference)
0
200
400
600
800
1000
1200
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0
200
400
600
800
1000
1200
RMBS spread range auto loans consumer loans commercial mortgage-backed securities
0
300
600
900
1200
1500
1800
2100
2400
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0
300
600
900
1200
1500
1800
2100
2400
Source: Bloomberg. Source: JPMorgan Chase & Co. Note: In the case of residential mortgage-backed securities (RMBSs), the spread range is the range of available individual country spreads in Greece, Ireland, Spain, Italy, the Netherlands, Portugal and the United Kingdom.
S 8 ECB Financial Stability Review November 2014
S.2.9 Price/earnings ratio for the euro area stock market
S.2.10 Equity indices
(3 Jan. 2005 - 14 Nov 2014; ten-year trailing earnings) (2 Jan. 2001 - 14 Nov 2014; index: Jan. 2001 = 100)
main index banking sector non-financial corporations insurance sector
0
5
10
15
20
25
30
2006 2008 2010 2012 2014 Aug Sep Oct 2014
0
5
10
15
20
25
30
Standard & Poor’s 500 index Standard & Poor’s 500 Banks index KBW Bank Sector index Dow Jones EURO STOXX 50 index Dow Jones EURO STOXX Banks index
0
20
40
60
80
100
120
140
160
2002 2004 2006 2008 2010 2012 2014 Aug Sep Oct 2014
0
20
40
60
80
100
120
140
160
Sources: Thomson Reuters and ECB calculations. Note: The price/earnings ratio is based on prevailing stock prices relative to an average of the previous ten years of earnings.
Source: Bloomberg.
S.2.11 Implied volatilities
S.2.12 Payments settled by the large-value payment systems
TARGET2 and EURO1 (2 Jan. 2001 - 14 Nov 2014; percentages) (Jan. 2004 - Sep. 2014; volumes and values)
Standard & Poor’s 500 index KBW Bank Sector index Dow Jones EURO STOXX 50 index Dow Jones EURO STOXX Banks index
0
20
40
60
80
100
120
140
2002 2004 2006 2008 2010 2012 2014 Aug Sep Oct 2014
0
20
40
60
80
100
120
140
100
150
200
250
300
350
400
450
500
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 0
500
1000
1500
2000
2500
3000
3500
volume EURO1 (thousands, left-hand scale) volume TARGET2 (thousands, left-hand scale) value EURO1 (billions, right-hand scale) value TARGET2 (billions, right-hand scale)
Sources: Bloomberg and Thomson Reuters Datastream. Source: ECB. Notes: TARGET2 is the real-time gross settlement system for the euro. TARGET2 is operated in central bank money by the Eurosystem. TARGET2 is the biggest large-value payment system (LVPS) operating in euro. The EBA CLEARING Company’s EURO1 is a euro-denominated net settlement system owned by private banks, which settles the final positions of its participants via TARGET2 at the end of the day. EURO1 is the second-biggest LVPS operating in euro.
STAT IST ICAL ANNEX
S 9 ECB
Financial Stability Review November 2014
S.2.13 Volumes and values of foreign exchange trades settled
via the Continuous Linked Settlement Bank
S.2.14 Value of securities held in custody by CSDs
and ICSDs (Jan. 2004 - Sep. 2014; volumes and values) (2013; EUR trillions; settlement in all currencies)
0
100
200
300
400
500
600
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 0
500
1000
1500
2000
2500
volume of transactions (thousands, left-hand scale) value of transactions (billions equivalent, right-hand scale)
2
3
4
5
6
7
8
9
10
11
1 2 3 4 5 6 7 8 2
3
4
5
6
7
8
9
10
11
Source: ECB. Notes: The Continuous Linked Settlement Bank (CLS) is a global financial market infrastructure which offers payment-versus-payment (PvP) settlement of foreign exchange (FX) transactions. Each PvP transaction consists in two legs. The figures above count only one leg per transaction. CLS transactions are estimated to cover about 60% of the global FX trading activity.
Source: ECB. Notes: CSDs stands for central securities depositaries and ICSDs for international central securities depositaries. 1 - Euroclear Bank (BE); 2 - Clearstream Banking Frankfurt - CBF (DE); 3 - Euroclear France; 4 - Clearstream Banking Luxembourg-CBL; 5 - CRESTCo (UK); 6 - Monte Titoli (IT); 7 - Iberclear (ES); 8 - Remaining 40 CSDs in the EU.
S.2.15 Value of securities settled by CSDs and ICSDs
S.2.16 Value of transactions cleared by central
counterparties (2013; EUR trillions; settlement in all currencies) (2013; EUR trillions)
0
50
100
150
200
250
300
350
1 5 3 4 7 6 2 8 0
50
100
150
200
250
300
350
0
100
200
300
1 2 3 4 5 6 0
100
200
300
Source: ECB. Note: See notes of Chart S.2.14.
Source: ECB. Notes: 1 - EUREX Clearing AG (DE); 2 - ICE Clear Europe (UK); 3 - LCH Clearnet Ltd; 4 - LCH Clearnet SA (FR); 5 - CC&G (IT); 6 - Others. The chart includes outright and repo transactions, financial and commodity derivatives.
3 FINANCIAL INSTITUTIONS
S 10 ECB Financial Stability Review November 2014
S.3.1 Return on shareholders' equity for euro area
significant banking groups
S.3.2 Return on risk-weighted assets for euro area
significant banking groups (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range distribution across significant banking groups) distribution across significant banking groups)
-60
-50
-40
-30
-20
-10
0
10
20
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
-60
-50
-40
-30
-20
-10
0
10
20
-83 -69
median for euro area large and complex banking groups median for global large and complex banking groups
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
-8
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
median for euro area large and complex banking groups median for global large and complex banking groups
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. Quarterly figures are annualised.
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. Quarterly figures are annualised.
S.3.3 Breakdown of operating income for euro area
significant banking groups
S.3.4 Diversification of operating income for euro area
significant banking groups (2010 - Q2 2014; percentage of total assets; weighted average) (2010 - Q2 2014; individual institutions’ standard deviation dispersion; 10th and 90th
percentile and interquartile range distribution across significant banking groups)
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
2010 2012 Q2 13 Q4 13 Q2 14 2011 2013 Q3 13 Q1 14
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
loan loss provisions net interest income net fee and commission income net trading income net other operating income
12
16
20
24
28
32
36
40
2010 2012 Q2 13 Q4 13 Q2 14 2011 2013 Q3 13 Q1 14
12
16
20
24
28
32
36
40
median for euro area large and complex banking groups median for global large and complex banking groups
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. Quarterly results are annualised. Annual and quarterly indicators are based on common samples of 66 and 27 significant banking groups in the euro area, respectively.
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. A value of "0" means full diversification, while a value of "50" means concentration on one source only. Annual and quarterly indicators are based on common samples of 68 and 27 significant banking groups in the euro area, respectively.
STAT IST ICAL ANNEX
S 11 ECB
Financial Stability Review November 2014
S.3.5 Actual and forecast earnings per share for euro area
significant banking groups
S.3.6 Lending and deposit spreads of euro area MFIs
(Q1 2008 - Q2 2015; EUR) (Jan. 2004 - Sep. 2014; percentage points)
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
2008 2009 2010 2011 2012 2013 2014 2015 -4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
median for the significant banking groups median for the euro area large and complex banking groups
-3
-2
-1
0
1
2
3
4
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 -3
-2
-1
0
1
2
3
4
lending to households lending to non-financial corporations deposits with agreed maturity by non-financial corporations deposits with agreed maturity by households
Sources: SNL Financial and ECB calculations. Note: The shaded area indicates the interquartile ranges across the diluted earnings per share of selected significant banking groups in the euro area.
Sources: ECB, Thomson Reuters and ECB calculations. Notes: Lending spreads are calculated as the average of the spreads for the relevant breakdowns of new business loans, using volumes as weights. The individual spreads are the difference between the MFI interest rate for new business loans and the swap rate with a maturity corresponding to the loan category’s initial period of rate fixation. For deposits with agreed maturity, spreads are calculated as the average of the spreads for the relevant break-downs by maturity, using new business volumes as weights. The individual spreads are the difference between the swap rate and the MFI interest rate on new deposits, where both have corresponding maturities.
S.3.7 Net loan impairment charges for euro area significant
banking groups
S.3.8 Total capital ratios for euro area significant banking
groups (2010 - Q3 2014; percentage of net interest income; 10th and 90th percentile (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range and interquartile range distribution across significant banking groups) distribution across significant banking groups)
0
30
60
90
120
150
180
210
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
0
30
60
90
120
150
180
210
median for euro area large and complex banking groups median for global large and complex banking groups
8
10
12
14
16
18
20
22
24
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
8
10
12
14
16
18
20
22
24
median for euro area large and complex banking groups median for global large and complex banking groups
Sources: SNL Financial and ECB calculations. Note: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis.
Sources: SNL Financial and ECB calculations. Note: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis.
S 12 ECB Financial Stability Review November 2014
S.3.9 Core Tier 1 capital ratios for euro area significant
banking groups
S.3.10 Contribution of components of the core Tier 1 capital
ratios to changes for euro area significant banking groups (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range (2010 - Q2 2014; percentages) distribution across significant banking groups)
5
7
9
11
13
15
17
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
5
7
9
11
13
15
17
median for euro area large and complex banking groups median for global large and complex banking groups
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
- -
- -
- - - - -
2010 2012 Q2 13 Q4 13 Q2 14 2011 2013 Q3 13 Q1 14
7
8
9
10
11
12
13
risk-weighted assets (left-hand scale) core Tier 1 capital (left-hand scale)
- core Tier 1 ratio (mean; right-hand scale)
Sources: SNL Financial and ECB calculations. Note: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis.
Sources: SNL Financial and ECB calculations. Note: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. Annual and quarterly indicators are based on common samples of 53 and 27 significant banking groups in the euro area, respectively.
S.3.11 Non-performing loan ratios for euro area significant
banking groups
S.3.12 Leverage ratios for euro area significant banking
groups (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range distribution across significant banking groups) distribution across significant banking groups)
0
5
10
15
20
25
30
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
0
5
10
15
20
25
30
median for euro area large and complex banking groups median for global large and complex banking groups
0
2
4
6
8
10
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
0
2
4
6
8
10
median for euro area large and complex banking groups median for global large and complex banking groups
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. The non-performing loan ratio is defined as the ratio of impaired customer loans to total customer loans.
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly frequency. Leverage is defined as the ratio of shareholder equity to total assets.
STAT IST ICAL ANNEX
S 13 ECB
Financial Stability Review November 2014
S.3.13 Risk-adjusted leverage ratios for euro area significant
banking groups
S.3.14 Liquid assets ratios for euro area significant banking
groups (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range (2010 - 2013; percentage of total assets; 10th and 90th percentile distribution across significant banking groups) and interquartile range distribution across significant banking groups)
2
4
6
8
10
12
14
16
18
20
22
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
2
4
6
8
10
12
14
16
18
20
22
median for euro area large and complex banking groups median for global large and complex banking groups
0
5
10
15
20
25
30
35
40
45
2010 2011 2012 2013 0
5
10
15
20
25
30
35
40
45
median for euro area large and complex banking groups median for global large and complex banking groups
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. Risk-adjusted leverage is defined as the ratio of shareholder equity to risk-weighted assets.
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements. Liquid assets comprise cash and cash equivalents as well as trading securities. Quarterly data are not included on account of the inadequate availability of interim results on the date of publication.
S.3.15 Customer loan-to-deposit ratios for euro area
significant banking groups
S.3.16 Interbank borrowing ratio for euro area significant
banking groups (2010 - Q3 2014; multiple; 10th and 90th percentile and interquartile range (2010 - Q3 2014; percentage of total assets; 10th and 90th percentile distribution across significant banking groups) and interquartile range distribution across significant banking groups)
0.5
1.0
1.5
2.0
2.5
3.0
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
0.5
1.0
1.5
2.0
2.5
3.0
median for euro area large and complex banking groups median for global large and complex banking groups
0
5
10
15
20
25
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
0
5
10
15
20
25
median for euro area large and complex banking groups median for global large and complex banking groups
Sources: SNL Financial and ECB calculations. Note: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis.
Sources: SNL Financial and ECB calculations. Note: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis.
S 14 ECB Financial Stability Review November 2014
S.3.17 Ratios of short-term funding to loans for euro area
significant banking groups
S.3.18 Issuance profile of long-term debt securities by euro
area significant banking groups (2010 - Q3 2014; percentages; 10th and 90th percentile and interquartile range (Oct. 2013 - Apr. 2015; EUR billions) distribution across significant banking groups)
0
10
20
30
40
50
60
2010 2012 Q3 13 Q1 14 Q3 14 2011 2013 Q4 13 Q2 14
0
10
20
30
40
50
60
median for euro area large and complex banking groups median for global large and complex banking groups
-100
-80
-60
-40
-20
0
20
40
60
2014 -100
-80
-60
-40
-20
0
20
40
60
corporate bonds medium-term notes asset-backed instruments net issuance
Sources: SNL Financial and ECB calculations. Notes: Includes publicly available data for significant banking groups that report annual financial statements and a subset of those banks that report on a quarterly basis. Interbank funding is used as the measure of short-term funding.
Sources: Dealogic DCM Analytics and ECB calculations. Notes: Net issuance is the total gross issuance minus scheduled redemptions. Dealogic does not trace instruments after their redemption, so that some of the instruments may have been redeemed early. Asset-backed instruments encompass asset-backed and mortgage-backed securities, as well as covered bond instruments.
S.3.19 Maturity profile of long-term debt securities for euro
area significant banking groups
S.3.20 Issuance of syndicated loans and bonds by euro area
banks (2006 - Oct. 2014; EUR billions) (Q1 2004 - Q3 2014; EUR billions)
0
100
200
300
400
500
600
700
1 year 3 years 5 years 7 years 9 years 2 years 4 years 6 years 8 years 10 years
0
100
200
300
400
500
600
700
average 2006-08 2009 2010 2011 2012 2013 Oct. 2014
0
50
100
150
200
250
300
350
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 0
50
100
150
200
250
300
350
bonds (excluding covered bonds, ABS and MBS) covered bonds syndicated loans granted to banks asset-backed and mortgage-backed securities (ABS and MBS)
Sources: Dealogic DCM Analytics and ECB calculations. Notes: Data refer to all amounts outstanding at the end of the corresponding year/month. Long-term debt securities include corporate bonds, medium-term notes, covered bonds, asset-backed securities and mortgage-backed securities with a minimum maturity of 12 months.
Sources: Dealogic DCM Analytics and Loan Analytics, ECB calculations.
STAT IST ICAL ANNEX
S 15 ECB
Financial Stability Review November 2014
S.3.21 Investment income and return on equity for a sample
of large euro area insurers
S.3.22 Gross-premium-written growth for a sample of large
euro area insurers (2011 - Q3 2014; percentages; 10th and 90th percentile and interquartile range (2009 - Q3 2014; percentage change per annum; 10th and 90th percentile and distribution) interquartile range distribution)
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
2011 2013 Q3 14 2012 Q2 14 2012 Q2 14 2011 2013 Q3 14
-15
-10
-5
0
5
10
15
20
Investment income (percentage of total assets)
Return on shareholders equity (percentages)
-20
-15
-10
-5
0
5
10
15
20
2009 2011 2013 Q2 14 2010 2012 Q1 14 Q3 14
-20
-15
-10
-5
0
5
10
15
20
Sources: Bloomberg, individual institutions’ reports and ECB calculations. Notes: Based on available figures for 21 euro area insurers and reinsurers.
Sources: Bloomberg, individual institutions’ reports, and ECB calculations. Note: Based on available figures for 21 euro area insurers and reinsurers.
S.3.23 Distribution of combined ratios for a sample of large
euro area insurers
S.3.24 Capital distribution for a sample of large euro area
insurers (2009 - Q3 2014; percentages; 10th and 90th percentile and interquartile range (2009 - Q3 2014; percentage of total assets; 10th and 90th percentile and interquartile distribution) range distribution)
86
88
90
92
94
96
98
100
102
104
106
108
2009 2011 2013 Q2 14 2010 2012 Q1 14 Q3 14
86
88
90
92
94
96
98
100
102
104
106
108
5
10
15
20
25
30
35
2009 2011 2013 Q2 14 2010 2012 Q1 14 Q3 14
5
10
15
20
25
30
35
Sources: Bloomberg, individual institutions’ reports and ECB calculations. Notes: Based on available figures for 21 euro area insurers and reinsurers.
Sources: Bloomberg, individual institutions’ reports and ECB calculations. Notes: Capital is the sum of borrowings, preferred equity, minority interests, policyholders’ equity and total common equity. Data are based on available figures for 21 euro area insurers and reinsurers.
S 16 ECB Financial Stability Review November 2014
S.3.25 Investment distribution for a sample of large euro
area insurers
S.3.26 Expected default frequency for banking groups
H1 2013 - H1 2014; percentage of total investments; minimum, maximum and (Jan. 2004 - Sep. 2014; percentages; weighted average) interquartile distribution)
0
10
20
30
40
50
60
70
80
H113 H114 H213 H113 H114 H213 H113 H114 H213 H113 H114 H213 H113 H114 H213
0
10
20
30
40
50
60
70
80 Government
bonds Corporate
bonds Structured
credit Equity Commercial
property
0
1
2
3
4
5
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 0
1
2
3
4
5
euro area significant banking groups euro area large and complex banking groups global large and complex banking groups
Sources: Individual institutions’ financial reports and ECB calculations. Notes: Equity exposure data exclude investments in mutual funds. Data are based on available figures for 14 euro area insurers and reinsurers.
Sources: Moody’s KMV and ECB calculations. Note: The weighted average is based on the amounts of non-equity liabilities.
S.3.27 Credit default swap spreads for euro area significant
banking groups
S.3.28 Credit default swap spreads for a sample of large
euro area insurers (1 Jan. 2008 - 14 Nov 2014; basis points; senior debt; five-year maturity) (3 Jan. 2007 - 14 Nov 2014; basis points; senior debt; five-year maturity)
median for euro area significant banking groups median for euro area large and complex banking groups median for global large and complex banking groups
0
200
400
600
800
1000
1200
1400
1600
1800
2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0
200
400
600
800
1000
1200
1400
1600
1800
median
0
100
200
300
400
500
600
700
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0
100
200
300
400
500
600
700
Sources: Thomson Reuters, Bloomberg and ECB calculations. Note: The hatched/shaded areas indicate the minimum-maximum and interquartile ranges across the CDS spreads of selected large banks. For presentational reasons, this chart has been truncated.
Sources: Thomson Reuters, Bloomberg and ECB calculations. Note: The hatched/shaded areas indicate the minimum-maximum and interquartile ranges across the CDS spreads of selected large insurers.
STAT IST ICAL ANNEX
S 17 ECB
Financial Stability Review November 2014
S.3.29 Stock performance of the euro area significant
banking groups
S.3.30 Stock performance of a sample of large euro area
insurers (3 Jan. 2007 - 14 Nov 2014 ; index: 2 Jan. 2007 = 100) (3 Jan. 2007 - 14 Nov 2014 ; index: 2 Jan. 2007 = 100)
median for euro area significant banking groups median for euro area large and complex banking groups median for global large and complex banking groups
0
20
40
60
80
100
120
140
160
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0
20
40
60
80
100
120
140
160
median
0
20
40
60
80
100
120
140
160
180
200
2007 2008 2009 2010 2011 2012 2013 2014 Aug Sep Oct 2014
0
20
40
60
80
100
120
140
160
180
200
Sources: Thomson Reuters , Bloomberg and ECB calculations. Note: The hatched/shaded areas indicate the minimum-maximum and interquartile this chart has been truncated.
Sources: Thomson Reuters , Bloomberg and ECB calculations. Note: The hatched/shaded areas indicate the minimum-maximum and interquartile ranges across equities of selected large insurers.
- FINANCIAL STABILITY REVIEW - NOVEMBER 2014
- CONTENTS
- FOREWORD
- OVERVIEW
- 1 MACRO-FINANCIAL AND CREDIT ENVIRONMENT
- 1.1 ONGOING MODERATE RECOVERY, BUT DOWNSIDE RISKS ON THE RISE
- Box 1 DOES THE GROWING IMPORTANCE OF EMERGING MARKET BANKS POSE A SYSTEMIC RISK?
- 1.2 STRUCTURAL REFORM AND FISCAL CONSOLIDATION NEEDS REMAIN HIGH, DESPITE CONTAINED SOVEREIGN STRESS
- 1.3 GRADUALLY IMPROVING FINANCING CONDITIONS IN THE NON-FINANCIAL PRIVATE SECTOR, BUT VULNERABILITIES REMAIN
- 2 FINANCIAL MARKETS
- 2.1 INTERBANK ACTIVITY IN EURO AREA MONEY MARKETS CONTINUES TO NORMALISE, BUT FRAGMENTATION REMAINS
- 2.2 YIELDS AT RECORD LOWS AMID A SLIGHT INCREASE IN CREDIT RISK PREMIA
- Box 2 STRUCTURAL AND SYSTEMIC RISK FEATURES OF EURO AREA INVESTMENT FUNDS
- Box 3 FINANCIAL MARKET VOLATILITY AND BANKING SECTOR LEVERAGE
- 3 EURO AREA FINANCIAL INSTITUTIONS
- 3.1 BALANCE SHEET REPAIR CONTINUES, BUT WEAK PROFITABILITY PERSISTS IN THE EURO AREA BANKING SECTOR
- Box 4 THE ECB’S COMPREHENSIVE ASSESSMENT EXERCISE
- Box 5 DO CONTINGENT CONVERTIBLE CAPITAL INSTRUMENTS AFFECT THE RISK PERCEPTIONS OF SENIOR DEBT HOLDERS?
- 3.2 THE EURO AREA INSURANCE SECTOR: RESILIENCE AMID CONTINUED HEADWINDS
- 3.3 MACRO-PRUDENTIAL POLICY MEASURES ANNOUNCED IN SEVERAL COUNTRIES
- 3.4 RESHAPING THE REGULATORY FRAMEWORK FOR FINANCIAL INSTITUTIONS, MARKETS AND INFRASTRUCTURES
- Box 6 REGULATORY INITIATIVES TO ENHANCE OVERALL LOSS-ABSORPTION CAPACITY
- SPECIAL FEATURES
- A FIRE-SALE EXTERNALITIES IN THE EURO AREA BANKING SECTOR
- Box A.1 THEORETICAL FRAMEWORK
- B CAPTURING THE FINANCIAL CYCLE IN EURO AREA COUNTRIES
- C INITIAL CONSIDERATIONS REGARDING A MACRO-PRUDENTIAL INSTRUMENT BASED ON THE NET STABLE FUNDING RATIO
- Box C.1 WHAT IS THE NET STABLE FUNDING RATIO?
- STATISTICAL ANNEX
- 1 MACRO-FINANCIAL AND CREDIT ENVIRONMENT
- 2 FINANCIAL MARKETS
- 3 FINANCIAL INSTITUTIONS
Foreign Portfolio Investors and Financial Sector Stability in Asia.pdf
Foreign Portfolio Investors and Financial Sector Stability in Asia Author(s): Jeong Yeon Lee Source: Asian Survey, Vol. 47, No. 6 (November/December 2007), pp. 850-871 Published by: University of California Press Stable URL: http://www.jstor.org/stable/10.1525/as.2007.47.6.850 .
Accessed: 13/02/2015 16:20
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
University of California Press is collaborating with JSTOR to digitize, preserve and extend access to Asian Survey.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
850
Asian Survey
, Vol. 47, Issue 6, pp. 850–871, ISSN 0004-4687, electronic ISSN 1533-838X. © 2007 by The Regents of the University of California. All rights reserved. Please direct all requests for permis- sion to photocopy or reproduce article content through the University of California Press’s Rights and Permissions website, at http://www.ucpressjournals.com/reprintInfo.asp. DOI: AS.2007.47.6.850.
Jeong Yeon Lee is Associate Professor at the Graduate School of Inter-
national Studies, Yonsei University, Seoul. The author would like to thank Nicholas Hope, Julie
Altman, and an anonymous referee for their helpful comments. Email:
�
�
.
1. Speech to the East Asia Economic Summit in Putrajaya, Malaysia, October 6, 2002.
FOREIGN PORTFOLIO INVESTORS AND FINANCIAL SECTOR STABILITY IN ASIA
Jeong Yeon Lee
Abstract
A careful review of the Asian crisis in 1997 reveals that despite widespread denunciation, hedge funds and other portfolio investors played only a minor role in the making of the crisis. To maintain financial sector stability, therefore, governments should focus on avoiding inconsistent policies rather than try to identify a “bad” class of investors and limit their activities.
Keywords: Asian crisis, global capital mobility, foreign portfolio investment, institutional investors, hedge funds
1. Introduction
In his speech to the East Asia Economic Summit in Octo- ber 2002, Mahathir Mohamad, then-prime minister of Malaysia, argued that “[h]edge funds had made arrangements to have huge sums at their disposal and even more that they could leverage from the friendly banks. And so began the rampage of the currency traders. Any country was fair game, but most of all the newly emerging economies, rich enough to be fleeced, but not powerful enough to fight back. . . . Perfectly good countries with enormous resources can be truly and really bankrupted.”
1
Dr. Mahathir’s remarks don’t always inspire widespread sympathy, but his denunciation of hedge funds tends to resonate well with audiences in developing countries. In fact, disparaging hedge funds seems to hold great populist appeal even beyond developing countries. In 2005, Gerhard Schröder, then chancellor of Germany, criticized hedge funds during the controversy surrounding the
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE
851
leadership of Deutsche Börse, the German stock exchange, saying: “There might be an argument in favor of closer scrutiny of hedge funds to check whether their philosophy is compatible with that of our society.”
2
Franz Müntefering, then chairman of the Social Democratic Party of Germany, even compared international investment companies to “locusts” picking off vulner- able German companies.
3
Although foreign direct investment (FDI) is generally regarded as a kind of long-term, stable investment flow and basically good news for economic growth, portfolio investors and hedge funds in particular tend to be viewed with mixed feelings. They are believed to bring economic benefits generally associated with cross-border capital flows as well as the signs of investor con- fidence that emerging market economies crave. At the same time, they are seen as a greedy, fickle type of investor that often profits from destabilizing finan- cial markets of host countries.
In the past few years, foreign portfolio investors have returned in droves to the Asian economies beset by the 1997 financial crisis. After years of slow and sometimes even negative inflows, since 2003 foreign portfolio investment has visibly rebounded in Malaysia, South Korea,
4
Indonesia, and Thailand. This rising presence of foreign investors obviously causes some jitters, as many ob- servers there feel a strong sense of déjà vu. Didn’t the Asian financial crisis occur in the first place because people like George Soros moved large sums of money into crisis countries and then quickly sucked them out? Shouldn’t gov- ernments therefore at least try to sort out those bad apples and ban them, in order to avoid another financial nightmare?
If any country ever tries to attain peace of mind simply by adopting policy measures targeted at potential speculators, it should think twice, because a careful review of the Asian crisis indicates that such a move would be largely off base. As it transpires, potential speculators such as hedge funds and other portfolio investors were not a primary mover of violent tides of capital flows in the region at the time. Rather, international banks were the real barbarians who led the havoc. In fact, one important lesson of the Asian crisis is that any investors can turn into barbarians once they perceive the existence of a one- way bet. This transformation may well be rooted in the very nature of cross- border capital flows.
2. Global Capital Flows
Economic theory clearly points to the tremendous potential benefits of cross-
border capital flows. Although it is not easy to present actual gains in tangible
2. “Schröder Orders Hedge Fund Controls Review,”
Financial Times,
May 14–15, 2005.
3. Ibid.
4. Hereafter, “Korea.”
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
852
ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
numbers, historical evidence supports the notion that considerable benefits
stem from global capital mobility. However, there are also risks associated
with global capital flows, as recent financial crises, which have become more
frequent and extensive over the years, plainly illustrate. Like the protagonist
with a dual personality in Robert Louis Stevenson’s famous novel,
The Strange Case of Dr. Jekyll and Mr. Hyde
, largely beneficial capital flows could all of a
sudden morph into a real menace under asymmetric information and imperfect
contract enforcement, which tend to be the norm rather than the exception in
the world in which we live.
Pros
Global capital flows obviously generate various economic benefits. Free flows
of capital across borders promote a more efficient allocation of world resources
by allowing savings to find their most productive use beyond their national bor-
ders. Consequently, countries with expanding investment opportunities can fi-
nance those opportunities without rapidly raising their domestic savings. As
the national income identity indicates, the financing of domestic investment is
not constrained by national saving when capital inflows from abroad are available.
In addition, cross-border capital movements provide investors with opportuni-
ties to increase risk-adjusted returns through international diversification
of portfolios.
Global capital flows also serve as a conduit for intertemporal trade and in-
ternational risk sharing. When a country is hit by negative shocks to domestic
production and income, the availability of foreign investors willing to lend can
reduce the need for households and firms to contract consumption and invest-
ment spending. Foreign investor lending can thus increase the scope for inter-
national risk sharing and smooth out consumption and hence business cycles.
5
Theoretical economics literature suggests that the welfare gains from interna-
tional risk sharing can indeed be substantial,
6
but official restrictions on capi-
tal mobility may prevent these potential gains from being fully realized. Karen
Lewis in fact confirms that consumption and output are more likely to move
together in countries with heavier restrictions on capital flows.
7
5. See Bruce Greenwald, Joseph Stiglitz, and Andrew Weiss, “Information Imperfections in
the Capital Markets and Macroeconomic Fluctuations,”
American Economic Review
74, Papers
and Proceedings (May 1984), pp. 194–99.
6. For example, see Maurice Obstfeld, “Risk-Taking, Global Diversification and Growth,”
ibid., 84:5 (December 1994), pp. 1310–29.
7. See “What Can Explain the Apparent Lack of International Consumption Risk-Sharing?”
Journal of Political Economy
104 (April 1996), pp. 267–97; and “Are Countries with Official In-
ternational Restrictions ‘Liquidity Constrained’?”
European Economic Review
41:6 (June 1997),
pp. 1079–1109.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE
853
In securities markets, capital flows from abroad can enlarge the investor
base and help deepen liquidity. Foreign investors have assumed particular im-
portance in expanding the investor base for government bonds in industrial
countries and reducing borrowing costs of those countries by broadening the
markets.
8
Furthermore, foreign institutional investors can contribute to the pro-
cess of financial innovation in domestic markets by introducing sophisticated
trading and investment strategies. Arbitrage activities resulting from those
strategies also can create additional liquidity.
9
Free capital flows tend to promote specialization in the production of many
financial services—and hence international trade in these services—since
their production generally exhibits economies of scale and scope.
10
If foreign
investors provide financial services as well as capital flows, the import of fi-
nancial services can bring additional gains to the domestic financial sector. It
can bring valuable know-how and foster the spread of good practices. Further-
more, the entry of foreign providers can make domestic producers of financial
services more efficient by enhancing competitive pressure.
11
Cons
Notwithstanding these substantial benefits, there are also risks associated with
cross-border capital flows. Incomplete information can make investors likely
to engage in herding behavior, causing steep market movements.
12
To the ex-
tent that foreign investors have an information disadvantage as compared with
domestic investors, the foreign presence can increase the extent of herding.
Some theoretical studies suggest that faced with investment opportunities
stretching over many foreign countries and a fixed cost of information about
each country, widely diversified investors may have very weak incentive to be
8. Ismail Dalla,
Asia’s Emerging Bond Markets
(Hong Kong: FT Financial Publishing, 1997).
9. Institutional investors themselves are much interested in market liquidity, and rarely invest
in illiquid markets. See Hans Blommenstein, “Institutional Investors, Pension Reform, and Emerg-
ing Securities Markets,” Working Paper, no. 359, Office of the Chief Economist, Inter-American
Development Bank (Washington, D.C., 1997).
10. Barry Eichengreen, Michael Mussa, et al.,
Capital Account Liberalization: Theoretical and Practical Aspects
, International Monetary Fund (IMF) Occasional Paper, no. 172 (Washington,
D.C., 1998).
11. However, it is also plausible that increased competition may lead to decreases in franchise
value and thus give domestic banks an incentive to assume excessive risks. Therefore, suddenly
and fully opening a weak banking sector to foreign competition may cause a crisis. See ibid.
12. Herding behavior is not necessarily “irrational.” The behavior is perfectly rational in an
economics sense if there exist (1) payoff externalities, where the payoffs of taking a position in-
crease with the number of others taking the same position; (2) principal-agent problems, where
fund managers find it in their interest to “hide in the herd” to make the assessment of their perfor-
mance more difficult; and/or (3) information cascades, where investors infer information from the
positions taken by others and optimally follow them. See ibid.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
854
ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
informed about all the countries in which they invest.
13
Woochan Kim and
Shang-Jin Wei provide evidence consistent with models suggesting informa-
tion disadvantages of foreign investors.
14
Cross-border capital flows may also play a key role in the contagion of crisis.
One possibility is what Paul Masson calls the “monsoonal effect” caused by
sudden shifts in economic conditions of industrial countries.
15
These shifts
may, just like the monsoon, simultaneously trigger crises in a number of emerg-
ing market economies, not least because of the importance of external “push”
factors in determining capital flows to these economies. Another possible ex-
planation for contagion in the context of emerging market securities is treat-
ment by institutional investors of these securities as a separate asset class.
16
If
equity investors do not sufficiently discriminate among risks in individual
emerging markets, a crisis in one country may lead to across-the-board sales
of emerging market stocks.
Other possible mechanisms through which foreign investors can contribute
to contagion include demonstration effects. Information revealed about one as-
set or country can prompt foreign investors to reassess other similar assets or
countries.
17
Particularly in the Asian crisis, these demonstration effects appear
to have been a most plausible channel of contagion. Effectively, the Thai crisis
“woke up” foreign investors to reexamine the creditworthiness of other Asian
countries exhibiting similar weaknesses.
18
An alternative explanation is that
13. For example, see Guillermo Calvo and Enrique Mendoza, “Rational Herd Behavior and the
Globalization of Securities Markets,” Institute for Empirical Macroeconomics Discussion Paper,
no. 120–1, Federal Reserve Bank of Minneapolis (August 1997). In their model, international in-
vestors experience weaker incentive to acquire information when financial markets in many coun-
tries are being liberalized at the same time. In this kind of model, financial market liberalization
and the presence of foreign investors increase the extent of herding permanently. However, it is
also possible that incomplete information is not a permanent problem but exists only temporarily.
For example, international investors possess little information about recently liberalized markets,
but they overcome this problem through learning over time. See Philippe Bacchetta and Eric van
Wincoop, “Capital Flows to Emerging Markets: Liberalization, Overshooting, and Volatility,”
National Bureau of Economic Research (NBER), Working Paper, no. 6530 (Cambridge, Mass.:
NBER, 1998).
14. Woochan Kim and Shang-Jin Wei, “Foreign Portfolio Investors Before and During a Cri-
sis,” ibid., Working Paper, no. 6968 (1999).
15. Paul Masson, “Contagion: Monsoonal Effects, Spillovers, and Jumps between Multiple
Equilibria,” IMF Working Paper, no. 98/142 (Washington, D.C.: IMF, 1998).
16. For statistical evidence on the hypothesis of investors’ treating emerging market stocks as
a separate asset class, see IMF,
Private Market Financing for Developing Countries
(Washington,
D.C., 1995).
17. See Mervyn King and Sushil Wadhwani, “Transmission of Volatility between Stock Mar-
kets,”
Review of Financial Studies
3:1 (Spring 1990), pp. 5–33.
18. See Morris Goldstein,
The Asian Financial Crisis: Causes, Cures, and Systemic Implica- tions
(Washington, D.C.: Institute for International Economics, 1998). Domestic developments
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE
855
although investors were awake, they expected that governments would rescue
local institutions once in trouble. The Thai case “demonstrated” the limited
room for government intervention during a systemic crisis.
A crisis can sometimes spread to unexpected places through demonstration
effects. For example, if asset prices are determined by common factors as well
as idiosyncratic factors, then even a shock to an idiosyncratic factor in one market
may induce the adjustment of positions in other markets among uninformed
investors who are uncertain about the nature of the shock that has occurred.
19
Alternatively, if informed investors trade not only based on new information
but also for non-informational reasons, then it is likely that uninformed investors
may follow informed investors even when the trade of informed investors was
driven by non-informational reasons and there was effectively no new infor-
mation revealed about fundamentals.
20
Margin calls are often cited as one of the
possible non-informational reasons.
One should also note that the presence of foreign investors can amplify
risks associated with domestic financial liberalization. Domestic financial lib-
eralization can narrow the margins of financial institutions by sharpening
competition among them. If the net worth of financial institutions becomes
negative because of increased competition, they would have an incentive to
gamble for redemption by pursuing risky activities. When not accompanied by
adequate prudential oversight, domestic financial liberalization can allow fi-
nancial institutions in distress to engage in such behavior more easily. Foreign
investors can further facilitate gambling for redemption by offering additional
offshore funding.
21
3. Foreign Portfolio Investors
Global capital flows mainly consist of cross-border flows of direct investment,
portfolio investment, and international bank lending. The past couple of de-
cades have seen an explosive growth of these global capital flows, with total
inflows exceeding $4 trillion in 2004. Of these inflows, portfolio investment
flows account for more than 40%. Which is the major force behind these huge
flows of portfolio capital across borders? Although hedge funds have attracted
much attention—from Dr. Mahathir and others—our typical foreign portfolio
investors are traditional institutional investors from the Organization for
also contributed to this process as witnessed in Korea when Hanbo’s collapse in January 1997
raised investor awareness about the financial vulnerability of other highly leveraged chaebols and
Korean banks with great exposure to them.
19. See King and Wadhwani, “Transmission of Volatility between Stock Markets.”
20. See Guillermo Calvo, “Contagion in Emerging Markets,” mimeo (College Park, Md.:
University of Maryland, 1999).
21. See Eichengreen, Mussa, et al.,
Capital Account Liberalization
.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
856
ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
Economic Cooperation and Development (OECD), comprising pension funds,
insurance companies, and mutual funds. These three main classes of institu-
tional investors hold total assets of over $45 trillion, about 15% of which are
foreign assets. Their total assets in fact amount to as much as 150% of the total
gross domestic product (GDP) of OECD countries. In contrast, hedge funds
controlled total assets of about $930 billion as of 2004, about 2.8% of the
total OECD GDP (see Table 1).
Of the three main classes of institutional investors, pension funds and insur-
ance companies were traditionally the more dominant players in OECD capital
markets. For example, their combined assets were almost four times as large
as mutual fund assets in the mid-1990s. However, mutual fund assets have
grown much faster over the past decade and their total asset size is now more
or less on a par with that of pension funds or insurance companies. The data
on portfolio capital inflows in Korea also show the growing importance of
mutual funds in recent years (see Table 2).
22
Over 2001–04, investment com-
panies including mutual funds accounted for about two-thirds of cumulative
table
1
Assets of Institutional Investors
1990 1995 2000 2001 2002 2003 2004
(In trillions of U.S. dollars)
Institutional investors 13.8 23.5 39.0 39.4 36.2 46.8 ––
Pension funds 3.8 6.7 13.5 12.7 11.4 15.0 15.3
Insurance companies 4.9 9.1 10.1 11.5 10.2 13.5 14.5
Investment companies
a
2.6 5.5 11.9 11.7 11.3 14.0 16.2
Hedge funds 0.03 0.10 0.41 0.56 0.59 0.80 0.93
Other institutional investors 2.4 2.2 3.1 3.0 2.7 3.4 ––
(In % of total OECD GDP)
Institutional investors 77.6 97.8 152.1 155.3 136.4 157.2 ––
Pension funds 21.2 27.8 52.6 50.1 42.9 50.4 46.4
Insurance companies 27.8 37.8 39.4 45.3 38.4 45.4 44.0
Investment companies 14.8 22.7 46.3 45.9 42.7 47.2 49.0
Hedge funds 0.1 0.4 1.6 2.2 2.2 2.7 2.8
Other institutional investors 13.6 9.1 12.3 11.7 10.1 11.5 ––
SOURCE: IMF, Global Financial Stability Report (September 2005).
a
Investment companies include mutual funds, closed-end and managed investment companies,
and unit investment trusts.
22. The breakdown of net portfolio inflows in Korea by investor types is only available for
recent years starting from 2001.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE
857
net portfolio inflows in Korea. Pension funds were also important foreign
portfolio investors in Korea during that time, with their share of cumulative
net portfolio inflows exceeding 13%. In contrast, the share of hedge funds was
only about 1%, suggesting that hedge funds are relatively minor players among
foreign institutional investors in Korea.
Institutional Investors from the OECD
Pension funds, insurance companies, and mutual funds operate under vastly
different investment guidelines, reflecting their particular investment objectives
and fiduciary mandates. Furthermore, they are subject to different regulatory
and tax treatment and have different risk appetites. The different classes of in-
stitutional investors also have different structures of liabilities. These internal
and external factors exert strong influence on asset allocations of institutional in-
vestors. Because these investors’ decisions on foreign holdings are just part of
the overall asset allocation process, the aforementioned factors are also of great
importance in understanding forces behind global portfolio flows. This section
reviews some of these institutional factors, both external and internal, which
each type of OECD institutional investor faces. In addition to the three main
classes of institutional investors, hedge funds are also covered.
Pension funds.
Many OECD countries impose regulatory caps on pension fund
holdings of certain asset classes deemed to be relatively risky, with a view to
protecting pension fund beneficiaries or benefit insurers.
23
There are often
ceilings on foreign asset holdings as well, motivated by concerns about insuf-
ficient local information, the different regulatory standards for issuing securities,
and other additional risks associated with foreign investment.
24
The need to
table
2
Net Portfolio Capital Inflows in Korea
(in billions of U.S. dollars)
2001 2002 2003 2004
Institutional investors 7.3
�
0.6 13.9 8.9
Pension funds 1.7 0.4 0.1 1.7
Insurance companies 0.7
�
1.8
�
0.6
�
0.9
Investment companies 3.8 1.4 10.2 4.3
Hedge funds 0.5
�
0.4 0.4
�
0.2
Other institutional investors 0.7
�
0.3 3.8 4.0
SOURCE: Bank of Korea.
23. See Blommenstein, “Institutional Investors.”
24. See E. Philip Davis,
Pension Funds: Retirement-Income Security and Capital Markets: An International Perspective
(Oxford: Clarendon Press, 1995).
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
858
ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
create stable demand for government securities also motivates governments in
some countries to resort to quantitative regulation of pension fund holdings.
25
In other OECD countries where outright quantitative regulation is not in
place, the “prudent man rule” often applies instead.
26
This rule generally requires
pension fund managers to diversify and to follow “seasoned” courses that other
careful professionals would favor. As a result, this rule tends to lead to the con-
centration of pension fund investments in listed securities of domestic blue-
chip corporations. It may also act to reduce the diversity of market behavior.
Pension funds hold long-term liabilities that entail little liquidity risk but
high longevity and inflation risks.
27
Given the nature of their liabilities, pen-
sion funds have traditionally concentrated their portfolios on equities, with the
goal of taking advantage of a long-term equity premium above bond yields
that, they believe, could serve as an extra cushion for the associated risks.
28
The OECD average of the equity share in pension fund assets is close to 50%.
29
Pension funds also tend to hold a higher share of foreign assets than other
types of institutional investors.
30
In fact, the growing pension sector in the
OECD has been the major source of steady portfolio flows into developing
countries.
31
Insurance companies.
In contrast to pension funds, whose liabilities are locked
in for long periods, insurance companies face substantial liquidity risk through
premature withdrawals and policy loans. As a result, insurance companies are
generally subject to much stricter portfolio regulations than pension funds. In
all OECD countries, insurance companies are only allowed to hold assets that
have been approved by regulatory authorities.
32
In addition, regulatory ceilings
on individual asset classes are often in place even though the actual asset alloca-
tions of insurance companies usually stay far below these regulatory caps.
33
Regulatory authorities in most OECD countries also carefully review the
maturity profiles of assets and liabilities when they examine the solvency of
insurance companies.
34
A serious maturity gap would be particularly problem-
atic for life insurance companies under falling market interest rates, because
they tend to have long-term liabilities, often with implicit interest guarantees.
25. See Blommenstein, “Institutional Investors.”
26. Ibid.
27. IMF,
Global Financial Stability Report
(Washington, D.C., September 2005).
28. Ibid. (September 2004).
29. See ibid. (September 2005).
30. Ibid.
31. Blommenstein, “Institutional Investors.”
32. Ibid.
33. Ibid.
34. Ibid.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE
859
The need for maturity matching has motivated insurance companies to con-
centrate their portfolios on fixed-income instruments. The IMF estimation in-
dicates that the average bond shares in the total financial assets of OECD
insurance companies remained somewhat stable at 57% between 1997 and 2003,
whereas equities made up only about 24% during the same period.
35
Almost all OECD countries have instituted explicit regulatory requirements
for some degree of currency matching of assets and liabilities.
36
These regula-
tory factors partly contribute to a somewhat lower level of global diversification
of portfolios of insurance companies compared with those of other institutional
investors. According to IMF estimates, insurance companies in major OECD
countries held only about 8% of their financial assets in foreign bonds and eq-
uities in 1997.
37
The share rose to about 14% in 2003 but still lagged behind
those of pension funds and mutual funds.
38
For non-life insurance companies,
currency matching is particularly important because of the highly uncertain
timing of claim payments.
39
Mutual funds.
Mutual funds are not only institutional investors but also in-
vestment vehicles for retail investors as well as other institutional investors.
40
Unlike pension funds and insurance companies, mutual funds invest assets they
manage on behalf of their shareholders. That is why mutual fund activities are
subject to heavy regulation in many OECD countries. Regulations usually cover
the following aspects of mutual fund activities: self dealings and affiliated
party transactions, management fees of professional fund managers, capital
structures, investment objectives and policies, protection of physical integrity of
the asset pool, fair valuation of investor purchase and redemption, and disclo-
sure of reliable information to investors.
41
In some OECD countries, restrictions
are also in place on the distribution of mutual fund products based on investor
protection concerns.
42
Although the liability structure plays a key role in the asset allocations of pen-
sion funds and insurance companies, mutual funds have no investment-linked
liabilities. The asset allocations of mutual funds therefore simply reflect the
demands of their shareholders. According to IMF estimates, mutual funds and
other investment vehicles such as closed-end funds and unit investment trusts
35. IMF, Global Financial Stability Report (September 2005).
36. See Blommenstein, “Institutional Investors.”
37. IMF, Global Financial Stability Report (September 2005).
38. Ibid.
39. Blommenstein, “Institutional Investors.”
40. On such dual nature of investment funds, see IMF, Global Financial Stability Report (Sep-
tember 2005).
41. Blommenstein, “Institutional Investors.”
42. Ibid.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
860 ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
collectively held about 47%–48% of their assets in equities and about 41%–
43% in bonds during 1997–2003.43 Foreign equities and bonds represent about
13%–17% of their total securities holdings.44
The mutual fund industry has seen assets under its management grow ex-
plosively over the past decade. Assets are now about three times as large as 10
years ago (see Table 1). When provided with tax incentives, households have
increasingly seen mutual funds as attractive investment vehicles for the pur-
pose of retirement savings. About 24% of the $12.9 trillion placed in U.S. retire-
ment accounts was under the management of mutual funds in 2004.45 Besides
retail investors, pension funds and other institutional investors are also in-
creasingly important customers of mutual funds.46
Hedge funds. Although total assets held by hedge funds have more than
doubled since 2000, they are still small relative to the assets in the hands
of traditional institutional investors. Hedge fund-held assets amounted to only
2% of the combined assets of pension funds, insurance companies, and mutual
funds in 2004 (see Table 1). As such, hedge funds may not be the biggest group
of portfolio investors, but they surely look the most colorful. Unlike most
other investment funds, their managers are paid fees based on fund perfor-
mance; consequently, some successful hedge fund managers seem to command
eye-popping compensation. Hedge funds generally use a wide range of inno-
vative and sophisticated investment techniques, including the use of short sales
and financial derivatives, to make typical mutual funds look like a lackluster
bunch.
Hedge funds also make frequent use of leverage to amplify returns on their
investment. According to one estimate, seven out of 10 hedge funds obtain le-
veraging by buying securities on margin or using collateralized borrowing.47
As the Long-Term Capital Management (LTCM) debacle in 1998 shows, they
sometimes use leverage very aggressively. The leverage ratio of LTCM was
later found to be more than 20. Although such overleveraging appears to be ex-
ceptional rather than typical, it is difficult to get reliable data on the actual lever-
age ratios of individual funds. In fact, the information about most activities of
hedge funds is very limited, contributing to the aura of secrecy surrounding them.
The distinctive features of hedge funds derive mainly from their legal status.48
Hedge funds are established as private partnerships and often located offshore.
43. IMF, Global Financial Stability Report (September 2005).
44. Ibid.
45. Ibid.
46. Ibid.
47. Barry Eichengreen, Donald Mathieson, et al., Hedge Funds and Financial Market Dynam- ics, IMF Occasional Paper, no. 166 (Washington, D.C., 1998).
48. Ibid.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE 861
Because the share ownership of hedge funds is normally distributed among high net worth individuals and institutions through private placements, hedge funds are largely exempt from disclosure and regulation requirements based on investor protection concerns. The loosely regulated hedge funds also face few restrictions posed by their prospectuses on their choice of portfolios and trans- actions. In contrast, other portfolio investors such as pension funds, insurance companies, and mutual funds generally face tight and detailed regulations and prospectuses in this area.
However, hedge funds are by no means monolithic within the industry. They exhibit a multitude of different investment styles. Nonetheless, they are broadly grouped into two main types—macro hedge funds and relative value funds. Macro hedge funds try to identify misaligned macroeconomic or finan- cial variables and take large directional bets, whereas relative value funds take arbitrage positions on the relative prices of securities.49 Macro funds tend to make more aggressive use of leverage. According to one estimate, their lever- age ratios are four to seven on average, while the ratio does not exceed one for almost 85% of all hedge funds.50
One should note that the line between hedge funds and traditional institu- tional investors is becoming increasingly blurred. Traditional institutional in- vestors now frequently undertake many of the same transactions as hedge funds.51 In particular, mutual funds are increasingly pursuing more flexible and often hedge fund-like strategies as competition from hedge funds intensifies. Furthermore, a growing number of traditional institutional investors are turn- ing to hedge funds in their search for asset returns that are not highly corre- lated.52 Although high net worth individuals continue to provide a steady supply of hedge fund capital, institutional investors including pension funds, mutual funds, insurance companies, and university endowments now represent the more important investor base for hedge funds.53
International Diversification of Portfolios Institutional investors diversify their portfolios beyond national borders by in-
vesting in foreign securities. By taking advantage of international diversification,
institutional investors are expected to increase risk-adjusted returns. However, it
is well known that institutional investors as a group are much less internation-
ally diversified than would be true of a global portfolio. The reasons for this
“home bias” include regulatory factors such as regulatory limits on foreign asset
49. Ibid.
50. Ibid.
51. See IMF, Global Financial Stability Report (September 2005) for details.
52. Ibid.
53. Eichengreen, Mathieson, et al., Hedge Funds.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
862 ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
holdings, regulatory requirements for currency matching of assets and liabili-
ties, and less favorable regulatory treatment of foreign assets. In addition,
home bias may stem from the widespread use of domestic benchmarks in as-
sessing fund managers’ performance, and higher transaction and information
costs of international asset trades. Furthermore, exchange rate risk is an obvi-
ous source of home bias.54
The potential gains of diversification for OECD institutional investors would
be greater from diversifying into emerging market securities whose asset returns
are less likely to be correlated with those of domestic assets. However, emerg-
ing market securities may entail additional risks in comparison with their ma-
ture market counterparts, leading to a higher degree of home bias with regard
to emerging market investment. These additional risks include transfer risk,
settlement risk, and liquidity risk.55 Furthermore, exchange rate risk may be
greater for emerging market securities. For example, financial derivatives that
can be used as hedging instruments are not always available for these securi-
ties.56 And even if they are, their prices may be too high or they may cover
only short periods.
Nevertheless, the degree of home bias has declined persistently over the past
15 years as regulatory hurdles of foreign asset holdings have been noticeably
reduced, global benchmarks have been increasingly adopted, and higher risk-
adjusted returns have been actively sought after.57 A useful index for measur-
ing home bias is the foreign asset acceptance ratio (FAAR), which measures
how much the share of foreign assets in a portfolio deviates from the share of
foreign assets that would prevail in a truly global portfolio.58 A FAAR of 100%
corresponds to no home bias. The IMF estimation of FAARs for major OECD
economies—the United States, Japan, the United Kingdom, France, Germany,
and the Netherlands—clearly shows the declining home bias of OECD port-
folios. The aggregate FAAR for equities jumped from 8% to 30% between 1990
and 2003.59 The figure for bonds also increased, but more slowly, with its ag-
gregate FAAR reaching about 20% in 2003.60
Against the backdrop of declining home bias, the portfolio flows to devel-
oping countries also have been on a rising path over the past couple of decades,
54. See IMF, Global Financial Stability Report (September 2005) for a review of these and
other possible sources of home bias.
55. Hans Blommenstein, “Impact of Institutional Investors on Financial Markets,” Institutional Investors in the New Financial Landscape (Paris: OECD, 1998).
56. See Bank for International Settlements (BIS), Recent Innovations in International Banking (the Cross Report) (Basle, 1986).
57. IMF, Global Financial Stability Report (September 2005).
58. Ibid.
59. Ibid.
60. Ibid.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE 863
albeit with swings of ups and downs around the general trend. However, devel-
oping countries still attract a very limited share of portfolio capital. Whereas
gross inflows of portfolio investment to industrial countries (the U.S., Canada,
Japan, U.K., and Eurozone) amount to $1.6 trillion in 2004, the corresponding
figure for developing countries was only $170.1 billion.61 Furthermore, most
of these flows were destined for a small group of developing countries. Such a
limited extent, coupled with institutional investors’ increasing search for un-
correlated asset returns, suggests the tremendous growth potential for portfo-
lio investment flows to developing countries.
4. The Asian Financial Crisis The dual personality of foreign investors couldn’t manifest itself more vividly
than in East Asia during the 1990s. The decade started with net private capital
inflows to developing Asia of less than $20 billion in 1990 but then saw a sharp
rise of these flows, surpassing $110 billion in 1996. About two-thirds of them
were concentrated in the five countries—Korea, Malaysia, Indonesia, Thailand,
and the Philippines—whose impressive economic booms had partly been fu-
eled by these capital flows from abroad. Then all of a sudden came the sharp
turn of the tide in cross-border capital flows in 1997. Throughout the year, net
private capital inflows in the five countries turned negative, hurling the coun-
tries into severe economic crisis (see Table 3). The Asian crisis can be consid-
ered essentially a two-step process comprising the Thai currency crisis and its
contagion. The crisis originated in Thailand when several waves of pressure
on the country’s currency, spurred by growing financial instability, eventually
led to the baht devaluation of July 1997. Then this currency crisis quickly
spread to neighboring countries. Portfolio investors and hedge funds in partic-
ular are frequently painted as speculators that played a major role in both steps
by leading so-called hot money flows. The popular perception suggests that
these investors not only precipitated the baht devaluation but also fueled the
contagion through the huge volume of their own transactions or by simply lead-
ing the herd. How accurate a description is this of what really happened? The
available evidence tends to raise doubts about its validity.
Portfolio Investment Flows
Prelude to crisis. Capital flows figures shown in Table 3 indicate that OECD
institutional investors built up substantial equity and bond holdings in Korea,
Malaysia, Indonesia, Thailand, and the Philippines prior to the crisis. Net portfolio
investment inflows in these countries exceeded $20 billion in 1996, accounting
61. Ibid.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
864 ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
for almost 30% of total net private capital inflows in the countries. However,
investment in emerging financial markets by OECD institutional investors was
still a relatively recent development. It had been almost unheard of until the
mid-1980s, when a small group of closed-end investment funds including
country funds started investing in emerging stock markets. The portfolio flows
to developing countries then grew explosively in the 1990s. As can be seen
from Table 3, net portfolio inflows in the five crisis countries, which made up
only about 5% of their total net private capital inflows at the beginning of the
1990s, increased more than fifteenfold in the run-up to the crisis.
This rising interest in capital markets in those five countries coincided with
some notable developments in major OECD economies, including falling interest
table 3 Private Capital Flows to Emerging Markets (in billions of U.S. dollars)
1990 1991 1992 1993 1994 1995 1996 1997
Emerging markets Total net private capital inflowsa 31.0 126.9 120.9 164.7 160.5 192.0 240.8 173.7
Net FDI 17.6 31.3 37.2 60.6 84.3 96.0 114.9 138.2
Net portfolio investment 17.1 37.3 59.9 103.5 87.8 23.5 49.7 42.9
Otherb �3.7 58.4 23.8 0.7 �11.7 72.5 76.2 �7.3
Net external borrowing from
official creditors 22.2 25.7 17.6 18.7 �2.5 34.9 �9.7 29.0
Total net capital inflows 53.2 152.7 138.5 183.4 158.0 226.9 231.1 202.7
Asia Total net private capital inflows 19.1 35.8 21.7 57.6 66.2 95.8 110.4 13.9
Net FDI 8.9 14.5 16.5 35.9 46.8 49.5 57.0 57.8
Net portfolio investment �1.4 1.8 9.3 21.6 9.5 10.5 13.4 �8.6
Other 11.6 19.5 �4.1 0.1 9.9 35.8 39.9 �35.4
Net external borrowing from
official creditors 5.6 11.0 10.3 8.7 5.9 4.5 8.8 28.6
Crisis countries’ net private
capital inflowsc 24.9 29.0 30.3 32.6 35.1 62.9 72.9 �11.0
Net FDI 6.2 7.2 8.6 8.6 7.4 9.5 12.0 9.6
Net portfolio investment 1.3 3.3 6.3 17.9 10.6 14.4 20.3 11.8
Other 17.4 18.5 15.4 6.1 17.1 39.0 40.6 �32.3
Crisis countries’ net external
borrowing from official
creditors 0.3 4.4 2.0 0.8 0.7 1.0 4.6 25.6
SOURCE: IMF, International Capital Markets: Developments, Prospects, and Key Policy Issues (Washington, D.C., September 1998). a Net FDI plus net portfolio investment plus net other investment. b “Other” flows largely consist of bank lending. c Crisis countries include Korea, Malaysia, Thailand, Indonesia, and the Philippines.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE 865
rates and slowing industrial production. These so-called push factors—factors
in the global economy—stimulated interest in emerging markets among OECD
institutional investors in accord with the overall trend of declining home bias
described in the previous section. In fact, empirical evidence generally points
to the importance of push factors in determining portfolio flows to emerging
market economies.62 In the Asian crisis countries, however, country-specific
“pull” factors appear to have been at least as important, since before the crisis
these nations continued to achieve rapid economic growth, liberalize their trade
and financial markets, and hold fairly balanced fiscal positions. These coun-
tries in fact were viewed by OECD institutional investors as the most appeal-
ing among emerging market economies in the first half of the 1990s.63
The Thai crisis. Starting with Malaysia in 1991, international banks were
heavily involved in what was known as the Asian carry trade. That is, they
tried to arbitrage huge spread differences by obtaining short-term funding in
U.S. dollars or yen and on-lending the proceeds in East Asian currencies in short
terms at much higher interest rates. Such arbitrage transactions remained prof-
itable as long as the values of East Asian currencies did not move against
those of the U.S. dollar or yen. However, the profitability of this carry trade in
Thailand was thrown into doubt when a growing number of investors began to
question the stability of the baht in the wake of rising financial instability and
disappointing macroeconomic performance there. After having weathered the
first wave of pressures on the currency in mid-1996, the Thai authorities en-
countered a second wave of serious pressures in early 1997 as international
banks increasingly closed out their positions in the carry trade.64
The account à la Dr. Mahathir of what followed next suggests that specula-
tors such as macro hedge funds, having smelled blood, jumped on short sales
of the baht on a grand scale, thereby playing a key role in precipitating the
baht devaluation of July 1997. However, the evidence seems to indicate that
the role of hedge funds was nowhere near dominant, given the relatively lim-
ited forward positions they collectively took. At the end of July 1997, transac-
tions taken directly with hedge funds are estimated at only one-quarter of $28
billion in forward contract sales to the Bank of Thailand.65 At the same time,
62. For example, see Punam Chuhan, Stijn Claessens, and Nlandu Mamingi, “Equity and Bond
Flows to Latin America and Asia: The Role of Global and Country Factors,” Journal of Develop- ment Economics 55:2 (April 1998), pp. 439–63.
63. Goldstein, The Asian Financial Crisis. 64. For a more detailed account of the course of events, see IMF, International Capital Markets
(September 1998).
65. Instead of taking positions directly with the central bank, hedge funds may also have sold
the baht forward through third parties that then offset their positions with the central bank; it is impos-
sible to figure out the extent of these transactions. See Eichengreen, Mathieson, et al., Hedge Funds.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
866 ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
hedge funds do not appear to have led the herding behavior, either. They appear
to have taken the bulk of their short positions only in May 1997 after pressure
on the baht had already mounted considerably for some time.66
Once the baht was allowed to float, domestic corporations with unhedged
foreign currency exposure scrambled to cover their exposure.67 These domestic
entities, along with international banks that closed out their existing credit
lines, appear to have been more responsible for the baht’s continued fall after
devaluation than were foreign portfolio investors. In fact, many hedge funds
appear to have quickly switched to long positions on the baht following its ini-
tial float.68 Similarly, in the Mexican crisis of 1994–95, the leading role of do-
mestic residents rather than foreign investors was also evident.69
Contagion. The currency crisis that originated in Thailand quickly spread to
neighboring countries. The affected countries suffered a free fall in the values
of their currencies before they finally bottomed out in January 1998. By that
time, the value of the Indonesian rupiah—relative to its level of July 1, 1997—
had dropped by as much as 81%, the Malaysian ringgit by 46%, and the Phil-
ippine peso by 41%. The corresponding figure for the Thai baht was 56%. The
value of the Korean won fell by 55% from October 1 to its low point in late
December.70
Foreign portfolio investors, especially hedge funds, are prominently fea-
tured in Dr. Mahathir’s narrative of this course of events as well. According to the
argument, portfolio investors were betting vast sums of money against the cur-
rencies of neighboring countries, thereby leading the herd and fueling the conta-
gion. Again, the available evidence points to the contrary on this. Instead of
leading the contagion, most foreign portfolio investors appear to have been
completely overwhelmed by the magnitude of the contagion. Even many hedge
funds, arguably the most astute of the investors, failed to anticipate the sharp
decline of other Asian currencies.
Although hedge funds built up large positions on the Indonesian rupiah,
these were primarily long positions taken shortly after the currency had begun
to depreciate.71 Most hedge funds obviously failed to anticipate a trend of con-
tinued exodus from the currency, initially led by international banks and later
accompanied by domestic institutions trying to hedge their external debts and
66. Ibid.
67. IMF, International Capital Markets. 68. Eichengreen, Mathieson, et al., Hedge Funds.
69. See Jeffrey Frankel and Sergio Schmukler, “Country Fund Discounts and the Mexican Cri-
sis of December 1994: Did Local Residents Turn Pessimistic Before International Investors?”
Open Economies Review 7:Supp.1 (1996), pp. 511–34.
70. IMF, International Capital Markets.
71. Eichengreen, Mathieson, et al., Hedge Funds.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE 867
option positions.72 In other crisis countries, there are very few indications of a
significant buildup in hedge fund positions on their currencies during that time.73
Real Barbarians The review of the Asian crisis thus far provides little evidence that the volume
of transactions carried out by hedge funds was large enough to lead the course
of the crisis. In fact, systematic evidence seems to indicate that hedge funds as
a group took sizable short positions only for the baht among the currencies in
the region at the time.74 Even in Thailand, however, the size of these short po-
sitions was quite limited in comparison with the total volume of forward con-
tract sales of the baht, as described above. There is also little evidence that
hedge funds led herding behavior in Thailand or any other countries affected
by the crisis.
Capital flows figures in Table 3 confirm that not just hedge funds but in fact
foreign portfolio investors as a group were an unlikely candidate to be the
dominant force in the making of a crisis. Although foreign portfolio inflows in
the five Asian crisis countries visibly slowed down in 1997, they still remained
substantially positive throughout the year. In contrast, foreign bank loans to
the five countries swung sharply from net inflows of more than $40 billion in
1996 to net outflows of more than $32 billion in 1997. Apparently, foreign
banks, not portfolio investors, were the real barbarians that led the sharp re-
versal of foreign capital flows in these countries.
The vast majority of bank loans to the crisis countries were of short term and
denominated in foreign currency, mostly U.S. dollars.75 Under normal circum-
stances, these short-term loans would continue to be rolled over at their maturity.
But once trouble loomed, foreign banks refused to roll over their loans to avoid a
default, thereby prompting capital outflows. In particular, Japanese banks, hav-
ing been the largest lender to the region, led the sharp withdrawal of credit from
the crisis countries as they tried to cope with domestic financial turbulence and
meet minimum capital standards.76 Their poor financial health made them more
responsive to any hint of trouble and subsequently quicker to act.
The cover story of Euromoney Magazine in 1998 vividly illustrates that it
was bank lending that drove massive capital outflows during the crisis. Fol-
lowing the sharp decline in the Hong Kong stock market in October 1997,
capital fled Korea rapidly; the country’s international reserves were being
72. Indonesian banks and corporations had sold options against the rupiah’s depreciation
whose premiums served as a lucrative source of income. See IMF, International Capital Markets. 73. Eichengreen, Mathieson, et al., Hedge Funds.
74. Ibid.
75. See Goldstein, The Asian Financial Crisis.
76. IMF, International Capital Markets.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
868 ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
depleted rapidly. Although IMF-led financial support was extended to Korea
and money market interest rates rose to 25% in line with the IMF program,
none of these measures was sufficient to stop the hemorrhaging. The rapid
outflow of capital was arrested only after major international banks on De-
cember 29 agreed to temporarily maintain their lines of credit to Korea.77
5. Conclusions What are the lessons of the Asian crisis for the region’s governments trying to
steer clear of another financial meltdown? They certainly do not include support
for any hysteria surrounding hedge funds. Hedge funds, as it happens, played
only a minor role in the making of the crisis in Asia. There are now, of course,
more avenues available for taking short positions on currencies and securities in
Asia than in 1997, and one should not discount the systemic risks posed by
hedge funds because of their aggressive use of leverage and derivatives, as the
LTCM fiasco in 1998 clearly indicates. Nevertheless, the systematic evidence
demonstrates that the pattern seen in Asia—i.e., hedge funds were not really the
first-mover to lead the herd throughout the initial crisis—is not unusual.78
The proper lesson we should draw from the Asian crisis is that any inves-
tors, not just hedge funds, could turn into barbarians at the hint of a sharp rise
in economy-wide risk. In Asia, international banks were the major barbarians
that spurred havoc. Other barbarians include such unexpected faces as domes-
tic corporations and banks that rushed to cover their unhedged exposures. It is
in fact meaningless to distinguish between these entities, trying to hedge their
exposures by liquidating their long positions, and speculators who are actively
taking short positions at the hint of, say, currency weakness, because both ac-
tions are exerting the same downward pressure on the currency.
In their efforts to minimize risks associated with foreign investors, there-
fore, governments should not waste so much of their energy in identifying a
“bad” class of investors and limiting their activities. They should instead focus
on avoiding inconsistent policies that present opportunities for one-way bet-
ting. Faced with the buildup of considerable financial imbalances that are not
sustainable indefinitely, or the existence of an exchange-rate peg that is not cred-
ibly defensible, market participants will expect the market to eventually move
in one direction, with the probability of it moving in the other quite negligible.
Once such a market expectation firmly sets in, it can prompt the collective actions
77. The rollover of loans falling due at year-end was agreed upon among major U.S., U.K., and
German banks in the meeting on December 29, providing much needed time for a more compre-
hensive debt rescheduling. For a detailed account of the negotiations, see “Korea Stares into the
Abyss,” Euromoney (March 1998), pp. 32–37.
78. According to the econometric evidence provided by Eichengreen, Mathieson, et al., Hedge Funds, there is no indication that other investors regularly follow the positions taken by hedge funds.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE 869
of market participants of all shades, both speculators and hedgers, of both foreign
and domestic origin, with destabilizing effects on the financial sector.
The Asian financial crisis is a classic example of such a process. As can be seen in Table 4, substantial financial imbalances were piling up in the run-up to the crisis even though the seriousness of these imbalances was masked at that time by the crisis countries’ fast economic growth. For example, all coun- tries experienced sharp real appreciation of their currency and bloated current account deficits prior to the crisis. These imbalances made the exchange-rate peg of such currencies as the Thai baht increasingly less credible and eventu- ally took their toll by prompting the collective actions of private capital flows. The extent of reversals in net private capital flows reached as much as 14% of GDP in Thailand over 1996–97.
Collective actions do not necessarily lead to a full blown crisis, but they are more likely to do so if the economy is inside the danger zone where self- fulfilling market actions can arise. One example of such a zone is the assump- tion of large amounts of short-term debt, denominated in foreign currency, by the governments or domestic financial institutions and corporations. Suppose
table 4 Selected Macroeconomic Indicators in Crisis Countries
1994 1995 1996 1997 1998 1999 2000
Thailand Real GDP growth (% per year) 9.0 9.2 5.9 �1.4 �10.5 4.4 4.8
Net private capital flows (as % of GDP) 8.2 11.8 9.2 �5.3 �16.3 �13.5 �10.7
Real exchange rate growth (% per year) 0.1 �1.8 6.8 �7.0 �15.5 5.1 �3.1
Current account (as % of GDP) �5.4 �7.9 �7.9 �2.1 12.8 10.2 7.6
Korea Real GDP growth (% per year) 8.5 9.2 7.0 4.7 �6.9 9.5 8.5
Net private capital flows (as % of GDP) 2.7 0.7 4.1 �4.6 �2.7 3.3 1.8
Real exchange rate (growth, % per year) 0.8 1.2 3.5 �6.0 �25.6 13.5 8.1
Current account (as % of GDP) �1.0 �1.7 �4.1 �1.6 11.7 5.5 2.4
Indonesia Real GDP growth (% per year) 7.5 8.2 7.8 4.7 �13.1 0.8 5.4
Net private capital flows (as % of GDP) 3.9 6.2 6.3 7.1 �3.0 �3.6 �0.9
Real exchange rate growth (% per year) �0.7 �3.4 5.1 �5.6 �51.6 45.2 �2.0
Current account (as % of GDP) �1.5 �3.0 �2.9 �1.6 3.8 3.7 4.8
Malaysia Real GDP growth (% per year) 9.2 9.8 10.0 7.3 �7.4 6.1 8.9
Net private capital flows (as % of GDP) �0.3 �0.4 0.8 0.2 �0.3 �0.3 2.2
Real exchange rate growth (% per year) �2.6 0.6 4.3 �2.4 �20.5 2.9 2.6
Current account (as % of GDP) �7.6 �9.7 �4.4 �5.9 13.2 15.9 9.4
SOURCE: IMF, World Economic Outlook Database; International Financial Statistics Database;
Balance of Payments Statistics Database.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
870 ASIAN SURVEY, VOL. XLVII, NO. 6, NOVEMBER/DECEMBER 2007
a country with substantial short term, foreign-currency-denominated debt runs into a sudden loss of investor confidence in its creditworthiness. Then the country may soon face accelerating demands for repayment of its maturing short-term debt and may be forced to meet them from its international reserves. The rapidly dwindling reserves may signal a crisis, further intensifying capital outflows from the country and exacerbating its economic contraction. This course of events can vindicate the initial pessimism about the country’s creditworthiness, mak- ing a crisis self-fulfilling.
The Asian crisis certainly holds the self-fulfilling element just described.
Table 5 shows several indicators of maturity and currency mismatches as
of June 1997 for the crisis countries. These indicators suggest that the maturity
and currency mismatches were prevalent in the foreign borrowing of banks
and corporations in most countries prior to the crisis. The mismatches of Korea
particularly stand out. The sequencing of capital account liberalization in fact
led to a concentration of short-term foreign debt among financial institutions
there. Short-term credit flows were liberalized first, while substantial restrictions
were retained on such long-term capital inflows as FDI and portfolio investment
in listed stocks.79
In Korea and other crisis countries, the rollover of short-term foreign bank
loans eventually became problematic, resulting in rapid depletion of interna-
tional reserves and destabilization of exchange rates. In contrast, portfolio in-
vestors stayed relatively quiet, suggesting that they had an incentive to wait
for prices to rebound rather than sell into a falling market. Such an incentive is
79. See Yung Chul Park and Chi-Young Song, “Managing Foreign Capital Flows: The Expe-
riences of Korea, Thailand, Malaysia and Indonesia,” The Jerome Levy Economics Institute of
Bard College, Working Paper, no. 163 (Washington, D.C., 1996); and Barry Johnston, Salim Darbar,
and Claudia Echeverria, “Sequencing Capital Account Liberalization: Lessons from the Experi-
ences in Chile, Indonesia, Korea, and Thailand,” IMF Working Paper, no. 97/157 (Washington,
D.C., 1997).
table 5 Maturity and Currency Mismatches, June 1997
Ratio of Short-Term Debt to
International Reserves
Short-Term Debt as a % of Total Debt
Ratio of Broad Money to
International Reserves
Korea 3.0 67 6.2
Indonesia 1.6 24 6.2
Thailand 1.1 46 4.9
Philippines 0.7 19 4.9
Malaysia 0.6 39 4.0
SOURCE: Goldstein, The Asian Financial Crisis.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
JEONG YEON LEE 871
not unusual for a foreign investor in local equities and bonds because the in-
vestor has no way of exiting the declining market under the falling currency
other than incurring a large loss.
Of different types of portfolio investors, hedge funds are even better able to
wait for prices to rebound than traditional institutional investors such as mu-
tual funds, because the shareholders of hedge funds are generally locked in for
a considerable period. Mutual funds, on the other hand, can face immediate
shareholder withdrawals when making losses. Hedge funds in fact can be the
first to take long positions in depressed markets with the intent of acquiring
assets on the rebound. This is because in searching for above-normal returns,
they often need to be contrarian rather than following others. Shutting out hedge
funds is therefore likely to do more harm than good to the stability of financial
markets. Governments should in fact promote diversity among investors rather
than curbing the entry of certain types of investors. The liquidity in financial
markets is, after all, a function of a diversity of opinions.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:20:10 PM All use subject to JSTOR Terms and Conditions
Improving Financial Stability Uncertainty versus Imperfection.pdf
Improving Financial Stability: Uncertainty versus Imperfection Author(s): Éric Tymoigne Source: Journal of Economic Issues, Vol. 41, No. 2, Papers from the 2007 AFEE Meeting (Jun., 2007), pp. 503-510 Published by: Association for Evolutionary Economics Stable URL: http://www.jstor.org/stable/25511203 .
Accessed: 13/02/2015 16:23
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Association for Evolutionary Economics is collaborating with JSTOR to digitize, preserve and extend access to Journal of Economic Issues.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
IPJ JOURNAL OF ECONOMIC ISSUES Jul Vol.XLI No. 2 June 2007
Improving Financial Stability: Uncertainty versus Imperfection
trie Tymoigne
For most contemporary economists, financial instability is an exceptional event that
results from exogenous shocks, market imperfections, and/or price instability. Thus,
in order to have financial stability, we should promote market mechanisms and correct imperfections, and should have a central bank that guarantees price stability.
This position, however, ignores the important contributions of Hyman P. Minsky,
John Kenneth Galbraith, John Maynard Keynes and other authors. They show that there are deep causes of financial instability rooted in psychological, sociological, and
political forces, which effects are multiplied by the way the capitalist economic system works.
This paper critically analyzes the imperfection view of financial stability by focusing on its analysis of individual behaviors and of the treatment of information.
In addition, an alternative way to analyze and to solve problems related to financial
stability, the uncertainty view, is presented. The first part of the paper briefly reviews
the views on financial stability in the imperfection view. The second part presents some of the limitations of the imperfection view and details some of the key points of the uncertainty view. The third part extracts from the uncertainty view some policy recommendations to promote financial stability.
Market Imperfection as the Source of Instability
In a pure and perfect competitive setting, a "well-behaved" economic system is at
equilibrium now and forever. Indeed, economic agents are assumed to make
"informed" decisions based on "fundamentals" provided by the economic, political, and financial outlooks. The latter are strong a priori attractors that nobody can ignore
without experiencing harsh consequences in a more or less immediate future (and the
The author is an Assistant Professor in the Department of Economics at California State University, Fresno. This
paper was presented at the Annual AFEE conference in Chicago, January 2007.
503 ?2007, Journal of Economic Issues
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
504 Eric Tymoigne
more efficient a market, the sooner harsh consequences materialize). Thus, in a pure
and perfect competitive setting, "both market participants and supervisors will see
hints of errant investment strategies of financial problems as they begin" (Shinasi 2006, 179) and so "market discipline" will do its job to eliminate any unsustainable
decisions, that is decisions not in accordance with the fundamentals.
Behind this reasoning lie specific assumptions about the nature of human
rationality and the nature of the information provided to economic agents. First,
following methodological individualism, economic agents are assumed to be robot-like
beings that respond to price signals according to their own preference, without any
considerations for their social environment. It is possible to partly deviate from this
assumption by having economic agents that follow Bayesian probabilities (as the "cascade of information" literature does), or by taking into account some of the
"cognitive biases" or "anomalies" observed (as behavioral finance does). However,
over time the bounded rationality of individuals with cognitive biases is corrected by additional information. Second, the information received is a neutral input directly usable to feed the decision process in order to generate an output (buy, sell, or stay
neutral). Third, changes in prices are assumed to give a direct understanding of the
underlying motivations of economic agents.
In this context, it is very difficult to explain financial instability as a process and to understand why there are recurring periods of growing financial fragility followed by financial instability. Financial instability is an exogenous state created by the bounded rationality of economic agents and the imperfection of information.
Indeed, without the capacity to input information and/or "without sufficient timely
information, the market disciplining mechanisms [. . . ] might not produce the
appropriate self-corrective adjustments" (Schinasi 2006, 167), which will lead economic agents to undertake unsustainable decisions and thus prevent the efficient
allocation of real resources from savers to investors.
In terms of policy, this implies that financial stability can be improved by allowing market mechanisms to work more fully. This can be done by improving the
quality and quantity of information, improving the computing capacity of individuals,
and educating economic agents so that their cognitive biases are eliminated (Shiller
2000; Schinasi 2006; Warneryd 2001). Public authorities can help to ensure financial
stability, but not by intervening directly in the pricing mechanism. Indeed, they are
viewed as incompetent to evaluate the fundamentals:
Central bankers have no particular expertise in valuing future
corporation earnings, that is, in pricing equities, which is a full-time
job carried on by armies of stock analysts and investors. On this
basis, central bankers should feel no obligations to make public their personal views on equity prices. (Goodfriend 1998, 18)
The best a central bank can do is "to take an active role in investing in the
development of the base of professionals -
actuaries, financial analysts, appraisers,
and solvency experts - to perform these sophisticated functions" (Hunter, Kaufman,
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
Improving Financial Stability 505
and Pomerleano 2003, xxiii). In addition, as the Basel II accords suggest, market
participants should do most of the regulation. They should design the tools and methods of regulation and supervision, and the "public sector role could be limited to
assessing and monitoring the quality of risk management and control systems more
systematically and thoroughly, and to defining how information is used, plus ensuring that counterparty disclosure is adequate" (Schinasi 2006, 225). This last quote, however, loosens the position of the imperfection view of financial stability. Indeed, it
recognizes that public institutions have a role in the treatment of information.
Information is no longer a neutral input but must be interpreted.
Criticisms and Implications
Several types of criticisms can be made about the previous way of conceptualizing and
ensuring financial stability (Tymoigne 2006a). In this paper, we focus on criticisms related to information, fundamentals and the individual decision-making process in
an uncertain environment.
The first criticism concerns the relation between information and decision.
Psychologists have observed that:
The relationship between availability of information and confidence is surprising. Intuitively, it is assumed that more
information leads to better decisions, and more confidence in
judgments and decisions. Only the second part is confirmed by research. More information inspires more confidence, but the
quality of decisions tends to increase only up to a point and then
deteriorates. (Warneryd 2001, 168)
Thus, more information does not mean more "rational" choices, even though confidence increases. The cognitive "biases" will not be corrected by providing more
information. In addition, research shows that sometimes, it is hard to interpret the
motivation behind a decision by looking at the financial position taken by economic
agents. People may not remember why they acted the way they did (Shiller 2000) and, at other times, they do not have any- specific motivations:
Having worked with investments for 10 years, I can tell you a lot of
people who trade bonds don't have anything in mind; they are just throwing pieces of paper around. (Seger 1987)
This leads to the second problem: information is not a neutral data. Any information needs to be interpreted and the latter is more important than the information itself. This has two implications. The first implication is that individuals tend to use heuristics to make decisions. These heuristics do not reflect irrationality;
they reflect the way decisions are made in an uncertain world (Kahneman and Tversky
1973; Tversky and Kahneman 1974; 1983; Harvey 1998). Information is interpreted
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
506 E^c Tymoigne
through those heuristics and a different presentation of the information leads to a
different interpretation. The second implication is that information can be twisted to reflect the wishes and interests of economic agents, either by ignoring or discounting bad news, or by transforming the latter into good news. As Galbraith notes:
Above all, it is evident that the capacity of the financial community for ignoring evidence of accumulating trouble, even of wishing devoutly that it might go unmentioned, is as great as ever.
(Galbraith 1961, xxi)
This was illustrated recently with the Long-Term Capital Management (LTCM) case:
Although it is easy in retrospect to question why LTCM's
counterparties did not demand more information, in a competitive environment, cost considerations must have weighed heavily.
Clearly, LTCM's counterparties thought the cost of more
information was too high, and walking away from deals was not in
their interests. (Schinasi 2006, 221)
A central reason for this behavior is that, because of the competitive nature of
capitalist economies, market participants are more sensitive to profitability than to
risk. Thus, even if they are concerned with the latter, it is always cast into strategies
oriented toward the former, which has two important consequences. First, the
riskiness of a situation has a tendency to be judged a posteriori. Thus, if an agent, or a
group of economic agents, takes large risks but is successful, nobody questions the
risks taken, and, on the contrary, others follow. Thus, if LTCM had been successful,
others would have praised its "innovative," "bold" and "highly sophisticated" fund
management strategies, and more financial market participants would have followed
them despite the large leveraging those strategies implied (LTCM leveraged 98% of its
capital). The second consequence is that profitability considerations reinforce the
psychological tendencies of human beings to ignore outlier events and to focus on the
recent past in order to judge the riskiness of an economic decision.
A third problem of the imperfection approach is that "fundamentals" are
not a priori variables that represent invariable long-term anchors toward which the
economic system must tend. There are two central reasons for this. The first one is
related to what we stated previously:
The fundamentals are a social construct. They evolve through the interaction of convention, information-gathering, market activity,
investor psychology, and economic events. (Goldstein 1995, 723)
The mass escape into make-believe, so much a part of the true
speculative orgy, started in earnest. It was still necessary to reassure
those who required some tie, however tenuous, to reality. And, [. . .]
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
Improving Financial Stability 507
the process of reassurance - of inventing the industrial equivalents
of the Florida climate - eventually achieved the status of a
profession. (Galbraith 1961, 16-17)
Fundamentals are conventions used to analyze the economic outlook. These
conventions are social norms backed by a specific vision of the future that define what
is a normal, i.e. sane, economic decision. They usually provide an a posteriori
justification of what is going on in the economy. For example, the main pattern found
in each financial bubble is that a "new era" explanation emerges "as reporters
scramble to justify stock market price moves" (Shiller 2000, 99). The fundamental value adjusts to inflated market prices. The second reason fundamentals are not
invariable anchors is that in an uncertain world, economic reality is changed and
created by the implementation of decisions (Davidson 2002, 64; Bernstein 1993
(1993), 76). In the end, therefore, we reach a different view of the decision-making
process of economic agents. In an uncertain world, economic agents follow a social
rationality and in a capitalist economic system, this social base of justification is
combined with an individual search for accumulation. This combination of social
rationality and individual profitability has large theoretical and policy implications.
The Problem of Financial Stability
In terms of theory, the normal leveraging ratio, the correctness of a certain price level
and other anchors used to make a decision, are heavily influenced by the psych ological and social setting of economic agents. We, thus, need to study more
thoroughly the role and formation of those conventions, and to integrate the latter in
formal models. This does not mean that agents are irrational or have cognitive biases; it just means that, given the uncertainty of the world, decision-making based on
conventions is the best method use to make decisions. Our psychological set up
pushes human beings to rely extensively on the present situation in order to
extrapolate the future, and to heavily discount the past. Our social set up pushes human beings to look for others to confirm the adequacy of our decisions.
Social influence is not only the strongest when the individual feels uncertain and finds no direct applicable earlier experience of her/ his own, but such situations may lead to an active search for some
kind of social support and confirmation. (Warneryd 2001, 203)
Minsky has given a coherent explanation of how these social and psychological characteristics combined with economic and political forces leads to the endogenous generation of financial instability (Tymoigne 2006b). There is no need for the
superficial lack or inaccuracy of information, irrationality, "noise" or other ad hoc
explanations. Information problems and irrationality may be a source of problems, but even if economic agents had all the information they needed and could compute it, this would not eliminate financial instability.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
508 Eric Tymoigne
In terms of policy, the implications are also important because a social
justification of individual accumulation is a recipe for the socialization of losses and the privatization of gains. Public authorities have a larger role than in the
imperfection view to improve the privatization of losses and the socialization of gains.
For example, the goal of central banks should be to guarantee a smooth functioning of the financial system while at the same time limiting the development of speculative and Ponzi financial positions. This would set the foundation to reach and maintain
full employment. To do that, the central bank should participate in the formation of the
conventions used by private economic agents to justify their decisions. Public authorities have a central role in the evaluation of the future that goes beyond
educating economic agents who rationally ignore fundamentals or created new ones, to
justify their choices. Indeed, by being removed from profitability considerations, and
by having a more macroeconomic view based on long-term experience, public
authorities are in a unique key position. Thus, public authorities provide an
alternative view that can be used to influence and manage the social base of decisions
of private economic agents.
More precisely, public authorities can influence conventions by using moral
suasion, by giving their opinion about the appropriateness of the financial position of some economic agents, by encouraging the creation of financial instruments that
promote smoother financial constraints on borrowers, and by being highly involved in
the restructuration of troubled financial institutions that received public assistance.
At the core of this policy should be a different understanding of the
"unsustainability" of economic decisions from the imperfection view. Instead of the
non-accordance of decisions with a priori fundamentals, unsustainable economic
decisions manifest themselves through a continuous weakening of the financial
positions of economic agents. Following Minsky, the fragility of financial positions can be checked by looking at the timing and specificity of three positions: the liability and asset positions ("balance-sheet" position);1 the cash-flow positions induced by the
previous positions; and the position-making activities.
This method of checking the sustainability, and so stability, of the financial
system has several important implications. The first one is that what matters is not the
probability of occurrence of an event. This ultimately is unknown and puts the focus on a number and away from the process that leads to this probability. Stated another
way, what is crucial is not the actual cause of the crisis but the trend of the financial
positions that lead to the possibility of a crisis. A detailed analysis of the trend will
give an understanding of the potential sources of the problem. Then, public
authorities could take proactive measures to limit or modify the use of certain
financial instruments; instead of acting as lender of last resort ex post and implicitly
validating whatever unsustainable practices the private sector developed.
This view of supervision and regulation does not say that economic agents
should not take risks. They should and may use probabilities in order to optimize returns for each risk level. However, public authorities should be concerned with
maintaining global financial instability, without profitability as a primary or even
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
Improving Financial Stability 509
secondary concern. What they should influence is the social base of decisions used in
private decisions. This would lead the supervisory authorities to develop alternative methods to analyze what is going in the economy. Their aim should be to minimize
surprise by looking at the previous elements and the implied interrelation they generate. In order to implement this strategy, existing data should be improved and
new data sets should be created (Tymoigne 2006a). As an illustration, the uncertainty view states that to comment about stock
market prices by claiming there is a bubble is an inappropriate way to deal with a
possible problem. What public authorities should focus on is the impact stock prices had on the financial positions of economic agents. The public authorities could then
argue in a more systematic way against the trend in the stock market by looking at the
implications of a change in stock market prices on financial positions. By claiming that there is a bubble, the public authorities will face large protests from financial
market participants and will be urged to justify systematically something that cannot be justified systematically: who knows what the future will be and what the
appropriate fundamentals are. In addition, the public authorities will point toward
the wrong problem by claiming that there is a bubble because even if there is actually no bubble, problems can still occur: what matters is the sensitivity of financial
positions to the change in share prices.
Note
1. Liabilities are defined in a broad sense: off-balance sheet obligations are included.
References
Bernstein, Peter. L. "Is Investing for the Long Term Theory or Just Mumbo-Jumbo?" Journal of Post
Keynesian Economics 15, 3 (1993): 387-393. Reprinted in Can the Free Market Pick Winners edited
by Paul Davidson, 75-81. Armonk: M. E. Sharpe, 1993.
Davidson, Paul. Financial Markets, Money, and the Real World. Northampton: Edward Elgar, 2002.
Galbraith, John K. The Great Crash. 3rd edition. Boston, Houghton Mifflin, 1961.
Goldstein, Don. "Uncertainty, Competition, and Speculative Finance in the Eighties." Journal of Economic Issues 29, 3 (1995): 719-746.
Goodfriend, Marvin. In Asset Prices and Monetary Policy: Four Views, edited by the Centre for Economic
Policy Research, 10-19. London: Centre for Economic Policy Research. 1998.
Harvey, John. "Heuristic Judgment Theory." Journal of Economic Issues 32, 1 (1998): 47-64.
Hunter, William C, George G. Kaufman, and Michael Pomerleano (eds.). Asset Price Bubbles. Cambridge: MIT Press. 2003.
Kahneman, Daniel, and Amos Tversky. "On the Psychology of Prediction." Psychological Review 80, 4 (1973): 237-251.
Schinasi, Gary J. Safeguarding Financial Stability: Theory and Practice. Washington, D.C.: International
Monetary Funds, 2006.
Seger, Martha R. Transcripts of the FOMC Meetings. Washington D.C.: Federal Reserve System, May 1987. Shiller, Robert J. Irrational Exuberance. Princeton: Princeton University Press, 2000.
Tversky, Amos and Daniel Kahneman. "Judgment Under Uncertainty: Heuristic and Bias." Science 185
(September 27, 1974): 1124-1131. -. "Extensional versus Intuitive Reasoning: The Conjunction Fallacy in Probability Judgment."
Psychological Review 90, 4 (1983): 293-315.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
510 Eric Tymoigne
Tymoigne, Eric. Central Banking Asset Prices and Financial Fragility. Ph.D. dissertation, University of Missouri
Kansas City, 2006a. -. "The Minskian System, Part II." Levy Economics Institute of Bard College, working paper no. 452,
2006b.
Warneryd, Karl-Erik. Stock-Market Psychology. Northampton: Edward Elgar, 2001.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:23:01 PM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 503
- p. 504
- p. 505
- p. 506
- p. 507
- p. 508
- p. 509
- p. 510
- Issue Table of Contents
- Journal of Economic Issues, Vol. 41, No. 2, Papers from the 2007 AFEE Meeting (Jun., 2007), pp. 311-624
- Front Matter
- The Veblen-Commons Award
- The 2007 Veblen-Commons Award Recipient: Richard R. Nelson [pp. 311-311]
- Institutions and Economic Growth: Sharpening the Research Agenda: Remarks upon Receipt of the Veblen-Commons Award [pp. 313-323]
- Presidential Address: The Revival of Veblenian Institutional Economics [pp. 324-340]
- Clarence Ayers Memorial Lecture
- Clarence Ayres Memorial Lecture (2007): Evolutionary Institutional Economics [pp. 341-350]
- The Evolutionary Policy Maker [pp. 351-358]
- Using the Social Fabric Matrix to Analyze Institutional Rules Relative to Adequacy in Education Funding [pp. 359-367]
- Institutionalist Perspectives on Immigration Policy [pp. 369-374]
- The Employment Relationship and the Social Costs of Labor [pp. 375-382]
- Tied to the Past - Bound to the Future: Ceremonial Encapsulation in a Maine Woods Land Use Policy [pp. 383-390]
- The French Competitiveness Clusters: Toward a New Public Policy for Innovation and Research? [pp. 391-398]
- China's Technological Emergence and the Loss of Skilled Jobs in the United States: Missing Link Found? [pp. 399-408]
- Reciprocal Transactions, Social Capital, and the Transformation of Bedouin Agriculture [pp. 409-416]
- Sweden's Economic Recovery and the Theory of Comparative Institutional Advantage [pp. 417-426]
- Globalization and the Integration-Assisted Transition in Central and Eastern European Economies [pp. 427-434]
- Institutional Change in Post-Socialist Regimes: Public Policy and Beyond [pp. 435-442]
- The Hospital Industry: The Consequences of the Reforms in Eastern Germany [pp. 443-450]
- Impact of Ideology on Institutional Solutions Addressing Women's Role in the Labor Market in Poland [pp. 451-458]
- The Global Spread of AIDS and HIV [pp. 459-468]
- The Transition from Planning to Markets in National Health Policy for Acute Care Hospitals: The Pittsburgh Experience [pp. 469-476]
- Economic Institutions under Disaster Situations: The Case of Hurricane Katrina [pp. 477-483]
- Can the Fed Target Inflation? Toward an Institutionalist Approach [pp. 485-494]
- Institutional Adjustment Planning for Full Employment [pp. 495-502]
- Improving Financial Stability: Uncertainty versus Imperfection [pp. 503-510]
- Should the Oracle Have a Moral Compass? Social Justice and Recent Federal Reserve Policy [pp. 511-517]
- Non-Trade Concerns in Agricultural and Environmental Economics: How J. R. Commons and Karl Polanyi Can Help Us [pp. 519-527]
- European Contributions to Evolutionary Institutional Economics: The Cases of 'Cumulative Circular Causation' (CCC) and 'Open Systems Approach' (OSA). Some Methodological and Policy Implications [pp. 529-537]
- Toward a Political Institutionalist Economics: Kapp's Social Costs, Lowe's Instrumental Analysis, and the European Institutionalist Approach to Environmental Policy [pp. 539-546]
- Minimizing Missed Opportunities: A New Model of Choice? [pp. 547-556]
- Revisiting Institutionalist Law and Economics: The Inadequacy of the Chicago School: The Case of Personal Bankruptcy Law [pp. 557-565]
- Credit Card Use and Abuse: A Veblenian Analysis [pp. 567-574]
- Deficits and Institutional Theorizing about Households and the State [pp. 575-582]
- Need or Want: What Explains the Run-Up in Consumer Debt? [pp. 583-591]
- Capital, Power and Knowledge According to Thorstein Veblen: Reinterpreting the Knowledge-Based Economy [pp. 593-600]
- The Consequences of Peace: Veblen on Proper Policy to Support Capitalist Economic Relations [pp. 601-608]
- Constitutional Economics and Its Policy Agenda: A Veblen-Inspired Critique [pp. 609-615]
- Veblen's "Theory of Business Enterprise" and Keynes's Monetary Theory of Production [pp. 617-624]
- Back Matter
International Standards and Codes and Financial Stability.pdf
International Standards and Codes and Financial Stability Author(s): A. Vasudevan Source: Economic and Political Weekly, Vol. 36, No. 20 (May 19-25, 2001), pp. 1733-1737 Published by: Economic and Political Weekly Stable URL: http://www.jstor.org/stable/4410636 .
Accessed: 06/02/2015 10:53
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to Economic and Political Weekly.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:05 AM All use subject to JSTOR Terms and Conditions
International Standards and Codes
and Financial Stability Efforts at implementing standards and codes without concomitant adoption of
sound macroeconomic and structural policies are unlikely to ensure growth with financial stability. Emerging market economies undertaking financial reforms have to adopt not one or
two but a requisite set of standards and codes in order to promote the twin objectives of growth and financial stability. The process of transmission discussed here shows that
information and expectations play a major role in market calculus in undertaking investment decisions and resource allocative functions.
A VASUDEVAN
he Mexican crisis of end-1994 and the currency and financial crises in some parts of south-east Asia,
Brazil and Russia between 1997 and 1999 have heightened unusually high interest in the adoption of international stan- dards and codes by all 'systemically important' countries.1 The international community has recognised that the grow- ing internationalisation of financial markets and the application of computer networking systems have unbounded the dimension and intensity of the crises. It is also being increasingly contended that at least the 'core' or key international standards and codes should be imple- mented in order to help contain the finan- cial stresses and reduce financial system vulnerabilities. It is generally accepted that financial stability should be a goal of public policy, since it is regarded as essential for achieving sustained rapid growth.
Against these evolving developments, it is necessary to have an appropriate understanding of the linkage between the implementation of international standards and codes and financial stability. In order to appreciate this linkage, it would be useful to have at the outset a brief idea of what financial stability means and what are the 'core' standards. Thereafter, we shall discuss the processes and mecha- nisms through which the implementation of standards and codes is said to foster financial stability. In the absence of ad- equate data base and historical experi- ences, we will not undertake a quantitative exercise to establish the linkage between the two.
Financial Stability: Meaning and Content
The expression, 'financial stability' cannot be easily defined: it can only be described. This is because it cannot be measured in some definite units and made time-invariant. Nor can there be one strat- egy to realise it, even where there is an elaborate explanation of what financial stability means. This is because financial innovations emerge without giving any notice, and financial market behaviour depends not only on local business prac- tices, and laws but also on the very struc- ture of the markets, and the risk-return perceptions of transactions. Besides, there is the difficulty in defining financial sta- bility arising from lack of clear under- standing about how much of volatility in asset prices and financial flows/transac- tions should be considered as destabilising.
Given the complexity of the factors influencing financial stability, the phe- nomenon is sought to be explained with reference to (a) financial institutions and (b) financial markets.2 Financial institu- tions are regarded as stable so long as there is public confidence that institutions will meet contractual obligations without in- terruption. The critical issue is: how to ensure that public confidence in the insti- tutions is maintained. It would depend largely on the performance of the institu- tions themselves. It could also be argued that apart from good financial results, transparency about the, institutions' poli- cies and practices and dissemination of requisite information to the wider public would greatly contribute to building up of
confidence in the institutions. Among the financial institutions, the ones which are most critical from the view point of build- ing up public confidence in the very pro- cesses of financial intermediation are commercial banks with which most house- holds, firms and public sector bodies including governments would interface. Besides, it is important to closely monitor banks' performance and behaviour because banks often operate payments systems, carry an array of exposures, face liquidity shortages in the short run, and are con- nected with hosts of institutions both domestically and internationally. It is for this reason all banks irrespective of their 'size' need to be protected from 'spill over' effects or 'contagion'. Banks carry liabilities that are repayable on demand and assets that could to a large extent be illiquid and unserviceable in the short run. The low quality of assets created either because of 'adverse selection' based on asymmetric information or 'herd' behaviour, could quickly precipitate li- quidity crisis and lead to lack of public confidence.3 In such an event, the prob- ability of occurrence of runs on banks could be high and could, in the absence of 'safety nets' such as liquidity facilities or lender-of-last-resort mechanisms, and resolution mechanisms such as direct budget allocations, turn out to be well-nigh possible. The impact of such an occur- rence will not be limited only to banks: it will be systemic in the sense that it will spread to other financial institutions as well via the payments failures.
With regard to financial market stability, it is generally taken to be assured if market
Economic and Political Weekly May 19, 2001 1733
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:05 AM All use subject to JSTOR Terms and Conditions
participants could transact with confidence at prices that reflect true economic funda- mentals. The confidence with which trans- actions could be undertaken would depend upon the efficiency and robustness of the market infrastructure. A sound payment and settlement system covering all the fi- nancial markets - money, capital, debt and
foreign exchange markets - with risk
management practices and policies em- bedded in it could help build up confi- dence about accomplishing the finality of transactions in relatively quick time, if not real time in all cases. This, combined with information dissemination would promote market efficiency and help secure finan- cial market stability.
While the payment system could be rendered efficient through synergic actions on the technological, legal and banking fronts, it is important to note that the prices at which transactions are effected may not necessarily reflect the 'fundamentals'. This could be for several reasons. First, the competitive forces may be weak, and may be influenced by one or a few participants. Collusion of interests may therefore not be ruled out. Secondly, information about prices and quantities available may not be accessible to all consumers as well as
producers. There may not be adequate legal enforcement about disclosures and dis- semination of information. Thirdly, regu- lations on financial markets may be weak, and where they are sufficiently strong, their enforceability may be limited by weak supervision and monitoring. Finally and perhaps more fundamentally, prices may not reflect the contemporary forces of demand and supply. Historical experience may well dictate participants to form an
expectation of where prices would be
positioned in the near future. As future
prices themselves tend to fluctuate, the
expectation itself would be in the form of a range of likely prices. There are, how- ever, no normative constructs to suggest that price expectations would be formed either by past experience or by computa- tion of all the information by market
participants in the same way as some of the applications of econometric methods would suggest.4 The processes by which
expectations of asset prices are formed could vary. However, it is generally be- lieved that asset prices often get set in the first instance at levels that are adjusted by a few market makers. Information on such asset prices is regularly disseminated by computers to market agents (e g, jobbers
in stock exchanges) who through inter- actions with other participants arrive at
prices that could be regarded as generally 'acceptable'. To elaborate on this point, institutions, be they stock exchanges or public debt offices, provide information that could be utilised by market agents to
develop expectations of other participants and influence their future behaviour. In this process, agents could adopt different strategies including the one for gaining maximum pay-off under a game-theoretic approach.5 The strategy that would be eventually adopted would depend upon the importance that markets attach to uncertainty and surprise.6 In general, prices can hardly be stable and expected prices of those assets which are deemed to be critical may not be in a range that could form a possible norm.
The interactions between financial insti- tutions and financial markets influence the decisions of agents and other actors in their pursuit of consumption or investment. As this is a continuous process, the 'informa- tion' base itself would get revised every now and then. It would be therefore dif- ficult to distinguish the influence of 'in- formation' from that of institutional struc- tures and actions on the behaviour of markets. In fact, both the institutional actions and information feed upon each other and interact, with the result the markets would have to constantly seek a range of possible outcomes that would appear as if they are spontaneously formed7 so as to be regarded as what may be called the 'good market practices'.
Identification of 'Core' Standards
The idea of having international stan- dards and codes was first thought of after the Mexican crisis of 1994 when the In- ternational Monetary Fund (IMF) and most of the market participants were found to be wanting in foreseeing it even in late 1994.8 The IMF on its part felt that the supply of quality information on right macro indicators was inadequate, resulting in its inaccurate analysis. While this alone is not the reason for the inability to foresee crises, the need for countries to provide correct data on a wide range of variables to IMF for dissemination to private investors has been accepted, leading to the first of the standards relating to data dissemination being evolved. There are two variants of this standard. The Special Data Dissemination Standards (SDDS) are one of them. They
are supposed to be more stringent than the General Data Dissemination Standards (GDDS), and are relevant for all those economies which have access to international capital markets. The IMF 'issues' these standards and encourages all its member countries to subscribe to either one of them.
The Thai crisis of mid-1997 and the subsequent Korean and Indonesian crises pointed to the need for extension of the idea of application of standards. The crisis situations showed that financial stability needs to be pursued through adoption of sound macroeconomic policies, of sound institutional practices and policies by fi- nancial institutions, especially banks, as also corporates, and other critical markets dealing with securities and insurance. Since balance sheets of most of the financial institutions depend a great deal on how they are prepared, the need for adopting standards in the area of accounting and auditing was felt, notwithstanding the sterile controversy on which of the two - the US GAAP and the European account- ing practices should be regarded as stan- dards. As the efficiency of the financial system could be under stress in the absence of a sound financial market infrastructure, and as cross-border transactions require a clear understanding of the processes that underlie the finality of settlements, con- sideration of standards/codes in respect of payment and settlement systems became critical. As market infrastructure and prac- tices and financial institutional soundness often depend upon 'contracts' and the power of enforceability, robustness of law and its implementation would be vital for implementation of standards/codes.
Standards and codes are required to (a) address the need for having symmetrical information; (b) adopt good practices of the market;9 and (c) meet the requirement of institutional and organisational sound- ness. Not all the standards and codes, however, are measurable or quantifiable, and several of them are mere statements of principles.
The development of literature on 'lead- ing indicators'l0 and on comparative fi- nancial systems has given further impetus to the need for instituting standards that are relevant to the countries' specific economic circumstances. There are 43 standards which have been identified11 but it is important to note that most econo- mies even in the industrialised world may not be able to adopt with all the 43 stan- dards at the same time.
1734Economic and Political Weekly May 19, 2001
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:05 AM All use subject to JSTOR Terms and Conditions
Standard setting can be a market-driven process or a process driven by a committee or an official agency.12 In the case of the former, the standard would become de facto and in respect of the latter, it would be de jure. If market practices in fact form the basis for standards to be evolved, market participants would be implementing them. But this does not seem to be the case.13 To a large extent, the standards are set by official agencies by combining observed practices in some markets with intuitive reasoning for realising financial stability. As a result, there are few standards that are products of pure market driven pro- cesses. Where they are driven largely by committees or agencies, their acceptability and ownership could be a severe problem. Again, as not all of them would be rel- evant, national authorities will have to identify the 'core' relevant ones for their economies. Most countries would make such identification on judgments about the institutional and legal capacities to moni- tor, supervise and enforce standards with speed at relatively affordable cost.
The Financial Stability Forum's Task Force on Implementation of Standards (chairman: Andrew Sheng)14 favoured twelve standards as 'core' or 'key' stan- dards deserving of priority implementa- tion by both industrialised and emerging market economies. The twelve standards identified by the Sheng Task Force relate to three subject areas : macroeconomic policy and data transparency; institutional and market infrastructure and financial regulation and supervision. The standards are given in Exhibit 1.
The Sheng Task Force cautioned that identification and implementation of stan- dards as key ones would depend on coun- try circumstances. This qualification was made essentially because standards could be many and heterogeneous, requiring of different specialised skills. Besides, coun- tries would be at different stages of eco- nomic reforms. In emerging market econo- mies, some of the financial markets are not yet well developed. In certain cases, laws are yet to be evolved to ensure that stan- dards are enforceable. Some economies have different degrees of capital account liberalisation and technological skills.
Both industrialised countries and emerg- ing market economies (including poten- tially important market economies) could, in the first instance, consider the twelve 'core' standards identified by the Sheng Task Force, as requiring of implementation within a medium term period. They may,
if they so desire, bring in certain adapta- tions to the standards to suit the country's institutional environment.15 However, where cross-border transactions or consid- erations exist, the international standards will have to be recognised and adopted. It is only in respect of domestic transac- tions, the international standards could be imbued with the domestic environment and circumstances. Forinstance, the do- mestic standards in respect of accounting or auditing or corporate governance or insolvency laws or payment settlement systems could differ from the international standards without being inconsistent.
The twelve standards recommended by the Sheng Task Force seem to receive a good degree of international endorsement and support, as they are broadly regarded as representing 'minimum requirements for good practice'. Some of the standards have relevance for more than one area. For example, the 'Code of Good Practices on Transparency in Monetary and Financial Policies' issued by the.IMF, which is a part of the twelve standards recommended by the Sheng Task Force, has relevance for more than one key standard namely, pay- ment and settlement as well as supervision of (a) banking; (b) securities; and (c) insurance.
The Linkage
The linkage between implementation of the international standards and codes and financial stability is generally taken for granted with the former leading to the latter. It is not however, empirically estab- lished as yet, in view of (a) the almost negligible observations of compliance of standards; and (b) the imprecise nature of measurement of financial stability. These limitations are serious and the linkage therefore may have to be set out from an analytical angle. But this does not mean that once an appropriate measure of finan-
cial stability is obtained, countries would be able to realise financial stability, by adopting international standards and codes.
Let us first examine the operational content of financial stability from the viewpoint of policy-makers. Most of the major macroeconomic indicators and microprudential indicators including risk evaluation exercises figure in thejudgments that policy-makers tend to make about financial stability. The microprudential indicators could be detailed institution- wise but they could also be aggregated (a) for each segment of the sector, such as banking or insurance, or (b) for a group of institutions selected on the basis of a criterion (such as the size or type of ownership of banks). Many countries periodically collect and put out aggregated microprudential indicators but it will be a great mistake to think that macroeco- nomic and aggregated microprudential indicators are two distinct and mutually exclusive categories. In fact, they would be mutually reinforcing but in the absence of knowledge about the strength of this interaction, and as the data on prudential indicators can be rendered available with lags, it would be analytically convenient to identify the major macroeconomic in- dicators as the main source for making judgments of financial stability, with some critical prudential indicators as supple- mental information set to enhance the value of such judgments.
The market participants on their part track macroeconomic and to the extent rendered accessible aggregated micro- prudential indicators for assessing pros- pects of financial stability before taking any investment decisions. They may, where relevant, monitor the published informa- tion on the financial position of individual financial entities and the markets' percep- tions of their performance and prospects.
Assuming for purposes of analysis that the list of indicators tracked by both of-
Exhibit 1
Characteristics Core Standard
I Data and Policy Transparency 1 Code of Good Practices on Transparency in Monetary and Financial Policies.
2 Code of Good Practices in Fiscal Transparency. 3 SDDS/GDDS
II Market Practices 4 International Accounting Standards 5 International Standards on Auditing
III Market Integrity and 6 Core Principles for Systemically Important Payment Systems. Regulatory Infrastructure 7 Core Principles for Effective Banking Supervision.
8 Objectives and Principles of Securities Regulation. 9 Insurance Supervisory Principles.
IV Economic Justice 10 Principles and Guidelines on Effective Insolvency Systems. 11 Principles of Corporate Governance. 12 The Forty Recommendations of the Financial Action Task Force
on Money Laundering.
Economic and Political Weekly May 19, 2001 1735
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:05 AM All use subject to JSTOR Terms and Conditions
ficial and private sector is the same, it would still be necessary to show as to how a movement in an indicator read along with movements in the remaining indicators could be attributable to compliance of a standard or code. In other words, the question remains as to whether compliance to a code or a set of codes would lead to financial stability as reflected in some well defined and specified movements of the identified indicators and if so, how. Again, the judgments on financial stability would need to be parameterised in terms of certain values for each of the identified indicators in the first place. However, does one arrive at those values is the main issue.
For this purpose, it would be necessary to have a well manageable macrocconomic model that is internally consistent. From this model would be derived the first set of optimal values for each identified in- dicators. The internal consistency exercise will take care of problems of reconciliation of and trade-offs as between the indicators. The macro model alone will not do: it will have to be interfaced with and supported by 'secteral' (e g, banking) and satellite or indicator models with adequate abilities to track behavioural relations and to pro- vide robust forecasts. The forecasted val- ues that indicator models provide will generally depend on past trends while the changes and events that occur in different sectors including the behavioural relations in them will provide important inputs to the macro model so that the internally consistent optimal values will be contem- poraneous and plausible.
Simultaneously we shall undertake the distinct task of classifying the 12 key standards of Sheng Task Force by their main functional characteristic and see how compliance to each of the standards would transmit influence on the identified indi- cators. Of the 12 core standards, one is on data dissemination (SDDS/GDDS), and two on transparency (see Exhibit 1). These three may be regarded as a block repre- senting the functional characteristic of data and policy transparency. The good market practices are reflected in two other stan- dards relating to accounting and auditing. Market integrity and infrastructure is sought to be achieved by committee or official entity approach to standards in the case of four other standards pertaining to pay- ments, banking and insurance supervision and securities regulation. Economic jus- tice characterises the remaining three stan- dards concerning corporate governance, insolvency and money laundering.
Understanding of the processes through which the application of different blocks of standards and codes would transmit their effects is important in order to ap- preciate the complex relationship between the objectives and policy measures. The data and policy transparency block, it may be expected, would help to contain, if not completely eliminate the possibility of emergency of asymmetric information. As is well known, such an outcome would help reduce adverse selection and im- prove macroeconomic adjustments in the public sector - private sector inter- face. Informed investments including those stimulated by credit enhancement are likely to occur, as a result. Given the productivity, investment growth should lead to output growth and perhaps be- cause of it, price stability would be fostered. As monetary stability is nece- ssary for financial stability in that the sta- bility of asset prices could proiAde ex- pected future streams of income and a basis for allocation of future income as between consumption and investment.
But it is important to recognise that commodity price stability can at best be a necessary but not a sufficient condition for financial stability. One could even go to the extent of arguing that where hed- ging instruments are available, financial stability could be at least theoretically, secured irrespective of commodity price stability. Again, it could be contended that transparency is a double-edged weapon: it could contribute to financial stability and could endanger it at times. 'Herd behaviour' could make an otherwise sustainable equi- librium, an unsustainable one. Similarly, credit growth which could occur due to better information, could, owing to better implementation of supervision standards and regulatory scrutiny be hit hard and lead to alteration in lending behaviour of financial institutions, especially banks. As a result, institutions may be forced to concentrate on investments in risk-free securities and shrink loan portfolios.
These counter arguments are as important as the ones that point to positive transmis- sion process of effecting data and policy transparency. However, once markets are integrated and market players are aware of standards, scepticism about their positive impact may be expected to be very weak.
The market practices block provides the investors a good basis for better valuations of businesses, and for taking decisions relating to allocation of their resources. For financial and business entities, the block implies accountability to their share- holders and better resources management. The block also helps entities to access markets, improve corporate governance and chalk out strategies for enhancing production. The block provides both the entities and authorities with the knowl- edge of the parameters of market behaviour to policy and exogenous developments. This will help to assess the liquidity needs of entities. A timely positive policy re- sponse to such needs will help deepen and develop financial markets, which in turn would help improve allocative efficiency, and growth prospects. Here too the effects on financial market stability flow from the policies to improve growth.
The block relating to market integrity and regulatory infrastructure arrangements essentially pertains to regulatory and super- visory mechanisms that the authorities work out. The wide dissemination of the core principles would enable the markets to have 'information' of not only the authori- ties' concerns but also the penalties for non- compliance. The certainty of actions of the authorities would provide the confidence that the financial soundness of entities could be secured if the core principles are complied with. Implicitly, it would mean that where financial soundness is not secured, the authorities will provide the necessary safety valves. There is moral hazard implied here but economic entities could make informed portfolio choices, depending on the macroeconomic environ- ment. Such actions could raise overall
Exhibit 2
Standards Probability of Realisation of Optimal Value Output Prices CAR -- N
(X) (X) (X) --- (X)
1 1,1 1,2 1,3 1,N * * * * *
12 12,1 12,2 12,3 12,N
Notes: 1 Bracketed numbers, the X are optimal values derived from consistency macro model. 2 N is the finite number of the indicators chosen. 3 In 1,1 the first 1 will represent the relevant standard and the second 1 will represent the probable
value. Similarly in 1,2, the 1 will represent the relevant standard and 2 will be represented by a probable value. And so on.
1736 Economic and Political Weekly May 19, 2001
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:05 AM All use subject to JSTOR Terms and Conditions
investments in the economy, fostering in the process, growth. In this block, the direct effects are expected to be felt first on financial. institutional and market stability. Growth would in the eventemerge as a bye-product.
The economic justice block helps im-
prove the application of the rule of law for better governance. Compliance can be assured if enforceability of law is certain. Symmetry of treatment of entities would be secure where economic justice prevails. Risks can be better managed and economic structure can be re-engineered to improve competitiveness and allocative efficiency. Here again the direct effects would be more pronounced on growth, and stability would emerge as a consequential outcome.
The theoretical expectations of outcomes from the application of individual blocks of standards and codes may not turn out to be significantly different when all the blocks are simultaneously applied or pursued. But the point to note is that with the exception of the block on market integrity and regulatory infrastructure, the rest of the blocks seem to suggest that the direct effects of their appli- cation will be pronounced on growth. One may contend that financial stability would follow growth but this cannot be taken to be automatic. For securing stabi- lity together with or via growth is not possible unless economies undertake appropriately sequenced macroeconomic and structural reform policies. In other words, the issue of implementation of standards and codes should be placed squarely in the context of economic reforms and financial development.
The authorities cannot be content with the theoretical expectations of outcomes of implementation of standards and codes. They would still need to work out the probability values of realising the assigned optimal value of the economic indicators with the implementation of the identified standards and codes. This task is not going to be easy, for the probabilities may them- selves have to be suitably weighted. Since each of the 12 standards are not single -
point standards, and each standard is a composite of action points, it would be imperative for the national authorities to identify, with reference to their institu- tional and legal arrangements, the critical action points and focus their implemen- tation efforts on them. In other words, those which are deemed to be not so critical may not be assigned any weight or may be set aside from consideration. The prob-
ability values in respect of the identified critical action points could be given in the range of 0.1 to 1.0. The optimal internally consistent values against each indicator may also be provided. Exhibit 2 provides a simple representation of the probable values of attaining the optimal values vis-a-vis the application of one standard at a time.
As implementation of standards and codes may not in cases of some economies, be complete at a discrete point of time, the probability values of realisation of optimal values would itself differ from time to time. The probability values will give an idea of the strength of the linkage between the particular standard/code and the rel- evant indicator. The higher the value, the stronger the linkage, but the-relevant in- dicator may not necessarily reflect all the elements that are required for financial stability to materialise.
In reality, countries would be adopting more than one standard and code at a time. In the event of simultaneous application of standards and values, the probability values of realising the optimal values would undergo a change. Such an exercise would become complex requiring a large compu- tational capability when the number of standards and codes is expanded beyond a reasonable limit. In any case, it is im- portant for countries to work out series of matrices of relationships among (a) the standards themselves; and (b) between the standards and indicators both when indi- vidual standards are applied and when all the identified standards are applied. It is only then one could establish the empirical truth as to whether the linkage between the application of standards and codes and financial stability is likely to be strong.
Conclusion
Our analysis shows that efforts at imple- menting standards and codes without the concomitant adoption of sound macroeco- nomic and structural policies would hardly ensure that growth and financial stability could be co-terminus. This would, imply that while emerging market economies undertake financial reforms, they may have to adopt not one or two, but a requisite set of standards and codes in order to promote the twin objectives of optimal growth and financial stability. The process of trans- mission that we discussed above shows that information and expectations play a major role in the market calculus in un- dertaking investment decisions and resource allocative functions. E1
Notes [Views expressed here are the author's own. He acknowledges with appreciation the useful comments on the initial draft from Mridul Saggar and Partha Ray. However, the author alone is responsible for any errors in interpretation and in recounting of facts.]
1 The words 'systemically important' are used to represent both industrialised and emerging market economies.
2 Andrew, Crockett, 'Why is Financial Stability a Goal of Public Policy?' in Maintaining Financial Stability in a Global Economy, The Federal Reserve Bank of Kansas City, 1997.
3 Federic S Mishkin, 'The Causes and Propagation of Financial Instability: Lessons for Policy Markets' in Maintaining Financial Stability in a Global Economy, ibid.
4 Geoffrey Hodgson, EconomicsandInstitutions, Polity Press, Cambridge, UK, 1988 for adiscus- sion on 'markets as institutions', pp 172-194.
5 A Schotter, Free Market Economics, St Martin's Press, New York, 1985.
6 G L S, Shackle, Epistemics and Economics, Cambridge University Press, Cambridge, UK, 1972.
7 See F A Hayek, Law, Legislation and Liberty, Vol I, Rules and Order, Routledge and Kegan Paul, London, 1973.
8 The writing on this subject has been vast. It was more seriously discussed after the crisis in Thailand. Apart from the spate of literature on 'leading' indicators, there were many writings on multiple equilibrium analysis of vulner- ability. Insofar as foreseeing of the Mexican Crisis is concerned, refrence may be made to J Sachs, A Tornell and A Velafco, 'The Mexican Peso Crisis: Sudden Death or Death Foretold', mimeo, Harvard University, 1995; Also R Dorn- busch and A Werner, 'Mexico Stabilisation, Reform and No Growth', Brookings Papers on Economic Activity, No 1, 1994.
9 'Good' practices as against 'best' practices is an important concept since the latter gives the impression that 'one size fits all' approach is correct. Besides, as standards are likely to evolve, it is useful to go in for 'good' practices rather than 'best' practices.
10 See M Obstfeld, 'Models of Currency Crisis with Self-Fulfilling Features', NBER Working Paper No 5285, European Economic Review, 1995. Also, G Kaminsky, S Lizondo and C Reinhart, 'Leading Indicators of Currency Crisis', IMF Staff Papers, Volume 45, 1998.
11 See Financial Stability Forum, Report of the Task Force on Implementation of standards (Chairman: Andrew Sheng), March 2000. (Available on website of FSF: <<FSForum.org>>
12 C P Kindleberger, 'Standards as Public, Collective and Private Goods', Kyklos, Vol 36, 1983. Also, J Farell, and G Saloner, 'Coordin- ation through Committees and Markets', RAND Journal of Economics, Vol 19, 1988. Also, P A David and S Greenstein, 'The Economics of Compatibility Standards: An Introduction to Recent Research', Economics of Innovation and New Technology, Vol 1, 1990.
13 The Follow Up Group on Incentives to Foster Implementation of Standards, Financial Stability Forum, BIS, Basle (Switzerland), September2000. (Available on website ofFSF).
14 See footnote 11 above. 15 The Indian experiment is to have a Standing
Committee on International Standards and Codes in order to ensure that the evolved stan- dards could be adopted to take into account the institutional and legal arrangements unique to the Indian economy. Similar experiments could be conducted in otheremerging market economies.
Economic and Political Weekly May 19, 2001 1737
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:05 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 1733
- p. 1734
- p. 1735
- p. 1736
- p. 1737
- Issue Table of Contents
- Economic and Political Weekly, Vol. 36, No. 20 (May 19-25, 2001), pp. 1657-1768
- Front Matter [pp. 1657-1732]
- Letter to Editor
- Budget and Pensioners [p. 1658]
- Editorials
- A Mixed Bag [pp. 1659-1660]
- Belated Reform [p. 1660]
- Credibility Gap [pp. 1660-1661]
- Illusive Budgeting [pp. 1661-1662]
- Wages of Poor Governance [p. 1662]
- Current Statistics [pp. 1663-1664]
- Companies
- Strategic Restructuring [p. 1665]
- Cost Control Helps [p. 1665]
- Inhospitable Times [pp. 1665-1666]
- Money Market
- Entering a Low Interest Rate Regime [pp. 1667-1673]
- Commentary
- Panchayat Elections in Kashmir: A Paper Exercise [pp. 1674-1677]
- UP on the Financial Brink: State Government's Budget 2001-2002 [pp. 1677-1680]
- Panchayat Elections: Overcoming State's Resistance [pp. 1681-1684]
- Census of India 2001 and After [pp. 1685-1687]
- Chinese Workers in Israel: A Bizarre Tale [p. 1688]
- Perspectives
- Transfer Pricing: Impact on Trade and Profit Taxation [pp. 1689-1692]
- Reviews
- Review: Contempt of Court and Free Speech [pp. 1693-1694]
- Review: Theory Ahead of Application [pp. 1695-1696]
- Review: Social Work as Praxis [p. 1696]
- Special Articles
- Capital Formation in Indian Agriculture: Re-Visiting the Debate [pp. 1697-1708]
- Between Dialogue and Conflict: Deendar Anjuman, 1920s-2000 [pp. 1709-1717]
- Questionable Economics of LNG-Based Power Generation: Need for Rigorous Analysis [pp. 1718-1725]
- International Standards and Codes and Financial Stability [pp. 1733-1737]
- Gujarati Business Communities in East African Diaspora: Major Historical Trends [pp. 1738-1747]
- Caste and Agrarian Class: Errata [p. 1747]
- Special Statistics-29: Finances of State Governments: A Time Series Presentation [pp. 1748-1761]
- Back Matter [pp. 1762-1768]
IslamabadSafeguarding Financial Stability Theory and Practice.pdf
Pakistan Institute of Development Economics, Islamabad
Safeguarding Financial Stability: Theory and Practice by Garry J. Schinasi Review by: Kalbe Abbas The Pakistan Development Review, Vol. 44, No. 2 (Summer 2005), pp. 223-225 Published by: Pakistan Institute of Development Economics, Islamabad Stable URL: http://www.jstor.org/stable/41260716 .
Accessed: 06/02/2015 10:54
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Pakistan Institute of Development Economics, Islamabad is collaborating with JSTOR to digitize, preserve and extend access to The Pakistan Development Review.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:54:16 AM All use subject to JSTOR Terms and Conditions
Book Reviews 223
Garry J. Schinasi. Safeguarding Financial Stability: Theory and Practice. Washington, D. C: IMF. 2006. xv+311 pages. $28.00.
Financial stability plays an important role in economic growth and leads to sustained development. As such, financial stability is gradually emerging »as a distinct policy function. Safeguarding Financial Stability, by Garry Schinasi, explains why financial stability matters, what it means, and what challenges are involved in securing it.
The book is divided into three parts and contains twelve chapters. Chapter 1 gives the outline of the book and provides answers to various related questions. It highlights the increasing importance of financial stability issues, pinpoints the need for an analytical framework, and describes the specific objectives and organisation of the book.
Part 1, titled "Foundations" and comprising three chapters, reviews the important logical foundations that show how the process of finance is related to the real economic process and why financial stability is viewed as providing public goods and requiring forms of private-collective public policy actions. An effective process of finance requires extensive private-collective and public policy involvement to capture social economic benefits. The chapters in this part provide a logical foundation for thinking about financial stability issues. Chapter 2 in this part discusses the huge economic benefits provided by an effective process of finance that improves economic efficiency and facilitates resource and risk allocation, wealth accumulation, growth, and social prosperity. It also examines the characteristics of finance that reduce private and social benefits and create financial and economic instability. Chapter 3 explores the public policy aspects of finance. It provides information on the sources of imperfections in finance including externalities, public goods, and incomplete markets. The chapter identifies sources of market imperfections in finance, justifies the role for both private-collective and public policy involvement, and argues that both fiat money and finance have the potential to convey significant positive externalities and the characteristics of a public good. The chapter relates the economics of public sector to finance that is associated with significant positive externalities and requires a balance between maximising social benefits and minimising social costs. Chapter 4 briefly defines efficiency and stability from an economic perspective. The implicit and practical importance of this distinction is that not all market imperfections in finance may necessitate a private- collective or public policy response. Whether intervention is desirable or necessary depends on the size and importance of the imperfection with regard to its impact on efficiency. It concludes that deviation from the efficient outcome should be part of the decision to intervene, though it is difficult to measure in practice. This chapter distinguishes between volatility, fragility, and instability by drawing on the experience in the 1990s and early 2000s marked by market instability and country
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:54:16 AM All use subject to JSTOR Terms and Conditions
224 Book Reviews
crises. The chapter briefly illustrates the distinction between inefficiency and
instability through supply and demand diagrams. Part II, titled "Towards a Framework for Financial Stability" and comprising
Chapter 5 through 7, presents a comprehensive and practical framework for safeguarding financial stability for both the prevention and the resolution of financial imbalances, problems, and crises. This part provides a set of definitions, concepts, and organising principles that impose discipline on the analysis of financial system. Chapter 5 develops a working definition of financial systems, financial stability, and systematic risk in terms of measurable economic process and identifies several practical implications of the definitions for financial stability work. It discusses the strengths and weaknesses of finance, brings financial system and systematic risk into focus, defines financial stability, and examines its key implications. Chapter 6 provides a generic framework for financial-stability monitoring, assessment, and policy element of financial stability. It develops a comprehensive framework for safeguarding financial stability through resolving financial imbalances and crises. The framework describes all the important aspects of financial systems such as institutions, markets, and infrastructure. Its implementation involves monitoring, analytical assessment, and policy adjustment. This chapter also identifies the remaining analytical and measurement challenges. Chapter 7 discusses the role of central banks in ensuring financial stability by controlling other supervisory and regulatory authorities. It explores what central banks need to do and how far central banks have gone to ensure financial stability through examples from the Euro zone, Japan, the United Kingdom, and the United States.
Finally, Part III, titled "The Benefits and Challenges of Modern Finance", Chapter 8 through 12, identifies and analyses the ongoing challenges to financial efficiency and stability posed by relatively recent structural changes in national and global finance. Each of these structural changes is improving financial and economic efficiency at the cost of new risks involved. Chapter 8 examines the process and the challenges posed by financial globalisation and its impact on market dynamics and international financial system. The chapter describes the changed nature of threats to financial stability and systematic risk. In Chapter 9, the potential for instability in national and global financial markets related to the growing reliance on over-the- counter (OTC) derivative instruments and markets is examined in detail. It describes exchange versus derivative markets, organisation of markets, sources of volatility and potential fragility in OTC derivatives activities and markets, and the weaknesses in the infrastructure. Chapter 10 discusses market tests during the slowdown of global growth in 2001-2002 and corporate downgrades in 2005. The industry challenges and implications for retail investors are also discussed in this chapter. Chapter 11 examines the supervisory, regulatory, and systematic challenges raised by the greater role of insurance companies in financial and capital market activities. It draws systematic implications of financial market activities of insurance and
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:54:16 AM All use subject to JSTOR Terms and Conditions
Book Reviews 225
reinsurance companies that are important for a growing class of financial market participants. Chapter 12 summarises the main challenges to financial stability that are discussed in the book and are likely to be faced in the future. It concludes that further continuous reforms are desirable. The authors rightly maintain that in advanced countries with mature markets, more reliance on market discipline is desirable. However, in developing countries, with poorly developed markets, strong efforts need to be made to improve the financial infrastructure through private- collective and government expenditures.
The book is very useful reading for anyone interested in the subject of financial stability, but it is meant particularly for practitioners as well as policy- makers, and will generate debate, and perhaps further research. It combines a large literature on financial stability with solutions for a number of crucial problems.
Pakistan Institute of Development Economics, Islamabad.
Kalbe Abbas
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:54:16 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 223
- p. 224
- p. 225
- Issue Table of Contents
- The Pakistan Development Review, Vol. 44, No. 2 (Summer 2005), pp. 117-232
- Front Matter
- P-Star Model: A Leading Indicator of Inflation for Pakistan [pp. 117-129]
- Reproductive Tract Infections among Women in Pakistan: An Urban Case Study [pp. 131-158]
- The Primary Sectors of the Economy and the Dutch Disease in Nigeria [pp. 159-175]
- Real Exchange Rate, Exports, and Imports Movements: A Trivariate Analysis [pp. 177-195]
- The Northern Immigration Policy in a North-South Economy Model [pp. 197-218]
- Book Reviews
- Review: untitled [pp. 219-222]
- Review: untitled [pp. 223-225]
- Review: untitled [pp. 226-228]
- Shorter Notices [pp. 229-232]
- Back Matter
Macroeconomic Stability, Financial Stability and Monetary policy.pdf
International Finance 15:2, 2012: pp. 205–224
DOI: 10.1111/j.1468-2362.2012.01302.x
Macroeconomic Stability, Financial Stability, and Monetary
Policy Rules∗
Pierre-Richard Agénor† and Luiz A. Pereira da Silva‡
†University of Manchester; Centre for Growth and Business Cycle Research; and FERDI (Fondation pour la Recherche et le
Développement International), and ‡Central Bank of Brazil
Abstract
This paper reviews arguments for and against attributing an explicit finan- cial stability objective to monetary policy. The discussion is conducted from the perspective of middle-income countries (MICs), where bank credit plays a critical role both on the supply and demand sides. It also discusses, on the assumption that a more proactive role is desirable, what monetary policy should react to and to what extent it should be combined with macropru- dential regulation. There are robust arguments in favour of monetary policy reacting in a state-contingent fashion to a measure at the private-sector credit gap, not only because of financial stability considerations but also
∗This paper dwells in part on some of our previous papers, including joint work with Koray Alper (Central Bank of Turkey). We are grateful to Koray Alper, Karim El Aynaoui, participants at the Inter-American Development Seminar for Central Banks and Finance Ministries (Washington, DC, 21–23 September 2011) and two anonymous referees for helpful discussions and comments. However, we bear sole responsibility for the views expressed here. A more detailed version of this paper is available upon request.
C© 2012 Blackwell Publishing Ltd. 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
206 Pierre-Richard Agénor and Luiz A. Pereira da Silva
because of the high degree of uncertainty regarding real-time estimates of the output gap in MICs. Nevertheless, monetary policy is not a substitute for macroprudential regulation; in particular, it cannot address the cross- section dimension of systemic risk.
I. Introduction
The global financial crisis has led to both a reassessment of financial regulatory systems worldwide and renewed calls for central banks to consider more ex- plicitly and systematically financial stability considerations in setting monetary policy. On the regulatory side, a number of proposals aimed at strengthening the financial system and encouraging more prudent lending behaviour in upturns have been put forward. In particular, it has been argued that by raising capital requirements in a contra-cyclical way, regulators could help to choke off asset price bubbles – such as the one that developed in the US housing market – before a crisis develops.1 Along these lines, and after months of internal debate, on 12 September 2010 the Basel Committee on Banking Supervision (BCBS) released a new capital framework, which not only strengthens the definition of capital but also recommends the implementation of both a capital conservation buffer and a countercyclical capital buffer.
On the monetary policy side, it has been argued that central banks should con- sider more systematically potential trade-offs between the objectives of macroe- conomic stability and financial stability.2 One reason for this is the growing concern among academics and policy makers that the achievement of price sta- bility may have been associated with an increased risk of financial instability. Indeed, it has been argued that financial imbalances may build up even in an environment of stable prices; low and stable rates of inflation may foster asset price bubbles, due for instance to excessively optimistic expectations about fu- ture economic prospects or to increased incentives to take on more risk. Thus, price stability may not be a sufficient condition for financial stability. At the same time, however, several observers have argued that trying to stabilize asset prices per se is problematic for a number of reasons – in particular because it is almost impossible to know for sure whether a given change in asset values results from
1 See Financial Services Authority (2009) and Brunnermeier et al. (2009). Agénor and Pereira da Silva (2010, 2012) offer a developing-country perspective.
2The debate actually predates the global financial crisis and initially focused on the extent to which monetary policy should respond to (or ‘lean against’) perceived misalignments in asset prices, such as real estate and equity prices, as opposed to ‘cleaning up after’. See Wadhwani (2008) for a review.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 207
changes in underlying fundamentals, non-fundamental factors, or both. Some observers have argued that, instead of getting into the tricky issue of deciding to what extent asset price fluctuations reflect changes in the economy’s fundamen- tals, central banks should focus on the implications of asset price movements for credit expansion and aggregate demand, and thus inflationary pressures.
This paper focuses on the second issue – the extent to which monetary policy should be concerned explicitly with financial stability objectives and, if so, to what financial indicators it should be made responsive. We do so in the context where macroprudential regulation is also a component of the policy framework aimed at preventing disruptive and costly financial crises.3 To conduct this analysis, and in contrast to much of the existing literature, we focus on middle-income countries (MICs) only. We do so for several reasons. First, financial markets in many of these countries remain underdeveloped. In most MICs, commercial banks continue to dominate the financial system.
Second, and related to the lack of financial diversification, bank credit has an important impact on the supply side of the economy. Firms borrow short term to finance their working capital needs (such as labor inputs and raw materials) prior to the sale of output. Third, the financial system in MICs is often highly vulnerable to small domestic or external disturbances. Abrupt reversals in short-term capital movements tend to exacerbate financial volatility.4 A number of studies have indeed documented a positive relation between the increasing international capital flows due to greater integration with world financial markets and the vulnerability to sudden reversals in capital flows. Forbes and Warnock (2012), for instance, found that global factors play an important role in explaining ‘waves’ of international capital flows. The more open and integrated a country is to global financial markets, the deeper are the channels through which reversals in capital flows will impact both the real economy and the financial system – and the more critical the policy response becomes to ensure macroeconomic and financial stability.
Fourth, MICs have suffered many costly crises over recent decades, with large drops in output, persistent credit crunches and sharp increases in unemployment and poverty. Although the exact trigger to these crises can be a wide range of events (including political turmoil, a real-estate crash, a sharp decline in the terms of trade or contagion from other economies), making it hard to predict
3Our discussion of macroprudential policy is thus focused on the extent to which it interacts with monetary policy. For a more general discussion, including coordination issues between these two policies, see Committee on the Global Financial System (2010), Financial Stability Board (2011), Galati and Moessner (2011) and International Monetary Fund (2011b).
4See Agénor (2012) for a thorough review of the evidence on, and the challenges posed by, international financial integration.
C© 2012 Blackwell Publishing Ltd
208 Pierre-Richard Agénor and Luiz A. Pereira da Silva
their exact timing, they are often preceded by sustained imbalances. Thus, any measure that can help to identify sources of weaknesses, prevent these imbalances from emerging, and minimize the chances of a crisis occurring may have large welfare benefits.
The remainder of this paper proceeds as follows. Section II considers the case against using monetary policy to react directly to financial instability; from our perspective, this is tantamount to arguing that macroprudential tools, possibly supplemented by capital controls, are enough, or more than enough to mitigate systemic risk. Section III considers the case for a more proactive monetary policy in response to perceived risks to financial stability, above and beyond the conventional objectives of price and output stability. Section IV discusses what monetary policy should react to, assuming indeed that a more proactive role is desirable. Section V addresses the issue of whether monetary policy should be combined with macroprudential regulation (and possibly capital controls) using a rule-based approach. The last section offers some concluding remarks.
II. The Case Against a More Proactive Role for Monetary Policy
There are a number of arguments that militate against using monetary policy to directly address financial stability concerns.
The first is the so-called Tinbergen principle, which states that to attain a given number of independent policy objectives, there must be at least an equal number of instruments.5 For the issue at hand, with macroeconomic stability and financial stability being the two objectives, this means that two separate tools are needed – the policy interest rate and a macroprudential tool. Put dif- ferently, policy makers necessarily need a tool other than the interest rate – particularly if there are potential trade-offs between policy objectives. With an additional instrument, and in a deterministic environment, the central bank can achieve exactly, and continuously (through dynamic rules) its targets; the two instruments are necessarily complements. From this perspective, the issue of whether monetary policy should respond to financial stability concerns is simply not relevant; it must be combined with macroprudential policy, regardless. This is, implicitly at least, the argument put forward by Svensson (2010). In prac- tice, however, central banks operate in a stochastic world and aim to minimize deviations from their targets rather than achieving them exactly and continu- ously; and because each instrument, manipulated independently, may affect both
5Tinbergen’s principle is concerned with the existence and location of a solution to the system; it does not assert that any given set of policy responses will, in fact, lead to that solution. To assert this, it is necessary to investigate the stability properties of a dynamic system.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 209
targets in the same direction (thereby reducing volatility in both cases), they may be substitutes. This issue is discussed further in Section IV.
The second argument is that, to the extent that it affects all lending activities (regardless of whether they represent a risk to stability), the policy interest rate is too blunt an instrument to be useful in addressing financial stability concerns, which often have a sectoral dimension – such as, for instance, overheating of the housing market. From that perspective, imposing a cost on the entire economy is not warranted – even though there is evidence to suggest a high correlation between credit expansion, which depends on the cost of borrowing and thus the policy rate, and house price inflation (see Glindro et al. 2008; Goodhart and Hoffman 2008; Claessens et al. 2011). Because the effect of higher policy rates on bank risk taking may depend on each institution’s initial capital position, the net aggregate effect may be limited. Banks with a low capital base (or less to lose), for instance, may try to ‘gamble’ by expanding the asset side of their balance sheets, by lending to increasingly riskier borrowers, whereas highly capitalized banks may choose to diversify their portfolios towards less risky assets. In addition, trying to ‘prick’ a developing housing price bubble through a (possibly very large) economy-wide increase in the cost of borrowing could have an immediate adverse effect on the supply side, given the importance (as indicated earlier) of bank credit in financing working capital needs. In turn, this may increase macroeconomic volatility. Under such conditions, sectoral prudential tools (such as changes in loan-to-value ratios, debt-to-income ratios, countercyclical capital requirements for real-estate lenders, and so on) may be more appropriate to prevent risk concentration.6
The third argument goes even further – depending on the nature of shocks, monetary policy may need to be conducted with caution, because of potentially undesirable side effects. This is what occurs when a country is confronted with a sudden flood of private capital, that is, large inflows induced by changes in external market conditions (Agénor et al. 2012). Indeed, sudden floods have on numerous occasions been a source of macroeconomic instability in many MICs, having led to rapid credit and monetary expansion (due to the difficulty and cost of pursuing sterilization policies), asset price pressures, real exchange rate appreciation and widening current account deficits. This occurred in the aftermath of the surge in capital flows to MICs during 2008–09 and in previous episodes.7 At the same time, the scope for responding to the risk of macroeconomic and financial instability through monetary policy is limited because higher domestic interest
6However, it is important to recognize at the same time that targeted tools, although they may be less costly than an economy-wide increase in interest rates, could be easier to circumvent than broader measures.
7See Jongwanich (2010) and Furceri et al. (2011) for instance.
C© 2012 Blackwell Publishing Ltd
210 Pierre-Richard Agénor and Luiz A. Pereira da Silva
rates vis-à-vis interest rates in advanced economies may simply exacerbate the flood of private capital. Put differently, monetary policy loses its effectiveness and other instruments (macroprudential tools, capital controls) must be used to manage capital flows and mitigate their destabilizing effects on the domestic economy.
A fourth and related argument is that strengthening macroprudential rules, using both ‘old’ instruments (such as liquidity or leverage ratios, loan-to-value and debt-to-income ratios and so on) and ‘new’ tools, such as countercyclical capital buffers linked to a measure of excessive credit expansion (as envisaged un- der Basel III) and dynamic provisioning, offers a better alternative to monetary policy. In fact, both types of instruments have been used in MICs for years. The Central Bank of Brazil introduced a capital charge in 2000, through a mechanism that links the deviation of credit growth relative to GDP growth. More recently, dynamic provisioning rules have been introduced in several Latin American countries (see Wezel 2010). In addition to reducing balance sheet vulnerabili- ties, these instruments have helped to reduce risk taking and strengthened the financial sector (at least in the case of dynamic provisions), explaining in part why MICs were able to weather the recent global financial crisis with limited strain. As documented by Montoro and Moreno (2011), for instance, reserve requirements were used in Latin America in a countercyclical fashion to smooth the expansion phase of the cycle and to tighten monetary conditions without attracting capital inflows. During the global financial crisis, reserve requirements were lowered, in order to inject liquidity rapidly in local and foreign currency, and restore market activity affected by sudden reversals in capital inflows.8 In another study on Latin America, Terrier et al. (2011) provided a broader review of microprudential policy tools used or available to policy makers in the region to mitigate the procyclical effects of financial cycles. They concluded that, although mainly microprudential in nature, when appropriately calibrated and used in combination over the financial cycle these tools may prove effective for macro- prudential purposes and could contribute significantly to addressing systemic risk.
A fifth argument is that if financial imbalances are related to excessive credit growth, and if credit expansion is fuelled by capital inflows (as is often the case in MICs), then a more effective policy could be to complement macroprudential tools – at least temporarily – with capital controls. The evidence regarding the effectiveness of capital controls is, at best, mixed. In the 1990s, capital controls were only temporarily able to drive a wedge between foreign and domestic interest rates and to reduce pressures on the exchange rate in countries such as Brazil, Chile, Colombia, Malaysia and Thailand (Ariyoshi et al. 2000). More
8Some countries in the region (namely, Brazil and Colombia) also resorted to capital controls.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 211
recent reviews, which include the Committee on the Global Financial System (2010), Agénor (2012), Habermeier et al. (2011) and the International Monetary Fund (2011a), reached similar conclusions: capital controls appear to have had little effect on overall capital flows, although they may have had some success in altering the composition of these flows.9 In most cases, controls have not been successful at mitigating currency appreciation. Specific econometric estimates on the effectiveness of capital controls covering four MICs during the 2000s (Brazil, Colombia, Korea and Thailand) confirm that controls have met with mixed success. It also appears that the effectiveness of any given measure decays over time. Nevertheless, temporary effectiveness may well be all that policy makers need when faced with sudden floods and neither monetary policy nor macroprudential policy can respond quickly.
A sixth argument is that if the central bank lacks credibility, adding a financial stability objective to monetary policy may confuse markets, weaken perceived commitment to price stability and destabilize expectations – thereby making it more difficult to maintain low inflation. In such conditions, there may be a stabilization cost associated with using monetary policy in a proactive manner. Suppose for instance that policy makers are faced with a negative demand shock that lowers both output and inflation. In an inflation-targeting regime, the correct policy response is to lower the policy rate; there is no trade-off between macroeconomic objectives. But if the central bank is concerned with systemic risk (perhaps because of the belief that low interest rates may promote risk taking motivated by a ‘search for yield’, as discussed later), a conflict between macroeconomic and financial stability objectives emerges: keeping interest rates high means that the risk of deflation must be accepted.
Under such conditions, some observers have proposed as a policy response to lengthen the horizon for achieving the inflation target. This is the same response typically advocated in the case of a (persistent) supply shock, which entails a trade-off between output and inflation. However, concerns about systemic risk, which includes both time and cross-sectional dimensions, may be difficult to convey to agents. Indeed, even though substantial progress has been achieved in recent years, there is still no consensus on defining ‘financial stability’ and how to measure it in its various dimensions. Consequently, lengthening the target horizon may have adverse effects on inflation expectations and central bank credibility. Similar reasoning suggests that allowing instead a wider fluctuation band for the inflation target could have equally adverse effects on credibility.
9See recent studies by Gochoco-Bautista et al. (2010), McCauley (2008), Habermeier et al. (2011) and Jongwanich et al. (2011).
C© 2012 Blackwell Publishing Ltd
212 Pierre-Richard Agénor and Luiz A. Pereira da Silva
III. The Case for a More Proactive Monetary Policy
There are also a number of arguments that militate in favour of making monetary policy more directly responsive to a financial stability objective.
The first argument is that monetary policy, precisely when it is successful at maintaining low and stable prices, may itself induce boom–bust cycles in asset prices; low interest rates may encourage increased risk taking, excessive leverage and promote a ‘search for yield’.10 If so, then there may be a trade-off between macroeconomic and financial stability. This argument has been used in part to highlight a contributing factor to the recent financial crisis: the low interest rates and low inflation that have been associated with the Great Moderation created in advanced economies an environment encouraging increased risk taking – with a switch from lower yielding safe assets into higher yielding risky assets, driving their prices up in the process – and more leveraging, which subsequently led to asset price bubbles. Bean et al. (2010) and Ahrend (2010) found indeed that, during periods when short-term interest rates have been persistently and significantly below what Taylor rules would prescribe, monetary policy has had a significant effect on increases in asset prices, especially housing prices.11
However, the fact that an accommodative policy stance may have an impact on asset prices and credit growth does not mean that monetary policy should respond directly to these variables: if increases in asset prices and credit expansion are expected to lead to an expansion in aggregate demand (through wealth and direct effects on private spending), a policy that reacts to the output gap and (expected) inflation would naturally lead to an endogenous policy response. There would be no need to respond directly to these variables. Put differently, excessive asset prices and credit growth matter only to the extent that they affect the future path of output and inflation. And to the extent that there are trade-offs between (future) financial (in)stability and present macroeconomic stability, they should be addressed through more targeted macroprudential measures, rather than tighter monetary policy.
It is also important to note that there is no evidence that (loose) monetary policy has been a systematic cause of boom–bust cycles in credit and asset prices in MICs. To begin with, very few MICs maintained policy interest rates at low levels for extended periods, so identifying periods during which the correlation
10See Rajan (2005) for the ‘search for yield’ argument. Bean et al. (2010) provide a brief review of the alternative channels through which loose monetary policy may encourage increased risk taking. Gambacorta (2009) provides evidence on the risk channel for industrial countries.
11See, however, Bernanke (2010) for an alternative view in the case of the United States. Svensson (2010) also rejects the view that the financial crisis was caused by an excessively accommodative monetary policy stance.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 213
between low interest rates and risk taking can be studied in large samples is difficult. A more substantive reason for the lack of evidence on this correlation is that banks in these countries have for years maintained capital ratios well above those required by international standards, as documented by Agénor and Pereira da Silva (2010) and Fonseca et al. (2010). In a sense, having more ‘skin in the game’ reduced incentives to gamble and may have prevented a weakening of balance sheets through imprudent lending practices. A third reason is the fact that in many countries sectoral (micro) prudential tools were actively used to mitigate excessive risk taking. In addition, with non-competitive credit markets (a common characteristic of banking in MICs), low policy rates may mean higher bank spreads, higher profits and possibly less risk. Put differently, if there is no evidence that monetary policy has potentially perverse side effects on financial stability, there should be less concern in attributing a financial stability target to it.
In general, excessive risk taking has to do with procyclicality, which is itself driven by optimistic expectations and the tendency by lenders to relax lending standards and underprice risks in good times. It is indeed well documented that bank intermediation is highly procyclical in MICs (see Claessens et al. 2011; Calderón and Fuentes 2011). Under such conditions, monetary policy – possibly in combination with some specific macroprudential tools – could help to mitigate procyclicality and thereby address the time dimension of systemic risk, through its effect on the economy-wide cost of borrowing.
A second and related argument is that while monetary policy should not be used to ‘prick’ stock market bubbles, it could be quite effective at deflating debt- financed bubbles, especially if they are credit-financed – a common scenario in MICs.12 By inducing a direct and across-the-board increase in the cost of borrowing, monetary policy may be more powerful than macroprudential policy in these circumstances.
A third argument is that it is not obvious that macroprudential policy was all that successful prior to the crisis. Indeed, in several MICs macroprudential measures did not prevent rapid credit growth in the lead-up to the crisis. Prior to the onset of the global financial crisis, credit growth was accelerating in many countries in Latin America, including Brazil, Colombia, Peru and Venezuela.13
A good question is whether these countries would have faced a crisis, even
12Blinder (2010) and Mishkin (2011) have both emphasized the distinction between credit- fuelled bubbles (such as house price bubbles) and equity-type bubbles (in which credit plays only a minor role) in their analysis of post-crisis monetary policy. However, they are fairly agnostic as to whether the central bank should try to limit credit-based bubbles through regulatory instruments or interest rates.
13See for instance the April 2011 issue of the IMF’s World Economic Outlook, pp. 5, 76–9.
C© 2012 Blackwell Publishing Ltd
214 Pierre-Richard Agénor and Luiz A. Pereira da Silva
without turmoil in advanced economies; if history is any guide, the likelihood appears to be quite high. But rather than an argument in favour of greater reliance on monetary policy, this evidence may be construed as a call for using macroprudential tools more aggressively or for adding new tools to the arsenal of policy makers. Indeed, Colombia (between July 2007 and July 2008) and Peru (in November 2008) both introduced dynamic loan provisioning systems in the aftermath of the global financial crisis. At the same time, the less effective macroprudential tools are, the greater the potential role of monetary policy in contributing to the maintenance of financial stability.
A fourth argument is that macroprudential policy is more subject to lobbying and political pressure than monetary policy. A case in point is the worldwide reaction of the financial sector to the proposed new Basel rules for higher capital requirements, even though research (for the United States and other countries) shows that this policy is likely to lead to only a modest increase in the cost of credit.14
A fifth argument is that too much reliance on macroprudential policy, to the extent that it limits bank credit availability or leads to higher borrowing costs, may foster financial disintermediation by promoting the development of shadow banking and the informal sector – making it in turn difficult to maintain financial stability. From this perspective, the scope and bluntness of the policy rate could be an advantage over macroprudential measures, as it is more difficult to circumvent a general increase in borrowing costs induced by a monetary policy contraction.
A sixth argument is that some of the ‘new’ macroprudential tools envisaged in Basel III are largely untested. There is no clear consensus yet on what tools will work and there is very little evidence on their effectiveness. For instance, regarding the performance of dynamic loan provisioning systems, much of the evidence relates to the Spanish case (see Saurina 2009); yet the conclusion from most studies is that even though these systems may succeed in making banks more resilient, they appear to have limited effectiveness when it comes to re- straining credit expansion.15 Similarly, the introduction of countercyclical capi- tal buffers (just like other macroprudential tools) may create serious operational and institutional challenges, especially in countries where the supervisory envi- ronment is weak to begin with – as is the case in many MICs. It is also not clear what variables they should be related to during the buildup and release phases.
14See Admati et al. (2011) for the impact of capital requirements on the cost of equity and Igan and Mishra (2011) for a discussion of the connection between financial lobbying and financial legislation in the lead-up to the US financial crisis.
15As noted earlier, several countries in Latin America have introduced dynamic loan provisioning systems in recent years, but the experience is too recent to provide new insights.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 215
Interactions among macroprudential tools are also not well understood; a case in point is the interaction between bank capital requirements and dynamic loan provisioning systems.16 Finally, and quite importantly, some macroprudential tools may alter the way the monetary transmission mechanism operates (see Agénor and Pereira da Silva 2011). What this all means is that there is a good case, if only for a transitory period (during which a better understanding of these issues can be acquired), to rely more on monetary policy to respond to financial stability concerns.
A seventh argument is that both macroeconomic instability and financial instability tend to increase in the lead-up to financial crises.17 This creates a case for monetary policy to react promptly, in normal times, to indications of growing financial vulnerability. By ‘leaning against the financial cycle’, a more active monetary policy may help to stabilize conventional targets (output and inflation). In that case then there could be a stabilization dividend. Indeed, a stable and sound financial system can contribute to macroeconomic stability by facilitating the transmission of monetary policy actions and cushioning the impact of macroeconomic shocks through the financial sector. In addition, a stable and sound financial system may decrease the incidence of financial stress and lead to less disruption in economic activity, which in turn contributes to price stability.
A final argument is that the view, according to which adding a financial stability objective may adversely affect central bank credibility, depends in part on initial conditions. If, for instance, inflation is initially above target, a rise in the policy rate motivated by systemic risk concerns may actually be beneficial. What the ‘credibility problem’ means is that there are new challenges for central banks in terms of transparency and communication of its policy decisions, and the indicators upon which they are based, but these are not insurmountable. After all, when some central banks in MICs initially adopted a measure of ‘core’ inflation, as opposed to headline inflation, as their measure of price stability, they faced significant problems in conveying to the public the nature of their objective, and the reasons for making their particular choice; over time, with communication improving, these issues became better understood. There is no reason to believe
16The common view is that bank capital should cover for unexpected credit losses, whereas dynamic loan loss provisions are intended to cover expected credit losses. However, introducing either one of those regulatory regimes while the other is present may change the behaviour of banks and thus the effectiveness of both types of tools. This may occur, for instance, if the reasons why banks hold (excess) capital buffers are altered by the introduction of loan loss provisions, and if capital buffers have a signalling effect that translates into changes in their market borrowing costs.
17See Demirguc-Kunt and Detragiache (2005) for a review of the evidence for developing countries.
C© 2012 Blackwell Publishing Ltd
216 Pierre-Richard Agénor and Luiz A. Pereira da Silva
that the same may not occur with a financial stability target – even though, as noted earlier, there is no consensus yet on how to measure financial stability. A good point of departure would therefore be to begin with a definition of financial stability as a final target. Because the concept has proved elusive, this is not a simple task; a sensible strategy perhaps is to follow an operational approach and respond to an intermediate financial target, as discussed in the next section.
IV. What Should Monetary Policy React to?
Assuming that the balance of arguments is in favour of a more proactive role for monetary policy – if only for a transitory period, as noted earlier – in addressing financial stability concerns, what should central banks react to? Many MICs have adopted a flexible inflation-targeting regime in recent years, with much success prior to the crisis. In these regimes, the optimal interest rate policy is a Taylor-type rule, which involves linking the policy interest rate to current or expected inflation and the output gap.18
Our view is that in the context of MICs, there is much merit in augmenting the interest rate rule by adding a measure of the private-sector credit gap, defined either in terms of growth rates (as the difference between the actual growth rate of that variable and a ‘reference’ growth rate), or in terms of deviations of the credit-to-GDP ratio with respect to a reference ratio. This would allow monetary policy (which can address only the time dimension of systemic risk, as noted earlier) to help to counter accelerator mechanisms that inflate credit expansion and asset prices, which are common manifestations of financial imbalances. In particular, rapid credit expansion tends to go hand-in-hand with a deterioration in lending origination standards and credit quality (see Dell’Ariccia and Mar- quez 2006). During upturns, credit standards tend to be more lenient, both in terms of screening of borrowers and in collateral requirements. As a result, a greater number of riskier borrowers are able to secure bank loans, whereas the share of collateralized loans tends to decrease. During boom times, the adverse selection problems created by informational asymmetries between lenders and borrowers are therefore magnified. In turn, the weakening of lending standards may increase vulnerability to financial distress when the economy experiences a downturn.
In addition, although credit and asset price cycles often exacerbate each other, several studies have found that credit is also a useful leading indicator of asset price busts; by contrast, there is no strong evidence that asset prices (in particular,
18See Svensson (1997) for a formal analysis. Taylor rules, sometimes augmented with an exchange rate pressure variable, appear to perform fairly well in practice for some MICs; see for instance de Mello and Moccero (2011) for Latin America.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 217
equity prices) are good out-of-sample predictors. More generally, rapid credit expansion – often associated with episodes of large capital inflows in MICs, as documented earlier – is often a warning sign of financial instability; even though not all episodes of credit booms end up in crises, almost invariably crises are preceded by episodes of credit booms. There is indeed robust evidence that credit booms significantly raise the likelihood of an asset price bust or a financial crisis in MICs.19 Recessions whose origin is the collapse of credit- fuelled bubbles – periods during which banks make loans that appear to have abnormally low expected returns – also tend to be more severe and longer lasting than those generated by ‘normal’ monetary policy contractions aimed at curbing inflationary pressures.
A third consideration is that most MICs do not have reliable data on land and property prices, and equity prices tend to be highly volatile. By contrast, credit data are readily available and usually subject to only small revisions (if at all) over time. In practice, many central banks in MICs are already paying much attention to credit growth – undoubtedly because of the importance of banks in the financial system, as discussed earlier.
Another important argument for responding to a credit growth gap is that this could be desirable not only for macroprudential reasons, but also because of the unreliability of real-time (preliminary) output gap measures in MICs. Differences in output gap measures based on real-time and final real GDP estimates can be quite substantial, as shown by Cusinato et al. (2010) for Brazil, with errors going in both directions. Similar results have been obtained for other countries. In the presence of large errors in the measurement of output gaps, it may in fact be optimal to reduce the weight attached to the output gap in a ‘real-time’ Taylor-type policy rule. At the same time, if the credit gap is closely related to final estimated output, the weight of that variable should be increased.
In a sense, the credit gap can be viewed as an intermediate target, concerns about which are easier to convey than those about a multi-faceted and hard- to-define final target, financial stability. In this approach, there is therefore an asymmetry in defining the central bank’s policy loss function, because infla- tion and output are final targets. Because of the difficulty of defining finan- cial stability as a final target (at least in the current state of affairs), using an intermediate target that is easier to identify may facilitate communication with the public and alleviate, to some extent, the credibility issues mentioned earlier.
19See International Monetary Fund (2009), Claessens et al. (2011) and Calderón and Fuentes (2011). Gerdesmeier et al. (2010) found that credit aggregates also play a significant role in predicting asset price busts in industrial countries.
C© 2012 Blackwell Publishing Ltd
218 Pierre-Richard Agénor and Luiz A. Pereira da Silva
The practical implementation of this ‘augmented’ policy rule needs of course to be thought out carefully. A first issue is whether the central bank should consider a real or a nominal credit gap (if it is measured in terms of growth rates), and whether it should consider a broad measure of aggregate credit or only a component of total credit. As noted earlier, working capital loans are related to changes in the supply side, not the demand side, of the economy; if the credit gap is to be used in part as a substitute to the output gap as a measure of excess aggregate demand, it might be argued that these loans should be excluded from the measure to which the central bank should respond to. However, it may also be argued that working capital loans are substitutes for firms’ internal resources (or cash flows), which can now be used to finance longer term investment – thereby indirectly affecting aggregate demand. This would militate in favour of using a broad aggregate. Fungibility and evergreening problems are also important considerations in choosing between narrow and broad credit aggregates.
A second issue is whether the ‘reference’ growth rate or credit ratio should be calculated as a trend (as proposed for instance in the calculation of the counter- cyclical capital buffer under Basel III) or rather on the basis of an equilibrium value that is related to some fundamental determinants, such as population growth, urbanization and so on. This second approach may be more appro- priate for MICs, because it would help to account for financial inclusion – an important consideration for many countries where the scope of the formal finan- cial system, and access to credit and other financial services, are limited to begin with. The implicit view here is that financial inclusion, by reducing reliance on the unregulated financial system, and increasing opportunities for risk sharing and consumption smoothing, helps to promote financial stability in the longer run (see Hawkins 2006).
At the same time, it is important to keep in mind that the credit gap is still a noisy indicator; false signals are inevitable and may raise the risk of policy errors. A policy response should be contingent on the magnitude of the credit gap, that is, it should occur only if the gap exceeds a certain threshold. In so doing, the primacy of the macroeconomic stability objective in ‘normal times’ would be maintained and credibility problems mitigated.
Yet, during episodes of sudden floods induced by external shocks, raising policy interest rates to account for excessive credit expansion may exacerbate the problem by triggering more inflows, as discussed earlier. Both points are arguments for combining (an augmented) monetary policy rule with macro- prudential tools. Indeed, given that monetary policy cannot address the cross- section dimension of systemic risk (that is, how risk is distributed within the financial system at a point in time), a combination of these two policies may be inescapable.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 219
V. How Should Macroprudential Regulation and Monetary Policy Be Combined?
An important practical issue for central banks is how an augmented monetary policy rule (of the type discussed earlier) and macroprudential rules should be combined. To address it requires understanding how the two policies interact. As noted earlier, even though the Tinbergen principle implies that in a deterministic world the two policies are complements if the two objectives (macroeconomic stability, financial stability) are to be achieved exactly, they may be substitutes if the central bank’s goal (in a stochastic environment) is to minimize deviations from targets over time, rather than achieve them exactly and continuously. This would occur if each instrument affects both targets in the same direction (lower volatility); due to decreasing marginal returns to each instrument, they may reinforce each other. It is therefore important to study jointly augmented monetary policy rules and macroprudential rules to understand how they should be combined.
Studies along these lines for MICs include those of Agénor et al. (2011, 2012), which focus on a Basel III type countercyclical regulatory rule and a monetary policy rule augmented with a credit gap variable (measured in terms of deviations of the growth rate of loans for investment from its steady- state value). Thus, the central bank sets its policy instrument in part to ‘lean against financial winds’ in a systematic fashion. Capital adequacy requirements are decomposed into a deterministic (minimum) requirement and a cyclical component related again to deviations in the growth rate of credit for invest- ment from its steady-state value. Macroeconomic stability is defined in terms of the volatility of a combination of the output gap and inflation, whereas fi- nancial stability is defined in terms of the volatility of a composite indicator that includes real house prices, bank loan spreads and the credit-to-GDP ratio. A composite index of economic stability is also defined, under the assump- tion that the central bank (still) attaches more importance to macroeconomic stability.
In response to a positive housing demand shock (meant to capture a housing boom) or a sudden flood (large capital inflows induced by external shocks), the analysis shows that the two instruments are complementary rather than substitutes; even with an aggressive interest rate response to inflation and credit gaps, it is optimal to also rely on the countercyclical regulatory rule under most circumstances. This complementarity is particularly important dur- ing episodes of sudden floods where, as indicated earlier, the central bank has limited ability to respond to inflationary pressures by raising interest rates.
C© 2012 Blackwell Publishing Ltd
220 Pierre-Richard Agénor and Luiz A. Pereira da Silva
VI. Concluding Remarks
A key issue on the agenda of policy makers, in industrial and middle-income developing countries alike, relates to the roles of monetary policy and macro- prudential rules in mitigating procyclicality and promoting macroeconomic and financial stability. In this paper, we focused the discussion on the arguments for and against attributing an explicit financial stability objective to monetary policy – as a complement, or substitute, to macroprudential policy. This dis- cussion was conducted from the perspective of MICs, where banks continue to dominate the financial system and bank credit plays a critical role both on the supply and demand sides. We also discussed, assuming that a more proac- tive role is desirable, what monetary policy should react to, and to what extent it should be combined with macroprudential regulation and possibly capital controls.
The findings in this paper bear on the broader debate, sparked by the global financial crisis, about the role of monetary policy and macroprudential regu- lation – viewed independently and jointly – in achieving macroeconomic and financial stability in both industrial and developing countries. Our review of the various arguments that have been put forward indicates that, on balance, there may be a good case for monetary policy in MICs to be more proactive and address the time dimension of systemic risk – if only during a transitory period, as more is learnt about the implementation and performance of the new macroprudential rules that are currently being discussed, as part of the Basel III agreement and in other policy circles. In particular, there are robust arguments in favour of monetary policy in MICs reacting to a measure of the private-sector credit gap because of concerns about financial stability. The credit gap acts as an intermediate target, which is relatively easy to calculate (given a reference growth rate or credit-to-GDP ratio) and easier to explain to the public than the more elusive final target of financial stability. By making the policy response contingent on the magnitude of the credit gap itself, the primacy of the macroeconomic stability target in ‘normal’ times would be maintained and credibility problems mitigated. Another important argument for responding to the credit gap is the high degree of uncertainty in these countries about real-time estimates of the output gap.
Nevertheless, our analysis also implies that there is no escape from the fact that monetary policy in MICs needs to be combined with macroprudential regu- lation – because monetary policy cannot, in any event, address the cross-section dimension of systemic risk, and because these countries often face circumstances (such as sudden surges in capital flows) where interest rate policy may have unde- sirable side effects that may be detrimental to both macroeconomic and financial stability.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 221
Pierre-Richard Agénor School of Social Sciences Oxford Road University of Manchester Manchester M13 9PL UK [email protected]
References
Admati, A. R., P. M. DeMarzo, M. F. Hellwig and P. Pfeiderer (2011), ‘Fallacies, Irrel- evant Facts, and Myths in the Discussion of Capital Regulation: Why Bank Equity is Not Expensive’, Unpublished, Stanford University.
Agénor, P.-R. (2012), ‘International Financial Integration: Benefits, Costs, and Policy Challenges’, in H. K. Baker and L. A. Riddick (eds), Survey of International Finance. Oxford: Oxford University Press.
Agénor, P.-R., K. Alper and L. Pereira da Silva (2009), ‘Capital Requirements and Business Cycles with Credit Market Imperfections’, Journal of Macroeconomics, Policy Research Working Paper No. 5151, World Bank, Forthcoming.
Agénor, P.-R., K. Alper and L. Pereira da Silva (2011), ‘Capital Regulation, Monetary Policy and Financial Stability’, International Journal of Central Banking, Working Paper No. 154, Centre for Growth and Business Cycles Research, Forthcoming, .
Agénor, P.-R., K. Alper and L. Pereira da Silva (2012), ‘Sudden Floods, Macroprudential Regulation and Stability in an Open Economy’, Working Paper No. 267, Central Bank of Brazil.
Agénor, P.-R., and L. Pereira da Silva (2010), ‘Reforming International Standards for Bank Capital Requirements: A Perspective from the Developing World’, in S. Kim and M. D. McKenzie (eds), International Banking in the New Era: Post-Crisis Challenges and Opportunities. IFR Vol. No. 11, Bingley: Emerald, pp.191–254.
Agénor, P.-R., and L. Pereira da Silva (2011), ‘Macroprudential Regulation and the Monetary Transmission Mechanism’, Working Paper No. 254, Central Bank of Brazil.
Agénor, P.-R., and L. Pereira da Silva (2012), ‘Cyclical Effects of Bank Capital Re- quirements with Imperfect Credit Markets’, Journal of Financial Stability, 8, 43– 56.
Ahrend, R. (2010), ‘Monetary Ease: A Factor behind Financial Crises? Some Evidence from OECD Countries’, Economics 4, 2010–12.
Ariyoshi, A., et al. (2000), Capital Controls: Country Experiences with their Use and Liberalization, Occasional Paper No. 190, Washington, DC: International Monetary Fund.
C© 2012 Blackwell Publishing Ltd
222 Pierre-Richard Agénor and Luiz A. Pereira da Silva
Basel Committee on Banking Supervision (2010), ‘Basel III: A Global Regulatory Framework for more Resilient Banks and Banking Systems’, Report No. 189, Basel Committee on Banking Supervision.
Bean, C., M. Paustian, A. Penalver and T. Taylor (2010), ‘Monetary Policy after the Fall’, Unpublished, Bank of England.
Bernanke, B. S. (2010), ‘Monetary Policy and the Housing Bubble’, Speech to the American Economic Association, Atlanta, Georgia.
Blinder, A. S. (2010), ‘How Central should the Central Bank Be?’ Journal of Economic Perspectives, 48, 123–33.
Brunnermeier, M., A. Crockett, C. Goodhart, A. D. Persaud and H. Shin (2009), The Fundamental Principles of Financial Regulation. Geneva: International Center for Monetary and Banking Studies.
Calderón, C., and J. R. Fuentes (2011), ‘Characterizing the Business Cycles of Emerging Economies’, Unpublished, World Bank.
Claessens, S., M. A. Kose and M. E. Terrones (2011), ‘How do Business and Financial Cycles Interact?’, Working Paper No. 11/88, International Monetary Fund.
Committee on the Global Financial System (2009), ‘Capital Flows and Emerging Market Economies’, CGFS Working Paper No. 33, Bank for International Settlements.
Committee on the Global Financial System (2010), ‘Macroprudential Instruments and Frameworks: A Stocktaking of Issues and Experiences’, CGFS Paper No. 38, Bank for International Settlements.
Cusinato, R. T., A. Minella and S. da Silva Porto Jr. (2010), ‘Output Gap and GDP in Brazil: A Real-Time Data Analysis’, Working Paper No. 203, Central Bank of Brazil.
Dell’Ariccia, G., and R. Marquez (2006), ‘Lending Booms and Lending Standards’, Journal of Finance, 61, 2511–46.
de Mello, L., and D. Moccero (2011), ‘Monetary Policy and Macroeconomic Stabil- ity in Latin America: The Cases of Brazil, Chile, Colombia and Mexico’, Journal of International Money and Finance, 30, 229–45.
Demirguc-Kunt, A., and E. Detragiache (2005), ‘Cross-Country Empirical Studies of Systemic Bank Distress: A Survey’, National Institute Economic Review, 192, 68–83.
Financial Services Authority (2009), The Turner Review – A Regulatory Response to the Global Banking Crisis. London: FSA.
Financial Stability Board (2011), ‘Macroprudential Policy Tools and Frameworks: Up- date for G20 Finance Ministers and Central Bank Governors’, Unpublished.
Fonseca, A. R., F. González and L. Pereira da Silva (2010), ‘Cyclical Effects of Bank Capital Buffers with Imperfect Credit Markets: International Evidence’, Working Paper No. 216, Central Bank of Brazil.
Forbes, K. J., and F. E. Warnock (2012), ‘Capital Flow Waves: Surges, Stops, Flight and Retrenchment’, Journal of International Economics, Forthcoming.
C© 2012 Blackwell Publishing Ltd
Macroeconomic Stability, Financial Stability, and Monetary Policy Rules 223
Furceri, D., S. Guichard and E. Rusticelli (2011), ‘The Effect of Episodes of Large Capital Inflows on Domestic Credit’, Working Paper No. 864, OECD Economics Department.
Galati, G., and R. Moessner (2011), ‘Macroprudential Policy – A Literature Review’, Working Paper No. 337, Bank for International Settlements.
Gambacorta, L. (2009), ‘Monetary Policy and the Risk-Taking Channel’, BIS Quarterly Review, 43–53.
Gerdesmeier, D., H.-E. Reimers and B. Roffia (2010), ‘Asset Price Misalignments and the Role of Money and Credit’, International Finance, 13, 377–407.
Glindro, E. T., T. Subhanij, J. Szeto and H. Zhu (2008), ‘Determinants of House Prices in Nine Asia-Pacific Economies’, Working Paper No. 263, Bank for International Settlements.
Gochoco-Bautista, M. S., J. Jongwanich and J.-W. Lee (2010), ‘How Effective are Capital Controls in Asia?’ Working Paper No. 224, Asian Development Bank.
Goodhart, C., and B. Hofmann (2008), ‘House Prices, Money, Credit and the Macroe- conomy’, Working Paper No. 888, European Central Bank.
Habermeier, K., A. Kokenyne and C. Baba (2011), ‘The Effectiveness of Capital Con- trols and Prudential Policies in Managing Large Inflows’, Staff Discussion Note No. SDN/11/14, International Monetary Fund.
Hawkins, P. (2006), ‘Financial Access and Financial Stability’, in Central Banks and the Challenge of Development. Basel: Bank for International Settlements, pp. 59–80.
Igan, D., and P. Mishra (2011), ‘Making Friends’, Finance and Development, 48, 27– 29.
International Monetary Fund (2011), ‘Toward Operationalizing Macroprudential Poli- cies: When to Act?’ in Global Financial Stability Report, Chapter 3. Washington, DC: International Monetary Fund, pp. 1–45.
International Monetary Fund (2011a), ‘Recent Experiences in Managing Capital In- flows – Cross-Cutting Themes and Possible Policy Framework’, Unpublished, Strategy, Policy and Review Department.
International Monetary Fund (2011b), ‘Macroprudential Policy: An Organizing Frame- work’, Unpublished, Monetary and Capital Markets Department.
Jongwanich, J. (2010), ‘Capital Flows and Real Exchange Rates in Emerging Asian Countries’, Working Paper No. 210, Asian Development Bank.
Jongwanich, J., M. S. Gochoco-Bautista and J.-W. Lee (2011), ‘When are Capital Con- trols Effective? Evidence from Malaysia and Thailand’, Working Paper No. 251, Asian Development Bank.
McCauley, R. N. (2008), ‘Managing Recent Hot Money Inflows in Asia’, Discussion Paper No. 99, Asian Development Bank Institute.
Mishkin, F. S. (2011), ‘Monetary Policy Strategy: Lessons from the Crisis’, Working Paper No. 16755, National Bureau of Economic Research.
C© 2012 Blackwell Publishing Ltd
224 Pierre-Richard Agénor and Luiz A. Pereira da Silva
Montoro, C., and R. Moreno (2011), ‘The Use of Reserve Requirements as a Policy Instrument in Latin America’, Quarterly Review, Bank for International Settlements, (March) 53–65.
Rajan, R. G. (2005), ‘Has Financial Development Made the World Riskier?’ in The Greenspan Era: Lessons for the Future. Kansas City, MO: Federal Reserve Bank of Kansas, pp. 1–42.
Saurina, J. (2009), ‘Dynamic Provisioning: The Experience of Spain’, Crisis Response Note No. 7, International Finance Corporation.
Svensson, L. E. O. (1997), ‘Inflation Forecast Targeting: Implementing and Monitoring Inflation Targets’, European Economic Review, 41, 1111–46.
Svensson, L. E. O. (2010), ‘Inflation Targeting and Financial Stability’, Keynote Lecture at the CEPR/ESI 14th Annual Conference, Hosted by the Central Bank of Turkey.
Terrier, G., R. Valdés, C. E. Tovar, J. Chan-Lau, C. Fernández-Valdovinos, M. Garcı́a- Escribano, C. Medeiros, M.-K. Tang, M. V. Martin and C. Walker (2011), ‘Policy Instruments to Lean Against the Wind in Latin America’, Working Paper No. 11/159, International Monetary Fund.
Wadhwani, S. (2008), ‘Should Monetary Policy Respond to Asset Price Bubbles? Revis- iting the Debate’, National Institute Economic Review, 206, 25–34.
Wezel, T. (2010), ‘Dynamic Loan Provisions in Uruguay: Properties, Shocks, Absorp- tion Capacity and Simulations Using Alternative Formulas’, Working Paper No. 10/125, International Monetary Fund.
C© 2012 Blackwell Publishing Ltd
Macroprudential and monetary policies Implications for financial.pdf
Journal of Banking & Finance 49 (2014) 326–336
Contents lists available at ScienceDirect
Journal of Banking & Finance
journal homepage: www.elsevier .com/locate / jbf
Macroprudential and monetary policies: Implications for financial stability and welfare
http://dx.doi.org/10.1016/j.jbankfin.2014.02.012 0378-4266/� 2014 Elsevier B.V. All rights reserved.
⇑ Corresponding author. Tel.: +44 (0)1159514768. E-mail addresses: [email protected] (M. Rubio), jose.carrasco@
urjc.es (J.A. Carrasco-Gallego). 1 See, for instance, Abraham et al. (2008) and Duca et al. (2011).
Margarita Rubio a,⇑, José A. Carrasco-Gallego a,b
a University of Nottingham, School of Economics, University Park, Nottingham NG7 2RD, UK b Departamento de Economía Aplicada I, Universidad Rey Juan Carlos, P. de los Artilleros, 28032 Madrid, Spain
a r t i c l e i n f o
Article history: Received 15 July 2013 Accepted 16 February 2014 Available online 6 March 2014
JEL classification: E32 E44 E58
Keywords: Macroprudential Monetary policy Welfare Financial stability Loan-to-value Kaldor–Hicks efficiency
a b s t r a c t
In this paper, we analyze the implications of macroprudential and monetary policies for business cycles, welfare, and financial stability. We consider a dynamic stochastic general equilibrium (DSGE) model with housing and collateral constraints. A macroprudential rule for the loan-to-value ratio (LTV), which responds to credit growth, interacts with a traditional Taylor rule for monetary policy. We compute the optimal parameters of these rules both when monetary and macroprudential policies act in a coor- dinated and in a non-coordinated way. We find that both policies acting together unambiguously improves the stability of the system. In both cases, this interaction is welfare improving for the society, especially in the case of the non-coordinated game. There is though a trade-off between borrowers and savers. However, borrowers can compensate the saver’s welfare loss �a la Kaldor–Hicks to achieve a Par- eto-superior outcome.
� 2014 Elsevier B.V. All rights reserved.
‘‘Normally, however, the policy rate is not the only available tool, and much better instruments are available for achieving and main- taining financial stability. Monetary policy should be the last line of defence of financial stability, not the first line.’’ Svensson (2012)
1. Introduction
The housing sector is key to understand how the recent finan- cial crisis developed and, therefore, crucial for designing recovery and prevention policies. The financial crisis was born in the hous- ing sector, grew in the financial sector and had its final conse- quences in the real sector. Financial innovations made the financial system increasingly complex and interconnected, leading to an expansion of systemic risk, especially through the mortgage market. In this context, when house prices collapsed, micro-pru- dential policies, those dedicated to prevent the risk from each com-
pany, had not managed to avoid the contagion to the real sector, and the crisis spread across the financial system to the real econ- omy. Then, a great recession affected the whole economy, causing a high level of unemployment. Thus, from a policy perspective, tra- ditional measures have not seemed to be sufficient to, first, avoid the crisis and, second, have a fast and effective recovery.
As a result, several institutions have implemented macropru- dential tools in order to explicitly promote the stability of the financial system in a global sense, not just focusing on individual companies. The goal of this kind of regulation is to avoid the trans- mission of financial shocks to the broader economy. Some exam- ples of macroprudential tools are asset-side tools (loan-to-value (LTV) and debt-to-income ratio caps), liquidity-based tools (coun- tercyclical liquidity requirements), or capital-based tools (counter- cyclical capital buffers, sectorial capital requirements or dynamic provisions).
The LTV requirement is a limit on the value of a loan relative to the underlying collateral (e.g. residential property). Several studies have pointed out that higher LTV ratios combined with higher risk mortgages contributed to the mortgage crisis.1 The LTV is nowa-
M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336 327
days described as one of the main macroprudential instruments to ‘‘mitigate and prevent excessive credit growth and leverage’’ by the European Systemic Risk Board.2 Within the EU, LTV limits are available in the national prudential framework of 16 Member States.3
The aim of this paper is to evaluate the implications of a macro- prudential LTV tool for business cycles, financial stability, and wel- fare, as well as its interaction with monetary policy. In order to do that, we use a dynamic stochastic general equilibrium (DSGE) model which features a housing market.
The modelling framework consists of an economy composed of borrowers and savers. In particular, our model imposes a limit on borrowing, that is, loans need to be collateralized by a proportion of the value of the assets that the borrower owns. This proportion can be interpreted as an LTV. The macroprudential tool we propose is a rule that automatically reduces loan-to-values when there is a credit boom, therefore limiting the expansion of credit. We assume that there exists a macroprudential Taylor-type rule for the LTV ra- tio, so that it responds to credit growth, in the spirit of the Basel III regulation which aims at avoiding episodes of excessive credit growth. The monetary policy literature has extensively shown that simple rules result in a good performance; therefore, it seems sen- sible to apply these kinds of rules to macroprudential supervision. This microfounded general equilibrium model allows us to explore all the interrelations that appear between the real economy and the credit market. Furthermore, such a model can deal with wel- fare-related issues.
In the context of this model, we address several research ques- tions. First, we study the welfare gain for each agent and for the aggregate both for different levels of a static LTV and for different values of the reaction parameters of the macroprudential rule. In this way, we discuss the welfare trade-offs that may appear be- tween borrowers and savers. Second, we analyze the combination of monetary and macroprudential policy parameters that maxi- mize welfare when the macroprudential regulator and the central bank are coordinated and when they are not. Third, we discuss a Pareto-superior outcome to overcome this trade-off by a system of transfers �a la Kaldor–Hicks. Then, we study the dynamics of the model under the optimal parameters. Finally, we graphically convey our results to highlight the effects on macroeconomic and financial stability of introducing a new macroprudential policy based on the LTV ratio.
The rest of the paper continues as follows: Section 1.1 reviews the literature. Section 2 describes the model. Section 3 presents the welfare analysis. Section 4 computes the optimal parameter combination of the different policies in a coordinated and in a non-coordinated situation. It also develops a rule to obtain a Pare- to-superior outcome, presents results from simulations, and con- veys the results graphically to show the effects of the macroprudential policy on financial and macroeconomic stability. Section 5 concludes.
1.1. Related literature
Our paper fits into the literature that introduces a macropru- dential rule and studies its effects using a DSGE model. Other examples are, for instance, Antipa et al. (2010), who uses a DSGE model to show that macroprudential policies would have been effective in smoothing the past credit cycle and in reducing the intensity of the recession. Another example is Borio and Shim (2007), which emphasizes the complementary role of macropru- dential policy to monetary policy and its supportive function as a
2 See Recommendation of the European Systemic Risk Board (2013). 3 More world results are available in Lim et al. (2011).
4 Borio et al. (2001) also evaluated limits on the LTV. 5 See Borio and Shim (2007) for a distinction between rules and discretion in
calibrating the tools of macroprudential policy. 6 See Galati and Moessner (2013) for an extensive review.
built-in stabilizer. As well, N’Diaye (2009) shows that monetary policy can be supported by countercyclical prudential regulation. Angelini et al. (2012) uses a DSGE model with a banking sector and shows interactions between capital requirement ratios as a macroprudential tool and monetary policy; they find that macro- prudential policies are most helpful to counter financial shocks that lead the credit and asset price booms. We find in our paper that macroprudential policies moderate credit booms. Further- more, for housing demand shocks, the combination of the macro- prudential and the monetary policies manages to control credit without moderating the real effects of the boom.
Since there is an extensive consensus that the origin of the last crisis is related to real estate booms and busts, we have focused on the effects of a macroprudential tool that has to do with the hous- ing sector. However, while most papers in the field tend to analyze macroprudential policies through the lens of a countercyclical bank leverage rule (e.g. Angelini et al., 2012; Christensen et al, 2011), in our paper, we study how a key element of the real estate sector, namely the LTV, can serve as a macroprudential tool to im- prove financial stability.4 With a macroprudential orientation, Kan- nan et al. (2012) also examines a monetary policy rule that reacts to prices, output and changes in collateral values with a macropruden- tial instrument based on the LTV; they remark on the importance of identifying the source of the shock of the housing price boom when assessing policy optimality. Funke and Paetz (2012) considerers a non-linear version of a macroprudential rule for the LTV. Following this literature, we propose a macroprudential policy based on a Tay- lor-type automatic rule.5 By analogy with monetary policy, rule- based macroprudential tools – for example, automatic stabilizers – appear appealing (Goodhart, 2004).
One question that arises from the topic is what the objective of the macroprudential authority should be. In recent years, research on macroprudential issues has been wide and intense6 and there is an increasing consensus among academics and policy makers that ‘‘the ultimate objective of macroprudential policy is to contribute to the safeguard of the stability of the financial system as a whole’’ (Recommendation of the European Systemic Risk Board, 2013). In this way, Almeida et al. (2006) has studied the effect on the ampli- tude of the credit cycle results from the mitigating impact of more stringent LTV ratios on the ‘financial accelerator’ mechanism. They find that when a positive income shock leads to an increase in hous- ing prices, the increase in borrowing is expected to be lower in coun- tries with lower LTV ratios. Gelain et al. (2013) evaluates different policy actions that might be used to dampen the resulting excess volatility, including a direct response to house-price growth or credit growth in the central bank’s interest rate rule, the imposition of a more restrictive loan-to-value ratio, and the use of a modified collat- eral constraint that takes into account the borrower’s wage income. We contribute to this line of research, finding that when we use the macroprudential policy based on the LTV, both the macroeconomy and the financial system become more stable. To illustrate that, we construct policy frontiers (Taylor curves) including not only the tra- ditional objectives of monetary policy but also the objective of the macroprudential regulator: financial stability. As a measure of finan- cial stability we propose the variability of borrowing. This three- dimensional policy frontier shows graphically that the macropru- dential policy unambiguously helps to achieve a more stable finan- cial and macroeconomic situation.
A central issue that we cover in our paper is the interaction be- tween monetary and macroprudential policies. There is no consen- sus on whether both policies should act in a coordinated or in a
328 M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336
non-coordinated way. For instance, Bean et al. (2010), with a DSGE model adapted from Gertler and Karadi (2011), studies how the use of a macroprudential policy tool based on a lump-sum levy or subsidy on the banking sector might affect the conduct of mon- etary policy. Their results suggest that monetary and macropru- dential policies should be coordinated, since they are not merely substitutes, but they mention that the issue of coordination needs to be studied further. Beau et al. (2012) claims that it is preferable to have a combination of separate objectives for monetary and macroprudential policies, with monetary policy taking the macro- economic effects of macroprudential policy into account in choos- ing interest rates, that is, the non-coordinated case would be preferable. Angelini et al. (2012) studies the coordination issue in a context in which the macroprudential regulator uses capital requirements as a tool to achieve financial stability. They find that lack of cooperation between a macroprudential authority and a central bank may actually generate conflicting policies and, there- fore, cooperation is preferred. In our paper, we also distinguish be- tween the cases of coordination and non-coordination to try to shed some light on this issue. As argued by Svensson (2012), we find that the non-coordination game delivers higher social welfare and then is preferable. When each authority focuses on its own objective, they are more effective in minimizing both macroeco- nomic and financial variability.
Finally, measuring the potential welfare improvement of mac- roprudential policies has deserved the special attention of academ- ics. Some papers have found that the macroprudential reaction to exogenous shocks can make some people better off (typically bor- rowers), but not every type of household, or not in all cases. For in- stance, Lambertini et al. (2013) extends the Iacoviello and Neri (2010) model to incorporate news shocks and a macroprudential rule on the LTV. They find that an optimized LTV-ratio rule that re- sponds to credit growth is a Pareto-improving policy compared to the use of a constant LTV ratio. Campbell and Hercowitz (2009) performes a welfare analysis in a DSGE model with borrowers and savers and determines that, although high LTV ratios have a di- rect positive effect on welfare through constraint relaxation, other indirect effects may dominate. Angelini et al. (2012) also discusses the issue and concludes that there is no regime that makes all agents better-off. They claim that the optimal (from a welfare per- spective) monetary and macroprudential policies may depend on which agent’s welfare is used as objective in the computation of the policies, and also on the type of shock considered. In our paper, we actively contribute to this discussion. We focus on highlighting the welfare trade-offs between agents in order to carefully charac- terize the conditions under which there is room for Pareto improvements. By analyzing welfare for a static LTV, we find an LTV threshold below which there is room for Pareto-improving solutions. However, for higher values the trade-off between bor- rowers and savers appears. Since a plausible value for the LTV tends to be higher than this value, we also observe this trade-off when calculating the optimal macroprudential rule. Thus, we pro- pose a system of transfers �a la Kaldor–Hicks in which borrowers would compensate savers so that they are indifferent between hav- ing the macroprudential policy or not. In this way, we obtain a Par- eto-superior outcome.7
2. Model setup
The modelling framework is a DSGE model with a housing mar- ket, following Iacoviello (2005). The model is solved by log-linear-
7 This is the first time that this criterion has been applied in the macroprudential context, albeit it is widely used in regulatory analysis in Law and Economics. See for instance Posner (2007).
izing the equilibrium equations around a well-defined steady state. The use of DSGE models for the study of macroprudential policies has some limitations and deserves some discussion. When using DSGE models for monetary policy evaluation, the dynamics of the model are matched with the monetary policy transmission mechanism found in the data. However, for macroprudential poli- cies, empirical applications are rare. Furthermore, the macropru- dential analysis often refers to the vulnerability of the financial system to exceptional events related to non-equilibrium, which cannot be captured by a DSGE model. At the same time, a drawback of DSGE models is that they are infinite horizon models and, there- fore, are not well suited to incorporate state contingency in a meaningful way. As a result, DSGE models have problems of mod- elling financial intermediation and frictions (Bean, 2009). However, regardless of these limitations, DSGE models are often used for macroprudential analysis since they count with other advantages: First, they can be compared with a benchmark in which there is only monetary policy. Second, they include many sources of shocks that can be used to check for different economic trajectories. More- over, they rely on general equilibrium analysis and are suitable for simulations to study the impact of new policy instruments. Also, calibrated parameters can be altered to test for alternative policy scenarios. And finally, since DSGE models are microfounded, they are suitable to study welfare issues.8
In our model, the economy features patient and impatient households, a final goods firm, and a central bank which conducts monetary policy. Households work and consume both consump- tion goods and housing. Patient and impatient households are sav- ers and borrowers, respectively. Borrowers are credit constrained and need collateral to obtain loans. The representative firm con- verts household labor into the final good. The central bank follows a Taylor rule for the setting of interest rates. The macroprudential authority sets the LTV following a Taylor-type rule.
2.1. Savers
Savers maximize their utility function by choosing consump- tion, housing and labor hours:
max Cs;t ;Hs;t ;Ns;t
E0
X1 t¼0
bt s log Cs;t þ jt log Hs;t �
ðNs;tÞg
g
� � ;
where bs 2 ð0;1Þ is the patient discount factor, E0 is the expectation operator and Cs;t; Hs;t and Ns;t represent consumption at time t, the housing stock and working hours, respectively. 1=ðg� 1Þ is the la- bor supply elasticity, g > 0. jt represents the weight of housing in the utility function. We assume that log ðjtÞ ¼ log ðjÞ þ uJt , where uJt follows an autoregressive process. A shock to jt represents a shock to the marginal utility of housing.
Subject to the budget constraint:
Cs;t þ bt þ qtðHs;t � Hs;t�1Þ ¼ Rt�1bt�1
pt þws;tNs;t þ Ft; ð1Þ
where bt denotes bank deposits, Rt is the gross return from deposits, qt is the price of housing in units of consumption, and ws;t is the real wage rate. Ft are lump-sum profits received from the firms. The first order conditions for this optimization problem are as follows:
1 Cs;t ¼ bsEt
Rt
ptþ1Cs;tþ1
� � ; ð2Þ
ws t ¼ ðNs;tÞg�1Cs;t; ð3Þ jt
Hs;t ¼ 1
Cs;t qt � bsEt
1 Cs;tþ1
qtþ1: ð4Þ
8 See Brázdik et al. (2012) for further discussion.
M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336 329
Eq. (2) is the Euler equation, the intertemporal condition for con- sumption. Eq. (4) represents the intertemporal condition for housing, in which, at the margin, benefits for consuming housing equate costs in terms of consumption. Eq. (3) is the labor-supply condition.
2.2. Borrowers
Borrowers solve:
max Cb;t ;Hb;t ;Nb;t
E0
X1 t¼0
bt b log Cb;t þ jt log Hb;t �
ðNb;tÞg
g
� � ;
where bb 2 ð0;1Þ is the impatient discount factor, subject to the budget constraint and the collateral constraint:
Cb;t þ Rt�1bt�1
pt þ qtðHb;t � Hb;t�1Þ ¼ bt þWb;tNb;t; ð5Þ
Et Rt
ptþ1 bt ¼ ktEtqtþ1Hb;t ; ð6Þ
where bt denotes bank loans and Rt is the gross interest rate. kt can be interpreted as a loan-to-value ratio. The borrowing constraint limits borrowing to the present discounted value of their housing holdings. The first order conditions are as follows:
1 Cb;t ¼ bbEt
Rt
ptþ1Cb;tþ1
� � þ ktRt ; ð7Þ
wb;t ¼ ðNb;tÞg�1Cb;t; ð8Þ jt
Hb;t ¼ 1
Cb;t qt � bbEt
1 Cb;tþ1
qtþ1
� � � ktktEtðqtþ1ptþ1Þ; ð9Þ
where kt denotes the multiplier on the borrowing constraint.9 These first order conditions can be interpreted analogously to those of savers.
2.3. Firms
2.3.1. Final goods producers There is a continuum of identical final goods producers that
operate under perfect competition and flexible prices. They aggre- gate intermediate goods according to the production function
Yt ¼ Z 1
0 YtðzÞ
e�1 e dz
� � e e�1
; ð10Þ
where e > 1 is the elasticity of substitution between intermediate goods. The final good firm chooses YtðzÞ to minimize its costs, resulting in demand of intermediate good z:
YtðzÞ ¼ PtðzÞ
Pt
� ��e
Yt : ð11Þ
The price index is then given by:
Pt ¼ Z 1
0 PtðzÞ1�e dz
� � 1 e�1
: ð12Þ
2.3.2. Intermediate goods producers The intermediate goods market is monopolistically competitive.
Following Iacoviello (2005), intermediate goods are produced according to the production function:
YtðzÞ ¼ AtNs;tðzÞaNb;tðzÞð1�aÞ ; ð13Þ
where a 2 ½0;1�measures the relative size of each group in terms of labor.10 This Cobb–Douglas production function implies that labor
9 Through simple algebra it can be shown that the Lagrange multiplier is positive in the steady state and thus the collateral constraint holds with equality.
10 Notice that the absolute size of each group is one.
11 It could also be interpreted as the savers being older than the borrowers therefore more experienced.
12 Symmetry across firms allows us to write the demands without the index z. 13 Variables with a hat denote percent deviations from the steady state.
efforts of constrained and unconstrained consumers are not perfect substitutes. This specification is analytically tractable and allows for closed form solutions for the steady state of the model. This assumption can be economically justified by the fact that savers are the managers of the firms and their wage is higher than that of the borrowers.11
At represents technology and it follows the following autore- gressive process:
log ðAtÞ ¼ qA log ðAt�1Þ þ uAt; ð14Þ
where qA is the autoregressive coefficient and uAt is a normally dis- tributed shock to technology. We normalize the steady-state value of technology to 1.
Labor demand is determined by:
ws;t ¼ 1 Xt
a Yt
Ns;t ; ð15Þ
wb;t ¼ 1 Xt ð1� aÞ Yt
Nb;t ; ð16Þ
where Xt is the markup, or the inverse of marginal cost.12
The price-setting problem for the intermediate good producers is a standard Calvo-Yun setting. An intermediate good producer sells its good at price PtðzÞ, and 1� h;2 ½0;1�, is the probability of being able to change the sale price in every period. The optimal re- set price P�t ðzÞ solves:
X1 k¼0
ðhbÞkEt Kt;k P�t ðzÞ Ptþk
� e=ðe� 1Þ Xtþk
� � Y�tþkðzÞ
� � ¼ 0; ð17Þ
where e=ðe� 1Þ is the steady-state markup. The aggregate price level is then given by:
Pt ¼ hP1�e t�1 þ ð1� hÞ P�t
� 1�e h i1=ð1�eÞ
: ð18Þ
Using (17) and (18), and log-linearizing, we can obtain a stan- dard forward-looking New Keynesian Phillips curve p̂t ¼ bEtp̂tþ1 � wx̂t þ upt , that relates inflation positively to future inflation and negatively to the markup ðw � ð1� hÞð1� bhÞ=hÞ. upt
is a normally distributed cost-push shock.13
2.4. Monetary policy
We consider a Taylor rule which responds to inflation and out- put growth:
Rt ¼ ðRt�1ÞqððptÞð1þ/R pÞðYt=Yt�1Þ/
R y RÞ
1�q eRt; ð19Þ
where 0 6 q � 1 is the parameter associated with interest-rate inertia, /R
p P 0 and /R y P 0 measure the response of interest rates
to current inflation and output growth, respectively. eRt is a white noise shock with zero mean and variance r2
e .
2.5. A macroprudential rule for the LTV
In standard models, the LTV ratio is a fixed parameter which is not affected by economic conditions. However, we can think of reg- ulations of LTV ratios as a way to moderate credit booms. When the LTV ratio is high, the collateral constraint is less tight. And, since the constraint is binding, borrowers will borrow as much as they are allowed to. Lowering the LTV tightens the constraint and therefore restricts the loans that borrowers can obtain. Recent
,
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 -1
-0.5
0
0.5
1
1.5
2
2.5
3
3.5
4
LTV
W el
fa re
g ai
n (C
E)
Total Savers Borrowers
Fig. 1. Welfare gains from increasing the LTV ratio, everything else constant. Benchmark case: no macroprudential regulator.
330 M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336
research on macroprudential policies has proposed Taylor-type rules for the LTV ratio so that it reacts inversely to variables such as the growth rates of GDP, credits, the credit-to-GDP ratio or house prices. These rules can be a simple illustration of how a mac- roprudential policy could work in practice. Here, we assume that there exists a macroprudential Taylor-type rule for the LTV ratio, so that it responds to credit growth, in the spirit of the Basel III reg- ulation which aims at avoiding episodes of excessive credit growth14:
kt ¼ kSS bt
bt�1
� ��/k b
; ð20Þ
where kSS is a steady state value for the loan-to-value ratio, and /k
b P 0 measures the response of the loan-to-to value to the credit growth. This kind of rule would deliver a lower LTV ratio in booms, when there is excessive credit growth, therefore restricting the credit in the economy and avoiding a credit boom derived from good economic conditions (and symmetrically for recessions).15
2.6. Equilibrium
The market clearing conditions are as follows:
Yt ¼ Cs;t þ Cb;t : ð21Þ
The total supply of housing is fixed and it is normalized to unity:
Hs;t þ Hb;t ¼ 1: ð22Þ
3. Welfare
3.1. Welfare measure
To assess the normative implications of macroprudential and monetary policies, we numerically evaluate the welfare derived in each case. As discussed in Benigno and Woodford (2008), the two approaches that have recently been used for welfare analysis in DSGE models include either characterizing the optimal Ramsey policy, or solving the model using a second-order approximation to the structural equations for given policy and then evaluating welfare using this solution. As in Mendicino and Pescatori (2007), we take this latter approach to be able to evaluate the welfare of the two types of agents separately.16 The individual welfare for sav- ers and borrowers, respectively, as follows:
Ws;t � Et
X1 m¼0
bm s log Cs;tþm þ j log Hs;tþm �
ðNs;tþmÞg
g
� � ; ð23Þ
Wb;t � Et
X1 m¼0
bm b log Cb;tþm þ j log Hb;tþm �
ðNb;tþmÞg
g
� � : ð24Þ
Following Mendicino and Pescatori (2007), we define social welfare as a weighted sum of the individual welfare for the differ- ent types of households:
Wt ¼ ð1� bsÞWs;t þ ð1� bbÞWb;t : ð25Þ
14 See Kannan et al. (2012) for a similar specification. 15 The feasibility of implementing an LTV rule at a quarterly frequency may be
questionable in practice. However, as the Committee on the Global Financial System (2012) suggests, once the legal and operational infrastructure is in place, LTV changes can be implemented rather rapidly, given that many jurisdictions have ample experience with these tools at the practical level.
16 We used the software Dynare to obtain a solution for the equilibrium implied by a given policy by solving a second-order approximation to the constraints, then evaluating welfare under the policy using this approximate solution, as in Schmitt- Grohe and Uribe (2004). See Monacelli (2006) for an example of the Ramsey approach in a model with heterogeneous consumers.
Each agent’s welfare is weighted by her discount factor, respec- tively, so that all the groups receive the same level of utility from a constant consumption stream.
However, in order to make the results more intuitive, we pres- ent welfare changes in terms of consumption equivalents. The con- sumption equivalent measure defines the constant fraction of consumption that households should give away in order to obtain the benefits of the macroprudential policy. A positive value means a welfare gain, that is, how much the consumer would be willing to pay to obtain the welfare improvement. Then, when there is a wel- fare gain, households would be willing to pay in consumption units for the measure to be implemented because it is welfare improv- ing. We use as a benchmark the welfare evaluated when the mac- roprudential policy is not active and compare it with the welfare obtained when such policy is implemented. The derivation of the welfare benefits in terms of consumption equivalent units is as follows:
CEs ¼ exp ð1� bsÞ WMP s �W�
s
�h i � 1; ð26Þ
CEb ¼ exp ð1� bbÞ WMP b �W�
b
�h i � 1; ð27Þ
where the superscripts in the welfare values denote the benchmark case when macroprudential policies are not introduced and the case in which they are, respectively.17
3.2. Welfare trade-offs
The literature typically finds that the macroprudential reaction to exogenous shocks can make some people better off (typically borrowers), but not every type of household, or not in all cases. This is why, welfare comparisons should not only be made on the basis of an ad hoc aggregate welfare function but disaggregat- ing welfare between agents, to highlight the trade-offs that may appear between them.
In this section, we first compute welfare for each individual and for the aggregate, when we have a static LTV. Then, we numerically evaluate welfare gains when we introduce a macroprudential rule, given the Taylor rule.
Fig. 1 presents welfare gains, in consumption equivalents, for different values of the LTV, when there is no macroprudential rule in place. Here, we observe that up to a threshold LTV value, there is
7 We follow Ascari and Ropele (2009).
1
0 0.5 1 1.5 2 2.5 3 3.5 -0.5
0
0.5
1
1.5
2
2.5
phib
W el
fa re
g ai
n (C
E)
Total Savers Borrowers
Fig. 2. Welfare gains from introducing the macroprudential rule, given monetary policy (different values of the reaction parameter for borrowing).
M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336 331
room for Pareto optimal policies. However, starting from a value of 0.55, there is a trade-off between borrowers and savers in terms of welfare when we keep increasing the LTV. Large values of the LTV harm borrowers while savers benefit from the increase. Social wel- fare decreases. This result is in line with that of Campbell and Hercowitz (2009), who performes a welfare analysis in a DSGE model with borrowers and savers and determined that although high LTV ratios have a direct positive effect on welfare through constraint relaxation, other indirect effects may dominate. Notice that k, the LTV ratio, is a parameter that strongly affects the collat- eral constraint. A small change in this parameter can cause very large changes in borrowing that can be excessive. Higher LTVs lead to higher consumption levels, because borrowing constraints are always binding: the more borrowers are offered, the more they take. But this in turn, as shown in Campbell and Hercowitz (2009), changes relative prices. In particular, higher consumption levels imply higher interest rates. This could lead to a situation of overindebtedness in the sense that high repayments could offset the positive effects on constraint relaxation. Then, higher interest rates imply higher returns on saving for savers. Smith (2009) shows that these results do not rely on the specific assumptions of Campbell and Hercowitz (2009); even in the simplest model with borrowers, savers, and collateral constraints, this effect takes place.18
Fig. 2 shows the welfare gains from introducing a macropruden- tial tool in the economy, given the Taylor rule. We use a steady- state value of the LTV of 0.9, as in Iacoviello (2005) and Iacoviello (2013). Therefore, we are in a region in which trade-offs should ap- pear. Leaving fixed monetary policy, we present welfare for a con- tinuum of values of the reaction parameters in the LTV rule, from a less to a more aggressive rule. The figure is very informative be- cause it shows welfare gains for each agent in the economy and for the aggregate. The conclusions we can obtain from the figure are the following: using both policy measures at the same time is unambiguously welfare enhancing, as we can observe from the solid line. We can see that welfare increases by more, the larger the response of the LTV to credit growth is, but up to a point at which welfare stops increasing. The figure also shows the trade- off between borrowers’ and savers’ welfare, illustrated by the dif-
18 Huggett (1997) also found a similar result, but in this case is the reduction in the precautionary motive for saving, driven by the looser borrowing constraints, wha leads to the increase in the interest rate.
19 Andres et al. (2013) find that optimal monetary policy may involve a trade-of between the stabilization of inflation, output gap, consumption gap and the distribution of the collateral asset between constrained and unconstrained
t
ference between the two dashed lines. Borrowers’ welfare in- creases with the introduction of the macroprudential rule because tightening the collateral constraint avoids situations of overindebt- edness in which debt repayments are a burden for them. Further- more, borrowers can benefit from more financial stability in the economy, as we will show later on. Notice that borrowers have a collateral constraint which is always binding and this does not al- low them to make consumption smoothing. They do not have an Euler equation to smooth consumption as savers do. A more stable financial system smooths their consumption path thus mitigating the negative effects of the collateral constraint. This welfare gain is at the expense of savers, who lose from having this measure in the economy, given that they are not financially constrained. How- ever, the borrower’s welfare gain compensates the loss of the sav- ers and globally, the measure is welfare increasing.
The next section performs an optimal policy analysis in order to assess which are the combination of values of the reaction param- eters which would maximize welfare and make policy recommen- dations on this issue.
4. Optimal policy analysis
4.1. Optimal parameters
In this section, we aim to find the optimal combination of policy parameters that maximizes welfare. For this purpose, we consider three different cases. A benchmark case in which there is only a monetary authority that acts in the traditional way, using the interest rate as an instrument. Then, we include a macroprudential authority that introduces an extra instrument, the LTV ratio. We study the interaction between the two authorities from two per- spectives, when they act both in a coordinated and in a non-coor- dinated way.
The optimal policy analysis, in models with financial frictions, deserves some discussion. In the standard new Keynesian model, the central bank aims at minimizing the variability of output and inflation to reduce the distortion introduced by nominal rigidities and monopolistic competition. However, in models with collateral constraints, welfare analysis and the design of optimal policies in- volves a number of issues not considered in standard sticky-price models. In models with constrained individuals, there are two types of distortions: price rigidities and credit frictions. This cre- ates conflicts and trade-offs between borrowers and savers. Savers may prefer policies that reduce the price stickiness distortion. However, borrowers may prefer a scenario in which the pervasive effect of the collateral constraint is softened. Borrowers operate in a second-best situation. They consume according to the borrowing constraint as opposed to savers that follow an Euler equation for consumption. Borrowers cannot smooth consumption by them- selves, but a more stable financial system would provide them a setting in which their consumption pattern is smoother. Therefore, in order to assess the optimality of policies, factors that help bor- rowers smooth their consumption should be included. Studies show that, in these kind of models, financial variables should be in- cluded in the loss function that the policy maker aims at minimizing.19
In the standard sticky-price model, the Taylor rule of the central bank is consistent with a loss function that includes the variability of inflation and output. In order to rationalize the Taylor rule of the macroprudential regulator, we follow Angelini et al. (2012) in which they assume that the loss function in the economy also con-
consumers.
f
Table 1 Optimal macroprudential and monetary policy mix.
Benchmark Coordinated Non-coordinated
/k� b
– 0.8 0.7
1þ /R� p
16.1 1.3 1.7
/R� y
8.2 0.5 1
Social Welfare Gain – 0.024 0.041 Borrowers Welfare Gain – 0.22 0.31 Savers Welfare Gain – �0.16 �0.21 r2
b 1.4308 1.1861 1.1767
r2 p 0.2183 0.3481 0.3025
r2 y 1.9113 1.7877 1.8087
2 Notice that here, we are considering that the central bank acts in a traditional
332 M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336
tains financial variables, namely borrowing variability, as a proxy for financial stability. Then, there would be a loss function for the economy that would include not only the variability of output and inflation but also the variability of borrowing: L ¼ r2
p þ kyr2 y þ r2
b where r2 p;r2
y and r2 b are the variances of infla-
tion, output and borrowing. ky P 0, represents the relative weight of the central bank to the stabilization of output.20
If the central bank and the macroprudential regulator coordi- nate, they would aim at jointly minimizing the loss function each one with its own instrument. The problem becomes analogous to the Mundell’s assignment rule in which each arm of policy concen- trates on a single task, addressing the issue it cares most about, and making coordination of policy trivial.21 Following this line of argu- ment, we consider a case in which we jointly optimize the parame- ters of both rules.
However, Svensson (2012) argues that conducting monetary policy and financial stability policy in an integrated way may be inappropriate, since monetary policy and financial-stability policy are distinct and separate policies with different objectives and dif- ferent instruments. Tinbergen (1952) put forth what we now call the ‘Tinbergen principle,’ that policymakers need at least one inde- pendent policy instrument for each policy objective. Since the pol- icy interest rate is used by monetary policymakers to achieve the objective of price stability, at least one other instrument is required to achieve the additional objective of financial stability of macro- prudential policy. Svensson (2012) suggests that monetary policy should be in charge of price stability while macroprudential policy needs to address financial stability. He argues that monetary policy should be conducted taking the macroprudential policy into ac- count, and vice versa, as in a Nash equilibrium rather than a coor- dinated equilibrium. Therefore, we study a second case in which the central bank and the macroprudential regulator play a non- coordinated game. The central bank would find the optimal param- eters in its policy rule, taking the macroprudential regulator behav- ior as given. Similarly, the macroprudential authority would find the best response given monetary policy. The intersection of these two best responses would give us the Nash equilibrium.
In order to contribute to the discussion and evaluate the welfare gains of introducing macroprudential polices, we first compute the optimal parameters of the Taylor rule for monetary policy, assum- ing that there is no macroprudential regulator. Then, we compute the optimal monetary and macroprudential policies for the coordi- nated and the non-coordinated game.
Table 1 shows the optimal parameter values and the welfare gains in consumption equivalents, taking as a benchmark the situ- ation without macroprudential policy. We also present the implied volatilities.
20 This loss function would be consistent with studies that make a second-order approximation of the utility of individuals and find that it differs from the standard case by including financial variables.
21 See Mundell (1962).
As expected, when a macroprudential regulator does not exist, the central bank needs to act in a very aggressive way, given that it only counts with a single instrument to minimize the loss func- tion.22 We take this case as a benchmark, both for welfare and for macroeconomic and financial volatilities (presented in the first column).
The second column presents the case in which there is a macro- prudential regulator that acts in a coordinated way with the cen- tral bank. We see that adding this extra instrument produces a welfare gain in the economy. In this case, monetary policy does not need to be as aggressive as in the benchmark case because it counts with the help of the macroprudential policy. However, as al- ready pointed out, there is a trade-off between borrowers and sav- ers and, while borrowers are better-off, savers are not. Furthermore, if we compare the volatilities that this combination of policies generates, with respect to the benchmark case, we ob- serve that the standard deviation of borrowing decreases, which is what makes borrower’s welfare increase. In terms of the macro- economic volatilities, we see that the volatility of output decreases, but this comes at the expense of a higher inflation volatility.23 This higher inflation volatility contributes to a decrease in savers’ welfare.
Nevertheless, if both authorities act in a non-coordinated way, social welfare gains are even higher. As Svensson (2012) argues, letting each regulator focus on its own objective, leads to more effective results in reducing volatilities. In this case, monetary pol- icy acts in a more aggressive way, favoring the reduction of the vol- atility of inflation. The macroprudential authority reaction parameter does not need to be as high as in the previous case to obtain a lower standard deviation of borrowing. As usual, we also observe the same trade-off between borrowers and savers.
4.2. Pareto-superior outcomes
Results from optimal policy analysis show that trade-offs be- tween the two agents appear. However, if the welfare gain that borrowers obtain is large enough, there could be room for Pare- to-superior outcomes.
In order to do that, we apply the concept of Kaldor–Hicks effi- ciency, also known as Kaldor–Hicks criterion.24 Under this criterion, an outcome is considered more efficient if a Pareto-superior out- come can be reached by arranging sufficient compensation from those that are made better off to those that are made worse off so that all would end up no worse off than before. The Kaldor–Hicks cri- terion does not require the compensation actually being paid, merely that the possibility for compensation exists, and thus need not leave each at least as well off.
In our case, our measure for welfare presented in consumption equivalents is given by Eqs. (26) and (27). Since, there is a trade-off between savers and borrowers, introducing the macroprudential policy, both in coordination and non-coordination with monetary policy, produces CEb > 0 and CEs < 0.
Thus, a Kaldor–Hicks improvement to a obtain Pareto-superior outcome would be one in which:
CEb � eb P 0
and
CEs þ eb ¼ 0:
ay, we are excluding the possibility that financial variables enter in the Taylor rule r the central bank. For further discussion on interactions between different rules, e Kannan et al. (2012) or Rubio and Carrasco-Galllego (2013). 3 This result is consistent with other studies on macroprudential policies. See for stance, Mendicino et al. (2013). 4 See Scitovsky (1941).
2
w fo se
2
in 2
Table 2 Optimal macroprudential and monetary policy mix (Kaldor–Hicks improvement).
Benchmark Coordinated Non-coordinated
/k� b
– 0.8 0.7
1þ /R� p
16.1 1.3 1.7
/R� y
8.2 0.5 1
Social Welfare Gain – 0.024 0.041 Borrowers Welfare Gain – 0.06 0.10 Savers Welfare Gain – 0 0
Table 3 Parameter values.
bs :99 Discount factor for savers bb :98 Discount factor for borrowers j :1 Weight of housing in utility function g 2 Parameter associated with labor elasticity k :9 Loan-to-value ratio a :64 Labor share for savers X 1:2 Steady-state markup h :75 Probability of not changing prices qA :9 Technology persistence qj :95 Housing demand shock persistence q :8 Interest-rate-smoothing parameter in Taylor rule
M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336 333
Then,
eb P 1� exp ð1� bsÞ WMP s �W�
s
�h i : ð28Þ
That is, a system of transfers in which the borrowers would compensate the savers with at least the amount they are losing, so that they are at least indifferent between having the macropru- dential policy or not. Then, the new outcome would be desirable for society and there would be no agent that would lose with the introduction of the new policy. Then, if Eq. (28) holds with equal- ity, the borrower compensates the saver with the exact welfare that she is losing. Then, in our case, after the compensations are made, the final result is presented in Table 2.
4.3. Impulse responses
In order to understand the dynamics of the model and how the LTV rule interacts with monetary policy, in this section, we simu- late the impulse responses of the model, using the optimized parameters we found in the previous section. We compare the benchmark (no macroprudential policy) with the case in which monetary and macroprudential policies coexist, both in a coordi- nated and in a non-coordinated game. We consider a technology shock and a housing demand shock.
The discount factor for savers, bs, is set to 0.99 so that the an- nual interest rate is 4% in steady state. The discount factor for the borrowers is set to 0.98.25 The steady-state weight of housing in the utility function, j, is set to 0.1 in order for the ratio of housing wealth to GDP to be approximately 1.40 in the steady state, consis- tent with the US data. We set g ¼ 2, implying a value of the labor supply elasticity of 1.26 For the parameters controlling leverage, we set kSS to 0.90, in line with the US data.27 The labor income share for savers is set to 0.64, following the estimate in Iacoviello (2005). For the Taylor rule, we consider the optimized parameters found in the previous section. For q we use 0.8, which also reflects a real- istic degree of interest rate smoothing.28
We assume that technology, At , follows an autoregressive pro- cess with 0.9 persistence and a normally distributed shock. We also assume that the weight of housing on the utility function is equal to its value in the steady state plus a shock which follows an auto- regressive process with 0.95 persistence.29 For the reactions param- eter in the LTV rule, we use the optimized parameters both for the coordination and non-coordination with monetary policy. Table 3 presents a summary of the parameter values used:
4.3.1. Technology shock Fig. 3 presents impulse responses to a 1% shock to technology.
Given the technology shock, output increases and inflation decreases.
In the benchmark case, when there is only monetary policy, the interest rate increases, while the LTV remains at its steady state. Since output is increasing, monetary policy reacts in a contractive way. Given the expansion in the economy, borrowing and housing demand increase, leading to an increase in house prices.
However, when the macroprudential rule interacts with mone- tary policy, the reaction of the interest rate is not as strong, given that the optimal parameters of the Taylor rule are lower. The LTV
25 Lawrance (1991) estimated discount factors for poor consumers at between 0.95 and 0.98 at quarterly frequency. We take the most conservative value.
26 Microeconomic estimates usually suggest values in the range of 0 and 0.5 (fo males). Domeij and Flodén (2006) show that in the presence of borrowing constraints these estimates could have a downward bias of 50%.
27 See Iacoviello (2013). 28 As in McCallum (2001). 29 The persistence of the shocks is consistent with the estimates in Iacoviello and
Neri (2010).
r
ratio decreases to cut credit because borrowing is growing follow- ing the boom. Therefore, when the macroprudential rule is in place, borrowing does not increase as much as in the benchmark, and this mitigates the effects of the boom.
Concerning the difference between the coordinated and the non-coordinated case, the pattern of the impulse responses is very similar. Nevertheless, the non-coordinated case is always slightly closer to the benchmark. This is due to the fact that the reaction parameters in the Taylor rule are higher for the coordinated case and thus more similar to the benchmark.
Notice that, interestingly, when monetary and macroprudential policies coexist, the interest rate decreases, focusing on stabilizing inflation, while the LTV is cut, to reach the financial stability objec- tive. The decrease in the interest rate contributes to increase bor- rowing while the decrease in the LTV cuts it.
4.3.2. Housing demand shock In Fig. 4, we see the effects of a 25% housing demand shock. Gi-
ven the increase in demand, house prices increase as well. This di- rectly affects the collateral constraint and borrowers are able to borrow more out of their housing collateral, which is worth more now. The wealth effect permits them consume both more houses and consumption goods. The increase in house prices is, therefore, transmitted to the real economy and output increases.
The raise in output generates inflation and the Taylor rule re- sponds with a higher interest rate. This is particularly true in the benchmark case in which monetary policy is more aggressive. On impact, this higher interest rate also dampens the increase in the price of the house, especially for the benchmark for the same rea- sons. Therefore, the initial shock is mitigated in the case in which monetary is the only policy in action.
When the macroprudential and the monetary policy interact, the LTV decreases to moderate the credit boom. This is the reason why, in this case, borrowing does not increase as much as in the benchmark. However, as we have seen, the increase in the interest rate is not as strong as in the benchmark and therefore the effects on real output of this demand shock are more noticeable. In the case of this shock, the combination of the macroprudential and the monetary policies manage to control credit without moderat- ing the real effects of the boom.
0 5 10 0
0.5
1 Output
% de
v. S
S 0 5 10
0
0.5
1 Borrowing
0 5 10 −0.5
0
0.5
Inflation %
de v.
S S
0 5 10 0
0.5
1
House Prices
0 5 10 −0.2
0
0.2
Interest Rate
quarters
% de
v. S
S
0 5 10 −0.2
0
0.2
LTV
quarters
Benchmark Coordinated Macropru Non−coord Macropru
Fig. 3. Impulse responses to a technology shock. Optimized parameters.
0 5 10 0
0.05
0.1 Output
% de
v. S
S
0 5 10 0
2
4 Borrowing
0 5 10 −0.1
0
0.1 Inflation
% de
v. S
S
0 5 10 0.2
0.3
0.4 House Prices
0 5 10 0
0.05
0.1 Interest Rate
quarters
% de
v. S
S
0 5 10 −0.5
0
0.5 LTV
quarters
Benchmark Coordinated Macropru Non−coord Macropru
Fig. 4. Impulse responses to a housing demand shock. Optimized parameters.
334 M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336
As in the previous case, and for the same reasons, the non-coor- dinated situation is closer to the benchmark.
0 See for instance Iacoviello (2005) that evaluates a Taylor rule responding to house rices with a policy frontier.
4.4. Financial and macroeconomic stability
Results from the optimal policy analysis have shown that the combination of macroprudential and monetary policies deliver a more stable financial and macroeconomic scenario. In order to show these results graphically, we plot an efficiency frontier that includes the three objectives that the policy makers aim at mini- mizing: variability of output, variability of inflation and variability of borrowing.
Policy analysis is usually done through policy frontiers, also known as Taylor curves or efficiency frontiers.30 This curve shows, given different parameters of the Taylor rule, the combination that delivers the lower output and inflation variability. Therefore, a Tay- lor curve which is closer to the origin would be more efficient. In or- der to include the objective of the macroprudential regulator, we present an extended Taylor curve in which we include the variability of borrowing, as a measure to capture financial stability.
We are aware that there is not a widely accepted definition of financial stability or systemic risk. Those are difficult concepts to
3
p
3 4
5 6
7 8
1.4
1.6
1.8
2 50
100
150
200
250
300
output varianceinflation variance
bo rro
w in
g va
ria nc
e
Benchmark Coordinated Macropru Non−coord Macropru
Fig. 5. Three dimensional efficiency frontier.
M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336 335
define and to measure. Many definitions include the interactions between the financial and the real sector.31 In our model, we char- acterize the financial sector implicitly: borrowers take credits from savers and sign mortgages to buy houses, the asset of our model. Therefore, the financial system can be proxied by the amount of bor- rowing that takes place. Within this framework, we propose a mea- sure for financial stability: a low variability of borrowing. In this sense, a lower variance of borrowing would imply a more stable financial system: if the variance of borrowing is lower, credit is smoother. A more stable financial system contributes to a lower sys- temic risk. Our model fits this idea. Borrowers do not have an Euler equation that allows them to smooth their consumption, as savers do. If the variability of the borrowing is lower then borrowers can sign mortgages in a smoother way and also can achieve a more sta- ble consumption. The financial sector will be more stable and also the real sector. The economy can benefit from a more stable financial system and a lower systemic risk with a higher welfare, as we proved in previous sections. However, if the situation is the opposite and there is a high variability of borrowing, the financial system will be more unstable: with credit being more variable, consumption would also be more variable, the systemic risk will increase and, therefore, welfare will be lower.
Fig. 5 presents our augmented policy frontier which is three- dimensional, since it takes into account three policy objectives: output, inflation and financial stabilization. The first two corre- spond to the standard objectives of the central bank, while the third one would be the objective of the macroprudential regulator. As in previous cases, we are comparing the macroprudential (coor- dinated and non-coordinated) with the no macroprudential sce- nario (benchmark). Here, curves are preferable, the lower (less borrowing variance) and closer to the inflation and output variance origin (less inflation and output variability) they are. We see that when we take the three dimensions together, macroprudential and monetary policies interacting with each other manage to deli- ver a more stable scenario, which includes not only macroeco- nomic stability but also financial stability. These results represent a way to convey the findings in previous sections, that is, the introduction of the macroprudential policy is welfare enhancing because it is delivering a more stable system.
5. Concluding remarks
In this paper, we analyze the impact of macroprudential and monetary policies on business cycles, welfare, and financial stabil-
31 See Galvão and Owyang (2013) for a discussion on the topic.
ity. In particular, we consider a macroprudential rule for the LTV ratio that responds to credit growth.
We compute the optimal parameters of the macroprudential and monetary rule both when monetary and macroprudential pol- icies act in a coordinated and in a non-coordinated way. We find that in both cases, this interaction is welfare improving for the society, especially in the case of the non-coordinated game. How- ever, there is a trade-off between the agents of the model and sav- ers lose from this new scenario. We find that by transfers �a la Kaldor–Hicks, so that borrowers can compensate the saver’s wel- fare loss, a Pareto-superior outcome can be obtained.
From a positive perspective, we show the dynamics of the mod- el under the optimal parameters that maximize welfare. We find that, given a positive technology or housing demand shock, the macroprudential authority would decrease the LTV to moderate the credit boom. In this way, it can achieve its ultimate goal: finan- cial stability.
We also show graphically, with a three dimensional policy fron- tier, that the interaction between monetary and macroprudential policies unambiguously enhances the stability of the economic system.
Acknowledgments
We would like to thank the discussants and participants of IRE- BS Conference 2012, 2012 Dynare Conference, ReCapNet Confer- ence 2013, CEUS Workshop 2013, IFABS Conference 2013, and AREUEA session at the ASSA Meetings 2014; as well as the seminar participants at the Bank of England, the Federal Reserve Board, the Federal Reserve Bank of St. Louis, the Central Bank of Luxembourg, the BBVA, and the University of Nottingham. Special thanks to Mat- teo Iacoviello, John Duca, William Dupor, Pau Rabanal, Carlos Tho- mas, Antonio Mele, Don Schlagenhauf, Daniel Fetter, Christopher Otrok, Rafael Repullo, Jagjit S. Chadha, and two anonymous refer- ees. All errors are our own. J. A. Carrasco-Gallego would also like to acknowledge the financial support of Universidad Rey Juan Car- los Fellowship for International Research Stays to visit the Univer- sity of Nottingham.
Appendix A. Main equations
1 Cs;t ¼ bsEt
Rt
ptþ1Cs;tþ1
� � ; ð29Þ
ws t ¼ ðNs;tÞg�1Cs;t; ð30Þ j
Hs;t ¼ 1
Cs;t qt � bsEt
1 Cs;tþ1
qtþ1; ð31Þ
1 Cb;t ¼ bbEt
Rt
ptþ1Cb;tþ1
� � þ ktRt; ð32Þ
wb;t ¼ ðNb;tÞg�1Cb;t ; ð33Þ j
Hb;t ¼ 1
Cb;t qt � bbEt
1 Cb;tþ1
qtþ1
� � � kb
t ktEtðqtþ1ptþ1Þ; ð34Þ
Et Rt
ptþ1 bt ¼ ktEtqtþ1Hb;t; ð35Þ
Cb;t þ qtHb;t þ Rt�1bt�1
pt ¼ qtHb;t�1 þwb;tLb;t þ bt ; ð36Þ
ws;t ¼ 1 Xt
a Yt
Ns;t ; ð37Þ
wb;t ¼ 1 Xt ð1� aÞ Yt
Nb;t ; ð38Þ
336 M. Rubio, J.A. Carrasco-Gallego / Journal of Banking & Finance 49 (2014) 326–336
p̂t ¼ bEtp̂tþ1 � wx̂t þ upt; ð39Þ
Ws;t � Et
X1 m¼0
bm s log Cs;tþm þ j log Hs;tþm �
ðNs;tþmÞg
g
� � ; ð40Þ
Wb;t � Et
X1 m¼0
bm b log Cb;tþm þ j log Hb;tþm �
ðNb;tþmÞg
g
� � ; ð41Þ
Wt ¼ ð1� bsÞWs;t þ ð1� bbÞWb;t : ð42Þ
References
Abraham, J., Pavlov, A., Wachter, S., 2008. Explaining the United States’ uniquely bad Housing Market. Wharton Real Estate Review 12 (1), 24–41.
Almeida, H., Campello, M., Liu, C., 2006. The financial accelerator: evidence from international housing markets. Review of Finance 10, 1–32.
Andres, J., Arce, O., Thomas, C., 2013. Banking competition, collateral constraints, and optimal monetary policy. Journal of Money, Credit and Banking 45 (2).
Angelini, P., Neri, S., Panetta, F., 2012. Monetary and Macroprudential Policies. Working Paper Series 1449, European Central Bank.
Antipa, P., Mengus, E., Mojon, B., 2010. Would Macroprudential Policy have Prevented the Great Recession? Mimeo, Banque de France.
Ascari, G., Ropele, T., 2009. Disinflation in a DSGE Perspective: Sacrifice Ratio or Welfare Gain Ratio? Kiel Institute for the World Economy Working Paper, 1499.
Bean, C., 2009. The great moderation, the great panic and the great contraction. In: Schumpeter Lecture delivered at the Annual Congress of the European Economic Association, Barcelona, 25 August 2009.
Bean, C., Paustian, M., Penalver, A., Taylor, T., 2010. Monetary policy after the fall. In: Paper presented at the Federal Reserve Bank of Kansas City Annual Conference, Jackson Hole, Wyoming.
Beau, D., Clerc, L., Mojon, B., 2012. Macro-prudential Policy and the Conduct of Monetary Policy. Mimeo, Bank of France.
Benigno, P., Woodford, M., 2008. Linear-Quadratic Approximation of Optimal Policy Problems. Mimeo.
Borio, C., Shim, I., 2007. What can (macro-)policy do to Support Monetary Policy? BIS Working Paper, 242.
Borio, C., Furfine, C., Lowe, P., 2001. Procyclicality of the financial system and financial stability: issues and policy options. In: Marrying the Macro- and Micro-prudential Dimensions of Financial Stability, BIS Papers, vol. 1. pp. 1–57.
Brázdik, F., Hlaváček, M., Maršal, A., 2012. Survey of research on financial sector modeling within DSGE models: what central banks can learn from it. Czech Journal of Economics and Finance 62 (3).
Campbell, J., Hercowitz, Z., 2009. Welfare implications of the transition to high household debt. Journal of Monetary Economics 56 (1), 1–16.
Christensen, I., Meh, C., Moran, K., 2011. Bank Leverage Regulation and Macroeconomic Dynamics, Cahiers de recherche 1140, CIRPEE.
Committee on the Global Financial System, 2012. Operationalising the Selection and Application of Macroprudential Instruments. CGFS Papers, 48.
Domeij, D., Flodén, M., 2006. The labor-supply elasticity and borrowing constraints: why estimates are biased. Review of Economic Dynamics 9, 242–262.
Duca, J.V., Muellbauer, J., Murphy, A., 2011. Shifting Credit Standards and the Boom and Bust in US House Prices. SERC Discussion Paper 76.
Funke, M., Paetz, M., 2012. A DSGE-Based Assessment of Nonlinear Loan-to-Value Policies: Evidence from Hong Kong. BOFIT Discussion Paper No. 11/2012.
Gertler, M., Karadi, P., 2011. A model of unconventional monetary policy. Journal of Monetary Economics 58 (1), 17–34.
Galati, G., Moessner, R., 2013. Macroprudential policy – a literature review. Journal of Economic Surveys 27 (5), 846–878.
Galvão, A.B., Owyang, M.T., 2013. Measuring Macro-Financial Conditions using a Factor-Augmented Smooth-Transition Vector Autoregression. Mimeo.
Gelain, P., Lansing, K., Mendicino, C., 2013. House prices, credit growth, and excess volatility: implications for monetary and macroprudential policy. International Journal of Central Banking 9 (2).
Goodhart, C.A.E., 2004. Some New Directions for Financial Stability? The Per Jacobsson Lecture, Zürich.
Huggett, M., 1997. Wealth distribution in life-cycle economies. Journal of Monetary Economics 38, 469–494.
Iacoviello, M., 2005. House prices, borrowing constraints and monetary policy in the business cycle. American Economic Review 95 (3), 739–764.
Iacoviello, M., 2013. Financial business cycles. Mimeo, Federal Reserve Board. Iacoviello, M., Neri, S., 2010. Housing market spillovers: evidence from an estimated
DSGE model. American Economic Journal: Macroeconomics 2, 125–164. Kannan, P., Rabanal, P., Scott, A., 2012. Monetary and macroprudential policy rules
in a model with house price booms. The B.E. Journal of Macroeconomics, Contributions 12 (1).
Lambertini, L., Mendicino, C., Teresa Punzi, M., 2013. Leaning against boom–bust cycles in credit and housing prices. Journal of Economic Dynamics and Control 37 (8), 1500–1522.
Lawrance, E., 1991. Poverty and the rate of time preference: evidence from panel data. The Journal of Political Economy 99 (1), 54–77.
Lim, C.H., Columba, F., Costa, A., Kongsamut, P., Otani, A., Saiyid, M., Wezel, T., Wu, X., 2011. Macroprudential policy: what instruments and how to use them? Lessons from country experiences. IMF Working Paper 11/238.
McCallum, B., 2001. Should monetary policy respond strongly to output gaps? American Economic Review 91 (2), 258–262.
Mendicino, C., Pescatori, A., 2007. Credit Frictions, Housing Prices and Optimal Monetary Policy Rules. Mimeo.
Monacelli, T., 2006. Optimal monetary policy with collateralized household debt and borrowing constraint. In: Campbell, J. (Ed.), Conference Proceedings Monetary Policy and Asset Prices.
Mundell, R., 1962. The Appropriate Use of Monetary and Fiscal Policy for Internal and External Stability. IMF Staff Papers 9, pp. 70–79.
N’Diaye, P., 2009. Countercyclical Macro Prudential Policies in a Supporting Role to Monetary Policy. IMF Working Paper.
Posner, R., 2007. Economic Analysis of Law, seventh ed. Wolters Kluwer, Austin, TX. Recommendation of the European Systemic Risk Board, 2013. Official Journal of the
European Union of 15.6.2013, on Intermediate Objectives and Instruments of Macro-prudential Policy.
Rubio, M., Carrasco-Galllego, J., 2013. Macroprudential Measures, Housing Markets, and Monetary Policy. Moneda y Crédito.
Schmitt-Grohe, S., Uribe, M., 2004. Solving dynamic general equilibrium models using a second-order approximation to the policy function. Journal of Economic Dynamics and Control 28, 755–775.
Scitovsky, T., 1941. A note on welfare propositions in economics. Review of Economic Studies 9 (1), 77–88.
Smith, A., 2009. Comment on welfare implications of the transition to high household debt by Campbell J., Hercowitz, Z. Journal of Monetary Economics 56 (1).
Svensson, L., 2012. Comment on Michael Woodford, ‘‘Inflation Targeting and Financial Stability’’ PENNING- OCH VALUTAPOLITIK 2012.
Tinbergen, J., 1952. On the Theory of Economic Policy. North Holland Publishing Company, Amsterdam.
- Macroprudential and monetary policies: Implications for financial stability and welfare
- 1 Introduction
- 1.1 Related literature
- 2 Model setup
- 2.1 Savers
- 2.2 Borrowers
- 2.3 Firms
- 2.3.1 Final goods producers
- 2.3.2 Intermediate goods producers
- 2.4 Monetary policy
- 2.5 A macroprudential rule for the LTV
- 2.6 Equilibrium
- 3 Welfare
- 3.1 Welfare measure
- 3.2 Welfare trade-offs
- 4 Optimal policy analysis
- 4.1 Optimal parameters
- 4.2 Pareto-superior outcomes
- 4.3 Impulse responses
- 4.3.1 Technology shock
- 4.3.2 Housing demand shock
- 4.4 Financial and macroeconomic stability
- 5 Concluding remarks
- Acknowledgments
- Appendix A Main equations
- References
Price and Financial Stability in Modern Central Banking.pdf
Price and Financial Stability in Modern Central Banking Author(s): JOSÉ DE GREGORIO Source: Economía, Vol. 13, No. 1 (Fall 2012), pp. 1-11 Published by: Brookings Institution Press Stable URL: http://www.jstor.org/stable/41756780 .
Accessed: 06/02/2015 10:59
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Brookings Institution Press is collaborating with JSTOR to digitize, preserve and extend access to Economía.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
JOSÉ DE GREGORIO
Price and Financial Stability
in Modem Central Banking
Although years were given
ago, the
it the
first was
monopoly not central
until banks the power
mid-nineteenth were
to issue
created
banknotes century more than
and that
to
three central act as
hundred
lender banks
Although years ago, it was not until the mid-nineteenth century that central banks were given the monopoly power to issue banknotes and to act as lender
of last resort. Thereafter, central banks played the role of liquidity provider and lender of last resort. These tasks were intended to allow a proper func- tioning of the payment system, so financial stability was implicitly a major concern for central banks. Over time, central banks moved toward achieving price stability, from monetary stability to controlling inflation. Financial sta- bility became a secondary goal, if a goal at all.
This has not been the case in emerging market economies, which have been affected by recurrent financial crises. Indeed, financial crises like those of Chile in the early 1980s or in Mexico and Asian countries in the 1990s are not radically different from the recent crisis in advanced economies. The complexity may have changed, but the original causes had many similarities.1 Some years ago it was much more frequent to find central bankers concerned about financial stability in emerging countries than in advanced ones. How- ever, as a consequence of the global financial crisis, the issue of financial stability has reemerged as a top priority for policymakers.
In this paper, I discuss the issue of price and financial stability in central banking. I first explore the conduct of central banks in achieving price sta- bility, in particular in the context of inflation targeting, and then move on to how the financial stability mandate has to be included as a key component of modern central banking. I end with a few concluding remarks.
De Gregorio was governor of the Central Bank of Chile from 2007 to 201 1 . He is currently with the University of Chile.
I would like to thank Rodrigo Cifuentes, Luis Oscar Herrera, Alejandro Jara, and Enrique Orellana for valuable discussion and comments.
1 . Reinhart and Rogoff (2009).
1
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
2 ECONOMIA, Fall 2012
Central Banks and Price Stability
As mentioned above, central banks in advanced economies have long been focused mainly on ensuring low inflation. Moreover, some scholars and prac- titioners argued that price stability should be the only objective of central banks, so that the goal would be more credible and monetary policy more effective in achieving stability. How exactly or operationally to achieve this target was an open question, however. Some central banks tried to target monetary aggregates, others to peg nominal exchange rates, and others to use an eclectic mix of indicators. Two decades ago, some central banks started conducting monetary policy targeting a specific value or range for the infla- tion rate. This trend started with New Zealand in 1990 and was followed by Canada, the United Kingdom, Australia, and Sweden in the early 1990s. This is a case in which policy development led academic advances. Progress on the academic front provided further impetus to the adoption of inflation targets as new models were developed to provide the theoretical underpinnings of inflation targets and the basis to conduct empirical work.2
This view was further justified by the success of monetary policy around the world in providing stability, not only on the inflation front, but also in activity and employment. The evidence that output volatility declined sig- nificantly in the United States after the mid-1980s was first reported by Kim and Nelson and later called the Great Moderation by Stock and Watson.3 Several factors could be behind this trend, such as technical progress, better policies, deeper financial markets, and sheer good luck. Although there is no final verdict, evidence points to the role of better macroeconomic policies.4 Emerging market economies also enjoyed a Great Moderation, but it came in the second half of the 1990s, much later than in developed economies. This coincided with the time in which inflation was conquered, supporting the hypothesis that it was good policies rather than good luck.5 It is easy to dis- credit the Great Moderation in the current juncture. However, the resilience of emerging market economies to the global crisis was impressive. Indeed, emerging markets had a recession, but much milder than in the past and with a remarkable recovery. This was, of course, the consequence of much better macroeconomic management.
2. Gali and Gertler (2007). 3. Kim and Nelson (1999); Stock and Watson (2003). A predecessor was Taylor (1998), who
called this period the long boom. 4. Gali and Gambetti (2009). 5. De Gregorio (2008).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
José de Gregorio 3
The case of Chile illustrates this point. The economy did suffer a reces- sion, but the size of the initial impact and the speed of the recovery were quite different from previous episodes. From the second quarter of 1998 to the second quarter of 1999, during the Asian crisis, Chilean GDP fell by 4.1 percent. Returning to the initial GDP level took one year. The unemploy- ment rate more than doubled between the beginning of 1998 and mid- 1999, reaching almost 12 percent and staying high, around 9.5 percent, until 2005. The effects of the 2008-09 crisis were very different. Economic activity fell, but less than in the previous episode, with a drop of 3.3 percent between mid-2008 and mid-2009. The recovery was much faster: GDP was compa- rable to its precrisis level by the end of 2009, only two quarters after the downturn. The subsequent growth rate was also different. Considering the first six quarters after output recovered to its initial level, following the Asian crisis the economy grew at an average annual rate of 4.4 percent, whereas following the 2008 financial crisis the economy grew at an average annual rate of 6.1 percent.
The behavior of the unemployment rate was totally different as well. After having risen from around 7.5 percent in mid-2008 to nearly 1 1.0 per- cent in mid-2009, it quickly descended to levels around 7.0 to 7.5 percent at the beginning of 201 1. The policy regime was crucial for this result. Fiscal policy implemented a sizable economic stimulus package. Monetary policy also gave a significant boost to the economy, as the Central Bank took the monetary policy interest rate to its minimum and implemented additional measures to ensure the effectiveness of its actions. The effects of the 2008-09 crisis were very substantial, but the resilience of the Chilean economy and the effectiveness of its macroeconomic policies were even stronger.
Over time, inflation-targeting regimes have evolved into what is now known as a flexible inflation target (FIT). In this scheme, the central bank sets an inflation target, which is intended to be achieved in a given time horizon. As shown by Svensson, inflation targeting implies inflation-forecast targeting.6 Thus, the central bank's inflation forecast at the policy horizon becomes the intermediate target.7 In the case of Chile, the inflation target is 3 percent, and the time horizon is two years. As long as this target is credible, monetary policy will not only achieve inflation stability, but will also reduce the volatility of the business cycle.
6. Svensson (1997). 7. To avoid indeterminacy or multiple equilibria, the forecast must be the central bank s
forecast and not that of the market (Bernanke and Woodford, 1997).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
4 ECONOMIA, Fall 2012
Analytically, the FIT regime is based on the idea that the policymaker mini- mizes a loss function that penalizes both inflation deviations from the target and deviations of output from full employment. The FIT thus optimizes the trade- off between output volatility and inflation volatility. The time horizon is what makes this scheme flexible. The target is not achieved in the short run since it also takes into account the output costs of achieving the target. A rigid inflation target is one in which the central bank cares only about inflation, so the horizon would be the shortest possible for monetary policy to affect output. The longer the time horizon, the higher the weight of output volatility in the loss func- tion.8 The time horizon typically extends to two years, or more ambiguously to the "medium term." As long as medium-term inflation expectations remain anchored, monetary policy helps to reduce the volatility of other variables.
A FIT regime also requires a flexible exchange rate, so that monetary pol- icy can be conducted independently. However, a proper FIT helps to stabilize the currency as long as monetary policy moves leaning against the wind. For example, a persistent depreciation of the currency, other things equal, increases the inflation forecast, although much more moderately than in rigid exchange rate systems. This effect calls for a tightening of monetary condi- tions, reducing pressures against the currency.
It is often asserted, especially in nonprofessional discussions, that infla- tion targets ignore output fluctuations. As I have just argued, however, this is a mistake. Flexible inflation targets take into account activity and employment, and this is implicit in the choice of the time horizon. Moreover, a credible inflation target is efficient in terms of minimizing the trade-off between out- put and inflation fluctuations, and it also helps to reduce real exchange rate volatility. Indeed, a flexible inflation-targeting regime can maximize welfare and perform much better than an exchange rate or monetary target.
What variables should a central bank consider when setting the interest rate? In the regime I just described, the answer to this question is pretty simple: anything affecting inflation over a two-year horizon. Variables such as inflation expectations, wages, output, unemployment, the exchange rate, commodity prices, and so on have important effects on inflationary forecasts and must be taken into account when deciding the future path of monetary policy. However, investors and wage and price setters must also understand the importance of these variables for the inflation process. Communication is essential, and that is the role of monetary policy or inflation reports, monetary policy statements, minutes, projections, speeches, and so forth.
8. De Gregorio (2007).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
José de Gregorio 5
A key question that arose during the financial crisis involved the extent to which central banks should react to asset prices, such as housing or stock prices. The answer from the perspective of inflation targeting is that as long as they affect the inflation forecast, they should be considered in the monetary policy reaction function.
Asset price bubbles or distortions that may threaten financial stability should be considered when evaluating financial vulnerabilities, but they should not influence monetary policy if they do not have an impact on inflation. It is not clear that an increase in interest rates will be capable of stopping an increase in asset prices. The required adjustments might be so large that they could end up unnecessarily generating high unemployment and an undesired drop in inflation. Under inflation targeting, any interest rate movements that are inconsistent with inflation converging to the target may undermine the cred- ibility of monetary policy, destabilizing inflationary expectations and weaken- ing the effectiveness of monetary policy.
Using monetary policy to burst a bubble in asset prices is particularly complicated in emerging market economies, since bubbles in domestic assets generally take the form of an exchange rate appreciation caused by large capi- tal inflows. Tightening monetary policy to burst the bubble may have perverse effects, since it induces further capital inflows and strengthens the currency. In this case, the interest rate is not the appropriate instrument. Exchange rate intervention could be a better policy tool.
Bringing Back Financial Stability to Central Banking
With the global financial crisis, there was a renewed discussion on the role of central banks in securing financial stability. This is not new. The first central banks were created in Sweden and England in the seventeenth century, but no proper role for central banks was established. Indeed, the Bank of England was founded to finance the war with France in the late seventeenth century.9 It was not until the nineteenth century that the Bank of England was given the monopoly for the issue of banknotes and assigned the role of lender of last resort. In the origins of central banking, its role was to secure the functioning of the payment system. Thus, financial stability was not new, although it was a secondary issue compared to the conduct of monetary policy, which gradu- ally turned to focus on price stability. Since central banks mainly focused on
9. Davies and Green (2010).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
6 ECONOMIA, Fall 2012
one instrument (either the interest rate or the money supply, but not both), financial stability was mostly ignored.
This is not the case with emerging market economies, which have suffered many financial crises. Consequently, the role of financial stability has always been central in policymaking. The most relevant aspect of financial stability is international financial transactions. For this purpose, central banks, most usu- ally, manage exchange rate policies, hold international reserves, and dictate norms to avoid currency mismatches and external payment crises. In Chile, the Central Bank's objectives are price and financial stability. The Constitu- tional Law explicitly establishes that the Central Bank's mission is "to safe- guard the stability of the currency and the normal functioning of internal and external payments." The resilience of financial systems in emerging markets during the global financial crisis owes much to the fact that financial stability was already an important piece of the policy framework.
I would like to discuss three issues regarding financial stability and mon- etary policy. First, was the crisis caused by monetary policy? Second, what are the instruments for financial stability? And third, what are the interactions between financial stability and monetary policy?
Regarding the cause of the crisis, I do not think monetary policy - conducted, for example, on the basis of a Taylor rule - was the main culprit. Very low interest rates in advanced economies induced high risk taking as financial institutions searched for yields. Excess liquidity may sow the seeds for asset price bubbles and financial vulnerabilities.10 However, to cause a huge finan- cial crisis, some serious distortions in the financial system are required. Countries like Australia and Canada had very low interest rates, but their financial systems responded appropriately to the financial crisis. Chile also followed an interest rate cycle similar to that of the United States and did not suffer a financial crisis. Some countries even had a housing bubble, but they did not have the degree of leverage recorded in the United States, which was central in triggering the crisis. Asset price bubbles do not necessarily cause a financial crisis, as was the case with the tech bubble in the early 2000s. The bad combination is asset price bubbles with high leverage in the banking system.
However, monetary policy played a role in the crisis in the way it dealt with bubbles, deviating from the prescriptions to pursue price stability. This was the so-called Greenspan put. The rule followed by the U.S. Federal Reserve was not to react to the formation of a bubble, but to mop up its effects after
10. See, for example, Acharya and Naqvi (2012).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
José de Gregorio 7
the bubble burst. This was done by providing an unlimited amount of liquidity and sharp reductions of interest rates. Recent evidence confirms that includ- ing asset price deflation in the equation improved the fit of the Taylor rule.11 Markets therefore expected monetary conditions to ease when asset prices declined significantly. This point was raised almost ten years ago by Miller, Weiler, and Zhang, who argue that eliminating the downside risk of asset prices feeds the bubble, so bubbles not only increased as the result of irratio- nal exuberance, but they were also "exaggerated by the faith in the stabilizing powers of Mr. Greenspan."12
Regarding the second question, the instruments for financial stability are what have been termed macroprudential tools, as opposed to micropruden- tial regulation, which targets the health of specific financial institutions. One of the first tools used for financial stability was the dynamic provisioning on housing loans implemented in Spain in 2000. It is still too early to fully evaluate this instrument, since it did not avoid a housing bubble, and many institutions dedicated to housing finance, the cajas de ahorro , went bust dur- ing the crisis.
On the time dimension, the idea of macroprudential tools is to avoid the buildup of financial vulnerabilities in the upturn of the business cycle and to have a cushion for the downturn. Financial systems tend to be procyclical, so some sort of break system should be implemented to avoid excessive risk taking. This underlies the new rules in Basel III, especially in the definition of the countercyclical buffer of new capital requirements.
On the cross-section dimension, extra capital has been proposed for sys- temic institutions to make them more resilient to financial turbulences. The definition of systemic institutions is still blurred, however, and given the evolution of financial innovation, an institution that is nonsystemic today may eventually become systemic. Indeed, a nonsystemic institution on a worldwide basis could be systemic from the point of view of particular economies.
Again, these issues are not new in emerging market economies, which in general have more capitalized banks. In Chile, most of the industry already satisfies the requirements that are supposed to be in place by 2019. Moreover, banks with a high market share are subject to even larger requirements.
Chile has also already made significant progress in areas such as restrictions on currency mismatches, liquidity management, and the use of derivatives.
11. Hall (2011). 12. Miller, Weiler, and Zhang (2002).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
8 ECONOMIA, Fall 2012
In all of these cases, the Central Bank of Chile has the authority to set prudential regulation. For banks, these are related to authorizing the use of derivatives and regulating market and liquidity risk, among others. The Central Bank also has a say in "systemic" regulation, such as overall limits for the pension funds. This scheme accommodates recent policy concerns, since it avoids the conflict of interest that arises from merging the supervi- sor of specific institutions with the monetary authority, while preserving an institution that provides a broad look at the stability of the financial system. This being said, however, there is a need to continue strengthening coor- dination with other regulators. The recent creation of a Financial Stability Committee in Chile represents a step in this direction. The Committee will also provide a clearer view on financial stability and risks. Over time, and as learning takes place, some legislation should be introduced to enhance the effectiveness of the Committee. In the Central Bank, an evaluation of financial vulnerabilities and strengths is performed semi-annually in the Financial Stability Report .
Finally, regarding the interactions between macroprudential policies and monetary policy, the traditional view has been influenced by the Tinbergen principle, by which there should exist as many instruments as policy targets.13 This view is reinforced by the fact that, as I argued above, the interest rate is too blunt of an instrument to deal with asset price bubbles and finan- cial dislocations. Therefore, the interest rate - that is, the monetary policy instrument used to achieve the inflation target - must be separated from macroprudential tools to deal with financial stability. However, and perhaps unfortunately, the separation is not that clear. As I discussed before, monetary policy actions, such as the Greenspan put, may create financial instability. The financial crisis also affected the business cycle and, hence, had implica- tions for monetary policy. Moreover, the transmission channels of monetary policy could break down during a crisis. Therefore, the state of the financial system should be taken into account in the conduct of monetary policy in economies following a FIT regime, since it affects the transmission channels and the business cycle.
Another issue is whether macroprudential tools may be used to comple- ment monetary policy. For example, adjusting capital requirements over the cycle or introducing dynamic provisioning may have effects on the output gap, inflation, and interest rates. This is similar to the case of automatic sta- bilizers of fiscal policy, which also have implications for monetary policy.
13. Bernanke (201 1) also discusses these issues.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
José de Gregorio 9
Therefore, conflicts should not arise between financial stability and price sta- bility tools, as long as financial regulation reduces the procyclicality of the banking system. From the point of view of the inflation target, this is just part of the environment in which decisions have to be made.
Perhaps a more controversial issue is the use of macroprudential tools as a substitute for monetary policy. Some emerging markets have recently begun using banks' reserve requirements to tighten credit by reducing the need to increase the interest rate, which has collateral effects on asset prices, espe- cially the exchange rate. The advantage of using the interest rate for monetary policy is that the transmission channels are relatively well known. Changes in monetary policy interest rates affect the cost of financing, asset prices, and the availability of credit. The macroeconomic consequences of changing regula- tion are less well understood. Tightening restrictions on banks may create disintermediation and move credit to unregulated segments of the market. In addition, the latitude of changes in regulation is much more limited than that of interest rates. A more constructive approach may be to design rule-based countercyclical regulation with the clear purpose of minimizing the risk of a financial crisis. Nevertheless, the interactions between monetary and financial policies need to be further explored and clarified.
Final Remarks
A natural reaction to a crisis is to think that everything is wrong and that all must be changed. Emerging markets, in particular Chile, performed well during the crisis and, above all, during the recovery. Therefore, a first lesson must be on the factors that produced these good results, and macroeconomic management was certainly central to the rapid recovery. The financial system was resilient, which shows that the regulatory framework was appropriate. However, the role of policymakers is not to congratulate themselves for past achievements, but to look at strengthening the macroeconomic framework.
As development proceeds, new challenges arise due to financial innovation. It is therefore extremely important to look for lessons while the crisis unfolds, to take advantage of financial development without risking financial stability. There is a need to study the interactions between monetary and financial poli- cies further, especially given the challenges stemming from the global outlook.
Implementing monetary policy in a flexible, inflation-targeting framework has shown its benefits; incorporating financial frictions and the appropriate policies should help us to navigate better in a very uncertain world.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
10 ECONOMIA, Fall 2012
The crisis also left some lessons for the profession and the way in which highly stylized models are used in policymaking. The narrow view is to think they are a precise description of how the world works, when in truth it is much more complicated. However, the other extreme view - and a very bad one for policymaking - is to disregard all that has been learned from academic work. It is necessary to be humble about the current state of eco- nomic knowledge, but it is also necessary to be rigorous and serious about policies.
References
Acharya Viral, and Hassan Naqvi. 2012. "The Seeds of a Crisis: A Theory on Bank Liquidity and Risk-Taking over the Business Cycle." Journal of Financial Eco- nomics (forthcoming).
Bernanke, Ben S. 2011. "The Effects of the Great Recession on Central Bank Doc- trine and Practice." Speech delivered at the Federal Reserve Bank of Boston 56th Economic Conference. Boston, Mass., 18 October.
Bernanke, Ben S., and Michael Woodford. 1997. "Inflation Forecasts and Monetary Policy." Journal of Money, Credit, and Banking 29(4): 653-84.
Davies, Howard, and David Green. 2010. Banking on the Future : The Fall and Rise of Central Banking. Princeton University Press.
De Gregorio, José. 2007. "Defining Inflation Targets, the Policy Horizon and the Output-Inflation Tradeoff." Working Paper 415. Santiago: Central Bank of Chile.
. 2008. "La gran moderación y el riesgo inflacionario: una mirada desde las economías emergentes." Estudios Públicos 110: 5-20.
Gali, Jordi, and Luca Gambetti. 2009. "On the Sources of the Great Moderation." American Economic Journal: Macroeconomics 1(1): 26-57.
Gali, Jordi, and Mark Gertler. 2007. "Macroeconomic Modeling for Monetary Policy Evaluation." Journal of Economic Perspectives 21(4): 25-45.
Hall, Pamela. 2011. "Is There Any Evidence of a Greenspan Put?" Working Paper 2011-6. Zurich: Swiss National Bank.
Kim, Chang-Jin, and Charles R. Nelson. 1999. "Has the U.S. Economy Become More Stable? A Bayesian Approach Based on a Markov-Switching Model of the Busi- ness Cycle." Review of Economics and Statistics 81(4): 608-16.
Miller, Marcus, Paul Weiler, and Lei Zhang. 2002. "Moral Hazard and the U.S. Stock Market: Analysing the Greenspan Put." Economic Journal 112(478): C171-86.
Reinhart, Carmen M., and Kenneth Rogoff. 2009. This Time Is Different: Eight Cen- turies of Financial Folly. Princeton University Press.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
José de Gregorio 1 1
Stock, James H., and Mark W. Watson. 2003. "Has the Business Cycle Changed and Why?" In NBER Macroeconomics Annual 2002 , vol. 17, edited by Mark Gertler and Kenneth Rogoff, pp. 159-230. MIT Press.
Svensson, Lars E. 0. 1997. "Inflation Forecast Targeting: Implementing and Monitor- ing Inflation Targets." European Economic Review 41(6): 111 1-46.
Taylor, John B. 1998. "Monetary Policy and the Long Boom: The Homer Jones Lec- ture." Review (November): 3-11. Federal Reserve Bank of St. Louis.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:59:36 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 1
- p. 2
- p. 3
- p. 4
- p. 5
- p. 6
- p. 7
- p. 8
- p. 9
- p. 10
- p. 11
- Issue Table of Contents
- Economía, Vol. 13, No. 1 (Fall 2012), pp. i-xi, 1-129
- Front Matter
- Editors' Summary [pp. vii-xi]
- Price and Financial Stability in Modern Central Banking [pp. 1-11]
- Foreign Entry and the Mexican Banking System, 1997-2007 [with Comments] [pp. 13-37]
- Temporal Aggregation in Political Budget Cycles [with Comment] [pp. 39-78]
- Distributional Effects of the Panama Canal Expansion [with Comment] [pp. 79-129]
- Back Matter
Reflections on the Conduct of Monetary and Financial Stability Policy.pdf
Reflections on the Conduct of Monetary and Financial Stability Policy Author(s): David A. Dodge Source: The Canadian Journal of Economics / Revue canadienne d'Economique, Vol. 43, No. 1 (Feb., 2010), pp. 29-40 Published by: Wiley on behalf of the Canadian Economics Association Stable URL: http://www.jstor.org/stable/40389554 .
Accessed: 06/02/2015 11:00
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Wiley and Canadian Economics Association are collaborating with JSTOR to digitize, preserve and extend access to The Canadian Journal of Economics / Revue canadienne d'Economique.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
Reflections on the conduct of monetary and financial stability policy
David A. Dodge Bennett Jones LLP
1. Introduction
Eleven years ago, after my term as deputy minister of finance, I had the privilege of giving the Purvis lecture (see Dodge 1998). In that lecture, I provided a practi- tioner's reflection on the role of fiscal policy in Canada over the post-war period and set out some guidelines for the conduct of fiscal policy over the period to 2012. Today, I want to provide a practitioner's reflection on the role and conduct of monetary and macrofinancial policy.
Over the longer term the best macroeconomic contribution that governments - and their agencies including central banks - can make is to preserve confidence in three things:
1 . the soundness of public finance; 2. the future value of money; and 3. the stability of the financial system.
Confidence in macroeconomic and macrofinanical stability is a necessary pre- condition for firms to innovate and invest and for households to work and to save. Confidence in macro stability also allows governments to pursue microeconomic policies that facilitate innovation and promote rapid adjustment to changing cir- cumstances. Sound macroeconomic and macrofinanical policy -just like the rule of law and security of the person and property - is thus a part of the overall Canadian governance framework.
The Doug Purvis Memorial Lecture, delivered at Toronto, 30 May 2009. Email: [email protected]
Canadian Journal of Economics / Revue canadienne d'Economique, Vol. 43, No. 1 February / février 2010. Printed in Canada / Imprimé au Canada
0008-4085 / 10 / 29^0 / ° Canadian Economics Association
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
30 D.A. Dodge
There is a symbiotic relationship between monetary and fiscal policy in achiev- ing longer-term macroeconomic stability and confidence. No central bank can pursue monetary and financial stability policy that maintains public confidence if governments at the same time recklessly create excessive public debt. In the end, poor fiscal policy will dominate good monetary policy. And that is why I argued eleven years ago that governments should aim to
1 . first reduce the ratio of public debt to GDP to a more sustainable level (I suggested 25% on a National Accounts Basis); and
2. operate thereafter to balance the budget over the business cycle.
A commitment to operating in this way would preserve public confidence in the stability of public finance over the long term while appropriately allowing the automatic fiscal stabilizers to work in the short term to reduce the amplitude of variations in output and employment. I argued that discretionary policy should be undertaken with extreme care. To the extent that discretionary fiscal stimulus was to be used in periods of very weak demand, I argued that it needed to be matched by a firm and specific commitment to discretionary reductions in expenditure or increase in taxes in periods of stronger demand. l
This practical approach to fiscal policy that I set out in 1998 is as applicable today as it was then. A Canadian federal deficit of 40 to 50 billion dollars in a year when the output gap is as large as it is likely to be in 2009 is certainly not inappropriate. But to preserve confidence in the future stability of public finances it is very important that Canadian governments (federal and provincial) now make commitments to reduce (eliminate) the new discretionary spending or to increase taxes in order to regain fiscal balance as quickly as possible after 20 10.2 This approach to prudent levels of public debt not only will contribute directly to the stabilization of output and employment in the short to medium term, but, most important, will bolster confidence in the future value of money and the stability of the financial system. I now turn to my reflections on monetary and financial stability policy.
The key long-term objectives of monetary and financial stability policy are to preserve confidence in the future value of money and confidence in the stability of the financial system. In Canada (as in many OECD countries) responsibility for preserving confidence in the future value of money has been assigned to the central bank. Responsibility for assuring financial stability has been divided between the Department of Finance, the Bank of Canada, and several prudential and market conduct regulatory authorities. The financial and economic turbulence of the last two years has called into question both the division of responsibilities and
1 Symmetrically, discretionary fiscal restraint in periods of excess demand needs to be accompanied by firm commitments to unwind the restraint later.
2 In light of our aging population, it is very important to enter the second half of the next decade with low levels of net public debt.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
Reflections on the conduct of monetary and financial stability policy 3 1
the policy frameworks that were in place prior to 2007. Major effort to rethink these frameworks is now under way globally and here in Canada. My objective today is to offer some limited reflections of a former practitioner on these policy frameworks. I will leave it to others to reflect on the division of responsibility between agencies.
2. Monetary policy
Let me begin with policy framework for assuring confidence in the future value of money. Confidence in the future value of money is important for three reasons. First, financial institutions and markets allocate capital most efficiently when there is a high degree of confidence in the future value of money. Longer-run price level certainty decreases the risk premium in longer term nominal financial contracts (see Crawford et al. 2009, 31). Second, certainty about the future value of money reduces the capricious redistribution of incomes that comes with un- expected inflation or deflation and thus promotes social stability (see Meh et al. 2009, 43). Third and most important, certainty about the future value of money actually serves to help stabilize economic output and employment in the short run, what Olivier Blanchard has called the 'divine coincidence' (for an exposition of this issue, see Dodge 2008b). Thus, certainty about the future value of money contributes to both the efficiency and stabilization responsibilities of the central bank.
In theory, it is possible to achieve simultaneously the twin goals of monetary policy - reasonable price stability and stabilization of output at a level consistent with economic potential. If demand can be kept growing at precisely the pace that the potential output of the economy is expanding, then the overall rate of inflation will remain roughly stable at the targeted rate. So the task of the monetary authority is, again in theory, relatively simple: keep the supply of money and credit growing at a rate that is consistent with allowing demand for goods and services to grow at the same rate as the real economy's ability to supply those goods and services. In practice, of course, this task is extremely complicated (see Dodge 2008a).
Given the inevitable uncertainties, a central bank needs a policy framework that minimizes the chances of making big errors in predicting the future course of output and inflation. The bank's policy framework must be communicated clearly, so that all actors - financial players, businesses, households, and governments - can reasonably be assured of the general direction of the actions the central bank will take to control inflation and stabilize output at levels close to potential over the medium term.
Since the end of the Second World War, central banks in OECD countries have struggled to find such an appropriate framework. Furthermore, central banks in both OECD and emerging economies continue to wrestle with this
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
32 D.A. Dodge
problem in an increasingly interconnected financial and economic world. As I have traced this search for a monetary policy framework in another lecture (see Dodge 2008b), I will not recount this history today. As is well known, the monetary policy framework adopted in 1991 by Canada is that the Bank of Canada should set the policy interest rate with the objective of keeping infla- tion (as measured by the rate of increase in the CPI) at 2% over the medium term. This framework is conducted in the context of a floating Canadian dollar.
This inflation targeting (IT) framework has the huge advantage that it an- chors expectations of inflation. Our own experience (as well as that of other IT countries) indicates that IT has been an important contributing factor to the stability of output and employment from 1991 to 2007 (see Bank of Canada 2008).
Does this framework represent the end of monetary policy history? In 2006, well before the current economic and financial turbulence, the Bank of Canada posed two questions to stimulate research on the issue:
1 . Is 2% the right target for inflation? 2. Are there advantages to moving to a price level (or price path) target?
The summary of the research work to date on these issues is contained in the excellent papers in the Bank of Canada Review, Spring 2009.
My reading of the research on the 2% target is that there is still no clear evidence that a lower inflation target would yield economic performance significantly bet- ter than that achieved with our 2% target, although there are some indications that it might do so. There do seem to be some indications that a price level or price path target (PT) would yield superior performance - especially at times when rates have hit the zero lower bound. I think the PT might well represent an improvement to our current IT framework. A major issue to be resolved however, is the period over which policy would aim to return to the agreed-on price path. There is also a potential additional benefit that could come from the government's agreeing to a price path - namely, this would provide additional incentive for fiscal pru- dence, as the government would not be able to reduce the real burden of the debt through a one-time burst of inflation.
One other key issue raised by the Bank in 2006 was whether or not the frame- work for setting the policy interest rate should include a focus on asset prices in addition to its focus on consumer price stability. I firmly believe the answer to this question is no. In setting the policy rate, the central bank must focus on consumer price stability, although a small amount 'leaning against the wind' of rapid and sustained asset price movements may be appropriate. But since this question can be adequately addressed only in the context of the overall framework policies to promote financial stability, I will now turn to these issues.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
Reflections on the conduct of monetary and financial stability policy 33
3. Policies to promote financial stability
Market participants will always be subject to bouts of excessive exuberance and deep pessimism.3 Over- and underreaction is inherent in financial markets. Mar- ket players inevitably look to the behaviour of other market players. If everyone else is betting on continuing asset price increases - and borrowing to purchase these assets - it is hard to resist doing the same thing. This is equally true for homeowners and sophisticated fund managers. For this reason, financial mar- kets will always be characterized by successive waves of exuberance (bubbles) and pessimism (crashes). Crashes are not black swans, but are rather the inevitable outcome of a previous wave of exuberance.
The stability role of the authorities is to establish and administer a policy framework that works to dampen waves of excessive exuberance and mitigate waves of excessive pessimism. As excessive exuberance has in the past shown up in rising leverage, rapid credit growth, unrealistically narrow credit spreads, and asset price inflation, the policy framework should include mechanisms which more or less automatically serve to dampen these excesses. Similarly the frame- work should include mechanisms that mitigate the liquidity squeezes, the market collapses, the blowout of credit spreads, and the credit contractions that are the result of waves of excessive pessimism.
The central bank shares the responsibility for macrofinancial stabilization with prudential and market conduct regulators and the Ministry of Finance. In its conduct of monetary policy, financial system analysis, and financial market operations, the central bank must keep financial stability concerns in mind. In both the design and operation of the prudential framework, prudential supervi- sors must take into account what is happening 'over the dyke' in markets beyond their regulatory fiat. Similarly, agencies responsible for the oversight of markets and market conduct need to have both the mandate and the capacity not only to ensure that basic principles of disclosure are applied in all financial markets, but also to ensure that markets are continuous, that the 'mechanics' of markets are sound, and that markets do not exacerbate pro-cyclicality.
Market conduct regulators need to cooperate with prudential supervisors, central banks, and ministries of finance. All agencies with oversight responsi- bilities (including CMHC) must assume some of the burden for stabilizing the
3 Of course, central bankers too can be subject to bouts of optimism and pessimism. Following the events of 9/1 1 in 2001, the Federal Reserve, and to a lesser extent we at the Bank of Canada, were overly pessimistic about demand growth and overly feared deflationary pressures. Hence, in the United States policy rates were held too low from mid-2002 through 2004 to achieve the Federal Reserve's twin objectives of low inflation and high unemployment. For a cogent analysis of this issue, see chapter 1 of John Taylor's excellent little book Getting Off Track (2009). Partly because the U.S. policy rate was so low, we at the Bank of Canada were somewhat hesitant to raise our policy rate further than we did, even though in retrospect a slightly higher rate might have been warranted to achieve our inflation target over the medium term.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
34 D.A. Dodge
financial system. It should be in their legislated mandate.4 In other words, agen- cies responsible for prudential regulation of financial institutions and oversight of mortgage and financial markets must, in cooperation with the central bank, focus on macrofinancial stabilization issues.
Coordination and cooperation at the international level are as essential as they are at the national level. Just as we recognize that implementation is the responsibility of national authorities, the same principles of regulation need to be applied across global markets (see Financial Stability Forum 2009; G20 Working Group 1 2009).
Now let me turn to the principles of financial stability policy and the roles of central banks and regulators.
4. Central banks
A major way central banks influence financial stability is, of course, through the setting of the policy interest rate. Some have argued that central banks should set the policy rate with a view to stabilizing asset prices rather than (or in addition to) consumer prices. They argue that failure to do so is a major contributing factor to instability in financial markets, instability that can then lead to output instability.5
I would argue that in setting the policy rate central banks should continue to focus on consumer prices over the medium term, not directly target asset prices.6 Nevertheless, it is certainly true that in setting the policy rate, that is, the price of overnight credit, the central bank should take into account the evolution of the financial system. As this system evolves, so may the relationship (spread) between the price of overnight credit and the effective price of credit to households and businesses. The problem is that, during times of increasing asset prices, the compression of spreads is small.7 Thus, only small adjustments in the policy rate would be warranted to take into account the changing financial structure and lean against the wind of rising asset prices. When credit markets are under severe stress, as they have been since mid-2007, credit spreads do widen dramatically, and a much more significant adjustment in the policy rate is warranted to offset the wider spreads and achieve the deserved effective price of credit to businesses and households.
4 Note that even the Bank of Canada does not have a specific financial stability mandate, although the preamble of the Bank of Canada Act calls on it to 'regulate credit and currency in the best interest of the economic life of the Nation.'
5 See, for example, annual reports of the Bank for International Settlements (BIS) from 2004 to 2007. Some of the earliest arguments can be found in Borio and Lowe (2002).
6 Of course, house prices are included m the Canadian CPI, and hence the Bank of Canada does respond to rapid increases in the price of owner-occupied housing. See the Bank of Canada Monetary Policy Reports of 2005 and 2006, especially the report of October 2006, 33.
7 hor example, the spread between A-rated corporate bonds and Government of Canada bonds declined only about 60 bps between 2002 and 2004, but increased by about 250 bps between June 2007 and September 2008. See Dodge (2008b), 10-1 1.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
Reflections on the conduct of monetary and financial stability policy 35
But changes in the structure of financial markets affect much more than the price of credit. They dramatically affect credit availability and the terms and conditions on which credit is made available. The rapid global escalation in asset prices from 2002 to 2007 (especially in house prices) was certainly an indication of danger ahead and that policy instruments that directly affected the availability of credit needed to be deployed. Similarly, the collapse of asset prices after mid- 2007 indicated the need to deploy policy instruments that directly affect liquidity and the availability of credit.8
Adjustment of the policy rate can work well to simultaneously achieve both reasonable stability of consumer prices and financial markets when the financial structure is stable and markets are normally liquid. But in times of a rapidly evolving financial structure or in times of excessively exuberant or deeply pes- simistic and illiquid financial markets, the appropriate degree of financial and economic stabilization cannot be achieved exclusively through adjustment of the policy rate. In periods of rapid asset-price inflation, some modest increase of the policy rate from the level that would be judged appropriate to stabilize output and consumer prices might be warranted as spreads narrow. But I would con- tinue to argue that large asset-price movements indicate the need for other policy instruments to be brought to bear - instruments that influence the availability of credit rather than just the price of credit.
The central bank does have an important role to play in improving macrofi- nancial stability, a role that does not involve the use of the policy rate. The Bank of Canada is the lender-of-last-resort to the banking system and hence has a key responsibility to ensure that commercial banks maintain adequate liquidity. It also is the provider of macrofinancial analysis through the Financial System Re- view and interaction with prudential and market regulators, governments, and the private sector. And the current crisis has demonstrated that in some countries the central bank has become de facto the market maker of last resort for systemically important financial markets. Whether this should be the case de jure, is a matter for further debate. But ways to ensure the continuity of markets must be found. I will return to this issue in my discussion of the role of securities commissions, below.
It is abundantly clear that central banks need to devote (and are devoting) more effort to monitoring and assessing financial market developments, including market and institutional liquidity issues. Central banks are in the best position to assess and analyze macrofinancial developments and to make this analysis available to other agencies and the private sector.
The Bank of Canada has done this since 2003 through its semi-annual Fi- nancial System Review and is devoting much more effort to this since 2007. On stability issues, the Bank cooperates very closely with its Financial Institutions
8 This distinction between the policy rate instrument that affects the price of credit and instruments that affect the availability of credit is important. In this regard, it is interesting to reread the Porter and Radcliffe reports and the report of the U.S. Commission on Money & Credit from the early 1960s.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
36 D.A. Dodge
Supervisory Committee partners - the Canada Deposit Insurance corporation (CDIC), the Office of the Superintendent of Financial Institutions (OSFI) and the Department of Finance - but even closer cooperation is desirable going for- ward. And some way must be found to enhance cooperation on stability issues with CMHC and, most important, securities commissions. Here the Bank of Canada has a real leadership role to play.
Some have argued that such cooperation is not enough, and that micropruden- tial supervision should return to central banks. I do not believe this is necessary or necessarily desirable. But what is necessary is that central banks have the capacity to absorb and make use of microfinancial data from prudential regulators and that prudential regulators absorb and make use of the macrofinancial analysis provided by central banks.
Let me now turn to the role of prudential regulators.
5. Prudential regulators
What the experience of the last two years has clearly demonstrated is the need for prudential regulators to adopt a regulatory framework that requires financial institutions to set aside more general reserves or increase capital during the up- swing and allow them to draw on those reserves or capital during the downturn. At the very least, prudential regulators need to use 'through the cycle' measures of risk in setting the minimum risk weighted capital standards.
My own view is that regulators should require financial institutions to set aside additional provisions on the upswing and at the top of the credit cycle and allow these reserves to be drawn upon in times of financial stress. Some 'automatic' form of countercyclical provisioning would clearly improve the stability of the financial system in the same way that the automatic tax and expenditure stabilizers improve economic stability.
But the establishment of 'automatic' standards by regulators is not easy, as Superintendent of Financial Institutions Julie Dickson has said. And just as automaticity should not replace all accounting judgment, so automatic rules cannot do the whole job for prudential regulators. But at least some general principles should automatically apply - and need to be applied internationally to prevent regulatory arbitrage.
Much work to create appropriate standards is currently under way at the Fi- nancial Stability Board, the Bank of International Settlements, the International Monetary Fund, and the national central banks and regulatory authorities. I believe useful general principles can be applied. As the superintendent said: 'It may be that the most promising avenues to explore are higher quality tier 1 capi- tal; leverage ratios; loan-to-value ratios; through-the-cycle estimates under pillar 2; and loan loss provisioning. The first four were very important in terms of the Canadian banking system and allowed the system to withstand the stress of global market turmoil and also successfully raise private capital. They help to
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
Reflections on the conduct of monetary and financial stability policy 37
make the system less pro-cyclical or more counter-cyclical' (Dickson 2009; see also G20 Working Group 1, 2009).9
These principles and others can be built into international standards (pillar 1). For example, loan-to- value ratios can be set to vary inversely with the rate of increase in the price of the asset class (e.g., residential mortgages). And I believe this can be done with appropriate flexibility. While precisely perfect automatic adjustment is not achievable, precise perfection should not become the enemy of the roughly good.
Let me re-emphasize that the build-up of excessive leverage earlier this decade was the fundamental cause of the problems we face today. Control of raw lever- age ratios for all financial institutions is key to ongoing financial stability. OSFFs 20-to-l maximum ratio in normal times has served Canada rather well. Gener- ally, leverage ratios for investment banks and banks outside North America are much higher. We need a set of international principles on acceptable levels of leverage.
To reduce the build-up of leverage in the cyclical upswing, higher minimum margins for financing securities need to be established by the market conduct regulator working in concert with the central bank. At the very least, required margins should be stable over the cycle. I believe that increases in these margins at times when asset prices are increasing rapidly would provide a powerful in- strument to dampen asset price volatility - a much more appropriate instrument than the use of the much blunter instrument of the policy rate.
The maximum loan-to-value ratio for insured mortgages, the terms and con- ditions of that insurance, and the minimum loan-to-value where insurance is required all are powerful tools in stabilizing housing prices. At best, these ra- tios and conditions should be adjusted to reduce procyclical behaviour. At the very least, these ratios and conditions should be stable over the cycle rather than being changed in a procyclical manner, as was the case earlier this decade.
Finally, let me emphasize that prudential regulators must pay much more attention to ensuring that banks have at all times adequate liquidity to meet con- ditions of extreme stress. In part this can be achieved through standardization of important securities and derivative instruments and ensuring continuous mar- kets for these instruments on back-stopped exchanges or clearing houses. But in the end, because it is expensive for banks to maintain adequate liquidity, much higher capital should be required for illiquid (or potentially illiquid) assets in their trading books than is currently the case, and symmetrically, lower capital charges for other assets. The banking system can function only if confidence and trust is maintained; liquidity is the fundamental bedrock upon which trust is built.
9 All the papers in the second half of the Bank of Canada's Financial System Review, June 2009, deal with the issue of procyclicality. Also see CGFS (2009).
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
38 D.A. Dodge
6. Securities regulators
Because as much as two-thirds of 'traditional banking business' is now conducted directly through financial markets, securities regulators have had thrust upon them a need for broadened financial market focus. At the moment, they are currently neither professionally equipped nor legally mandated to carry out this broadened focus. This needs to change. Additional focus on full disclosure and appropriate documentation for complex products will improve stability as well as market efficiency.
But fundamentally, governments throughout the world and here in Canada need to expand the mandate of securities regulators. The mandate should be that of 'financial markets oversight' - oversight of all aspects of markets and financial products. These market overseers need to have the mandate to go beyond disclosure for certain products. They also need to enforce disclosure in what are currently exempt products.
What is most important is that some agency - whether the securities regulator or the central bank - be given the mandate to force standardization of the most important complex products and derivatives that are currently traded over the counter. Exchanges or clearing houses need to be established and central coun- terparties - complete with risk-proofed clearing and settlements systems - need to be created and monitored.
And, what is very important, the securities regulator needs to work closely with the prudential regulator to ensure not only that certain products (such as credit default swaps) trade in a transparent organized market, but that the issuer (such as the seller of credit default insurance) has adequate capital to support the product when markets are under stress.
None of this is easy. But it is vital. Continuous markets for these products must be built and backstopped. It would be easier if we had one securities regulator and one government responsible for creating the mandate. But we do not and that shouldn't stop us cooperating to get the mandates right. In Canada we have had an exemplary stability record on the macrofinancial and prudential regulatory side. We have had almost 20 years of successful monetary policy. We have a fiscal record unequalled in the major OECD countries. And we have had unparalleled cooperation between Finance, the Bank, and the prudential regulator. Surely we have the ability to create and oversee transparent and continuous securities markets.
7. Conclusion
I have tried to provide a brief tour d'horizon of the issues in monetary and financial policy based on my own experience. At the end of my Purvis lecture in 1998, I advanced 10 guidelines for future fiscal policy. Let me conclude, here, with 8 guidelines for monetary and financial policy in the decade ahead. These guidelines
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
Reflections on the conduct of monetary and financial stability policy 39
are broad and general, reflecting my experience that in the area of monetary and financial policy it is far better to be 'general and roughly right 'than to be 'specific and precisely wrong.'
1 . The main long-run contribution a central bank can make to the welfare of citizens is to preserve confidence in the future value of money. While the Bank of Canada does need to be cognizant of changes in financial market structures in setting the policy rate, monetary policy should continue to be set to achieve consumer price stability.
2. Our general framework of inflation targeting is working reasonably well both in anchoring expectations and guiding policy to stabilize output at a level close to potential. This framework should be continued in the coming decade. Some modest improvements may be possible; in particular, price level or price path targeting should be considered in the Bank's next five- year review.
3. The Bank of Canada does have an important role to play in providing macrofinancial analysis at all times and in providing liquidity to financial institutions in times of stress. The Bank will have to devote more effort to the monitoring and analysis of systemic risks that are building in the financial system as innovation takes place, and regulators will have to become better equipped to use that analysis. The objective is not to stifle innovation but to ensure a better understanding of the systemic risks posed by innovation and devise ways to mitigate the systemic risks that will arise because ofthat innovation.
4. Prudential regulators will need to find some way to introduce counter- cyclical capital buffers or reserves for banks. At the very least, the procycli- cal bias in the current system needs to be eliminated. The basic principles applicable globally (pillar 1) should include some automatic adjustment, although such adjustment will be rough and not necessarily completely ad- equate. It will be up to national regulators to make additional judgments (pillar 2) with respect to what additional or lesser reserves (or capital) should be required.
5 . Minimum margins for financing securities transactions and minimum hair- cuts for derivatives need to be established and enforced. At the very leas, these should be stable over the cycle, although preferably counter-cyclical in their application. Loan-to-value ratios for residential mortgages and terms and conditions for mortgage insurance should also at the very least be stable over the cycle, although preferably counter-cyclical in operation.
6. Securities commissions will need a mandate to ensure that all issuers of all securities provide all purchasers with adequate information. Systemically important derivative products (including credit default swaps) need to be standardized. The securities regulator (or the central bank) will need a mandate to create (or cause to be created) and oversee continuous markets for these systemically important instruments.
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
40 D.A. Dodge
7. Financial institutions themselves need to manage risks better over the course of the cycle. Continuous stress testing using 'improbable' scenarios is essential. It is important that institutions are given incentives to improve their own practices but that these incentives not be stifled by overly detailed prescriptive rules.
8 . Finally, cooperation is essential - both nationally and globally. The Bank of Canada, regulators, and the private sector need to work together at home to build a stronger system. We have cooperated well in Canada but can do even better. And we can provide leadership to promote international cooperation to forge a better framework for global macrofinancial and prudential policies.
References
Bank of Canada (2006) Renewal of Inflation Control Target: Background Information, November
Borio, G, and P. Lowe (2002) 'Asset prices, financial and monetary stability: exploring the nexus,' BIS Working Paper 1 14, July
CGFS (2009) The Role of Valuation and Leverage in Procyclicality, Publication No. 34, April
Crawford et al. (2009) 'Price-level uncertainty, price level targeting and nominal debt contracts,' Bank of Canada Review, Spring
Dickson, Julie (2009) Remarks to the Asian Bankers Summit, OSFI, 12 May. www.osfi- bsif.gc.ca - under speeches
Dodge, David A. (1998) Reflections on the role of fiscal policy, Canadian Public Policy, September, 275-89
- (2008a) Eric Hanson Memorial Lecture, University of Alberta, February. http://www.uofaweb.ualberta.ca/economics2/pdfs/04-Feb-08-David-Dodge- Hanson-Lecture.pdf
- (2008b) 'Central banking at a time of crisis and beyond,' Benefactors Lecture, CD. Howe Institute, November, www.cdhowe.org - under search publications/dodge
Financial Stability Forum (2009) Report of the FSFon Addressing Stability in the Financial System, April.
O20 Working uroup 1 (2009) Enhancing Sound Regulation and Strengthening Trans- parency, Final Report, 25 March; G20 'Declaration on Strengthening the Financial System,' London, 2 April
King, Mervyn (1995) 'Do inflation targets work?' Bank of England Quarterly Bulletin Men et al. (2009) Unexpected inflation and redistribution of wealth in Canada Bank of
Canada Review, Spring Taylor, John B. (2009) Getting Off Track: How Government Actions and Interventions
Caused, Prolonged, and Worsened the Financial Crisis (Stanford, CA: Hoover Institu- tion Press)
Walsh, Carl (2008) 'Inflation targeting: what have we learned?' Bank of Canada, July
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 11:00:42 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. [29]
- p. 30
- p. 31
- p. 32
- p. 33
- p. 34
- p. 35
- p. 36
- p. 37
- p. 38
- p. 39
- p. 40
- Issue Table of Contents
- The Canadian Journal of Economics / Revue canadienne d'Economique, Vol. 43, No. 1 (Feb., 2010), pp. 1-403
- Front Matter
- Measuring the Gains from Trade under Monopolistic Competition [pp. 1-28]
- Reflections on the Conduct of Monetary and Financial Stability Policy [pp. 29-40]
- Structural Gravity Equations with Intensive and Extensive Margins [pp. 41-62]
- Trade Flows in a Spatial Oligopoly: Gravity Fits Well, but What Does It Explain? [pp. 63-96]
- U.S. Trade Remedy Law and Agriculture: Trade Diversion and Investigation Effects [pp. 97-126]
- Trade Diversion from Tomato Suspension Agreements [pp. 127-151]
- Does the Version of the Penn World Tables Matter? An Analysis of the Relationship between Growth and Volatility [pp. 152-179]
- Does FDI in Manufacturing Cause FDI in Business Services? Evidence from French Firm-Level Data [pp. 180-203]
- National Champions and Globalization [pp. 204-231]
- International Corporate Taxation and U.S. Multinationals' Behaviour: An Integrated Approach [pp. 232-253]
- Information Technology and Efficiency in Trucking [pp. 254-279]
- A Bioeconomic View of the Neolithic Transition to Agriculture [pp. 280-300]
- Equity-Regarding Poverty Measures: Differences in Needs and the Role of Equivalence Scales [pp. 301-322]
- The Role of Child Health and Economic Status in Educational, Health, and Labour Market Outcomes in Young Adulthood [pp. 323-346]
- The Evolution of Male-Female Earnings Differentials in Canadian Universities, 1970-2001 [pp. 347-372]
- Understanding the Wage Patterns of Canadian Less Skilled Workers: The Role of Implicit Contracts [pp. 373-403]
- Back Matter
Rough Sets and the role of the monetary policy.pdf
European Journal of Operational Research 181 (2007) 1554–1573
www.elsevier.com/locate/ejor
Rough Sets and the role of the monetary policy in financial stability (macroeconomic problem)
and the prediction of insolvency in insurance sector (microeconomic problem)
A. Sanchis b, M.J. Segovia a,*, J.A. Gil a, A. Heras a, J.L. Vilar a
a Department of Financial Economy and Accounting I, Facultad de Ciencias Económicas y Empresariales,
Universidad Complutense de Madrid 28223, Spain b Department of Business Economy, Universidad Carlos III de Madrid, Spain
Received 1 December 2004; accepted 1 January 2006 Available online 5 June 2006
Abstract
This paper faces two questions related with financial stability. The first one is a macroeconomic problem in which we try to further investigate the role of monetary policy in explaining banking sector fragility and, ultimately, systemic banking crisis. It analyses a large sample of countries in the period 1981–1999. We find that the degree of central bank independence is one of the key variables to explain financial crisis. However, the effects of the degree of independence are not linear. Surprisingly, either a high degree of independence or a high degree of dependence are compatible with a situation of finan- cial stability, while intermediate levels of independence are more likely associated with financial crisis. It seems that it is the uncertainty related with a non-clear allocation of monetary policy responsibilities that contributes to financial crisis episodes.
The second one is a microeconomic problem: the prediction of insolvency in insurance companies. This question has been a concern of several parties stemmed from the perceived need to protect general public and to minimize the costs associated such as the effects on state insurance guaranty funds or the responsibilities for management and auditors. We have developed a bankruptcy prediction model for Spanish non-life insurance companies and the results obtained are very encouraging in comparison with previous analysis. This model could be used as an early warning system for super- visors in charge of the soundness of these entities and/or in charge of the financial system stability.
Most methods applied in the past to tackle these two problems are techniques of statistical nature and, variables employed in these models do not usually satisfy statistical assumptions what complicates the analysis. We propose an approach to undertake these questions based on Rough Set Theory. � 2006 Elsevier B.V. All rights reserved.
Keywords: Rough Sets; Financial stability; Central bank independence; Insolvency; Insurance companies
0377-2217/$ - see front matter � 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejor.2006.01.045
* Corresponding author. Tel.: +34 91 3942569; fax: +34 91 3942570. E-mail address: [email protected] (M.J. Segovia).
1 In his words, ‘‘the issue of financial stability was part of the central banks’ genetic code’’.
2 For a description of the role of central banks in financial stability across regimes see Borio and Lowe (2002).
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1555
1. Introduction
The financial system plays a crucial role in eco- nomic development as responsible for the allocation of resources over time and among different alterna- tives of investment by pricing the postposition of consumption (free risk rate) and pricing the risk (risk premium). A correct functioning of the finan- cial system allows economies to reach higher levels of real growth as well as more stable macroeco- nomic conditions. In the last 20 years at least 10 countries have experienced the simultaneous onset of banking and currency crisis, with contractions in Gross Domestic Product of between 5% and 12% in the first year of the crisis, and negative or only slightly positive growth for several years there- after (Stiglitz and Furman, 1998). Therefore, pre- serving financial stability is one of the main goals for policy makers since the beginning of the mone- tary systems.
The especial role that banks play in the financial system and their specificities as money issuers explain why a great number of financial crisis had got the banking sector as protagonist. In the 1980s and 1990s several countries, including developed economies, developing countries, and economies in transition have experienced severe banking crises. Such proliferation of large scale banking sector problems has raised widespread concern, as banking crises disrupt the flow of credit to households and enterprises, reducing investment and consumption and possibly forcing viable firms into bankruptcy. Banking crises may also jeopardize the functioning of the payments system and, by undermining confi- dence in domestic financial institutions; they may cause a decline in domestic savings and/or a large scale capital outflow. Finally, a systemic crisis may force sound banks to go to bankrupt.
Preventing the occurrence of systemic banking problems is undoubtedly a chief objective for pol- icy-makers, and understanding the mechanisms that are behind the surge in banking crises in the last dec- ades is a first step in this direction. A number of studies have analyzed various episodes of banking sector distress in an effort to draw useful policy les- sons (González-Hermosillo, 1996; Kaminsky and Reinhart, 1999).
The goal of the first study we present in this paper is to identify the factors behind banking sec- tor fragility focusing on the role of monetary policy. Our panel includes all market economies for which data were available in the period 1981–1999. The
explanatory variables capture many of the factors suggested by the theory and highlighted by empiri- cal studies.
There is no clear consensus on how monetary policy and financial stability are related. In particu- lar, it is not clear whether there are any trade-offs or synergies between them. This issue is very impor- tant, since it could help to devise arrangements and policy responses to promote both monetary and financial stability. The design of monetary pol- icy should be particularly important since the cen- tral bank has a natural role in ensuring financial stability, as argued by Padoa-Schioppa (2002)1
and Schinasi (2003), and has virtually always been involved in financial stability, directly or indirectly.2
As important as to gain some insight on the macro factors which contribute to financial stability is to know fragilities that arise at the micro level. Although a sound macro economic and institutional environment is crucial to promote financial stability, supervisors have to perform a continuous basis oversigh on the individual elements that constitute the financial system: financial companies, markets, clearing and settlements institutions, etc. in order to guarantee an appropriate level of financial stability.
Although, as it is said above, many financial cri- sis are associated with the banking sector, globaliza- tion, the emergence of conglomerates, financial innovation, and system integration makes more and more difficult to isolate one part of the financial system from another. The nature of potential insta- bility may have already taken new forms as a conse- quence of the ongoing transformation of the financial system. Such recent changes in the finan- cial system might be summarized by the breakdown in the separations between financial institutions and financial markets, between the three main categories of financial institutions (banks, insurance compa- nies, and on-bank financial institutions), and between national financial systems. These separa- tions have been replaced by an increasing integra- tion of markets with banks, and of banks with other financial institutions, and by an increasing internationalization of the financial system. There- fore, new potential sources of disturbances can be identified that are closely related to this changed
1556 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
environment. Financial instability may result from market instability and also from other financial institutions (Padoa-Schioppa, 2002). For instance, in 2001 in Australia the collapse of a major insur- ance company led to a Royal Commission of Inquiry as contagion was spread through the small business sector via denial of insurance or claims for certain activities or incidents, or through higher premiums.
Among the components of the financial system in the second study we present in this paper we focus on the insurance sector because, although it plays a growing and crucial role in modern economies, it has received less attention from researches. More- over, given its business peculiarities it is not possible to translate the conclusions from banking sector analysis to the insurance sector and therefore a spe- cific analysis is needed.
The insurance industry is of fundamental eco- nomic and social importance. It has long been recog- nized that there needs to be some form of prudential supervision of such entities to attempt to minimize the risk of failure. Nowadays, Solvency II project is intended to lead to the reform of the existing sol- vency rules in European Union. Therefore, develop- ing new methods to tackle the problems we have mentioned above is a highly topical question.
In this paper we tried to show how the methodol- ogy we propose is flexible enough to tackle with both problems, the analysis at macro and micro level outperforming previous analysis.
In the past a large number of methods have been proposed to deal with these two matters. Most approaches applied are statistical techniques such as discriminant or logit analysis. In most cases, the attributes employed as explicative variables do not usually satisfy statistical assumptions. So in order to avoid these inconveniencies of statistical meth- ods, we propose an approach to predict insolvency of insurance companies and financial instability in a country based on Rough Set Theory (RS Theory).
Some of the advantages of this theory are: first, it is a useful tool to analyse information systems rep- resenting knowledge gained by experience; second, we can use qualitative and quantitative variables and it is not necessary that the variables employed satisfy any assumption; third, through this analysis the elimination of the redundant variables is got, so we can focus on minimal subsets of variables to evaluate insolvency or instability and, therefore the cost of the decision making process and time employed by the decision maker are reduced;
fourth, the analysis process results in a model con- sisted of a set of easily understandable decision rules so usually it is not necessary the interpretation of an expert and finally, fifth, these rules are based on the experience and they are well supported by a set of real examples so this allows the argumentation of the decisions we make.
In this paper we applied the RS analysis to get closer to the factors that can contribute to financial stability in a country. The RS analysis allows over- coming some of the rigidities of other methodolo- gies previously applied as logit and probit analysis. This paper also completes previous researches for prediction of banks bankruptcy based on RS Theory (Dimitras et al., 1999; Greco et al., 1998; Mckee, 2000; Slowinski and Zopounidis, 1995; Zopounidis and Dimitras, 1998) developing a prediction model for insurance companies. The results are very encouraging in comparison with more traditional techniques.
A financial crisis is the sum, significant enough, of several individual crises together. The time coinci- dence of these individual crises can be explained by common factors affecting the financial sector, as in any other sector, or by contagion. The peculiarity of the financial sector is that the contagion can arise through affecting consumer confidence. For instance, a sharp fall in depositor’s confidence can lead to an unexpected withdraw of deposits, which can pose problems even in otherwise sound financial companies. Moreover, there are other contagion mechanisms as the ownership links, commercial links, etc. Therefore, at the extent the crisis of a financial company can be contagious, this individual crisis could trigger a systemic crisis. Once a financial crisis is on going, the likelihood of a crisis in other financial companies also increases. This feedback mechanism is the typical feature of a financial crisis.
Although we treat in this paper both problems separately, an obvious step forward would be to develop an integrate model that accounts for the links between the crisis of an individual financial company and a systemic financial crisis. In the meantime, we think the use of the macro model to monitor the macro factors that can increase the like- lihood of a financial crisis together with the models at a micro level to monitor the probability of an individual crisis can help the competent authorities to prevent or, at least, mitigate the effects of a finan- cial crisis.
The rest of the paper is structured as follows: Sec- tions 2 and 3 introduce the theoretical models
3 Higher real interest rates are likely to hurt bank balance sheets even if they can be passed on to borrowers, as higher lending rates result in a larger fraction of non-performing loans.
4 For an in-depth discussion of the theory of bank runs, see Bhattacharya and Thakor (1994).
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1557
underlying the selection of explanatory variables for financial crisis and insurance insolvency, respec- tively. In Section 4 we explain the methodology. Section 5 describes the empirical models and the main results we obtained. Finally, Section 6 high- lights the main conclusions that can be outlined from the analysis.
2. The determinants of banking crises
Houben et al. (2004) define financial stability in terms of its ability to help the economic system to allocate resources, manage risks, and absorb shocks. Moreover, financial stability is considered a continuum, changeable over time and consistent with multiple combinations of its constituent ele- ments. In the same paper we can find an Appendix A that provides an overview of definitions or descriptions of financial stability by a selected group of officials, central banks and academics. We focused on banking crises, as a financial instability outcome, because monetary policy, which is the var- iable we want to focus, is more directly related to the functioning of the banking system than to the rest of the financial system.
The literature offers several definitions of bank- ing crises (Friedman and Schwartz, 1963; Bordo, 1986; Lindgreen et al., 1996; Caprio and Klingebiel, 1997; Gupta, 1996). However, none of which com- pletely solve the problem of how to summarize such description in one single quantitative indicator, or a set of them. Existing indicators, such as those men- tioned by Lindgreen et al. (1996), are not readily available for a large number of countries, or else there is the lack of comparable cross-country data to construct such indicator. The empirical literature has opted for identifying banking crises as events, expressed through a binary variable, constructed with the help of cross-country surveys (Lindgreen et al., 1996; Caprio and Klingebiel, 2003). This will be our approach as well.
Banks are financial intermediaries whose liabili- ties are mainly short-term deposits and whose assets are usually short and long-term loans to businesses and consumers. When the value of their assets falls short of the value of their liabilities, banks become insolvent. Moreover, the nature of their business results in banks are institutions heavily leveraged. Then the banks, apart from the common risks face by a company, also face some specifics risks.
The most characteristic risk faced by banks is credit risk, which is the risk that borrowers become
unable or unwilling to service their debt. The empir- ical literature has highlighted a number of economic shocks associated with the materialization of this risk: cyclical output downturns, terms of trade dete- riorations, declines in asset prices such as equity and real estate (Gorton, 1998; Caprio and Klingebiel, 1997; Lindgreen et al., 1996; Kaminsky and Rein- hart, 1999).
Another typical bank risk is the interest rate risk. Because the asset side of bank balance sheets usually consists of loans of longer maturity at fixed interest rates, the rate of return on assets cannot be adjusted quickly enough, and banks must bear losses. Thus, a large increase in short-term interest rates is likely to be a major source of systemic banking sector prob- lems. In turn, the increase in short-term interest rates may be due to various factors, such as an increase in the rate of inflation, a shift towards more restrictive monetary policy that raises real rates, an increase in international interest rates, the removal of interest rate controls due to financial liberaliza- tion (Pill and Pradhan, 1995), the need to defend the exchange rate against a speculative attack (Velasco, 1987; Kaminsky and Reinhart, 1999).3
Another risk banks face is currency risk, when banks borrow in foreign currency and lend in domestic currency. In this case, an unexpected depreciation of the domestic currency threatens bank profitability. Foreign currency debt was a source of banking problems in Mexico in 1995, in the Nordic countries in the early 1990s, and in Tur- key in 1994 (Mishkin, 1996).
Liquidity risk is one of the most characteristic risks of banks. When bank deposits are not insured, deterioration in the quality of a bank’s asset portfo- lio may trigger a run, as depositors rush to with- draw their funds before the bank declares bankruptcy. Because bank assets are typically illiq- uid, runs on deposits accelerate the onset of insol- vency. The possibility of self-fulfilling runs makes banks especially vulnerable financial institutions. A run on an individual bank should not threaten the banking system as a whole unless partially informed depositors take it as a signal that other banks are also at risk (contagion).4 In these circum- stances, bank runs turn into a banking panic.
1558 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
A sudden withdrawal of bank deposits with effects similar to those of a bank run may also take place after a period of large inflows of foreign short-term capital, as indicated by the experience of a number of Latin American, Asian, and Eastern European countries in the early 1990s.
The literature, therefore, suggests a variety of mechanisms that can bring about banking sector problems. In what follows, we attempt to use our data set to identify which of these mechanisms have played a major role in the crises of the 1980s and early 1990s.
Moreover, we want to focus on the role of mon- etary policy. The existing literature on monetary policy has concentrated on issues different than financial stability (mainly price stability but also output stabilization). The impact of the monetary policy design on financial stability is related to the very much debated question of the relation between price stability and financial stability. The economic literature is divided as to whether there are synergies or a trade-off between them (Mishkin, 1996; Cukier- man et al., 1992; Fisher, 1933). If synergies existed between the two objectives it would seem safe to argue that the same monetary policy design which helps to achieve price stability also fosters financial stability (Schwart, 1995; Padoa-Schioppa, 2002; Issing, 2003). However, if there were a trade-off, it would be much harder to establish an a priori on the impact of price stability on financial stability.
There is some empirical analysis, albeit still scarce, on the impact of financial instability, and in particular of banking crisis, on a country’s mon- etary policy. In particular, Garcı́a-Herrero (1997) and Martinez-Peria (2000) find empirical evidence that money demand is stable in the long run in countries having experienced systemic banking crisis. However, to the best of our knowledge only one study is available on the reverse causality, Garcı́a-Herrero and del Rio (2003). They apply a multivariate logit model to estimate the relation- ship between monetary policy design and finan- cial instability, controlling for other relevant variables.
Most of the empirical analyses conducted to find the determinants of financial instability (Demirgüç- Kant and Detragiache, 1997, 1998, 2000; Eichen- green and Rose, 1998; Frydl, 1999; Glick and Hutchinson, 1999; Gourinchas et al., 1999; Hardy and Pazarbasioglu, 1998; Rossi, 1999; Eichengreen and Arteta, 2000) or to analyse the role of a partic- ular variable in explaining financial instability, as
exchange rate regimes (Eichengreen, 1997, 2000; Kaminsky and Reinhart, 1999; Mendis, 1998; Domaç and Martı́nez Peria, 2000) or monetary pol- icy strategies (Garcı́a-Herrero and del Rio, 2003) are based on a classical probit or logit methodology (Eichengreen and Arteta, 2000). The main short- coming of this analysis is that the outcome is sum- marized in one single rule that averages the contribution of each significant variable to explain financial crisis. This methodology fails to capture the variety of rules that fully describe the reality. For instance one variable can only be significant in some circumstances while insignificant in others; or one variable can have a positive effect for some intervals of values and negative for others even non-monotonically, as we will see in our application to financial crisis.
3. Financial ratios as explicatives variables of the
insolvency in insurance sector
On the other hand, the other objective variable is insolvency in insurance sector. In general financial terms, insolvency can be referred as the impossibil- ity or inability of a firm to pay its debts. A prior per- iod of insolvency could be got over, for example, by means of the postponement in the payments of the debts. If the firm is unable to overcome this first per- iod, it can become bankrupt. Therefore, bankruptcy could be interpreted as the culmination of the insol- vency process. In any case, in this work we are inter- ested in looking for the minimal set of financial ratios that could anticipate possible insolvencies due to permanent financial problems.
There are several reasons that could explain why an insurance firm becomes insolvent (Bannister, 1997) but all of them are reflected in the financial statements. These statements are specifically affected in all items related to:
– Liquidity: One of the most important questions in order to assure the proper functioning of any firm is the need of having sufficient liquidity. But in the case of an insurance firm, the lack of liquidity should not arise due to ‘‘productive activity inversion’’ which implies that premiums are paid in before claims occur. If an insurance firm cannot pay the incurred claims, the clients and public in general could lose faith in that company. – Profitability: Profits guarantee the present and future viability of any firm. In order to measure
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1559
this variable we will consider the results obtained and the cashflow. Sometimes it would be better use the second one because is less manipulated than the first one. – Solvency ‘‘in a strict sense’’: We have to take into account the risk exposure of the insurance firm (through premiums or incurred claims) and the real financial support (through technical pro- visions together with capital and reserves). Their comparison demonstrates the need of having suf- ficient shareholder’ funds and the need of com- plying correctly with the technical provisions to guarantee the financial viability of the insurance company.
Therefore, these three questions will be consid- ered in order to define the financial ratios that will be employed in our research.
Previous researchers applied to predict insurance insolvency in Spain are usually based on discrimi- nant analysis: López et al. (1994), Mora (1994), Martı́n et al. (1999) and Sanchis et al. (2003). There- fore same considerations about methodology as in the systemic crisis problem apply here.
4. The methodology: Main concepts of the Rough Set
(RS) Theory
RS Theory was firstly developed by Pawlak (1991) in the 1980s as a mathematical tool to deal with the uncertainty or vagueness inherent in a deci- sion making process. Though nowadays this theory has been extended (Greco et al., 1998, 2001), we refer to classical approach that does not order attri- bute domains as it assumes that different values of the same attribute are equally preferable and that only the predictive value of the attribute, as revealed by the data, will be factored into the model.
The extended approach handle dominance rela- tions, in addition to indiscernibility relations, incor- porating data about the ordering properties of the attributes analyzed, if these exit and are known. For example, if it were known that a financial ratio that was high was preferable to a financial ratio that was low, the firm with a high ratio could be consid- ered to be preferred over the firm with a low ratio and indeed all the values of the ratio could be con- sidered to be ordered. The resulted model is poten- tially more compact since some rules conflicts for certain cases are eliminated. Therefore, it uses addi- tional information to generate a simpler final model, but the classical approach makes a less restrictive
data assumption than does the extended approach (Mckee, 2000, p. 162).
The wisdom about traditional financial ratios has considered them as attributes with ordered domains. For example, a high value for a profitabil- ity ratio (net income to total assets) is preferable to a low value. This fact would imply the monotonicity
between the condition attributes (financial ratios) and the decision attribute. Increasing monotonicity occurs when the larger the value of the financial ratio, the better the value of the decision attribute. Decreasing monotonicity occurs when the smaller the value of the financial ratio, the better the value of the decision attribute. However, this assumption could be questioned for purposes of insolvency pre- diction. Mckee and Lensberg (1999) employed genetic programming to develop a bankruptcy pre- diction model. The model found that bankruptcy probability could be predicted as a complex func- tion of three ratios and, further, it was found that whether a high value or low value in one of the three ratios could be considered as good news or bad news
depended on the level of the other two ratios. There- fore, a higher ratio in the case of increasing mono- tonicity (for example, a high value for profitability ratio decreased business failure probability, except when the profitability ratio was unusually high because in that case it increased the predicted bank- ruptcy risk) was not always better and consequently, it would seem appropriate not to assume that finan- cial ratios have a dominance relation.
This reasoning can be also considered for the pre- diction of financial stability and, accordingly, this paper uses the classic RS theory based on indiscern- ibility relations for both financial problems.
Therefore, RS theory is related in some aspects to other tools that deal with uncertainty. However, RS approach is somewhat different to statistical proba- bility, which deals with random events in nature or fuzzy set theory, which deals with objects that may not belong only to one category but may belong to more than one category by differing degrees. On the contrary, RS theory deals with the uncertainty pro- duced when some objects described by the same data or knowledge (so, they are indiscernible) can be classified into different classes, that is, there is not a unique inclusion of these indiscernible objects. This fact prevents their precise assignment to a set. These differences show one of the main advantages of RS theory: an agent is not required to assign pre- cise numerical values to express imprecision of his knowledge, such as probability distributions in
1560 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
statistics or grade of membership in fuzzy set theory (Nurmi et al., 1996).
This section presents some concepts of RS The- ory following Pawlak’s reference and some remarks by Slowinski (1993) and Dimitras et al. (1999).
The philosophy of this approach is based on the assumption that with every object of the universe we are considering we can associate knowledge, data. Knowledge is regarded as ability to classify objects. Therefore knowledge consists of a family of various classification patterns of a domain of interest. Objects described by the same data or knowledge are indiscernible in view of such knowledge. The indiscernibility relation leads to mathematical basis for the RS Theory. Vague information causes indis- cernibility of objects by means of data available and, as a result, this prevents their precise assign- ment to a set. Intuitively, a rough set is a set or a subset of objects that cannot be expressed exactly by employing available knowledge. If this informa- tion or knowledge consists of a set of objects described by another set of attributes, we consider a rough set as a collection of objects that, in general, cannot be precisely characterized in terms of the val- ues of the set of attributes.
RS Theory represents knowledge about the objects as a data table, that is, an information table. Rows of which are labelled by objects (states, pro- cesses, firms, patients, candidates, . . .) and columns are labelled by attributes. Entries of the table are attribute values. Therefore, for each pair object- attribute, x � q, there is known a value called descriptor, f(x,q). The indiscernibility relation would occur if for two objects, x and y, all their descriptors in the table have the same values, that is if, and only if, f(x,q) = f(y,q).
4.1. Approximation of sets, accuracy and quality of
approximation
In general, all properties of rough sets are not absolute, but are related to what we know about them. Indiscernible objects by means of attributes prevent their precise assignment to a class. There- fore, some categories (subsets of objects) cannot be expressed exactly by employing available knowl- edge and, consequently, the idea of approximation of a set by other sets is reached. A rough set is a pair of a lower and an upper approximation of a set in terms of the classes of indiscernible objects. That is, it is a collection of objects that, in general, cannot be precisely characterized in terms of the values of
the set of attributes, while a lower and an upper approximation of the collection can be. Therefore, each rough set has boundary-line cases, that is, objects that cannot be classified certainly as mem- bers of the set or of its complement and can be rep- resented by a pair of crisp sets, called the lower and the upper approximation. The lower approximation consists of all objects that certainly belong to the set and can be certainly classified as elements of that set, employing the set of attributes in the table (the knowledge we are considering). The upper approximation contains objects that possibly belong to the set and can be possibly classified as elements of that set using the set of attributes in the table. The boundary or doubtful region is the difference between the lower and the upper approximation and is the set of elements that cannot be certainly classified to a set using the set of attributes. There- fore, the borderline region is the undecidable area of the universe, that is, none of the objects belong- ing to the boundary can be classified with certainty into a set or its complement as far as knowledge is concerned.
Inexactness of a set is due to the existence of the boundary. The greater the doubtful region of a set is, the lower the accuracy of that set. The accuracy
of approximation is defined as the quotient between the cardinality of the lower approximation and the cardinality of the upper one. This ratio expresses the percentage of possible correct decisions when classifying objects employing knowledge available. Therefore, using the lower and the upper approxi- mation we can define precisely those subsets that cannot be expressed exactly using the available attributes.
Because we are interested in classifications, the quality of classification is defined as the quotient between the addition of the cardinalities of all the lower approximations of the classes in which the objects set is classified, and the cardinality of the objects set. It expresses the percentage of objects which can be correctly classified to classes employ- ing the knowledge available.
4.2. Reduction and dependency of attributes
A fundamental problem in the rough set approach is discovering dependencies between attri- butes in an information table because it enables to reduce the set of attributes removing those that are not essential (unnecessary) to characterize knowledge. This problem will be referred to as
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1561
knowledge reduction or, in more general terms, as a feature selection problem. The main concepts related to this question are the core and the reduct. A reduct is the minimal subset of attributes which provides the same quality of classification as the set of all attributes. If the information table has more than one reduct, the intersection of all of them is called the core and is the collection of the most relevant attributes in the table.
4.3. Decision rules
An information table which contains condition and decision attributes is referred as a decision table. A decision table specifies what decisions (actions) should be undertaken when some conditions are sat- isfied. So a reduced information table may provide decision rules of the form ‘‘if conditions then
decisions’’. These rules can be deterministic when the rules
describe the decisions to be made when some condi- tions are satisfied and non-deterministic when the decisions are not uniquely determined by the condi- tions so they can lead to several possible decisions if their conditions are satisfied. The number of objects that satisfy the condition part of the rule is called the strength of the rule and is a useful concept to assign objects to the strongest decision class when rules are non-deterministic.
The rules derived from a decision table do not usually need to be interpreted by an expert as they are easily understandable by the user or decision maker. The most important result in this approach is the generation of decision rules because they can be used to assign new objects to a decision class by matching the condition part of one of the deci- sion rule to the description of the object. Therefore, rules can be used for decision support.
RS Theory can analyse several multiattribute decision problems. It is especially well suited to classification problems. One of these problems is multiattribute classification problem which consists of the assignment of each object, described by values of attributes, to a predefined class or category.
We want to mention that rough set analysis has been performed using ROSE software provided by the Institute of Computing Science of Poznan Uni- versity of Technology. Any personal computer with a link to internet can access to the web http:// idss.cs.put.poznan.p1/site/rose.html where ROSE software and its manual can be downloaded. More
details about this software are given in Predki et al. (1998) and Predki and Wilk (1999).
5. Empirical results
As we have previously mentioned, RS approach is especially well suited to classification problems. One of these problems is a multiattribute classifica- tion problem which consists of the assignment of each object, described by values of attributes, to a predefined class or category. The two financial problems we are going to tackle are examples of this kind of problems. In financial instability prediction (Model 1), we try to assign countries described by a set of macroeconomic variables to a category (crisis or financial stability). In business failure prediction (Model 2), we try to assign firms (objects) described by a set of financial ratios (attributes) to a category (failed or ‘‘healthy’’ firm). In this stage of our research, we have proceeded to the election of the data and variables that will be used to develop our models.
5.1. The data
As for the data employed in Model 1, we have employed a sample of 79 countries in the period 1981–1999 (annual data). The dependent variable can be defined in this way: Systemic and non- systemic banking crises dummy equals one during episodes identified as in Caprio and Klingebiel (2003). The independent variables included are dic- tated by the theory on the determinants of banking crisis. We provide a detailed list of variables and sources in Data Appendix A. We included two types of variables in our estimations: macroeconomic variables and financial variables. Among the macro- economic variables we include: the real growth of GDP, the level of real GDP per capita, the inflation rate and the real interest rate to capture the external conditions that countries face. We have employed qualitative and quantitative variables. The possibil- ity of using both kinds of variables is one of the advantages of this methodology.
As for the data in Model 2, we have employed a sample of Spanish firms used by Sanchis et al. (2003). This data sample consists of non-life insur- ance firm data 5 years prior to failure. The firms were in operation or went bankrupt between 1983 and 1994. In each period, 72 firms (36 failed and 36 non-failed) are selected. As a control measure, a failed firm is matched with a non-failed one in
1562 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
terms of industry and size (premiums volume). In our analysis we have used data 1 year prior to fail- ure to obtain the decision rules and we have tested the rules with data from years 2, 3, 4 and 5 (Dimi- tras et al., 1999). Provided that we are looking for those financial ratios that could help the decision maker to anticipate possible insolvencies, we have to mention that our definition for insolvency (objec- tive variable) is made in strict sense, therefore insol- vent group consists of those firms that have been taken over by Spanish ministry of economy. So, the insolvent group consists of those firms that have disappeared due to permanent financial problems. This way we avoid working with firms that have temporary financial problems or that have disap- peared voluntarily. On the other hand we have checked that the firms of the solvent group have gone on working for several years after the sample period because it is possible that a firm, that has not been taken over by Spanish ministry of econ- omy can become bankrupt.
As for the variables in Model 2, we have to men- tion that each firm is described by 17 financial ratios that have come from a detailed analysis of the vari- ables, previous bankruptcy studies for insurance companies and our preferences and knowledge. These ratios (see Table A.1 in Appendix A) have been divided into four groups: A group contains ratios related to financial position; B group contains ratios related to operating account; C group con- tains ratios related to earnings and cash-flow and, finally, D group contains ratios related to provi- sions. We have to draw particular attention to the fact that special financial characteristics of insur- ance companies require general financial ratios as well as those that are specially proposed for evaluat- ing insolvency of insurance sector.
We are going to analyze results for the two mod- els separately.
5.2. Prediction of financial instability
If we developed a model and we test it with the same sample, the results obtained could be condi- tioned. So in order to avoid it, for this first model we have formed a training set, and a holdout sample to validate the obtained decision rules, i.e., the test set. Both sets have been randomly selected. The training information table consisted of 421 data from 79 countries in the period 1981–1997 (annual data) described by the variables explained in Section 4, and assigned to a decision class (crisis – 1 or not –
0). We have 293 objects for class 0 and 128 objects for class 1. The test information table consisted of 100 data described by the same variables in the per- iod 1997–1999 (36 objects for class 1, and 64 objects for class 0). So the training information table was entered into an input file in ROSE.
We have recoded the continuous variables into qualitative terms (low, medium, high and very high) with corresponding numeric values such us 1, 2, 3 and 4. This recoding has been made dividing the ori- ginal domain into subintervals. This recoding is not imposed by the RS theory but it is very useful in order to draw general conclusions from the ratios in terms of dependencies, reducts and decision rules (Dimitras et al., 1999).
The definition of the boundary values can influ- ence results of the RS analysis, in particular the quality of classification. There is not a general way to define the optimal boundary values. It is usually done by experts according to their experience, knowledge, habits or conventions, as in financial problems (Slowinski and Zopounidis, 1995; Dimi- tras et al., 1999). If there is not an expert to recode the variables that could follow their experience or standards of financial analysis, it is deemed desir- able to avoid subjective inputs to the extent possi- ble. Accordingly, the four subintervals are based on the quartiles for the actual variable values (year 1) for the whole sample because percentiles are fre- quently used in scientific researches to divide a domain into subintervals (Laitinen, 1992; Mckee, 2000). We choose this method because we have no a priori knowledge about some other partitions with more economic sense. Of course it could be justified to use fewer intervals in some variables and more in other cases, but we consider this type of discretiza- tion as a good first step in terms of the subsequent interpretation of the results. We want to mention that there are other approaches to discretize vari- ables. In fact, ROSE software has implemented an entropy-based method to get attributes with discrete domains. However, though we know that if we had employed other approaches, we would have obtained other models, this simple form of discreti- zation based on the quartiles of the distribution has provided good results in the validation tests. Never- theless, in future researches we will try other discret- ization strategies (see Table 1).
The first results obtained from RS analysis of the coded information table were: the approximation of the decision classes and their quality of classification were equal to one:
Class Number of objects
Lower approximation
Upper approximation
Accuracy of approximation
Quality of classification
0 (non-crisis) 293 293 293 1 1 1 (crisis) 128 128 128 1 1
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1563
These results show that the data are very well dis- criminated among them (so the boundary region is empty for the two decision classes). Yet, the fact that a set of attributes discriminates well objects does not necessarily imply that a good classifier can be constructed on this set.
The core is consisted of four attributes: Inflation, Domestic Credit Growth, Real GDP per capita, and Bank Foreign Liabilities to Foreign Assets, which represent the most relevant attributes in the table. This result shows the importance of these four variables to forecast financial instability in a country. Moreover, they are well in line with previ- ous research. Demirgüç-Kant and Detragiache (1997) found that crisis tend to erupt when growth is low and inflation is high. Eichengreen and Arteta (2000) discover among the robust causes of emerg- ing banking crisis a rapid domestic credit growth and large bank liabilities relative to reserves. We have obtained 19 reducts from the table which con- tain 9–10 attributes. We have selected the reduct consisted of Central Bank Independence, Inflation, Domestic Credit Growth, Real GDP growth, Bank Foreign Liabilities to Foreign Assets, Real GDP per capita, World Growth, Real Interest Rate and Previous Crisis. The model has been selected attending to its better performance in terms of cor- rectly classified firms as well as in terms of eco- nomic interpretation. So we have obtained a reduced table to obtain the decision rules. The strategy we have followed to obtain the decision
Table 1 List of subintervals (quartiles)
Variable 1st 2nd
CRECIM. (�1, 2.9] (2.9, CBANKINDEP (�1, 0.37] (0.37 REAL INTEREST (�1, 0.33] (0.33 DOM.CREDIT.GROWTH (�1, 7.90] (7.90 BANK CASH REV. (�1, 0.02] (0.02 FOR_LIAB_REV (�1, 0.36] (0.36 GDP_GROWTH (�1, 2] (2,4 INFLATION (�1, 2.81] (2.81 %NETKFLOWS (�1,�0.003] (�0. GDPPERHEAD (�1, 3144] (314
rules consists in the generation of a minimal subset of rules covering all the objects from the decision table. This strategy is implemented in the ROSE software. We have obtained 116 deterministic rules (63 for class 0 and the other ones for class 1). To interpret the rules, we have only selected the strongest rules (3.12%) for each decision class, thus we have only considered 40 rules. This way we have covered 85.5% objects in the table. Therefore, Table A.2 (see Appendix A) shows the strongest rules.
The decision model has been tested (using the 116 rules) on data from the testing test, i.e., on the 100 firms that have not been used to estimate the algo- rithm. The classification accuracy in percent of cor- rectly classified firms by this second set of rules is: 80%. This result is quite satisfactory comparing with previous analysis. Demirgüç-Kant and Detragiache (1997) obtained similar results and in general the corrected R-square is well below this percentage (Eichengreen and Arteta, 2000; Domaç and Martı́nez Peria, 2000, etc.).
Focusing in the role of the design of monetary policy in determining financial stability, we can observe that in 18 of the 19 reducts at least one of the three variables related with the design of the monetary policy (Exchange, Independence, and Monetarypol) appears. Given the co-linearity between them it is not strange that generally only one of them was chosen in each model. This result confirms the idea that the design of the monetary
3rd 4th
3.5] (3.5,4.6] (4.6,+1) ,0.59] (0.59,0.82] (0.82,+1) ,3.64] (3.64,6.15] (6.15,+1) ,15.88] (15.88,28.47] (28.47,+1) ,0.06] (0.06,0.15] (0.15,+1) ,0.52] (0.52,0.65] (0.65,+1)
] (4,6] (6,+1) ,7.25] (7.25,16.04] (16.04,+1) 003,0.896] (0.896,4724] (4724,+1) 4,8180] (8180,17,392] (17,392,+1)
Table 2 Number of rules that use the INDEPEN variable
Quartiles 1 2 3 4 Total
Non-crisis/crisis
D = 0 6 7 9 8 30 % 20.0 23.3 30.0 26.7 100.0 D = 1 5 7 7 3 22 % 22.7 31.8 31.8 13.6 100.0
Number of units classified by these rules
D = 0 74 29 34 55 192 % 38.5 15.1 17.7 28.6 100.0 D = 1 9 29 16 5 59 % 15.3 49.2 27.1 8.5 100.0
5 Non-linear probit/logit models can be developed by perform- ing a non-monotonic transformation of the variables or a transformation into nominal categorical variables.
1564 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
policy is a relevant variable in order to explain financial stability, as it was suggested in the paper of Garcı́a-Herrero and del Rio (2003).
Moreover, between the four variables that belong to the core there are two directly related with mon- etary policy: the level of inflation and the domestic credit growth.
Focusing in the chosen model we can see that the Central Bank Independence variable enters in 52 of the 116 rules. This represents the 45% of the total rules. However, in terms of objects covered by the rules (strength) represents the 59.6%. As we can see in Table 2, the percentage in terms of classified units in the rules for no crisis is 66% and in the crisis group 46%. So, results suggest that this variable play an important role in determining financial crisis.
But as shown in Table 2 not always a higher degree of independence is associated with financial stability. Seventy-four units with a degree of inde- pendency belonging to the lowest quartile showed financial stability, while 55 units in the highest quar- tile also showed stability. In these two extremes of the distribution we can see that only a reduced num- ber of crisis are associated, indicating that a clear independence or a clear dependency is associated almost always with financial stability. On the con- trary, the crisis are clearly associated with levels of independence in the second and third quartile of the distribution highlighting that no clear definition of the monetary policy objectives is a factor that contributes to the financial crisis. In other words, it is more important for financial stability that finan- cial agents know the reaction function on monetary policy rather than the function in itself. This result is independent on the level of inflation since there is a control variable accounting for this. A way to see
that central bank independence is not picking up the effect of the inflation variable is looking at the correlation between both variables. The coefficient of correlation, calculated with the original continu- ous variables is very low, 0.05. This coefficient could be influenced by outliers that play a different role when we use discrete variables. So we calculate a measure using the discretized variables and, although the result is not so strong (0.55% of cases show a distance lower than two in absolute terms) we can conclude that there are no clear correlation between central bank independence and inflation in our sample.
This result differs from the results obtained in previous researches by Garcı́a and del Rı́o who found a negative relationship between the degree of central bank independence and the emergence of financial crisis. A multivariate linear probit and logit models are used, respectively.5 Linear models do not have the flexibility to capture non-monotonic relationships as we have found, so it is reasonable that results differ as we have a much more flexible methodological approach.
5.3. Prediction of the insolvency of Spanish
non-life insurance companies
The information table for year 1 which consisted of 72 firms described with 17 ratios and assigned to a decision class (healthy – 1 or not – 0) was entered into an input file in ROSE. We have followed the same steps as in the first model. Therefore, the train- ing information table was entered into an input file in ROSE. In this model we have recoded the finan- cial ratios into qualitative terms (low, medium, high and very high) with corresponding numerical values such us 1, 2, 3 and 4 using the quartiles for the values of each variable. We want to make the same remarks described in Section 5.1 related to the selection of the discretization method (see Table 3).
The first results obtained from RS analysis of the coded information table were that the approxima- tion of the decision classes and their quality of clas- sification were equal to one:
Class Number of firms
Lower approximation
Upper approximation
Accuracy of approximation
Quality of classification
0 (failed firms) 36 36 36 1 1 1 (healthy firms) 36 36 36 1 1
Table 3 List of subintervals (quartiles) for financial ratios
Ratio 1� 2� 3� 4�
A1 (�1, 0.155] (0.155,0.385] (0.385,0.68] (0.68,+1) A5 (�1,�0.29] (�0.29,�0.005] (�0.005,0.195] (0.195,+1) A6 (�1, 0.52] (0.52,0.705] (0.705,0.875] (0.875,+1) B3 (�1, 0.325] (0.325,0.55] (0.55,0.96] (0.96,+1) B6 (�1, 0.07] (0.07,0.495] (0.495,1.35] (1.35,+1) B7 (�1, 0.635] (0.635,1.435] (1.435,3.185] (3.185,+1) B8 (�1, 0.775] (0.775,1.465] (1.465,2.485] (2.485,+1) C1 (�1,�0.04] (�0.04,0] (0,0.04] (0.04,+1) C4 (�1, 0.005] (0.005,0.13] (0.13,0.41] (0.41,+1) C5 (�1, 0.005] (0.005,0.095] (0.095,0.33] (0.33,+1) C6 (�1,�0.245] (�0.245,�0.025] (�0.025,0.05] (0.05,+1) C7 (�1,�0.03] (�0.03,0.01] (0.01,0.06] (0.06,+1) D1 (�1, 0.04] (0.04,0.295] (0.295,0.965] (0.965,+1) D4 (�1, 0.08] (0.08,0.785] (0.785,1.63] (1.63,+1) D6 (�1, 0.07] (0.07,0.565] (0.565,2.82] (2.82,+1) D7 (�1, 0] (0,0.01] (0.01,0.46] (0.46,+1) D8 (�1,0] (0,0.355] (0.355,0.435] (0.435,+1)
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1565
These results obtained show that the firms are very well discriminated among them (consequently, the boundary region is empty for the two decision classes). This fact can be explained because we have employed data 1 year prior to bankruptcy and therefore ratios for failed firms are substantially dif- ferent from the ratios of the healthy ones but, as we have previously mentioned for model 1, the fact that a set of attributes (ratios) discriminates well objects (firms) does not necessarily imply that a good clas- sifier can be constructed on this set. Another result is that none of the attributes are indispensable for the approximation of the two decision classes (so the core was empty). We have obtained 452 reducts from the table which contain 4–8 attributes. This result means that at least 9 attributes are redundant (and, therefore, they could be eliminated). Conse- quently, this fact shows the strong support of this approach in feature selection. The list of the fre- quencies of the attributes in the 452 reducts is the following one:
Attr. A1 A5 A6 B3 B6 B7 B8 C1 C
Freq 184 209 142 86 95 152 262 131 2
As we can see the ratios that have the highest fre- quency of occurrence (more than 40%) in reducts are B8, C5, D4, C4, A5, and C6. This fact indicates that these variables are highly discriminatory between solvent and insolvent firms in our sample confirming the importance, from a solvency view- point, of these questions: sufficient liquidity, correct rating, proper reinsurance and the need of having enough technical provisions.
These results are broadly in line with those obtained by Sanchis et al. (2003) that also high- lighted the role of liquidity, and correct rating and the need to have enough provisions when predicting insurance bankruptcy. However, reinsurance plays here a role that was not found by those authors.
We have selected the reduct consisted of A5, A6, B6, B8, C6, D8 taking into account two questions: the reduct should have a small number of attributes as possible and it should have the most significant attributes in our opinion for the evaluation of the companies (at least the selected reduct should
4 C5 C6 C7 D1 D4 D6 D7 D8
27 259 199 140 120 232 130 65 178
1566 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
contain one ratio of each group A, B, C or D and, if possible, the ratios with highest frequency of occur- rence within each group). So we have obtained a reduced table (see Table A.3 in Appendix A) (only six financial ratios) to obtain the decision rules. We have followed the same strategy to generate the decision rules as in the previous model. We have obtained 30 deterministic rules (see Table A.4 in Appendix A).
We want to mention that the large number of reducts obtained implies a very detailed analysis of the reducts to choose one of them in order to gener- ate the smallest number of stronger rules. Therefore, the first strategy we followed was to generate the minimal set of the decision rules considering the whole data set. We obtained 23 rules. Some of them were supported by only one object or by two objects and the classification accuracy of the 23-rules model was significantly worse. Consequently, we have decided to analyze the reducts and to obtain better classification results though the model employed contains more decision rules.
The 30-rules decision model has been tested on data from 2, 3, 4 and 5 years before the actual ratio values (year 1 or year prior to bankruptcy) that were used to obtain the decision rules (Dimitras et al., 1999). The classifications accuracies in percent of correctly classified firms by the set of 30 rules for the 5 years prior to the reference year (year 1) are shown in Table 4 at the end of Section 5.4.
5.4. Comparison of Rough Set approach with
discriminant analysis
We have compared rough set model with Dis- criminant Analysis (DA). Briefly, DA is a statistical technique used to classify objects into distinct groups on the basis of their observed characteristics. Basically, a linear discriminant function is devel- oped which will compute a ‘‘score’’ for an object. This function is a weighted linear combination of the object’s observed values on discriminating char- acteristics. These weights represent, essentially, the relative importance and impact of the various
Table 4 RS and DA results
Year 1 Year 2
Rough Set 100% 80.56% Linear discriminant function 81.86% 81.27% Quadratic discriminant function 67.48% 59.78%
characteristics. On the basis of its discriminant score, an object is then classified. Altman et al. (1981) provides a detailed description of DA and its financial applications.
RS and DA analysis require some assumptions but the ones required by RS approach are much weaker that the ones required by DA approach. This way, the discriminant analysis requires these restrictive assumptions: each group follows a multi- variate normal distribution, the covariance matrices of each group are identical, and, the mean vectors, covariance matrices, prior probabilities and misclas- sifications costs are known. These theoretical assumptions constitute ideal conditions in which DA should be applied and if they are violated, the methodology can be applied but the results obtained may be erroneous or inferior. Therefore, one advan- tage of RS theory is that it does not need restrictive assumptions and, consequently, it is more realistic which is manifested by better results of this approach.
Unfortunately in practice, violations of statistical assumptions of DA analysis occur regularly. How- ever, although these assumptions are not satisfied in the case of financial ratios, DA has provided good empirical results in real problems dealing with this kind of variables. This explains why this tech- nique is one of the most used in prediction problems and the reasons why it has been chosen.
To compare the two methods we have used two discriminant functions: a linear function and a qua- dratic function (Sanchis et al., 2003). Both functions have been derived using the original data table instead of the recoded one. The quadratic one has been developed due to covariance matrices are not equal, so results obtained by the linear model could be questioned. The classifications accuracies (prior probabilities and misclassifications costs are set for 0.5) in percent of correctly classified firms by the two discriminant functions and RS model for the 5 years prior to the reference year (year 1) are:
In general, results for the rough set model and linear discriminant model are quite similar and RS model has outperformed the quadratic model.
Year 3 Year 4 Year 5
76.36% 75.50% 65.85% 76.79% 75.34% 77.78% 73.02% 53.46% 75%
Table A.1 List of ratios
Ratio Definition
A1 (Capital + Reserves)/Total liabilities A5 Working capital/Total assets A6 Current assets/Total assets B3 Net premiums/Total assets B6 Provisions for benefit/Claims incurred B7 Net premiums/(Capital + Reserves) B8 (Capital + Reserves + Technical provisions)
/Earned premiums C1 Earnings before taxes/(Total liabilities �
Capital � Reserves) C4 Cash-flow/(Capital + Reserves) C5 Cash-flow/(Total liabilities � Capital � Reserves) C6 Accrued results/(Subscribed capital � Accrued results) C7 Earnings before taxes/(Capital + Reserves) D1 Provisions for benefit/Earned premiums D4 Technical provisions/Earned premiums D6 Technical provisions/(Capital + Reserves) D7 Technical provisions of cession/(Capital + Reserves) D8 Technical provisions for current risks/Earned premiums
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1567
6. Conclusions
We have presented a new approach to predict financial stability in a country and to predict insur- ance insolvency using rough sets. The results obtained for both models are quite satisfactory.
Through the exposition we have mentioned some advantages of this approach so we can conclude that this method is an effective tool for supporting managerial decision making in general. In the light of the experiments carried out, this method is a competitive alternative to existing prediction models for both problems that undoubtedly make it attrac- tive for application to the field of business classification.
Our empirical results in the insolvency prediction case show that rough set model offers better predic- tive accuracy than the quadratic discriminant model we have developed. The results obtained by the lin- ear discriminant model are comparable to the ones obtained by RS model. However, RS model does not require the pre-specification of a functional form, or the adoption of restrictive assumptions about the characteristics of statistical distributions of the variables and errors of the model. In short, by its nature, rough set approach makes working with imprecise variables possible. The flexibility of the decision rules with changes of the models over the time allows us to adapt them gradually to the appearance of new cases representing changes in the situation. Consequently, for some real-world problems, the method we have presented is more attractive than the discriminant analysis showing that it is a very robust technique especially in the areas of forecasting and classification decision problems.
In practical terms, the decision rules generated can be used to preselect companies or countries to examine more thoroughly, quickly and inexpen- sively, thereby, managing the financial user’s time efficiently. They can also be used to check and monitor insurance firms or countries as a ‘‘warning system’’ for insurance regulators, investors, man- agement, financial analysts, banks, auditors, policy holders and consumers.
Acknowledgements
We want to thank to the Institute of Computing Science of Poznan University of Technology for providing ROSE software and for helping us with the rough set analysis and to the anonymous
referees for the comments that have really improved this paper.
Appendix A. Data appendix
A.1. Financial crisis database
A.1.1. Dependent variable
Systemic and non-systemic banking crises dummy:
Equals one during episodes identified as in Caprio and Klingebiel (2003). They present information on 117 systemic banking crises (defined as much or all of bank capital being exhausted) that have occurred since the late 1970s in 93 countries and 51 smaller non-systemic banking crises in 45 coun- tries during that period. The information on crises is cross-checked with that of Domaç and Marti- nez-Peria (2000) and with IMF staff reports and financial news.
A.1.2. The Objective variables
* Monetary policy strategies: These variables (Exchange rate target, Monetary policy target) are dummies. The exchange rate target takes four values depending on the exchange rate regime: free float- ing, managed floating, pegged currencies and cur- rency board. The Monetary policy target equals one during periods in which targets were based on monetary aggregates, two when the objective was inflation, three when the two variables are into the objective function and zero in other cases, according
1568 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
to the chronology of the Bank of England survey of monetary frameworks, in Mahadeva and Sterne (2000). Since it provides a chronology for the 1990s, we have complemented it with information from other sources for the previous years. Regard- ing exchange rate arrangements, we use classifica- tions of exchange rate strategies in Reinhart and Rogoff (2002), Kuttner and Posen (2001), and Berg et al. (2002) for Latin America countries. Data for monetary and inflation targets were complemented with the information taken from Kuttner and Posen (2001) and Carare and Stone (2003). It should be
Table A.2 Decision rules for Model 1
# Rule Indep DCG Bank cash Liabil. Real GDP
Inflat. Inte
1 1 2 2 3 2 3 4 1 4 1 3 4 5 1 2 6 2 7 3 3 8 4 9 3 3 3 10 4 11 4 1 12 1 1 13 4 14 2 2 15 4 2 2 16 3 17 4 18 3 2 3 19 4 2 20 2 3 21 2 3 22 2 1 23 3 3 24 4 25 2 1 4 26 4 27 3 4 28 2 3 29 2 1 3 30 1 4 31 2 1 2 2 32 4 33 4 34 2 35 2 4 36 1 4 37 3 4 38 1 39 3 4 40 4 2 4
noted that some judgement has gone into the classi- fication of regimes.
* Central Bank Independence measures to what extent the central banks are legally independent according to their charters, following the approach of Cukierman et al. (1992). This variable goes from 0 (least independent) to 1 (most independent) and is taken from Cukierman et al. (1992), for the 1970s and 1980s.). For the 1990s, Mahadeva and Sterne (2000) and Cukierman et al. (2002). The index of independence is assumed to be constant through every year of each decade.
rest NKF GDP p/cap.
Prev. crisis World GDP
Class Streng.
3 0 21 0 30
0 0 20 0 0 9 0 0 15
2 0 0 25 1 0 10
3 0 0 10 0 9
4 1 3 0 10 0 0 9
1 0 0 13 2 1 0 10
0 0 16 0 11
3 4 0 15 3 1 0 11
0 10 0 0 9
1 0 0 9 1 1 5 3 2 1 5
1 1 1 6 3 3 1 10
2 1 4 1 4 1 4
1 3 1 6 2 1 5
1 4 3 1 7
1 4 1 2 1 1 4 1 1 1 1 4
4 3 1 6 0 3 1 4
1 1 4 4 3 1 5 1 2 1 4 2 1 5
1 4
Table A.3 Decision table for Model 2
Firms A5 A6 B6 B8 C6 D8 D
1 2 2 3 3 2 2 0 2 3 4 1 3 2 3 0 3 1 3 1 3 1 4 0 4 2 4 4 3 3 4 0 5 3 3 2 2 3 3 0 6 1 4 3 3 1 2 0 7 1 4 4 1 4 4 0 8 1 4 3 3 2 4 0 9 3 2 4 3 2 4 0 10 1 3 3 3 1 2 0 11 1 3 3 3 1 4 0 12 1 3 4 1 1 1 0 13 1 4 2 4 2 2 0 14 2 2 3 3 2 4 0 15 2 4 2 4 1 4 0 16 1 1 2 1 1 2 0 17 1 2 3 1 1 4 0 18 2 3 2 2 4 2 0 19 1 3 3 3 1 4 0 20 2 4 3 2 2 2 0 21 3 3 1 3 1 3 0 22 2 3 1 2 3 3 0 23 2 2 4 1 4 1 0 24 1 4 3 4 3 3 0 25 1 1 2 2 2 2 0 26 4 2 4 1 1 1 0 27 3 3 1 1 1 3 0 28 1 3 1 4 3 1 0 29 2 4 4 2 2 3 0 30 4 1 1 1 4 1 0 31 3 3 1 2 4 1 0 32 2 4 3 2 1 4 0 33 3 1 1 1 2 1 0 34 2 3 4 1 3 1 0 35 2 2 4 1 4 3 0 36 4 1 4 1 4 1 0 101 4 1 2 2 3 2 1 102 4 4 4 4 3 1 1 103 4 2 1 3 4 1 1 104 2 1 2 2 4 2 1 105 4 1 4 4 4 3 1 106 2 2 2 4 2 4 1 107 4 1 3 4 4 2 1 108 3 1 3 4 1 4 1 109 2 2 2 2 3 2 1 110 3 1 2 2 2 2 1 111 3 1 2 4 2 4 1 112 4 2 4 1 4 1 1 113 4 1 2 2 3 2 1 114 3 2 2 4 2 2 1 115 1 3 3 4 2 4 1 116 2 2 3 3 1 2 1 117 3 2 3 2 1 2 1 118 4 2 4 3 4 3 1 119 1 1 2 2 1 2 1 120 3 2 3 2 3 4 1 121 2 3 1 1 2 1 1
(continued on next page)
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1569
Table A.3 (continued)
Firms A5 A6 B6 B8 C6 D8 D
122 3 3 1 4 2 1 1 123 4 4 1 3 3 3 1 124 1 1 2 4 3 3 1 125 1 4 2 4 4 3 1 126 4 4 4 4 3 3 1 127 4 4 4 1 4 3 1 128 2 3 1 3 1 1 1 129 3 4 3 4 3 4 1 130 3 3 1 2 3 1 1 131 4 1 1 1 3 1 1 132 3 2 2 2 2 4 1 133 4 2 1 1 3 1 1 134 3 3 1 4 3 3 1 135 4 1 1 1 4 3 1 136 4 1 1 3 4 1 1
1570 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
A.1.3. Control variables
A.1.3.1. Macroeconomic variables. * Inflation: Per- centage change in the GDP deflator. Source: Inter- national Monetary Fund, International Financial Statistics, line 99bir.
Table A.4 Decision rules for Model 2
Rules A5 A6 B6 B8 C6 D8 D
1 1 3 1 0 2 2 4 0 3 1 4 3 0 4 1 3 0 5 3 2 0 6 1 2 2 0 7 1 1 4 1 0 8 3 4 0 9 1 1 0 10 2 4 0 11 2 3 0 12 4 2 0 13 4 1 0 14 1 1 0 15 3 1 0 16 4 3 1 17 3 4 1 18 2 2 1 19 4 3 1 20 4 3 1 21 1 4 2 1 22 1 2 3 1 23 2 2 1 24 2 1 1 1 25 3 1 2 1 26 2 2 1 1 27 4 2 4 1 28 2 3 1 1 29 2 3 1 1 30 3 4 2 1
* Real Interest Rate: Nominal interest rate minus inflation in the same period, calculated as the per- centage change in the GDP deflator. Source: Inter- national Monetary Fund, International Financial Statistics. Where available, money market rate (line
ecision Strength Firms
5 6, 10, 11, 17, 19 5 4, 23, 29, 34, 35 3 6, 8, 24 2 21, 27 5 1, 2, 8, 9, 14 1 25 2 30, 36 2 18, 31 4 7, 12, 16, 17 5 4, 15, 20, 29, 32 3 5, 22, 29 5 2, 8, 13, 20, 29 1 26 2 3, 28 2 27, 33 6 105, 118, 123, 126, 127, 135 6 108, 111, 114, 122, 129, 134 4 109, 117, 120, 132 4 103, 118, 123, 136 7 101, 102, 113, 123, 126, 131, 133 2 104, 107 2 124, 125 4 106, 109, 114, 132 2 121, 128 1 110 1 119 3 103, 112, 118 2 116, 128 1 130 1 115
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1571
60B); otherwise, the commercial bank deposit inter- est rate (line 60l); otherwise, a rate charged by the Central Bank to domestic banks such as the dis- count rate (line 60).
* Net Capital Flows to GDP: Capital Account + Financial Account + Net Errors and Omissions. Source: International Monetary Fund, International Financial Statistics, lines (78bcd + 78bjd + 78cad).
* Real GDP per capita in 1995 US dollars: This variable is expressed in US dollars instead of PPP for reasons of data availability. GDP per capita in PPP was available only for two points in time. Source: The World Bank, World Tables; and EBRD, Transition Report, for some transition countries.
* Real GDP growth : Percentage change in GDP Volume (1995 = 100). Source: International Mone- tary Fund, International Financial Statistics (line 99bvp) where available; otherwise, The World Bank, World Tables; and EBRD, Transition Report, for some transition countries.
* World Real GDP growth: Percentage change in GDP Volume (1995 = 100). Source: International Monetary Fund, International Financial Statistics (line 99bvp) where available; otherwise, The World Bank, World Tables; and EBRD, Transition Report, for some transition countries.
A.1.3.2. Financial variables. * Domestic Credit
growth: Percentage change in domestic credit, claims on private sector. Source: International Monetary Fund, International Financial Statistics, line 32d.
* Bank Cash to total assets: Reserves of Deposit Money Banks divided by total assets of Deposit Money Banks. Source: International Monetary Fund, International Financial Statistics, line 20 divided by lines (22a + 22b + 22c + 22d + 22f).
* Bank Foreign Liabilities to Foreign Assets:
Deposit money banks foreign liabilities to foreign assets. Source: International Monetary Fund, Inter- national Financial Statistics, lines (26c + 26cl) divided by line 21.
* Previous Crisis: This variable equals zero if the country has not previous crisis; one, if the country has suffered one previous crisis; two, in case of two or three previous crisis, and, three, otherwise.
Insurance insolvency database shown in Table A.1.
Decision tables and decision rules tables shown in Tables A.2 and A.3.
References
Altman, E.I., Avery, R., Eisenbeis, R., Stinkey, J., 1981. Application of classification techniques in business, banking and finance. Contemporary Studies in Economic and Finan- cial Analysis 9, 195–211.
Bannister, J., 1997. Insurance Solvency Analysis. LLP Limited. Bhattacharya, S., Thakor, A., 1994. Contemporary banking
theory. Journal of Financial Intermediation 3, 1. Berg, A., Borensztein, E., Mauro, P., 2002. An Evaluation of
Monetary Regime Options for Latin America. IMF WP, p. 211.
Bordo, M., 1986. Financial crises, banking crises, stock market crashes, and the money supply: Some international evidence, 1980–1933. In: Capie, F., Wood, G. (Eds.), Financial Crises and the World Banking System. St. Martin’s, New York.
Borio, C., Lowe, P., 2002. Asset Prices, Financial and Monetary Stability: Exploring the Nexus, BIS Working Papers, p. 114.
Caprio, G., Klingebiel, D., 1997. Bank Insolvencies: Cross- Country Experience, Policy Research Working Paper. The World Bank, Washington, DC, p. 1620.
Caprio, G., Klingebiel, D., 2003. Episodes of Systemic and Borderline Financial Crises, Dataset mimeo, The World Bank.
Carare, A., Stone, M., 2003. Inflation Targeting Regimes. IMF WP, p. 9.
Cukierman, A., Webb, S.B., Neyapti, B., 1992. Measuring the independence of central banks and its effect on policy outcomes. The World Bank Economic Review 6, 353–398.
Cukierman, A., Miller, G.P., Neyapti, B., 2002. Central bank reform, liberalization and inflation in transition economies – An international perspective. Journal of Monetary Economics 49, 237–264.
Demirgüç-Kant, A., Detragiache, E., 1997. The Determinants of Banking Crisis: Evidence from Developing and Developed Countries. IMF WP, p. 106 (September).
Demirgüç-Kant, A., Detragiache, E., 1998. Financial Liberaliza- tion and Financial Fragility. IMF WP, p. 83 (March).
Demirgüç-Kant, A., Detragiache, E., 2000. Does Deposit Insur- ance Increase Banking System Stability? IMF WP, p. 3 (January).
Dimitras, A., Slowinski, R., Susmaga, R., Zopounidis, C., 1999. Business failure prediction using Rough Sets. European Journal of Operational Research 114, 263–280.
Domaç, I., Martı́nez Peria, M.S., 2000. Banking Crises and Exchange Rate Regimes: Is there a Link? The World Bank WP, p. 2489.
Eichengreen, B., 1997. Exchange Rate Stability and Financial Stability. University of California, Berkeley WP, pp. C97–092 (June).
Eichengreen, B., 2000. When to Dollar Ice. Unpublished. University of California, Berkeley (January).
Eichengreen, B., Rose, A., 1998. Staying Afloat When the Wind Shifts: External Factors and Emerging-market Banking Crisis. NBER WP, p. 6370 (January).
Eichengreen, B., Arteta, C., 2000. Banking crisis in emerging markets: Presumptions and evidence. Center for International and Development Economics Research, University of California, Berkeley.
Fisher, I., 1933. The Debt-Deflation Theory of Great Depressions, Econometrica.
1572 A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573
Friedman, M., Schwartz, A.J., 1963. A Monetary History of the United States, 1867–1960. Princeton University Press, Prince- ton, NJ.
Frydl, E., 1999. The length and cost of banking crisis. IMF WP, p. 30 (March).
Garcı́a-Herrero, A., 1997. Monetary Impact of a Banking Crisis and the Conduct of Monetary Policy. IMF WP, p. 124.
Garcı́a-Herrero, A., del Rio, P., 2003. Financial Stability and the Design of Monetary Policy. Documento de trabajo Banco de España, p. 15.
Glick, R., Hutchinson, M., 1999. Banking and Currency Crisis: How Common are Twins? Unpublished. Federal Reserve Bank of San Fracisco and UC Santa Cruz (September).
González-Hermosillo, B., 1996. Banking Sector Fragility and Systemic Sources of Fragility. IMF WP, p. 12.
Gorton, G., 1998. Banking Panics and Business Cycles. Oxford Economic Papers, vol. 40. pp. 751–781.
Gourinchas, P., Valdes, R., Landerretsche, O., 1999. Lending Booms: Some Stylized Facts. Unpublished. Princeton Uni- versity and Central Bak of Chile (August).
Greco, S., Matarazzo, B., Slowinski, R., 1998. A new rough set approach to evaluation of bankruptcy risk. In: Zopounidis, C. (Ed.), New Operational Tools in the Management of Finan- cial Risks. Kluwer Academic Publishers, Dordrecht, pp. 121– 136.
Greco, S., Matarazzo, B., Slowinski, R., 2001. Rough sets theory for multicriteria decision analysis. European Journal of Operational Research 129 (1), 1–47.
Gupta, P., 1996. Currency crises, banking crises and twin crises: A comprehensive review of the literature, mimeo. Interna- tional Monetary Fund.
Hardy, D., Pazarbasioglu, C., 1998. Leading Indicators of Banking Crisis: Was Asia different? IMF WP, p. 91 (June).
Houben, A., Kakes, J., Schinasi, G., 2004. Toward a Framework for Safeguarding Financial Stability. IMF WP, p. 101.
Issing, O., 2003. Monetary and financial stability: Is there a trade- off? Conference on Monetary Stability, Financial Stability and the Business Cycle, March 28–29, 2003. Bank for International Settlements, Basle.
Kaminsky, G., Reinhart, C., 1999. The twin crises: The causes of banking and balance-of-payments problems. American Eco- nomic Review 89 (3), 473–500.
Kuttner, K.N., Posen, A.S., 2001. Beyond Bipolar: A Three- dimensional Assessment of Monetary Frameworks. Oester- reichische Nationalbank, WP, p. 52.
Laitinen, E.K., 1992. Prediction of failure of a newly founded firm. Journal of Business Venturing (July), 323–340.
Lindgreen, C.J., Garcı́a, G., Saal, M.I. (Eds.), 1996. Bank Soundness and Macroeconomic Policy. International Mone- tary Fund.
López, D., Moreno, J., Rodrı́guez, P., 1994. Modelos de Previsión del fracaso empresarial: aplicación a entidades de seguros. Revista española de Seguros 54 (abril–junio), pp. 71–110.
Mahadeva, L., Sterne, G. (Eds.), 2000. Monetary Policy Frame- works in a Global Context. Routledge.
Martı́n, M.L., Leguey, S., Sánchez, J.M., 1999. Solvencia y estabilidad financiera en la empresa de seguros: Metodologı́a y evaluación empı́rica mediante análisis multivariante. Cuad- ernos de la Fundación Mapfre Estudios, p. 49.
Martinez-Peria, S., 2000. Banking crisis and exchange rate regimes, is there a link? Policy Research, WP, p. 2489 (November).
Mckee, T., 2000. Developing a bankruptcy prediction model via rough sets theory. International Journal of Intelligent Systems in Accounting, Finance and Management 9, 159–173.
Mckee, T.E., Lensberg, T., 1999. Using a genetic algorithm to obtain a causally ordered model from a rough sets derived bankruptcy prediction model. Paper presented at The Inter- national Symposium on Audit Research, The University of Southern California.
Mendis, C., 1998. External Shocks and Banking Crisis in Small Open Economies: Does the Exchange Rate Regime Matter? Unpublished. Center for the Study of African Economies, Oxford (October).
Mishkin, F.S., 1996. Understanding Financial Crises: A Devel- oping Country’s Perspective. Nber. WP 5600, National Bureau of Economic Research, Cambridge, MA.
Mora, A., 1994. Los modelos de predicción del fracaso empres- arial: una aplicación empı́rica del logia. Revista Española de Financiación y Contabilidad 78 (enero–marzo), pp. 203–233.
Nurmi, H., Kacprzyk, J., Fedrizzi, M., 1996. Probabilistic, fuzzy and rough concepts in social choice. European Journal of Operational Research 95, 264–277.
Padoa-Schioppa, T., 2002. Central banks and financial stability: Exploring a land in between. Paper Presented at the Second ECB Central Banking Conference, The Transformation of the European Financial System, Frankfurt am Main, October.
Pawlak, Z., 1991. Rough SetsTheoretical Aspects of Reasoning About Data. Kluwer Academic Publishers, Dordrecht/Bos- ton/London.
Pill, H., Pradhan, M., 1995. Financial Indicators and Financial Change in Africa and Asia. IMF WP, p. 23.
Predki, B., Wilk, S., 1999. Rough set based data exploration using ROSE system. In: Ras, Z.W., Skowron, A. (Eds.), Foundations of Intelligent Systems, Lecture Notes in Artifi- cial Intelligence, vol. 1609. Springer-Verlag, Berlin, pp. 172– 180.
Predki, B., Slowinski, R., Stefanowski, J., Susmaga, R., Wilk, S., 1998. ROSE-software implementation of the Rough Set Theory. In: Polkowski, L., Skowron, A. (Eds.), Rough Sets and Current Trends in Computing, Lecture Notes in Artificial Intelligence, vol. 1424. Springer-Verlag, Berlin, pp. 605– 608.
Reinhart, C.M., Rogoff, K.S., 2002. The Modern History of Exchange Rate Arrangements: A reinterpretation, NBER WP 8963, National Bureau of Economic Research, Cambridge.
Rossi, M., 1999. Financial Fragility and Economic Performance in Developing Countries: Do Capital Controls, Prudential Regulation and Supervision Matter? IMF WP, p. 66 (May).
Sanchis, A., Gil, J.A., Heras, A., 2003. El análisis discriminante en la previsión de la insolvencia en las empresas de seguros no vida, Revista Española de Financiación y Contabilidad vol’XXXII, pp. 116, 183–233.
Schinasi, G.J., 2003. Responsibility of Central Banks for Stability in Financial Markets. IMF WP, p. 121.
Schwart, Z.A., 1995. Systemic risk and the macroeconomy. In: Kaufman, G. (Ed.), Banking Financial Markets and Systemic Risk, Research in Financial Services, Private and Public Policy, vol. 7. JAI Press Inc., Hampton.
Slowinski, R., 1993. Rough set learning of preferential attitude in multicriteria decision making. In: Komorowski, J., Ras, Z.W. (Eds.), Methodologies for Intelligent Systems, Lecture Notes in Artificial Intelligence, vol. 689. Springer-Verlag, Berlin, pp. 642–651.
A. Sanchis et al. / European Journal of Operational Research 181 (2007) 1554–1573 1573
Slowinski, R., Zopounidis, C., 1995. Application of the rough set approach to evaluation of bankruptcy risk. International Journal of Intelligent Systems in Accounting, Finance and Management 4 (1), 27–41.
Stiglitz, J., Furman, J., 1998. Evidence and insights from East Asia. Brookings Papers on Economic Activity 2.
Velasco, A., 1987. Financial crises and balance of payments crises. Journal of Development Economics 27, 263– 283.
Zopounidis, C., Dimitras, A., 1998. Multicriteria Decision aid Methods for the Prediction of Business Failure. Kluwer Academic Publishers.
- Rough Sets and the role of the monetary policy in financial stability (macroeconomic problem) and the prediction of insolvency in insurance sector (microeconomic problem)
- Introduction
- The determinants of banking crises
- Financial ratios as explicatives variables of the insolvency in insurance sector
- The methodology: Main concepts of the Rough Set (RS) Theory
- Approximation of sets, accuracy and quality of approximation
- Reduction and dependency of attributes
- Decision rules
- Empirical results
- The data
- Prediction of financial instability
- Prediction of the insolvency of Spanishnon-life insurance companies
- Comparison of Rough Set approach with discriminant analysis
- Conclusions
- Acknowledgements
- Data appendix
- Financial crisis database
- Dependent variable
- The Objective variables
- Control variables
- Macroeconomic variables
- Financial variables
- References
Safeguarding Financial Stability Theory and Practice.pdf
Safeguarding Financial Stability: Theory and Practice by Garry J. Schinasi Review by: Éric Tymoigne Journal of Economic Issues, Vol. 41, No. 1 (Mar., 2007), pp. 305-307 Published by: Association for Evolutionary Economics Stable URL: http://www.jstor.org/stable/25511176 .
Accessed: 13/02/2015 16:24
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Association for Evolutionary Economics is collaborating with JSTOR to digitize, preserve and extend access to Journal of Economic Issues.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:24:33 PM All use subject to JSTOR Terms and Conditions
Journal of Economic Issues vol. 41, no. 1 305
funded research institute for the study, prevention and treatment of tropical diseases was rejected by the international participants in the Dolder-Club Meetings because,
according to one author, "the pharmaceutical industry [had] not learnt to reconcile its
individualistic profit-seeking culture with broader visions of common societal and
political responsibility"(p. 32). Readers of this journal will find much of interest in this volume. It contains
a wealth of historical information and international perspective about the
pharmaceutical industry, the patient's rights movement, regulatory legislation, and
global health issues, just to name a few. Although this is my first foray into the field of
healthcare, I agreed to review the text because I was interested in using it for a
graduate course in contemporary political economy. Because of the complexity of the
issues debated, it deserves more time than I have. However, the volume has been
extremely useful in explaining a variety of institutional concepts such as culture,
instrumental valuation, and ceremonial behavior. For these reasons and others, I
highly recommend this book as a resource for both undergraduate and graduate courses.
Paulette Olson
Wright State University
Safeguarding Financial Stability: Theory and Practice, by Garry J. Schinasi. Washington, D.C.: International Monetary Fund. 2006. Paper: ISBN 1 58906 440 2, $28.00. 311 pages.
In this book, the author presents a broad framework to promote financial stability. In
doing so, he puts forward good points but one can see the tension that runs through the book because of his willingness to apply standard economic tools to the analysis of financial fragility. In the end, if Schinasi argues in favor of leaving most regulation of the financial system to the "market" by reducing market imperfections, one is forced
to recognize that his case rests on weak foundations.
The whole argument of the book is based on the standard contemporary approach to economics. The author views the financial system as a pooling mechanism that redistributes a "fixed supply of fiat money" (p. 55) by leveraging upon it via the creation of credit instruments. By doing this, the financial system facilitates
intertemporal choices through the allocation of real resources from savers to
investors, the allocation and assessment of risks, and the provision of liquidity (pp.
36, 85, 123). Market imperfections (externalities, public goods, asymmetries of
information), however, push the financial system above or below optimality (p. 59). This, combined with the intrinsic incompleteness of financial contracts induced by the uncertainty of human trust, makes regulation and supervision a must in order to
avoid financial instability.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:24:33 PM All use subject to JSTOR Terms and Conditions
306 Book Reviews
Given this traditional equilibrium/optimality analysis, the second part of the book provides a general policy framework to promote financial stability (financial
instability is defined as the capacity of the financial system to perform its three tasks
simultaneously (pp. 82ff.)). However, in doing so the author attempts to introduce
history into the equilibrium analysis. Indeed, "the traditional 'shock-transmission'
approach that is the basis of many existing policy-oriented framework" presumes that
a system remains "in a state (or path) of equilibrium if undisturbed or adjusts to a
different, perhaps less desirable, state (or path) of equilibrium" (p. 98). Instead, we
should approach financial stability "as a continuum" (p. 98) process in which "the
building up of vulnerabilities" (p. 98) is more of a concern than the shock triggering financial instability.
One should make an assessment of the strength of the financial system by defining a corridor of financial stability and by practicing a constant monitoring of the "individuals parts of the financial system (financial markets, institutions, and
infrastructure) and the real economy (households, firms, the public sector)." In
addition, the interlinkages between the financial positions of each sector of the
economy should be checked (p. 110). The whole point of the framework is to identify all endogenous and exogenous risks, check their possibility given the state of the
economy, check if they can generate systemic risk, calculate the expected loss
generated at the individual level and on the whole economy, and rank the plausible systemic risk by expected loss. Once that is done, one could implement policies that
prevent, remedy or resolve problems. Each stage of assessment has different
implications in terms of policy instruments, going from the use of market discipline to
discretionary governmental interventions, from the use of official communication
(moral suasion, restoring confidence) to change in the legal system (p. 116).
This pushes the reader to ask how to define the corridor of stability and who
should be in charge to do so. Regarding the latter, the author is clear: "The public sector role could be limited to assessing and monitoring the quality of risk
management and control systems [established by private financial institutions] more
systematically and thoroughly, and to defining how information is used, plus ensuring
that counterparty disclosure is adequate" (p. 225). Indeed, the private sector is prone
to extrapolate the recent past (p. 171), and may not assess risk correctly because the
cost of getting the information is too high given the competitive pressure existing in
the system (pp. 221, 225). Concerning the definition of the corridor, the author
interestingly notes that "financial-stability risks often reflect the far-reaching
consequences of unlikely events, implying that the focus is not the mean, median, or
mode of possible outcomes but the entire distribution of outcomes, in particular the
'left tail'" (p. 133). The last part of the book presents some challenges induced by the innovative
drive of financial institutions and by the internationalization of financial activities.
The author shows how private financial institutions have arbitraged on the loophole
of the regulatory framework. He also shows how this has reduced the information
provided by balance sheets and other traditional accounting methods concerning the
financial position of financial institutions. This clearly gives an example of where the
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:24:33 PM All use subject to JSTOR Terms and Conditions
Journal of Economic Issues vol. 41, no. 1 307
need to reduce the asymmetry and opacity of information is present. However, the
question becomes whether improving the quantity and quality of information will be
enough. The author believes that by improving transparency and quality of
information, "both market participants and supervisors will see hints of errant
investment strategies of financial problems as they begin" (p. 179) because "without
sufficient timely information, the market disciplining mechanisms [...] might not
produce the appropriate self-corrective adjustments" (p. 167).
This view is problematic. One central reason (among others) is that
experience shows that market participants are more sensitive to profitability than to
risk issues. Thus, even if they are concerned with the latter, it is always cast into
strategies oriented toward the former, which has two consequences. First, the riskiness
of a situation has a tendency to be judged a posteriori, that is, if ones takes large risks but is successful, nobody will put into question the risks taken, whereas, if the
enterprise fails, people will complain that the financial market participants took too much risk. An example of this is Enron: nobody questioned its leveraging position
(98% of its equity capital) until it was too late, and if Enron had been successful, praises would have been made for its innovative and highly sophisticated fund
management strategy; leading more financial market participants to follow it. Second, as the author partly notes, for reasons of profitability, financial system participants
may be lead to ignore some crucial information. This, however, goes further than cost
related issues because financial market participants may simply avoid looking at
important pieces of information, or may reinterpret them in a positive way. Many
authors, going from Keynes, to Galbraith, to Minsky, have explained the social, psychological and economic dynamics behind this phenomenon. All this shows that what really matters is not the information in itself but the way it is interpreted. This is
even more the case, as the author notes, that competition is ferocious and pushes to
follow the most successful leader (whatever its strategy).
Hyman Minsky would have found the current debate most stimulating, him
who worked all his life on this subject but never really found any large audience. He died just as economists became more aware of the importance of financial stability in
itself. If he is gone, his framework of analysis is still here and provides a good guide to
explain how to deal with financial stability. Instability is not the result of asymmetry of information, irrational behavior, or other imperfection, but is deeply rooted in the
psychology and social behavior of individuals as well as economic forces like
competition and economic policy (Tymoigne 2006a, 2006b). Eric Tymoigne
California State University, Fresno
References.
Tymoigne, E. 2006a. "The Minskian system, Part I." Levy Economics Institute of Bard College, working paper No. 452.
-. 2006b. "The Minskian system, Part II." Levy Economics Institute of Bard College, working paper No. 453.
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:24:33 PM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 305
- p. 306
- p. 307
- Issue Table of Contents
- Journal of Economic Issues, Vol. 41, No. 1 (Mar., 2007), pp. 1-310
- Front Matter
- Principles of Institutional-Evolutionary Political Economy: Converging Themes from the Schools of Heterodoxy [pp. 1-42]
- Economic Restructuring in Iraq: Intended and Unintended Consequences [pp. 43-60]
- Thailand's Financial Crisis: Its Causes, Consequences, and Implications [pp. 61-76]
- Industrial Policy and Economic Development: Korea's Experience [pp. 77-92]
- Macroeconomic Stabilization through an Employer of Last Resort [pp. 93-134]
- Turning Economics into an Evolutionary Science: Veblen, the Selection Metaphor, and Analogical Thinking [pp. 135-154]
- How Can Institutional Economics Be an Evolutionary Science? [pp. 155-179]
- The Decline of the Middle Class: An International Perspective [pp. 181-200]
- The Bubble Machine: Relative Capital Valuation, Distributive Shares and Capital Gains [pp. 201-220]
- Norm-Based Behavior and Corporate Malpractice [pp. 221-241]
- Notes and Communications
- Heterodox Theoretical Convergence: Possibility or Pipe Dream? [pp. 243-263]
- A Response to Christian Cordes and Clifford Poirot [pp. 265-276]
- Can a Generalized Darwinism Be Criticized? A Rejoinder to Geoffrey Hodgson [pp. 277-281]
- Book Reviews
- Review: untitled [pp. 283-287]
- Review: untitled [pp. 287-289]
- Review: untitled [pp. 289-291]
- Review: untitled [pp. 291-293]
- Review: untitled [pp. 294-295]
- Review: untitled [pp. 295-297]
- Review: untitled [pp. 297-299]
- Review: untitled [pp. 300-302]
- Review: untitled [pp. 302-305]
- Review: untitled [pp. 305-307]
- Review: untitled [pp. 308-309]
- Back Matter
Securitisation and Financial Stability.pdf
Securitisation and Financial Stability Author(s): Hyun Song Shin Source: The Economic Journal, Vol. 119, No. 536, Conference Papers (Mar., 2009), pp. 309-332 Published by: Wiley on behalf of the Royal Economic Society Stable URL: http://www.jstor.org/stable/20485321 .
Accessed: 06/02/2015 10:53
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Wiley and Royal Economic Society are collaborating with JSTOR to digitize, preserve and extend access to The Economic Journal.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
The EconomicJournal, 119 (March), 309-332. (D Th, Author(s). Journal compilation (? Royal Economic Society 2009. Published by
Blackwell Publishing, 9600 Garsington Road, Oxi ord OX4 2DQ UK and 350 Main Street, Malden, MA 02148, USA.
SECURITISATIOI4 AND FINANCIAL STABILITY*
Hyun Song Shin
A widespread opinion before the crc dit crisis of 2007/8 was that securitisation enhances financial
stability by dispersing credit risk. Aft 'r the credit crisis, securitisation was blamed for allowing the
'hot potato' of bad loans to be passe I to unsuspecting investors. Both views miss the endogeneity
of credit supply. Securitisation enab] es credit expansion through higher leverage of the financial
system as a whole. Securitisation by tself may not enhance financial stability if the imperative to
expand assets drives down lending tandards. The 'hot potato' of bad loans sits in the financial
system on the balance sheets of largz banks rather than being sold on to final investors, since the
aim of financial intermediaries is t(o expand lending in order to utilise slack in balance sheet
capacity.
There are two pieces of received wisdom concerning securitisation - one old and one
new. The old view (prevalent bel'ore outbreak of the credit crisis of 2007/8) emphas
ised the positive role played b y securitisation in dispersing credit risk, thereby
enhancing the resilience of the financial system to defaults by borrowers. The sub
sequent credit crisis has somewh2 t tarnished this positive image.' In its place, there is a
new received wisdom which emF hasises the distorted incentives that developed at all
stages of the securitisation proce; s, and which allowed the 'hot potato' of bad loans to
pass through the financial systerr to be held finally in the hands of unsuspecting final
investors.
Although both views of secur itisation (old and new, positive and negative) are
appealing at a superficial level, they both neglect the endogeneity of credit supply.
Financial intermediaries manage their balance sheets actively in response to shifts in
measured risks. The supply of ci edit is the outcome of such decisions, and depends
sensitively on key attributes of in termediaries' balance sheets.
Three attributes merit special mention - equity, leverage and funding source. The
equity of a financial intermedia y is its risk capital that can absorb potential losses.
Leverage is the ratio of total asset, to equity and is a reflection of the constraints placed
on the financial intermediary by i ts creditors on the level of exposure for each dollar of
its equity. Finally, the funding ,ource matters for the total credit supplied by the
financial intermediary sector as <. whole to the ultimate borrowers.
At the aggregate sector level (i e. once the claims and obligations between leveraged
entities have been netted out), the lending to ultimate borrowers must be funded
either from the equity of the inte rmediary sector or by borrowing from creditors outside
the intermediary sector. For any fixed profile of equity and leverage across the banks,
the supply of credit to ultimate t orrowers is larger when the banks borrow more from
creditors outside the banking syi tem.
* Presented as the Economic Journa l Lecture at the Royal Economic Society Conference in Warwick, March 17-8, 2008.1 am grateful to Frank I teinemann, Gara Minguez Afonso, Jean-Charles Rochet and Andrew
Scott for their comments on earlier versio is and, especially, to Tobias Adrian for allowing me to draw on work
from our on-going collaboration. 1
See BIS (2008), Brunnermeier (forth :oming), Greenlaw et al (2008) or IMF (2008) for an account of the
credit crisis of 2007/8.
[ 309 ]
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
310 THE ECONOMIC JOURNAL [MARCH
In a traditional banking system that intermediates between retail depositors and
ultimate borrowers, the total quantity of deposits represents the obligation of the
banking system to creditors outside the banking system. However, securitisation
opens up potentially new sources of funding for the banking system by tapping new
creditors. The new creditors are those who buy mortgage-backed securities (MBSs),
claims that are written on MBSs such as collateralised debt obligations (CDOs), and
(one step removed) those who buy the asset-backed commercial paper (ABCP) that
are ultimately backed by CDOs and MBSs.2 The new creditors who buy the secur
itised claims include pension funds, mutual funds and insurance companies, as well
as foreign investors such as foreign central banks. Indeed, we will see shortly that
foreign central banks have been an important funding source for residential mort
gage lending in the US. We will also examine some more partial evidence for the UK
from the balance sheet of Northern Rock, the UK mortgage bank which failed in
2007. Although securitisation may facilitate greater credit supply to ultimate borrowers at
the aggregate level, the choice to supply credit is taken by the constituents of the
banking system taken as a whole. For a financial intermediary, its return on equity is
magnified by leverage. To the extent that it wishes to maximise its return on equity, it
will attempt to maintain the highest level of leverage consistent with limits set by
creditors (for instance, through the 'haircuts' on repurchase agreements) or self
imposed risk constraints. As measured risk fluctuates, so will leverage itself. In benign
financial market conditions when measured risks are low, financial intermediaries
expand balance sheets as they increase leverage. Although the intermediary could
increase leverage in other ways - for instance, returning equity to shareholders, buying
back equity by issuing long-term debt - the evidence suggests that they tend to keep
equity intact and adjust the size of total assets; see Adrian and Shin (2007, 2008a). As
balance sheets expand, new borrowers must be found. When all prime borrowers have
a mortgage but balance sheets still need to expand, then banks have to lower their
lending standards in order to lend to subprime borrowers. The seeds of the subsequent
downturn in the credit cycle are thus sown.
When the downturn arrives, the bad loans are either sitting on the balance sheets of
the large financial intermediaries, or they are in special purpose vehicles (SPVs) that
are sponsored by them. This is so, since the bad loans were taken on precisely in order
to utilise the slack on their balance sheets. Although final investors such as pension
funds and insurance companies will suffer losses, too, the large financial intermediaries
are more exposed in the sense that they face the danger of seeing their capital wiped
out. The severity of the credit crisis of 2007/8 lies precisely in the fact that the bad loans
were not all passed on to final investors. Instead, the 'hot potato' sits inside the financial
system, on the balance sheet of the largest and most sophisticated financial inter
mediaries.
The outline of this article is as follows. I begin with some background on securitisa
tion, and construct an accounting framework of the financial system as a network of
inter-linked balance sheets. When this accounting framework is combined with a model
2 See Gorton and Souleles (2006) for a description of special purpose vehicles involved in the securitisation
process in the US.
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 311
of leverage based on value at risk (VaR), it is possible to model a lending boom fuelled
by declines in measured risks. I conclude with a discussion of the implications for
financial stability.
1. Background
Securitisation has played a key role in the growth of residential mortgage lending,
especially in the US. Figure 1 plots the total outstanding US home mortgage assets held
by various classes of financial institutions from 1980.
Even as recently as the early 1980s, banks and savings institutions held the bulk of
home mortgages. Since then, the mortgage pools of the government sponsored
enterprises (GSEs) such as Fannie Mae and Freddie Mac have become the largest
holder of residential mortgages. Also noticeable are the securitisation vehicles classified
under asset backed securities (ABS) issuers. The ABS issuers hold mortgages that do
not conform to the GSE standards, hence including the subprime mortgages as well as
large mortgages ('umbo' mortgages) that exceed the upper threshold on the GSE
conforming mortgages. Figure 2 is an aggregate series that distinguishes the 'bank-based' holdings of resid
ential mortgages from the 'market-based' holdings. The latter is the sum of the
holdings of the government sponsored enterprises, the GSE mortgage pools and the
private label ABS issuers. The bank-based series is the sum of the remaining three
categories. The market-based series overtook the bank-based series in 1990 and now
accounts for two thirds of approximately 11 trillion dollars' worth of residential
mortgages outstanding. Securitisation had long been seen as a positive development for the resilience of
the financial system by enabling the dispersion of credit risk. However, since the
4.5......... - Agency and GSE mortgage pools
-3 - Commercial banks
ABS issuers
: 2.5- -Savings institutions
ifl .0- GSEs f - Credit unions
1.5. ........
0.5 - - - -
0.0 --m e t
~ - m----C - ----lq -n .O- - -1 - - - -t -n -Ot
w t f 0
CN00% 0 ~ ON000 QN
e~ O 0% 0% 0% 0% 0% 0% 0% 0% CIA 07% 0% 0% 0% ON 0% 0% 0% 0% ON ON
Fig. 1. US Home Mortgage Assets (1980Q1-2008Ql): Flow of Funds, US Federal Reserve
?The Author(s). journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
312 THE ECONOMIC JOURNAL [MARCH
7.0
6.0.
5.0
a 5.0- _~~~Market-based f 040
3Bank-based
~3.0
2.0
1.0
0.0 - ........l l ul .............lag-lqlll?ll .............. ......r .....
\0V'V 0 %0%OQ0 .0C '.0 %0 '. ' .0 Dc%o%00O0000 0000000000 0 O00oO 0 %0%0 0 No 0 's o%0O'.% 00000000 0 i-''3W.tA0%4OO _ '.0, '-k)k .4.LtAQ% - '.0 t J LA %-.4
ctoIooocIcoc ccoto to #ocoocco8
Fig. 2. Bank-based and Market-based Home Mortgage Holdings (1980Q1-2008Q1): Flow of Funds, US Federal Reserve
onset of the credit crisis of 2007/8, a less sympathetic view of securitisation has
gained support that emphasises the multi-layered agency problems that took hold at
every stage of the securitisation process, starting with the origination of the loan to
the sale, warehousing and securitisation as well as the role of the credit rating
agencies in the process.3 We could dub this less charitable view the 'hot potato'
hypothesis, which has figured frequently in speeches given by policy makers on the
credit crisis.4 The motto would be that there is always a greater fool in the chain who
will buy the bad loan. At the end of the chain, according to this view, is the hapless
final investor who ends up holding the hot potato and suffers the eventual loss. A
celebrated anonymous cartoon strip has circulated widely on the internet5 depicting a
hapless official from a Norwegian municipality in conversation with a broker after
suffering losses on subprime mortgage securities. There is also mounting empirical
evidence that lending standards had been lowered progressively in the run-up to the
credit crisis of 2007; see Demyanyk and van Hemert (2007), Mian and Sufi (2007)
and Keys et al. (2007).
It is clear that final investors who buy claims backed by bad assets will suffer
losses. However, it is important to draw a distinction between selling a bad loan
down the chain and issuing liabilities backed by bad loans. By selling a bad loan,
you get rid of the bad loan from your balance sheet. In this sense, the hot potato is
passed down the chain to the greater fool next in the chain. However, the second
action has a different consequence. By issuing liabilities against bad loans, you do
not get rid of the bad loan. The hot potato is sitting in the financial system, on the
3 A comprehensive survey of the securitisation process for subprime mortgages is given by Ashcraft and
Schuermann (2008) who details the specific agency problems at seven points in the securitisation chain. 4
See, for instance, Gieve (2008) and Mishkin (2008) among others. n
For instance, http://bigpicture.typepad.com/comments/2008/02/how-subprime-re.html
? The Author(s). Journal compilation (? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 313
books of the special purpose vel ticles (SPVs). Although the special purpose vehicles
are separate legal entities from t -ie large financial intermediaries that sponsor them,
the finanical intermediaries hav_ exposures to them from liquidity enhancements
and various forms of retained i -lterest; see Gorton and Souleles (2006). Thus, far
from passing the hot potato dow n the chain to the greater fool next in the chain,
the large financial intermediari ~s end up keeping the hot potato. In effect, the
large financial intermediaries ar the last in the chain. They are the greatest fool.
While the final investors such is the famed Norwegian municipality will end up
losing money, the financial inter mediaries that sponsored the SPVs are in danger of
larger losses. Since the intermedliaries are leveraged, they are in danger of having
their equity wiped out.
Indeed, Greenlaw et al. (2008) report that of the approximately 1.4 trillion dollar
total exposure to subprime mortv ages, around half of the potential losses are borne by
US leveraged financial institutior s, such as commercial banks, investment banks and
hedge funds. When foreign levei aged institutions are included, the total rises to two
thirds. Gorton (2008) also argucs against the hot potato hypothesis by noting that
financial intermediaries have bo -ne a large share of the total losses. Hence, we are
faced with the following importa it question. Why did apparently sophisticated banks
act as the 'greatest fool'? In the rn st of the article, I outline a framework that addresses
this question.
2. An Accounting Framework
Financial intermediaries play a ro le both as a lender and also as a borrower. In what
follows, I describe an accounting Framework to take account of the interlocking claims
and obligations. There are n + [ entities in financial system, where n of them are
leveraged institutions (referred t:) as 'banks' for convenience) and one unleveraged
sector (indexed by n + 1), whict aggregates the balance sheets of unleveraged insti
tutions such as insurance comp tnies, pension funds and mutual funds, as well as
household investors or foreign ce -tral banks. There is also an 'end-user' sector who are
the ultimate borrowers. For these purposes, the ultimate borrowers may be considered
as households who buy a house f nanced with a mortgage.
Denote by y, the face value o ̂ claims held by bank i against such end-user bor
rowers. As well as the end-user 1)ans, there are also claims between members of the
financial system. The liability o f one party in the system will be the asset of
another party. Denote by xi the lace value of the obligation of bank i, and by 1rij the
share of bank i's obligations ti at are held by bank j. Then, denoting by ei the
notional value of equity of bank i, the balance sheet identity of bank i in terms of
face values is:
n
+ E x-j-gi =- X-i + e-i(1)
j=1
The left-hand side of (1) is the t )tal assets of bank i in notional values, consisting of
the loans made to end-users yi, and the claims held against the other leveraged
entities (the 'banks') in the fina-icial system, Znl xy. The right-hand side of (1)
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
314 THE ECONOMIC JOURNAL [MARCH
gives the total liabilities of bank i in notional values and consists of the total
promised repayment xi by bank i plus the notional equity e-i that equates the two
sides of the balance sheets. The interlocking claims and obligations can be depicted
in terms of the following table, where xpi denotes the notional value of bank i's
obligations to bank j.
bank 1 bank 2 ... bank n outside debt bank 1 0 X12 ... Xln Xn+ xl
bank 2 X21 0 X2n X2,n+l X2
bank n xnl Xn2 ... 0 xfn,n+l Xn
end-user loans Y1 Y2 ... Yn
total assets a] a2 an
Summing the ith row of the matrix gives the total liabilities of bank i, since it sums the
obligations of bank i to other banks and to the long-only investors (sector n + 1). The
sum of the entries in the ith column of the matrix gives the total notional assets of bank
i, since it sums the claims that bank i has on all other banks in the system, plus the loans
it has made to the end-users. The total notional assets of bank i are denoted as ai.
2.1. Credit Risk
To begin with, suppose there are two dates, date 0 and date 1. Loans are made at
date 0 and are repaid at date 1. The loans made to the end-users are risky
and banks face credit risk. Credit risk follows the familiar Vasicek (2002) one
factor model, which is widely used and has been adopted as the backbone of the
Basel II capital regulations.6 Under the Vasicek one factor model, the end-user
borrower j of bank i repays the loan when the realisation of random variable Zij is
non-negative, where Zij is defined as
Zij = _q- ID (pi + V/Y + V"1-- ij (2)
where (D () is the c.d.f. of the standard normal, Y and {Xij} are mutually independent
standard normal random variables, and p and pi are constants. Yhas the interpretation
of the common risk factor and Xij is the idiosyncratic risk factor. The probability of
default of any borrower j of bank i is pi since
Pr( Zij < 0) P Pr [ Y + T-p Xj < 'D (pi)1
= [(D-(p-)] = p
Conditional on the common factor Y, defaults are independent across borrowers, and
the parameter p gives the ex ante correlation in defaults between any two loans made by
bank i.
6 See also Alizalde and Repullo (2006) for an application of the Vasicek model in a model of banking
competition.
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 315
Suppose that bank i's portfolio includes N loans to end-users each with face value
yi/N. But letting Nbecome large, 1 he loan portfolio to end-users consists of many small
loans whose defaults are independ ent conditional on the realisation of Y. By the law of
large numbers, the repayment wi on the loan book of face value 5i then becomes a
determinstic function of Y. In oth er words,
Wi (Y)-= U1 (Zij > ? I Y)
-y-p r
ii[vf ]-
~
The c.d.f. over the repayment on bank i's loan book is thus
Fi(z) =Prwit(Y) < z]
Pr[Y < wA(z)]
@ @1(pi) + @-1 ( /- )])
Note the following features. A change in pi (the probability of default on a
particular loan made by bank i) implies a first degree stochastic dominance shift in
the repayment density. A fall in pi pushes down the c.d.f., implying a first-degree
stochastic shift to the right in repayments. When pi is fixed, the mean repayment
remains unchanged. However, a change in the parameter p keeping pi fixed implies
a second degree stochastic domi-iance shift in the repayment density. An increase in
p is associated with a mean-pres rving spread of the repayment density, making the
loan book more risky.
2.2. Realised Values of Debt
The realised value of repayment on the loans to end-users will determine the realised
value of the claims held betweer the banks, since the ability of one bank to fulfil its
promise will depend on the resoi irces it has to meet its obligations. Let us use the hat
notation '^' to denote realised v; lues at date 1. Thus, ji is the realised repayment on
bank i's loans to end-users, x, is the realised repayment by bank i and so on. Assume
that all debt is of equal seniority so that if xi < xi, then bank j receives share rij of x
Creditors receive the full value of the assets of the bank if the realised value of the assets
fall short of the face value. Hen( e, realised values of debt satisfy
X1 min (a, (X~), Xl )
X2 min(a2(&) x2)
(4)
Jn R minE(ani (Sc), Xn
(3 The Author(s). Journal compilation (Cc Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
316 THE ECONOMIC JOURNAL [MARCH
where x = (xl, x2, , kn) is the profile of realised values of debt. There is non
decreasing function F() that maps realised asset values to the realised asset values that
result when debts are settled. The ex post allocation is a fixed point of the mapping F( ).
Eisenberg and Noe (2001) showed that under mild regularity conditions, there is a
unique fixed point of this mapping F(Q); see Shin (2008a) for a simple exposition.
Moreover, given the unique fixed point of F( ), the realised value of bank i's debt can
be written as a function of the realised repayments from the loans to end-users
= (i... iyn). Since the realised values {ji} are determinstic functions of Y, I can
write the realised value of the assets of bank i as a deterministic function of the
common factor Y. Hence,
ai Y) = (Y) + EZiT j (sY)] (5) j
Moreover, the comparative statics result on lattices due to Milgrom and Roberts (1994,
theorem 3) ensures that the unique fixed point of the mapping F( ) is increasing in the
realised repayments { j }, so that each %j(y) is an increasing function of y; see Eisenberg
and Noe (2001). In this way, for each bank i, the realised value of its assets &i is a
well-defined, increasing function of Y.
2.3. Market Values
Market values are defined as the expected values (seen from date 0) of the possible
realised values at date 1 where the expectations is taken with respect to the distri
bution of loans losses given by the Vasicek model. I use the notation yi (without any
hats or bars) to denote the market value of bank i's loans to end-users. Similarly, xi
is the market value of bank i's debt, given by the expected value of realised debt
values xcj at date 1. The total marked-to-market value of assets of bank i can then be
written as
ai = yi + E xjsi. (6)
The balance sheet identity for bank i in market values is
yi + EZxjscyiv= ei + xi- (7)
The left-hand side is the market value of assets and the right-hand side is the market
value of the liabilities side of the balance sheet, where ei denotes the market value of
equity of bank i. The matrix of claims and obligations between banks can then be
written in market values, as below. The ith row of the matrix can be summed to give the
market value of debt of bank i, while the ith column of the matrix can be summed to
give the market value of total assets of bank i.
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATI DN AND FINANCIAL STABILITY 317
ba 2k 1 bank 2 ... bank n outside debt bank 1 0 X12 ... x1n Xl,n+l xI bank 2 -21 0 X2n X2,n+1 X2
bank n nl Xn2 0 * Xn,n+l Xn
end-user loans = Y2 ... Yn
total assets 11 a2 an
From the balance sheet identity (7), I can express the vector of debt values across the
banks as follows, where H is the n x n matrix where the (i, j) th entry is 7Tij.
[XI, , X.] = [XI , Xn] [H] + y1, , Yn] - [el, , en] (8)
or more succinctly as
x=xH +y-e. (9)
Equation (9) shows the recursive lature of debt in a financial system. Each bank's debt
value is increasing in the debt value of other banks. Solving for y,
y - e + x(I
- H).
Define the leverage of bank i as -he ratio of the market value of assets to the market
value of its equity. Denote leveral re by 4j. Then, leverage is defined as
a, (10) ei
Since xiei= j- 1, x = e(A-I) where A is the diagonal matrix whose ith diagonal
entry is 4j. Thus
y -= e + e(A -I)(I -H). (1)
Thus, the profile of total lendii ig by the n banks to the end-user borrowers depends on the interaction of three featur -s of the banking system - the distribution of equity e
in the banking system, the prof le of leverage A and the structure of the financial
system given by H. Total lending o end users is increasing in equity and in leverage, as
one would expect. More subtle is -he role of the financial system, as given by the matrix
H. Define the vector z as
z-(I- H)u (12)
where
u
so that zi 1 _- 7 li . In oth r words, zi is the proportion of bank i's debt held by
the outside claimholders - the se, tor n + 1. Then, total lending to end-user borrowers
iy can be obtained by post-mu tiplying (11) by u so that
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
318 THE ECONOMIC JOURNAL [MARCH
n n n
Eyi' = E ei + E ejzz-(AI- 1). (13) i=l i1= iZ=
Equation (13) is the balance sheet identity for the financial sector as a whole, where all
the claims and obligations between banks have been netted out. The left-hand side is
the total lending to the end-user borrowers. The first term on the right-hand side of
(13) is the total equity of the banking system, and the second term is the total funding
to the banking sector provided'by the outside claimholders (note that the second term
can be written as n=L x-z-). Thus, the importance of the structure of the financial
system for the supply of credit is clear from (13). Ultimately, credit supply to end-users
must come either from the equity of the banking system or the funding provided by
non-banks.
2.4. Financial System Leverage
A given degree of leverage for the financial system as a whole is consistent with a wide
range of leverage levels for the individual banks. This is true both in terms of the face
values of claims, as well as market values. First consider face values. A financial system in
face values can be represented as the array (ey, x, H) that satisfies the balance sheet
identity:
x=xfl+y-e. (14)
Then, for positive constant 0, we can construct a financial system where the aggregate
equity, lending and leverage are all unchanged but where the debt to equity ratio of all
individual banks is q times as large. Specifically, consider the financial system
(e', y', x',H H') where e' e, x' Orx and H' is any matrix of interbank claims whose ith
row sum to 1 - zJ/q. Finally, y' is defined as
y' = e' + x'(I-tI'). (15)
Then, aggregate lending is given by
n
Ey. e
eu+ x(I -
I')u
n n
= /e + E Xi Z, n n
+ Se i+ Xizi i=l i~=l
n
=Syj. i=l
Hence, aggregate notional leverage in both financial systems is j= y-I/ >i> ei. How
ever, by construction, the debt to equity ratio of all individual banks is 0 times larger in
the second financial system. The only restriction on the constant 0 comes from the
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISAT- ON AND FINANCIAL STABILITY 319
feature that the ith row of H' s -rms to 1 - z So, 0 should not be so small that
1 - zJ/0 < 0 for some i. This pui s a lower bound on 0. But there is no upper bound. I
can construct a financial system where aggregate notional leverage is unchanged but
where individual bank notional l verage can be as high as we want. The intuition is that
the banks can lend and borrow fi om each other in large amounts so that their leverage
can be raised, without altering tl e aggregate relationship between the banking sector
with the ultimate creditors.
The construction presented above can also be made for the balance sheet quantities
expressed in market values but vith one difference. It is still true that two financial
systems can have the same aggregate market leverage and where the individual market
leverage for the banks differ by L positive factor 0. However, for market leverage, the
constant factor 0 cannot be chos4 n arbitrarily large. This is because the market value of
debt xi cannot be larger than tt e market value of assets ai, and the market value of
assets is underpinned by the valr e of fundamental assets {Yk}. Thus, there is an upper
bound in choosing the constant factor 0. Subject to this condition, the construction
follows the exactly analogous pr(cess.
The leverage of the aggregat banking sector itself is related to the leverage of
individual banks in the following way. If I denote the leverage of the banking sector as a
whole by L, I can write it as
,n
L n e
I + Ei=1 eizi(Qi - 1) (16)
Ki=l ei
where (16) follows from (13). EFhus, other things being equal, the leverage of the
banking sector as a whole is incre using in the amount of funding obtained from outside
claimholders, as given by the qu intities {zi}.
PROPOSITION 1. For any given profile of leverage for individual banks, the leverage of the
financial intermediary sector as a wi ole is increasing in the proportion offunding obtained from
creditors outside the financial intern ediary sector.
3. Value at Risk
Up to this point, I have confined myself to manipulating balance sheet identities. I now
turn to the decision rule followe(. by the banks so that I can address comparative statics
questions on how lending to er d-users depends on the underlying parameters that
drive credit risk (see Figure 3). 1 or this purpose, I employ the notion of value at risk.
For bank i its value at risk at conf dence level c relative to the face value of its assets ai, is
the smallest non-negative numb( r Vi such that
P ̂ (ai < ai - Vi) < 1 - c. (17)
Value at risk Vi is the 'approxim. Ltely' worst case loss that can be suffered by the bank,
where 'approximately worst cas Z' is defined so that anything worse happens with
? The Author(s). Journal compilation (C Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
320 THE ECONOMIC JOURNAL [MARCH
-a)~ ~ ~ 1
a i > ai 0 ~~~~~~~~~~~~~~Vi
Fig. 3. Value at Risk
probability smaller than the benchmark 1 - c. The concept of value at risk has been
adopted widely, both by the private sector and regulators, and is the bedrock of the
capital regulations adopted by Basel regulations. The 1996 Market Risk Amendment of
the original 1988 Basel Accord is based on the notion of value at risk, and the Basel II
regulations have further built on the notion of value at risk. There is an important open
question of how well grounded is the notion of value at risk from a microeconomic
perspective. Adrian and Shin (2008a) provide one possible approach in terms of a
model of a contracting problem in which value at risk can be shown to arise as part of
the optimal contract between a bank and its creditors in a repurchase agreement.
For the exercise here, let us simply assume that banks behave according to the
prescriptions that flow from the notion of value at risk and investigate the conse
quences of such actions. In particular, assume that bank i aims to set market equity ei to
its value at risk Vi, so that
ei= V, (18)
3.1. Decrease in Default Probability
In this context, let us examine consequences of more favourable macroeconomic
conditions as reflected in the decline of default probabilities {pi} in the Vasicek
one-factor model. For simplicity, let pi= p for all i, and we suppose that p has
fallen. Recall that the c.d.f. for realised repayments yi on bank i's loans to end-users is given by
[D-' (p) + 1= p(D1(zl (19) (D Y~~ p ' (z/9))
Notice that when the default probability pi for bank i declines, there is a rightwards shift in the density over the realised loan values ji in the sense of first degree stochastic
dominance. Moreover, since the value of interbank claims are increasing in the
? The Author(s). Journal compilation (? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 321
|~~~~~~~a ,a i
0 a
ri~~~~~~~~~~~~~~~~~~~~e
, ,, ,,
density over realised values of intcrbank asset E,= nj,j held by bank i. Hence, there is
a first-degree stochastic dominan, -e shift in the density over bank i's total asset value hi.
Figure 4 illustrates the shift.
The market value of assets following the fall in p is given by a', and the market equity is given by ei. We have e' > ei, since the ex post value of equity at the terminal date is
increasing in the realised values and there is a first-degree stochastic dominance shift
in firAt the same time, there is o decline in the value at risk of bank i. This is because
the c.d.f. over asset values shifts lower following the fall in p. Therefore, the (1 - c)
quantile of the realised asset volue shifts upward. The value at risk is smaller than
before, and is given by Vi'. Thus. following the decline in p, we have
e' > e, > V,' (20)
so that e' > Vi'. Hence, bank i has surplus equity in the sense that its market equity is
too large relative to the equity 1hat is required to meet its value at risk. The surplus
equity could, in principle, be r medied by paying a dividend to shareholders, or by
buying back equity by issuing more debt. However, in practice, the evidence points to
banks remedying surplus equity b y raising the size of total balance sheets instead, rather
than paying out the surplus eqi Lity.7 Consistent with this evidence, we assume that if
bank i has surplus equity, it exp; Lnds its balance sheet by increasing the notional value
of debt xi and using the procee Is to take on more assets.
7 See Adrian and Shin (2007, 2008a).
? The Author(s). Journal compilation 0 ) Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
322 THE ECONOMIC JOURNAL [MARCH
ASSUMPTION 1. When e' > Vi' after the decline in p, bank i increases the face value of its
debt Xs.
As banks raise new debt, they will acquire assets with the proceeds. The interbank
claims matrix H will therefore change. Since our focus here is on the effect on
aggregate lending, the exact way in which the interbank claims matrix changes is not of
direct interest. Suppose the new interbank claims matrix is given by H* after the
adjustment of face values and the profile of market value of debt is given by x* after the
adjustment of face values. The comparative statics result due to Milgrom and Roberts
(1994, theorem 3) for the fixed point of increasing functions on complete lattices
implies that when the face value of bank i's debt increases, the market values of debt is
increasing for all banks.8 Hence, given Assumption 1, we have
x* > x. for all i. (21)
Let us make one further assumption. As banks increase their borrowing in response
to the appearance of surplus equity, they will search for new sources of funding. If
financial innovation through securitisation is available, the banks may tap new sources
of funding by borrowing from the outside creditor sector-sector n + 1 in my notation.
I therefore make the following assumption.
ASSUMPTION 2. When banks increase notional debt in response to a fall in p, the proportion of
funding raised from the outside creditor sector is non-decreasing.
This assumption places a restriction on the new interbank claims matrix 11* so that
the sum of the ith row of HI is no larger than the sum of the i's row of the initial
interbank matrix H. In other words,
(I- H*)u > (I -
H)u. (22)
We will see shortly some empirical evidence that bears on Assumption 2.
PROPOSITION 2. When p falls, the value of aggregate lending to end-users increases, both in
notional values and in market values.
The argument for this Proposition starts with the balance sheet identities before and
after the change in face values of debt. The balance sheet identities in face values are
y = e + x(I-II) (23) y*= * + -* (I- T)
where * indicates variables after the change. The face value of equity remains
unchanged (e* = e), so that the change in aggregate notional lending is given by
8 See Eisenberg and Noe (2001).
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 323
The first term on the right-hand side is positive from our assumption that banks react
to surplus equity by expanding th4 ir balance sheets, while the second term on the right
hand side is positive from our assumption (22) that an increasing proportion of the
funding comes from the outside,i ector. Thus, (y* - y)u > 0 , so that total lending to
end-user borrowers in terms of n tional values increases.
The argument for the increase i n the market value of loans to end-users following the
decline in p is similar. The balan ce sheet identities in market values before and after
the change are
y = e + x(I-[H) (25)
V*=e* +x*(I-Il*).
The change in the market value Af loans to end-users is
(y- y)u = (e* -
4)u + (x* - x)(I - H)u + x*(H - H*)u. (26)
Equation (26) differs from the analogous one for face values in that the banks' balance
sheets now reflect the capital gain on their loan portolio as given by (e* - e)u, where
e', where e' is the value given in (20). The increased equity is an additional
funding source when loans are v ilued at market values. All three terms on the right
hand side of (26) are positive, ar d so (y* - y)u > 0.
4. Lending Boom
I can now sketch the scenario fo- a lending boom by using the results derived so far.
The first ingredient is the relatioi iship between the probability of default p on the loan
book and the aggregate lending to end-user borrowers, who may be interpreted as
being households who borrow in order to buy a house. Proposition 1 gives a declining
function that maps p to total len ling.
Figure 5 depicts the negative ' elationship between total (notional) lending and p,
where the total lending appears on the horizontal axis. The arrows indicate that for
each level of p, there is an associ ited level of total lending > Iyi. If I further suppose that there i; a macroeconomic feedback going from total lending
to the probability of default, th, n I may expect amplifications that result from the
interplay between strengthening balance sheets and increased lending.9 If increased
loan supply feeds through to mor buoyant aggregate conditions, it is possible to sketch
a scenario for a lending boom. T iUs, for the purpose of illustration, suppose there is a
mapping g which maps aggregat, lending E y to the probability of default p. To be
consistent with the interpretation of higher credit supply leading to more buoyant
conditions, the function g(-) sho Mld be decreasing.
Figure 6 superimposes the fu iction g( ) on Figure 5. Now consider the scenario
where the advent of securitisatior shifts the mix between internal and external funding
that banks use toward greater us( of funding from outside creditors. We may interpret
this scenario as a decline in th e entries of the interbank matrix H such that the
9 Adrian and Shin (2008?>) exhibit evi( lence that expansions of intermediary balance sheets help explain
future growth of GDP components such is housing investment and durable good consumption.
? The Author(s). Journal compilation Cc Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
324 THE ECONOMIC JOURNAL [MARCH
1
p
Fig. 5. Aggregate Lending is Decreasing in p
p
0
xiyl
Fig. 6. Initial Point
proportion of funding raised from outside creditors increases. As argued in the pre
vious Section, such a development increases the aggregate lending to the end-user
borrowers even if the leverage of individual banks (and their value at risk) is
unchanged. In terms of the diagram, the shift to greater use of outside funding can be
represented as a shift to the right of the credit supply function. Figure 7 depicts the
shift in credit supply that results and the consequences of such a shift.
The initial shock from the greater use of outside funding results in a rightwards shift
in the supply of credit curve. The new intersection point is to the bottom right-hand
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 325
p
0 xiYi
Fig. 7. Lending Boom
side of the initial point, associated with a lower probability of default p and greater total
lending to the end-user sector.
Although there are no explicit dynamics in our framework, it is illuminating to trace
out the step-wise adjustment resulting from the one-off shift in the use of outside funding.
The initial shift is a rightwards shift in the credit supply curve which results in higher
aggregate lending for a fixed p . However, the macro feedback effect of greater loan
supply then kicks in, resulting in a decrease in the probability of default. This adjustment
is depicted by the first downward sloping arrow in Figure 7. However, the fall in p results
in greater lending according to the argument for Proposition 1. Greater lending then
feeds to lower p and so on. The new settling point given in Figure 7 is associated with a
substantially lower probability of default as well as a large stock of lending.
4.1. Subprime Lending
At the cost of some additional complexity, it would be possible to incorporate subprime
lending into the story. Suppose that the population of prime borrowers is small relative
to the expansion of total lending as implied by the new crossing point between the
credit supply curve and the macro feedback function g( ) in Figure 7. Then, once all
the prime borrowers have been granted a mortgage, the banking system has to find
additional means of creating assets. One way would be for the banks to lend to each
other. However, as discussed earlier, the aggregate lending of the banking system to
mortgage borrowers must equal the sum of the equity and the borrowing from outside
creditors. Since it is the borrowing from the outside creditors which is increasing, the
funding must ultimately find its way to an end-user borrower.
Once all the prime borrowers in the population have a mortgage, the banks must
find new borrowers in order to expand their balance sheets. The only way they can do
this is to lower their lending standards. Subprime borrowers will then start to receive
? The Author(s). Journal compilation (?) Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
326 THE ECONOMIC JOURNAL [MARCH
funding. The mechanical nature of our framework in which banks simply choose their
balance sheet size masks important questions concerning the short-termist nature of
such lending to subprime borrowers. The answer as to why banks would lower their
lending standards in order to lend to subprime borrowers must appeal to other fric
tions within the banking institutions that allows such short-termism. Distorted incent
ives and shortened decisions horizons induced by agency problems within the bank
would be part of the overall story. See Rajan (2005) and Kashyap et al. (2008) for
discussion of such incentives.
5. Empirical Evidence
Aggregate lending to end-user borrowers by the banking system must be financed
either by the equity in the banking system or by borrowing from creditors outside the
banking system. The empirical counterpart to the sector described as the 'banking
system' is the whole of the leveraged financial sector, which includes the traditional
commercial banking system, but also encompasses the market-based financial system
that plays a role in extending credit to banks and non-banks by borrowing from outside
creditors. In this sense, the leveraged financial sector should be conceived broadly to
include all leveraged institutions, such as investment banks, hedge funds and (in the
US especially) the government sponsored enterprises (GSEs) such as Fannie Mae and
Freddie Mac.
5.1. Evidence from US GSE Mortgage-backed Securities
A complete disaggregation of the funding source for the leveraged financial sector is
not possible due to the lack of detailed breakdowns in the data between funding from
leveraged and unleveraged creditors. A partial picture can be obtained, however, by
examining the holding of US agency and GSE-backed securities.
Figure 8 plots the total holding of US agency and GSE-backed securities broken
down into the identity of the creditor at the end of each year from 2001 to 2007. The
data are from the US Flow of Funds accounts compiled by the Federal Reserve
(table L.210). Leveraged financial institutions include commercial banks, broker
dealers and other securitisation vehicles. The non-leveraged financial institutions
include mutual funds, insurance companies and pension funds. The 'non-financial
sector' includes household, corporate and government sectors. Finally, the 'rest of the
world' category indicates foreign creditors, especially foreign central banks or other
official sector holders. Figure 9 charts the holders by percentage holdings.
The key series for our purposes is the proportion held by other leveraged financial
institutions. We see that US leveraged institutions have been holding a declining
proportion of the total. At the end of 2002, leveraged financial institutions held 48.4%
of the total but by the end of 2007 that percentage had dropped to 36.7%. There has
been a consequent increase in the funding provided by the non-leveraged sector. In
terms of the model, this translates to an increase in the z vector of proportions raised
from outside the 'banking sector' of the model. Notably, the holdings of the 'rest of the
world' category (which itself is mostly accounted for by foreign central banks) has more
than tripled from $504 billion at the end of 2001 to $1,540 billion at the end of 2007.
(? The Author(s). Journal compilation (? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 327
8.0 -
7.0 t rfilfifll
Rest of te world
{ E~~~~~~. .. . . . ...........X
5.0 -: t E3 Non-leveraged
_ <=. I
~~~~~~~~~~~~~financial institutions
4.0 - . . . . . . . . . o :.:....:.:..:: .: . .:.:: .::.:.:.:.:.:.: 0 Leveraged financial . _ , * - -*.. . . . . . . . . -.-. -. . . . . . . . . . . . . .
:::: : . ....:: : : : :.:: :. :.-:. ::.:. :.:. :. :. ::. : institutions 3 0 - . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . ................... . . . . . . . . . . . . .
.. .. . . . . .
20
1.0- _
0.0-.
2001 2002 2003 2004 2005 2006 2007
Fig. 8. Holdinkg of GSE-backed Securities
IOS 00 ,,,,,,,,,,X..Yto IlllllElslXla| ?lllEElXX[|nul?
90%_ nlllntl""||"lTnnTltTlT 80% X|||iE is
Ii Re ist of the |||ss| world sX|s |it
. . . . . . . . . . . . . . ~ ~ . .. .. . .. .. .. .. . . . . . . . _ . .
, , . . , . . . . , , ,,....................I
1:::::::::::::::::::::::::::::: : u Non-leveraged financial
60% :::::::::::::::::: i nstitutions .......:::::.:.::::.:.::::::::::::::::::
. ........... .:.:.::.:.:.:. :-::::::::::::::::.. 1:::::::::::::::::::::::::::::::::::::: 1:::::::::::::::::::::::::::::::..1....era.e finnc.a 5 0 0 :: :: ::: ::: ::: ::: :: ::: :: ::: ::: ::: : : : : : : : : . . . . ... . ,. , . .. . ... .. . . . . . . . .
.....................:.:.:.:.:.:.:.:.:.:.:...... ....... ......::::::::::::::::I... . . . . . . . . . . . . . . . . . . . . . . . . . .
.. .. ~ ~ ~ . . . . . . . .
_ . .. _ _ .. _ _ .. _ _ .. _ _ . ..__ _ _ . . ._ _ _. .._ _ _ .
40% 1 1 ,.I
2001 2002 2003 2004 2005 2006 2007
Fig. 9. Holding of GSE-barked KSecufities (percenta>ges)
C)The Auithor(s). Journal compilation (C Royal Economic Society 2()09
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
328 THE ECONOMIC JOURNAL [MARCH
Recall that the increased proportion of the funding coming from outside the banking
system plays a key role in the development of the lending boom scenario of the
previous Section. We see that the assumption has some empirical support.
5.2. Evidence from the UK
The increased importance of securitisation can be found also in the UK, from the
balance sheet series of Northern Rock, the UK bank that failed in 2007. Northern Rock
was a building society (i.e. a mutually owned savings and mortgage bank) until its
decision to go public and float its shares on the stock market in 1997. In the nine years
from June 1998 (the first year after demutualisation) to June 2007 (on the eve of its
crisis), Northern Rock's total assets grew from 17.4 billion pounds to 113.5 billion
pounds (a constant equivalent annual growth rate of 23.2%). By the eve of its crisis,
Northern Rock was the fifth largest bank in the UK by mortgage assets. Northern
Rock's liabilities reflect both the funding constraints it faced, as well as the way it
overcame those constraints. Figure 10 charts the composition of Northern Rock's
liabilities from June 1998 to June 2007.
Traditional deposit funding did not keep pace with total assets and the gap was
made up primarily by securitised notes and other forms of non-retail funding. The
'other liabilities' category includes interbank funding, short-term notes and covered
bonds. Covered bonds are long-term liabilities written against segregated mortgage
assets. The breakdown between leveraged and non-leveraged holders of these
liabilities is not available, but two points are worthy of mention. First, the securitised
120
U~~~~~~~~~~~~ Real Deost Fig. 1. N n c a a
? e AEurition ? l E i S Retail~~~~ ~ ~~ DelplsitI
ffi~ ~ ~~1- I=" 10 I-I Irv
^~~ ~Fg 10 lote Roc Anulan-Imm eot
The ~ ~ Auhr. jouma coplto I Roa EcnmcScey20
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATION AND FINANCIAL STABILITY 329
notes were of long maturity, with the maturity being around two years; see Shin
(2008b). Thus, it is quite plausib e that a substantial part of these notes were held by
non-leveraged financial institutikns. If this is the case, then Figure 10 would be an
illustration of the increased furding obtained from creditors outside the banking
sector.
5.3. Liquidity Crisis
The consequences of the increa,ed funding of assets by creditors from outside the
banking sector are felt most a(utely when the lending boom turns to bust. The
framework sketched so far is a static one in which loans are made at date 0 and
repaid at date 1. This assumptio i masks the maturity mismatch that can build up in
the aggregate balance sheet whe i the loans to the end-user borrowers are long term,
while the debt is short term. In 1he expansion stage, the maturity mismatch does not
show up but the mismatch mal;es the contraction stage more painful due to the
irreversibility of long-term loans. The contraction stage must make reference to more
than two dates.
The simplest extension would l e to have three dates, 0, 1 and 2, where loans {yi} are
granted at date 0, and are repaid tt date 2 but then banks experience an increase in the
probability of default p in the Vc sicek credit risk model at date 1. Then, value at risk
increases following the paramet lr shift, while the market value of equity decreases.
Indeed, there is the chain of ine iualities:
e' < e, < V,' (27)
which is the mirror image of th inequalities (20) that hold during the boom. The
counterpart to Assumption 1 wc uld be that the banks attempt to reduce their total
balance sheet size by reducing t e size of their notional debt level -i. However, if the
loans to end-users -i are long ter -n, then banks are not free to reduce the size of their
balance sheets flexibly. The con traint will bind harder if the proportion of assets of
this type constitute a large fraction of total assets.
In aggregate, the total long-tei m lending to end-users is mirrored by the size of the
funding obtained from lenders from outside the banking sector. Thus, at an aggregate
level, the increased use of securi tisation is associated with a tighter constraint against
rapid reductions in balance shef t size. When value at risk increases, banks must cut
back the size of their balance shi ets. Some banks will be able to reduce their balance
sheets flexibly by not rolling ovc r short-term assets and short-term liabilities. But not
every bank can do this, since thc financial system as a whole holds long-term illiquid
assets financed by short-term liqL id liabilities. There will be 'pinch points' that are thus
exposed when value at risk incrc ases. These pinch point banks will suffer a liquidity
crisis.
Northern Rock is a good exar iple of such a pinch point. Its assets were almost all
long-term residential mortgages. Thus, it had very little scope to reduce its balance
sheet in a flexible way once the :risis struck.
Crucially, Northern Rock was v alnerable to the tick-up in value at risk due to its high
leverage. Figure 11 plots the l verage series from June 1998 to December 2007
C The Author(s). Journal compilation (C Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
330 THE ECONOMIC JOURNAL [MARCH
90
80
70 -
Leverage on total equity/5
60 - ? Leverage on shareholder equity
50 Leverage on common equity
40
30
20
10
\0 \, C) C) g) o ? oO ?) o o ? o
Fig. 11. Northern Rock's Leverage June 1998 - December 2007
according to three different measures of equity. In the early years, there was no dis
tinction between total equity, shareholder equity and common equity. All equity was
common equity. However, in 2005, the total equity series included for the first time
736.5 million pounds worth of subordinated debt, as well as 299.3 million pounds
worth of reserve notes. Both of these items had been issued much earlier (in 2001) but
they were included in the equity series in the annual report for the first time in 2005.10
The inclusion of these subordinated debt items introduced a jump up in the equity
series for Northern Rock, and accounts for the sharp jump down in the leverage series
in June 2005 in Figure 11. However, as we can see also from Figure 11, when the
subordinated debt items are excluded, and equity is construed just as shareholder
equity, the leverage series continues to move up in 2005. On the eve of its crisis in 2007,
the leverage ratio on common equity stood at almost 60. This made Northern Rock
particularly vulnerable among UK banks to the credit crisis that erupted in August
2007.11
Although the conventional notion of equity in bank regulation is as a buffer against
losses to depositors, the relevant equity measure for funding constraints is the stake
held by those who control the assets - i.e. the common equity holders. The distinction
is seen most clearly for repurchase agreements (repos), where the haircuts applied by
creditors determine the permitted leverage; see Adrian and Shin (2008a). Thus, even
though the Basel-style leverage ratios were kept low by issuing subordinated debt and
10 See Shin (2008a). 11 See Yorulmazer (2008) for an empirical analysis of UK banks at the time of the Northern Rock crisis.
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
2009] SECURITISATI )N AND FINANCIAL STABILITY 331
preferred equity, the effective lev( rage for funding purposes was climbing very rapidly
in the case of Northern Rock. Its run in the summer of 2007 should be seen in this
context.
6. Related Literature and Con clusions
The importance of securitisation for financial stability derives from the ability of the
shadow banking system to increase total supply of credit to end-users. When there is a
decline in the riskiness of fundimental assets (in our case, through a fall in the
parameter p), the risk-taking capX city of the shadow banking system increases.
The idea that the changes in the lender's balance sheet is important in determining
the supply of credit has also figured in an earlier literature that has emphasised the
liquidity structure of the banks' b ilance sheets (Bernanke and Blinder, 1988; Kashyap
and Stein, 2000), or the cushionmng effect of the banks' regulatory capital (Van den
Heuvel, 2002). The supply-side mechanism fo- the growth of credit should be distinguished from
the larger literature on the fluct, ations in credit due to the shifts in the demand for
credit, as emphasised for instanc by Bernanke and Gertler (1989) and Kiyotaki and
Moore (1998, 2001). The key to the demand-side explanations of the fluctuation of
credit is the changing strength of the borrower's balance sheet and the resulting
change in the creditworthiness o the borrower.
The mechanism proposed her, for the origin of the subprime crisis has more in
common with the supply side explanation. The greater risk-taking capacity of the
shadow banking system leads to an increased demand for new assets to fill the
expanding balance sheets and ar increase in leverage. The picture is of an inflating
balloon which fills up with new as: ets. As the balloon expands, the banks search for new
assets to fill the balloon. They look for borrowers that they can lend to. However, once
they have exhausted all the good borrowers, they need to scour for other borrowers -
even subprime ones. The seeds ol the subsequent downturn in the credit cycle are thus
sown. According to the picture pair ted here, the subprime crisis has its origin in the
increased supply of loans - or ec uivalently, in the imperative to find new assets to fill
the expanding balance sheets. Ir this way, it is possible to explain two features of the
subprime crisis - first, why appart ntly sophisticated financial intermediaries continued
to lend to borrowers of dubious -reditworthiness and, second, why such sophisticated
financial intermediaries held the bad loans on their own balance sheets, rather than
passing them on to other unsu specting investors. Both facts are explained by the
imperative to use up slack in bo lance sheet capacity during an upturn in the credit
cycle.
Princeton University
References
Adrian, T. and Shin, H. S. (2007). 'Liqu dity and leverage', Journal of Financial Intermediation, forthcoming, available at http://www.princeton.ee u/~hsshin/working.htm
?D The Author(s). Journal compilation (C Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
332 THE ECONOMIC JOURNAL [MARCH 2009]
Adrian, T. and Shin, H. S. (2008a). 'Financial intermediary leverage and value at risk', working paper, Federal
Reserve Bank of New York and Princeton University, available at http://www.princeton.edu/~hsshin/
working.htm Adrian, T. and Shin, H. S. (2008&). 'Financial intermediaries, financial stability and monetary policy',
paper for the Federal Reserve Bank of Kansas City Symposium at Jackson Hole, 2008, available at http://
www.princeton.edu/~hsshin/working.htm Alizalde, A. and Repullo, R. (2006). 'Economic and regulatory capital in banking: what is the difference?',
working paper, CEMFI.
Ashcraft, A. and Schuermann, T. (2008) 'Understanding the securitization of subprime mortgage credit', Staff
Report No. 318, Federal Reserve Bank of New York, available at http://www.newyorkfed.org/research/
staff_reports/sr318.pdf Bank for International Settlements (2008). 78th Annual Report, Basel: BIS.
Bernanke, B. and Blinder, A. (1988). 'Credit, money and aggregate demand', American Economic Review, vol. 78, pp. 435-9.
Bernanke, B. and Gertler, M. (1989). 'Agency costs, net worth, and business fluctuations', American Economic
Review, vol. 79, pp. 14-31.
Brunnermeier, M. (forthcoming). 'De-ciphering the credit crisis of 2007', Journal of Economic Perspectives.
Demyanyk, Y and van Hemert, O. (2007). 'Understanding the subprime mortgage crisis', working paper, New
York University, Stern School of Business.
Eisenberg, L. and Noe, T. H. (2001). 'Systemic risk in financial systems', Management Science, vol. 47, pp. 236 49.
Gieve, J. (2008). 'The return of the credit cycle: old lessons in new markets', speech at the Euromoney bond
investors congress, February 27, available at http://www.bankofengland.co.uk/publications/speeches/
2008/speech338.pdf Gorton, G. (2008). 'The panic of 2007', paper for the Federal Reserve Bank of Kansas City Symposium at
Jackson Hole.
Gorton, G. and Souleles, N. (2006). 'Special purpose vehicles and securitization', in (R. Stulz and M. Carey, eds), The Risks of Financial Institutions, pp. 549-97, Chicago: University of Chicago Press.
Greenlaw, D., Hatzius, J., Kashyap, A. and Shin, H. S. (2008). 'Leveraged losses: lessons from the mortgage market meltdown', Report of the US Monetary Monetary Form, No. 2, available at http://www.chica
gogsb.edu/usmpf/docs/usmpf2008confdraft.pdf International Monetary Fund (2008). Global Financial Stability Report, April, Washington DC: IMF.
Kashyap, A., Rajan, R. and Stein, J. (2008). 'Rethinking capital regulation', paper for the Federal Reserve Bank
of Kansas City Symposium at Jackson Hole.
Kashyap, A. and Stein, J. (2000). 'What do a million observations on banks say about the transmission of
monetary policy?', American Economic Review, vol. 90, pp. 407-28.
Keys, B., Mukherjee, T., Seru, A. and Vig, V. (2007). 'Did securitization lead to lax screening? Evidence from
subprime loans', working paper, University of Chicago GSB.
Kiyotaki, N. and Moore, J. (1998). 'Credit chains', LSE working paper, available at http://econ.lse.ac.uk/
staff/kiyotaki/creditchains.pdf.
Kiyotaki, N. and Moore, J. (2001) 'Liquidity and asset prices', LSE working paper, available at http://
econ.lse.ac.uk/staff/kiyotaki/liquidityandassetprices.pdf. Mian, A. and Sufi, A. (2007). 'The consequences of mortgage credit expansion: evidence from the 2007
mortgage default crisis', working paper, University of Chicago GSB.
Milgrom, P. and Roberts, J. (1994). 'Comparing equilibria', American Economic Review, vol. 84, pp. 441-59.
Mishkin, F. (2008). 'On leveraged losses: lessons from the mortgage market meltdown', discussion at 2nd US
Monetary Policy Forum, New York, February 27, available at http://www.federalreserve.gov/newsevents/
speech/mishkin20080229a.htm
Rajan, R. (2005). 'Has financial development made the world riskier?', Proceedings of the Federal Reserve
Bank of Kansas City Symposium at Jackson Hole, available at http://www.kc.frb.org/publicat/sympos/
2005/sym05prg.htm Shin, H. S. (2008a). 'Risk and liquidity in a system context', Journal of Financial Intermediation, vol. 17, pp. 315
29.
Shin, H. S. (2008?). 'Reflections on Northern Rock: the bank run that heralded the global financial crisis',
Journal of Economic Perspectives, forthcoming. Van den Heuvel, S. (2002). 'The bank capital channel of monetary policy', working paper, Wharton School,
University of Pennsylvania, available at http://finance.wharton.upenn.edu/~vdheuvel/BCC.pdf Vasicek, O. (2002). 'The distribution of loan portfolio value', available at http://www.moodyskmv.com/
conf04/pdf/papers/dist_loan_port_val.pdf Yorulmazer, T. (2008). 'Liquidity, bank runs and bailouts: spillover effects during the Northern Rock episode',
working paper, Federal Reserve Bank of New York.
? The Author(s). Journal compilation ? Royal Economic Society 2009
This content downloaded from 147.143.2.5 on Fri, 6 Feb 2015 10:53:27 AM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 309
- p. 310
- p. 311
- p. 312
- p. 313
- p. 314
- p. 315
- p. 316
- p. 317
- p. 318
- p. 319
- p. 320
- p. 321
- p. 322
- p. 323
- p. 324
- p. 325
- p. 326
- p. 327
- p. 328
- p. 329
- p. 330
- p. 331
- p. 332
- Issue Table of Contents
- The Economic Journal, Vol. 119, No. 536, Conference Papers (Mar., 2009), pp. 309-642, i-vi
- Front Matter
- Securitisation and Financial Stability [pp. 309-332]
- On Reputation: A Microfoundation of Contract Enforcement and Price Rigidity [pp. 333-353]
- Diverse Beliefs, Survival and the Market Price of Risk [pp. 354-376]
- Learning, Adaptive Expectations and Technology Shocks [pp. 377-405]
- Did the Single Market Cause Competition in Excise Taxes? Evidence from EU Countries [pp. 406-429]
- Job Satisfaction and Co-Worker Wages: Status or Signal? [pp. 430-447]
- You're Fired! The Causal Negative Effect of Entry Unemployment on Life Satisfaction [pp. 448-462]
- Fired or Retired? A Competing Risks Analysis of Chief Executive Turnover [pp. 463-481]
- Valuing Air Quality Using the Life Satisfaction Approach [pp. 482-515]
- Changes in Compulsory Schooling, Education and the Distribution of Wages in Europe [pp. 516-539]
- Teachers' Training, Class Size and Students' Outcomes: Learning from Administrative Forecasting Mistakes [pp. 540-561]
- Implications of Endogenous Group Formation for Efficient Risk-Sharing [pp. 562-591]
- Homo Reciprocans: Survey Evidence on Behavioural Outcomes [pp. 592-612]
- The Role of Information Revelation in Elimination Contests [pp. 613-641]
- Back Matter
Should Financial Stability Be Assigned to Public Policy.pdf
Should Financial Stability Be Assigned to Public Policy? Author(s): Michael Debabrata Patra Source: Economic and Political Weekly, Vol. 38, No. 23 (Jun. 7-13, 2003), pp. 2271-2275+2277- 2283 Published by: Economic and Political Weekly Stable URL: http://www.jstor.org/stable/4413656 .
Accessed: 13/02/2015 16:12
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].
.
Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend access to Economic and Political Weekly.
http://www.jstor.org
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
Special articles
Should Financial Stability Be Assigned to Public Policy?
In the light of the experience with the severe financial crises of the 1990s, the responsibility for financial stability has implicitly been assigned to public policy, overturning, in a sense, the
dominant paradigm until then of regarding financial development, including stability, as a function best performed by the financial markets. This paper undertakes a critical examination of this assignment, its magnitude and quality, by questioning its analytical underpinnings. The
paper examines the search for the appropriate international financial architecture as the virtuous approach to the assignment and concludes that the identification of international
standards and codes for adoption by countries may be a suboptimal approach. On the other hand, establishment of an international bankruptcy mechanism holds promise of filling a major
gap in the efforts to strengthen the international financial architecture.
MICHAEL DEBABRATA PATRA
Introduction In the aftermath of the global financial crises of the late 1990s,
financial stability dominated the agenda of discussions among academics, policy practitioners and the private sector alike.
Albeit temporarily overshadowed by the worldwide downturn in economic activity, it returned to centre stage with the collapse of the Argentine currency peg, and the spate of financial collapses in the US, including the spectacular bankruptcy of the energy giant, Enron. Several interrelated fora are engaged in the dis- cussion on financial stability - the IMF, the World Bank, the Financial Stability Forum conveying essentially the G-7 view- point, the G-20, the G-22 and G-24 incorporating poor country perspectives, the International Financial Institution Advisory Commission authorised by the US Congress (which produced the Meltzer Report in March 2000), several vistas of independent work being undertaken under the aegis of the Group of Thirty, the Institute of International Economics, the Independent Task Force on the Future International Financial Architecture spon- sored by the Council on Foreign Relations, to name a few. Efforts have gone into developing leading indicators of financial crises. And by no means the least, there is the Bank for International Settlements (BIS) and the Basel process, with Basel I entrenched in almost all countries across the world as subordinated national regulations for almost a decade. In the absence of the appropriate international architecture for ensuring financial stability, intense efforts have been undertaken unilaterally to strengthen and monitor domestic financial systems and in general to seek the world's best standards and codes in various areas relating to the operation of financial systems. Due lip service is being paid to the need for adapting these best practices to the specifics of the country situation. Furthermore, when crises have struck, sovereigns have pre-emptively and prospectively taken recourse to discretionary action exemplified by capital controls and bail-outs for national
financial institutions. All these efforts implicitly assume that public policy, whether national or international, has a respon- sibility for ensuring financial stability.
At one end of the spectrum is the United Nations (UN), principally through its conference on Trade and Development (UNCTAD), which questions the possibility of the appropriate international financial architecture vis-a-vis the developing countries' interests and urges alternative possibilities - inter- national bankruptcy procedures, standstills, orderly debt workout procedures and innovative creditor-debtor relationships - which essentially relate to burden sharing arrangements initially en- forced by public policy rather than an abiding responsibility. Recently an advocacy for this view has emerged from an un- expected quarter - the IMF - with deputy managing director, Anne Krueger's impassioned statements in favour of an inter- national bankruptcy mechanism. At the other end, there is sub- stantially strong advocacy for self-regulation and discipline, industry standards and pre-commitment approaches assigning the responsibility for financial stability predominantly to the private sector through market pressure and limiting the role of public policy to essentially the provision of incentives for prudent behaviour with risk insurance for only the extremely contagious situations. These articulations are often found in reports and occasional papers issuing from the Group of Thirty and the Global Governance journal. The US financial system is often held up as a virtuous model of this 'laissez faire' approach.
The theory and practice of financial development has under- gone major paradigm shifts in the last three decades. Since the 1970s, coincident with the dominance of private capital in international financial flows and the intellectual rationale pro- vided by the McKinnon-Shaw hypothesis (and the subsequent work by Maxwell Fry), many countries abandoned erstwhile dirigiste regimes, dismantling instruments of financial repression to participate in the process of globalisation and to accelerate the pace of national economic growth. At the turn of the century,
Economic and Political Weekly June 7, 2003 2271
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
the paradigm seems poised on the threshold of a major shift. Increasingly, efficiency in financial intermediation that powered the liberalisation of domestic financial systems through the second half of the 1990s has come to be overshadowed by concerns relating to financial stability. By the end of the 1990s, financial crises had engulfed almost three-fourths of the membership of the IMF, including the US in the late 1980s. Given the pervasive nature of financial instability and the massive costs of restoration and losses of welfare - ranging between 3 per cent of GDP in the US to 40 per cent of GDP in Chile - which inevitably involves the sovereign governments, the view gathering critical mass is that financial stability involves collective action. Beyond this point, however, any semblence of a consensus disintegrates and questions relating to rebalancing the relationship between the State and the markets, the nature and extent of public policy intervention in financial market processes, the role of the private sector and the validity of the search for and establishment of the international architecture and its intrusiveness, continue to raise heat and dust. Within the intense debate, there is also the nagging worry that developments are going forward without the intel- lectual backing of innovations in economic theory and that the chasm between theory and reality is widening beyond repair.
It is against this backdrop that this paper seeks to address the subject of financial stability and the role of public policy. At the outset it is recognised that the subject is lacerated with deep- seated disagreements on even conceptual and definitional issues - an agreement on what is meant by financial stability has proved elusive. Conscious of dealing with unsettled issues shrouded in various shades of grey, the paper attempts to take an eclectic and non-partisan approach. More often than not, the approach is clearly agnostic and dwells at length on arguing the counterfactual. The purpose is an assimilation of ideas, or rather an assimilation of questions that are being asked relating to the theme, under the belief that if the right questions are raised, the answers cannot be far behind. The methodology underlying this approach is a combination of personal interviews with a wide spectrum of participants in the financial system, academics and public institutions in the US, and a critical assessment of the proliferating literature on the subject. Although the paper is based primarily on the financial system in the US, it has a developing country focus. The questions raised and the issues examined relate equally to all countries.
The rest of the paper is divided into five sections. Section II deals with the issue of whether or not financial stability should be a goal of public policy. Section III assesses the spectrum of opinion on the nature and extent of public intervention in finan- cial markets and institutions. Section IV examines the efforts at constructing the appropriate international architecture including an examination of the recent focus on the implementation of international standards and codes as a generic form which has revealed Darwinian 'survival-of-the-fittest' characteristics. The final section contains concluding observations.
II Why Should Financial Stability be a Goal
of Public Policy? There is a broad consensus in the responses received affirming
the conventional assignment of financial stability to public policy. At one level, this convention seems to be derived from at least five centuries of history and tradition. Indeed, this is the reason why several public institutions - notably central banks, and in view of the recent recognition of core competencies, agencies like Securities and Exchange'Commissions and other supervisors
of specific aspects of the financial system - are established. Abstracting from principal-agent relationships, the provision of financial stability by public policy is seen as analogous to the guaranteeing of price stability, i e, societies desire a money they can trust. Financial systems are equipped with an awesome leverage - the ability to convert relatively illiquid assets into multiples of highly liquid liabilities - and their constituents are able to take on significant risks and dispel asymmetries in in- formation availability. But when disaster strikes, all this adds up to a recipe for systemic explosion that can wipe out portions of national income and welfare, and even global growth. Ac- cordingly, the national regulators and the multilateral institutions regard as a fait accompli the role of public policy in providing, or more correctly, in ensuring the conditions that secure financial stability. Surprisingly, similar strains emanate from the specialised research institutions although the heaviness of the public policy hand upon the financial system is argued in various shades.
Beyond this point, however, there is a pursuit of self-interest. Private investors view public policy involvement in the provision of financial stability as necessary for creating investor confidence by creating level playing fields (to be read as "opening up" or removing barriers to markets rather than any determined pursuit of rules for fair play), ensuring free entry, full repatriability of capital, international rates of taxation on profits and capital gains, and by providing bail-outs. Internationally active banks respond with greater caution and conventional wisdom recognising the special role of banks in the financial system and the possibility of single big failures starting off cascades of defaults which only entities with vastly superior leverage could stem. The role of public policy, however, is in facilitating rather than providing financial stability. Moreover they see financial markets as not self regulatory and, therefore, checks and balances become necessary although, a balance is required between regulation and intervention.
The rating agencies exhibit similarities in responses, but at the other end of the swing. They see no role for public policy in the entrenchment of financial stability. The involvement of public policy is viewed as a justification for the existence of public institutions, as having turned the quest for stability into a political process and that public policy will do what it has to do. Occasional financial market failures are regarded as Schumpeterian 'cre- ative' destruction. Policy involvement has tended to generate greater instability as in the case of Japan. Even in the US, regulatory public authorities are seen as too interventionistic. The rating agencies hold that the interests of financial stability are best served under a system of self regulation, self assessment and risk management.
The UN's approach to financial stability has been primarily in the context of the issues relevant in financing economic development, without finance endangering the prospects of growth in the developing world. This has often led it to adopt positions that oppose the Washington consensus and the G-7 point of view. The UN's approach has a clear historical priority dating from 1977,' The essence of this approach is that it is difficult for govern- ments not to get involved in the responsibility for financial stabi- lity since the private sector cannot provide it. Governments have to provide public money in the resolution of instability and expe- rience has shown that the costs of failure are larger than the costs of prevention. The UN clearly stands at the opposing side of those who advocate market solutions or market-based deployment of public policy. Its stand is captured in the Trade and Development Report, 2001: markets can and do get it wrong, and for developing and developed countries alike. The onus is still on policy makers to find preventive measures and appropriate remedies.
2272 Economic and Political Weekly June 7, 2003
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
All this, however is not convincing enough for making financial stability a goal of public policy. The symposium sponsored by the Federal Reserve Bank of Kansas City in 1997 may be taken as a convenient starting point for examining the recent literature on the theme. The case for public policy has been made on the propositions that the financial system is prone to bouts of in- stability and instability generates large negative externalities (contagion) [Crockett 1997, 1998,2000,2001]. Taking the second issue first, the defence could perhaps rest if financial stability can be defined as a public good like public health or global peace, thus validating the involvement of public policy in its provision. It partakes of at least one characteristic of i public good, i e, as mentioned earlier, there are externalities involved. It is non- rival and non-excludable in its consumption. In this sense strictly, it is a global public good in that stable financial systems and well-functioning financial markets have benefits that reach across borders, generations and economic groupings. Therefore, it should be the concern of global public policy and typical of the provision of public goods, collective action through international coopera- tion is required to ensure financial stability. This is even more valid in the age of globalisation and geography-defying capital flows. So far, however, the world has not been able to assign financial stability to collective action through international cooperation and the search for the appropriate international financial architecture is still on. In the interregnum, financial stability is being assigned to national policies and immediately the case for financial stability being a public good breaks down. For in this context, national governments correspond to self interest. The essence of developing countries' policies is to delay the inexorable convergence to steady state in the face of massive poverty and therefore, squiggles like crony capitalism, over- investment booms and imprudent lending create temporary surges of growth and promises of returns to internationally mobile capital. In the same vein, much of the problems of debt and currency risk exposure in these countries can be traced to the failure of internationally coordinated actions to remove misalignments in the G-3 currencies. Once this argument is allowed to enter, the case for public policy intervention in ensuring financial stability recedes even further, for collective action at the national level to provide financial stability can involve combinations of private initiatives as interested/affected parties, partnerships between public and private parties with the public policy as facilitator of a coalition of private forces a la Long-Term Capital Markets (LTCM). Several combinations of affected parties can be conceived but the case for financial stability as a goal of public policy does not emerge.
In several senses, financial stability has characteristics of a private good in that it is a market outcome. Hedging and risk management instruments are traded in markets, risks are priced. Market processes are excludable in thesense that there have to be risk takers just as there are risk transferers and both are elements of competitive and stable equilibria. Markets are made on two way bets. If there was only one view, there would not be markets. In many ways, public intervention, by allowing and guaranteeing leverage, could interfere with stability that markets could inherently provide. If there was no leverage, the risk takes would absorb the losses from risk taking and this would virtually eliminate the negative externalities associated with financial instability [Greenspan 1997]. Accordingly, failures (without large scale spillovers) are an important part of the market process and provide 'discipline and information'. In the US financial system, bankruptcy procedures are lauded and upheld as the virtuous model of financial stability. Yet as in the cases of Costa Rica in 1982 and more recently of Peru in 1997 which are cited later
on in this paper, it has become clear that financial stability is excludable when it involves parties affected beyond borders, and there is obviously national ownership in US bankruptcy proce- dures. The same argument could be made about countries adopt- ing capital controls, or even discretionary monetary and exchange rate policies. Thus, there seems to be a flaw in the argument for financial stability as a goal of public policy. Equity must be an important element in the provision of a global public good and therefore, financial stability is best treated as ajoint global public good, if at all.
Financial stability requires stable financial institutions as well as stable financial markets. The advocacy for making financial stability a goal of public policy is persuasive in offering extremely valid arguments why financial institutions should be intervened - vulnerability of institutions to runs of public confidence, depositor protection, contagion, budgetary, GDP and social costs of resolution including into the future - but discourages inter- vention in financial markets which suffer from the same insta- bility bias and contagious effects [Crockett 1997, 2001]. There is a contradiction in terms here. Instability in financial institutions occurs because of their inability to handle large fluctuations in market prices. And at least some part of the movement in financial market prices is acknowledged to be often out of alignment with fundamentals [Crockett 1997]. Financial markets are prone to bouts of instability, and yet or even so, markets are better equipped for determining prices than public policy! An important element of the quest for financial standards and codes is based on marking portfolios to market notwithstanding the ideosyncratic values that the financial markets throw up!
Yet another flaw in the argument for preserving the competitive flavour of market processes in determining financial prices is the assumption that the law of large numbers prevails. Existing risk management and asset pricing techniques employed by the market assume normal distributions which is actually convenient since it enables doing away with specification and testing of the complex underlying relationships. What is more is obtaining the 'right signs' empirically. The biggest problem in the evaluation of risk is the fat-tail problem and "...once you start putting in non-normality assumptions, which unfortunately is what characterises the real world, then these issues become extremely difficult" [Greenspan 1997]. The growing complexity of finan- cial innovations makes it even more difficult to understand intuitively the risks facing financial institutions from markets. Mathematics of existing market practices seem to have come unstuck from the realities of the market place.
Finally, in the absence of a consistent and coherently crafted theoretical framework free of contradictions, an incremental approach is being adopted - of making graduated progress in several directions. This runs the high risk of producing multiple equilibria and instability as the process enforces what are being seen as the right practices into market behaviour under the garb of creating the right incentives for prudent market agents. And one wonders whether at the end of it all, we are not heading our financial systems back to financial repression regimes that characterised the world till the 1970s. They produced large bureaucracies, massive corruption, inefficiency in pricing and allocation of resources and scope for the extraction of revenue for the government, but with one important difference from the financial systems of the present day: they provided financial stability; episodes of financial distress were extremely limited!
It needs to be mentioned in the nature of a postscript to this section that the existence of provisions of the Community Reinvestment Act in the US and budgetary preemptions for certain sections of society is a tacit recognition of the possibility
Economic and Political Weekly June 7, 2003 2273
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
of market failure. Such interventions exist in almost every society in the world today. Indeed even in the US there is gloved support for the Fed's interventions in the money, credit and foreign exchange markets; however, interventions in the equity markets raises heckles.
Ill Nature and Extent of Public Intervention
in Financial Systems All countries adopt policies to ensure and promote financial
stability. Underlying the deployment of these policies is recog- nition of what markets cannot do. The foremost among the limitations of financial markets is that they cannot, by themselves, guarantee financial stability. Indeed, in each strength attributed to the functioning of markets, there is a limitation and a rec- ognition of these limitations determines the nature and content of policy intervention.
Markets are efficient in the allocation of resources mainly because market prices are a congellation of all available infor- mation which is required by market agents to make saving- investment and allocation decisions. There is, however, a fun- damental tendency, and particularly among failing entities, to under-supply information to the market. Disclosures by Enron employees under oath to congressional committees are an example. Consequently, public policies are required to ensure effective and comprehensive disclosure of information. More- over, public authorities have access to privileged and classified information. In terms of supply considerations alone, therefore, public policies can be proactive and preemptive in creating information conditions for markets to work efficiently, and more importantly, in averting market failures. At this stage itself, it is necessary to take on board the view emerging out of the office of the comptroller of the currency in the US that markets do not efficiently allocate resources in a manner which maximises social welfare and, therefore, have to be intervened in the interest of the greater common good. This intrusion, however, needs to be restricted to the banking segment alone and in this sense, the Gramm-Leach-Bliley Act does a disservice by expanding the scope of banking activity. Non-bank financial companies, secu- rities and insurance firms cannot be subjected to this onerous social responsibility. Accordingly, while the comptroller allowed in December 1996 national banks to own subsidiaries in any segment of the financial services industry, they were required to deduct any investments in their subsidiaries from regulatory capital.
Markets are able to assess risk-return trade-offs, to price expectations into valuations, to obtain assessments of fair value through appropriate processes of discounting. This is, however, not a specific superiority of markets. There is no evidence to suggest that market participants process information better than public authorities. In fact, as mentioned earlier, existing models of risk evaluation are so flawed that they do not reflect market realities. Often, herds of market participants can make similar mistakes in the processing of information and 'bubbles' occur. Public authorities have no recognised superiority in the process- ing of information over market participants. Undoubtedly, they may be inefficient - they are grossly underpaid for the same skill - and regulatory requirements may sometimes be uneconomical by market standards. Moreover, they may not be able to keep pace with innovations in the market place; one example is that of the identification and measurement of risks. Experience suggests that they have been able to at least enforce prudential behaviour in market participants by this function. The challenge before
public policy authorities in a world of fast paced changes is to constantly balance the conflicting pulls of overbearance and forbearance to compensate for possible shortfalls in the ability to 'read' information.
Public policy intervention in providing the appropriate infra- structure for the efficient functioning of markets - legal, insti- tutional, informational and 'plumbing', i e, the payment and settlement system - is virtually unquestioned. Indeed, this aspect of public policy is regarded as developmental, irrespective of the level of development of the country in question. Markets, instruments and practices are constantly evolving and in order that these changes are frictionless and merge into a continuum, public action is warranted even if in a vanguard role, as it typically is.
Perhaps the most significant aspect of public policy engage- ment in the responsibility for financial stability is that of regu- lation and supervision of the financial system. Here the debate revolves not so much on 'why' but 'how', i e, the context and therefore, the content. Depositor (or consumer) protection and the problem of systemic risks constitute the dominant rationale for public regulation of the financial system, buttressed by the requirements of monetary stability and efficient financial inter- mediation. Increasingly in the recent period, however, there is a sensitivity to what regulation should not do - prevent badly managed institutions from failing, substitute for market processes in evaluating risk-return relationships and in setting financial prices, distort the playing field in favour of interest or pressure groups.
The content of regulation, as it has existed so far, goes under a generic nomenclature provided in an important work contrib- uted by the Brookings Institution - 'the prevention safety-net approach' to maintaining financial stability [Litan 1997]. It contains crisis prevention policies as well as safety nets, the latter including deposit insurance and emergency liquidity injections which represent the lender of last resort function going back to the time of Bagehot (1873). Crisis prevention measures have taken the form of activity restrictions (as under the Glass-Steagal Act in the US), competition shelters, credit and interest rate controls and it is only since the 1980s, that these instruments are being replaced by risk-based capital adequacy. In the US, numerical minimum capital-to-asset ratios were first specified in 1981 and made generally applicable in 1983; however, total assets were not risk-adjusted and this led to increasing exposure to off-balance sheet transactions [Wall and Patterson 1996]. In response came the internationally (primarily industrial country sponsored) agreed Basel capital standards in 1988 that took full effect in 1992. The combined effect of the Basel Accord and the Prompt Corrective Action introduced under the Federal Deposit Insurance Corporation Improvement Act of 1991 made capital ratios the primary instrument of financial regulation not only in the US and other industrialised countries, but all over the world, irrespective of the fact that they were originally intended for internationally active banks. Capital ratios were also a defence against the marauding competition from Japanese banks faced primarily by European banks.
It is important to recognise that the capital standards contain as much a component of moral hazard as implicit in the lender- of-last-resort function. All prevention measures including capitalisation are intended to economise on provision of emer- gency liquidity so that lender-of-last resort functions can be reserved for 'catastrophic financial insurance coverage' [Greenspan 1997]. As the experience of the 1990s has shown, there is a greater propensity for riskiness - capital standards have provided strong incentives to allow banks' capital ratios to fall and their portfolio risks to rise, and depositors show less incentive
2274 Economic and Political Weekly June 7, 2003
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
to monitor banks' activities once capital ratios are achieved. The early experience with the capital standards in the 1980s, particu- larly in the US, and in other countries in more recent times, indicates that these prevention strategies encouraged forbearance among the regulatory authorities as well. Moreover, the Basel norms set a minimum capital ratio, not a maximum insolvency probability [Greenspan 1998]. The ratios explicitly account for credit risk and market risk, but not other types of risk, notably operating risk. Moreover they do not take into account hedging and portfolio management strategies (except in trading books). Banks are indulging in cosmetic changes to their portfolios by reducing total assets to improve the capital-assets ratio while increasing portfolio risk since the risky assets stay on. They also exploit the differences between regulatory (required) capital and economic (true value) capital by securitising and removing the safest assets from the balance sheet as also by refusing to recognise reductions in the market value of assets.
It is necessary also to take note of the fact that while bail-outs can be reserved for the gravest financial disorders or the ones with maximum systemic portent in developed countries, it is often required to be the first line of defence in developing countries. The importance of public policy in a developmental role in these countries makes public policy intervention an act of faith, a measure of restoring investor confidence. Accordingly, private capital flows are willing to visit these countries when they open up only under some form of guarantee, explicit and implicit, that their claims would be regarded as senior and that they would be bailed out in an emergency. This has led developing countries, individually, in the G-77 and most notably in the UN, to question the asymmetry in the international financial architecture in which imprudent borrowing is always punished while imprudent lend- ing goes scot-free. If public money has to be inevitably pumped in by governments in developing countries to ensure financial stability, then these governments would need to have a say in the ways in which financial institutions and even financial markets function. The exercise of capital controls, exchange and payments restrictions are ways in which market seizures are confronted. In the interest of financial stability, should governments in developing countries have oversight over cross-border financial transactions involving national entities? Can they, for instance, abrogate financial contracts that are perceived to have disastrous systemic consequences in the future?
Quo Vadis Basel II?
Criticisms of the Basel norms have been levied not only by official regulators and supervisors, but also by independent bodies such as the US Shadow Financial Regulatory Committee and in the empirical literature on bank capital regulation (Santos, 2000 provides a comprehensive summary). This has also led the Basel Committee on Banking Supervision to propose a new capital adequacy framework in June 1999, subsequently revised in the light of comments and released in January, 2001 with a timeframe for formal implementation by 2005. The new proposal has been widely read and commented upon. Each of the pillars on which it rests - minimum capital standards, supervisory review and market discipline - reflects an awareness and an accommo- dation of the rapidly changing financial environment in which the new standards are evolving and the decade of experience with the Basel I standards. Comments and criticisms from over 140 countries and a large number of participants in the financial system are available in the public domain.
Besides the 'dense, regulatory style' [Greenspan 1998] and the maze of regulatory complexity it entails, some specific criticisms
are noteworthy, especially those that emanate from market participants. Despite the enlargement in the number of risk buckets, the differences in risk weights across risk buckets are felt to be disproportionate to credit spreads, especially for cor- porate bonds. Variations within risk buckets remain large relative to variations across risk buckets - 'A' rated companies have the same risk weights as companies rated lower. Many participants in US markets feel that unrated companies are being treated as favourably as rated companies, presumably to enlist support from countries where firms are not rated. Moreover, the same sum- mation-of-risks approach to measuring the risk of a portfolio, which characterised Basel I, is retained. The Institute of Inter- national Finance (IIF) has objected to capital cushions for ex- pected losses and unexpected losses, on the ground that expected losses tend to be covered by a combination of pricing at the front end of a transaction and general provisioning. It has also objected to the capital market bias in the new proposal, overlooking the "macroeconomic importance of bank lending to small and medium- sized companies which are engines of growth in any economy" [IIF 2001].
The proposal to rely on external assessments of credit risk - by agencies such as Standard and Poors and Moody's Investor Services - for the assignment of loans to risk buckets has attracted the severest criticism generally. Almost the entire developing world has raised objections on the grounds that these rating agencies have shown an imperviousness to developments in underlying fundamentals. Their ratings have an implicit bias towards short-term debt over long term, which is dangerous in a developing country context. They are slow to move and have in fact, started cascades by mis-timed blowing of the whistle. The UN has pointed to trenchant criticism of the pro-cyclicity of the ratings of the major agencies, calling into question their usefulness, and more seriously, their adverse impact by exa- cerbating the instability of bank lending [UNCTAD 2001]. As the US Shadow Financial Regulatory Committee (SFRC) points out "the record of the rating agencies before the recent Asian financial crisis was particularly poor"(p 9, March 2000).
The use of private credit ratings seems to be an implicit acceptance of the inability of regulators to cope with the com- plexity of the new proposal. By shifting the burden of the assessment of the quality of assets to rating agencies, however, regulators risk undermining the quality of credit ratings. Rating agencies would be exposed to incentives to engage in favourable grade enhancements for banks wanting to duck the capital re- quirements and who are willing to 'pay'. As abnormal profits in the rating industry rise, they would attract new entrants and standards would inevitably deteriorate. In this context, it is also necessary to take note of the proposal for domestic rating agen- cies, on grounds of not merely understanding the underlying terrain but also in terms of coverage. In India, out of 9,640 borrowers enjoying fund-based working capital facilities from banks only 300 had been rated by any of the major agencies [UN 2001]. The IIF also underscores a more fundamental concern that the new framework will be expensive for all banks around the world, regardless of their sophistication, as well as for regulators. The cost of implementing the internal rating based approaches may create a disincentive for banks to use internal ratings. Validation requirements also arouse cost concerns.
Notwithstanding the existing acrimony that developing coun- tries bear towards rating agencies, there is a strengthening call for rating the raters, subjecting them to standards of market disclosure and supervisory review. This is strongly resisted by the rating agencies, and they are even inclined to resist their involvement in capital regulation as proposed in Basel II! Moody' s
Economic and Political Weekly June 7, 2003 2275
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
have cited 'unintended consequences' arising from the use of third-party ratings: "In particular, the use of ratings within a regulatory regime could erode rating objectivity as a result of regulatory influence and rating shopping....For example, there is a risk that the national ECAI recognition process could be used to reward and punish rating agencies for their ratings. Ideally, we would argue for a centralised, global process housed within the Committee. Failing that, we would request that the Committee establish clear and fair recognition criteria that emphasise the quality of an ECAI's ratings. Quality measures should include an analysis of the outputs [italics Moody's] of the ratings process - based on default studies and transition matrices - rather than relying only on inputs, such as practices and rating methodolo- gies [italics author's]." (Moody's Investor Services in Comments on the Second Consultative Package of the New Basel Accord, May 2001). Standard and Poor's are sceptical about the eligibility criteria for external credit rating agencies (ECAIs) and the use of ratings, dubbing them as 'sensitive subjects'. They have emphasised that "in order to retain credibility in the capital markets, ECAIs must remain independent of regulatory influence...Such influence could occur in subtle ways and can create pressure on an ECAI to change operating procedures that are fundamental to its business, thereby affecting the ratings assigned....Given the divergence in rating scales and what they denote, even incremental disclosure of both qualitative and quantitative information by ECAIs will make comparisons dif- ficult and the comparability of disclosures across banks prob- lematic [italics author's]" (Standard and Poor's in Response to the New Basel Capital Accord, May 2001).
The consultative document released in 2001 retains the defi- nition of items counting as capital. There are, however, a number of alternative options for numerical standards, increased risk- sensitivity for risk weights, explicit recognition of operational and interest rate risk in banking books and new approaches to the treatment of asset securitisation. The links between pillars have been strengthened. A fuller version of the internal ratings based approach is provided - foundation and advanced variants - and internal risk measurement is integrated in a relatively fuller manner. Nevertheless, the concerns remain. Moreover, it is impossible not to worry about supervisory capacities as also the potential for regulatory arbitrage in such a diversified approach. Under Pillar III the Basel Committee on Banking Supervision (BCBS) itself notes the concern that "the release of too much information could blur the key signals to the market".
What are the alternatives? Several possibilities are being offered in the recent period which attempt to correct for the felt inade- quacies of the capital based standards, i e, the lack of economic foundations, the dulled sensitivity of capital ratios to risks in the market place and the relatively low weightage given to financial innovation and diversification. One approach emanating out of the Brookings Institute loosely integrates aspects of other alter- natives under what is termed as the 'competition-containment paradigm'[Litan 1997]. It cobbles together an advocacy for prompt corrective action provisions of the FDICIA, allowing for some failures, Real Time Gross Settlement (RTGS) in clearing and payments and shortening of settlement periods, backing of assets with subordinated debt, i e, unsecured uninsured debt that is subordinate to the interests of depositors, and external super- vision. It does not regard as necessary harmonisation in disclosure and accounting practices, legal infrastructure and umbrella supervision, as advocated by the Group of Thirty [GoT 1997].
A fuller and more integrated alternative centres around the use of subordinated debt for making portfolios more risk sensitive is proposed by the US SFRC. It consists of five independent but
mutually reinforcing proposals: (i) bank capital should be measured on the basis of market valuation rather than book values of assets and liabilities - bank capital should be the difference between the market values of assets and senior (insured) liabilities which essentially becomes an argument for subordinated debt; (ii) the risk-weighted capital to assets ratio should be replaced by a simple leverage ratio which has the advantage of simplicity and reduces the incentive to shift assets among risk categories to circumvent the capital ratios (such a ratio is applied in the US even today; however, the office of the Comptroller of the Currency reports a step-up in the leverage ratio after 1991, coinciding with the adoption of the Basel I capital standards); (iii) the unweighted minimum capital ratio should be raised to 10 per cent with subordinated debt qualifying as eligible capital; (iv) subordinated debt should count as capital, the advantages being that it dis- courages inappropriate risk taking; interest yields and the ease or difficulties of issuing subordinated debt provide early signals of market perception of risk, and it promotes disclosure (no ceiling is proposed provided (a) such debt is not convertible; (b) cannot be collateralised; (c) has a minimum residual maturity; (d) is sold at arm's length and in large denominations to indicate its unsecured and subordinated nature; and (e) is issued with a covenant that debt servicing can be withheld if the capital ratio falls below the minimum; (v) prompt corrective action through structured early intervention and resolution.
The questions that arise are who would buy the relatively un- attractive subordinated debt? What would happen to issuing costs? If issuance is regulated what happens to the price signals? Can subordinated debt be arbitraged, including off the balance sheet?
Another class of alternatives that is currently under scrutiny draws from the recent tradition of 'incentive-compatible regulation'[Greenspan 1996]. Here, the focus is on the internal risk measurement and management processes in which financial institutions invest heavily to develop and adapt. The objective is to induce banks and financial institutions to reveal their superior information about the riskiness of their portfolios. In the process, regulation is expected to become more risk-focused and driven by market incentives for quantifying, pricing and managing risk. It also enables regulators to ride the pace of sophistication of risk management technologies.
Under one variant, i e, the standardised approach or the full models approach, the Basel approach to credit risk is extended to market risk - assets are categorised and capital charges corresponding to the riskiness of each category are imposed. Banks are required to be able to estimate the probability density function for losses stemming from each of the risk categories. The regulator would then set the capital requirement so that a solvency standard is met. The difficulty is in getting the bank to reveal its true probability density function to make the back testing validation procedure effective. Feasibility constraints, especially with regard to trading books, and the intrusiveness involved are other criticisms levied against the standardised approach. In all fairness, the Basel II proposals are reported to have given this approach due consideration before opting to incorporate the internal models approach [BIS 2000].
Under the internal models approach, the capital requirement for a portfolio is calculated using the internal risk management model of the bank. The internal model is used to calculate the value at risk (VaR) - a measure of potential losses over a time period that would only be exceeded with a given probability (a 1 per cent VaR of $ 1 million implies that losses would exceed $ 1 million 1 per cent of the time). The VaR can be calculated parametrically under the assumption that future returns follow a particular distribution (typically normal) or non-parametrically
Economic and Political Weekly June 7, 2003 2277
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
by generating a simulated time series of profits and losses that would have occurred if the portfolio is held over a particular time period. The capital requirement is set equal to a multiple of the VaR. The Basel Accord specifies a probability of 1 per cent, a period of 10 days and a multiple of 3. What gives, then? The key point is to ensure that the internal model used to calculate the VaR is accurate. Banks could potentially develop models that produces low capital charges. At this point in time, a full-fledged bankwide internal models approach could require a substantial amount of time to develop.
The pre-commitment approach is based on the established economic principle of menu of contracts. Banks are required to choose a level of capital to back their trading books for market risk. If the cumulative trading losses over some period exceed the commitment of capital, penalties are imposed. The prospect of future penalties would induce the banks to commit an amount of capital that reflects a more accurate internal perception of risk. Regulation is 'hands-off' since the regulator is not required to estimate the level of risk. The task before the regulator, however, is to choose an appropriate schedule of penalties. And therein lies the major difficulty of the pre-commitment approach. A penalty proportional to the amount by which losses exceed the capital commitment would operate in a one-size-fits-all manner. Penalties would have to be bank-specific to be effective and, if so, they cannot be put on the menu. It is also argued that there could be principal (shareholder)/agent (bank management) prob- lems arising out of the soft link between risk and capital which could work against the pre-commitment approach viz, Barings [Daripa and Varetto 1998]. Furthermore, responding to ex post penalties assumes that banks are forward looking and take the potential penalties into account while making capital allocations [Parkinson 1998]. The debate rages on and in the interregnum, considerable work is going into integrating the internal models approach and the pre-commitment approach by modifying trouble- some aspects of both. The fear lurks that when the dust settles and light breaks through, will the regulators be sufficiently empowered in terms of skills, or will they be destined for back- testing and hindsight!
V Ongoing Search for International Financial
Architecture Moving from the plumbing to the architecture the debate
encounters highly unsettled conditions. When Robert Rubin, then Secretary of the US Treasury, spoke of the need to strengthen the 'architecture of the international financial system' at the Brookings Institution in April 1998 he provided a new name to a quest which should have begun nearly 20 years ago. Coinciding with the onset of the age of globalisation - the integration of financial markets across borders on the back of the dominance of private capital flows - financial crises have occurred with disturbing intensity and frequency across the world. The debt crisis of the early 1980s cost Latin America 'a lost decade'. Over the last 20 years, more than 125 countries including the US have encountered the consequences of financial instability - banking crises, currency crises, debt crises and even a combination of them all - which have included insolvency of entire banking systems, resolution costs going up to 40 per cent of GDP, plummeting equity and currencies, evaporation of wealth, un- employment, poverty and sharp setbacks to living standards.
It is the severity and spectacular dimensions of the financial crises of the 1990s that has provided the greatest urgency for global collaboration for ensuring financial stability, a term which
has come to encompass both crisis prevention and crisis man- agement. In the 1992-93 crisis, when countries of the European Monetary System lost between US $ 150-200 billion in an unsuccessful defence of currency parities; the Mexican crisis of 1994-95 which caused that economy to contract by as much as 6 per cent in its worst recession in six decades; the Asian crisis of 1997-98 which had global ramifications with the worst affected countries continuing to remain depressed in all aspects of social and economic activity; the decade of near-zero activity in the Japanese economy; the Russian default of 1998 and the collapse of the Long Term Capital Management reverberating in Wall Street; the Brazilian currency crisis of 1999 with contagion spreading over Latin America; and more recently, the Argentine crisis of 2001-02 have all brought home the lesson that financial stability is a global responsibility. Thus, when former US presi- dent Clinton characterised the Asian crisis as the greatest finan- cial challenge facing the world in the last half century in his speech before the Council on Foreign Relations in September 1998, he was underscoring the point that the US is not immune to financial crises abroad. Ensuring financial stability worldwide not only raises global prosperity; it also safeguards domestic employment, output, social welfare and national security.
The costs of the crises are not the only reason for global collective action in the form of a public responsibility for financial stability. The globalisation of finance has opened up domestic financial systems faster than national economies. Financial institutions are required to be internationally active, in various degrees, irrespective of the national position. Accordingly, they are required to operate in the international payment and settlement systems where 'Herstatt' risks can have huge consequences. There is also some evidence to suggest that domestic financial markets perform better when regulated, and even laissez-faire protagonists accept this. International regulation could ensure the stability of both institutions and markets. Furthermore, there are no institutional mechanisms that exist today which can be as- signed dejure or even de facto jurisdiction to safeguard financial stability. Ad hoc assignment cobbled by stretching responsibility for 'monetary and exchange rate arrangements' (assigned to the IMF) has brought forth sharp criticisms of either too little or too much. Moreover, global governance issues have strong under- currents of political economy and the concerns for the greater common good can easily be hijacked by self-interest. Where you stand on reforming the global financial architecture depends on where you sit [Armijo 2001]. For international bankers and investors, reform of the architecture means guarantees of repatriability, even market risk, if possible and international rates of taxation. To many members of the US Congress, it means leaner and less wasteful IMF and World Bank. To Japan and West Europe, it means whittling down the hegemony of the US. To the very poor countries, it means debt forgiveness. To the emerging markets, it means the creation of a lender of the last resort with deeper pockets than the IMF and less or no condi- tionality. And to the doctors themselves - the IMF, the FSF, the BIS, etc - it has come to mean consensual implementation of best practices and standards with an emphasis on information. Intellectually, broad divisions can be superimposed a la Armijo: (i) laissez-faire liberalisers (Milton Friedman, the Meltzer Re- port, the American Enterprise Institute, the Cato Institute, the Institute for International Finance): markets are autonomous; regulation does more harm than good; no safety nets; no capital controls; either gold standard or freely floating exchange rates; close down IMF? (ii) Transparency protagonists (OECD, Financial Stability Forum, G-22, G-20, US Council on Foreign Relations, IMF):
2278 Economic and Political Weekly June 7, 2003
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
improve regulation, transparency and reporting; limited IMF lending into arrears; loan agreements that ensure creditor par- ticipation in case of borrower default (no enforcement mecha- nism proposed so far, however); tax on short-term capital flows. (iii) Financial stabilisers (Institute for International Economics, Joseph Stiglitz, World Bank, James Tobin, George Soros, UNCTAD, countries outside the OECD, Japan): global financial regulation with equity; pre-emptive capital controls; managed floats; better allocation of IMF credits and lender-of-last-resort function; limits on short-term capital flows; global bankruptcy court? (iv) Antiglobalisers (intellectuals, politicians, organised labour in the US, farmers in Western Europe): anti-capitalism; anti- multinational corporations; anti-free trade; anti-Bretton Woods institutions.
It is worthwhile to draw from Kenen (2001) and review the evolution of the recent quest for the architecture to see the inter- play of these actors and what it has achieved. The origins of the exercise can be traced to the Mexican crisis of 1994-95. Failing in its bid to obtain congressional support for a bail-out for Mexico, the US Treasury moved the IMF to organise the biggest financial assistance in its history until that date (US $ 18 billion), supported by the US Exchange Stabilisation Fund (US $ 20 billion). Acerbic criticism of the Mexican bail-out - seen as increasing moral hazard, unfair since it bailed out only the holders of tesobonos, and weakening the IMF in terms of support to other crisis affected countries - led up to the Halifax communique of the G-7 which recommended an early warning system for crisis prevention based on strengthened IMF surveillance and enhanced information disclosure, emergency financing facility to be set up by the IMF to provide faster access to financing with a doubling of the Fund's resources underits General Agreement to Borrow (GAB), strength- ening of international cooperation in regulation and supervision, review of legal and other issues posed by debt crises. Also in 1995, a proposal was made for the establishment of an inter- national bankruptcy regime to meet the problems of sovereign debtors [Sachs 1995].
An Emergency Financing Mechanism was soon established by the IMF and the Special Data Dissemination Standard as well as the less demanding General Data Dissemination Standard were developed. By the end of 1996, the IMF's GAB was supple- mented by the New Arrangements to Borrow (NAB) to deal with exceptional situations that pose a threat to the stability of the international monetary system. Although progress on regulatory and supervisory issues was slower, the development of an in- ternational banking standard was proposed [Goldstein 1997] and this culminated in the Core Principles for Effective Banking Supervision under the Basel Committee on Banking Supervision.
The Rey Report (G 10, 1996) rejected the Sachs proposal for an international bankruptcy regime, citing legal and practical constraints, although it found praiseworthy the work of the Paris Club, which restructures sovereign debt to official creditors, and the London Club, which restructures sovereign debt to commer- cial banks. The Report was not generally in favour of large scale financial support to a sovereign debtor and expressed strong reservations about interrupting servicing of private sector debt.
The Asian crisis brought forth massive official financing orchestrated by the IMF - under criticism of delayed diagnosis and delayed disbursals - and some changes in the IMF itself. The Supplemental Reserve Facility was set up in December 1997 to support countries facing sudden and disruptive loss of con- fidence reflected in pressure on the capital account and the reserves. Although this moved the IMF closer towards Bagehot's classic lender of the last resort since it enabled large drawals
for relatively short periods at penalty rates of interest, the con- ditionality attached to such drawings has been roundly criticised especially when economies which were soundly managed were subjected to creditor panic, and conditionality aggravated the depressions that followed. The IMF's lending policies have been viewed as an attack on Asian capitalism (food subsidies, tariffs, domestic monopolies, directed lending, corporate governance). Indeed, some performance criteria, notably the fiscal targets were relaxed but structural reforms were defended and persevered with. In 1999, following the Russian crisis, the IMF undertook to provide Contingent Credit Lines. When it was opened, how- ever, "no one was waiting outside to apply. The preconditions were daunting, and activation was not automatic" [Kenen 2001].
Adverse reactions to the IMF's response to the crises of the 1990s are also captured in the rejection of the proposal in 1999 to amend its Articles of Agreement so as to give it jurisdiction over the capital account of the balance of payments. The Report of the Independent Task Force sponsored by the US Council on Foreign Relations on 'Safeguarding Prosperity in a Global Financial System: The Future International Financial Architec- ture' recommended that the IMF should lend on more favourable terms to those countries that take steps to reduce crisis vulner- ability and its assessments should be published. This has hap- pened in a limited way in the context of Article IV consultations and Public information Notices. The report also recommended that the IMF should focus on monetary, fiscal, exchange rate and financial sector policies, not on structural reforms and the World Bank should focus on developmental issues and not on crisis management. For crisis situations the IMF should adhere to normal limits of lending and not undertake large-scale rescue packages. Other recommendations are taxes on capital flows, collective action clauses in sovereign bond contracts and aban- donment of pegged exchange rates.
The Meltzer Report issued by the US International Financial Institution Advisory Commission in 2000 carried the criticism even further, echoing these sentiments in offering proposals for reform of the IMF and the World Bank, essentially in terms of narrowing down their focus. The IMF should be a quasi lender of the last resort, providing liquidity when markets close. Pre- conditions (replacing conditionality) to IMF assistance must be 'straight-forward, clear, easily monitored and enforced': ad- equately capitalised financial system, prudent financial policies, prompt availability of information on external debt, fair com- petition between local and foreign banks (this would reduce moral hazard), either fixed or floating exchange rates. If these pre- conditions are met, IMF assistance should be provided imme- diately. The Meltzer Report also underscored the information supply role of the IMF as helping to sustain market discipline.
The fate of the Meltzer Report remains uncertain; however, the IMF itself has moved in 2001 under its new managing director to limit the number and scope of policy conditions in new programmes. The quest for the appropriate international financial architecture continues to travel through inertial dynamics sur- rounding the reform of the Bretton Woods institutions and very little else. The reality that crises are and that they recur with devastating consequences has kept the search alive. The answers remain elusive; the questions accumulate.
International Standards and Codes
Two different directions that the search has taken deserve close questioning. The urgent need to avert financial instability or at least to manage it better has provoked, almost concomitantly, a suboptimal course of action to proxy the appropriate architecture.
Economic and Political Weekly June 7, 2003 2279
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
In early 1998, the US treasury brought together the Group of 22 - a group of 'systemically significant' countries. During that year, the Group issued reports on transparency and accountability, strengthening national financial systems, and managing interna- tional financial crises. In 1999, the G-7 countries established the Financial Stability Forum to enable information exchange and cooperation between national authorities responsible for finan- cial stability, the international financial institutions, sector specific regulators, supervisors and standard-setting bodies and central banking experts. The Forum produced a compendium of inter- national standards and codes to strengthen financial systems in April 2001 and also set up a task force on the implementation of standards. Twelve key areas out of a larger set of more than 60 have been identified for standardisation relating to macro- economic policies and data transparency, institutional and market infrastructure, and financial regulation and supervision. Closely in sync, the IMF and the World Bank introduced the Financial Sector Assessment Programme in 1999 to undertake voluntary assessments of the strengths and weaknesses of na- tional financial systems. Since these assessments also included appraisals of observance of the FSF's standards and codes, Reports on Observance of Standards and Codes (ROSCs) are generated and published with the members' consent (more than 70 ROSCs are available at the web sites of these institutions). The quest for the international architecture began moving sub- global. A new strategy of putting together the pieces on the basis of national efforts towards implementing industry-led best prac- tices has begun to emerge as a second-best approach to crisis prevention. The international financial institutions are currently in a rearguard role, initially benign and merely reporting obser- vance of compliance, but their past and their intrinsic character - international bureaucracy - weighs heavily on their footfalls.
It needs to be recognised at the very outset that the standards and codes approach is primarily governed by self-interest. In fact, this is recognised by the FSF's compendium as the key motivation for implementing standards. The approach is principally directed at countries receiving capital flows and since these countries are the ones most vulnerable to financial crises as the past has shown, they should have a strong interest in implementing standards. Indeed, underneath the fine print is the dictum that if emerging market economies seek to participate in international capital markets they must accept some standardisation [Kenen 2001]. Clearly there is self-interest on the other side too, or rather self- preservation against southern flus. As mentioned earlier, once self-interest enters the picture, the public good content of finan- cial stability diminishes and so does the case for its assignment to public policy.
Secondly, most of the relevant standards aim at defining best practices rather than minimally acceptable practices. The drafters defined best practices by contemplating the sophisticated finan- cial systems of the industrial countries. Moreover, there are significant differences among the industrial country regimes. Even the most positively disposed developing country needs to worry about goodness of fit; illustratively, while the US bank- ruptcy code may qualify as a world's best practice, implementing it in a country without the adequate legal infrastructure is a recipe for disaster. Furthermore, financial development has taken vari- ous different forms in the post-World War II period, shaped by history, and spawning a variety of indigenous institutional mechanisms and processes against which there is no evidence of their being sub-standard. In many countries, the indigenous architecture has served them well in terms of the country-specific needs and priorities. Consequently, it is but natural that some countries view the standards and codes as being disruptive and
dangerous and lip service has to be paid to these concerns by the reformers, viz, 'voluntary adoption' (IMF, FSF), 'balance between international and domestic considerations' [Sheng 2000], 'progress rather than absolute compliance (FSF)'. In this vein, the UNCTAD makes a fundamental point. The standards and codes approach does not entail any change in the policies and practices of the countries sending capital flows and therefore, leave the emerging economies vulnerable to supply-driven fluc- tuations in international capital flows [UNCTAD TDR 2001].
Thirdly, the FSF suggests that markets are likely to provide the strongest incentive for the implementation of standards. In fact, the emphasis in almost all the standards identified seems to be on transparency and disclosure so that it 'would help lenders, investors and intermediaries to use that information in their risk analyses'. Better informed market participants may be able to minimise being taken by surprise but this in no way minimises risk taking. Arguably, purveyors of private capital knew about crony capitalism and implicit guarantees in south-east Asia for long enough and yet they pumped in money into 'miracle' terrain for several years before the denouement. After all, the Eurpoean crosses were virtually dead after the advent of the ERM. If there was money to be made, it was in Asia. Most market participants do not currently take account of a country's adherence to stan- dards and codes while making risk assessments. If they ever will, they would worry about end results, not a country's progress in implementing standards and codes. In time, this may produce dangerous actions such as official incentives to encourage implementation of standards and codes and this is what the UNCTAD worries about - that they could become features of IMF conditionality and surveillance.
The link between standards and codes and financial stability is neither theoretically established nor backed by sufficient empirical evidence [Reddy 2001]. Financial markets are noto- riously idiosyncratic and have shown no marked superiority in processing information that will flow out of the implementation of standards and codes. The FSF has itself noted that certain commonly employed risk management techniques.. .can have the effect of adding to the volatility of both prices and flows in the international capital markets (FSF, 2000). Brazilian bonds are used as instruments to hedge positions in Russian debt [UNCTAD 2001 ]. No attempt is being made to ascertain the degree of overlap that the standards and codes could create potentially in terms of market responses or to assess them all as part of an integrated system, opening up the possibility of multiple equilibria, as mentioned earlier. Moreover, the standards must themselves have a time dimension since it is innovation which marks the evolution of financial intermediation. A decade after Basel I, countries with stretched resources are bracing up for Basel II and the distressing point is that like in the previous case, they will have no choice. It is often remarked that finance is the derivative of the real economy. Excoriating financial instability will eventually depend on ironing out cycles in real activity and standards and codes can hardly be expected to accomplish that! Thus the second-best approach starts out with large potential costs of implementation and assessment, but its effectiveness remains to be proved.
International Bankruptcy Court
The most stimulating advancement in the search for the appro- priate international financial architecture, and from the point of view of this study the most promising, took root at a dinner at the National Economists' Club, American Enterprise Institute on November 26, 2001 when first deputy managing director of the IMF Anne Krueger proposed a formal Sovereign Debt Restructuring
2280 Economic and Political Weekly June 7, 2003
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
Mechanism (SDRM) based on the model of a domestic bank- ruptcy court. The proposal could not have had the blessings of the US treasury as it ran counter to the US position on extending its own bankruptcy procedures to insolvents beyond its borders. It also flung down the gauntlet at notable creditor associations such as the Institute for International Finance as well as leading think-tanks making intellectual opinion in the US such as the Brookings Institution and the Institute of International Economics. Quite clearly at that dinner, Anne Krueger stood alone!
To be sure, the proposal itself is not new. Indeed the origins of the call for international cooperation for resolving debt prob- lems of developing countries along these lines can be traced back to the report of the Intergovernmental Group of Experts on the external indebtedness of developing countries convened by the UNCTAD in 1977. The UNCTAD's trade and development board resolution of September 1980 proposed the creation of a multinational forum agreed upon by the debtors and creditors in which the chairman would conduct the debt operation in a fair and impartial manner in accordance with the agreed objec- tives - expeditious and timely international action to restore the development prospects and the capacity to service short and long term debt by the debtor country, and to protect the interests of the debtors and creditors equitably. The Annex to the UNCTAD's Trade and Development Report of 1986 contained a proposal drawing on the analysis of a New York law firm for international debt reform based on Chapter 11 of the US Bankruptcy Reform Act of 1978, and this has been endorsed by the Trade and Development Reports of 1988 and 2001. At this point, a digres- sion on the US bankruptcy Code may not be out of place. The basic underlying premise is that the value of an entity as a going concern exceeds its value if it is liquidated. Accordingly an orderly debt workout takes place in three stages. First, on filing a bankruptcy petition by the debtor, there is an automatic standstill on debt servicing and creditors are not allowed to approach courts of law to enforce repayment. All claims on the debtor are fixed and claims for future interest on pre-petition indebtedness cease to accrue as of the petition date. This gives the debtor breathing space to formulate a reorganisation plan. Secondly, the Code grants a senior status to debt contracted after the filing of the petition. This allows the debtor access to working capital to continue operations and reorganisation plans which is indepen- dent of the will of existing creditors. The third stage is the reorganisation of assets and liabilities of the debtor with an emphasis on speedy resolution - no requirement of unanimous support from creditors, and debtor can obtain court approval of reorganisation plan. Chapter 11 of the Code applies to private debt while Chapter 9 applies to public debt under which the Orange County workout is the most successful example. Accord- ing to the UNCTAD proposal, the international bankruptcy court could apply the US Bankruptcy Code at the international level or at least apply the key principles, i e, standstill, debtor in- possession financing, and debt restructuring.
What are the problems in elevating the US Code to the in- ternational level? Anne Krueger cited the example of Elliott Associates, a holdout or vulture creditor which had bought commercial loans guaranteed by Peru in 1997. When Peru tried to restructure its debt, Elliott obtained a court order attaching Peruvian assets used for commercial activity in the US. Indeed, she took this case as the basis for recommending a formal debt restructuring mechanism over voluntary and market-oriented solutions. Rogue creditors prefer disorderly restructuring and hope to profit through litigation unless restrained from approaching courts in the US or adopting other legal avenues. In 1982, Costa Rica suspended debt servicing by three state-owned banks. The
case opened by the creditors was initially dismissed on the grounds that the Costa Rican action was consistent with the law and policy of the US, with reference to Chapter 11 of the Bankruptcy Code. In 1984, the same court reversed its decision when it was told by the US government that it had incorrectly interpreted US policy (Financial Times, May 24, 1984). These and other similar rulings have given rise to the opinion that the US bankruptcy Code cannot easily be raised to the level of an international bankruptcy practice for fear that foreign govern- ments may act unilaterally and arbitrarily in matters relating to US banks and indirectly affect the stability of the US banking system [UNCTAD 1988].
In 1997, the UN Conference on International Trade Law (UNCITRAL) adopted a model law on cross-border insolvency with the objectives of cooperation between courts and competent authorities across borders, legal certainty and protection for trade, investment and employment, fair and efficient administration of cross-border insolvencies that protects the interests of creditors and debtors, protection and maximisation of the debtor's assets. The model law has been benchmarked as an international standard by the FSF. It is also regarded as compatible with the London Club approach towards the restructuring of sovereign debt owed to commercial banks.
Under Anne Krueger's proposal, a country facing payment difficulties could come to the IMF and request a temporary (some months) standstill on repayments. During this interregnum, it would reschedule or restructure its debt, providing assurance of absence of capital flight to creditors through the imposition of exchange controls. While this would create a mandatory process for restructuring the outcome would be left to negotiations between creditors and debtors. The mechanism would prevent creditors from disrupting debt restructuring by seeking repay- ments through national courts. This will hold off vulture creditors. The mechanism would provide creditors with some guarantee of repayments - through appropriate economic policies by the debtor, equal treatment of creditors, and measures for rebuild- ing of confidence. New lenders would be given seniority or preferred creditor status. The mechanism would also enforce the 'majority' ruling on the 'minority' creditors. The mechanism would have to have the force of law in all countries. The IMF would be involved in the operation of the mechanism from the point of view of endorsing applications for standstill and periods for which standstills are applicable, ensuring appropriate economic policies during the standstill, etc, but adjudication of disputes would lie beyond the IMF's competencies. Limited IMF financing is envisaged after the restructuring to rebuild reserves and maintain essential imports but not to pay off creditors. The standstill would apply equally to sovereign debt owed to residents and to external debt owed to non-sovereign creditors.
The principal merit of the proposal, i e, that it would provide a fairer and more efficient process of debt workouts which, in effect, provides the missing piece in the efforts to strengthen the international architecture for dealing with financial crises. This is disputed mainly by the creditors and their intellectual advocates in the Institute of International Economics. The view here is that the proposal does not protect the interests of the creditors and if implemented, would lead to a receding of private capital flows from the developing world. It encourages imprudent borrowing and debt management and may even encourage speedier debt restructuring. It is also argued that the cost of borrowing by developing countries would rise as creditors factor in the delays and lock-ins involved in orderly restructuring processes and the loss of value of assets trading in secondary markets.
Economic and Political Weekly June 7, 2003 2281
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
The objection of the UNCTAD lies primarily in giving the IMF the role that is envisaged in the mechanism. Involvement of the IMF would formalise the international bankruptcy procedure when what is sought is the benefit of the mechanism without a formal judicial structure. Moreover, the IMF's involvement would take away the voluntary nature of the decisions to approach for restructuring and to bargain between the negotiating parties. It has also been argued that the executive board of the IMF is n9t a neutral body and therefore, not competent to act as inde- pendent arbiter. The IMF is a creditor and by becoming the arbiter there would be a conflict of interest [Raffer 1990; Sachs 1998; Eichengreen and Portes 1995]. Besides these arguments, there are technical questions raised. The IMF has set out a fuller articulations of its proposal following the Spring meeting of 2002. The Brookings Institution is planning to focus the forthcoming volume of the Brookings Papers on Economic Activity on the international bankruptcy court. The issue was discussed by the G-20 in November, 2001 in Ottawa and the UNCTAD proposes to discuss an alternative proposal in its forthcoming meeting in Monterrey. This builds on UN precedence on the issue that is closer to the London club approach including combining private and sovereign debt workout procedures, an independent mediator and a non-statutory basis for the mechanism albeit in coordination with the IMF. And suddenly, in the quest for the international architecture, the horizon is lighting up!
V Concluding Observations
The assignment of the responsibility for ensuring financial stability, or at least creating the enabling conditions, to public policy seems to draw largely from tradition rather than logic or economic principles. Indeed ensuring financial stability has come to be the reason why certain public institutions and public policies exist at all. In this assignment, however, the public good char- acteristic of financial stability on the consumption side seems to be stressed, i e, that there are large scale externalities involved both in its presence and absence. What needs to be stressed for the cause of appropriate assignment is the production characteris- tic, i e, that the supply of financial stability hinges upon collective action - in this case, by national public authorities. Aggregation of national responsibilities cannot produce the optimal outcome since it cannot take away vestiges of self-interest.
Financial stability involves both markets and institutions. If institutions need to be intervened in the cause of financial stability, so do markets, since the stability of financial institutions depends upon their ability to adapt to movements in markets. There is some, if not comprehensive, evidence suggesting that financial markets, which are more prone to seizures than others, actually perform better when appropriately intervened. The quality of public intervention in markets may vary widely from provision of incentives to information supply to allocation and price discovery inducements, but public policy wielding entities should not hesitate from providing to the markets their views on prices and market fundamentals. It is possible, given the growing might of markets, that these views may be rejected and public authorities may have to tactically withdraw. Nevertheless, the opinion that market valuations are out of alignment with fundamentals should not be left to hindsight alone. Where public authorities are en- gaged in market development and in provision of the institutional infrastructure, intervention goes unquestioned. In some sense, public authorities, irrespective of the level of sophistication of the markets, have to contend with innovation and therefore, deregu- lation. While frauds are unavoidable and can occur in any clime,
they more often than not expose what missed the all-seeing eye. Financial institutions in developing countries engage in inter-
national transactions regardless of the degree of the openness of their host economies; in doing so, they operate in the inter- national payment and settlement system and are therefore subject to the rules set in market places of the developed countries. Accordingly, they are automatically disadvantaged into being rule takers and are treated implicitly as potential purveyors of systemic risk. In this sense, there is no such thing as a best practice and this is true of all identified standards and codes. They are at best outcomes of exercises in comparative statics and relative to the time period in which they apply. They cannot be tailored; history and institutions have a major role in their emergence and application. Standards do not stand alone; they are inexorably wrought into a complex panoply of institutions, processes, his- torical precedents and social structures. There is no evidence linking them to financial stability since even the most developed of financial systems have suffered crises, sometimes repetitively. They have never been tested for system stability analogous to a system of equations or parametric constraints. Yet the reality is that they are applied and often at considerable cost to devel- oping country financial systems, and voluntary adoption is only a figure of speech. The most visible case is that of the Basel capital standards. The alternatives are out there, some of them well-grounded in theory and practice, unlike the a theoretic Basel pillars.
Much has been said on the need to strengthen the international architecture. As regards what is missing, the new proposal for a formal sovereign debt restructuring mechanism has consider- able virtue in filling the 'gaping hole'. In the final analysis, the counterarguments are technical and require constructive delib- eration. Parallels exist in the Paris Club and the London Club; a New York Club, say, would not be too far-fetched. Ultimately, the best approach would be one marked by voluntariness, but in the interregnum, public policy intervention may be warranted to produce the voluntary outcome. Ii
Address for correspondence: [email protected]
[The views and opinions expressed in this paper are those of the author only.]
References American Enterprise Institute for Public Policy Research (2000): 'Reforming
Bank Capital Regulation: A Proposal by the US Shadow Financial Regulatory Committee', Washington DC, March.
Armijo, Leslie Elliott. (2001): 'The Political Geography of World Financial Reform: Who Wants What and Why?' Global Governance, Vol 7, No 4, Oct-Dec.
Crockett, Andrew (2000): 'Progress Towards Greater International Financial Stability', Speech at the End-of-Programme Conference of the GEI Programme on 'Reforming the Architecture of Global Economic Institutions', May 5, London.
-(2001): 'Market Discipline and Financial Stability', speech at the Conference on 'Banks and Systemic Risk' of the Bank of England, May 23, London.
Daripa, Arupratan and Simone Varetto (1998): 'Value at Risk and Precommitment: Approaches to Market Risk Regulation', Federal Reserve Bank of New York Economic Policy Review, October.
Federal Reserve Bank of Boston (2000): Conference on 'Building an Infrastructure for Financial Stability', June, Boston.
Federal Reserve Bank of Kansas City (1997): Symposium on 'Maintaining Financial Stability in a Global Economy', August 28-30, Jackson Hole, Wyoming. (References to Crockett 1997; Litan 1997; and Greenspan 1997 are from this Symposium Volume.)
2282 Economic and Political Weekly June 7, 2003
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
Financial Stability Forum (2001): 'Compendium of Standards: International Standards and Codes to Strengthen Financial Systems', April.
Greenspan, Alan (1997): 'The kole of Capital in Optimal Banking Supervision and Regulation', Federal Reserve Bank of New York Economic Policy Review, October.
Group of Thirty (1997): 'Global Institutions, National Supervision and Systemic Risk', A Study Group Report, Washington DC.
Hooks, Linda M (1994): 'Capital, Asset Risk and Bank Failure' Occasional Paper No 47, Group of Thirty, Washington DC.
Institute for International Economics (1999): Report of an independent task force sponsored by the Council on Foreign Relations on 'Safeguarding Prosperity in a Global Financial System: The Future International Financial Architecture', Washington DC.
Institute of International Finance (2001): Report of the steering committee on Regulatory Capital: Response to the Basel Committee on Banking Supervision Regulatory Capital Reform Proposals, Washington DC, May; Principles for Private Sector Involvement in Crisis Prevention and Resolution, Washington DC, January; Report of the Working Group on Capital Adequacy: Response to the Basel Committee on Banking Supervision Regulatory Capital Reform Proposals, Washington DC, May.
Kenen, Peter B (2001): 'The International Financial Architecture: What's New? What's Missing', Institute for International Economics, Washington DC, November.
Krueger, Anne (2001): 'International Financial Architecture for 2002: A New Approach to Sovereign Debt Restructuring', address at the National Economists' Club, American Enterprise Institute, Washington DC, November 26.
Large, Andrew (1997): 'The Future of Global Financial Regulation', Occasional Paper No 57, Group of Thirty, Washington DC.
Meltzer, Alan H (2001): 'Reforming the International Financial Institutions: A Plan for Financial Stability and Economic Development', Economic
Perspectives, International Information Programmes, US Department of State, February.
Moody's Investor Services (2001): 'Comment on the Second Consultative Package of the New Basel Capital Accord', May.
Olson, Mark, V (2002): 'The Financial Services Industry and Public Policy', Speech at the Centre for the Study of Mergers and Acquisitions, University of Miami School of Law, Miami Beach, Florida, February 7.
Parkinson, Patrick (1998): 'Incentive Compatible Regulation: Views on the Pre-Commitment Approach', Federal Reserve Bank of New York Economic Policy Review, Vol 4, No 3 October.
Reddy, Y V (2001): 'Issues in Implementing Financial Standards and Codes', Public Lecture at the Centre for Banking Studies of the Central Bank of Sri Lanka, Colombo, June 28.
Santos, Joao A C (2000): 'Bank Capital Regulation in Contemporary Banking Theory: A Review of the Literature', BIS Working Paper No 90, September.
Scott, Hal S (1992): 'Supervision of International Banking Post BCCI', Georgia State University Law Review, Vol 8, No 3, June.
Scott, Hal S and Shinsaku Iwahara (1994): 'In Search of a Level Playing Field: The Implementation of the Basel Capital Accord in Japan and the United States', Occasional Papers No 46, Group of Thirty, Washington DC.
Standard and Poor's (2001): 'Standard & Poor's Response to the New Basel Capital Accord' May.
United Nations Conference on Trade and Development (1977): Report of the Intergovernmental Group of Experts on the External Indebtedness of Developing Countries, Geneva, September.
- Trade and Development Reports, 1980, 1986, 1988, 1997, 2001. - (2001): 'The Basel Committee's Proposals for Revised Capital Standards:
Mark 2 and the State of Play', Discussion Paper No 156, Geneva; 'How Risky is Financial Liberalisation in the Developing Countries?' G-24 Discussion Paper No 14, Geneva and New York.
The Bilingual Family The Dravidian Languages Edith Harding-Esch and Philip Riley Bhadriraju Krishnamurti This is the second edition of the best-selling The Dravidian languages are spoken by over book that has provided practical advice to 200 million people in South Asia and in thousands of parents who want their children Diaspora communities around the world, to grow up bilingual. It still gives parents up- and constitute the world's fifth largest to-date information and advice they need to language family. It consists of about 26 make informed decisions about what language languages in total including Tamil, 'policy' to adopt with their children. This new Malayalam, Kannada and Telugu, as well as edition also looks at cases of single-parent k over 20 non-literary languageas. In this book, families with bilingual children, as well as Bhadriraju Krishnamurti, one of the most
schooling and bi-literacy issues. The authors help parents identify eminent Dravidianists of our time, provides a comprehensive study the factors that will influence their decision to bring up their children of the phonological and grammatical structure of the whole as bilinguals. The second part consists of case studies of bilingual Dravidian family from different aspects. He describes its history families, which illustrate a wide range of different 'solutions'. The and writing systems, discusses its structure and typology, and third part is an alphabetical reference guide providing answers to considers its lexicon. Distant and more recent contacts between the most frequently asked questions about bilingualism. Bringing Dravidian and other language groups are also discussed. With its
up bilingual children is a challenge and this book helps parents comprehensive coverage this book will be welcomed by all students meet that challenge. of Dravidian languages and will be of interest to linguists in various
Original Price ? 12.95 PB 0-521-00464-0 Rs. 695.00 branches of the discipline as well as Indologists. 206pp (Special price for limited stock only) Rs. 950.00 HB 0-521-771 1 -0 571 pp
FOUNDATION BOOKS -4764/2A, 23 Ansari Road, Daryaganj, New Delhi 110 002 Tel: 23277332, 23285851 / 52 Fax: 23288534 E-mail: [email protected]
C-22, 'C' Block, Brigade MM, K. R. Road, Jayanagar, Bangalore 560 082 Tel: 6764817, 6762764 Fax: 6761322 E-mail: [email protected] - 21/1 (New No.49) 1st Floor, Model School Road, Thousand Lights, Chennai 600 006 Tel: 28291294, 52146807 Fax: 28291295 E-mail: [email protected]
House No.3-5-874/6/4 (Near Apollo Hospital), Hyderguda, Hyderabad 500 029 Tel: 23244458 Fax: 23244459 E-mail: [email protected] 60, Dr. Sundari Mohan Avenue, First Floor, Kolkata 700 014 Tel: 22845725 / 26 Fax: 22845727 E-mail: [email protected]
Plot No. 80, Service Industries Shirvane, Sector-1, Nerul, Navi Mumbai 400 706 Tel: 27709172, 27713810 Fax: 27709173 E-mail: [email protected]
Economic and Political Weekly June 7, 2003 2283
This content downloaded from 147.143.2.5 on Fri, 13 Feb 2015 16:12:50 PM All use subject to JSTOR Terms and Conditions
- Article Contents
- p. 2271
- p. 2272
- p. 2273
- p. 2274
- p. 2275
- p. 2277
- p. 2278
- p. 2279
- p. 2280
- p. 2281
- p. 2282
- p. 2283
- Issue Table of Contents
- Economic and Political Weekly, Vol. 38, No. 23 (Jun. 7-13, 2003), pp. 2233-2336
- Front Matter [pp. 2233-2306]
- Letter to Editor
- Imperialist Ignorance [p. 2234]
- Editorials
- CAS without Convergence [pp. 2235-2236]
- Wheel Turns Full Circle [p. 2236]
- Distribution Woes [pp. 2236-2237]
- Southern Initiative [p. 2237]
- TARU Leading Edge: Correction [p. 2237]
- Current Statistics [pp. 2238-2239]
- Calcutta Diary [pp. 2240-2241]
- Commentary
- Twelfth Finance Commission and Panchayat Finances [pp. 2242-2243]
- Shifting Political Equations in UP [pp. 2244-2246]
- Localisation as an Alternative to Globalisation? [pp. 2246-2247]
- Developing a History of Science and Technology in South Asia [pp. 2248-2251]
- Alice Stewart, MD (1906-2002): A Tribute [pp. 2252-2254]
- Argentina: 18 Months of Popular Struggle [pp. 2255-2260]
- Perspectives
- Iraq, UN and Changing Bases of World Order [pp. 2261-2266]
- Reviews
- Review: Revisiting Agrarian Bihar [pp. 2267-2268]
- Review: The World's Largest Hypocrisies [pp. 2269-2270]
- Special Articles
- Should Financial Stability Be Assigned to Public Policy? [pp. 2271-2275+2277-2283]
- Stagnation and Revival of Kerala Economy: An Open Economy Perspective [pp. 2286-2294]
- Interrogating the Nation [pp. 2295-2302]
- The Political Economy of Drug Quality: Changing Perceptions and Implications for Indian Pharmaceutical Industry [pp. 2303-2305+2307-2309]
- Discussion
- Subaltern Fantasies [pp. 2310-2311]
- Back Matter [pp. 2312-2336]
Systemic Risk Monitoring and Financial Stability.pdf
NELLIE LIANG
Systemic Risk Monitoring and Financial Stability
This discussion briefly outlines key elements of a systemic risk monitor to help identify risks to financial stability. The monitor distinguishes shocks, which are varied and difficult to predict, from vulnerabilities, which can am- plify shocks and lead to instability. Better data and models of amplification channels, and better communication among different authorities, are needed to be effective.
JEL codes: E58, G01, G23, G28 Keywords: systemic risk, financial stability.
LIKE OTHER CENTRAL BANKS, the Federal Reserve (Fed) has responsibility for financial stability, primarily as the lender of last resort. In addition, the Fed has regulatory and supervisory responsibilities for many large banking in- stitutions. The Dodd–Frank Act, enacted in 2010 in response to the financial crisis, expands those responsibilities.1 It requires that the Fed be the primary supervisor for the largest bank holding companies and nonbank financial institutions designated as systemically important, and adopt a macroprudential approach to supervision and regulation. Such an approach would supplement the focus on the safety and soundness of an individual institution with explicit consideration of threats to the stability of the financial sector as a whole. The Dodd–Frank Act also imposes greater accountabil- ity for financial stability on regulators by creating the Financial Stability Oversight Council (Council). The Chairman of the Board of Governors of the Federal Reserve System is a voting member of the new Council, which has nine other voting members and is chaired by the Secretary of the Treasury. As a member of the Council, the Fed is obligated to identify structural weaknesses and emerging risks in the financial system, and recommend financial regulatory and macroprudential policies to increase its resilience.
The views presented here are my own and do not necessarily reflect those of the Board of Governors or its staff.
NELLIE LIANG is the Director at the Office of Financial Stability Policy and Research, Federal Reserve Board (E-mail: [email protected]).
Received September 19, 2012; and accepted in revised form February 12, 2013.
1. “Dodd-Frank Wall Street Reform and Consumer Protection Act,” HR 4173, 111th Congress, 2nd Session, Passed by the House in December 2009 and by the Senate in May 2010.
Journal of Money, Credit and Banking, Supplement to Vol. 45, No. 1 (August 2013) C© 2013 The Ohio State University No claim to original US government works
130 : MONEY, CREDIT AND BANKING
I briefly outline the key elements of a new systemic risk monitor and accompany- ing policy framework being developed at the Fed to meet its new financial stability responsibilities. This work is ongoing and will require progress on new models, data, and organizational structures to implement effectively. Importantly, it will require a culture that promotes better communication across many groups, including su- pervisors, microeconomists, and macroeconomists within the Fed, as well as across regulatory agencies. It builds on and complements traditional work related to bank holding company supervision and regulation, financial regulatory reforms, and mon- etary policy.
1. ASSESSMENT OF SYSTEMIC RISK
The recent financial crisis demonstrates vividly that there are many channels through which seemingly small losses can become systemic and threaten finan- cial stability. Systemic risk arises when shocks are amplified and inflict signifi- cant damage on the broader financial system and broader economy. The goal of a financial stability authority is to identify shocks and vulnerabilities—the poten- tial amplification channels—and to preemptively address these vulnerabilities in order to reduce the frequency and severity of crises in the future. The financial system performs effectively, and is stable, when it is sufficiently resilient to ab- sorb shocks and perform its function of allocating capital and credit, and facilitating payments.
Many researchers have looked at how losses in the relatively small subprime mortgage market could have triggered such a severe financial crisis and the Great Recession. Recent papers highlight multiple potential vulnerabilities, including weak financial firms, substantial interlinkages across these firms, complex financial prod- ucts, and excessive leverage and maturity mismatch fueled by the shadow banking system (see, e.g., Brunnermeier 2009, Adrian and Shin 2010, Acharya, Schnabl, and Suarez 2013, Covitz, Liang, and Suarez 2013, Gorton and Metrick 2012b). These vulnerabilities amplified the shock of subprime losses from a fall in house prices through direct counterparty losses, and through indirect losses from fire sales, contagion, and deleveraging. A simple stylized narrative of the story illus- trates some of the various amplification mechanisms.2 Underwriting standards in the lead-up to the crisis were lax as the unregulated shadow banking system bid ever more aggressively for securitized subprime mortgage assets. As house prices fell, losses in the value of subprime mortgages were amplified because they were financed in short-term funding markets, like the asset-backed commercial paper market, which then were run on by investors. Bank sponsors came under pres- sure to support some of these conduits, shrinking their capacity to provide other credit, and the decline in the value of mortgages and other assets from house price
2. See Gorton and Metrick (2012a) for a more detailed narrative based on recent research papers, and Bernanke (2010).
NELLIE LIANG : 131
declines led to deleveraging. In addition, the securities also were used as collateral by broker-dealers and other market participants in repo markets. Falling collateral values then sparked a run as investors ran on broker-dealers financed in the repo mar- ket. The failure of the government-sponsored enterprises (Fannie Mae and Freddie Mac), the failure of Lehman Brothers, and the “break the buck” event at the Primary Reserve Fund, led to further pullbacks in funding, actual or potential fire sales of sub- prime MBS that drove prices down further, and intensified deleveraging by financial intermediaries.
An effective systemic risk monitoring effort is based on identifying a range of possible shocks and assessing vulnerabilities—the channels that can amplify shocks. Vulnerabilities can be structural, present in all conditions, such as interconnections and common exposures, or they can be cyclical, which vary with financial and economic conditions, such as increasing leverage and maturity transformation. The main purpose of the monitoring effort is to evaluate how possible shocks, if amplified, could disrupt financial intermediation and real economic activity. The success of this effort does not rest on unique foresight; instead, it explicitly builds on the view that it would be impossible to predict how any future crisis could fully play out.
While the approach is simple, the monitoring efforts are analytically intensive. They require more data and advanced techniques to develop better measures of systemic risks, based on interconnections among firms and markets that likely vary over time. They require a better understanding of linkages between the financial sector and the macroeconomy, and more focus on considering adverse outcomes that may not be the most likely, but certainly plausible.
Starting with the large and complex financial firms, a monitor should include measures of the expected financial conditions and tail risks of these financial firms. Researchers have developed a number of systemic risk measures for firms based on their stock prices and CDS premiums. These measures are based on the insight that firms with high covariance with the broader markets in bad states of the world will be more systemically risky. For example, one measure, conditional value at risk (Co- VaR), estimates the increase in the value at risk of the financial system conditional on a firm’s distress.3 Two other measures look at a firm’s returns conditional on weak returns for the broader markets.4 Some have criticized these market-based measures on the grounds that investors might not really know much about the interconnec- tions of the firms and the sensitivity to strains in broader markets. Nonetheless, these types of measures capture how markets perceive how specific firms would perform in bad states of the world. Moreover, even without full information, investors’ per- ceptions can raise funding and capital costs for these firms, and thus have important implications for their viability.
3. Specifically, CoVaR measures the increase in the value at risk of the financial system conditional on a firm becoming distressed, for example, a 5th percentile bad event (Adrian and Brunnermeier 2011).
4. The systemic expected shortfall (SES) reflects losses borne by equity holders conditional on a large-tail return for the broader equity markets (Acharya et al. 2010). The distressed insurance premium (DIP) measures a hypothetical insurance premium against catastrophic losses in a portfolio of financial institutions. The systemic importance of an institution is measured by its marginal contribution to the aggregate insurance premium (Huang, Zhou, and Zhu 2009).
132 : MONEY, CREDIT AND BANKING
A promising measure of the systemic risk of the largest institutions is based on regular supervisory stress tests, such as the type originated in the Supervisory Capital Assessment Program in 2009. These measures will be based on firm-specific detailed asset information collected by supervisors and will be a forward-looking assessment of firms’ exposures to a scenario with adverse macroeconomic or financial risks. The firms’ assessments will be evaluated jointly and so reflect losses when multiple firms are stressed at the same time. This rigorous summary measure of the vulnerability of firms to stressed economic and financial conditions is a complement to market- based indicators because supervisors have better information than outside market participants.
Network analysis built from data on direct interconnections between firms is po- tentially quite valuable to evaluate the potential for systemic risk. It would allow regulators to estimate how the distress of a given firm would directly affect the other firms in the network, and also to simulate follow-on effects, which can be very sig- nificant. Because good data on connections among firms are quite limited, however, only a few countries’ banking systems, such as those in Austria and Canada, have been mapped. In addition, some researchers have usefully applied network analysis to cross-country banking data, and document a significant increase in interconnections since the late 1980s.5
Imbalances in financial markets are another important amplification mechanism. Thus, a systemic risk monitor reports need to encompass more than institutions. Financial imbalances could be reflected in narrow risk premiums, fueled by high leverage and short-term funding, and the proliferation of new products and innova- tions that operate in the shadows beyond regulatory boundaries.
On this front, a monitor should include risk premiums for major asset classes. Some aggregate financial conditions indexes, based on principal components of various risk spreads, may also capture the extent to which low discount rates are common to many asset classes. Options on asset prices can also reveal a skew in investors’ views about downside risks and a rapid unwind that could cause markets to become dysfunctional. High or increasing leverage and maturity mismatch in the financial and nonfinancial sectors of the economy can also pose risks. A new survey of major dealers was initiated by the Fed about a year ago to gain insight into the availability and terms of credit for securities financing and over-the-counter derivatives.6 At the same time, the Fed is better utilizing its existing data collection efforts, such as the Flow of Funds accounts, to measure the reliance of the debt of the nonfinancial sector on unstable short-term funding sources.
In addition, as the central bank, the Fed is especially interested in evaluating how possible risks could disrupt the availability and terms of credit, and thus have adverse effects on the real economy. Structural macroeconomic models of the U.S.
5. See, for example, Elsinger, Lehar, and Summer (2006), Gauthier, Lehar, and Souissi (2010), and Garratt, Mahadeva, and Svirydzenka (2011).
6. Senior Credit Officer Opinion Survey on Dealer Financing Terms, http://www.federalreserve. gov/econresdata/releases/scoos.htm.
NELLIE LIANG : 133
economy and foreign economies are used to better understand the interaction of the financial system and real activity, to simulate plausible adverse scenarios, and to help evaluate potential policy tools. Much work is ongoing at the Fed and in the academic community to enrich the financial sectors in these models.
2. MACROPRUDENTIAL POLICIES
In considering macroprudential policies, one has to recognize that systemic risk is inherently a negative externality, arising from fire sales or coordination failures, for example. Macroprudential policies are justified because firms lack private incentives to mitigate this externality. As research indicates, the recent crisis had multiple causes, and reflected interactions between built-up cyclical imbalances and structural risks. Macroprudential policies can be designed to address both types of systemic risks. Policies can be aimed at vulnerabilities that build with extended periods of favorable economic and financial conditions. Such policies are designed to “lean against the wind” to prevent, for example, credit-fueled asset bubbles that could unwind in destabilizing ways, but perhaps more importantly to bolster the resilience of financial institutions when bubbles inevitably burst. Policies also can be aimed at reducing structural vulnerabilities, such as those arising from regulatory gaps or weak business models, to increase the resilience of the financial system. Key examples of structural risks are complex linkages across firms, the perception of too-big-to-fail institutions, and the unstable business model of money market mutual funds.
To mitigate threats posed by the too-big-to-fail problem, the Fed is developing a package of enhanced prudential standards, including higher capital and leverage requirements, liquidity standards, stress testing, and recovery and resolution plans for large complex financial institutions. It is also working with the Council to designate systemically important nonbank financial institutions and financial market utilities that would become subject to enhanced prudential oversight to help reduce the sys- temic risk consequences were such firms to fail. In addition, the Fed proposed rules to set margin requirements for over-the-counter derivatives, with more stringent re- quirements for contracts between certain parties whose default would have greater risk of triggering a cascade.
In other efforts to reduce systemic risk, the Council has recommended that the SEC continue to pursue reform alternatives, such as mandatory floating net asset value, capital buffers to absorb fund losses, or deterrents to redemptions. In addition, the Fed is working to increase the stability of the tri-party repo market by limiting intraday credit exposures and strengthening collateral management practices.
At times when build-ups appear on track to becoming excessive, policymakers may want to take pre-emptive actions to mitigate costs of a possible disorderly unwind. Policymakers can make public statements about possible excesses or can increase supervisory attention and increase guidance if financial imbalances are emerging. There are other targeted macroprudential tools designed to lean against the wind,
134 : MONEY, CREDIT AND BANKING
one being the countercyclical capital buffer in Basel III should aggregate credit grow “too quickly.” Others include higher loan-to-value or debt-to-income ratios when borrowers begin to get too levered. Evidence to date is limited on the efficacy of such tools, and there are many implementation hurdles, such as when to act, with what force, and how to coordinate globally. While advocates recognize the limited evidence of benefits for such targeted tools—much based on lessons gleaned from experiences in other countries—they view the alternative of no action as too costly.
3. CHALLENGES
There are many significant challenges to identifying threats to financial stability and appropriate policy actions. I have discussed some of those associated with devel- oping a robust systemic risk monitoring effort. The information needs are immense. Research on how risks are propagated is still in early stages, because of both insuf- ficient data and complexity of linkages. In addition, in considering potential policies to address building vulnerabilities, financial stability authorities would be, almost by definition, taking away the punch bowl just as the party’s getting fun. While this stance may not be unusual for a central banker, there is less experience than for monetary policy, where over time, tools and targets have become well defined and policy formulation quite sophisticated.
A final challenge is to create a culture of more effective communication across many groups, including supervisors, microeconomists, and macroeconomists within the Fed, and across regulatory agencies that do not regularly share information. In addition, it is important to engage actively with the outside community of researchers and analysts. To that end, the Fed will need more standardized and timelier data, and greater disclosure to encourage interactions. The analysis and data disclosure in the Financial Stability Oversight Council’s annual report is a step in the right direction.7
There is a long and successful history of engagement between academics and poli- cymakers in setting monetary policy, and I hope that promoting such engagement for financial stability will produce a similarly successful payoff.
LITERATURE CITED
Acharya, Viral, Lasse Pedersen, Thomas Philippon, and Matthew Richardson. (2010) “Mea- suring Systemic Risk.” Working Paper, May.
Acharya, Viral V., Philipp Schnabl, and Gustavo Suarez. (2013) “Securitization without Risk Transfer.” Journal of Financial Economics, 103, 515–36.
Adrian, Tobias, and M. Brunnermeier. (2011) “CoVaR.” Federal Reserve Bank of New York Staff Report No. 348, September.
7. Financial Stability Oversight Council (2012).
NELLIE LIANG : 135
Adrian, Tobias, and Hyun Song Shin. (2010) “Liquidity and Leverage.” Journal of Financial Intermediation, 19, 418–37.
Bernanke, Ben. (2010) “Causes of the Recent Financial and Economic Crisis.” Statement before the Financial Crisis Inquiry Commission, September 2010. http://www.federalreserve.gov/ newsevents/testimony/bernanke20100902a.htm.
Board of Governors of the Federal Reserve System. (2012). “Senior Credit Officer Opinion Survey on Dealer Financing Terms.” http://www.federalreserve.gov/econresdata/ releases/scoos.htm.
Brunnermeier, Markus. (2009) “Deciphering the Liquidity and Credit Crunch of 2007-2008.” Journal of Economic Perspectives, 23, 77–100.
Covitz, Daniel, Nellie Liang, and Gustavo Suarez. (2013) “The Evolution of a Financial Crisis: Collapse of the Asset-Backed Commercial Paper Market.” Journal of Finance, 68, 815–848.
Elsinger, Helmut, Alfred Lehar, and Martin Summer. (2006) “Risk Assessment for Banking Systems.” Management Science, 52, 1301–14.
Financial Stability Oversight Council. (2012) “Annual Report.” http://www.treasury.gov/ initiatives/fsoc/Documents/2012%20Annual%20Report.pdf.
Garratt, Rodney J., Lavan Mahadeva, and Katsiaryna Svirydzenka. (2011) “Measuring Sys- temic Risk in the International Banking Network.” Bank of England Working Paper, No. 413.
Gauthier, Céline, Alfred Lehar, and Moez Souissi. (2010) “Macroprudential Regulation and Systemic Capital Requirements.” Bank of Canada Working Paper 2010-4.
Gorton, Gary, and Andrew Metrick. (2012a) “Getting Up to Speed on the Financial Crisis: A One-Weekend-Reader’s Guide.” Journal of Economic Literature, 50, 128–50.
Gorton, Gary, and Andrew Metrick. (2012b) “Securitized Banking and the Run on Repo.” Journal of Financial Economics, 104, 425–51.
Huang, Xin, Hao Zhou, and Haibin Zhu. (2009) “Assessing the Systemic Risk of a Hetero- geneous Portfolio of Banks during the Recent Financial Crisis.” Federal Reserve Board Finance and Economics Discussion Series 2009-44.
The Quarterly Journal of Economics-2012-Stein-57-95.pdf
MONETARY POLICY AS FINANCIAL STABILITY REGULATION∗
JEREMY C. STEIN
This articledevelops a model that speaks tothegoals andmethods of financial stabilitypolicies. Therearethreemainpoints. First, froma normativeperspective, the model defines the fundamental market failure to be addressed, namely, that unregulated private money creation can lead to an externality in which intermediaries issue too much short-term debt and leave the system excessively vulnerable to costly financial crises. Second, it shows how in a simple economy where commercial banks are the only lenders, conventional monetary policy tools such as open-market operations can be used to regulate this externality, whereas in more advancedeconomies it may be helpful tosupplement monetary policy with other measures. Third, from a positive perspective, the model provides an account of how monetary policy can influence bank lending and real activity, even in a world where prices adjust frictionlessly and there are other transactions media besides bank-created money that are outside the control of the central bank. JEL Codes: E58, G01.
I. INTRODUCTION
The modern literature on monetary policy emphasizes the central bank’s role in fostering price stability.1 Historically, however, a dominant concern for central bankers has been not just price stability but also financial stability. Goodhart (1988) argues that the original motivation for creating central banks in many countries was totemper the financial crises associated with unregulated “free banking” regimes:
In the nineteenth century, the advocates of free banking argued that the banking system could be trusted to operate effectively without external con- straints or regulation. . . . [But] experience suggested that competitive pressures in a milieu of limited information (and, thence, contagion risks) would lead to procyclical fluctuations punctuated by banking
∗Eduardo Davila and Fan Zhang provided outstanding research assistance. I am grateful for comments from Robert Barro, Effi Benmelech, Ricardo Caballero, Emmanuel Farhi, MarkGertler, MarvinGoodfriend, RobinGreenwood, Sam Hanson, Larry Katz, Arvind Krishnamurthy, Jamie McAndrews, David Scharfstein, Andrei Shleifer, Robert Vishny, the referees, and seminar partici- pants at numerous institutions. Thanks also to Mary Goodman, Sam Hanson, Matt Kabaker, Andrew Metrick, Charlie Nathanson, Larry Summers, and Adi Sunderam for a series of early conversations that helped shape the ideas in this article.
1. See, for example, Goodfriend (2007) for a recent articulation of this view.
c© The Author(s) 2012. Published by Oxford University Press, on the behalf of President and Fellows of Harvard College. All rights reserved. For Permissions, please email: journals. [email protected]. The Quarterly Journal of Economics (2012) 127, 57–95. doi:10.1093/qje/qjr054. Advance Access publication on January 6, 2012.
57
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
58 QUARTERLY JOURNAL OF ECONOMICS
panics. It was this experience that led to the forma- tion of noncompetitive, non-profit maximizing Central Banks. (p. 77)
A related emphasis on crisis mitigation is evident in Bagehot’s (1873) famous discussion of the lender-of-last-resort function.2 Certainly, recent events have served to underscore the importance of the central bank’s role in preserving financial stability.
In this article, I develop a model that speaks to the goals and methods of central bank financial stability policies. The first step is to define the fundamental market failure that needs to be addressed. I begin with an unregulated banking system in which banks raise financing from households toinvest in projects. Banks can raise this financing in the form of either short-term or long-term debt. Households are risk-neutral with respect to fluctuations in their consumption, but derive additional monetary services from holding any claim that is entirely riskless—with the notion being that riskless claims are easy to value and hence facilitate exchange among households. I show that banks can manufacturesomeamount of riskless private“money”of this sort, therebyloweringtheirfinancingcosts. Moreover, theycandosoin greater quantity by issuing short-term debt, because it is harder for long-term bank debt to be made risk-free.
The role for financial stability policy arises because the pri- vate choices of unregulated banks with respect to money creation are not in general socially optimal. When banks issue cheaper short-term debt, they capture its social benefits, namely, the monetary services it generates for households. However, they do not always fully internalize its costs. In an adverse “financial crisis” state of the world, the only way for banks to honor their short-term debts is by selling assets at fire-sale prices. I showthat in equilibrium, the potential for such fire sales may give rise to a negative externality. Thus, left to their own devices, unregulated banks may engage in excessive money creation and may leave the financial system overly vulnerable to costly crises.3
2. Tucker (2009) paraphrases Bagehot’s (1873) dictum as follows: “to avert panic, central banks should lend early and freely (i.e., without limit) to solvent firms, against good collateral, and at ‘high rates.”’
3. Gersbach (1998) and Hart and Zingales (2011) are other papers in which unregulatedbanks create a socially excessive quantity of private money. However, the externalities in these papers are unrelated to financial stability.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 59
There are a variety of ways for a regulator to address this externality. One possibility is the use of conventional monetary policytools, that is, open-market operations. Toseehowmonetary policy might be of value, note that a crude approach to dealing withtheexternalitywouldbefortheregulatortojust imposea cap oneachbank’s total moneycreation. However, whentheregulator is imperfectly informed about banks’ investment opportunities, he will not know where to set the cap, since it is desirable for banks with stronger investment opportunities to do more money creation. In this setting, the regulator can dobetter with a flexible “cap-and-trade” system in which banks are granted tradable per- mits, each of which allows them to do some amount of money creation.4 The market price of the permits reveals information about banks’ investment opportunities to the regulator, who can then adjust the cap accordingly—when the price of the permits goes up, this suggests that banks in the aggregate have strong investment opportunities, so the regulator should loosen the cap by putting more permits into the system.
All of this maysoundabit likesciencefiction; wedon’t observe cap-and-trade regulation of banks in the real world. However if banks’ short-term liabilities are subject to reserve requirements, it turns out that monetary policy can be used as a mechanism for implementing the cap-and-trade approach. When the central bank injects reserves into the system, it effectively increases the number of permits for private money creation. The nominal inter- est rate, which captures the cost of holding reserves, functions as the permit price. Thus, open-market operations that adjust aggregate reserves in response to changes in short-term nominal rates can be use to achieve the cap-and-trade solution.
An interesting benchmark case is where reserve require- ments apply to the money-like liabilities of all lenders in the economy. This allows the central bank toprecisely control private money creation with monetary policy alone. Although this case may roughly capture the situation facing central banks at an earlier period in history, it is less realistic as a description of modern advanced economies. Nowadays there are a range of short-term financial intermediary liabilities that are not subject to reserve requirements, and yet may both provide monetary services and create fire-sale externalities. For example, Gorton
4. Kashyap and Stein (2004) suggest using an analogous cap-and-trade approach to implement time-varying bank capital requirements.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
60 QUARTERLY JOURNAL OF ECONOMICS
and Metrick (2011), and Gorton (2010) argue that an important fractionofprivatemoneycreationnowtakes placeentirelyoutside of the formal banking sector, via the large volumes of short-term collateralized claims created in the “shadow banking” sector.
In this richer environment, monetary policy as convention- ally practiced is generally not sufficient to rein in excessive money creation. Continuing with the foregoing example, it may additionally be necessary to regulate the volume of activity in the shadow banking sector, either by expanding the reach of reserve requirements or by some other means. Thus the model helps make clear the circumstances under which monetary policy needs to be supplemented with other measures. Moreover, it sug- gests that these other measures lie squarely in the central bank’s traditional domain, totheextent that theyarealsotargetedat the fundamental externality associated with excessive private money creation. This is of interest in light of the ongoing debate over the appropriate mix of central bank tools for achieving financial stability.5
In addition to its normative implications, the model is also relevant from a positive perspective. It provides a coherent account of how monetary policy “works”—that is, of how open market operations lead to changes in bank lending and output— in an environment that is arguably more realistic than in other theories. In contrast to the usual model, prices are fully flexible. Moreover, I do not need to assume that the central bank has monopoly control over all forms of transactions media. The model is unchanged if one introduces a set of nonreservable securities that providethesamemonetaryservices as bank-createdmoney.6
Indeed, I consider the limiting case where the interest rate spread between money and bonds is fixed and unresponsive totheir rela- tive supplies. Monetary policy works in this case not by changing real interest rates but through a pure quantity effect: a loosening of policy allows banks to finance themselves with more of the cheaper money, which encourages them to do more lending.
5. See, for example, Adrian and Shin (2008) and Ashcraft, Garleanu, and Pedersen (2010).
6. To be clear on the distinction: my model assumes that the central bank acts as a regulator, controlling those forms of private money creation that lead to negative externalities—in particular, short-term bank debt that finances risky long-term assets. However, it does not require the central bank to control more benign forms of money creation, for example, money market fund accounts backed exclusively by Treasury bills.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 61
The ideas in this article connect to several strands of pre- vious work. First, the basic model of fire sales that creates the rationale for policy intervention draws on Shleifer and Vishny (1992, 1997).7 Second, the insight that banks create a valuable transactions medium by issuing low-risk claims is formalized in Gorton and Pennacchi (1990). Third, the notion that central bank reserves can be thought of as permits that allow banks to do more of a particular kind of cheap financing appears in Stein’s (1998) elaboration of the bank lending channel of monetary policy transmission.8
Finally, to focus on the financial stability consequences of monetary policy, it helps to set aside its effects on price stability. I dosobyappealingtothefiscal theoryof thepricelevel, according to which the price level is determined not by the monetary base but by total outstanding nominal government liabilities—that is, by the sum of Treasury securities and the monetary base.9 This enables openmarket operations that changethe mix of Treasuries and bank reserves (while keeping their sum constant) to have real effects on bank investment and financing behavior, even in a world where all prices are perfectly flexible. However, I also discuss how the model’s conclusions carry over to an alternative NewKeynesian setting with sticky prices, where price stability is governed by a version of the “Taylor rule” (Taylor 1993, 1999).
Therest of thearticle is organizedas follows. Section II devel- ops thebasicmodel ofprivatemoneycreationbybanks. Section III compares banks’ financing choices to the social planner’s solution andclarifies theconditions underwhichbanks engageinexcessive money creation. It also shows that a cap-and-trade approach to regulation can be useful when the social planner has imper- fect information. Section IV demonstrates how the cap-and-trade approach can be implemented with open market operations. Section V explores a number of other complementary policy tools; these include liquidity regulation, deposit insurance, and a lender-of-last-resort function, as well as regulation of the shadow
7. On fire sales, see also Kiyotaki and Moore (1997), Gromb and Vayanos (2002), Morris and Shin (2004), Allen and Gale (2005), Fostel and Geanakoplos (2008), Brunnermeier and Pedersen (2009), Stein (2009), Caballero and Simsek (2010), and Geanakoplos (2010).
8. Forearlyworkonthebanklendingchannel, seealsoBernankeandBlinder (1998, 1992); Kashyap, Stein, and Wilcox (1993); and Kashyap and Stein (2000).
9. The fiscal theory is developed in Leeper (1991), Sims (1994), Woodford (1995), and Cochrane (1998). My own adaptation of the theory is particularly indebted to Cochrane’s exposition.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
62 QUARTERLY JOURNAL OF ECONOMICS
banking sector. Section VI discusses how the model differs from other accounts of the monetary transmission mechanism. Conclu- sions are in section VII.
II. A MODEL OF PRIVATE MONEY CREATION
The model features three sets of actors: households, banks, and“patient investors.”I beginbydescribingeachof thesegroups, and then turn to the optimization problem faced by the banks.
II.A. Households
There are three dates, 0, 1, and 2. At time 0, households have an initial endowment of the one good in the economy. They can either consume this endowment at time 0, or invest some of it in financial assets and consume the proceeds from investment at time 2. They have linear preferences over consumption at these two dates. In addition to consumption, households also derive utility from monetary services. The key assumption is that monetary services can be provided by any privately created claim on time 2 consumption, so long as that claim is completely riskless.10 Thus the utility of a representative household is given by:
(1) U = C0 + βE(C2) + γM,
where M represents the household’s time 0 holdings of privately created“money.”11 Tobeclearonthenotational convention, when a household has M units of money at time 0, this means that it holds claims guaranteedtodeliver M units of time2 consumption.
10. This assumption is meant to capture the spirit of Gorton and Pennacchi (1990) and Dang, Gorton, and Holmstrom (2009). These papers argue that information-insensitive securities are an attractive medium of exchange because they eliminate the potential for adverse selection between transacting parties. My formulation implies that households do not derive monetary services from state- contingent deposits (as in Hellwig 1994). If they did, efficiency might be improved by having banks issue claims that pay a lower return in bad states of the world. However, if some agents have a better ability than others toforecast when the bad state is coming, such state-contingent claims would be subject toadverse selection problems.
11. In a similar formulation, Krishnamurthy and Vissing-Jorgensen (2010) put the stock of Treasury securities directly into the representative agent’s utility function. As one rationale for doing so, they cite the “surety” of Treasuries—that is, the fact that Treasuries are riskless. Like I do, they posit that surety has an extra value above and beyond what is captured in a standard asset pricing model. See also Sidrauski (1967) for an early model with money in the utility function.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 63
Given their linear form, household preferences pin down two real rates. The first is the (gross) real return on risky “bonds” that pay off at time 2, given by RB = 1
β . The second is the
(gross) real return on riskless “money,” given by RM = 1 (β+γ), where
β + γ < 1. Thelatterfollows fromtheobservationthat a household is always indifferent between having: (1) β + γ units of time 0 consumption; or (2) a riskless claim that promises one unit of time 2 consumption, since such a claim delivers β of utility from expected future consumption, along with an additional γ of utility in monetary services. The bottom line is that because riskless money offers households a convenience yield that risky bonds do not, in equilibrium it must have a lower rate of return.
The idea that money has a lower return in equilibrium than bonds is standardin textbook models. But here, the return spread is fixed and independent of the quantities of money and bonds, thanks to the linear preferences on the part of households. This feature is not necessary for anything that follows and is easily relaxed. However, it serves tohighlight a keynoveltyof mymodel: here, changes in central bank policy work not by altering the real rates oneithertypeof claimbut byvaryingtheproportions of each that banks use. In other words, looser policy encourages banks to lend more by enabling them to tilt their capital structure toward cheap money financing, thereby lowering their weighted average cost of funds.
II.B. Banks
Households cannot invest their time 0 endowments directly in physical projects, because they do not have the monitoring expertise todoso. This investment must be undertaken by banks, who in turn issue financial claims—in the form of either riskless money or risky bonds—to households. There is a continuum of such banks, with total mass of one. Each bank faces the following investment opportunities. If an amount I is invested at time 0, and the good state prevails, which happens with probability p, total output at time 2 is given by the concave function f (I) > I. If instead the bad state prevails, total expected output at time 2 is λI ≤ I, and there is a positive probability that output collapses all the way tozero. In particular, in the bad state, output is either λI
q with probability q, or zero with probability (1− q).
At time 1, there is a public signal that reveals whether the good or bad state will be realized at time 2. At time 1 it is also
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
64 QUARTERLY JOURNAL OF ECONOMICS
possible for a bank to sell any fraction of its existing physical assets toa patient investor.12 If a fractionΔ of the assets are sold, total proceeds to the bank are given by Δ kλI, where 0 ≤ k ≤ 1, andtheremainingunsoldassets yieldoutput at time2 tothebank of (1−Δ)λI. Thus k is a measure of the discount toexpected value associated with a time 1 asset sale. A central feature of the model is that k is endogenous and depends on total asset sales by all banks in the economy. The equilibrium determination of k will be discussed shortly.
Other than their access to investment opportunities, banks have no initial endowments, and hence must raise the entire amount I externally. They can do so by issuing either short-term (maturingat time1) or long-term(maturingat time2) debt claims to households. Note that if they finance with long-term debt, no amount of this debt can ever be riskless, because there is a positive probability of the assets yielding zero output at time 2. By contrast, short-term debt can be made riskless, if not toomuch is issued. This is because by forcing an asset sale on seeing a bad signal at time 1, short-term creditors can escape early with a sure value equal to the proceeds from the sale.
These assumptions are starker than they need to be. In a more general model where the lowest possible value of output at time 2 is greater than zero, banks can issue some riskless long- term debt—so there is no longer a one-to-one mapping between debt maturity and the ability for debt to be made risk-free. Nevertheless, it will always be the case that banks can create a larger quantity of riskless claims by issuing short-maturity debt; the early escape intuition still holds. Because there is a fixed premium on riskless claims, banks will continue to be tempted to issue short-term debt in this more general version of the model, and all the qualitative results that follow will continue to apply.
The model can alsobe extendedsothat monetary services are providednot onlybyentirelyriskless assets but byanyclaims that are sufficiently low risk—that is, by any claims whose worst-case payoffis at least x cents onthedollar. What is critical is that there still be a violation of the Modigliani and Miller (1958) conditions, sothat as a bank manufactures more of these low-risk money-like claims, it does not have to pay more for its remaining long-term debt, which becomes riskier. This M-M violation is captured here
12. Because households only consume at time 0 and time 2, they do not consume the proceeds of any time 1 asset sales until time 2. One can think of them as simply sitting on these proceeds in the interim.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 65
in the assumption that the return on nonmonetary claims, RB, is a constant.
In any of these formulations, the key trade-off is this: on one hand, banks have an incentive to issue some short-term debt, because more of this debt can be made low-risk—and hence by virtue of its money-ness, represents a cheap form of finance.13
On the other hand, what keeps short-term debt safe is the bank’s ability to sell assets in the bad state. As will become clear, these sales of existing assets can lead tosocial costs that are not always fullyinternalizedbyindividual banks whentheypicktheircapital structures. As a result, there may be excessive private money creation by banks.
Supposethat a bankraises a fraction m of its total investment of I byissuingshort-termdebt. If this short-termdebt canbemade riskless, it will carry a rate of return of RM, and the bank will owe its short-term creditors a repayment of mIRM ≡ M. Can it meet this promise in the bad state by selling assets if necessary? From before, if it sells a fractionΔ of its assets, total proceeds areΔ kλI, so we require that:
(2) Δ kλI = mIRM, orΔ = mRM
kλ .
Since Δ ≤ 1, there is an upper bound on private money creation given by:
(3) mmax = kλ RM
.
Thus, thepotential forasset sales makes it possiblefora bank to create riskless private money by issuing short-term debt—as long as the amount issued is not too large.
Is it also the case that asset sales are an unavoidable con- sequence of money creation? One might think that since holding on to assets is positive NPV relative to selling them at time 1, it might be possible for a bank to raise new funding at time 1 to pay off the departing short-term creditors, and thereby avoid forced sales. However, if one assumes that any new funding must be subordinated to existing long-term debt, such new funding may be blockaded by a severe debt overhang problem (Myers 1977),
13. Other theories of short-term financing include Flannery (1986), Diamond (1991), and Stein (2005), who stress its signaling properties, and Diamond and Rajan (2001), who argue that short-term debt is a valuable disciplining device, particularly for financial intermediaries.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
66 QUARTERLY JOURNAL OF ECONOMICS
given the low value of the assets in the bad state relative to the total face amount of already-issued debt.14 Thus, under plau- sible circumstances, private money creation inevitably leads to some amount of asset sales.15
Note that I assume that banks invest all the resources that they raise at time 0 in real projects and do not hold any back as a buffer against a loss of funding at time 1. However, this is without loss ofgenerality, sinceabankalways has theoptiontochangethe mix of its short-term versus long-term debt, which has a similar buffering benefit and is more cost-effective. In other words, there are two ways to reduce asset sales in the bad state by $1: either borrow an extra dollar of long-term debt at time 0 and park the proceeds in storage, or borrow an extra dollar of long-term debt at time 0 so as to reduce the amount of short-term borrowing by $1. The net cost of the former transaction is (RB − 1), and the net cost of the latter is (RB − RM). Hence, the latter approach is strictly preferred, and banks endogenously choose not to engage in storage.
Before moving on, it is worth fleshing out an issue of inter- pretation about the banks in the model. In the real world, banks donot invest in physical projects directly, but lendtofirms whoin turn do the project selection. Abstracting away from this extra layer of activity, as I do here, is tantamount to assuming that there are no contracting frictions between operating firms and banks, that is, that firms can costlessly pledge all of their output to the banks. This raises the question of whether it is appropri- ate to interpret what I label “banks” as really being financial
14. In particular, denoting the face value of the existing long-term debt by B, it must be that M + B > I, for the bank tohave raised I at time 0 by issuing money and bonds. If the bank now wants to raise an amount M to pay off the short-term creditors in the bad state at time 1, it must do so by issuing new claims that are junior to the existing long-term debt. But given that they are junior, the value of these claims in the bad state is only q(λI/q− B). For q large enough (certainly for q > λ) the value of the new claims is necessarily less than M, so refinancing the short-term debt is impossible.
15. This line of argument leaves open the question of why the original long- term financing for the bank is in the form of senior debt, as opposedto, say, equity, or some other junior security that allows for new financing to come in on top of it. Following Hart and Moore (1995), it may be that this seniority of the long- term debt represents a valuable precommitment in the more likely good state of the world. For example, it may prevent managers from using assets in place as collateral for empire-building investments. Thus, as in Hart and Moore, senior long-term debt is a double-edged sword: it serves to discipline wayward managers in the good state, but forces underinvestment (here, in the form of asset sales) in the bad state.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 67
intermediaries, as opposedtooperating firms that borrowdirectly from households in the securities market.
To create a meaningful distinction, suppose that any individ- ual operating firm, once funded, always has some probability of immediate (i.e., before time 1) idiosyncratic failure, in which case it becomes publicknowledgethat its output will bezeroinboththe goodandbadstates. This riskof failuremakes it impossible foran operating firm toever issue riskless claims in any amount. Banks, on the other hand, represent highly diversified portfolios of such firm-level projects, andthereforetheirassets always havepositive expected value as of time 1, as assumed. The diversification associated with banks is thus a necessary condition for them to create riskless claims.16
II.C. Patient Investors
Patient investors (PIs) are another type of intermediary, and as such, any output that they produce reverts to the household sector at time 2. As a group, PIs are endowed with resources of W at time 1. For simplicity, I treat this endowment as exogenous for now, but it can be endogenized by allowing the PIs toraise the W from the household sector at time 0 by issuing risky long-term claims. In this case, the PIs choose an optimal level of W at time 0 that equates the expected return on their time 1 investments to the cost of capital RB. Imposing this ex ante breakeven condition does not affect the qualitative results of the model, soI set it aside for the time being.
What is crucial is that when time 1 rolls around and the state of the world is realized, W is fixed. Thus, although it is fine to think of PIs as having full access to financial markets at time 0, they cannot go back and raise more at time 1 once they know the state. In other words, W is an unconditional war chest, with the same amount available to PIs in the good and bad states. This assumption can be thought of as a crude stand- in for the phenomenon of “slow-moving capital” (Duffie 2010). A more explicit micro-foundation might involve an information asymmetry between PIs and households at time 1—for example, the PIs get a private signal about the quality of their investment
16. Thus, as in other models of intermediation, both pooling (i.e., diversifica- tion) andtranching(i.e., theissuanceof properlystructuredseniorsecurities) have roles to play in creating low-risk claims. See, for example, Gorton and Pennacchi (1990), DeMarzo and Duffie (1999), and DeMarzo (2005). Diamond (1984) also emphasizes the importance of diversification to the process of intermediation.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
68 QUARTERLY JOURNAL OF ECONOMICS
opportunities at this time, which creates an adverse selection problem for any further attempts to raise financing.
PIs can do one of two things with their resources at time 1. First, they can invest in new, late-arriving real investment projects. Irrespective of the state of the world, an investment of K in such new projects at time 1 yields expected gross output of g(K) at time 2, where g( ) is a concave function. Alternatively, PIs can absorb assets being sold by banks at time 1. In the good state, there are noasset sales, sothe PIs invest all of W in newprojects, yielding g(W). Inthebadstate, banks havetosell enoughassets to repay short-term creditors the M they have promised them. Thus in equilibrium, PIs spend M on asset purchases, and invest only (W − M) in new projects, yielding g(W − M). For the PIs to be willing to allocate their endowment in this way, it must be that the marginal return on new projects is the same as the marginal return from buying existing assets from banks. This is what pins down the fire-sale discount k. In particular, we have that:
(4) 1 k
= g′(W −M).
Equation(4) makes clearthereal costs of firesales, andhence of short-term debt financing by banks. The greater is M, and hence the more bank assets that the PIs have to absorb in the bad state at time 1, the less they have left over for investment in new projects. With scarce PI capital, the return on secondary market arbitrage opportunities (buying up fire-sold assets) also becomes the hurdle rate for new investment, a point emphasized by Diamond and Rajan (2009) and Shleifer and Vishny (2010).
For expositional purposes, I treat the PIs and the banks as two distinct categories of intermediaries. This is not necessary; one could alternatively merge them into a single entity that has investment opportunities at both time 0 and time 1, issues some short-term debt at time 0, and also holds liquidity W in reserve at time 0. This reinterpretation of the model is innocuous, subject to one caveat: it is crucial that the merged entities behave not as autarkic islands but as price-takers who can transact in the asset market at time 1. Thus, even if a bank satisfies most of its departing creditors by drawing down on its own stock of liquidity at time 1, it must continue to consider the possibility of asset sales to another bank. This feature emerges naturally if we move away from the knife-edge case where the scales of time 0 and time 1 investment are in identical proportions across all banks.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 69
If so, those that have relatively bigger time 1 scale will tend to stockpile more W relative to their short-term debts, and hence will be buyers of assets from those who have bigger time 0 scale. My two-categories formulation can be thought of as capturing an extreme case of this heterogeneity.
II.D. The Bank’s Optimization Problem
I formulate the optimization problem for a bank that invests an amount I and finances it with some fraction m ≤ mmax of money. The bank’s expected net profits at time 2 are given by:
(5) Π = {pf (I) + (1− p)λI − IRB} + mI(RB − RM)− (1− p)zmIRM,
where I have defined z = (1−k) k as the net rate of return on
fire-sold assets. (Note that higher values of z correspond to larger fire-sale discounts, and z = 0 is the case where there is no discount.) The three terms in equation (5) are easily interpreted. The first, {pf (I) + (1 − p)λI − IRB}, is the NPV of investment assuming that it is entirely financed at the higher bond market rate—and hence that there is no need to ever sell assets. The secondterm, mI(RB−RM), is the financing cost savings associated with using a fraction m of money in the capital structure. The last term, (1 − p)zmIRM, captures the expected fire-sale losses associated with this riskier short-term capital structure.
Each bank picks m and I to maximize equation (5), subject to the collateral constraint that m ≤ mmax = kλ
RM . I assume that each bank treats the fire-sale discount k as a fixed constant—that is, they do not internalizxe the incremental impact of their choices on the fire-sale outcome. By contrast, when I examine the social planner’s problem, the key difference will be that the planner takes into account the dependence of k on the capital structure of the banks. The Lagrangian for the bank’s problem is thus: (6)
LB={pf (I) + (1−p)λI−IRB} + mI(RB−RM)−(1−p)zmIRM−η(m− kλ RM
),
where η is the shadow value of the collateral constraint. Taking the first-order condition with respect to m, we have:
(7) I{(RB − RM)− (1− p)zRM} = η.
It follows that the collateral constraint binds, andthe bank is at a corner, setting m = mmax, if (RB−RM)> (1−p)zRM, that is, if the equilibrium spread between bonds and money is sufficiently
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
70 QUARTERLY JOURNAL OF ECONOMICS
large. Alternatively, if the spread is smaller in equilibrium (that is, if (RB − RM) = (1 − p)zRM), then the bank chooses an interior value of m, and η = 0.
The first-order condition with respect to I yields:
(8) pf ′(I) + (1− p)λ− RB + m(RB − RM)− (1− p)zmRM = 0.
Using equation (7), we can rewrite equation (8) as follows:
(9) pf ′(I) + (1− p)λ− RB = −ηm
I .
There are two ways that equation (9) can be satisfied. First, the bank can be at an interior solution with respect to m, in which case η = 0, and therefore pf ′(I) + (1− p)λ = RB. Alternatively, the bank can be at a corner with m = mmax, and η > 0, in which case it follows that pf ′(I) + (1 − p)λ < RB. This reasoning leads to the following proposition.
PROPOSITION 1. Define IB as the optimal level of investment for a bank that finances itself exclusively in the long-term bond market: pf ′(IB) + (1− p)λ−RB = 0. The solution to the bank’s problem involves two regions. In the low-spread region (for (RB − RM) relatively small) the bank chooses m < mmax and I∗ = IB. In the high-spread region (for (RB − RM) relatively large) the bank chooses m = mmax and I∗ > IB.
The point to take away from the proposition is that in the low-spread region, a bank’s investment and financing choices are decoupled, whereas in the high-spread region they are interde- pendent. This is because when m < mmax, a bank’s ability to tap low-cost money financing is not constrained by the amount of investment it does. Bycontrast, inthehigh-spreadregioninwhich m=mmax, a bank faces a binding collateral constraint—it can only issue more money if it increases the quantity of physical assets backing its debts. This is what ties investment and financing decisions together. If money financing is cheapenough that banks want to do a lot of it, and they begin to bump up against the collateral constraint, they will be induced to invest more so as to loosen the constraint.
III. SOCIALLY EXCESSIVE MONEY CREATION: A ROLE FOR
REGULATION
The next step in the analysis is to identify the circumstances in which the process of private money creation already described
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 71
involves an externality—that is, when the level of money creation chosen by banks exceeds that preferred by a benevolent social planner.
III.A. The Social Planner’s Problem
Given that all output of the banks and the PIs ultimately accrues to the household sector, the social planner seeks to maximize the utility of a representative household, as given by equation (1). It is easily shown that, disregarding constants, this utility, expressed in units of time 2 consumption, is equivalent to:17
U = {pf (I) + (1− p)λI − IRB} + M (RB − RM)
RM + pg(W)(10)
+(1− p){g(W −M) + M} −WRB.
Comparing this tothe bank’s expected profits in equation (5), we can see that the first two terms coincide. The difference is in the latter three terms: the planner does not care about expected fire-salelosses perse, becausetheseonlyrepresent atransferfrom thebanks tothePIs. However, theplannerdoes careabout thenet expectedreturns toinvestment by the PIs, as capturedby pg(W)+ (1− p){g(W −M) + M} −WRB.
Theplannerfaces thesamecollateral constraint as thebanks, namely, that m ≤ mmax = kλ
RM . Denoting the shadow value of the constraint in this case by ηP, and recalling that M = mIRM, the Lagrangian for the planner’s problem is given by:
LP = {pf (I) + (1− p)λI − IRB} + mI (
RB − RM )
+ pg(W)(11)
+ (1− p){g(W −mIRM) + mIRM} −WRB − ηP (
m− kλ RM
) .
In taking the first-order conditions for this problem, it is important to note that unlike an individual bank, the planner recognizes the dependence of k on the average behavior of all banks—he understands that, as per equation (4), k = 1
g′(W−mIRM).
17. Inparticular, supposehouseholds havea fixedtime0 endowment of Y, and that they invest I of this endowment with the banks and W with the PIs at time 0. It follows that C0 = Y − I −W, and that C2 = f (I) + g(W) with probability p, and C2 = λ I + g(W −M) + M with probability (1− p). The expression in equation (10) then follows from also including the monetary services γM in the utility function, and multiplying time 0 values by RB to put everything in common units of time 2 consumption.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
72 QUARTERLY JOURNAL OF ECONOMICS
Using this fact, the first-order condition with respect to m can be written as:
(12) I{(RB − RM)− (1− p)zRM} = ηP
(
1− g′′( ∙ )
(g′( ∙ ))2 λI
)
.
Similarly, the first-order condition with respect to I can be expressed as: (13)
pf ′(I)+(1−p)λ−RB +m(RB−RM)− (1−p)zmRM =−ηP g′′( ∙ ) (g′( ∙ ))2
λm.
Comparing equations (7) and (12), and equations (8) and (13), we can see that the bank’s private solution coincides exactly with the social planner’s solution in the low-spread region where (RB − RM) = (1− p)zRM, and where the collateral constraint is nonbinding, that is, where η = ηP = 0. In this case, equation (13) reduces to equation (8), meaning that the planner chooses the same level of I as the bank.
By contrast, in the high-spread region where the constraint binds, so that ηP > 0, the term on the right-hand side of equation (13), −ηP g′′(∙)
(g′(∙))2λm, describes the wedge between the bank’s solu- tion and the planner’s solution. Since g′′( ∙ ) < 0, this term is positive, which implies that the marginal product of investment is higher in the social planner’s solution, or alternatively that I is lower. In other words, in this region, the social planner would like to restrain investment, and hence money creation, relative to the private outcome.
The following proposition summarizes the analysis.
PROPOSITION 2. Denote the private and socially optimal values of investment I by I∗ and I∗∗, respectively, and similarly for the private and socially optimal values of money creation M. In the low-spread region, I∗ = I∗∗, and M∗ = M∗∗. In the high- spread region, I∗ > I∗∗, and M∗ > M∗∗.
Thus banks may create a socially excessive amount of money, but this happens only if the spread between money and bonds (RB−RM) is highenough. If thespreadis solowthat anyindividual bank choose an interior value of money creation m < mmax, there is no divergence between private and social incentives.
EXAMPLE 1. Pick these functional forms and parameter values: f (I) = ψlog(I) + I, g(K) = θlog(K), RB = 1.04; RM = 1.01; ψ = 3.5; θ=150; λ=1; W=140; andp=0.98. Forthesevalues, theprivate
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 73
FIGURE I
Private and Socially Optimal Outcomes versus the Money-Bond Spread
The figure plots private and socially optimal values of money creation M and investment I as a function of RM . Functional forms and parameter values are as follows: f (I) = ψlog(I) + I; g(K) = θlog(K); RB = 1.04; ψ = 3.5; θ = 150; λ = 1; W = 140; and p = 0.98. RM varies between 1.0 and 1.035.
optimum is in the high-spread region and involves banks choosing M∗ = 57.6 and I∗ = 104.9, with an associated rate of return on fire-sale assets of z = 82.1% (k = 0.549). By contrast, in the social optimum, the planner chooses M∗∗ = 55.2 and I∗∗ = 97.7, leading to a rate of return on fire-sale assets of z = 77.0% (k = 0.565).
Figure I expands on Example 1, keeping all of the other parameter values the same as before, but allowing RM to vary between 1.00 and 1.035, thereby causing the bond money spread (RB − RM) to vary between 50 and 400 basis points. As can be seen, for lowvalues of the spread, the private andsocially optimal values of M and I coincide. But as the spreadwidens, these values diverge further and further from one another.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
74 QUARTERLY JOURNAL OF ECONOMICS
III.B. Understanding the Nature of the Externality
At first glance, it may not be clear why fire sales create a divergence between private and socially optimal outcomes. After all, the price impact of liquidations is a pecuniary externality, and pecuniary externalities by themselves need not lead to violations of the standard welfare theorems. The result in Proposition 2 is a specific case of the generic inefficiency result in economies with incomplete markets (Geanakoplos and Polemarchakis 1986; Greenwald and Stiglitz 1986). Perhaps the closest analogs are Caballero and Krishnamurthy (2003) and Lorenzoni (2008), who also show how there can be socially excessive borrowing in economies with various financial frictions. In the current setting, the key friction is the presence of a binding collateral constraint. When this constraint is operative, any one agent’s impact on mar- ket prices affects other agents not only by altering their budget constraints but also by loosening or tightening their collateral constraints. The first welfare theorem effectively says that pecu- niary externalities that operate solely through prices in budget constraints do not lead to inefficiencies, but when prices show up elsewhere, this conclusion no longer holds.
The importance of the collateral constraint can be seen in the expression for the wedge between the bank’s first-order con- dition and that of the planner; as noted, this wedge is given by: −ηP g′′(∙)
(g′(∙))2λm. Thus when the collateral constraint does not bind,
that is,when ηP = 0, there is no wedge, and the private and social solutions coincide. By contrast, when the collateral constraint binds, there is a wedge to the extent that g′′(∙)
(g′(∙))2 < 0, that is, to the extent that an increase in liquidations widens the fire-sale discount, or equivalently, raises the marginal product of time 1 investment by the PIs.
The intuition behind this result can be understood as follows. When the constraint does not bind, equation (7) tells us that in decidinghowmuchmoneytocreate, eachbanktrades offthelower financing cost (RB−RM) associatedwith money against the poten- tial for greater fire-sales discounts (1 − p)zRM. But according to equation (12), this is exactly the same trade-off the planner faces in attempting to balance the marginal value of monetary services to households against the marginal cost of underinvestment by the PIs. Hence in this case, everything is well internalized.
By contrast, when the constraint binds, and each bank is setting m = mmax, an incremental increase in money creation by
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 75
any one bank has an added effect: by reducing the equilibrium value of k, it effectively lowers the collateral value of all other banks’ assets, thereby tightening their collateral constraints and impinging on their ability to create money. Thus, when any one bank creates an additional unit of money andcaptures the private benefit for doing so, the social benefit is less than that one unit of money, because other banks can nolonger produce as much M for a given level of I.18
The result that there is no externality in the low-spread region when m < mmax is dependent on the strong assumption that whenthePIs invest inreal projects, theycaptureall thesocial surplus associated with these projects. If one adds another finan- cial friction to the model, and makes this surplus only partially pledgeable, private money creation is always socially excessive, irrespective of parameter values. In particular, suppose that the social returntoaninvestment project financedbyaPI is still given byg(K), but onlyϕg(K)canbepledgedtothePI, withϕ < 1. Inthis case, the equilibrium determination of k in (4) is altered so that 1 k =ϕg′(W−M). That is, a given amount of underinvestment by the PIs is now associated with a smaller fire-sale discount. Hence, a bank’s aversion to fire sales no longer leads it to fully internalize the social costs of underinvestment.
This imperfect pledgeability variant of the model is briefly explored in the Appendix. Because it is possible to make many of the key normative points that follow without introducing imperfect pledgeability, there is a certain minimalist appeal to focusing on the perfect pledgeability limit of ϕ = 1, as I do in the remainder of the text. However, if one is interested in generating more realistic comparative statics along some dimensions, the augmented version of the model that allows for ϕ < 1 may be better suited to doing so. For example, I show in the Appendix that theperfect pledgeabilityversionof themodel yields thesome- what counterintuitive implication that the central bank should lower nominal interest rates when the risk of a financial crisis is greater. If instead we posit that ϕ < 1, this result can easily be reversed.
18. Think of two banks, A and B, as factories that each have a technology for producing money out of physical assets. When the collateral constraint binds, an incremental increase in money production by A is equivalent toa form of pollution that gums upB’s production technology, since it reduces the amount of money that B can manufacture out of a given stock of physical assets.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
76 QUARTERLY JOURNAL OF ECONOMICS
III.C. A “Cap-and-Trade” Approach to Bank Liquidity Regulation
The analysis thus far makes clear that in some cases banks will choose tocreate more money than is socially optimal, thereby inflicting inefficiently high levels of fire sales on the economy. This suggests a role for regulation. In the full information case, in which the regulator observes all the relevant parameters of the model, the social optimum can be easily implemented with a cap on money creation: each bank can simply be prohibited from issu- ing more short-term claims than the desired level of M∗∗, which the regulator can directly compute from equations (12) and (13).
However, if the regulator is imperfectly informed, it becomes more challenging to set the cap appropriately.19 Consider a situ- ation in which banks know the productivity of their investment opportunities—that is, they know what the function f (I) looks like—but the regulator does not. As can be seen from equation (13), the value of I∗∗, and hence the value of M∗∗, depends on the marginal product of investment f ′(I). Intuitively, it makes sense to allow banks to create more cheap money financing when they have better investment opportunities. Thus without knowledge of the value of f ′(I), it is impossible for the regulator to target the socially optimal level of money creation with a simple cap.
One way for the regulator to generate the required informa- tion is through a system of cap and trade. In particular, each bank can be grantedpermits that allowit toissue some amount of money; bypickingtheaggregatequantityofpermits, theregulator can, as before, effectively target the total amount of money M in theeconomy. Moreover, if thepermits canbetradedamongbanks, their market-clearing price P(M) (per unit of money creation allowed) will equal the shadow value of the M-constraint to the banks:
P(M) = dΠ dM = 1
mRM dΠ dI .20 Conditional on the regulator knowing
the other parameters of the model, observing dΠ dI allows him to
infer the value of f ′(I). It follows from this reasoning that the regulator can imple-
ment the M∗∗ solution by making the permits tradable, and then
19. Weitzman (1974) is the seminal paper on regulation in the face of param- eter uncertainty.
20. Note that because the banks in the model are all identical, the volume of trade in the permits is zero. Nevertheless, there is a unique equilibrium price, given by the common shadow value of the M-constraint.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 77
targeting the appropriate price for these permits by varying the available quantity. That is, the regulator adjusts the quantity of permits, looking for a fixed point where the market-clearing price P(M) equals a target value PT(M) that itself depends on the quantity of permits. To calculate this target value, recall that in the high-spread region when m = mmax, the social optimum involves dΠ
dI = −ηP g′′(∙) (g′(∙))2λm, which would imply setting PT(M) =
1 mRM
dΠ dI = −ηPλ
RM . g′′(∙) (g′(∙))2 . Using equation (12), we can substitute for
ηP to obtain the following result.
PROPOSITION 3. A regulator who is imperfectly informed about the nature of bank lending opportunities can implement the desired level of money M∗∗ with a system of tradable permits for money creation. This involves adjusting the number of permits such that their observed market-clearing price P(M) equals the following target price PT(M):
(14) PT(M) =
{ (RB − RM)
RM − (1− p)z
}
−λI g′′(∙) (g′(∙))2
( 1− λI g′′(∙)
(g′(∙))2
)
.
To be clear on the implementation, suppose the regulator picks an initial trial value of M. At this value, the regulator can calculate the target price of permits PT(M) from equation (14), basedon his knowledge of M andthe other observable parameters of the model—as can be seen from equation (14), he does not need toknowanything about the value of f ′(I) toevaluate PT(M). If the market price of permits P(M) turns out to be higher than PT(M), the regulator increases M, and vice versa. The optimum M∗∗ is that value of M where the target price in equation (14) coincides with the market price.
EXAMPLE 2. Keep everything the same as in Example 1: f (I) = ψlog(I)+ I, g(K)= θlog(K), RB =1.04; RM =1.01; ψ=3.5; θ=150; λ = 1; W = 140; and p = 0.98. At the social optimum of M∗∗ = 55.2, the price of permits is P=0.0056. Nowsuppose there is a positive productivity shock, andψ rises to4.0. If the capis not adjusted, the price of permits spikes to P = 0.0146. However, this price increase reveals the newvalue of ψ tothe regulator, whocanincreasethenumberof permits inthesystem, raising the quantity of money in the system to its new optimal value of M∗∗ = 58.9. At this new optimum, the price of permits is given by P = 0.0054.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
78 QUARTERLY JOURNAL OF ECONOMICS
The example suggests that in the face of productivity shocks, it is optimal for the regulator to actively lean against incipient changes in the price of permits. When a positive shock pushes the price of permits up, the regulator should increase the supply of permits, thereby driving their price back down. In fact, optimality in this setting requires the supply response to be sufficiently strong that the equilibrium price of permits actually falls slightly as productivity rises.
III.D. Relationship to Pigouvian Taxation
A handful of recent papers have suggested that a system of Pigouvian taxes might be used to force banks to properly internalize any systemic externalities they create (e.g., Jeanne and Korinek 2010; Kocherlakota 2010; Perotti and Suarez 2010). In the current context, this would amount to imposing a tax τ on eachunit of moneycreatedbybanks. A coupleof points about such taxes are worth noting.
First, in the full information case where the planner observes everything needed to compute the socially optimal level of money creation M∗∗, this outcome can be achieved equally well either with a regulatory cap on money creation, or by picking the correct valueof thetaxτ. Indeed, givenfull information, theregulatorcan implement M∗∗ simply by setting τ = PT(M∗∗), that is, the target price of permits given by equation (14), calculated at the desired value of M∗∗. So Pigouvian taxes can be used, but in this setting they donot add any value relative tomore conventional quantity- based regulation.
Second, intheincompleteinformationcasewheretheplanner does not knowenoughtopicktheright level of thecap, healsodoes not know enough to set the correct value of the tax τ, because the optimal tax depends on M∗∗. Thus, an optimal system of Pigouvian taxation still requires a mechanism toelicit the private information. So the cap-and-trade design remains useful, for the same reasons as before. Indeed, one can interpret the cap-and- trade approach as a “smart” system of Pigouvian taxation, since foranyindividual bankthepermit price is identical totheoptimal tax on money creation.21
21. This is not to say that cap-and-trade is the unique way of implementing the optimal scheme. An alternative would be an iterative form of taxation: the regulator announces a trial value of the tax rate. He then observes the quantity of M chosen by banks and uses this to infer the productivity of their investment opportunities. With these data, he can then set the optimal tax rate. Thus rather
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 79
IV. IMPLEMENTING THE CAP-AND-TRADE APPROACH WITH
MONETARY POLICY
The cap-and-trade approach to bank regulation may seem alien—it does not have any direct counterpart in the real world. However, I argue that the cap-and-trade approach can be imple- mented with something that looks very much like conventional monetary policy—with open market operations in which the cen- tral bank adjusts the quantity of nominal reserves in the banking system. Inthis setting, reserves playtheroleof permits formoney creation, giventheexistenceofabindingreserverequirement. And the nominal interest rate corresponds tothe price of the permits.
In drawing this analogy, one wrinkle is that I have so far been working in an entirely real economy. To introduce a central bank and a role for monetary policy, I need to bring in a set of nominallydenominatedgovernment liabilities, andthenpindown the price level. To do so, I rely on the fiscal theory of the price level (Leeper 1991; Sims 1994; Woodford 1995; Cochrane 1998). In particular, the government is assumed to issue two types of nominal liabilities: Treasury bills and bank reserves. According to the fiscal theory, the sum of these two nominal liabilities is what is relevant for determining the price level. Given the sum, the composition of these liabilities is a real variable, since only reserves canbeusedtosatisfyreserverequirements. Thus holding fixed total government liabilities, when there are more reserves, banks are able to create more money, that is, to finance a greater fraction of their operations with short-term debt. Hence, reserves correspond exactly to the concept of regulatory permits in the real model.22 By contrast, if Treasury bills could also be used to satisfy reserve requirements, there would be nothing special about reserves, and open market operations would have noeffect.
To operationalize the fiscal theory, I assume that the govern- ment anticipates real tax revenues of T at time 2, and the value of T is exogenously fixed. At time 0, the government has total nominal liabilities outstanding of l0, composed of Treasury bills b0, and bank reserves r0. Thus l0 = b0 + r0. The time 0 price level Λ0, is then determined by the requirement that the real value of
than setting quantities and learning from market prices, the regulator sets prices (taxes) and learns from market-determined quantities.
22. Since the price level is pinned down by fiscal considerations, the goal of achieving price stability cannot be the central bank’s job. Rather, the central bank is left with just the role of financial stability regulator.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
80 QUARTERLY JOURNAL OF ECONOMICS
the government’s obligations must equal the present value of its future tax revenues:
(15) l0
Λ0 =
T RM
.
Two points are worth noting here. First, the relevant real discount rate for the government is RM, given that its obligations are riskless: when households own Treasury bills, they derive the same monetary services from these bills that they do from privately created bank money, so the return on Treasury bills is equal to RM. Second, to keep real tax revenues fixed at T as the composition of government liabilities varies, I assume that thegovernment rebates anyseignoragerevenues derivedfromthe issuance of non–interest-bearing reserves in a lump-sum fashion to the household sector.23
Again, the key distinction between Treasury bills and bank reserves is that only the latter can be used to satisfy reserve requirements. In particular, any bank wishing to issue a dollar of short-term debt must hold ρ dollars of reserves, where ρ is the fractional reserve requirement. Hence the net amount of short- term debt financing made possible by $1 of reserves is (1−ρ)
ρ
dollars.24 It follows that in real terms, the total amount of M that can be created by the banking sector is now given by:
(16) M = (1− ρ)r0
ρΛ0 =
(1− ρ)T ρRM
r0
l0 .
This expression makes it clear that the ratio of r0 to l0— namely, the composition of the government’s nominal liabilities— is a real variable, and is the means by which the government can target total real money creation by banks. An open market operation that increases the supply of reserves relative to T-bills is isomorphic to an increase in the regulatory limit on M in the all-real cap-and-trade version of the model.
23. Without this assumption, the composition of government liabilities would influence real tax revenues. In particular, as the government issued more non– interest-bearing reserves and fewer interest-bearing bills, its effective tax rev- enues would go up through a seignorage mechanism. The assumption can be loosely motivatedby the idea that the government has some kindof social compact with its citizens that prevent it from letting total tax revenues—no matter how they are raised—go above T.
24. As an example, suppose ρ = 0.10. In this case, with $1 of reserves, a bank is allowed toraise $10 of short-term debt. But given that it must hold the reserves as an asset, only $9 represent net financing that is available to fund new loans.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 81
Moreover, as noted, the analog to the price of permits is the current setting is the nominal interest rate. This is because when banks want tocreate money, they are forced tohold non–interest- bearing reserves, and the nominal interest rate represents the opportunity cost of doing so.
Denoting the nominal interest rate by i, one can express the time 2 price level as:
(17) Λ2 = Λ0(1 + i)
RM .
Now suppose a bank wishes to increase its net issuance of real M by one unit at time 0, thereby increasing its real time 2 profits by dΠ
dM . To do so, it must increase net nominal M by Λ0
units, which requires it tohold ρΛ0/(1−ρ) of nominal reserves. To finance these reserve holdings, it must pay ρiΛ0/(1−ρ) of nominal financing costs at time 2. The real time 2 value of these financing costs is therefore ρiΛ0
(1−ρ)Λ2 or, using equation (17), ρiRM
(1−ρ)(1+i). For a bank to be indifferent, it must be that these real costs are equal to dΠ
dM . Thus it follows that the nominal interest rate is given by:
(18) i
(1 + i) =
(1− ρ) ρRM
dΠ dM
.
EXAMPLE 3. Keep everything the same as in Example 1: f (I) = ψlog(I)+ I, g(K)= θlog(K), RB =1.04; RM =1.01; ψ=3.5; θ=150; λ=1; W=140; andp=0.98. At thesocial optimumof M∗∗=55.2, we had that dΠ
dM = 0.0056. With a reserve requirement of ρ = 0.10, if this optimum is implemented with monetary policy, the nominal interest rate is given by i = 5.25%. (Since the nominal rate exceeds the riskless real rate of 1.0%, the implied rate of inflation between time 0 and time 2 is 4.25%.) Ifwekeepall elsethesamebut set RM=1.02, thenewoptimum involves M∗∗ = 52.5, which is implemented with a nominal rateof i=1.81%. Intuitively, as thespreadbetweenmoneyand bonds shrinks, banks have a weaker desire to create private money. So the nominal interest rate, which is equivalent to the value of a permit for money creation, falls as well.
V. OTHER POLICY TOOLS
V.A. Liquidity Regulation
I have thus far taken the time 0 liquidity stockpile W of the PIs to be exogenous. This does not affect any of the conclusions
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
82 QUARTERLY JOURNAL OF ECONOMICS
in the foregoing analysis regarding the socially optimal quantity of money, because these conclusions hold for any value of W such that there is a scarcity of PI resources in the bad state at time 1. However, I now pose two related questions about W. First, if the PIs are allowed to choose W optimally, what value will they pick? Second, if the social planner is allowed to choose W, will his choicedifferfromthat of thePIs?Inotherwords, is thereacasefor regulationof liquidityholdings, inadditiontoregulationof money creation?
The privately optimal choice of W, denoted by W∗, is deter- mined by the following first-order condition:
(19) pg′(W) + (1− p)g′(W −M) = RB.
The logic is straightforward. PIs raise W at time 0, paying a gross interest rate of RB.25 With probability p, the good state ensues, and the marginal return on their investment is g′(W). With probability (1 − p), the bad state ensues, and the marginal return on investment is g′(W − M). One interesting feature of this solution is that the more unlikely the bad state, the lower the equilibrium value of W∗, and the deeper the fire-sale discount when the bad state does in fact occur.
To solve for the socially optimal value of W, denoted by W∗∗, we can return tothe planner’s Lagrangian from equation (11) and take the first-order condition with respect to W, which yields:
(20) pg′(W) + (1− p)g′(W −M) = RB + ηP λg′′( ∙ ) RM(g′( ∙ ))2
.
Comparing equations (19) and (20), we can see that the pri- vateandsocial solutions onceagaindivergeonlywhenηP > 0, that is, when the collateral constraint binds. Moreover, when this does happen, the additional term in equation (20), ηP λg′′(∙)
RM(g′(∙))2 , is nega- tive, meaning that the planner prefers a lower marginal product of W, oralternatively, a higherlevel of W. Thus optimal regulation takes the form of a floor on liquidity holdings by the PIs.26
25. The assumption that the PIs’ cost of capital is RB, rather than RM , is tantamount tosayingthat theyareunabletoissueanyriskless debt. This wouldbe the case if their production technology g(K) were risky, and had some probability of delivering zero output at time 2. However, even if the PIs could issue riskless claims, the rest of the analysis would be little changed, since the return on these claims is already pinned down independent of their quantity. Thus the marginal appeal to the banks of issuing short-term debt against their long-term assets is unaffected if the PIs do some additional riskless financing on the side.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 83
This result connects to Farhi, Golosov, and Tsyvinski (2009), who also develop a rationale for liquidity regulation. However, the mechanism in FGT is quite different. Following Diamond and Dybvig (1983), Jacklin (1987), Bhattacharya andGale (1987), and AllenandGale (2004), theymodel banks as providers of insurance to consumers with unpredictable liquidity needs. As this litera- ture has shown, incentive-compatible insurance can be frustrated by the existence of securities markets, since “late” consumers may be tempted to mimic “early” consumers by withdrawing their money from the bank prematurely and reinvesting it at the market rate of interest. The insight of FGT is that liquidity requirements can be used to depress the security market rate, thereby reducing the temptation for late consumers to withdraw early. By contrast, in my model, real rates are pinneddown by the linear preferences of households and thus unaffected by liquidity requirements. Instead, the rationale for a liquidity requirement reflects a desire to reduce the equilibrium fire-sale discount.27
Although liquidity requirements arise naturally in my frame- work, there are a couple of caveats. First, the liquidity require- ments envisioned by the theory may be difficult to enforce. To implement them efficiently, they have to be imposed on PIs at time 0, in proportion to the scale of each PI’s time 1 investment opportunities. But a regulator may not know at time 0 what the distribution of time 1 projects across PIs looks like. By contrast, the monetary regulation does not face this enforcement problem, since short-term debt issuance is contemporaneously observable at time 0.
Second, in the limited set of numerical experiments that I have tried, the planner’s utility gain from imposing liquidity regulationturns out tobemuchsmallerthanthat fromregulating the creation of private money. Combined with the enforcement problem, this helps explain why the primary focus of the analysis
26. As emphasized, liquidity requirements can never obviate the need for regulation of money creation. One way to show this formally is to note that according to equations (19) and (20), liquidity regulation is only ever worth using when the collateral constraint binds in equilibrium. However, as we have seen, when the collateral constraint binds, it is always desirable to regulate money creation.
27. At a more abstract level, the twomodels are similar in the following sense. In both models there is an additional constraint beyond the budget constraint: an incentivecompatibilityconstraint inFGT, anda collateral constraint inmymodel. Moreover, in both models, a market price (the interest rate, or the fire-sale price) enters into the constraint; this is what motivates the planner to intervene, in an effort to change the market price and thereby relax the constraint.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
84 QUARTERLY JOURNAL OF ECONOMICS
in this article has been on the latter. The following example is illustrative of the magnitudes that arise.
EXAMPLE 4. Keep everything the same as in Example 1, except allow W to be chosen endogenously: f (I) = ψlog(I) + I, g(K) = θlog(K), RB =1.04; RM =1.01; ψ=3.5; θ=150; λ=1; and p=0.98. The PIs’ optimal choice of W is given by W∗ = 146.31, whereas the social optimum is given by W∗∗ = 147.04. Compared to a benchmark case with no regulation at all, the following regulatoryconfigurations producetheseincreases intheplan- ner’s utility: (i) regulationonlyof moneycreation: +0.0148; (ii) regulation of both money creation and liquidity: +0.0167; and (iii) regulationof just liquidity: +0.0014. Thus inthis example, the benefit of liquidity regulation is approximately one-tenth that which comes from regulating money creation.
V.B. Deposit Insurance and Lender of Last Resort
In the baseline version of the model, the only way for banks to pay off their short-term creditors in the crisis state is by fire- selling their assets, and the only role for policy is to control the amount of short-term debt that is created ex ante. An alternative approachwouldbeforthegovernment totrytostemtheamount of sociallycostlyfiresales that occurfora given amount ofshort-term bankdebt. This couldbedoneeitherwitheitherdeposit insurance or a lender-of-last-resort policy.
Unlike in the classic framework of Diamond and Dybvig (1983), such policies are not costless to the government in equi- librium, because here, in the crisis state, there is a probability (1−q) that the banks’ assets will turn out tobe entirely worthless. So there is always a chance that taxpayers will be left on the hook. If taxpayer-financed bailouts create deadweight losses, the overall optimum set of policies may have the realistic feature that: (1) some fraction of banks’ money-like claims are insured by the government; (2) the remainder are uninsured, and hence still subject to fire-sale risk; and (3) as before, it makes sense for the regulator to control the total quantity of bank-created money.
To see this explicitly, consider a case where the deadweight costs of taxationtakethefollowingform: there is nocost toraising any amount less than L to pay for a bailout, but it is infinitely costly to raise anything more than L. It follows that the amount of government-insured money that can be created, MI, is bounded by MI ≤ L, and it will in fact always be optimal to set MI = L.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 85
Note, too, that if thegovernment offers insuranceonsomeamount of bank deposits, it will have to put in place a rule to prevent banks from selling all of their assets in a crisis state to satisfy the demands of uninsured depositors; otherwise banks will create just as much uninsured money as before, and the deposit insurer will always be left holding an empty shell in the crisis state. A simple version of such a rule—which can effectively be thought of as a ban on fraudulent conveyance—is a requirement that the fraction of assets sold in a crisis, Δ, not exceed the relative proportion of uninsured deposits. Thus the requirement that goes along with insurance is thatΔ ≤ MU
MU+MI , where MU is the quantity of uninsured money created by the bank.
It follows that the total amount of money—insured plus uninsured—that can be created must satisfy the same collateral constraint as before: M = MU + MI ≤ kλI. The only thing that is changed is the determination of the fire-sale discount k. Since insured depositors are protected and do not need to demand repayment at time 1, only uninsured deposits give rise to fire sales. Thus k is now given by:
(21) 1 k
= g′(W −MU) = g′(W −M + L).
In other words, the outcome in a world with limited deposit insurance is equivalent to that in a world with no deposit insurance, but where the wealth of the PIs is augmented from W to (W + L). A given amount of total money creation now causes less fire-sale damage, and as a result, more money can be created in equilibrium.
Equation (21) also makes clear the close connection between deposit insurance and a lender-of-last-resort function. Given that thegovernment canneverput itself ina positiontolosemorethan L, an alternative to deposit insurance would be for it to leave all deposits uninsuredbut tocommit tostepinandinvest L alongside thePIs intheevent ofafiresale. This wouldhaveexactlythesame effect—it wouldreducethefire-salediscount perequation(21)and thereby allow for more total money creation.
The bottom line is that one can add deposit insurance to the model in such a way as to make it more realistic, without changing any of its qualitative properties. The optimal policy mix will involve limited use of deposit insurance or, equivalently, limiteduse of a lender-of-last-resort function. Banks will continue to issue uninsured money-like claims alongside insured deposits,
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
86 QUARTERLY JOURNAL OF ECONOMICS
and hence will continue to create some degree of fire-sale risk. Thus, as before, there will continue to be a motive for regulating the creation of these uninsured short-term claims.
V.C. Regulating the Shadow Banking System
The model also assumes that all private money is manu- factured by commercial banks that are subject to reserve re- quirements. Hence, private money creation can be completely controlled by conventional open market operations. Though this may be an adequate representation of an earlier periodin history, it omits an important form of money creation in the modern econ- omy. As Gorton andMetrick (2011) and Gorton (2010) emphasize, privatemoney—inpreciselythesensemeant here—is alsocreated by the unregulated shadow banking system, via the large volume of short-term claims that are collateralized by securitized loan pools of one form or another.
This observation suggests that commercial banks and shadow banks should be regulated in a symmetric fashion. According to the logic of the model, the ideal way to do this would be to broaden the reach of reserve requirements, so that the cap-and-trade regime covers all the short-term liabilities of both commercial banks and shadow banks. If, due to some political constraint outside the model, the liabilities of shadow banks cannot be subjected to reserve requirements, an alternative approach might be toimpose a regime of “haircut” regulation. For example, the central bank could specify the maximum fraction of short-term financing that could be issued against a given amount of collateralizable assets. Moreover, just as the optimal quantity of bank-created money M∗∗ varies with economic conditions, optimal haircuts would respond to these conditions as well. The Appendix provides a brief analysis of haircut regulation. It turns out that although such regulation is indeed useful, it is strictly less efficient than direct control of the quantity of privately created money via, for example, the sort of reserve requirements–based mechanism already outlined.
V.D. Government Debt Maturity
As we have seen, the magnitude of the externality associated with private money creation is related to the bond-money spread (RB − RM): when the spread widens, the wedge between the social and private returns to money creation goes up. Thus an alternative way tomoderate the externality would be tocompress
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 87
thespread. Inthecurrent versionof themodel this is impossible— given the assumption of linear preferences, the spread is exoge- nously fixed and insensitive to asset supplies.
However, if one changes the model so that the monetary ser- vices enjoyedbyhouseholds area concavefunctionof thesupplyof money—that is, there is diminishing marginal utility of money— then it becomes possible for the government to act on the bond- money spread. For example, since short-term Treasury bills are riskless, they can provide the same monetary services as short- term bank debt. Hence, an increase in the supply of Treasury bills will, in this modified setting, reduce the bond-money spread.
One appeal of dealing with the externality in this fashion is that unlike some other regulatory approaches, it does not invite evasion. For example, if the scope of reserve requirements were broadened, private actors might try to get around limits on their ability to use short-term debt by using various forms of hidden borrowing, forexample, byembeddingtheborrowinginanopaque derivative contract. In contrast, when the relative cost of short- term borrowing goes up—because the market has been saturated with riskless short-term claims—the incentive to create private money is blunted.
In Greenwood, Hanson, and Stein (2010), we use this obser- vation as the point of departure for a normative theory of govern- ment debt maturity. We argue that the government shouldchoose a shorter debt maturity—and in particular, should issue more riskless T-bills—than it otherwise might, in an active effort to crowd out the short-term debt of financial intermediaries. The ar- gument is based on a principle of comparative advantage. On the one hand, tilting its issuance toward short-term debt is not with- out cost for the government, since with stochastic interest rates this increases the variability of future interest payments andulti- matelydisrupts efforts tosmoothtaxrates overtime. Ontheother hand, short-term government debt, unlike the short-term debt of financial intermediaries, does not create fire-sale risk. To the extent that the fire-sale externality is more costly to the economy at the margin than the disruption of tax smoothing, it can make sense for the government to take on a bigger role in providing the short-term riskless claims that the economy demands.28
28. Tothe extent that monetary services reflect an ability totransact between time 0 and 1 without threat of adverse selection, the relevant notion of risk is short-horizon risk—that is, the potential for loss between time 0 andtime 1. While long-termTreasuries offercertainultimatepayoffs, theyarenot riskless overshort
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
88 QUARTERLY JOURNAL OF ECONOMICS
Of course, precisely because of tax-smoothing considerations, it will not generally be optimal for the government to tilt so strongly toward short-maturity issuance as to entirely eliminate the bond-money spread in equilibrium. Rather, optimal behavior bythegovernment onthis dimensionwill typicallyinvolveleaving the spread only partially compressed. Although government debt maturity may be one helpful tool in addressing the problem of excessive private money creation, it is not a panacea, and it is unlikely to eliminate the usefulness of the other tools.
V.E. Interest on Reserves
I have thus far assumed that the price level is determined outside the central bank, by the fiscal theory mechanism. Though this is a convenient modeling device, it is not an essential piece of the story. An alternative approach, in the New Keynesian spirit, would be to model prices as being anchored by the central bank’s adherence toa “Taylorrule”(Taylor1993, 1999) which dictates its path for the short-term nominal rate.
However, this raises a potential problem of there being more objectives than tools. If the short-term nominal rate must satisfy a Taylor rule to maintain price stability, how can it also satisfy equation (18), which specifies its optimal value from a regulatory perspective? One way out of this box is via the payment of interest on reserves (IOR), which many central banks around the world have been doing for years, and which the U.S. Federal Reserve first took up in October 2008. As Goodfriend (2002) points out, with IOR, there are two distinct methods for raising short-term nominal rates: byincreasingtheinterest paidonreservebalances, or by draining reserves from the system, thereby increasing their scarcityvalue. Thesemethods arenot equivalent, becauseonlythe latter scarcity-based approach increases the effective “reserves tax” paid by banks, which has been the focus of the foregoing analysis.
Buildingonthis observation, KashyapandStein (2012) argue that IOR allows the central bank to simultaneously accomplish two goals: (1) set the short-term nominal rate in accordance with a Taylor rule, and (2) implement an optimal regulatory scheme
horizons if interest rates are stochastic. Hence, they can create adverse-selection problems in trade if one party to a transaction has a better ability to forecast changes in rates than the other.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 89
of the sort described in this article.29 They note that in a regime with IOR, one can decompose the nominal federal funds rate f as follows:
(22) f = rIOR + ySVR,
where rIOR is the level of interest paid on reserves, and the ySVR is the quantity-mediated scarcity value of reserves. The latter term corresponds exactly tothe variable i in equation (18), as it reflects the opportunity cost to a bank of holding reserves.
For example, suppose that an analysis of the sort suggested by equation (18) yields the conclusion that for regulatory pur- poses, the optimal value of i (or equivalently, of ySVR) is 2.0%, whereas an application of the Taylor rule implies that the optimal value of f is 5.0%. In this case, the central bank should set rIOR to 3.0%, and then adjust the quantity of reserves in the system until f equilibrates at 5.0%.
VI. A DISTINCTIVE ACCOUNT OF THE MONETARY TRANSMISSION
MECHANISM
Much of the discussion has focused on the normative im- plications of the model. But the model is also of interest as a positive account of the monetary transmission mechanism. Three of its properties are particularly noteworthy in this regard. First, monetary policy has real effects even though all prices are per- fectly flexible. Second, monetary policy works entirely through a quantitative effect on bank lending. That is, the real rates on both moneyandbonds arefixedandindependent of thestanceof policy; an easing of policy impacts bank lending only because it enables banks to use more of the former, relatively cheaper funding source. This is a pure version of the bank lending channel, and as such helps explain how monetary policy can have important real effects evenwhenit does not movelong-termopenmarket interest rates by much, or when firm investment is not very responsive to such open market rates.
Third, the model has the property that the central bank does not lose control of monetary policy when other, nonreservable forms of money are introduced. Consider what happens if there is, in addition to the risky production technology already in
29. See Woodford (2011) for a more complete treatment of these issues in a dynamic New Keynesian model.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
90 QUARTERLY JOURNAL OF ECONOMICS
the model, a safe storage technology. Claims to this technology are riskless, and hence circulate as an alternative transactions medium alongside bank-created money, bearing the same gross interest rate of RM. They are also not subject to reserve requirements. (To be more concrete, one can interpret these claims as money market fund deposits backed by Treasury bills.) Even if the volume of these claims is large, nothing in the model changes. All real rates are already pinned down by the linearity of householdpreferences andare therefore unaffectedby the total quantity of money in circulation.
The distinctive feature of the model in this regard is that the central bank’s ability to influence real outcomes derives not from its control over the total quantity of transactions facilitating claims available to households, but from the fact that it is the unique provider of permits that allow banks to issue short-term debt andhencefinancethemselves morecheaply. Simplyput, only central bank–provided reserves can be used to satisfy the reserve requirements that constrain short-term debt issuance by banks. This “permits” aspect of monetary policy is also emphasized in Stein (1998), though the model in that paper differs significantly on other dimensions.30
VII. CONCLUSIONS
The basic message of this article can be summarized as fol- lows. Banks and other financial intermediaries like tofund them- selves with short-term debt. With sufficient collateral backing it, this short-term debt can be made into riskless money, which, because of the transactions services it generates, represents a cheap source of finance for banks. While society benefits from this private money creation, banks’ private incentives lead them to overdo it, since they do not fully internalize the fire-sales costs that are a by-product of their maturity transformation activities. The externality associated with excessive private money creation provides a fundamental rationale for financial stability regula- tion, and arguably, for the existence of central banks.
In a sufficiently simple institutional environment, the ex- ternality can be addressed with conventional monetary policy,
30. In Stein (1998), reserves are effectively permits that allowbanks toaccess the deposit insurance fund. Because banks face an adverse selection problem in raisinguninsuredfinance, anincreaseinthequantityofreserves canmovelending closer to the first-best level.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 91
complemented by either deposit insurance or a lender-of-last- resort facility. Indeed, this is one interpretation of what central banks have done for much of their history. In a more realistic modern-day setting, where a substantial shadow banking sector exists alongsidetraditional commercial banks, othertools, suchas expanded reserve requirements, or haircut regulation, may also benecessary. If so, central banks shouldnot bereluctant todeploy these tools—to the extent that they do so in an effort to contain excessive private money creation, they can be said to be pursuing oneof theirtraditional coremissions inamorecomprehensiveand effective manner.
APPENDIX
A. A Variant of the Model with Imperfect Pledgeability
As noted in the text, the result that there is no externality in the lowM region when m < mmax is dependent on the assumption that whenthePIs invest inreal projects, theycaptureall thesocial surplus associatedwith these projects. An alternative approach is toassumethat thesocial returntoa project financedbya PI is still given by g(K), but that onlyϕg(K) can be pledgedtothe PI. In this case, the equilibrium determination of k in equation (4) is altered so that 1
k = ϕg′(W −M). Equation (7), the bank’s first-order condition with respect to
m, still holds as stated. If the collateral constraint is not binding, so that η = 0, this condition reduces to:
(23) (RB − RM)− (1− p)zRM = 0.
However, the planner’s first-order condition for m in equation (12) is nowmodified, becausewecannolongersubstitute 1
k =g′(W−M). Instead, if ηP = 0 this condition can be written as:
(24) (RB − RM)− (1− p)zRM − (1− p)(1− ϕ)g′(W −M)RM = 0.
Thus even in the low spread region where m < mmax and I = IB, there is now a wedge of (1 − p)(1 − ϕ)g′(W − M)RM
between the private andsocial first-order conditions. This implies that the optimal price of permits will now be strictly positive in this region. Alternatively, in the monetary policy implementation of the optimum, the nominal interest rate will now be strictly positive for all parameter values.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
92 QUARTERLY JOURNAL OF ECONOMICS
Another noteworthy feature of this version of the model is that it implies different comparative statics than the baseline model with respect to the ex ante probability of a financial crisis, as captured by (1 − p). Here, if we are in the low-spread region, an increase in (1 − p) increases the wedge, and hence raises the optimal value of the permit price P, or equivalently, the nominal interest rate. By contrast, in the baseline model with perfect pledgeability, equation (14) says that an increase in (1−p) lowers the desired permit price. Intuitively, the difference is that in the baselineversionofthemodel, banks doabetterjobof internalizing thesocial costs offiresales. Indeed, whentheriskofafiresalegoes up, banks become sufficiently more cautious about using short- termdebt that theybecomebetteralignedwiththesocial planner, which in turn implies that there is less need to rein them in by raising permit prices/interest rates. However, with imperfect pledgeability, there is an effect in the opposite direction, because banks tend tounderweight the social costs of fire sales even when the collateral constraint is not binding.
B. Haircut Regulation
To see the effects of haircut regulation most transparently, consider the imperfect pledgeability version of the model just described. Suppose that we are in a “shadow banking” economy where all else is the same as before, with one exception: it is impossible to regulate the absolute quantity of privately created money M directly—say, because shadow banks cannot be sub- jected to reserve requirements—but it is possible to impose a cap mcap < mmax on the fraction of investment that is money financed.
It turns out that this form of haircut regulation, though useful, is a second-best means of intervention as compared to controlling the aggregate quantity of money. This is because the social costs of fire sales are a function of M, so this is the item the planner would ideally like to control. Trying to do this indirectly, by picking a value of mcap, will nowhave the undesired side effect of encouraging shadow banks to raise their investment above the optimal level of IB. (I assume that we are in the low spread region of the parameter space, so that absent haircut regulation, shadowbanks would choose I = IB.) Intuitively, haircut regulation always gives shadowbanks theoptiontocreatemorecheapmoney financing at the margin, as long as they are willing to raise the level of investment.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 93
This can be seen formally by considering the first-order con- dition with respect to I for a shadow bank facing binding haircut regulation:
(25) dΠ dI
=pf ′(I)+(1−p)λ−RB +mcap{(RB−RM)−(1−p)zRM}=0.
It follows that it is impossible to use haircut regulation to im- plement the social optimum described in equation (24). For if equation (24) is satisfied with I = IB, it must be that (RB − RM)− (1−p)zRM = (1−p)(1−φ)g′(W−M)RM > 0. But then for equation (25) to be satisfied, that is, for the shadow bank to be optimizing given the haircut constraint, we require pf ′(I)+ (1−p)λ−RB < 0, which means that I > IB.
HARVARD UNIVERSITY AND NATIONAL BUREAU OF
ECONOMIC RESEARCH
REFERENCES
Adrian, Tobias, and Hyun Song Shin, “Financial Intermediaries, Financial Stabil- ity and Monetary Policy,” in Maintaining Stability in a Changing Financial System, 287–334 (Kansas City: Federal Reserve Bank of Kansas City, 2008).
Allen, Franklin, and Douglas Gale, “Financial Intermediaries and Markets,” Econometrica, 72 (2004), 1023–1061.
———, “From Cash-in-the-Market Pricing to Financial Fragility,” Journal of the European Economic Association, 3 (2005), 535–546.
Ashcraft, Adam, NicolaeGarleanu, andLasseHejePedersen, TwoMonetaryTools: Interest Rates and Haircuts, Working Paper, New York University, 2010.
Bagehot, Walter, Lombard Street: A Description of the Money Market, (New York: Scribner’s, 1873).
Bernanke, Ben S., and Alan Blinder, “Credit, Money, and Aggregate Demand,” American Economic Review, 78 (1988), 435–439.
———, “The Federal Funds Rate and the Channels of Monetary Transmission,” American Economic Review, 82 (1992), 901–921.
Bhattacharya, Sudipto, and Douglas Gale, “Preference Shocks, Liquidity and Central Bank Policy, in New Approaches to Monetary Economics, W. Barnett and K. Singleton, eds. (Cambridge: Cambridge University Press, 1987).
Brunnermeier, Markus K., andLasseH. Pedersen, “Market LiquidityandFunding Liquidity,” Review of Financial Studies, 22 (2009), 2201–2238.
Caballero, Ricardo J., and Arvind Krishnamurthy, “Excessive Dollar Debt: Finan- cial Development and Underinsurance,” Journal of Finance, 58 (2003), 867– 893.
Caballero, Ricardo J., and Alp Simsek, Fire Sales in a Model of Complexity, Working Paper, MIT, 2010.
Cochrane, John, “A Frictionless Model of U.S. Inflation,” NBER Macro Annual, 13 (1998), 323–384.
Dang, Tri Vi, Gary B. Gorton, and Bengt Holmstrom, Opacity and the Optimality of Debt for Liquidity Provision, Working Paper, Yale University, 2009.
DeMarzo, PeterM., “ThePoolingandTranchingof Securities: A Model of Informed Intermediation,” Review of Financial Studies, 18 (2005), 1–35.
DeMarzo, Peter M., and Darrell Duffie, “A Liquidity-Based Model of Security Design,” Econometrica, 67 (1999), 65–99.
Diamond, Douglas W., “Financial Intermediation and Delegated Monitoring,” Review of Economic Studies, 51 (1984), 393–414.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
94 QUARTERLY JOURNAL OF ECONOMICS
———, “Debt Maturity Structure and Liquidity Risk,” Quarterly Journal of Eco- nomics, 106 (1991) , 709–737.
Diamond, Douglas W., and Philip Dybvig, “Bank Runs, Deposit Insurance and Liquidity,” Journal of Political Economy, 91 (1983), 401–419.
Diamond, Douglas W., and Raghuram G. Rajan, “Liquidity Risk, Liquidity Cre- ation, and Financial Fragility: A Theory of Banking,” Journal of Political Economy, 109 (2001), 287–327.
———, Fear of Fire Sales and the Credit Freeze, Working Paper, University of Chicago, 2009.
Duffie, Darrell, “Presidential Address: Asset Price Dynamics with Slow-Moving Capital,” Journal of Finance, 65 (2010), 1237–1267.
Farhi, Emmanuel, Mikhail Golosov, and Aleh Tsyvinski, “A Theory of Liquidity and Regulation of Financial Intermediation,” Review of Economic Studies, 76 (2009), 973–992.
Flannery, Mark J., “Asymmetric Information and Risky Debt Maturity Choice,” Journal of Finance, 41 (1986), 19–37.
Fostel, Ana, and John Geanakoplos, “Leverage Cycles and the Anxious Economy,” American Economic Review, 98 (2008), 1211–1244.
Geanakoplos, John, “The Leverage Cycle,” NBER Macro Annual, 24 (2010), 1–65. Geanakoplos, John, and Heraklis Polemarchakis, “Existence, Regularity and Con-
strained Suboptimality of Competitive Allocations When the Asset Market Is Incomplete,” in Essays in Honor of Kenneth Arrow, vol. 3, W. Heller, R. Starr, and D. Starrett, eds., 65–95 (Cambridge: Cambridge University Press, 1986).
Gersbach, Hans, “Liquidity Creation, Efficiency, and Free Banking,” Journal of Financial Intermediation, 7 (1998), 91–118.
Goodfriend, Marvin, “Interest on Reserves and Monetary Policy,” Federal Reserve Bank of New York Economic Policy Review, 8 (2002), 13–29.
———, “How the World Achieved Consensus on Monetary Policy,” Journal of Economic Perspectives, 21 (2007), 47–68.
Goodhart, Charles, The Evolution of Central Banks (Cambridge, Mass.: MIT Press, 1988).
Gorton, Gary B., Slapped by the Invisible Hand: The Panic of 2007 (New York: Oxford University Press, 2010).
Gorton, GaryB., andAndrewMetrick, “SecuritizedBankingandtheRunonRepo,” Journal of Financial Economics, forthcoming, 2011.
Gorton, Gary B., and George Pennacchi, “Financial Intermediaries and Liquidity Creation,” Journal of Finance, 45 (1990), 49–72.
Greenwald, Bruce, andJosephE. Stiglitz, “Externalities inEconomies withImper- fect Information and Incomplete Markets,” Quarterly Journal of Economics, 101 (1986), 229–264.
Greenwood, Robin, Samuel Hanson, and Jeremy C. Stein, A Comparative Ad- vantage Approach to Government Debt Maturity, Working Paper, Harvard University, 2010.
Gromb, Denis, and Dimitri Vayanos, “Equilibrium and Welfare in Markets with Financially Constrained Arbitrageurs,” Journal of Financial Economics, 66 (2002), 361–407.
Hart, Oliver, andJohnMoore, “Debt andSeniority: AnAnalysis of theRoleof Hard Claims in Constraining Management,” American Economic Review, 85 (1995), 567–585.
Hart, Oliver, and Luigi Zingales, Inefficient Provision of Inside Money, Working Paper, Harvard University, 2011.
Hellwig, Martin, “LiquidityProvision, Banking, andtheAllocationof Interest Rate Risk,” European Economic Review, 38 (1994), 1363–1389.
Jacklin, Charles J., “Demand Deposits, Trading Restrictions, and Risk Shar- ing,” in Contractual Arrangements for Intertemporal Trade, E. Prescott and N. Wallace, eds., 26–47 (Minneapolis: University of Minnesota Press, 1987).
Jeanne, Olivier, and Anton Korinek, Managing Credit Booms and Busts: A Pigouvian Taxation Approach, Working Paper, University of Maryland, 2010.
Kashyap, Anil K, andJeremy C. Stein, “What Doa Million Observations on Banks Sayabout theTransmissionofMonetaryPolicy?,”American Economic Review, 90 (2000), 407–428.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
MONETARY POLICY AS FINANCIAL STABILITY 95
———, “Cyclical Implications of the Basel-II Capital Standards,” Federal Reserve Bank of Chicago Economic Perspectives, 28, Q1 (2004), 18–31.
———, “The Optimal Conduct of Monetary Policy with Interest on Reserves,” American Economic Journal: Macroeconomics, forthcoming (2012).
Kashyap, Anil K, Jeremy C. Stein, and David Wilcox, “Monetary Policy and Credit Conditions: Evidence from the Composition of External Finance,” American Economic Review, 83 (1993), 78–98.
Kiyotaki, Nobuhiro, and John Moore, “Credit Cycles,” Journal of Political Econ- omy, 105 (1997), 211–248.
Kocherlakota, Narayana, “Taxing Risk and the Optimal Regulation of Financial Institutions,”EconomicPolicyPaper10-3, Federal ReserveBankofMinneapo- lis, 2010.
Krishnamurthy, Arvind, and Annette Vissing-Jorgensen, The Aggregate Demand for Treasury Debt, Working Paper, Northwestern University, 2010.
Leeper, Eric, “Equilibria under Active and Passive Monetary Policies,” Journal of Monetary Economics, 27 (1991), 129–147.
Lorenzoni, Guido, “Inefficient Credit Booms,” Review of Economic Studies, 75 (2008), 809–833.
Modigliani, Franco, and Merton H. Miller, “The Cost of Capital, Corporation Finance and the Theory of Investment,” American Economic Review, 48 (1958), 261–297.
Morris, Stephen, andHyun Song Shin, “Liquidity Black Holes,” Review of Finance, 8 (2004), 1–18.
Myers, Stewart C., “Determinants of Corporate Borrowing,” Journal of Financial Economics, 5 (1977), 147–175.
Perotti, Enrico, and Javier Suarez, Regulation of Liquidity Risk, Working Paper, University of Amsterdam, 2010.
Shleifer, Andrei, and Robert W. Vishny, “Liquidation Values and Debt Capacity: A Market Equilibrium Approach,” Journal of Finance, 47 (1992).
———, “The Limits of Arbitrage,” Journal of Finance, 52 (1997), 35–55. ———, “Unstable Banking,” Journal of Financial Economics, 97 (2010), 306–318. Sidrauski, Miguel, “Rational Choice and Patterns of Growth in a Monetary Econ-
omy,” American Economic Review, 57 (1967), 534–544. Sims, Christopher A., “A Simple Model for the Determination of the Price Level
andtheInteractionof MonetaryandFiscal Policy,” Economic Theory, 4 (1994), 381–399.
Stein, Jeremy C., “An Adverse-Selection Model of Bank Asset and Liability Man- agement with Implications for the Transmission of Monetary Policy,” RAND Journal of Economics, 29 (1998), 466–486.
———, “WhyAreMost Funds Open-End?CompetitionandtheLimits ofArbitrage,” Quarterly Journal of Economics, 120 (2005), 247–272.
———, “Presidential Address: Sophisticated Investors and Market Efficiency,” Journal of Finance, 64 (2009), 1517–1548.
Taylor, John B., “Discretion versus Policy Rules in Practice,” Carnegie-Rochester Series on Public Policy, 39 (1993), 195–214.
———, “AnHistorical Analysis ofMonetaryPolicyRules,”in Monetary Policy Rules, John B. Taylor, ed. (Chicago: University of Chicago Press, 1999).
Tucker, Paul, “The Repertoire of Official Sector Interventions in the Financial Aystem: Last Resort Lending, Market-Making, andCapital,”remarks at Bank of Japan International Conference, May 27–28, 2009.
Weitzman, Martin L., “Prices vs. Quantities,” Review of Economic Studies, 41 (1974), 477–491.
Woodford, Michael, “Price Level Determinacy without Control of a Monetary Aggregate,” Carnegie-Rochester Conference Series on Public Policy, 43 (1995), 1–46.
———, Monetary Policy and Financial Stability, lecture notes, Columbia Univer- sity, 2011.
at B angor U
niversity on February 6, 2015 http://qje.oxfordjournals.org/
D ow
nloaded from
- Introduction
- A Model of Private Money Creation
- Households
- Banks
- Patient Investors
- The Bank's Optimization Problem
- Socially Excessive Money Creation: A Role for Regulation
- The Social Planner's Problem
- Understanding the Nature of the Externality
- A ``Cap-and-Trade'' Approach to Bank Liquidity Regulation
- Relationship to Pigouvian Taxation
- implementing the cap-and-trade approach with monetary policy
- Other Policy Tools
- Liquidity Regulation
- Deposit Insurance and Lender of Last Resort
- Regulating the Shadow Banking System
- Government Debt Maturity
- Interest on Reserves
- A Distinctive Account of the Monetary Transmission Mechanism
- Conclusions
- A Variant of the Model with Imperfect Pledgeability
- Haircut Regulation
Time-varying monetary-policy rules and financial stress Does financial.pdf
T i
J a
b
c
a
A R R 2 A A
J E E E
K F T M T E
1
e s l l s w a l f D
i l t 2
1 d
Journal of Financial Stability 9 (2013) 117– 138
Contents lists available at SciVerse ScienceDirect
Journal of Financial Stability
journal homepage: www.elsevier.com/locate/jfstabil
ime-varying monetary-policy rules and financial stress: Does financial nstability matter for monetary policy?
aromír Baxaa,b, Roman Horvátha,∗, Bořek Vašíčekc
Institute of Economic Studies, Charles University, Prague, Czech Republic Institute of Information Theory and Automation, Academy of Sciences of the Czech Republic, Czech Republic Czech National Bank, Czech Republic
r t i c l e i n f o
rticle history: eceived 3 January 2011 eceived in revised form 4 September 2011 ccepted 5 October 2011 vailable online 12 October 2011
EL classification: 43 52 58
a b s t r a c t
We examine whether and how selected central banks responded to episodes of financial stress over the last three decades. We employ a recently developed monetary-policy rule estimation methodology which allows for time-varying response coefficients and corrects for endogeneity. This flexible frame- work applied to the USA, the UK, Australia, Canada, and Sweden, together with a new financial stress dataset developed by the International Monetary Fund, not only allows testing of whether central banks responded to financial stress, but also detects the periods and types of stress that were the most worrying for monetary authorities and quantifies the intensity of the policy response. Our findings suggest that central banks often change policy rates, mainly decreasing them in the face of high financial stress. How- ever, the size of the policy response varies substantially over time as well as across countries, with the 2008–2009 financial crisis being the period of the most severe and generalized response. With regard to
eywords: inancial stress aylor rule onetary policy
the specific components of financial stress, most central banks seemed to respond to stock-market stress and bank stress, while exchange-rate stress is found to drive the reaction of central banks only in more open economies.
© 2011 Elsevier B.V. All rights reserved.
p s a m s l a d f
i i o a b
ime-varying parameter model ndogenous regressors
. Introduction
The recent financial crisis has intensified the interest in xploring the interactions between monetary policy and financial tability. Official interest rates were driven sharply to historical ows, and many unconventional measures were used to pump iquidity into the international financial system. Central banks pur- ued monetary policy under high economic uncertainty coupled ith large financial shocks in many countries. The financial crisis
lso raised new challenges for central bank policies, in particu- ar the operationalization of issues related to financial stability or monetary-policy decision making (Goodhart, 2006; Borio and rehmann, 2009).
This paper seeks to analyze whether and how monetary policy nterest rates evolved in response to financial instability over the
ast three decades. The monetary policies of central banks are likely o react to financial instability in a non-linear way (Goodhart et al., 009). When a financial system is stable, the interest-rate-setting
∗ Corresponding author. E-mail address: [email protected] (R. Horváth).
d n s
m
572-3089/$ – see front matter © 2011 Elsevier B.V. All rights reserved. oi:10.1016/j.jfs.2011.10.002
rocess largely reflects macroeconomic conditions, and financial tability considerations enter monetary policy discussions only to
limited degree. On the other hand, central banks may alter their onetary policies to reduce financial imbalances if these become
evere. In this respect, Mishkin (2009) questions the traditional inear-quadratic framework1 when financial markets are disrupted nd puts forward an argument for replacing it with non-linear ynamics describing the economy and a non-quadratic objective unction resulting in non-linear optimal policy.
To address the complexity of the nexus between monetary pol- cy and financial stability as well as to evaluate monetary policy n a systematic manner, this paper employs the recently devel- ped time-varying parameter estimation of monetary-policy rules, ppropriately accounting for endogeneity in policy rules. This flexi- le framework, together with a new comprehensive financial stress
ataset developed by the International Monetary Fund, will allow ot only testing of whether central banks responded to financial tress, but also quantification of the magnitude of this response
1 That is, linear behavior of the economy and a quadratic objective function of the onetary authority.
1 ncial
a w
c e t ( t f H a h t r s
e S w r
2
e ( m
2 t
l m s s
b t o c t i t ( e fi m e t c a a S 2
b
t i B a
a i s p p l r fi n i a o s A i p a s a m i r w r t c a a a c c b e r d t r c l w t w i ( e w a a s s n
i p
18 J. Baxa et al. / Journal of Fina
nd detection of the periods and types of stress that were the most orrying for monetary authorities.
Although theoretical studies disagree about the role of finan- ial instability for central banks’ interest-rate-setting policies, our mpirical estimates of the time-varying monetary-policy rules of he US Fed, the Bank of England (BoE), the Reserve Bank of Australia RBA), the Bank of Canada (BoC), and Sveriges Riksbank (SR) show hat central banks often alter the course of monetary policy in the ace of high financial stress, mainly by decreasing policy rates.2
owever, the size of this response varies substantially over time s well as across countries. There is some cross-country and time eterogeneity as well when we examine central banks’ considera- ions of specific types of financial stress: most of them seemed to espond to stock-market stress and bank stress, and exchange-rate tress drives central bank reactions only in more open economies.
The paper is organized as follows. Section 2 discusses related lit- rature. Section 3 describes our data and empirical methodology. ection 4 presents our results. Section 5 concludes. An appendix ith a detailed description of the methodology and additional
esults follows.
. Related literature
First, this section gives a brief overview of the theory as well as mpirical evidence on the relationship between monetary policy rules) and financial instability. Second, it provides a short sum-
ary of various measures of financial stress.
.1. Monetary policy (rules) and financial instability – some heories
Financial friction, such as unequal access to credit or debt col- ateralization, is recognized as having important consequences for
onetary policy transmission, and Fisher (1933) has already pre- ented the idea that adverse credit-market conditions can cause ignificant macroeconomic disequilibria.
During the last two decades, the effects of monetary policy have een studied mainly within New Keynesian (NK) dynamic stochas- ic general equilibrium (DSGE) models, which assume the existence f nominal rigidities. The common approach to incorporating finan- ial market friction within the DSGE framework is to introduce he financial accelerator mechanism (Bernanke et al., 1996, 1999), mplying that endogenous developments in credit markets work o amplify and propagate shocks to the macro economy. Tovar 2009) emphasizes that the major weakness of the financial accel- rator mechanism is that it only addresses one of many possible nancial frictions. Goodhart et al. (2009) note that many NK DSGE odels lack the financial sector completely or model it in a rather
mbryonic way. Consequently, more recent contributions within his stream of literature have examined other aspects of finan- ial friction, such as balance sheets in the banking sector (Choi nd Cook, 2004), the portfolio-choice issue with complete (Engel nd Matsumoto, 2009) or incomplete markets (Devereux and
utherland, 2007), and collateral constraints (Iacovello and Neri, 010).3
A few studies focus more specifically on the relationship etween the monetary-policy stance (or the monetary-policy rule)
2 Our choice of countries is based on data availability and on the suitability of he data for our econometric framework. Due to limited data availability, we do not nclude the Reserve Bank of New Zealand, the ECB, and emerging countries. The ank of Japan could not be included either, given that its policy rates were flat for n extended period. 3 A survey of this literature is provided by Tovar (2009).
t fi t o c t d b a a s
Stability 9 (2013) 117– 138
nd financial stability. However, they do not arrive at a unan- mous view of whether a monetary-policy rule should include ome measure of financial stability. Brousseau and Detken (2001) resent an NK model where a conflict arises between short-term rice stability and financial stability due to a self-fulfilling belief
inking the stability of inflation to the smoothness of the interest- ate path and suggests that monetary policy should react to nancial instability. Akram et al. (2007) investigate the macroeco- omic implications of pursuing financial stability within a flexible
nflation-targeting framework. Their model, using a policy rule ugmented by financial-stability indicators, shows that the gains f such an augmented rule vis-à-vis the rule without financial- tability indicators highly depends on the nature of the shocks. kram and Eitrheim (2008) build on the previous framework, find-
ng some evidence that the policy response to housing prices, equity rices or credit growth can cause high interest-rate volatility and ctually lower financial stability in terms of indicators that are ensitive to interest rates. Cecchetti and Li (2008) show, in both
static and dynamic setting, that a potential conflict between onetary policy and financial supervision can be avoided if the
nterest-rate rule takes into account (procyclical) capital-adequacy equirements, in particular, that policy interest rates are lowered hen financial stress is high. Bauducco et al. (2008) extend the cur-
ent benchmark NK model to include financial systems and firms hat require external financing. Their simulations show that if a entral bank responds to financial instability by policy easing, it chieves better inflation and output stabilization in the short term t the cost of greater inflation and output volatility in the long term, nd vice versa. For the US Fed, Taylor (2008) proposes a modifi- ation of the standard Taylor rule to incorporate adjustments to redit spreads. Teranishi (2009) derives a Taylor rule augmented y the response to credit spreads as an optimal policy under het- rogeneous loan-interest-rate contracts. He finds that the policy esponse to a credit spread can be both positive and negative, epending on the financial structure. However, he also proposes hat when nominal policy rates are close to zero, a commitment ather than a discretional policy response is the key to reducing redit spreads. Christiano et al. (2008) suggest augmenting the Tay- or rule with aggregate private credit and find that such a policy
ould raise welfare by reducing the magnitude of the output fluc- uations. Cúrdia and Woodford (2010) develop a NK DSGE model ith credit friction to evaluate the performance of alternative pol-
cy rules that are augmented by a response (1) to credit spreads and 2) to aggregate the volume of private credit in the face of differ- nt shocks. They argue that the response to credit spreads can be elfare improving, but the optimal size of such a response is prob-
bly rather small. Like Teranishi (2009), they find little support for ugmenting the Taylor rule by the credit volume, given that the ize and even the sign of the desired response is sensitive to the ources of shock and their persistence, which is information that is ot always available during operational policy making.
A related stream of literature focuses on the somewhat narrower ssue of whether or not monetary policy should respond to asset rices. Bernanke and Gertler (1999, 2001) argue that the stabiliza- ion of inflation and output provides a substantial contribution to nancial stability and that there are few, if any, gains to responding o asset prices. Faia and Monacelli (2007) extend the model devel- ped by Bernanke and Gertler (2001) by a robust welfare metric, onfirming that strict inflation stabilization offers the best solu- ion. Cecchetti et al. (2000) take the opposite stance, arguing that evelopments in asset markets can have a significant impact on
oth inflation and real economic activity, and central banks might chieve better outcomes by considering asset prices provided they re able to detect asset-price misalignments. Borio and Lowe (2002) upport this view, claiming that financial imbalances can build up
ncial
e t d p b a t R f
2 e
c w B s a C 2 t f t i r i a r r s s r p t F e m
o a s c o
u ( B F G fi t i w i C m r t H a
r
b J r b d f i t s m p i a m t s
2
b d b e a R ( s
d s f c c t r c a
i i a b i t b a h e i E A
T m a r f
J. Baxa et al. / Journal of Fina
ven in a low-inflation environment, which is normally favorable o financial stability. The side effect of low inflation is that excess emand pressures may first appear in credit aggregates and asset rices rather than consumer prices, which are normally considered y policy makers. Gruen et al. (2005) argue that responding to an sset bubble is feasible only when the monetary authority is able o make a correct judgment about the process driving the bubble. oubini (2006) and Posen (2006) provide a summary of this debate
rom a policy perspective.
.2. Monetary policy (rules) and financial instability – empirical vidence
The empirical evidence on central banks’ reactions to finan- ial instability is rather scant. Following the ongoing debate about hether central banks should respond to asset-price volatility (e.g. ernanke and Gertler, 1999, 2001; Cecchetti et al., 2000), some tudies have tested the response of monetary policy to different sset prices, most commonly stock prices (Rigobon and Sack, 2003; hadha et al., 2004; Siklos and Bohl, 2008; Fuhrer and Tootell, 008). They find some evidence either that asset prices entered he policy-information set (because they contain information about uture inflation) or that some central banks were directly trying o offset these disequilibria.4 All of these papers estimate time- nvariant policy rules, which means that they test a permanent esponse to these variables. However, it seems more plausible that f central banks respond to asset prices, they do so only when sset-price misalignments are substantial; in other words, their esponses are asymmetric. There are two additional controversies elated to the effects of asset prices on monetary-policy deci- ions. The first concerns the measure, in particular whether the tock-market index that is typically employed is sufficiently rep- esentative, or whether some other assets, in particular housing rices, should be considered as well. The second issue is related o the (even ex-post) identification of asset-price misalignment. inally, it is likely that the perception of misalignments is influ- nced by general economic conditions and that a possible response ight evolve over time. Detken and Smets (2004) summarize some stylized facts
n macroeconomic and monetary-policy developments during sset-price booms. Overall, they find that monetary policy was ignificantly looser during high-cost booms that were marked by rashes of investment and real-estate prices in the post-boom peri- ds.
A few empirical studies measure the monetary-policy response sing broader measures of financial imbalances. Borio and Lowe 2004) estimate the response of four central banks (the Reserve ank of Australia, the Bundesbank, the Bank of Japan, and the US ed) to imbalances proxied by the ratio of private-sector credit to DP, inflation-adjusted equity prices, and their composite. They nd either negative or ambiguous evidence for all countries except he USA, confirming that the Fed responded to financial imbalances n an asymmetric and reactive way, i.e., that the federal funds rate
as disproportionately lowered in the face of imbalance unwind- ng, but was not tightened beyond normal as imbalances built up. ecchetti and Li (2008) estimate a Taylor rule augmented by a easure of banking stress, in particular the deviation of leverage
atios (total loans to the sum of equity and subordinated debt;
otal assets to the sum of bank capital and reserves) from their odrick–Prescott trend. They find some evidence that the Fed djusted the interest rate to counteract the procyclical impact of a
4 A similar but somewhat less polemic debate applies to the role of exchange ates, especially for small, open economies (Taylor, 2001).
t o i m F p t
Stability 9 (2013) 117– 138 119
ank’s capital requirements, while the Bundesbank and the Bank of apan did not. Bulíř and Čihák (2008) estimate the monetary-policy esponse to seven alternative measures of financial-sector vulnera- ility (crisis probability, time to crisis, distance to default or credit efault swap spreads) in a panel of 28 countries. Their empirical ramework is different in the sense that the monetary-policy stance s proxied along the short-term interest rate by measures of domes- ic liquidity, and external shocks are controlled for. In the panel etting, they find a statistically significant negative response to any variables representing vulnerability (policy easing) but, sur-
risingly, not in country-level regressions. Belke and Klose (2010) nvestigate the factors behind the interest-rate decisions of the ECB nd the Fed during the current crisis. They conclude that the esti- ated policy rule was significantly altered only for the Fed, and
hey put forward that the ECB gave greater weight to inflation tabilization at the cost of some output loss.
.3. Measures of financial stress
The incidence and determinants of different types of crises have een typically traced in the literature by a means of narrative evi- ence (expert judgment). This has sometimes been complemented y selected indicators (exchange rate devaluation or the state of for- ign reserves) that point to historical regularities (e.g., Eichengreen nd Bordo, 2002; Kaminsky and Reinhart, 1999; Reinhart and ogoff, 2008; Laeven and Valencia, 2008). The empirical studies e.g., Goldstein et al., 2000) used binary variables that were con- tructed based on these narratives.
Consequently, some contributions strived to provide more ata-driven measures of financial stress. Most of the existing tress indices are based on high-frequency data, but they dif- er in the selected variables (bank capitalization, credit ratings, redit growth, interest rate spreads or volatility of different asset lasses), country coverage, and the aggregation method. An impor- ant advantage of continuous stress indicators is that they may eveal periods of small-scale stress that did not result in full-blown rises and were neglected in studies based on binary crisis vari- bles.
The Bank Credit Analyst (BCA) reports a monthly financial stress ndex (FSI) for the USA that is based on the performance of bank- ng shares compared to the whole stock market, credit spreads nd the slope of the yield curve, and new issues of stocks and onds and consumer confidence. JP Morgan calculates a Liquid-
ty, Credit and Volatility Index (LCVI) based on seven variables: he US Treasury curve error (the standard deviation of the spread etween on-the-run and off-the-run US Treasury bills and bonds long the entire maturity curve), the 10-year US swap spread, US igh-yield spreads, JP Morgan’s Emerging Markets Bond Index, for- ign exchange volatility (the weighted average of the 12-month mplied volatilities of several currencies), the Chicago Board of xchange VIX equity volatility index, and the JP Morgan Global Risk ppetite Index.
Illing and Liu (2006) develop a comprehensive FSI for Canada. heir underlying data cover equity, bond, and foreign exchange arkets as well as the banking sector. They use a standard measure
nd refined measure of each stress component, where the former efers to the variables and their transformations that are commonly ound in the literature, while the latter incorporates adjustments hat allow for better extraction of information about stressful peri- ds. They explore different weighting schemes to aggregate the ndividual series (factor analysis, the size of the corresponding
arket for total credit in the economy, variance-equal weighting). inally, they perform an expert survey to identify periods that were erceived as especially stressful, confirming that the FSI matches hese episodes very well.
1 ncial
a t t d a fi m i C a e i o p o a
l ( d t a v t c r t v m o e e d a h m
3
3
i i u v t 1
r e T m p u m D m i a
(
w b ( g o i p f a c b r 2 t s
a s u i t b d t d O a
b f c m f ( c
spread (the difference between short-term and long-term government bonds), TED spread (the difference between inter- bank rates and the yield on Treasury bills), banking beta
6 Borio and Disyatat (2009) characterize unconventional policies as policies that affect the central bank’s balance sheet size and composition and that can be insu- lated from interest rate policy (the so-called “decoupling principle”). One common example of such a policy (not necessarily used during times of crisis) is sterilized exchange-rate intervention. Given that we are looking not at a single episode of stress, but rather want to identify whether monetary authorities deviated from sys- tematic patterns (the policy rule) during these periods (by responding to indicators of financial stress), we need to use a consistent measure of policy action that is adjusted during periods of financial stress, though other measures may be in place as well. Therefore, we assume that the monetary-policy stance is fully reflected in the interest rate, and we are aware that it might be subject to downward bias on the financial-stress coefficient. The reader may want to interpret our results on the importance of financial stress for interest-rate setting as a conservative estimate.
7 There are other policy measures that can be used as a reactive or pre-emptive response to financial stress, such as regulatory or administrative measures, although their effects are likely to appear only in the longer term and cannot be reasonably included in our empirical analysis.
8
20 J. Baxa et al. / Journal of Fina
For the Fed Board of Governors, Carlson et al. (2009) propose framework similar to the option-pricing model (Merton, 1974) hat aims to provide the distance-to-default of the financial sys- em, the so-called Index of Financial Health. The method uses the ifference between the market value of a firm’s assets and liabilities nd the volatility of the asset’s value to measure the proximity of a rm’s assets to being exceeded by their liabilities. They apply this easure to 25 of the largest US financial institutions, confirming
ts impact on capital investments in the US economy. The Kansas ity Fed developed the Kansas City Financial Stress Index (Hakkio nd Keeton, 2009), which is published monthly and is based on leven variables (seven spreads between different bond classes by ssuers, risk profiles and maturities, correlations between returns n stocks and Treasury bonds, expected volatility of overall stock rices, volatility of bank stock prices, and a cross-section dispersion f bank stock returns) that are aggregated by principal component nalysis.
Finally, the International Monetary Fund (IMF) recently pub- ished financial stress indices for various countries. Cardarelli et al. 2011) propose a comprehensive index based on high-frequency ata where the price changes are measured with respect to heir previous levels or trend values. The underlying variables re standardized and aggregated into a single index (FSI) using ariance-equal weighting for each country and period. The FSI has hree subcomponents: the banking sector (the slope of the yield urve, TED spread, and the beta of banking-sector stocks), secu- ities markets (corporate bond spreads, stock-market returns and ime-varying volatility of stock returns) and exchange rates (time- arying volatility of NEER changes). Balakrishnan et al. (2009) odify the previous index to account for the specific conditions
f emerging economies, on the one hand including a measure of xchange rate pressures (currency depreciation and decline in for- ign reserves) and sovereign debt spread, and on the other hand ownplaying the banking-sector measures (slope of the yield curve nd TED spread).5 We will use the former index, given its compre- ensiveness as well as its availability for different countries (see ore details below).
. Data and empirical methodology
.1. The dataset
Given the frequency of monetary policy committee meetings n most central banks, we use monthly data (due to unavailabil- ty of all monthly series for a sufficiently long time period, we se quarterly data for Sweden and Canada). The sample periods ary slightly due to data availability (the US 1981:1M–2009:6M; he UK 1981:1M–2009:3M; Australia 1983:3M–2009:5M; Canada 981:1Q–2008:4Q; Sweden 1984:2Q–2009:1Q).
The dependent variable is typically an interest rate closely elated to the official (censored) policy rate, in particular the fed- ral funds rate (3M) for the USA, the discount rate (three-month reasury bills) for the UK, Canada, and Sweden, and the three- onth RBA-accepted bills rate for Australia. It is evident that the
olicy rate is not necessarily the only instrument that central banks se, especially during the 2008–2009 global financial crisis, when any unconventional measures were implemented (see Borio and isyatat, 2009; Reis, 2010). To address this issue in terms of esti-
ated policy rules, for a robustness check we use the interbank
nterest rate (at a maturity of three months). While both rates re used in empirical papers on monetary-policy rule estimation
5 The IMF Financial Stress Index has recently been applied by Melvin and Taylor 2009) to analyze exchange rate crises.
o m
g v t a v
Stability 9 (2013) 117– 138
ithout great controversy, the selection of the interest rate ecomes a more delicate issue during periods of financial stress Taylor, 2008). While the former is more directly affected by enuine monetary-policy decisions (carried out by open market perations), the latter additionally includes liquidity conditions on nterbank markets and, as such, can be affected by unconventional olicies, though these are usually insulated (often intentionally) rom policy interest rates.6 This is a drawback but also a potential dvantage of this alternative dependent variable. On the one hand, hanges in official policy rates may not pass through fully to inter- ank interest rates, in particular when the perceived counterparty isk is too high and credit spreads widen (see Taylor and Williams, 009). On the other hand, the interbank rate may also incorporate he impact of policy actions, such as quantitative easing aimed at upplying additional liquidity into the system.7
Inflation is measured as the year-on-year change in the CPI, part from for the United States, where we use the personal con- umption expenditures (PCE) price index, and Sweden, where nderlying CPIX inflation (which excludes households’ mortgage-
nterest expenditures and the direct effects of changes in indirect axes and subsidies from the CPI) is used.8 The output gap is proxied y the gap of the seasonally adjusted industrial production index erived by the Hodrick–Prescott filter with a smoothing parame- er set to 14,400.9 For Sweden and Canada, where we use quarterly ata, the output gap was taken as reported in the OECD Economic utlook (production function method based on NAWRU – non- ccelerating wage rate of unemployment).
We proxy financial stress by means of the FSI provided recently y the IMF (Cardarelli et al., 2011), which is a consistent measure or a wide range of countries but, at the same time, is sufficiently omprehensive to track stress of a different nature. It includes the ain components of financial stress in an economy and is available
or a reasonably long period to be used for our empirical analysis see Fig. 1). We use both the overall index, which is a sum of seven omponents, as well as each sub-index and component separately:
(i) Banking-related sub-index components: the inverted term
For Australia, the monthly CPI is not available because both the Reserve Bank f Australia and the Australian Bureau of Statistics only publish quarterly data. The onthly series was obtained using linear interpolation of the CPI index. 9 The industrial production cycle had to be used as a proxy for the output gap
iven that GDP data are not available at monthly frequency. Though a bit more olatile, it is highly correlated with the output gap from GDP (comparison at quar- erly frequency). Moreover, industrial production data tend to be revised less often nd to a lesser extent than the GDP data, which reduced the problem of real-time s. ex-post data present in the GDP data.
J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138 121
USA
-10
-5
0
5
10
15
20
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
UK
-10
-5
0
5
10
15
20
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Swede n
-10
-5
0
5
10
15
20
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Canada
-10
-5
0
5
10
15
20
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Australia
-10
-5
0
5
10
15
20
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Note: The figure presents the evolut ion of the IMF stress index over time. Higher numbers indicate more stress (see Cardarelli et al., 2011).
Fig. 1. IMF financial stress indicator. Note: The figure presents the evolution of the IMF stress index over time. Higher numbers indicate more stress (see Cardarelli et al., 2
(
c w t t
011).
(12-month rolling beta, which is a measure of the correlation of banking stock returns to total returns in line with the CAPM).
(ii) Securities-market-related sub-index components: corporate bond spread (the difference between corporate bonds and long-term government bond yields), stock-market returns
(monthly returns multiplied by −1), time-varying stock-return volatility from the GARCH(1, 1) model.
iii) Foreign-exchange-related sub-index: the time-varying volatil- ity of monthly changes in NEER, from the GARCH (1, 1) model.
v 8 o i
We examined various alternative methods of aggregating the omponents – simple sum, variance-equal weighting, and PCA eighting – but failed to uncover any systematic differences among
hese in terms of the values of the overall index and consecutively in he empirical results. Cardarelli et al. (2011) confirm that extreme
alues of this indicator correctly identify almost all (approximately 0–90%) of the financial crises (including banking, currency, and ther crises, along with stock and house-price boom and busts) dentified in previous studies.
1 ncial
a f o n 2 d s t l i p t t H o v d t s a
3
a w n
r
w r t u c o
a d p e s
r
w i p a p p a p b o m v i
r
i ( m
c s 2 o p t t o a m s a
s s
r
W r m x i m e f s t i t a s b o t 6 t t
f t t t alternative methods for modeling structural changes in monetary- policy rules that occur on an unknown date: (i) regime-switching models, in particular state-dependent Markov switching models
12 More precisely, i equals 6 when we use monthly data and 2 for quarterly data. Although the targeting horizon of central banks is usually somewhat longer (4–8 quarters), as in the other papers in this stream of literature, we prefer to proxy inflation expectations by inflation in t + 2 quarters for the following reasons. First, the endogeneity correction requires a strong correlation between the endogenous regressor and its instruments. Second, the prediction error logically increases at longer horizons. Most importantly, the choice of i is in line with the theory. Batini and Nelson (2001) show that i = 2 in their baseline model of an optimal policy hori-
22 J. Baxa et al. / Journal of Fina
The use of a composite index has a number of benefits. First, it pproximates the evolution of financial stress caused by different actors and thus is not limited to one specific type of instability. Sec- nd, the inclusion of additional variables in the stress index does ot affect the evolution of the indicator markedly (Cardarelli et al., 011). Third, the composition of the indicator allows for breaking own the reactions of the central bank with respect to different tress subcomponents. Nevertheless, one has to be cautious about he interpretation. The composite indicator might suggest a mis- eading interpretation as long as the stress is caused by variables not ncluded in the FSI but rather highly correlated with some subcom- onent. An example is the case of Sweden during the ERM crisis. At he time of the crisis, Sweden maintained a fixed exchange rate, and he Riksbank sharply increased interest rates to sustain the parity. owever, this is not captured by the exchange-rate subcomponent f the FSI, which measures exchange-rate volatility, because the olatility was actually close to zero. A closer examination of the ata shows that this period of stress is captured by the inverted erm structure; hence, it is incorrectly attributed to bank stress. A imilar pattern can be observed for the UK, where the FSI increases fter the announcement of withdrawal from the ERM.
.2. The empirical model
Following Clarida et al. (1998a,b), most empirical studies ssume that the central bank sets the nominal interest rate in line ith the state of the economy typically in a forward-looking man- er:
∗ t = r̄ + ˇ(E[�t+i|˝t] − �∗
t+i) + �E[yt+j|˝t] (1)
here r∗t denotes the targeted interest rate, r̄ is the policy neutral ate,10 �t+i stands for the central bank forecast of the yearly infla- ion rate, i indicates periods ahead based on an information set ˝t
sed for interest-rate decisions available at time t, and �∗ t+i is the
entral bank’s inflation target.11 yt+j represents a measure of the utput gap.
Nevertheless, Eq. (1) was found to be too restrictive to provide reasonable description of actual interest-rate setting. Notably, it oes not account for interest-rate smoothing by central banks, in articular the practice whereby the central bank adjusts the inter- st rate sluggishly to the targeted value. This is tracked in empirical tudies by the simple partial-adjustment mechanism:
t = �rt−1 + (1 − �)r∗t (2)
here � ∈ [0, 1] is the smoothing parameter. There is an ongo- ng controversy as to whether this parameter represents genuine olicy inertia or reflects empirical problems related to omitted vari- bles, dynamics or shocks (see, e.g., Rudebusch, 2006). The linear olicy rule in Eq. (1) can be obtained as the optimal monetary- olicy rule in the LQ framework, where the central bank aims only t price stability and economic activity. Bauducco et al. (2008) ropose an NK model with a financial system where the central ank has privileged information (given its supervisory function) n the health of the financial sector. In such a setting, the com-
on policy rule represented by Eq. (1) will be augmented by
ariables representing the health of the financial sector. Follow- ng this contribution, we consider the forward-looking rule where
10 The policy-neutral rate is typically defined as the sum of the real equilibrium ate and expected inflation. 11 An explicit definition of an inflation target exists only for countries with an nflation-targeting (IT) regime. Most empirical studies assume, in line with Taylor 1993), that this target does not vary over time and can be omitted from the empirical
odel.
z t r o t t t i t M b l r
Stability 9 (2013) 117– 138
entral banks may respond to a comprehensive measure of financial tress rather than stress in a particular segment (Bulíř and Čihák, 008). In practice, the augmented rule can be of some interest to utsiders because inflation expected by the individual monetary- olicy committee members is unobservable to the public (even hough some central banks publish figures that may be very close to he unobserved expected inflation, such as staff inflation forecasts r inflation forecasts stemming from interactions between staff nd monetary-policy committee members). In such case, outsiders ay benefit from including additional indicators such as financial
tress in the policy rule to predict the central bank’s behavior more ccurately.
Therefore, we substitute Eq. (2) into Eq. (1), eliminate unob- erved forecast variables and include measures of the financial tress described above, which results in Eq. (3):
t = (1 − �)[ ̨ + ˇ(�t+i − �∗ t+i) + �yt+j] + �rt−1 + ıxt+k + εt (3)
hile in Eq. (1) the term ̨ coincides with the policy-neutral rate ¯, its interpretation is not straightforward once the model is aug-
ented by additional variables. Note that the financial stress index t+k does not appear within the square brackets. This is because t is typically not included in the loss function of central banks’
onetary policy but it is rather a factor such as the lagged inter- st rate, i.e., it may explain why the actual interest rate rt deviates rom the target. Moreover, by placing it in the regression at the ame level as a lagged interest rate, we can directly test whether his variable representing ad hoc policy decisions decreases the nterest-rate inertia �, as suggested by Mishkin (2009). At the same ime, the response on the coefficient ı can increase, as central banks re more likely to react to financial stress when stress is high. Con- equently, it is possible that � and ı move in opposite directions ecause the central bank either smoothes the interest-rate changes r adjusts the rates in the face of financial stress. In the latter case, he response is likely to be quick and substantial. We set i equal to , j equal to 0 and k equal to −1.12 Consequently, the disturbance erm εt is a combination of forecast errors and is thus orthogonal o all information available at time t (˝t).
The empirical studies on monetary-policy rules have moved rom using time-invariant estimates (Clarida et al., 1998a,b) hrough sub-sample analysis (Taylor, 1999; Clarida et al., 1998a,b) oward more complex methods that allow an assessment of he evolution of the conduct of monetary policy. There are two
on. However, alternative specifications of their model show some sensitivity in erms of what is the optimal i. Nevertheless, employing different i’s for regression esults left the results in most cases unchanged, to a large extent. In the case of the utput gap, we instead assume a backward-looking reaction. The reason is that in he absence of real-time data, we have to rely on the output-gap construction by sta- istical methods such as HP filter. It is arguable that aside from the prediction error, here is also a construction error that might be magnified if an unobserved forecast s substituted by the output-gap estimate for future periods. Finally, we assume hat central bankers’ response (if any) to financial stress is rather immediate (see
ishkin, 2009). Therefore, we use one lag of the FSI and its subcomponents in the enchmark case. However, as a robustness check, we allow for different lags and
eads, allowing the central bankers’ response to financial stress to be preemptive ather than reactive.
ncial
( a b K B a a g i T s s (
m U b w u o a o n i f c c b i a a s K
e v e r p e r e f
r
˛
ˇ
�
ı
�
�
y
x
s c c e
T i t o p i ( c f a ( d � ϕ i ( s E r g i
r
A s fi “ a l s s
∑
w s t p s a w i
•
•
J. Baxa et al. / Journal of Fina
Valente, 2003; Assenmacher-Wesche, 2006; Sims and Zha, 2006) nd (ii) state-space models, where the changes are characterized y smooth transitions rather than abrupt switches (Boivin, 2006; im and Nelson, 2006; Trecroci and Vassalli, 2010). As argued in axa et al. (2010), we consider the second approach to be prefer- ble for the estimation of policy rules, given that it is more flexible nd allows for the incorporation of a simple correction of endo- eneity (Kim, 2006; Kim and Nelson, 2006), which is a major issue n forward-looking policy rules estimated from ex-post data.13
he state-space approach, or time-varying coefficient model, also eems suitable when one wants to evaluate the effect of factors uch as financial stress that can, for a limited length of time, alter rather than permanently change) monetary-policy conduct.
State-space models are commonly estimated by means of a aximum likelihood estimator via the Kalman filter or smoother. nfortunately, this approach has several limitations that can ecome problematic in applied work. First, the results are some- hat sensitive to the initial values of the parameters, which are sually unknown, especially in the case of variables whose impacts n the dependent variable are not permanent and whose sizes re unknown, which is the case for financial stress and its effect n interest rates. Second, the log likelihood function is highly on-linear, and in some cases optimization algorithms fail to min-
mize the negative of the log likelihood. In particular, it can either ail to calculate the Hessian matrix throughout the iteration pro- ess, or, when the likelihood function is approximated to facilitate omputations, the covariance matrix of observation vectors can ecome singular for the starting values provided. The alternative
s a moment-based estimator proposed by Schlicht (1981, 2005) nd Schlicht and Ludsteck (2006), which is employed in our paper nd briefly described below. This framework is sufficiently flexible uch that it incorporates the endogeneity correction proposed by im (2006).
Kim (2006) shows that the conventional time-varying param- ter model delivers inconsistent estimates when explanatory ariables are correlated with the disturbance term and proposes an stimator of the time-varying coefficient model with endogenous egressors. Endogeneity may arise not only in forward-looking olicy rules based on ex-post data (Kim and Nelson, 2006; Baxa t al., 2010) but also in the case of variables that have a two-sided elationship with monetary policy. Financial stress unquestionably nters this category. Following Kim (2006), we rewrite Eq. (3) as ollows:
t = (1 − �t)[˛t + ˇt(�t+i) + �tyt+j] + �trt−1 + ıtxt+k + εt (4)
t = ˛t−1 + ϑ1,t, ϑ1,t∼i.i.d. N(0, �2 ϑ1
) (5)
t = ˇt−1 + ϑ2,t, ϑ2,t∼i.i.d. N(0, �2 ϑ2
) (6)
t = �t−1 + ϑ3,t, ϑ3,t∼i.i.d. N(0, �2 ϑ3
) (7)
t = ıt−1 + ϑ4,t, ϑ4,t∼i.i.d. N(0, �2 ϑ4
) (8)
t = �t−1 + ϑ5,t, ϑ5,t∼i.i.d. N(0, �2 ϑ5
) (9)
t+i = Z ′ t−m� + �ϕϕt, ϕt∼i.i.d. N(0, 1) (10)
t+j = Z ′ t−m + � t, t∼i.i.d. N(0, 1) (11)
t+k = Z ′ t−mo + ���t, �t∼i.i.d. N(0, 1) (12)
13 The time-varying parameter model with specific treatment of endogeneity is till relevant when real-time data are used (Orphanides, 2001). The real-time fore- ast is not derived under the assumption that nominal interest rates will remain onstant within the forecasting horizon (Boivin, 2006) or in the case of measurement rror and heteroscedasticity (Kim et al., 2006).
•
c
u
Stability 9 (2013) 117– 138 123
he measurement Eq. (4) of the state-space representation s the monetary-policy rule. The transitions in Eqs. (5)–(9) describe he time-varying coefficients as a random-walk process with- ut drift.14 Eqs. (10)–(12) track the relationship between the otentially endogenous regressors (�t+i, yt+j, and xt+k) and their
nstruments, Zt. We use the following instruments: �t−1, �t−12 �t−4 for CAN and SWE), yt−1, yt−2, rt−1, the foreign interest rate for ountries other than the United States (the three-month EURIBOR or SWE and UK, and the US three-month interbank rate for CAN nd AUS). Unlike Kim (2006), we assume that the parameters in Eqs. 10)–(12) are time-invariant. The correlation between the stan- ardized residuals ϕt, t, and �t and the error term εt is �ϕ,ε, � ,ε, and �,ε, respectively (note that �ϕ , � , and �� are the standard errors of t, t, and �t, respectively). Consistent estimates of the coefficients
n Eq. (4) are obtained in two steps. In the first step, we estimate Eqs. 10)–(12) and save the standardized residuals ϕt, t, and �t. In the econd step, we estimate Eq. (13) along with Eqs. (5)–(9). Note that q. (13) now includes bias correction terms, i.e., the (standardized) esiduals from Eqs. (10)–(12), to address the aforementioned endo- eneity of the regressors. Consequently, the estimated parameters n Eq. (13) are consistent, as t is uncorrelated with the regressors.
t = (1 − �t)[˛t + ˇt�t+6 + �tyt−1] + �trt−1 + ıtxt−1 + �ϕ,ε�εϕt
+ � ,ε�ε t + ��,ε�ε�t + t, t∼N(0, (1 − �2 ϕ,ε − �2
v,ε − �2 �,ε)�
2 ε,t)
(13)
s previously noted, instead of the standard framework for second- tep estimation, the maximum likelihood estimator via the Kalman lter (Kim, 2006), we use an alternative estimation framework, the varying coefficients” (VC) method (Schlicht, 1981, 2005; Schlicht nd Ludsteck, 2006). This method is a generalization of the ordinary east squares approach that, instead of minimizing the sum of the quares of the residuals
∑T t=1
2, uses minimization of the weighted um of the squares:
T
t=1
2 + �1
T∑
t=1
ϑ2 1 + �2
T∑
t=1
ϑ2 2 + · · · + �n
T∑
t=1
ϑ2 n (14)
here the weights �i are the inverse variance ratios of the regres- ion residuals εt and the shocks in time-varying coefficients ϑt, hat is, �i = �2/�2
i . This approach balances the fit of the model and
arameter stability. Additionally, the time averages of the regres- ion coefficients, estimated by a weighted least squares estimator, re identical to their GLS estimates of the corresponding regression ith fixed coefficients, that is, (1/T)
∑T t=1ât = âGLS.15 The method
s useful in our case because:
it does not require knowledge of initial values even for non- stationary variables prior to the estimation procedure. Instead, both the variance ratios and the coefficients are estimated simul- taneously; the property of the estimator that the time averages of the esti- mated time-varying coefficients are equal to its time-invariant counterparts, permits easy interpretation of the results in relation to time-invariant results;
it coincides with the MLE estimator via the Kalman filter if the time series are sufficiently long and if the variance ratios are properly estimated.16 However, this method suffers from
14 Note that while a typical time-invariant regression assumes that at = at−1, in this ase, it is assumed that E[at] = at−1. 15 See Schlicht and Ludsteck (2006) and Baxa et al. (2010) for more details. 16 The Kalman filter as implemented in common econometric packages typically ses the diffusion of priors for its initiation, but it still produces many corner
1 ncial
t t o p q b k i t i r t fi r e i W w i l a i
c fi f t r fi i i n c F s c o i o b s
s c t m u t M O c i t i
t c t
4
s o s o m a o
4
i s a a t s t g o r a t S s s a
i s c a f f ( o m r v c a ( T b
24 J. Baxa et al. / Journal of Fina
certain limitations of its own. In particular it requires that: (a) the time-varying coefficients are described as random walks, and (b) the shocks in time-varying coefficients ϑt are minimized (see Eq. (14)).
While this does not represent a major problem for the estima- ion of the coefficients of common variables such as inflation, where he monetary-policy response is permanent, it can lead to a loss f some information about ad hoc response factors in monetary olicy making that are considered by central bankers only infre- uently; however, once they are in place, the policy response can e substantial. The financial stress indicator xt+k seems to be this ind of factor. One way to address this problem is by estimation- ndependent calibration of the variance ratios in Eq. (14), such that he estimated coefficient is consistent with economic logic, i.e., it s mostly insignificant and can become significant (with no prior estriction on its sign) during periods of financial stress, i.e., when he financial stress indicator is different from zero. Therefore, we rst estimate Eq. (13) using the VC method and study whether the esulting coefficients in the FSI correspond to economic intuition, specially whether the coefficient is not constant or slowly mov- ng (the so-called pile-up problem, see Stock and Watson, 1998).
hen this problem occurs, we compare the results with models here k belongs to (−2, −1, 0, 1, 2) and calibrate the variance ratios
n Eq. (13) by the variance ratios estimated for the model with the argest variances in the FSI. This step was necessary for Australia nd Sweden. The Taylor-rule coefficients were compared with the nitial estimates and were consistent in both cases.17
The results of our empirical analysis should reveal whether entral banks adjusted their interest-rate policies in the face of nancial stress. However, the time-varying framework also allows
or inferring whether any response to financial stress led to the emporal dismissal of other targets, in particular the inflation ate. Therefore, we are mainly interested in the evolution of the nancial-stress coefficient ıt. We expect it to be mostly insignif-
cant or zero, given that episodes of financial stress are rather nfrequent, and even if they occur, the monetary authorities may ot always respond to them. Moreover, the size of the estimated oefficient does not have any obvious interpretation because the SI is a composite indicator normalized to have a zero mean. Con- equently, we define the stress effect as a product of the estimated oefficient ıt and the value of the IMF’s FSI xt+k. The interpretation f the stress effect is straightforward: it shows the magnitude of nterest-rate reactions to financial stress in percentage points or, in ther words, the deviation from the target interest rate, as implied y the macroeconomic variables, due to the response to financial
tress.
olutions and often does not achieve convergence. Schlicht and Ludsteck (2006) ompare the performance of the moment estimator and the Kalman smoother in erms of the mean squared error on simulated data, and they conclude that the
oment estimator outperforms the Kalman filter on small samples with a size of p to 100 observations. For comparison, we estimated Eq. (12) using the conven- ional Kalman filter in the GROCER software using the tvp function (Dubois and
ichaux, 2009). We parameterized the model by initial conditions taken from the LS estimates of the parameters on the full sample and the initial forecast error ovariance matrix set to 0. The matrix of the residuals of time-varying coefficients s assumed to be diagonal, as in the VC method. The results were very similar to hose obtained from the VC method when the estimated variances were the same n both methods. 17 Stock and Watson (1998) propose a medium-unbiased estimator for variance in he time-varying parameter model, but its application is straightforward only in the ase of one time-varying coefficient, and more importantly, it requires the variables o be stationary.
i a a s i
s a a
e w
e s p r
Stability 9 (2013) 117– 138
. Results
This section summarizes our results on the effect of financial tress on interest-rate setting. First, the results on the effect of the verall measure of financial stress on interest-rate setting are pre- ented. Second, the effect of specific components of financial stress n monetary policy is examined. Third, we briefly comment on the onetary-policy rule estimates that served as the input for the
ssessment of financial-stress effects. Finally, we perform a series f robustness checks.
.1. Financial-stress effect
Fig. 2 presents our results on the effect of financial stress on nterest-rate setting in all five countries (referred to as the financial- tress effect hereinafter).18 Although there is some heterogeneity cross countries, some global trends in the effect of financial stress re apparent. Whereas in good times, such as in the second half of he 1990s, financial stress has virtually no effect on interest-rate etting or is slightly positive,19 the reaction of monetary authori- ies to financial stress was highly negative during the 2008–2009 lobal financial crisis. While the previous evidence on the effect f financial stress on monetary policy is somewhat limited, our esults broadly confirm the time-invariant findings of Cecchetti nd Li (2008), who show that the US Fed adjusted interest rates to he procyclical impact of bank capital requirements in 1989–2000. imilarly, Belke and Klose (2010) estimate the Taylor rule on two ub-samples (before and during the 2008–2009 global financial cri- is) and find that the Fed reacted systematically not only to inflation nd the output gap, but also to asset prices, credit, and money.
The size of financial-stress effects on interest-rate setting dur- ng the recent financial crisis is somewhat heterogeneous, with the trongest reaction found for the UK. The results suggest that all entral banks except the Bank of England maintain policy rates at pproximately 50–100 basis points lower compared to the counter- actual policy of no reaction to financial stress. The size of this effect or the UK is assessed to be approximately three times stronger i.e., 250 basis points). This implies that approximately 50% of the verall policy-rate decrease during the recent financial crisis was otivated by financial-stability concerns in the UK (10%–30% in the
emaining sample countries), while the remaining half falls to unfa- orable developments in domestic economic activity. This finding omplements previous results suggesting that the BoE’s consider- tion of expected inflation over the last decade has been very low as found by Baxa et al., 2010, using the time-varying model and by aylor and Davradakis, 2006, in the context of the threshold model) y evidence that it further decreased during the current crisis. It
s also evident that the magnitude of the response is unusual for ll five central banks. However, the results for Australia, Canada, nd Sweden show a similar magnitude of response to financial tress during the recent financial crisis compared to that observed n previous periods of high financial stress.
Given that the 2008–2009 global crisis occurred at the end of our
ample (there is a peak in the stress indicator of five standard devi- tions that has not returned to normal values yet), we performed an dditional check to avoid possible end-point bias. In particular, we
18 Given that the magnitude of the financial-stress effect differs across countries, specially due to the high positive peak for Sweden and negative peak for the UK, e use different scales for different countries.
19 Note that the positive effect of financial stress on interest-rate setting is to some xtent a consequence of scaling the financial-stress indicator; its zero value corre- ponds to the long-run average stress. Hence, we do not pay much attention to ositive values of stress unless caused by a temporarily positive and significant egression coefficient associated with the FSI.
J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138 125
USA
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
UK
-2.5
-2
-1.5
-1
-0.5
0
0.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Sweden
-2
-1
0
1
2
3
4
5
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Canada
-0.5 -0.4 -0.3 -0.2 -0.1
0 0.1 0.2 0.3 0.4 0.5 0.6
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Australia
-0,8
-0,6
-0,4
-0,2
0
0,2
0,4
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Notes: The figu re depicts the evolution of the financial-stress effect. The stress effect (y-axis) is coefficient estimated the of product the as defined the in indicator financial-stress the on
shows t he ma gnitude of the i nte rest -rate r eaction to financial stress in percentage points.
F ts the p ary-po s points
r 2 f r C q t
( t 1
ig. 2. The effect of financial stress on interest-rate setting. Notes: The figure depic roduct of the estimated coefficient on the financial-stress indicator in the monet hows the magnitude of the interest-rate reaction to financial stress in percentage
an our estimation excluding the observation from the period of the 008–2009 crisis. These results were practically indistinguishable rom the full sample estimation. With regard to the effect of the cur-
ent crisis, the largest uncertainty is associated with the results for anada, for which the shortest data sample – ending in the fourth uarter of 2008 – was available. When the possibility of a preemp- ive reaction of the central bank to financial stress is considered
t t
i
evolution of the financial-stress effect. The stress effect (y-axis) is defined as the licy rule and the value of the IMF financial-stress indicator (ıx). The stress effect .
see the robustness checks below), the effect of financial stress in he current crisis is estimated for Canada at somewhere between % and 2% (see Appendix 3). These additional results suggest that
he response of the Bank of Canada in the benchmark model is likely o be underestimated.
The question of which components of financial stress influence nterest-rate setting is addressed in Fig. 3. In this case, we estimate
126 J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138
USA
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3 19
81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Ban k stres s Stoc k market stress Exchan ge rate stres s
UK
-2
-1.5
-1
-0.5
0
0.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Ban k stres s Stock market stres s Exchan ge rate stres s
Sweden
-3 -2
-1 0 1
2 3 4
5 6
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Bank stress Stock mar ket stre ss Exchange rate stress ERM crisis
Canada
-1.5
-1
-0.5
0
0.5
1
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Ban k stres s Stock market stres s Exchan ge rate stres s
Australia
-2
-1,5
-1
-0,5
0
0,5
1
1,5
2
2,5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Bank stress Stock market stress Exchang e rate stress
Notes: The f igure depicts the evolution of the components of the financial-stress effect, namely, the bank-s tress effect, the exchange- rate st ress effect, and the stock-market stress effect. The
defined is (y-axis) effect stress given the on coefficient estimated the of product the as component of the financ ial-stress indicator in the monetary-policy rule and the value of the
Fig. 3. The effect of financial stress components on interest-rate setting: bank stress, exchange-rate stress, and stock-market stress. Notes: The figure depicts the evolution o xchan d nanc c gnitu
t e e e i b
c e
f the components of the financial-stress effect, namely, the bank-stress effect, the e efined as the product of the estimated coefficient on the given component of the fi omponent of the IMF financial-stress indicator (ıx). The stress effect shows the ma
he model using each FSI subcomponent separately (the bank stress ffect, the exchange-rate stress effect, and the stock-market stress
ffect) instead of the overall FSI and report the financial-stress ffect attributable to each subcomponent. Some heterogene- ty across countries is again apparent, although it seems that ank stress and stock-market stress dominated central bankers’
C
i m
ge-rate stress effect, and the stock-market stress effect. The stress effect (y-axis) is ial-stress indicator in the monetary-policy rule and the value of the corresponding de of the interest-rate reaction to financial stress in percentage points.
onsiderations in less open economies. On the other hand, xchange-rate stress matters in more open economies such as
anada and Sweden.
Specifically, the US Fed seemed to be worried about financial nstability, especially during the 1980s. We can observe that the
ain concern in the early 1980s was banking stress, which is
ncial
a w m r
t s t e t t 2 i e u a t o t o i p i u h w l i i r o k p
e F fi 2 c o s t a s t
R u R t l e a a
t u m o l r i p i
d s i
i t h e t t c
s n o c c t
t m i e
4
t e p F v u s
c p d t T c l 2 b t d b w
i
J. Baxa et al. / Journal of Fina
rguably related to the Savings and Loans crisis. Another concern as that of stock-market stress, in particular during the stock- arket crash of 1987, when interest rates were 30 b.p. lower with
espect to the benchmark case. The Bank of England was, in general, much more perceptive
o financial stress. We find its response mainly to stock-market tress again, notably, in 1987. Interestingly, we find little response o exchange-rate stress, not even during the 1992 ERM crisis. Nev- rtheless, it has to be emphasized that the interest-rate reaction to his speculative attack was subdued in comparison to, for example, he Riksbank (Buiter et al., 1998). The base rate was increased by
p.p. to 12% on September 16, 1992. Despite a promise of further ncreases up to 15%, traders continued selling the pound. On the vening of the same day, the UK left the ERM with interest rates nchanged; on the following day, the base rate decreased to 10.5%; nd at the end of September, the base rate was 9%, lower than at he beginning of the month. Therefore, despite huge open market perations, the response of the interest rate was moderate, with he monthly interest-rate average practically unaffected. Hence, ur framework does not detect any effect of financial stress on the nterest rate during the ERM crisis. Since the devaluation of the ound sterling in September 1992, the effect of financial stress on
nterest-rate setting approaches zero from originally negative val- es. Aside from this, the response of the Bank of England to inflation as decreased. From this perspective, it seems the pound sterling’s ithdrawal from the ERM allowed for both a more rule-based and
ess restrictive monetary policy. With respect to the banking crisis n the late 2000s, the Bank of England provided liquidity support n its earlier stage in 2007 with the fall of Northern Rock. Policy ates remained constant until late 2008, despite the bankruptcy f Lehman Brothers in the US in September 2008. The reason for eeping policy rates constant was related to concerns regarding otential inflationary pressures from rising oil and food prices.
The interest-rate effect of the banking crisis in Sweden in the arly 1990s is estimated to be slightly over 1% in absolute terms (see ig. 2). The crisis began in September 1990, when the non-banking nancial institution Nyckeln unexpectedly collapsed (Jennergren, 002). The Riksbank did not decrease interest rates sharply because oincidental international factors, in particular the reunification f Germany, forced interest rates upwards. Despite facing reces- ion, the government attempted to defend the peg of the krona o ECU and decided to prevent the spread of the banking crisis by nnouncing a blanket guarantee for the liabilities of the banking ector (Jonung, 2009). Hence, interest-rate cuts were not a primary ool chosen for resolution of the crisis.
In comparison to the United Kingdom, the reaction of the iksbank to the ERM crisis was different. First, after a series of spec- lative attacks on the Swedish krona in mid-September 1992, the iksbank still attempted to maintain the fixed exchange rate, and he marginal interest rate jumped up 500% to offset the outflow of iquidity and other speculative attacks (see the large positive stress
ffect on the interest rate in 1992 in Fig. 2). However, not even such n increase was sufficient, and the fixed exchange rate had to be bandoned later, in November.20
20 For Sweden, we add a dummy variable for the third quarter of 1992 (ERM crisis) o Eq. (13). At this time, the Swedish central bank forced short-term interest rates pward in an effort to keep the krona within the ERM. From the perspective of our odel, it was a case of a strong positive reaction to the actual stress that lasted
nly one period. When this dummy variable was not included, the model with a agged value of the FSI was unable to show any link between stress and interest ates, and the estimates of other coefficients were inconsistent with economic intu- tion. Clearly, since we use data at monthly and quarterly frequency, this limits the ossibility to detect and properly analyze day-to-day dynamics of some short-term
nstability events.
f h p a c o i l t e l R i o
Stability 9 (2013) 117– 138 127
The Reserve Bank of Australia significantly loosened its policy uring the 1980s. This can be attributed to stress in the banking ector with the exception of the reaction to the stock-market crash n 1987 (see Fig. 3).
The exchange rate as well as bank stress seems to matter for nterest-rate considerations at the Bank of Canada. Interestingly, he results suggest that the Bank of Canada often responded to igher exchange-rate stress by monetary tightening. A possible xplanation for this finding might be that given the openness of he Canadian economy, its central bank tightened the policy when he currency stabilized at the level that the monetary authority onsidered to be undervalued.
We would like to highlight a comparison of Figs. 2 and 3. First, it hould be noted that a positive response to one stress subcompo- ent may cancel out in the face of a negative response to another ne, making the response to the overall stress negligible (as in the ase of Canada). Second, the stress effects related to individual sub- omponents do not necessarily sum up to the stress effect related o the entire FSI.
Overall, the results suggest that the central bank tends to react o financial stress, and different components of financial stress
atter in different time periods. The effect of financial stress on nterest-rate setting is found to be virtually zero in good times and conomically sizable during periods of high financial stress.
.2. Monetary policy rule estimates
Given that our main interest lies in the interest-rate response o financial stress, we comment on the other monetary-policy rule stimates only briefly. The plot of the evolution of the estimated arameters over time for all countries is available in Appendix 1. irst of all, it should be noted that most coefficients do indeed ary over time, which is consistent with previous evidence and nderlines the fact that monetary-policy conduct has evolved sub- tantially in recent decades.
In general, the responses to inflation (ˇ) are positive, and the oefficient is often above one, consistent with the Taylor princi- le. Nevertheless, we find that in the last decade the coefficient ecreased somewhat, and during the recent financial crisis it even urned slightly negative (in the US and UK; more on this below). he decrease of the inflation response during the last decade is typi- ally attributed to well-anchored inflation expectations as well as a ow-inflation environment (Sekine and Teranishi, 2008; Baxa et al., 010). The finding of negative ˇ during the recent crisis is likely to e related to the fact that central banks were decreasing policy rates o historical lows in the face of exceptionally high financial stress, espite inflation expectations being largely unchanged, rather than eing an indication that policy rates were systematically decreased hen inflation expectations increased.
For the United States, our results show that the response to nflation was highest in the early 1980s, and except for the period ollowing the recession of 1990–1991 the estimated coefficient is igher or very close to one. This value is slightly lower in com- arison to Kim and Nelson (2006), who found the response to be round 1.5 and almost invariant since 1981. Given the size of the onfidence intervals, it is, however, difficult to determine whether ur results differ significantly. Kim and Nelson (2006) estimate the nterest-rate smoothing coefficient to be higher than 0.8, i.e., in ine with what time-invariant estimates of monetary-policy rules ypically suggest (see, for example, Clarida et al., 1998a,b). Our stimates indicate that the interest-rate smoothing is somewhat
ower (0.5–0.6). This finding is in line with the recent critique by udebusch (2006), who argues that the practical unpredictability of
nterest-rate changes over a few quarters suggests that the degree f interest-rate smoothing is rather low. Interestingly, we find that
128 J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138
Response to inflation ( )
-3
-2
-1
0
1
2
3
4
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Response to output gap ( )
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Interest rate smoothin g ( )
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
1
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Response to financial stress ( )
-0.12
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Note: The estima ted coefficie nts o f the time -varying monetary policy rule are depicted with a
F ts of t
t i ( t r m t ( t r p w t t a f s a
f a w u c R d
t f
p t t e e c t w b d t c
s l s l a t f t a a
95% confidence interval.
ig. A1.1. Time-varying monetary policy rules: USA. Note: The estimated coefficien
he response to inflation decreases substantially after the terror- st attacks on September 11, 2001. This complies with Greenspan 2007), who argued in that case that the Fed was concerned about he US economy spiraling downward into recession after the ter- orist attacks. Later, Greenspan himself acknowledged that the onetary policy was somewhat loose, but ex ante optimal, given
he increased uncertainty after the attacks. In a similar vein, Taylor 2010) compares the actual values of the federal funds rate and he counterfactual values predicted by the (time-invariant) Taylor ule, finding that in 2002–2005 interest rates were too low com- ared to predictions and this deviation from a rules-based policy as “larger than in any period since the unstable decade before
he Great Moderation” (p. 167). Negative estimates of the response o inflation in this particular period are reported also by Trecroci nd Vassalli (2010). The response to the output gap is significant or nearly the whole sample, although the values close to 0.2 are omewhat lower than in Kim–Nelson (2007), but similar to Trecroci nd Vassalli (2010).
The results for countries that currently have an explicit target or inflation share several features. The interest-rate smoothing is gain found to be lower in comparison to time-invariant estimates, ith midpoints around 0.5. The exception is Canada, where the val- es fluctuate around zero and are insignificant. Moreover, for some
entral banks, such as the RBA and the BoE in 2010 or the Sveriges iksbank in late 1980, we find that central banks are less inertial uring crises.21 Second, the response of interest rates to inflation is
21 Indeed, the correlation coefficient of the estimated time-varying coefficient of he lagged interest rate � and the financial-stress index ı is −0.79 for Australia, 0.21 or Canada, −0.20 for Sweden, −0.68 for the UK, and 0.60 for the US.
t t p
4
a
he time-varying monetary policy rule are depicted with a 95% confidence interval.
articularly strong during the periods when central bankers want o break a record of high inflation, such as in the UK or Australia at he beginning of the 1980s, and is less aggressive in a low-inflation nvironment with subdued shocks and well-anchored inflation xpectations (Kuttner and Posen, 1999). In this respect, our results onfirm the findings of Taylor and Davradakis (2006), who argue hat the response of the Bank of England to inflation is insignificant hen the inflation rate is close to its target. Third, some central
anks (Australia and Canada) are also found to react to output-gap evelopments, with the parameter estimated to be slightly posi- ive on average, whereas the parameter is insignificant with wide onfidence intervals in Sweden and the United Kingdom.
The results show that the interest-rate response to financial tress is insignificant most of the time, at the 95% significance evel. This is in line with our expectations, i.e., that the coefficients hould be insignificant in periods when stress is low. Neverthe- ess, the coefficient on financial stress is statistically significant t the 95% level during the recent financial crises for most coun- ries. The importance of financial stress for interest-rate setting is urther confirmed using the GMM estimation, which shows that he financial-stress index is significant, in fact, in all countries. In ddition, when the one-standard-deviation quantile is taken into ccount instead of the more usual two-standard-deviation quan- ile, the periods when we can identify any interest-rate response o financial stress become more evident. We present a list of these eriods in Table A5.2.
.3. Robustness checks
In terms of the financial-stress effect estimates, we perform battery of robustness checks. First, following the argument put
J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138 129
Response to inflation ( )
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Response to output gap ( )
-0.3 -0.25 -0.2
-0.15 -0.1
-0.05 0
0.05 0.1
0.15 0.2
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Interest rate smoothin g ( )
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Response to financial stress ( )
-0.25 -0.2
-0.15 -0.1
-0.05 0
0.05 0.1
0.15 0.2
0.25
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Note: The estima ted coef ficients o f the time -varying monetary policy rule are depicted with a 95% confidence interval.
Fig. A1.2. Time-varying monetary policy rules: UK. Note: The estimated coefficients of the time-varying monetary policy rule are depicted with a 95% confidence interval.
Response to inflation ( Response to output gap () )
Interest rate smoothin g ( Response to financial stress () )
Note: The estima ted coefficie nts o f the time -varying monetary policy rule are depicted with a 95% confidence interval.
0
0.5
1
1.5
2
2.5
3
3.5
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
Fig. A1.3. Time-varying monetary policy rules: Sweden. Note: The estimated coefficients of the time-varying monetary policy rule are depicted with a 95% confidence interval.
130 J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138
Response to inflation (β)
0
0.5
1
1.5
2
2.5
3
3.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Response to output gap (γ)
Interest rate smoothin g ( ρ)
-0.2
0
0.2
0.4
0.6
0.8
1
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Response to financial stress (δ)
Note: The estima ted coefficie nts o f the time -varying monetary policy rule are depicted with a 95% confidence interval.
-0.2
0
0.2
0.4
0.6
0.8
1
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
-0.12
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Fig. A1.4. Time-varying monetary policy rules: Australia. Note: The estimated coefficients of the time-varying monetary policy rule are depicted with a 95% confidence interval.
Response to inflation (β)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
Response to output gap (γ)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
Interest rate smoothin g ( ρ)
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
Response to financial stress (δ)
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
19 81
19 84
19 87
19 90
19 93
19 96
19 99
20 02
20 05
20 08
Note: The estimated coef ficients o f the time -varying monetary policy rule are depicted with a 95% confidence interval.
Fig. A1.5. Time-varying monetary policy rules: Canada. Note: The estimated coefficients of the time-varying monetary policy rule are depicted with a 95% confidence interval.
J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138 131
USA
-2.5
-2
-1.5
-1
-0.5
0
0.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
UK
-2.5
-2
-1.5
-1
-0.5
0
0.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Sweden
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Canada
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Australia
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Notes: The figu re depicts the evolution of the financial-stress effect. The stress effect (y-axis) is coefficient estimated the of product the as defined the in indicator financial-stress the on
monetary-po licy rule and the va lue of the IMF financial-stress indicator ( x). The stress effect shows t he ma gnitude of the i nte rest -rate r eaction to financial stress in percentage points.
Fig. A2.1. The effect of financial stress on interest-rate setting. Notes: The figure depicts the evolution of the financial-stress effect. The stress effect (y-axis) is defined as t etary- s points
f b u a a t b
e a c
he product of the estimated coefficient on the financial-stress indicator in the mon hows the magnitude of the interest-rate reaction to financial stress in percentage
orward above that the interbank rate may occasionally provide a etter signal of monetary-policy intentions than the policy rate, we se interbank interest rates as a dependent variable. These results
re reported in Figs. A2.1 and A2.2. We can observe that the over- ll stress effect on the interbank rate was larger for the US during he current crisis, where it explains 2% of the decrease of the inter- ank interest rate. For Sweden, we found a strong positive effect of
r v
t
policy rule and the value of the IMF financial-stress indicator (ıx). The stress effect .
xchange rate volatility in the late 1980s; this might be linked to the im of the central bank to keep the exchange rate fixed. In other ases, there is no substantial difference between the benchmark
esults and the results obtained using this alternative dependent ariable.
Second, in the benchmark model and all of the results reported hus far, we use the first lag of the FSI in the policy-rule estimation.
132 J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138
USA
-2
-1.5
-1
-0.5
0
0.5
1 19
81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Ban k stres s Stoc k market stress Exchan ge rate stres s
UK
-1.6 -1.4 -1.2
-1 -0.8 -0.6 -0.4 -0.2
0 0.2 0.4
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Ban k stres s Stock market stres s Exchan ge rate stres s
Swede n
-1.5
-1
-0.5
0
0.5
1
1.5
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Ban k stres s Stoc k market stress Exchan ge rate stres s
Canada
-1.5
-1
-0.5
0
0.5
1
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Ban k stres s Stock market stres s Exchan ge rate stres s
Australia
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Ban k stres s Stock market stres s Exchan ge rate stres s
Notes: The f igure depicts the evolution of the components of the financial-stress effect, namely, the bank stress effect, the exchange-rate stress effect, and the stock-market stress effect. The
defined is (y-axis) effect stress given the on coefficient estimated the of product the as component of the financ ial-stress indicator in the monetary-policy rule and the value of the corresponding component of the IMF financial-stress indicator ( x). The stress effect shows the
to reaction interest-rate the of magnitude in stress financial points.percentage
Fig. A2.2. The effect of financial stress components on interest-rate setting: bank stress, exchange-rate stress and stock-market stress. Notes: The figure depicts the evolution of the components of the financial-stress effect, namely, the bank stress effect, the exchange-rate stress effect, and the stock-market stress effect. The stress effect (y-axis) is d nanc c gnitu
W o a t
efined as the product of the estimated coefficient on the given component of the fi omponent of the IMF financial-stress indicator (ıx). The stress effect shows the ma
e motivate this choice by the use of monthly data, the frequency f monetary-policy meetings of most central-bank boards, and the ssumption that policy actions are likely to be implemented in a imely fashion. In addition, we employ different lags and leads, in
t r a n
ial-stress indicator in the monetary-policy rule and the value of the corresponding de of the interest-rate reaction to financial stress in percentage points.
he latter case allowing the policy to be preemptive rather than eactive. In this case, we use the future realized value of the FSI s a proxy for the central bank’s expectation (in a similar man- er as to how it is routinely executed for inflation expectations)
J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138 133
USA
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1,0
1984 1988 1992 1996 2000 2004 2008
t-2 t-1 t t+1 t+2
United Ki ngd om
-2,5
-2,0
-1,5
-1,0
-0,5
0,0
0,5
1,0
1984 1988 1992 1996 2000 2004 2008
t-2 t-1 t t+1 t+2
Swede n
-2
-1
0
1
2
3
4
5
6
1988 1992 1996 2000 2004 2008
t-2 t-1 t t+1 t+2
Canad a
-2,5
-2
-1,5
-1
-0,5
0
0,5
1
1,5
1984 1988 1992 1996 2000 2004 2008
t-2 t-1 t t+1 t+2
Australia
-2,5
-2
-1,5
-1
-0,5
0
0,5
1
1984 1988 1992 1996 2000 2004 2008
t-2 t-1 t t+1 t+2
1 vs.
a F b s t p u b fi e
l c i a a v r o b i
f
t s
e u i o i w c s t e a o n p fi
Fig. A3.1. The effect of financial stress (t −
nd, consequently, treat the FSI as an endogenous variable (see ig. A3.1 for the results). To obtain comparable results, we cali- rate the variance ratios with the same values as in the baseline pecification. Although we find rather mixed evidence on preemp- ive policy actions, which may also be related to the inadequacy of roxying the expected values of financial stress by the actual val- es of the financial-stress indicator as well as the fact that a central ank might not react to the stress preemptively, the reaction to nancial stress in the current crisis is strongly negative for both xpected and observed stress.
Third, we further break down the FSI sub-indices to each under- ying variable to evaluate their individual contributions.22 The orresponding stress effects appear in Figs. A4.1 and A4.2. Break- ng down stock-market-related stress, we find that the US Fed nd the BoC react to the corporate bond spread, whereas the BoE nd Sveriges Riksbank are more concerned with stock returns and olatility. While the RBA seems to be concerned with both corpo- ate bond spreads and stock-market volatility in the 1980s, the role
f stock-related stress had substantially decreased by then. As far as ank-related stress is concerned, the TED spread plays a major role
n all countries apart from the UK, where the largest proportion of
22 This applies only to the banking and stock-market subcomponents because the oreign-exchange subcomponent is represented by a single variable.
t e
5
i
t − 2, t, t + 1, t + 2) on interest-rate setting.
he effect on the interest rate can be attributed to an inverted term tructure.
Fourth, because the verifications related to comparing our conometric framework to obvious alternatives such as, first, the se of a maximum likelihood estimator via the Kalman filter
nstead of the moment-based time-varying coefficient framework f Schlicht and, second, the use of a Markov switching model nstead of a state-space model, were provided in Baxa et al. (2010),
e estimate simple time-invariant monetary-policy rules for each ountry by the generalized method of moments, including various ubsamples. This simple evidence reaffirms that the analyzed cen- ral banks seem to pay attention to overall financial stress in the conomy. The FSI is statistically significant, with a negative sign nd a magnitude of between 0.05 and 0.20 for all countries. On the ther hand, the coefficients of its subcomponents often are not sig- ificant, and the exchange-rate subcomponent in some cases has a ositive sign. These results, which are available upon request, con- rm that to understand the interest-rate adjustment in response o financial stress, one should rely on a model allowing for a differ- ntial response across time.
. Concluding remarks
The 2008–2009 global financial crisis generated significant nterest in exploring the interactions between monetary policy and
134 J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138
UKUSA
CanadaSweden
Australia
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Bank ing beta TE D spread Inverted term spread
-4
-3
-2
-1
0
1
2
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Bank ing beta TE D spread Inverted term spread
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Bank ing beta TE D spread Inverted term spread
-1.5
-1
-0.5
0
0.5
1
1.5 19
81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Bank ing beta TE D spread Inverted term spread
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Bank ing beta TE D spread Inverted term spread
k stre
fi m c o t o d fi F c e
u
m a n e f b o a t
Fig. A4.1. The effect of ban
nancial stability. This paper aimed to examine in a systematic anner whether and how the monetary policy of selected main
entral banks (the US Fed, the Bank of England, the Reserve Bank f Australia, the Bank of Canada, and Sveriges Riksbank) responded o episodes of financial stress over the last three decades. Instead f using individual alternative measures of financial stress in ifferent markets, we employed the comprehensive indicator of nancial stress recently developed by the International Monetary und, which tracks overall financial stress as well as its main sub-
omponents, in particular banking stress, stock-market stress and xchange-rate stress.
Unlike a few existing empirical contributions that aim to eval- ate the impact of financial-stability concerns on monetary policy
t t t m
ss on interest-rate setting.
aking, we adopt a more flexible methodology that not only llows for the response to financial stress (and other macroeco- omic variables) to change over time, but also addresses potential ndogeneity (Kim and Nelson, 2006). The main advantage of this ramework is that it not only enables testing of whether central anks responded to financial stress at all, but also detects the peri- ds and types of stress that were the most worrying for monetary uthorities. Our results indicate that central banks truly change heir policy stances in the face of financial stress, but the magni-
ude of such responses varies substantially over time. As expected, he impact of financial stress on interest-rate setting is essen- ially zero most of the time, when the levels of stress are very
oderate. However, most central banks loosen monetary policy
J. Baxa et al. / Journal of Financial Stability 9 (2013) 117– 138 135
UKUSA
CanadaSweden
Australia
-0.5 -0.4 -0.3 -0.2 -0.1
0 0.1 0.2 0.3 0.4 0.5 0.6
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Corpo rate spread Stock returns Stock volatili ty
-0.6 -0.5 -0.4 -0.3 -0.2 -0.1
0 0.1 0.2 0.3 0.4 0.5
19 81
19 83
19 86
19 89
19 92
19 95
19 98
20 01
20 04
20 07
Corpo rate spread Stock returns Stock volatili ty
-2
-1.5
-1
-0.5
0
0.5
1
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Corpo rate spread Stock returns Stock volatili ty
-2
-1.5
-1
-0.5
0
0.5
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Corpo rate spread Stock returns Stock volatili ty
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
19 81
19 83
19 84
19 87
19 89
19 91
19 93
19 95
19 97
19 99
20 01
20 03
20 05
20 07
20 09
Corpo rate spread Stock returns Stock volatili ty
arket
w c c t e o
a fi s e o g
p t s s o m t u U
Fig. A4.2. The effect of stock-m
hen the economy faces high financial stress. There is some cross- ountry and time heterogeneity when we examine central banks’ onsiderations of specific types of financial stress. While most cen- ral banks seem to respond to stock-market stress and bank stress, xchange-rate stress is found to drive the reaction of central banks nly in more open economies
Consistent with our expectations, the results indicate that a size- ble fraction of the monetary-policy easing during the 2008–2009 nancial crisis can be explained by a direct response to the financial
tress above what might be attributed to the decline in inflation xpectations and output below its potential. However, the size f the financial-stress effect differs by country. The result sug- ests that all central banks except the Bank of England kept their
e t l b
stress on interest-rate setting.
olicy rates at 50–100 basis points lower, on average, solely due o the financial stress present in the economy. Interestingly, the ize of this effect for the UK is assessed at about three times tronger (i.e., 250 basis points). This implies that about 50% of the verall policy-rate decrease during the recent financial crisis was otivated by financial-stability concerns in the UK (10%–30% in
he remaining sample countries), while the remaining half falls to nfavorable developments in domestic economic activity. For the S Fed, macroeconomic developments themselves (a low-inflation
nvironment and output substantially below its potential) explain he majority of the interest-rate policy decreases during the crisis, eaving any further response to financial stress to be constrained y zero interest rates.
1 ncial
t o i u s i
a i l a
T T
N 1 r ( B 2 i c A i
T P
36 J. Baxa et al. / Journal of Fina
Overall, our results point to the usefulness of augmenting he standard version of monetary-policy rules by some measure f financial conditions to obtain a better understanding of the nterest-rate-setting process, especially when financial markets are
nstable. The empirical results suggest that the central banks con- idered in this study altered the course of their monetary policy n the face of financial stress. The recent crisis seems truly to be
p s i
able A5.1 ime-invariant reaction functions, GMM estimates.
˛ ˇ �
United States 1 5.59 (1.42) 1.59 (0.99) 1.51 (0.53)
2 −9.42 (4.78) 0.45 (0.33) 0.94 (0.34)
United Kingdom 1 3.93 (1.31) 0.37 (0.52) 1.51 (0.39)
2 7.68 (2.78) −0.89 (0.74) 2.77 (0.71)
Sweden 1 −1.87 (0.86) 2.59 (0.46) −0.16 (0.22)
2 0.24 (0.53) 2.03 (0.39) −0.12 (0.16)
Australia 1 0.04 (0.79) 2.06 (0.3) −0.02 (0.1)
2 2.2 (0.93) 1.84 (0.22) 0.19 (0.14)
Canada 1 −0.33 (1.23) 2.07 (0.96) 0.87 (0.30)
2 1.21 (1.23) 1.67 (0.87) 0.75 (0.28)
United States: 1981:1–1999:12 sample 1 1.82 (1.27) 1.43 (0.57) 1.48 (0.27)
1* −0.25 (0.77) 2.18 (0.42) 0.3 (0.06)
United States: Clarida et al. (1998a,b) – 1982:10–1994:12 sample 2 −0.1 (1.54) 1.83 (0.45) 0.56 (0.16)
otes: Numbers in (·) are standard errors. The samples are as follows: United 983:3M–2009:5M, Sweden 1984:2Q–20091Q, Canada 1981:1Q–2008:4Q. Model 1: rt
t = (1 − �)( ̨ + ˇ�t+k + �yt) + �rt−1. k equals 6 for the USA, the UK and Australia, 2 for Swed the RM crisis) is included. The coefficient is significant at the 5% level, when the ratio of
oth models are estimated using the GMM. The list of instruments follows. United States without lags of FSI in a set of instruments. United Kingdom: lags of interest rate, outp
nflation, output gap, US money market rate and FSI (1–6, 9, 12). Sweden: lags of interes risis. Canada: interest rate, inflation, output gap, U.S. money market rate, and FSI (1–4). dditionally, we show the results for the USA estimated on the subsample 1981–1999. Th
ndustrial production in a similar fashion as in Clarida et al. (1998a,b). Their results are pr
able A5.2 eriods with significant responses to financial stress.
1980s
United States 2SD
1SD 1982:M11–1992
United Kingdom 2SD 1987:M08–1989:M11
1SD 1987:M01–1993
Sweden 2SD 1SD
Australia 2SD 1987M:04–1988:M10
1SD 1983:M07–1993
Canada 2SD 1SD 1982:Q3–1984:Q1
Stability 9 (2013) 117– 138
n exceptional period, in the sense that the response to financial nstability was substantial and coincided in all the countries ana- yzed, which is evidently related to intentional policy coordination bsent in previous decades. However, we have also observed that
revious idiosyncratic episodes of financial distress were, at least in ome countries, followed by monetary-policy responses of similar, f not higher, magnitude.
� ı J-Statistics p-Value
0.97 (0.01) −0.014 (0.006) 24.9808 0.6289 0.94 (0.01) 15.0451 0.8207
0.97 (0.01) −0.018 (0.004) 15.0423 0.5212 0.98 (0.01) 11.4534 0.4905
0.84 (0.04) −0.135 (0.029) 24.9808 0.6289 0.76 (0.05) 15.0451 0.8207
0.95 (0.01) −0.038 (0.006) 21.5261 0.9731 0.89 (0.02) 15.2464 0.9830
0.89 (0.04) −0.089 (0.023) 10.6859 0.8284 0.86 (0.05) 9.4332 0.7395
0.95 (0.01) −0.015 (0.007) 20.1672 0.8583 0.87 (0.02) −0.043 (0.012) 19.8946 0.8683
0.97 (0.03) 10.9000 0.9980
States: 1981:1M–2009:6M, United Kingdom: 1981:1M–2009:3M, Australia: = (1 − �)( ̨ + ˇ�t+k + �yt) + �rt−1 + ıxt−1. Model 2 does not contain financial stress: en and 4 for Canada. For Sweden, a dummy variable for the third quarter of 1992
coefficient to standard error is greater than 1.96. : lags of interest rate, output gap, inflation, and financial stress (1–6, 9, 12), model ut gap, inflation, and EURIBOR 3M (1–6, 9, 12), FSI (1–3). Australia: interest rate, t rate, inflation, output gap, EURIBOR 3M, and FSI (1–4)+ the dummy for the ERM
e model denoted as 1* has the output gap derived from the quadratic trend of log ovided for comparison with ours.
1990s 2000s
2008:M03–2009:M03 :M09 2007:M05–2009:M06
2007:M09–2009:M03 :M01 2006:M03–2009:M03
1990:Q2–1992:Q2 2001:Q2–2002:Q3 1993:Q1 2009:Q1 1999:Q4–2000:Q2
2008:M09–2009:M03 :M10 2002:M10–2009:M05
1996:M06–1997:M05
1992:Q3–1995:Q4
ncial
A
H s C A o T l l d t t
A (
A d
A F
A s
A e
R
A
A
A
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
C
C
C
C
C
C
C
C
C
C
D
D
D
E
E
F
F
F
G
G
G
G
G
H
I
I
J
J
K
K
K
K
K
L
J. Baxa et al. / Journal of Fina
cknowledgments
We thank Aleš Bulíř, Sofia Bauducco, Øyvind Eitrheim, Dana ájková, Bernhard Herz, Ekkehart Schlicht, Miloslav Vošvrda, and
eminar participants at the 7th Norges Annual Monetary Policy onference, the 15th International Conference on Macroeconomic nalysis and International Finance (Rethymno, Greece), the Bank f England, the Czech National Bank, the Institute of Information heory and Automation (Academy of Sciences of the Czech Repub- ic), Universitat de Barcelona, Universitat de Girona, Universitat de es Illes Balears, and Universidad Computense de Madrid for helpful iscussions. The views expressed in this paper are not necessarily hose of the Czech National Bank. This research was supported by he Grant Agency of the Czech Republic, project no. P402/11/1487.
ppendix 1. Time-varying monetary policy rule estimates Figs. A1.1–A1.5)
ppendix 2. The results with the interbank rate as the ependent variable in the policy rule
ppendix 3. The results with different leads and lag of the SI
ppendix 4. The results with individual variables of bank tress and stock-market stress
ppendix 5. Significance of the financial stress index in the stimated Taylor rules (Tables A5.1 and A5.2)
eferences
kram, Q.F., Bårdsen, G., Lindquist, K.-G., 2007. Pursuing financial stability under an inflation-targeting regime. Annals of Finance 3, 131–153.
kram, Q.F., Eitrheim, Ø., 2008. Flexible inflation targeting and financial stability: is it enough to stabilize inflation and output? Journal of Banking & Finance 32, 1242–1254.
ssenmacher-Wesche, K., 2006. Estimating central banks’ preferences from a time- varying empirical reaction function. European Economic Review 50, 1951–1974.
alakrishnan, R., Danninger, S., Elekdag, S., Tytell, I., 2009. The transmission of financial stress from advanced to emerging economies. IMF Working Paper No. 09/133.
axa, J., Horváth, R., Vašíček, B., 2010. How does monetary policy change? Evidence on inflation targeting countries. Czech National Bank Working Paper No. 2-2010.
atini, N., Nelson, E., 2001. Optimal horizons for inflation targeting. Journal of Eco- nomic Dynamics and Control 25, 891–910.
auducco, S., Bulíř, A., Čihák, M., 2008. Taylor rule under financial instability. IMF Working Paper No. 08/18.
elke, A., Klose, J., 2010. (How) Do the ECB and the Fed react to financial market uncertainty? – The Taylor rule in times of crisis. DIW Berlin Discussion Paper No. 972.
ernanke, B., Gertler, M., Gilchrist, S., 1996. The financial accelerator and the flight to quality. The Review of Economics and Statistics 78 (1), 1–15.
ernanke, B., Gertler, M., Gilchrist, S., 1999. The financial accelerator in a quantitative business cycle framework. In: Taylor, J.B., Woodford, M. (Eds.), Handbook of Macroeconomics. Amsterdam, North-Holland.
ernanke, B., Gertler, M., 1999. Monetary policy and asset price volatility. Economic Review (FRB of Kansas City), 17–51.
ernanke, B., Gertler, M., 2001. Should central banks respond to movements in asset prices? American Economic Review 91 (2), 253–257.
oivin, J., 2006. Has U.S. monetary policy changed? Evidence from drifting coef- ficients and real-time data. Journal of Money, Credit and Banking 38 (5), 1149–1173.
orio, C., Lowe, P., 2002. Asset prices, financial and monetary stability: exploring the nexus. BIS Working Paper No. 114.
orio, C., Lowe, P., 2004. Securing sustainable price stability: should credit come back from the wilderness? BIS Working Paper No. 157.
orio, C., Disyatat, P., 2009. Unconventional monetary policies: an appraisal. BIS Working Paper No. 292.
orio, C., Drehmann, M., 2009. Towards an operational framework for financial stability: ‘fuzzy’ measurement and its consequences. BIS Working Paper No. 284.
rousseau, V., Detken, C., 2001. Monetary policy and fears of financial instability. ECB Working Paper No. 89.
M
M
Stability 9 (2013) 117– 138 137
uiter, W.H., Corsetti, G.M., Pesenti, P.A., 1998. Interpreting the ERM crisis: country- specific and systemic issues. Princeton Studies in International Economics, No. 84, Princeton University.
ulíř, A., Čihák, M., 2008. Central bankers’ dilemma when banks are vulnerable: to tighten or not to tighten? IMF mimeo.
ardarelli, R., Elekdag, S., Lall, S., 2011. Financial stress and economic contractions. Journal of Financial Stability 7 (2), 78–97.
arlson, M.A., King, T.B., Lewis, K.F., 2009. Distress in the financial sector and eco- nomic activity. Finance and Economics Discussion Series 2009-01, Board of Governors of the Federal Reserve System.
ecchetti, S., Genberg, H., Lipsky, J., Wadhwani, S., 2000. Asset prices and central bank policy. Geneva Reports on the World Economy, No. 2.
ecchetti, S.G., Li, L., 2008. Do Capital Adequacy Requirements Matter For Monetary Policy? Economic Inquiry, Western Economic Association International 46 (4), 643–659.
hadha, J.S., Sarno, L., Valente, G., 2004. Monetary policy rules, asset prices and exchange rates. IMF Staff Papers 51 (3).
hristiano, L., Ilut, C., Motto, R., Rostagno, M., 2008. Monetary policy and stock market boom-bust cycles. ECB Working Paper No. 955.
hoi, W., Cook, D., 2004. Liability dollarization and the bank balance sheet channel. Journal of International Economics 64 (2), 247–275.
larida, R., Galí, J., Gertler, M., 1998a. Monetary policy rules in practice: some inter- national evidence. European Economic Review 42, 1033–1067.
larida, R., Galí, J., Gertler, M., 1998b. Monetary policy rules and macroeconomic stability: evidence and some theory. The Quarterly Journal of Economics 115, 147–180.
úrdia, V., Woodford, M., 2010. Credit spreads and monetary policy. Journal of Money, Credit and Banking 42 (s1), 3–35.
evereux, M., Sutherland, A., 2007. Country portfolios in open economy macro models. Journal of the European Economic Association 5 (2–3), 491–499.
etken, C., Smets, F., 2004. Asset price booms and monetary policy. ECB Working Paper No. 364.
ubois, É., Michaux, E., 2009. Grocer 1.4: An Econometric Toolbox for Scilab, Avail- able at http://dubois.ensae.net/grocer.html.
ichengreen, B., Bordo, M.D., 2002. Crises now and then: what lessons from the last era of financial globalization. NBER Working Paper No. 8716.
ngel, C., Matsumoto, A., 2009. The international diversification puzzle when prices are sticky: it’s really about exchange-rate hedging not equity portfolios. Amer- ican Economic Journal: Macroeconomics 1 (2), 155–188.
aia, E., Monacelli, T., 2007. Optimal monetary policy rules, asset prices and credit frictions. Journal of Economic Dynamics and Control 31 (10), 3228–3254.
isher, I., 1933. Debt-deflation theory of great depressions. Econometrica 1 (4), 337–357.
uhrer, J., Tootell, G., 2008. Eyes on the prize: how did the Fed respond to the stock market? Journal of Monetary Economics 55 (4), 796–805.
oldstein, M., Kaminsky, G.L., Reinhart, C.M., 2000. Assessing Financial Vulnerabil- ity: An Early Warning System for Emerging Markets. Institute for International Economics, Washington, DC.
oodhart, C., 2006. A framework for assessing financial stability? Journal of Banking and Finance 30 (12), 3415–3422.
oodhart, C., Osorio, C., Tsomocos, D., 2009. An analysis of monetary policy and financial stability: a new paradigm. CESifo Working Paper No. 2885.
reenspan, A., 2007. The Age of Turbulence: Adventures in a New World. Penguin Press, New York.
ruen, D., Plumb, M., Stone, A., December 2005. How should monetary pol- icy respond to asset-price bubbles? International Journal of Central Banking, 1–31.
akkio, C.S., Keeton, W.R., 2009. Financial stress: what is it, how can it be measured, and why does it matter? Economic Review (2Q 2009), Federal Reserve Bank of Kansas City, 1–50.
acovello, M., Neri, S., 2010. Housing market spillovers: evidence from an estimated DSGE model. American Economic Journal: Macroeconomics 2 (2), 125–164.
lling, M., Liu, Y., 2006. Measuring financial stress in a developed country: an appli- cation to Canada. Journal of Financial Stability 2 (3), 243–265.
ennergren, P., 2002. The Swedish finance company crisis – could it have been antic- ipated? Scandinavian Economic History Review 50 (2), 7–30.
onung, L., 2009. Financial crisis and crisis management in Sweden. Lessons for today. Asian Development Bank Institute Working Paper No. 156.
aminsky, G., Reinhart, C., 1999. The twin crises: the causes of banking and balance- of-payments problems. American Economic Review 89 (3), 473–500.
im, C.-J., 2006. Time-varying parameter models with endogenous regressors. Eco- nomics Letters 91, 21–26.
im, C.-J., Nelson, C.R., 2006. Estimation of a forward-looking monetary policy rule: a time-varying parameter model using ex-post data. Journal of Monetary Eco- nomics 53, 1949–1966.
im, C.-J., Kishor, K., Nelson, C.R., 2006. A time-varying parameter model for a forward-looking monetary policy rule based on real-time data. Mimeo.
uttner, K.N., Posen, A.S., 1999. Does talk matter after all? Inflation targeting and central bank behavior. Federal Reserve Bank of New York Staff Report No. 88.
aeven, L., Valencia, F., 2008. Systemic banking crisis: a new database. IMF Working
Paper No. 08/224.
elvin, M., Taylor, M.P., 2009. The crisis in the foreign exchange market. Journal of International Money and Finance 28 (8), 1317–1330.
erton, R., 1974. On the pricing of corporate debt: the risk structure of interest rates. Journal of Finance 29 (2), 449–470.
1 ncial
M
O
P
R
R
R
R
R
S
S
S
S
S
S
S
T
T
T
T
T
T
T
T
T
38 J. Baxa et al. / Journal of Fina
ishkin, F., 2009. Is monetary policy effective during financial crises? American Economic Review Papers & Proceedings 99 (2), 573–577.
rphanides, A., 2001. Monetary policy rules based on real-time data. American Economic Review 91, 964–985.
osen, A., 2006. Why central banks should not burst bubbles. International Finance 9 (1), 109–124.
einhart, C.M., Rogoff, K.S., 2008. This time is different: a panoramic view of eight centuries of financial crises. NBER Working Paper No. 13882.
eis, R., 2010. Interpreting the unconventional U.S. monetary policy of 2007–09. NBER Working Paper No. 15662.
igobon, R., Sack, B., 2003. Measuring the reaction of monetary policy to the stock market. The Quarterly Journal of Economics 118 (2), 639–669.
oubini, N., 2006. Why central banks should burst bubbles. International Finance 9 (1), 87–107.
udebusch, G., 2006. Monetary policy inertia: fact or fiction? International Journal of Central Banking 2 (4), 85–136.
ekine, T., Teranishi, Y., 2008. Inflation targeting and monetary policy activism. IMES Discussion Paper Series 2008-E-13, Bank of Japan.
iklos, P., Bohl, M., 2008. Asset prices as indicators of Euro area monetary policy: an empirical assessment of their role in a Taylor rule. Open Economies Review 20 (1), 39–59.
ims, C., Zha, T., 2006. Were there regime switches in US monetary policy. American Economic Review 96 (1), 54–81.
chlicht, E., 1981. A seasonal adjustment principle and a seasonal adjustment
method derived from this principle. Journal of the American Statistical Asso- ciation 76 (374), 374–378.
chlicht, E., 2005. Estimating the smoothing parameter in the so-called Hodrick–Prescott filter. Journal of the Japan Statistical Society 35 (1), 99–119.
T
V
Stability 9 (2013) 117– 138
chlicht, E., Ludsteck, J., 2006. Variance estimation in a random coefficients model. IZA Discussion Paper No. 2031.
tock, J.H., Watson, M.W., 1998. Median unbiased estimation of coefficient vari- ance in a time-varying parameter model. Journal of the American Statistical Association 93 (441), 349–358.
aylor, J.B., 1993. Discretion versus policy rules in practice. Carnegie-Rochester Con- ference Series on Public Policy 39, 195–214.
aylor, J.B. (Ed.), 1999. Monetary Policy Rules. The University of Chicago Press, Chicago.
aylor, J.B., 2001. The role of the exchange rate in monetary policy rules. American Economic Review 91 (2), 263–267.
aylor, J.B., February 26, 2008. Monetary Policy and the State of the Economy, Testi- mony before the Committee on Financial Services, U.S. House of Representatives.
aylor, J.B., 2010. Getting back on track: macroeconomic policy lessons from the financial crisis. Federal Reserve Bank of St. Louis Review 92 (May/June (3)), 165–176.
aylor, J.B., Williams, J., 2009. A black swan in the money market. American Eco- nomic Journal: Macroeconomics 1 (1), 58–83.
aylor, M.P., Davradakis, E., 2006. Interest rate setting and inflation targeting: evi- dence of a nonlinear Taylor rule for the United Kingdom. Studies in Nonlinear Dynamics & Econometrics 10 (4) (Article 1).
eranishi, Y., 2009. Credit spread and monetary policy. IMES Discussion Paper Series 2009-E-14, Bank of Japan.
recroci, C., Vassalli, M., 2010. Monetary policy regime shifts: new evidence from
time-varying interest-rate rules. Economic Inquiry 48 (4), 933–950.
ovar, C.E., 2009. DSGE models and central banks, economics: the open-access. Open-Assessment E-Journal 3, 200916.
alente, G., 2003. Monetary policy rules and regime shift. Applied Financial Eco- nomics 13, 525–535.
- Time-varying monetary-policy rules and financial stress: Does financial instability matter for monetary policy?
- 1 Introduction
- 2 Related literature
- 2.1 Monetary policy (rules) and financial instability – some theories
- 2.2 Monetary policy (rules) and financial instability – empirical evidence
- 2.3 Measures of financial stress
- 3 Data and empirical methodology
- 3.1 The dataset
- 3.2 The empirical model
- 4 Results
- 4.1 Financial-stress effect
- 4.2 Monetary policy rule estimates
- 4.3 Robustness checks
- 5 Concluding remarks
- Acknowledgments
- Appendix 1 Time-varying monetary policy rule estimates (Figs. A1.1–A1.5)
- Appendix 2 The results with the interbank rate as the dependent variable in the policy rule
- Appendix 3 The results with different leads and lag of the FSI
- Appendix 4 The results with individual variables of bank stress and stock-market stress
- Appendix 5 Significance of the financial stress index in the estimated Taylor rules (Tables A5.1 and A5.2)
- References
36.3ballen.pdf
p t t n nd F n n l t b l t
Franklin Allen, Douglas Gale
Journal of Money, Credit, and Banking, Volume 36, Number 3 (Part 2), June 2004, pp. 453-480 (Article)
P bl h d b Th h t t n v r t Pr DOI: 10.1353/mcb.2004.0038
For additional information about this article
Access provided by Bangor University (5 Mar 2015 18:52 GMT)
http://muse.jhu.edu/journals/mcb/summary/v036/36.3ballen.html
FRANKLIN ALLEN
DOUGLAS GALE
Competition and Financial Stability
Competition policy in the banking sector is complicated by the necessity of maintaining financial stability. Greater competition may be good for (static) efficiency, but bad for financial stability. From the point of view of welfare economics, the relevant question is: what are the efficient levels of competi- tion and financial stability? We use a variety of models to address this question and find that different models provide different answers. The relationship between competition and stability is complex: sometimes com- petition increases stability. In addition, in a second-best world, concentration may be socially preferable to perfect competition and perfect stability may be socially undesirable.
JEL codes: D4, D5, D6, G2 Keywords: crises, banking concentration, dynamic, spatial,
and Schumpeterian competition.
In the banking sector, unlike other sectors of the economy, competition policy must take account of the interaction between competi- tion and financial stability. Greater competition may be good for (static) efficiency, but bad for financial stability.1 In this paper, we shall argue that the relationship between competition and financial stability is considerably more complex than this simple “trade-off” suggests, but understanding why increasing competition might reduce economic stability is a good starting point.
In an important paper, Keeley (1990) provided a theoretical framework and empirical evidence that deregulation of the banking sector in the U.S. in the
1. See Canoy et al. (2001) and Carletti and Hartmann (2003) for excellent surveys of the literature on financial stability and competition.
Prepared for the World Bank and Federal Reserve Bank of Cleveland project on Bank Concentration. Presented at the April 3–4, 2003 conference at the World Bank and the May 21–23, 2003 conference at the Federal Reserve Bank of Cleveland. We are grateful to an anonymous referee, to our discussants Stephen Haber and Charles Kahn, and to Elena Carletti, Qian Liu, Lemma Senbet, Andrew Winton, and participants in the conferences.
Franklin Allen is a professor of finance in the Department of Finance, Wharton School, University of Pennsylvania. E-mail: allenf�wharton.upenn.edu Douglas Gale is a professor in the Department of Economics, New York University. E-mail: douglas.gale�nyu.edu
Received May 29, 2003; and accepted in revised form May 29, 2003.
Journal of Money, Credit, and Banking, Vol. 36, No. 3 (June 2004, Part 2) Published in 2004 by The Ohio State University Press.
454 : MONEY, CREDIT, AND BANKING
1970s and 1980s had increased competition and led to a reduction in monopoly rents. This reduction in “charter value” magnified the agency problem between bank owners and the government deposit insurance fund. The bank owners or managers acting on their behalf had an increased incentive to take on extra risk, given the guaranteed funds available to them because of deposit insurance. As in the agency problem identified by Jensen and Meckling (1976), if the gamble was successful the equity owners would obtain the rewards while if it was unsuccessful the cost would be born by the deposit insurance fund. The extra risk that banks took on as a result of this agency problem caused a dramatic increase in bank failures during the 1980s. The US is not the only country where there appears to be an empirical relationship between increased competition and financial instability. Beck, Demir- guc-Kunt, and Levine (2003) find using data from 79 countries that crises are less likely in more concentrated banking systems.
Various empirical studies have found that the cost of financial instability is high. For example, Hoggarth and Saporta (2001) find that the average fiscal costs of banking resolution across countries are 16% of GDP. For emerging countries the figure is 17.5% and for developed countries it is 12%. As Table 1 shows, the costs of banking crises alone are estimated at 4.5% of GDP. Although these costs are substantial, they are much lower than the costs, estimated at 23% of GDP, of banking and currency crises occurring together. A proportion of the fiscal costs are transferred, so these figures do not represent the deadweight economic costs. A number of studies measure the cumulative output loss resulting from a financial crisis by using the deviation from trend output. Table 2 gives estimates for these costs. The average cumulative output loss for all crises is 16.9% of GDP. Here the costs of twin crises are again higher; the loss caused by twin banking and currency crises is 29.9% of GDP versus 5.6% for banking crises alone. However, in contrast to fiscal costs,
TABLE 1
Average Cumulative Fiscal Costs of Banking Crises in 24 Crises, 1977–2000
Non-performing loans Fiscal costs of banking resolution Number of crises (percentage of total loans) (percentage of GDP)
All countries 24 22 16 Emerging market countries 17 28 17.5 Developed countries 7 13.5 12 Banking crisis alone 9 18 4.5 Banking and currency crises of which 15 26 23 Emerging market countries 11 30 25 Developed countries 4 18 16 Banking and currency crises with 11 26 27.5
previous fixed exchange rate of which
Emerging market countries 8 30 32 Developed countries 3 18 16
Note: Source: Hoggarth and Saporta (2001, p. 150).
FRANKLIN ALLEN AND DOUGLAS GALE : 455
TABLE 2
Output Losses Associated with Banking Crises, 1977–98
Average crisis length Average cumulative output losses Number of crises (years) (percentage of GDP)
All 43 3.7 16.9 Single banking crises 23 3.3 5.6 Twin banking and currency crises 20 4.2 29.9 Developed countries 13 4.6 23.8 Emerging market countries 30 3.3 13.9
Note: Source: Hoggarth and Saporta (2001, p. 155).
developed countries have a greater loss, 23.8% of GDP, than emerging countries, 13.9% of GDP.
The large literature on the efficiency of the banking industry (for a survey see, e.g., Berger and Humphrey, 1997) is mostly concerned with the cost- and profit- efficiency of retail banking. For example, Canoy et al. (2001) summarize the evidence as suggesting that the average bank operates at a cost level that is 10% or 20% above the best-practices level. This is just one (probably small) part of the total costs of deviations from perfect competition. Unfortunately, the total costs of a deviation from perfect competition have not been documented as carefully as the costs of financial instability.
Given the large and visible costs of financial instability, it is natural for policymak- ers to make the avoidance of financial crises a high priority. By contrast, the difficulty of measuring the efficiency costs of concentration may suggest that competition policy warrants a lower priority. In fact, the uncertainty about the costs of concentra- tion together with the perceived (negative) trade-off between competition and finan- cial stability may actually encourage policymakers to favor concentration at the expense of competition policy. This subordination of competition policy to financial stability may be unwise for a number of reasons, however. In the first place, the extent to which there is a negative trade-off between competition and financial stability may be questioned. The costs of financial crises are undoubtedly high, but it does not follow that it is necessary to reduce competition to avoid those costs. Secondly, the wide range of estimates of the efficiency costs from concentration is at least consistent with a high efficiency gain from greater competition. Thirdly, the costs of financial crises occur infrequently, perhaps every decade or few decades, whereas the inefficiency cost concentrations are born continuously.
The proper balance between competition and financial stability presupposes a framework in which we can identify the welfare costs and benefits of different levels of competition and financial stability. Our objective in this paper is to review a number of theoretical models as a prelude to the development of a theoretical framework in which the optimal policy can be identified. From the point of view of welfare economics, the relevant question is: What are the efficient levels of competition and financial stability? We use a variety of models to address this
456 : MONEY, CREDIT, AND BANKING
question and find that different models provide different answers. This should not be surprising. In a second-best world, concentration may be preferred to perfect competition (Schumpeter 1950) and perfect stability may be socially undesirable (Allen and Gale 1998). When we consider the relationship between competition and stability, we find that the idea of a simple negative trade-off is, again, too simple: sometimes competition decreases stability and sometimes perfect competition is compatible with the socially optimal level of stability.
We begin in Section 1 by describing the general equilibrium model of financial intermediaries and markets from Allen and Gale (2003a). They provide analogues of the classical theorems of welfare economics for a model of intermediation with asymmetric information. If financial markets are complete and contracts between intermediaries and their customers are complete, the perfectly competitive equilib- rium allocation is incentive-efficient. In this sense, perfect competition is socially optimal. There is no financial instability because contracts are completely contingent and hence there is no need to default. Similarly, if contracts are incomplete, the perfectly competitive equilibrium allocation is constrained-efficient, but now finan- cial instability is necessary for efficiency. If the banks cannot meet the fixed payments that they have promised, there is a financial crisis. A deviation from competition may increase financial stability, but cannot increase and is likely to reduce welfare. This result illustrates that, in general, there should be no presumption that reducing competition in order to increase financial stability is socially desirable.
In simple partial-equilibrium models, it is possible to generate a negative trade-off between competition and financial stability. However, even in this case, the nature of the trade-off between competition and stability is more complicated than was first thought. For example, Allen and Gale (2000a, chap. 8), Boyd and De Nicolo (2002), and Perotti and Suarez (2003) have identified a number of different effects of increased competition on financial stability. In some circumstances, increased competition can actually increase financial stability. These models are discussed in Section 2.
Introducing other kinds of frictions produces further complications to our picture of competition. On the one hand, as mentioned above, Allen and Gale (2000a, chap. 8) show that when search costs are introduced, competition among a large number of unitary banks may result in the monopoly price being charged. On the other hand, a system with two banks with branches at each location may result in the perfectly competitive price. A Hotelling-type model of spatial competition introduces a rich variety of effects concerning regional diversification and risk sharing. The profitability of banks is shown to be extremely sensitive to the precise form of local interactions.
Section 4 describes a model of Schumpeterian competition, in which firms compete by developing new products. The firm that makes the best innovation manages to capture the whole market. The equilibrium price equals the difference between the value of the successful firm’s product and the value of the second-best product. So the successful firm’s profit is equal to the social value of its innovation. This provides the firms with the right incentives to innovate efficiently. In this context, perfect competition is again desirable and can lead to efficiency. Clearly, this kind
FRANKLIN ALLEN AND DOUGLAS GALE : 457
of innovation process is not consistent with financial stability. The successful innova- tor will survive while the unsuccessful will fail. Again, as in the benchmark model, efficiency requires a combination of perfect competition and financial instability. If the government is concerned with financial stability it may ensure that banks survive by regulating entry in different submarkets. We consider a setting where banks are assured of a monopoly in their region and consider the incentives to innovate. It is shown that this enforced stability leads to a welfare loss as might be expected. Less obvious is the result that there may be too little or too much investment in innovation. There can be too little because each bank only obtains profits from its own region. There can be too much because the bank is assured of a return no matter what happens.
Contagion is another important source of financial instability. It occurs when some shock, possibly small, spreads throughout the financial system and causes a systemic problem. Allen and Gale (2000b) developed a model of contagion with a perfectly competitive banking sector. It was shown that a shock that was arbitrarily small relative to the economy as a whole could cause all the banks in the financial system to go bankrupt. The contagion spreads through the interbank market. Section 5 extends the Allen and Gale (2000b) model of contagion to allow for imperfect competition in the banking sector. It is shown that in this case the economy is not as susceptible to contagion as it is with perfect competition. Each oligopolistic bank realizes that its actions affect the price of liquidity. By providing sufficient liquidity to the market they can ensure that contagion and their own bankruptcy are avoided. In this case there is a trade-off between competition and stability.
Concluding remarks are contained in Section 6.
1. COMPETITION AND CRISES
In the Arrow–Debreu model of general equilibrium, the fundamental theorems of welfare economics show that perfect competition is a necessary condition for efficiency. Allen and Gale (2003a) show that analogous results hold for a model of financial crises with complete markets. In this setting, perfect competition is compati- ble with the efficient level of financial stability. In this sense, there is no “trade- off” between competition and stability. We begin by describing the model of perfect competition in a financial system consisting of financial intermediaries and markets and then summarize the theoretical results in Allen and Gale (2003a).
There are three dates t � 0,1,2 and a single good at each date. The good is used for consumption and investment.
The economy is subject to two kinds of uncertainty. First, individual agents are subject to idiosyncratic preference shocks, which affect their demand for liquidity (these will be described later). Second, the entire economy is subject to aggregate shocks that affect asset returns and the cross-sectional distribution of preferences. The aggregate shocks are represented by a finite number of states of nature. All agents have a common prior probability density over the states of nature. All
458 : MONEY, CREDIT, AND BANKING
uncertainty is resolved at the beginning of date 1, when the aggregate state is revealed and each agent discovers his/her individual preference shock.
Each agent has an endowment of one unit of the good at date 0 and no endowment at dates 1 and 2. So, in order to provide consumption at dates 1 and 2, they need to invest.
There are two assets distinguished by their returns and liquidity structure. One is a short-term asset (the short asset), and the other is a long-term asset (the long asset). The short asset is represented by a storage technology: one unit invested in the short asset at date t � 0,1 yields a return of one unit at date t � 1. The long asset yields a return after two periods. One unit of the good invested in the long asset at date 0 yields a random return of more than one unit of the good that depends on the aggregate state at date 2.
Investors’ preferences are distinguished ex ante and ex post. At date 0 there is a finite number n of types of investors, indexed by i � 1,…,n. We call i an investor’s ex ante type. An investor’s ex ante type is common knowledge and hence contractible.
While investors of a given ex ante type are identical at date 0, they receive a private, idiosyncratic, preference shock at the beginning of date 1. The date 1 preference shock is denoted by θi � Θi, where Θi is a finite set. We call θi the investor’s ex post type. Because θi is private information, contracts cannot be explicitly contingent on θi.
Investors only value consumption at dates 1 and 2. An investor’s preferences are represented by a von Neumann–Morgenstern utility function, ui(c1, c2; θi), where ct
denotes consumption at date t � 1,2. The utility function ui(·; θi) is assumed to be concave, increasing, and continuous for every type θi. Diamond and Dybvig (1983) assumed that consumers were one of two ex post types, either early diers who valued consumption at date 1 or late diers who valued consumption at date 2. This is a special case of the preference shock θi. The present framework allows for much more general preference uncertainty.
Allen and Gale (2003a) consider two different versions of the model, depending on the kind of contracts financial institutions offer to their customers. In the first version, contracts are completely contingent, subject only to incentive-compatibility constraints. More precisely, contracts are required to be incentive-compatible and are allowed to be contingent on the aggregate states η and individuals’ reports of their ex post types. Each intermediary offers a single contract and each ex ante type is attracted to a different intermediary.
One can, of course, imagine a world in which a single “universal” intermediary offers contracts to all ex ante type of investors. A universal intermediary could act as a central planner and implement the incentive-efficient allocation of risk. There would be no reason to resort to markets at all. Our world view is based on the assumption that transaction costs preclude this kind of centralized solution and that decentralized intermediaries are restricted in the number of different contracts they can offer. This assumption provides a role for financial markets in which financial intermediaries can share risk and obtain liquidity.
FRANKLIN ALLEN AND DOUGLAS GALE : 459
At the same time, financial markets alone will not suffice to achieve optimal risk sharing. Because individual economic agents have private information, markets for individual risks are incomplete. The markets that are available will not achieve an incentive-efficient allocation of risk. Intermediaries, by contrast, can offer individuals incentive-compatible contracts and improve on the risk sharing provided by the market.
In the Diamond and Dybvig (1983) model, all investors are ex ante identical. Consequently, a single representative bank can provide complete risk sharing and there is no need for markets to provide cross-sectional risk sharing across banks. Allen and Gale (1994) showed that differences in risk and liquidity preferences can be crucial in explaining asset prices. This is another reason for allowing for ex ante heterogeneity.
In the context of intermediaries with complete markets and complete contingent incentive-compatible contracts, Allen and Gale (2003a) prove the following result.
Proposition 1: Under the maintained assumptions, the equilibrium allocation of the model with complete markets and incomplete contracts is constrained-efficient.
Proposition 1 assumes that intermediaries use complete, incentive-compatible contracts. In reality, we do not observe such complex contracts, for reasons that are well documented in the literature, including transaction costs, asymmetric infor- mation, and the nature of the legal system. These frictions can justify the use of debt and many other kinds of incomplete contracts that intermediaries use in practice. The second version of the model presented by Allen and Gale (2003a) assumes that intermediaries are restricted to using a set of incompletely contingent contracts. This framework allows for many special cases, including at one extreme the earlier model with completely contingent contracts and completely non-contingent debt contracts. Note that this framework allows for a wide variety of assumptions about what is feasible, but takes the set of feasible contracts as given. To endogenize the set of feasible contracts one would have to appeal to factors such as transaction costs, non-verifiable information, and so on.
When contracts are complete, there is no incentive for intermediaries to enter into commitments that they cannot carry out. When contracts are constrained to be incomplete, it may be (ex ante) optimal for the intermediary to plan to default in some states. In the event of default, it is assumed that the intermediary’s assets, including the Arrow securities it holds, are liquidated and the proceeds distrib- uted among the intermediary’s investors. For markets to be complete, which is an assumption we maintain here, the Arrow securities that the bank issues must be default free. Hence, we assume that these securities are collateralized and their holders have priority. Anything that is left after the Arrow security holders have been paid off is paid out pro rata to the depositors. Allen and Gale (2003a) demonstrate the following result for the case when contracts are incomplete and take the form of deposit contracts.
Proposition 2: Under maintained assumptions, the equilibrium allocation of the model with complete markets and incomplete contracts is constrained-efficient.
460 : MONEY, CREDIT, AND BANKING
This is an important result. It shows that, in the presence of complete markets and perfect competition, the incidence of default is optimal in a laisser-faire equilib- rium. There is no scope for welfare-improving government intervention to prevent financial crises. In fact competition and financial instability are both necessary for constrained efficiency.
This result demonstrates that in a standard framework achieving optimality does not require trading off competition and financial stability. As we will see below this result extends to a number of other circumstances.
It is important to stress that the results in this section are simply benchmarks to illustrate what may happen. The only costs modeled are the losses to consumers from inefficient risk sharing. Many features that may be important in practice, such as unemployment and bankruptcy costs to firms are excluded. When these are taken into account, there may be a role for government intervention to reduce the incidence of financial crises. What the results do show is that the operation of the financial system and the occurrence of crises when there are complete markets are not the problem. There must be some form of market failure for financial crises to be undesirable.
2. AGENCY COSTS
Keeley (1990) developed a simple model of risk taking by banks with two dates and two states when there is deposit insurance. He showed that as competition increased risk taking by banks also increased. In fact, deposit insurance is not necessary for this effect to be present although it does exacerbate it. Allen and Gale (2000a, chap. 8) developed a simple model of competition and risk taking to illustrate the agency problem.
When firms are debt-financed, managers acting in the shareholders’ interests have an incentive to take excessive risks, because the debtholders bear the downside risk while the shareholders benefit from the upside potential. This well known problem of risk shifting is particularly acute in the banking sector where a large proportion of the liabilities are in the form of debt (deposits). The risk-shifting problem is exacerbated by competition. Other things being equal, greater competition reduces the profits or quasi-rents available to managers and/or shareholders. As a result, the gains from taking excessive risks become relatively more attractive and this increases the incentive to exploit the non-convexity in the payoff function. Any analysis of the costs and benefits of competition has to weigh this effect against the supposed efficiency gains of greater competition.
To illustrate these ideas, consider the problem faced by a banking regulator who controls entry into the banking industry by granting charters to a limited number of banks. We use a model of Cournot competition, in which banks choose the volume of deposits they want, subject to an upward sloping supply of funds schedule. Having more banks will tend to raise the equilibrium deposit rate and increase the tendency to shift risks. What is the optimal number of charters?
FRANKLIN ALLEN AND DOUGLAS GALE : 461
Should the regulator restrict competition by granting only a few charters or encourage competition by granting many?
2.1 A Static Model
Suppose that the regulator has chartered n banks, indexed i � 1,…,n. Each bank chooses a portfolio consisting of perfectly correlated risks. This as-
sumption is equivalent to assuming that the risk of each investment can be decom- posed into a common component and a purely idiosyncratic component. If there is a very large number of investments, the purely idiosyncratic components can be pooled perfectly. Then the idiosyncratic risks disappear from the analysis and we are left with a common component representing the systematic risks.
A portfolio is characterized by its size and rate of return. The bank’s investments have a two-point return structure: for each dollar invested, bank i will receive a return yi with probability p(yi); with probability (1 – p(yi)) they pay a return 0. The bank chooses the riskiness of its portfolio by choosing the target return yi on its investments. The function p(yi) is assumed to be twice continuously differentiable and satisfies
p(0) � 1, p(ȳ) � 0, and p′(yi) � 0, p″(yi) ≤ 0, ∀0 � yi � ȳ .
The higher the target return, the lower the probability of success and the more rapidly the probability of success falls. Because the investments have perfectly correlated returns, the portfolio return has the same distribution as the returns to the individual investments.
Let di ≥ 0 denote the total deposits of bank i, which is by definition the total number of dollars the bank has to invest. (For the moment, we ignore bank capital.) There is an upward sloping supply-of-funds curve. If the total demand for deposits is D � �i
di , then the opportunity cost of funds is R(D), where R(D) is assumed to be a differentiable function satisfying
R′(D) � 0, R″(D) � 0, R(0) � 0 and R(∞) � ∞ .
We assume that all deposits are insured, so the supply of funds is independent of the riskiness of the banks’ portfolios. In the sequel, we consider the case where banks bear the cost of deposit insurance.
The payoff to bank i is a function of the riskiness of its own portfolio and the demand for deposits of all the banks
πi(y, d) � p(yi)[yidi � R(D)di] ,
where d � (d1,…,dn) and y � (y1,…,yn). Note that we have ignored the cost of deposit insurance to the bank in calculating its net return.
Since a bank can always ensure non-negative profits by choosing di � 0, it will always earn a non-negative expected return in equilibrium, that is, yidi � R(D)di ≥ 0. There is no need to introduce a separate limited-liability constraint.
In a Nash–Cournot equilibrium, each bank i chooses an ordered pair (yi, di) that is a best response to the strategies of all the other banks. Consider an equilibrium
462 : MONEY, CREDIT, AND BANKING
(y, d) in which each bank i chooses a strictly positive pair (yi, di)[0. As a necessary condition for a best response, this pair must satisfy the following first-order conditions
p(yi)[yi � R(D) � R′(D)di] � 0 ,
p′(yi)[yi � R(D)]di � p(yi)di � 0 .
Assuming that the equilibrium is symmetric, that is, (yi, di) � (y, d) for every i, the first-order conditions reduce to
y � R(nd) � R′(nd)d � 0 ,
p′(y)[y � R(nd)] � p(y) � 0 ,
and this implies that
� p(y) p′(y)
� y � R(nd) � R′(nd)d .
Given our assumptions on p(y), an increase in y reduces � p(y)�p′(y). Suppose that there are two symmetric equilibria, (y, d) and (y′, d′). Then y � y′ implies that R′(nd)d � R′(nd′)d′ , which, given our assumptions on R(D), implies that d � d′ and R(nd) � R(nd′). Then clearly y � R(nd) � y′ � R(nd′), contradicting the first equa- tion. So there is at most one solution to this set of equations, which determines both the size and the riskiness of the banks’ portfolio in a symmetric equilibrium.
Proposition 3: Under the maintained assumptions, there is at most one symmetric equilibrium (y*, d*)[0 which is completely characterized by the conditions
� p(y) p′(y)
� y � R(nd) � R′(nd)d .
What can we now say about the effect of competition on risk taking? Suppose that we identify the degree of competitiveness of the banking sector
with the degree of concentration. In other words, the larger the number of banks, the more competitive the banking sector is. So a first attempt at answering the question would involve increasing n ceteris paribus and observing how the riskiness of the banks’ behavior changes.
With a fixed supply-of-funds schedule R(·) it is most likely that the volume of deposits will remain bounded as n increases. More precisely, if we assume that R(D) → ∞ as D → ∞ then it is clear that D → ∞ is inconsistent with equilibrium. Then the equilibrium value of D is bounded above (uniformly in n) and this implies that d ≡ D�n → 0 as n → ∞. This in turn implies that R′(nd)d → 0 from which it immediately follows that y � R(nd) → 0 and p(y) → 0 , or in other words, that y and R(nd) both converge to ȳ.
Proposition 4: If R(D) → ∞ as D → ∞ then in any symmetric equilibrium, y � R(nd) → 0 and y → ȳ as n → ∞.
The effect of increasing competition is to make each bank much smaller relative to the market for funds and this in turn reduces the importance of the price effect
FRANKLIN ALLEN AND DOUGLAS GALE : 463
(the R′(nd)d term) in the bank’s decision. As a result, banks behave more like perfect competitors and will increase their business as long as profits are positive. Equilibrium then requires that profits converge to zero, and this in turn implies that banks have extreme incentives for risk taking. In the limit as n → ∞, they will choose the riskiest investments possible in an attempt to earn a positive profit.
The effect of replicating the market. This exercise is enlightening but somewhat artificial since it assumes that we are dealing with a market of fixed size and increasing the number of banks without bound in order to achieve competition. Normally, one thinks of perfect competition as arising in the limit as the number of banks and consumers grows without bound. One way to do this is to replicate the market by shifting the supply-of-funds function as we increase the number of banks. Precisely, suppose that the rate of return on deposits is a function of the deposits per bank
R � R(D�n) .
In effect, we are assuming that, as the number of banks is increased, the number of depositors is increased proportionately, so that the supply of funds in relation to a particular bank is unchanged.
The effect of this change in the model is to make it more like the traditional model of a market in which, as the number of firms increases, the effect of any firm supply on the price of the product becomes vanishingly small. Here the effect of any bank’s demand for deposits on the equilibrium deposit rate becomes vanishingly small in the limit as the number of banks becomes unboundedly large. To see this, note that the first-order conditions become
y � R(d) � R′(d)dn�1 � 0
and
p′(y)[y � R(d)] � p(y) � 0 .
As before, we can ensure that d remains bounded as n → ∞ by assuming that R(d) → ∞ as n → ∞. Then the last term on the left hand side of the first equation will vanish as n → ∞ , leaving a limiting value of (y, d) that satisfies y � R(d). Substituting this in the second equation tells us that p(y) � 0. In other words, as the number of banks increases, the profit margins fall to zero, with the result that banks choose riskier and riskier investments.
Proposition 5: If R(D) → ∞ as D → ∞ then in any symmetric equilibrium, y � R(d) → 0 and y → ȳ as n → ∞.
This is a highly stylized model, so the results have to be taken with a grain of salt; nonetheless, they illustrate clearly the operative principle, which is that competi- tion, by reducing profits, encourages risk taking.
In this particular case, we have constant returns to scale in banking, so that in the limit, when there is a large number of individually insignificant banks, profits must converge to zero. In other words, banks will expand the volume of their
464 : MONEY, CREDIT, AND BANKING
deposits and loans until the deposit rate approaches the expected return on invest- ments. But this gives them an extreme incentive to shift risks to the depositors or the deposit insurance agency, since it is only by doing so that they can get positive profits at all.
With constant returns to scale, zero profit is always a necessary condition of equilibrium in a competitive industry. However, there are other ways of ensuring the same outcome even if constant returns to scale is not assumed. We replicated the market by increasing the number of banks and potential depositors in the same proportion. This is an interesting thought experiment, but it is not the same as the comparative static exercise the regulator is undertaking. Presumably, the regulator has to choose n optimally, taking as given the supply-of-funds schedule. Suppose that m is the number of depositors and n the number of banks. Then the market supply-of-funds schedule can be written as R(D/m) if R(·) is the individual supply- of-funds schedule. When m is very large, the supply of funds is elastic, other things being equal, so the banks will take the marginal cost of funds as being equal to the average cost R(D/m). However, increasing the number of banks n in relation to m will force profits down. If d remains bounded away from zero, the average cost of funds must increase to ∞ and if d goes to zero, profits will also go to zero. In this way, the regulator can achieve the effects of free entry, but there is no need to do this in order to ensure competition. Competition, in the sense of price-taking behavior, follows from having a large market, that is, a large value of m, independent of whether n is large or not. Clearly, the regulator does not want to drive profits to zero if it can be helped, because of the incentives for risk taking that that creates.
Cost of deposit insurance. The preceding analysis assumes that all deposits are insured and that the costs are not born by the banks. This is clearly unrealistic, so it makes sense to consider explicitly the cost of deposit insurance. We assume that the premium for deposit insurance is set before the banks choose their strategies and that it is the same for each bank, independently of the strategy chosen. In equilibrium, the premium accurately reflects the cost of deposit insurance provided by a risk neutral insurer.
Let π denote the premium per dollar of deposits. Then the objective function of bank i is p(yi)(yi � R(D) � p)di and the first-order conditions in a symmetric equilib- rium in which banks choose the strategy (y, d) will be
y � R(d) � π � R′(d)dn�1 � 0 ,
p′(y)[y � R(d) � π] � p(y) � 0 .
In equilibrium, the premium must be set so that the expected return on deposits is equal to the return demanded by depositors
R(d) � p(y)(R(d) � π) .
Substituting π � [(1 � p(y))�p(y)]R(d) into the first-order conditions yields
y � R(d)�p(y) � R′(d)dn�1 � 0 ,
p′(y)[y � R(d)�p(y)] � p(y) � 0 .
FRANKLIN ALLEN AND DOUGLAS GALE : 465
Let (yn, dn) be a symmetric equilibrium when there are n banks and suppose that (yn, dn) → (y0, d0) as n → ∞. Then the first-order conditions imply that
lim n→∞
yn � R(dn)�p(yn) � 0 ,
which is only possible if yn → ȳ and p(yn) → 0, as before. Efficiency. Let us leave distributional questions on one side for the moment,
although historically they have been at the center of the arguments for competition in banking, and suppose that the regulator is only interested in maximizing surplus. Reverting to the constant-returns-to-scale case, two necessary conditions for Pareto optimality are that the average cost of funds be equal to the expected return on investments, and that the expected return on investments should be a maximum
R(D�m) � p(y)y ,
p(y)y ≥ p(y′)y′, ∀y′�[0, ȳ] .
Neither of these conditions will hold in equilibrium when m and n are very large. The first condition requires that the volume of deposits expand until the cost of funds equals the expected value of investments. However, when the market is highly competitive, we have y � R(D�m) ≅ 0 , so that p(y)y � R(D�m) � 0. As we also saw, the second condition cannot be satisfied in equilibrium, since as the market grows large (m, n → ∞), we have y → ȳ � ∞ and p(y) → 0, so p(y)y → 0. This is not only sub-optimal but the worst possible outcome because it minimizes the total surplus.
Suppose instead that we hold the value of m fixed and adjust n to maximize total surplus, taking the equilibrium values (y(n), d(n)) as given functions of n determined by the equilibrium conditions
y � R(nd�m) � R′(nd�m)dm�1 � 0
and
p′(y)[y � R(nd�m)] � p(y) � 0 .
A “small change” in n will increase the expected revenue by
{p′(y(n))y(n) � p(y(n))}ny′(n) � p(y(n))y(n) ,
and the cost by
R(nd(n)�m){nd′(n) � d(n)} ,
so a necessary condition for an (interior) optimum is
{p′(y(n))y(n) � p(y(n))}ny′(n) � p(y(n))y(n) � R(nd(n)�m){nd′(n) � d(n)} .
This can be rewritten as
p(y(n))y(n) � R(nd(n)�m)d(n) � �{p′(y(n))y(n) � p(y(n))}ny′(n) � R(nd(n)�m)nd′(n) ,
466 : MONEY, CREDIT, AND BANKING
where the left hand side is the expected surplus generated by a single bank and the right hand side is the change in expected revenue per bank as n increases plus the change in the cost per bank of the funds borrowed. Since the left hand side is positive, the right hand side must be positive, too. But we know that the second term on the right must be negative since adding more banks reduces the volume of business each bank does, so the first term on the right is positive. We know that y′(n) is negative—increased competition leads to increased risk taking—so the term in braces must be positive. Assuming that p(y)y is concave in y, this tells us that n will be chosen so that y(n) is less than the value that maximizes expected revenue.
A fortiori, it will not, as we have seen, be as great as the value under free entry, since it is never optimal to let n → ∞. It may in fact, be optimal to let the number of banks remain quite small.
2.2 Loan Market Competition
The model in Allen and Gale (2000a) analyzes competition in the deposit market. The bank is assumed to invest deposits directly in a portfolio of assets with given risk characteristics. The bank directly determines the riskiness of its portfolio. As profits decline the bank’s preference for risk increases, so increasing competition leads to increasing risk and a decrease in stability. Boyd and De Nicolo (2002) point out that the assumption that banks invest directly in assets is crucial for the result. To show this, they extend the Allen and Gale model to include entrepreneurs. The entrepreneurs obtain loans from the banks and invest the money in risky ventures. Each entrepreneur chooses the riskiness of the venture he invests in. The entrepre- neurs, like the banks in the Allen and Gale model, have a greater incentive to take risk when profits are lower. However, the effect of competition among banks here is the opposite of what we observed in the Allen and Gale (2000a) model. Greater competition among banks reduces the interest rates that borrowers pay, increases the profitability of their ventures, and hence reduces the incentive to take risk. Thus, increased competition among banks leads to increased financial stability. The effect of competition in the deposit market is the same as before but Boyd and De Nicolo are able to show that the loan market effect dominates. The trade-off between competition and stability presented in the Allen and Gale model is reversed in the Boyd and De Nicolo model. As competition between banks increases the risks taken by borrowers is unambiguously reduced and financial stability is improved.
2.3 Dynamic Competition
The results in Section 2.1 demonstrate how the limited liability of managers and shareholders in a modern banking corporation can produce a convex objective function which in turn leads to risk-shifting behavior. This kind of behavior is most likely to occur when the bank is “close to the water line,” that is, when the risk of bankruptcy is imminent. For banks which are not in immediate danger of bankruptcy, the risk-shifting argument may be less relevant. However, even if a bank is not close to the water line, there may be other reasons for thinking that its objective
FRANKLIN ALLEN AND DOUGLAS GALE : 467
function is convex. Consider, for example, the winner-takes-all nature of competition. When banks compete for market share, the bank that ends up with the largest share may be able to exploit its market power to increase profitability. In this case, the profit function may be convex in market share, that is, doubling market share may more than double profits. Another reason is the presence of increasing returns to scale. If larger banks have lower average costs, then profits will be a convex function of the size of the bank. Either of these possibilities will give the bank an incentive to take riskier actions, even when the bank is not in immediate danger of bankruptcy.
These incentives for risk-taking behavior are naturally studied in a dynamic context. Suppose that a group of banks are competing over time. Their activities are constrained by the minimum capital ratio, so the only way to expand is to acquire more capital. Because of agency costs or the adverse signaling effects, it may be expensive for banks to raise capital from external sources, so they try to accumu- late capital by retaining earnings. The relative size of the bank matters, because it gives the bank a competitive edge over other banks. Because their reduced-form profit functions are convex and they are constrained by their capital, the game is a race to see who can accumulate capital or market share fastest.
When banks compete to capture greater market share or to reap economies of scale, they consider the effect of their actions not only on immediate profits but also on their future position in the market. How the bank’s current actions will affect its future position in the market depends on the nature of the risks involved and on the behavior of the other banks. Even if the profit function is only convex when the bank is close to the bankruptcy point, the bank’s objective function may be convex over a much wider region because the bank’s objective function incorporates or discounts future possibilities which are still far away. This may influence the shape of the bank’s objective function globally, through backward induction.
Allen and Gale (2000a, chap. 8) consider a variety of different models and show that risk taking can either be increased or decreased by competition in a dynamic setting. The first case analyzed involves a pair of duopolists who compete for market share. They play repeatedly for many periods. In each round market shares can go up or down a small amount. By taking a risky action instead of a safe action in any period they increase the variance of the change in market share. A crucial assumption is that there are “reflecting barriers” at the extreme values of market share. When a bank’s market share hits zero, it does not go out of business; at worst it will remain at zero for some period before bouncing back. This non-convexity is like having increasing returns in the neighborhood of zero. Similarly, when a bank’s market share hits 100% it must eventually bounce back, and this is like having locally decreasing returns. A simple numerical example is used to illustrate the effect of non-convexity on the bank’s behavior. When a bank’s market share is low, its objective function is convex and it has an incentive to take risk; when its market share is high, its objective function is concave and it has an incentive to avoid risk.
The second case analyzed assumes that there are “absorbing barriers.” Now when market share hits zero or one it stays the same with probability one. In this case it is shown that the incentive to take risks is eliminated. The reason is the assumption
468 : MONEY, CREDIT, AND BANKING
that the bank’s position can only change a little bit at a time. If the period length and the step size are made very small, the binomial process considered would approximate Brownian motion, which has continuous sample paths with probability one. It is the continuity of the movement of the bank’s market share over time which eliminates the usual incentive for risk taking. The bank becomes “bankrupt” as soon as its market share hits zero. It cannot go below the line and so it cannot shift risk to depositors or other creditors. Whether incentives to take risks are greater or less with reflecting barriers is ambiguous. While absorbing barriers eliminate the positive incentive for risk taking with reflecting barriers, they also eliminate the in- centive to avoid risk when market share is high that was found in the same model.
Perotti and Suarez (2002) consider the effect of dynamic competition in a model where there can be a banking duopoly or monopoly. If a bank fails when there is a duopoly the market structure switches temporarily to a monopoly. Banks can lend prudently in which case their portfolio of loans has a safe payoff. The alternative is to lend speculatively in which case there is some probability of a high payoff but the average payoff is low. Limited liability and deposit insurance mean lending speculatively can shift risks and be advantageous in the short run. However, a bank can be hit by a random solvency shock. If it lent prudently it will always survive this shock but if it lent speculatively it will fail unless loan returns are high. It is shown that this effect introduces an incentive for banks to lend prudently. A prudent bank will be less exposed to the risk of being driven out of business and will emerge as a monopolist if the other duopolist lent speculatively and is hit by a solvency shock. In their model this “last one standing effect” makes duopolistic banks unam- biguously more prudent and encourages stability.
The range of results obtained with these dynamic models illustrate how crucial the particular details of the model are in determining whether or not competition leads to more or less financial stability.
3. SPATIAL COMPETITION
So far, we have only considered competition in markets for homogeneous com- modities or services. Product differentiation occurs in the financial sector, just as it does in non-financial sectors, and it is important to consider the effect of product differentiation on the competitive process. Models of spatial competition are used to represent competition among firms with differentiated products and the same can be done for the banking sector. We can interpret the spatial dimension literally as representing banks with different locations or we can interpret it metaphorically as representing some other qualitative difference in the services provided. In either case, we find that the effects of concentration on competition in spatial models can be quite different from the results obtained for markets with undifferentiated products. In this section we consider two models of bank competition. The first model, from Allen and Gale (2000a), compares branch banking with unitary banking and shows that competition among a small number of banks (with many branches) may be
FRANKLIN ALLEN AND DOUGLAS GALE : 469
more aggressive than competition among a large number of unitary banks. We then consider a Hotelling-type model of spatial competition and again compare competi- tion among a small number of banks (with many branches) with competition among a large number of unitary banks. The results concerning the trade-off between competition and diversification are very sensitive to the spatial arrangement of the branches.2
3.1 Unitary Banking versus Branch Banking
The model presented below exploits two types of imperfections arising from asymmetric information. The first is the presence of “lock-in” effects. Information is costly for both banks and their customers, whether borrowers or depositors, and once relationship-specific investments in information have been made, the parties may find themselves “locked-in” to the relationship. For example, a borrower having incurred a fixed cost of revealing its type to a bank will suffer a loss if it switches to another bank. In addition there is the well known “lemons effect,” that arises if the borrower leaves the bank with which it has been doing business for many years. These lock-in effects will be modeled by simply assuming that there is a fixed cost of switching banks. As is well known (see, e.g., Diamond 1971), switching costs give the bank a degree of monopoly power, even if the bank is not “large.”
There are other reasons why banks are monopolistic competitors. For locational reasons their services are not perfect substitutes. Differences in size and products and specialized knowledge also make them imperfect substitutes. Here, we focus on the lock-in effect.
The second essential imperfection arises from the fact that a bank’s customers have incomplete information about the services offered by a bank and the prices at which these services are offered, at the time when the relationship has begun. In fact, the smaller the bank is, the less likely it is that the bank’s reputation will be an adequate source of information about the quality and prices of the bank’s products.
A third important feature of this model is the fact that banks offer a variety of services. The simplest example of this is the case of a bank with a large number of branches. Since customers have different preferences over branch location, branches at different locations are offering different services. This means that a bank with many branches is offering a bundle of different services to their customers.
Location is not the only dimension along which banks differ, of course. They will offer different menus of accounts, or concentrate on different types of lending business; they will attract a different mix of retail or wholesale funds; they may diversify into non-bank products such as insurance or mutual funds. Location is a convenient metaphor for these different dimensions.
By exploiting these three features of the model, lock-in effects, limited information, and product diversity, we can reverse the usual presumption that greater concentra- tion leads to more efficient outcomes.
2. For other welfare analyses involving spatial competition and risk see Besanko and Thakor (1992) and Matutes and Vives (1996, 2000).
470 : MONEY, CREDIT, AND BANKING
The model. There is a finite set of locations indexed by l � 1,…,L and at each location there are two banking offices j � 1,2. Time is divided into an infinite number of discrete periods t � 1,2,…. There is a large number of individuals allo- cated exogenously to the different locations. Each period, these individuals have a demand for a unit of banking services which provides them with a surplus v. The value of banking services to individuals is distributed according to the distribution function F(v), that is, F(v) is the fraction of the population with valuation less than or equal to v.
At each date consumers are randomly assigned to a new location. They have an equal probability of arriving at any location and we assume that the number of locations is so large that the probability of returning to the same location is negligible and can be ignored. This is an extreme assumption, to be sure, but it serves to eliminate inconvenient and apparently unimportant complications. Each location is also assumed to receive a representative sample of the different types of individuals so, whatever the number of individuals at a given location, the distribution of types is F(v).
For simplicity, banks are assumed to have a zero marginal cost of providing banking services. Profit is thus identical to revenue.
At each date, the market is assumed to clear as follows. First, the individuals who have gathered at a particular location choose which bank to patronize. They do this before they know the price that the bank will charge. Next, the bank sets the price for its product (the interest rate on loans or deposit accounts, or the fees for other bank services). Finally, the consumers make one of three choices: to purchase the current bank’s services at the quoted price, to switch to the other bank, whose price by now has been fixed, or to do without the services of a bank. There is a fixed cost c � 0 of switching from one banking office to the other.
We consider two limiting cases of bank organization. In the first, which we call unitary banking, each bank has a single branch. In other words, each banking office represents an independent bank. In the second case, which we call branch banking, there are only two banks, each owning one branch in each location. That is, all the banking offices are organized into two large networks. Regardless of the form of organization, we use the term banking office to denote the smallest unit of the bank, whether it constitutes the entire bank or a branch of a larger bank network.
Individuals are assumed to observe only what happens at their own bank in each period and, since they move to a different location in each period, they have no knowledge of the previous behavior of the bank they are patronizing in the current period. The banks themselves are assumed to condition their behavior in each location on their experience at the same location. This maintains an informational symmetry between the unitary- and branch-banking forms of industrial organization, i.e., unitary banking and branch banking, since in each case only local information is being used to condition the (local) pricing decision.
Unitary banking. In this form of industrial organization each bank consists of a single office. A bank sets the price of its product in each period to maximize the present value of profits. In the static version of this model, it is well known that
FRANKLIN ALLEN AND DOUGLAS GALE : 471
the unique equilibrium involves each banking office choosing the monopoly price. More precisely, suppose that there is a unique price pM such that
pM(1 � F(pM)) ≥ p(1 � F(p)), ∀p .
A bank will clearly never want to charge more than pM. If the two banks at some location happen to charge prices p ≤ p′ ≤ pM , where p � pM, the bank charging p can always raise its price by � without losing any customers, because of the fixed cost of switching. This will clearly increase its profits, so the only equilibrium is for both banking offices to charge pM.
In a dynamic context things are generally more complicated, because of the possibility of supporting a different equilibrium by means of punishment strategies. Under the maintained assumptions, however, there is no possibility of using such strategies to increase the set of equilibria. Because individuals only observe what happens at their own locations and never return to the same location, nothing that a banking office does in the current period will be observed by individuals who will visit that location in the future. Further, since banks at other locations do not condition their behavior on what happens at this location, there is no possibility of the bank’s future customers being indirectly informed of a deviation through another bank’s reaction to this bank’s current deviation. Our informational assumptions have effectively severed any possible feedback from a current change in price to a future change in demand, so the argument used in the static model continues to apply. We conclude, then, that the unique, subgame perfect equilibrium of the unitary banking game consists of each bank, in each location, charging the monopoly price pM
in every period. Consumers whose valuation v is greater than pM will purchase banking services; those whose valuation is less than pM will not. The equilibrium is inefficient for the usual reason: the monopoly price is too high and the monopoly quantity is too low.
Branch banking. Now suppose that banking offices are formed into two large networks. Each bank has one branch in each of the locations. Although consumers move from location to location they can stay with the same bank if they wish. The possibility of staying with the same bank generates a plethora of other equilibria. We describe one such equilibrium to illustrate the possibilities.
In each period, at each location, half the customers patronize each of the banks. Along the equilibrium path, the banks charge a price p � ε � 0 in every period. If, in any period, one of the banking offices has deviated from the equilibrium strategy, all the customers will leave the bank that last deviated and henceforth patronize branches belonging to the other bank. The banks continue to charge the same price. If no bank deviated, but some of the customers deviated in the past, the banks continue to charge the same price and the customers continue to patronize the two banks equally. The profits (for a single banking office) from deviating last one period and are less than or equal to (pM � ε)(1 � F(pM)); on the other hand, it loses the profits from these customers in each future period until they all disappear from the game. Customers only last a finite number of periods, but if the number of locations is large, they will be around on average for a very long time. Each period, a fraction
472 : MONEY, CREDIT, AND BANKING
l�1 of the customers dies and is replaced. So the equilibrium profits lost by a single banking office’s deviation are equal to [ε�(1 � l�1)(1 � δ)](1 � F(ε)). For δ suffi- ciently close to 1 and l sufficiently large, the profits from deviating are less than the profits of the equilibrium strategy. This shows that under branch banking, it is possible to support equilibria which are “more efficient” than the unique equilibrium in the case of unitary banking, where “more efficient” means that the sum of consumers’ and producers’ surplus is greater.
How should we interpret these results? The lock-in effects in the banking sector may be substantially greater than in most service industries and may be one of its distinguishing features. Small banks with a limited range of services and a limited geographical presence may have a greater incentive to exploit the lock-in effect than a large bank, because the large bank is always competing for the customer’s future business, in another product line or another location.
Empirical evidence. There is some empirical evidence in support of the view of competition presented above.3 Bordo, Rockoff, and Redish (1994) compare the Canadian and US banking systems from 1920 to 1980. During this period Canada had a few branch banks while the US had unit banking in many parts. It is found that the Canadian system outperformed the US system in a number of respects. First, in Canada the interest rates paid on deposits were generally higher and the income received by security holders was generally slightly higher than in the US. Second, the interest rates charged on loans were generally quite similar in the two countries. Finally, the returns on equity were generally higher in Canada. Taken together this evidence is consistent with the Canadian branch banking system being more competitive than the US unitary banking one.
Carlson and Mitchener (2003) also find evidence that branch banking is more competitive than unitary banking. Using data on national banks from the 1920s and 1930s, they compare states which just have unit banks with ones that have both branch banking and unitary banking. In the former states, many banks that charge high rates of interest on loans and pay low rates on deposits survive. However, in states with both types of banks this kind of local monopoly is eliminated. Both branch banks and unit banks price in exactly the same way.
3.2 Spatial Competition and Diversification
We next consider another model of spatial competition in the tradition of Hotelling. It is shown that competition can be consistent with diversification and hence stability. However, the precise way in which this works depends crucially on the particular assumptions made.
Let locations be denoted by n � 0, ±1, ±2,… and assume that there is a single bank at each location. Identical individuals are uniformly distributed on the real line. Each has one unit of a good that can be invested by the bank in a risky asset with return Rn. For simplicity assume that the returns Rn are i.i.d. with distribution
3. We are grateful to Stephen Haber for bringing this evidence to our attention.
FRANKLIN ALLEN AND DOUGLAS GALE : 473
Rn � {R w.pr. π 0 w.pr. 1 � π .
In a symmetric equilibrium a bank will draw clients from the interval [n � 1�2, n � 1�2] and offer them a risk sharing contract that promises a payment D if the project is successful and nothing otherwise (the banker is risk neutral but has limited liability). The expected utility of the contract is πU(D) � (1 � π)U(0) � πU(D) if U(0) � 0 .
The bank chooses D to maximize profits taking as given the expected utility ū offered by the adjacent banks. Assuming linear transportation costs of one utility per unit distance, the marginal agent m � n who goes to bank n is determined by the condition that
πU(D) � (n � m) � ū � (m � n � 1)
or
2(n � m) � πU(D) � ū � 1 .
The profit per agent is π(R – D) so total profits are
π(R � D)2(n � m) � π(R � D)(πU(D) � ū � 1) .
The first-order condition is
πU′(D) � 1
R � D .
If we allow a bank to occupy several locations, it can pool several independent risks, which allows it to offer better risk sharing to its customers. However, it may face less competition. Whether it does depends on precisely which locations a bank is allowed to occupy. For example, if two banks occupy alternate locations, we get improved risk sharing with no loss of competition. Here there is no trade-off between competition and stability. If a bank occupies a sequence of adjacent loca- tions, it can extract a large amount of surplus from the consumers located in the middle. Here there is a trade-off between competition and stability.
4. SCHUMPETERIAN COMPETITION
Technological innovation is one of the major sources of growth in welfare. As Schumpeter (1950) famously pointed out, perfect competition undermines the incen- tive to innovate (when intellectual property rights are weak) and in that sense imperfect competition may be more “efficient” than perfect competition. Similar ideas apply in the financial sector. If banks innovate they may be able to capture the market and drive other banks out of business. Thus Schumpeterian competition may be associated with financial instability (creative destruction). We start by considering
474 : MONEY, CREDIT, AND BANKING
a “winner-takes-all” model of competition based on the work of Allen and Gale (2000c) which has this feature.
4.1 Winner-Takes-All Competition
There is a finite number of locations i � 1,…, n with a bank at each location. There are two dates, t � 0,1. At date 0, each bank i invests ki ≥ 0 in the development of a product. The banks’ opportunity cost of capital is R. The value of the product developed by bank i is given by Vi(ki, ω). There is symmetric information all agents know the function Vi(·) and the investment ki and have the same continuous probability distribution F(·) over the states of nature ω. We assume that V(0, ω ) � 0 for all ω so capital is essential to the development of a useful product. In general, the more the capital that is provided the greater is the probability that the value of the product is high. We assume that once the new product is developed it can be produced at constant marginal cost and, without essential loss of generality, we set the marginal cost equal to zero.
At date 1 identical consumers are uniformly distributed on the line interval [0, n]. Each consumer wants to consume at most one unit of a new product.
4.2 Equilibrium
At date 0 the banks jointly choose their investment strategies k � (k1,…, kn). At the beginning of date 1, the state of nature ω is realized and the banks observe the quality of the product they have developed
V(k, ω ) � V1(k1, ω ),…, Vn(kn, ω )) .
Then the banks engage in Bertrand competition. Since the qualities are assumed to be continuously distributed (for ki � 0) the probability of ties can be ignored. Then Bertrand competition will lead to an outcome in which the best product captures the entire market and the price charged for this product is equal to the difference between the value of the first- and second-best products. The price for every other product is zero. At this price, consumers are indifferent between the first- and second- best products, but they will demand only the first-best product in equilibrium (if a positive fraction of consumers were expected to choose the second-best product, the firm with the first-best product would have chosen a slightly lower price to capture the entire market). Formally, for any bank i let
V�i(k�i, ω ) � (V1(k1, ω ), …,Vi�1(ki�1, ω ),Vi�1(ki�1, ω ), …,Vn(kn, ω ))
denote the vector of the qualities of products j ≠ i; let
k�i � (k1, …, ki�1, ki�1, …, kn)
denote the allocation of investment in products j ≠ i; and let
V* �i(k�i, ω ) � maxj≠i{Vj(kj, ω )}
denote the highest value in the vector V–i(k–i, ω). Then, in the second-period equilib- rium, the price charged for the ith product is denoted by pi(k, ω) and satisfies
FRANKLIN ALLEN AND DOUGLAS GALE : 475
pi(k, ω ) � max{Vi(k, ω ) � V* �i(k�i, ω),0}, ∀i .
Since the demand is equal to one for the best product and zero for the rest, the revenue of bank i is also equal to pi(k, ω).
At the first date, we look for a Nash equilibrium in the investment levels. The ith bank chooses ki to maximize E[pi(ki, k�i, ω )] � Rki , taking as given the invest- ment levels of the other banks, k–i. So a Nash equilibrium is a vector k* such that
k* i �arg maxki≥0{E[pi(ki, k*
�i, ω )] � Rki}
for each i.
4.3 Optimum
Since the cost of production at date 1 is zero, the surplus generated by consuming the ith product is Vi(ki, ω). Surplus is maximized by having all consumers consume the best product, so the total surplus at date 1 is
V*(k, ω ) ≡ maxi�1,…,n{Vi(ki, ω )} .
Assuming that the consumers are also risk neutral and that lump sum transfers are possible, the first-best efficient allocation is found by maximizing net surplus, that is, by solving the planner’s problem
maxk≥0V *(k) � R�
n
i�1 ki ,
where V*(k) ≡ E[V*(k, ω )] is the expected value of V*(k, ω). Define V*
�i(k�i) ≡ E[V* �i(k�i, ω )]. Then the objective function V*(k) can be written
equivalently as
V*(k) � E[max{Vi(ki, ω ) � V* �i(k�i, ω ), 0} � V*
�i(k�i, ω )] � E[max{Vi(ki, ω ) � V*
�i(k�i, ω ),0}] � V* �i(k�i)
and the planner’s problem can be rewritten equivalently as
maxk≥0E[max{Vi(ki, ω ) � V* �i(k�i, ω ), 0} � V*
�i(k�i) � R� n
i�1 ki
Suppose that k* is a solution to the planner’s problem above. A necessary condition is that ki
* maximizes
E[max{Vi(ki, ω ) � V* �i(k*
�i, ω ), 0}] � Rki � E[pi(ki, k* �i, ω )] � Rki .
But this means that ki * satisfies the equilibrium condition for the ith bank’s choice
of ki. Hence, we have the following result. Proposition 6: If k* is a solution to the planner’s problem, then k* is a Nash
equilibrium of the banks’ investment “game.” In this case the allocation produced by the winner-takes-all competition is efficient.
The innovating banks have exactly the right incentives to invest. However, only
476 : MONEY, CREDIT, AND BANKING
one bank survives in each period. In other words there is considerable financial instability. In order to prevent this instability the government may wish to restrict competition and give each bank a monopoly in a particular region. We turn next to this restricted competition.
4.4 Restricted Competition
Next assume that the bank located at n has a legal monopoly of the region [i � 1�2, i � 1�2]. Then it maximizes E[V(ki,ω ) � Rki]. Clearly, surplus must be lower, both because we are not providing the best product to all consumers but also because we are not using the right investment level.
Will investment be too high or too low? Examples can be constructed to show that the answer is ambiguous. On the one hand losers are protected under the “competitive” arrangement. On the other hand winners suffer because they cannot capture the entire market.
To see this consider the following simple example. There is a continuum of identical consumers with measure one and two banks. Each bank can invest 0 or 0 � I � 1�2. If a bank invests zero the value of its good is zero; if it invests I the value of its good is 1. If the market is unified, there are three equilibria, two asymmetric equilibria in which only one bank invests and a mixed strategy equilib- rium in which the probability of investment is λ � 1 � I. Now suppose we divide the market between the two banks, each getting 1/2 of the market. Then each bank will invest for sure, since I � 1�2.
Comparing equilibria, we see that total investment is greater in the divided market. Also, comparing the unique pure strategy equilibrium of the divided market with the mixed strategy of the unified market, we can see that (the probability of) innovation is strictly higher in the divided market. In the mixed strategy equilibria of the divided market the probability of innovation is the same.
For the case I � 1�2 there will be no investment, and hence no innovation, in the segmented market. In this case investment and innovation are unambiguously lower than in winner-takes-all case.
These examples are enough to indicate that even in simple models it is hard to obtain robust comparative static results. As with the other models we have examined, the relationship between competition and stability is complex and nuanced.
5. CONTAGION
One source of instability in financial systems is the possibility of contagion, in which a small shock that initially affects one region or sector or perhaps even a few institutions, spreads from bank to bank throughout the rest of the system, and then affects the entire economy. There are a number of different types of contagion that have been suggested in the literature. The first is contagion through interlinkages between banks and financial institutions (see, e.g., Rochet and Tirole, 1996a, 1996b, Freixas and Parigi, 1998, Freixas, Parigi, and Rochet, 2000, and Allen and Gale,
FRANKLIN ALLEN AND DOUGLAS GALE : 477
2000b, for theoretical analyses and Van Rijckeghem and Weder, 2000, for empirical evidence). The second is contagion of currency crises (see, e.g., Masson, 1999, Eichengreen, Rose, and Wyplocz, 1996, and Glick and Rose, 1999). The third is contagion through financial markets (see, e.g., King and Wadwhani, 1990, Kyle and Xiong, 2001, and Kodres and Pritsker, 2002).
The notion of financial fragility is closely related to that of contagion. When a financial system is fragile a small shock can have a big effect. A financial crisis may rage out of control and bring down the entire economic edifice (see, e.g., Kiyotaki and Moore, 1997, Chari and Kehoe, 2000, Lagunoff and Schreft, 2001, and Allen and Gale, 2003b).
In this section we are interested in the relationship between contagion and financial fragility and competition. Allen and Gale (2000b) develop a model of financial contagion through the interbank market. They assumed perfect competition. In that case a small aggregate shock in liquidity demand in a particular region can lead to systemic risk. Although the shock is small it may cause a bank to go bankrupt and liquidate its assets. This in turn causes other banks which have deposits in it to also go bankrupt and so on. Eventually all banks are forced to liquidate their assets at a considerable loss. Perfect competition in the interbank market plays an important role in this contagion. Since each bank is small, acts as a price taker and assumes its actions have no effect on the equilibrium, no bank has an incentive to provide liquidity to the troubled bank.
Saez and Shi (2004) have argued that if banks are limited in number they may have an incentive to act strategically and provide liquidity to the bank that had the original problem. This will prevent the contagion and make the banks providing funds in this way better off. The formal model of this phenomenon that captures the role of imperfect competition is similar to the model in Bagnoli and Lipman (1989). They consider the problem of the provision of public goods through private contributions. There is a critical level of resources required to provide the public good. Each person becomes pivotal and in this case the public good can be provided. Similarly, here each bank would be pivotal in the provision of liquidity to pre- vent contagion.
Suppose there are two banks, each holding A units of an illiquid asset and M units of money. One unit of the asset is worth one unit if liquidated today and R units if liquidated tomorrow. Money is storable, so one unit of money today is worth one unit tomorrow. Each bank owes one unit of money to a depositor who withdraws today and the liquidation value of the portfolio to a depositor who withdraws tomorrow. In addition, bank 1 owes bank 2 B units today. Suppose that bank 1 has λ early consumers and 1 � λ late consumers. If λ � B � M � A � ((1 � λ)�R), then bank 1 cannot meet its commitments on its own. If it fails, however, bank 2 can claim only a fraction (B�(1 � B))(M � A) � B. Bank 2 has several actions it can take. It can forgive part of its debt. It can wait for payment at date 2. It can make a cash transfer to bank 1 at date 1. If R is big enough, this may be better for bank 2 than enforcing its debt now.
478 : MONEY, CREDIT, AND BANKING
The same argument obviously applies to any number of banks, though the coordi- nation problem will become more severe as the numbers increase. If there is a continuum of banks or asymmetric information, it may be impossible to sustain the cooperative equilibrium. Note that even in the case with a finite number of banks there may be a coordination failure; if every bank thinks the others will not contribute anything, it may be optimal not to contribute anything (this may depend on the extensive form, e.g., simultaneous moves rather than sequential moves).
Gradstein, Nitzan, and Slutsky (1993) show that Bagnoli and Lipman’s result is not very robust to the introduction of uncertainty. It remains to be seen whether a model of contagion and imperfectly competitive banks would also not be robust.
A model of an imperfectly competitive interbank market may therefore be more stable than the case where there is perfect competition. As in the original agency model of Keeley (1990) there is again a trade-off between competition and finan- cial stability.
6. CONCLUDING REMARKS
In this paper we have considered a variety of different models of competition and financial stability. These include general equilibrium models of financial interme- diaries and markets, agency models, models of spatial competition, Schumpeterian competition, and contagion. There is a very wide range of possibilities concerning the relationship between competition and financial stability. In some situations there is a trade-off as is conventionally supposed but in others there is not. For example, with general equilibrium and Schumpeterian models efficiency requires the combination of perfect competition and financial instability.
There is a large policy literature based on the conventional view that there is a trade-off between competition and stability. Since competition is generally viewed as being desirable because it leads to allocational efficiency, this perceived trade-off lead to calls for increased regulation of the banking sector to ensure the coexistence of competition and financial stability. The most popular instrument for achieving this end was the imposition of minimum capital requirements on banks. If the owners of banks were forced to put up significant amounts of capital, they would be unwilling to take risks because they would again stand to loose large amounts of funds. The Basel Agreement of 1988 imposed capital controls on banks so that the incentive to take risks would be reduced and they could compete on equal terms. A large literature has developed analyzing the effect of capital controls. For example, Hellman, Mur- dock, and Stiglitz (2000) showed in the context of a simple model of moral hazard that capital controls were not sufficient. In addition to capital controls, deposit rate controls were also necessary to achieve Pareto efficiency.
Our analysis suggests that the issue of regulation and its effect on competition and financial stability is complex and multi-faceted. Careful consideration of all the factors at work both at a theoretical and empirical level is required for sound policy.
FRANKLIN ALLEN AND DOUGLAS GALE : 479
LITERATURE CITED
Allen, F., and D. Gale (1994). “Limited Market Participation and Volatility of Asset Prices.” American Economic Review 84, 933–955.
Allen, F. and D. Gale (1998). “Optimal Financial Crises.” Journal of Finance 53, 1245–1284.
Allen, F., and D. Gale (2000a). Comparing Financial Systems. Cambridge, MA: MIT Press.
Allen, F., and D. Gale (2000b). “Financial Contagion.” Journal of Political Economy 108, 1–33.
Allen, F., and D. Gale (2000c). “Corporate Governance and Competition.” In Corporate Governance: Theoretical and Empirical Perspectives, edited by X. Vives, pp. 23–94. Cambridge, UK: Cambridge University Press.
Allen, F., and D. Gale (2003a). “Financial Intermediaries and Markets.” Working Paper 00- 44-C, Wharton Financial Institutions Center. Econometrica, forthcoming.
Allen, F., and D. Gale (2003b). “Financial Fragility, Liquidity and Asset Prices.” Working Paper 01-37-B, Wharton Financial Institutions Center.
Bagnoli, M., and B. Lipman (1989). “Provision of Public Goods: Fully Implementing the Core through Private Contributions.” Review of Economic Studies 56, 583–601.
Beck, T., A. Demirguc-Kunt, and R. Levine (2003). “Bank Concentration and Crises.” Working Paper, World Bank.
Berger, A., and D. Humphrey (1997). “Efficiency of Financial Institutions: International Survey and Directions for Future Research.” European Journal of Operational Research 98, 175–212.
Besanko, D., and A. Thakor (1992). “Banking Regulation: Allocational Consequences of Relaxing Entry Barriers.” Journal of Banking and Finance 16, 909–932.
Bordo, M., H. Rockoff, and A. Redish (1994). “The U.S. Banking System from a Northern Exposure: Stability versus Efficiency.” Journal of Economic History 54, 325–341.
Boyd, J., and G. De Nicolo (2002). “Bank Risk-Taking and Competition Revisited.” Working Paper, Carlson School of Management, University of Minnesota.
Canoy, M., M. van Dijk, J. Lemmen, R. de Mooij, and J. Weigand (2001). Competition and Stability in Banking. The Hague, Netherlands: CPB Netherlands Bureau for Economic Policy Analysis.
Carletti, E., and P. Hartmann (2003). “Competition and Financial Stability: What’s Special about Banking?” In Monetary History, Exchange Rates and Financial Markets: Essays in Honour of Charles Goodhart, Vol. 2, edited by P. Mizen. Cheltenham, UK: Edward Elgar.
Carlson, M., and K. Mitchener (2003). “Branch Banking, Bank Competition and Financial Stability.” Working Paper, Department of Economics, Santa Clara University.
Chari, V., and P. Kehoe (2000). “Financial Crises as Herds.” Working Paper, Federal Reserve Bank of Minneapolis.
Diamond, P. (1971). “A Model of Price Adjustment.” Journal of Economic Theory 3, 156– 168.
Diamond, D., and P. Dybvig (1983). “Bank Runs, Deposit Insurance, and Liquidity.” Journal of Political Economy 91, 401–419.
Eichengreen, B., A. Rose, and C. Wyplocz (1996). “Contagious Currency Crises: First Tests.” Scandinavian Journal of Economics 98, 463–484.
Freixas, X., and B. Parigi (1998). “Contagion and Efficiency in Gross and Net Interbank Payment Systems.” Journal of Financial Intermediation 7, 3–31.
Freixas, X., B. Parigi, and J. Rochet (2000). “Systemic Risk, Interbank Relations and Liquidity Provision by the Central Bank.” Journal of Money, Credit, and Banking 32, 611–638.
480 : MONEY, CREDIT, AND BANKING
Glick, R., and A. Rose (1999). “Contagion and Trade: Why are Currency Crises Regional?” In The Asian Financial Crisis: Causes, Contagion and Consequences, edited by P. Agénor, M. Miller, D. Vines, and A. Weber, chap. 9. Cambridge, UK: Cambridge University Press.
Gradstein, M., S. Nitzan, and S. Slutsky (1993). “Private Provision of Public Goods under Price Uncertainty.” Social Choice and Welfare 10, 371–382.
Hellmann, T., K. Murdock, and J. Stiglitz (2000). “Liberalization, Moral Hazard in Banking, and Prudential Regulation: Are Capital Requirements Enough?” American Economic Review 90, 147–165.
Hoggarth, G., and V. Saporta (2001). “Costs of Banking System Instability: Some Empirical Evidence.” Financial Stability Review (June 2001), 148–165.
Jensen, M., and W. Meckling (1976). “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure.” Journal of Financial Economics 3, 305–360.
Keeley, M. (1990). “Deposit Insurance, Risk and Market Power in Banking.” American Economic Review 80, 1183–1200.
King, M., and S. Wadhwani (1990). “Transmission of Volatility between Stock Markets.” Review of Financial Studies 3, 5–33.
Kiyotaki, N., and J. Moore (1997). “Credit Chains.” Journal of Political Economy 105, 211–248.
Kodres L., and M. Pritsker (2002). “A Rational Expectations Model of Financial Contagion.” Journal of Finance 57, 768–799.
Kyle, A., and W. Xiong (2001). “Contagion as a Wealth Effect.” Journal of Finance 56, 1401–1440.
Lagunoff, R., and S. Schreft (2001). “A Model of Financial Fragility.” Journal of Economic Theory 99, 220–264.
Masson, P. (1999). “Contagion: Monsoonal Effects, Spillovers and Jumps between Multiple Equilibria.” In The Asian Financial Crisis: Causes, Contagion and Consequences, edited by P. Agénor, M. Miller, D. Vines, and A. Weber, chap. 8. Cambridge, UK: Cambridge University Press.
Matutes, C., and X. Vives (1996). “Competition for Deposits, Fragility and Insurance.” Journal of Financial Intermediation 5, 184–216.
Matutes, C., and X. Vives (2000). “Imperfect Competition, Risk Taking, and Regulation in Banking.” European Economic Review 44, 1–34.
Perotti, E., and J. Suarez (2003). “Last Bank Standing: What Do I Gain if You Fail?” European Economic Review 46, 1599–1622.
Rochet, J., and J. Tirole (1996a). “Interbank Lending and Systemic Risk.” Journal of Money, Credit, and Banking 28, 733–762.
Rochet, J., and J. Tirole (1996b). “Controlling Risk in Payment Systems.” Journal of Money, Credit, and Banking 28, 832–862.
Saez, L., and X. Shi (2004). “Liquidity Pools, Risk Sharing and Financial Contagion.” Journal of Financial Services Research 25, 5–23.
Schumpeter, J. (1950). Capitalism, Socialism and Democracy, 3rd edition. New York: Harper & Row.
Van Rijckeghem, C., and B. Weder (2000). “Spillovers through Banking Centers: a Panel Data Analysis.” IMF Working Paper WP/00/88, Washington, DC: International Monetary Fund.
Bank-competition-and-financial-stability-A-comparison-of-commercial-banks-and-mutual-savings-banks-in-Korea_2013_Pacific-Basin-Finance-Journal.pdf
Pacific-Basin Finance Journal 25 (2013) 253–272
Contents lists available at ScienceDirect
Pacific-Basin Finance Journal
j ourna l homepage: www.e lsev ie r .com/ locate /pacf in
Bank competition and financial stability: A comparison of commercial banks and mutual savings banks in Korea☆
Jin Q. Jeon a,⁎, Kwang Kyu Lim b
a Dongguk Business School, Dongguk University, 3-26 Pil-dong, Chung-gu, Seoul 100-715, South Korea b Economist, Macroprudential Analysis Department, The Bank of Korea, 110, Namdaemunro 3-Ga, Jung-Gu, Seoul 100-794, South Korea
a r t i c l e i n f o
☆ We thank seminar participants at the Financial St is grateful to the Dongguk University Research Fund ⁎ Corresponding author. Tel.: +82 2 2260 8911.
E-mail addresses: [email protected] (J.Q. Jeon),
0927-538X/$ – see front matter © 2013 Elsevier B.V. http://dx.doi.org/10.1016/j.pacfin.2013.10.003
a b s t r a c t
Article history: Received 5 February 2013 Accepted 8 October 2013 Available online 17 October 2013
In this study, we provide new evidence that the relationship between banking competition and financial stability varies depending on the characteristics of banks. By using a sample of two different types of banks, Korean commercial banks and mutual savings banks, we find that the non-linear relationship between competition and the stability of commercial banks reflects a trade-off between the interest effect and risk-shifting effect. However, consistent with Boyd and De Nicolo (2005), competition has a positive effect on the stability of mutual savings banks with greater business risk and weaker corporate governance. Our results provide important implications on banking competition policy.
© 2013 Elsevier B.V. All rights reserved.
JEL classification: G21 G23 L1
Keywords: Bank competition Stability Risk shifting Mutual saving banks Commercial banks
1. Introduction
Banking literature provides two alternative hypotheses regarding the relationship between bank competition and stability or risk-taking behavior. According to the conventional competition–fragility theory, higher competition in the financial services industry causes financial institutions to lose their market power, leading to a decrease in profitability. In order to recover from financial losses, financial institutions are more likely to invest in riskier portfolios. Consequently, this risk-taking behavior will undermine the stability of financial institutions (Keeley, 1990; Allen and Gale, 2000; Hellmann et al., 2000, etc.).
ability Forum supported by the Bank of Korea for helpful Comments. Jin Q. Jeon for financial support. Remaining errors are our own.
[email protected] (K.K. Lim)
All rights reserved.
254 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
Competition–stability theory, on the other hand, suggests that competitionhas a potentially positive effect on the stability of financial institutions. According to Boyd and De Nicolo (2005), bankswith a greatermarket share in the loan market, which experience lower competition, tend to impose higher interest rates on their loans. Higher interest rates being charged by banks in a less competitivemarket may increase the risk-taking behavior of borrowing firms. Boyd and De Nicolo consequently argue that since the risk is ultimately shifted from borrowers to banks in this situation, the default probability of banks increases in the riskiness of bank loans.
A recent study by Martinez-Miera and Repullo (2010) provides partial support for Boyd and De Nicolo's (2005) risk-shifting effect that the greater interest rates that exist in less competitive markets raise the risk of loans and thus make a bank bankruptcy more likely. They, however, additionally take into account the fact that greater interest rates also improve bank profitability, which is known as the interest effect, and suggest that there exists a U-shaped relationship between competition and the risk of bank failure. Thus far, empirical research has produced mixed results on the influence of bank competition on stability (e.g., Berger et al., 2009; Tabak et al., 2012a; Beck et al., 2013).
The purpose of this paper is to investigate the relationship between competition and stability by using a sample of two different types of banks: mutual savings banks (hereafter MSBs) and commercial banks in the Korean depository industry. Although previous studies have indicated the theoretical and empirical links between bank competition and stability, it has not as yet been investigated how this relationship varies across different types of banks. This paper attempts to fill this gap in the literature by examining whether there exist differences in terms of governance structure, loan characteristics and regulatory environments that cause banks to differently interact with industry competition.
The primary difference in the corporate governance structure between MSBs and commercial banks in the Korean markets is that MSB ownership is, in general, concentrated in the hands of a few individuals, who customarily belong to one family, while the ownership of commercial banks is widely dispersed.1 It is also important to note that since there are little mandatory requirements for disclosure, MSBs which provide little information on their financial situation are not affected by market discipline mechanisms. This creates a moral hazard problem in that the major shareholders and senior management are likely to take excessive risks in pursuit of private gains. Moreover, the borrowers of MSBs are mainly small and medium-sized enterprises (hereafter, SME), which possess greater business and credit risks than is true of the borrowers of commercial banks. Under this circumstance, MSBs in less competitive markets are more likely to be exposed to a greater degree of moral hazard and to charge higher interest rates on their loans.2 As a result, in light of the lower level of competition, MSB borrowers are more likely to choose risky projects to repay high interest rates, leading to the high possibility of defaults for both the borrowers and MSBs. As Martinez-Miera and Repullo observe, the risk-shifting effect may overwhelm the interest effect for MSBs.
This study investigates the hypothesis using the quarterly panel data of MSBs and commercial banks from 1999, the end of Asian financial crisis, through to 2011. We first estimate the level of bank competition. Although it is hard to directly measure competition, several possible approaches to this are presented in the banking literature. We follow Boone (2008a, 2008b) which measures competitiveness using operating efficiency, known as the Boone index. The conventional concentration ratio and the Herfindahl–Hirschman Index are also used as supplementary measures for bank competition. We employ the Z-score as a measure of bank stability which can capture how far a given bank is from insolvency. Finally, we estimate pooling regressions, panel regressions, and the difference-in-differences model to examine the relationship between the competitive levels and stability of MSBs and commercial banks.
Our empirical results generally support the hypothesis that the effect of bank competition on stability is different depending on the characteristics of the banks involved. The results of our multiple regressions show that competition has a significant and positive effect on the stability of MSBs with weak corporate governance. On the contrary, competition pressure significantly reduces the stability of commercial banks but this relationship is shown to be non-linear. Consistent with Tabak et al. (2012a), commercial banks are more stable at the very low or very high level of competition, while they are riskier at the medium level of
1 As of March 2011, the average ownership by the largest shareholders of large MSBs is 62.2%, whereas that of small and medium- sized MSBs is 70.4% (Korea Financial Supervisory Service, 2012).
2 It has been documented that the recent distress of MSBs has mainly been caused by excess risk-taking and, in some MSBs, unauthorized appropriation by major shareholders and senior management, and by the failure of effective monitoring on the part of internal and external governance mechanisms (See The Korean Herald, May 2, 2011; Korea Times, May 14, 2012).
255J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
competition. These results are robust across alternative proxies for competition and stability as well as across various model specifications. Therefore, we conclude that, for commercial banks at least, higher competition creates a trade-off between the risk-shifting effect and interest effect, but the risk shifting effect overwhelms the interest effect for MSBs when the market is less competitive. In addition, our difference-in-differences analysis shows that the positive effect of competition on the stability of MSBs became even greater after deregulations by the Korean government and after the global financial crisis.
The paper contributes to the literature in several ways. First, while most of previous studies examine the relationship using either domestic or international banking data and focusing only on the commercial bank industry, it is one of the first works to provide comprehensive evidence on the relationship between bank competition and stability, as conditional on different bank characteristics. Second, the paper can provide further understanding of the influence of bank competition on stability by showing that competition significantly decreases the risk-taking behaviors of MSBs with weaker corporate governance and greater business risk, while it has a non-linear relationship with stability for commercial banks. Third, our evidence can provide important implications on banking competition policy of which the main focus is to ensure financial stability. We argue that competition policy should be applied differently to commercial banks and MSBs. That is, either a very low or a very high degree of competition would be optimal for commercial banks, while a higher degree of competition will significantly decrease the default risk of MSBs. Fourth, it is the first paper to analyze Korean MSBs by using hand-collected data. Thus far, there has been no published empirical research directly examining MSBs, due to the lack of available data that results from the fact that most MSBs are non-listed.3 Finally, while the previous literature has used annual financial data in its analysis, this work aims to derive more reliable results from its empirical tests by using quarterly financial data, which enables us to analyze a larger number of observations. We should, however, be clear about the limitations of the work. Due to the lack of detailed information on the corporate governance of MSBs, the paper does not address how the corporate governance mechanisms of banks affect the relationship between competition and stability, which we leave for future research.
The paper is organized as follows. Section 2 provides a literature review. In Section 3, we describe our sample and empirical methodology. Section 4 analyzes the relationship between competition and stability, while Section 5 concludes.
2. Literature review
2.1. Bank competition and stability
According to the Bank for International Settlement (BIS) and International Monetary Fund (IMF), the world has seen a rapid consolidation of banks during the past decades, which results in a decrease in the number of banks and an increase in the average size of the banks.4 The Korean banking industry also has experienced unprecedented merger and restructuring activity during the last decade. As a result, the number of MSBs and commercial banks has significantly declined leading to a rise in the degree of concentration (for detailed discussions, see Section 2.2).
Several recent studies show that consolidation and the associated change in the size structure of the banking industry significantly affect not only the performance of individual banks but also market-wide competition and stability of banks. Berger et al. (2007) examine theeffect of the change inmarket structure and competition in the banking industry on the performance and stability of banks. They show that the increased presence of large multi-market banks in a local market significantly lowers loan rates indicating higher competition for small business loans. Using a theoretical framework to analyze competition between large multimarket banks and small local banks, Park and Pennacchi (2009) indicate that large banks are likely to set uniform loan rates across their markets, whereas small banks tend to set rates based on the competitive environmentwithin their respectivemarket. They conclude that competition from the largemultimarket banks tends to benefit borrowers but harm depositors. Erel (2011) find that bank mergers, on average, reduce loan
3 As of 2011, only 7 out of 105 MSBs were listed in the Korean exchange. 4 See “The banking industry in the emerging market economies: Competition, consolidation, and systemic stability.” (BIS paper 4,
2001) and “Financial sector consolidation in emerging markets” (International Capital Market Report published by the International Monetary Fund, 2001).
256 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
spreads and the effect is larger with an acquirer with larger declines in operating costs. However, the spread is larger when the acquirer and target have significantmarket overlap and increasedmarket power post merger. Cyree and Spurlin (2012) investigate the effect of the presence of a big bank on the performance of rural and small banks. They find lower profit efficiency for rural one-county banks, while they tend to achieve greater ROA and higher levels of interests and fees from their loans.
Turning now to the relationship between bank competition and stability, theoretical models and empirical results have provided conflicting evidence. According to the theory of competition fragility, a greater level of competition in the banking industry leads to more fragility. On the other hand, in a less competitive market, banks will have numerous lending opportunities and can increase profits and capital ratios, and as such they will be able to withstand any economic shocks and are less likely to take excessive risk. Keeley (1990) developed a theoretical model that predicts that banks with more pressures on their profits have higher incentives to take excessive risk, resulting in a greater probability of defaults. Allen and Gale (2000) report that the number of competitors in a loan market has a positive effect on bank defaults. Hellmann et al. (2000) examine the relationship between competition for deposits, risk taking and regulation, using a dynamic framework model. They argue that removal of interest ceilings on deposits erodes franchise value and motivates moral hazard behavior by banks. The framework of these models is characterized by the fact that bank competition is modeled on the liability side whereas banks' asset allocation decisions are modeled on the asset side, which is not affected by bank competition.
However, more recent papers have taken into account the relationship between banks and borrowers on the asset side of banks. Boyd and De Nicolo (2005), in their theoretical model, assume that borrowing firms entirely determine the risk of projects in the condition of the loan rates set by banks. This paper supports the competition–stability hypothesis by showing that greater concentration (less competition) causes banks to become riskier because the greater loan rate charged by banks with less competition implies a greater bankruptcy risk for borrowing firms. The empirical study by Boyd et al. (2006), who measure market structure by concentration indicators, also shows that, consistent with Boyd and De Nicolo (2005), the probability of bank failure increases by the level of concentration in the bank industry. However, Martinez-Miera and Repullo (2010) point out that the competition–stability view supported by Boyd and De Nicolo does not necessarily hold when the loan defaults are imperfectly correlated. Rather, greater competition decreases the risk taking by borrowing firms and thus reduces the probability of bank defaults, which is called the risk-shifting effect. On the other hand, a decrease in loan rates due to greater competition reduces bank profits, known as the margin effect, which is not considered in Boyd and De Nicolo. As a result, Martinez-Miera et al. suggest a non-linear U-shaped relationship between competition and bank risk taking.
Recent empirical literature on the effect of bank competition on stability has shownmixed results. Berger et al. (2009) examine 8235 banks in 23 developed economies and report the results that support both the competition–fragility and competition–stability views. Specifically, they find that banks with greater market power have riskier loan portfolios but that the risk is traded off by higher capital ratios or other risk-mitigating techniques. Tabak et al. (2012a) use bank data from 10 Latin American countries from 2003 to 2008 and find evidence that the relationship between competition and risk-taking is non-linear. That is, both high and low levels of competition significantly increase bank stability, while the opposite is true under moderate competition. Beck et al. (2013) study cross-country variation in the relationship between bank competition and stability. Specifically, they investigate how heterogeneous regulatory and institutional features affect this relationship across countries. Their paper shows that competition significantly decreases bank stability in countries with stronger activity restrictions and more homogenous market environments. They also find that the deposit insurance policy and efficiency of credit information sharing are important determinants of the negative relationship between stability and competition. Finally, Liu andWilson (2011) examine whether the effect of competition on stability varies depending on the characteristics of banks in the Japanesemarket. They found that competition enhances the stability of banks with a lower stability level, but damages the stability of banks with a higher stability level.
Several studies have extended the empirical framework by including the banking system and the government policies. For example, Hakenes and Schnabel (2011) analyze the effects of capital regulation on banking stability when banks compete for loans and deposit. They show that tighter capital regulation increase the probability of loan default, which increase the risk of banks. Tabak et al. (2012b) investigate how bank size and market competition affect bank stability in 17 Latin countries. They show that in concentrated markets a few dominant banks are likely to outperform other banks, while this unequal
257J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
banking system is detrimental for the stability and performance of small banks. Gropp et al. (2011) test the competitive effects of bail-out policies by the government. Their results suggest that government guarantees significantly increase the risk-taking of competitors, while they don't find any evidence on the greater default probability of the protected banks.
2.2. Mutual savings banks and commercial banks in Korea
Based on Mutual Savings Banks Act of 2001, a MSB is classified as a non-bank depositary institution. This type of institution was introduced by the Korean government in order to protect consumers by incorporating private lenders into the financial landscape and rationalizing their business, as well as to act as a financial intermediary for SMEs, which tend to have low credit ratings and sound solvency. MSBs charge higher interest rates than commercial banks so as to expand their business base and to compensate for greater perceived risk. Due to drastic deregulations by the Korean government in order to support microfinance and promote small loan markets, MSBs aggressively expanded their microcredit loans during the mid-2000s. Furthermore, they enjoyed a boom as real estate project financing loans rose sharply as a consequence of the real estate boom. However, from 2007, sluggishness in the real estate business and the global financial crisis increased the defaults in the real estate project financing loans extended by the MSBs. Consequently, as of the end of 2011, MSBs have greater risk and more vulnerable loans than is true of other financial institutions. For example, low-rated loans with a borrower's credit score greater than 7 account for half of the total household loans of MSBs, according to the Bank of Korea.5 The average credit rating of companies borrowing from MSBs is at CC. The proportion of loans given to those whose credit rating is B-rated and above, which is in general considered relatively good, accounts for only 20% of the total corporate loans of MSBs. It has been documented that MSB distress is mainly caused by excess risk-taking practices and, in some MSBs, unauthorized appropriation by major shareholders and senior management, as well as by the failure of effective monitoring on the part of internal and external governance mechanisms.
According to the Korea Federation of Saving Banks (KFSB), the number of MSBs showed a downward trend from 211 at the end of 1998 to 105 at the end of 2010.6 During this period, many MSBs were restructured by license revocations, mergers and acquisitions, as well as purchase and assumptions (P&A). Furthermore, 20 banks (e.g., Samhwa and Busan) were ordered to suspend business between 2011 and 2012. As a result, 93 MSBs were in business as of the end of June 2012. The default problems of these MSBs have not yet been solved in Korea because there are worries about additional defaults in real estate project financing loans and uncertainties throughout the industry, which have been caused by a series of business suspensions of some large-sized MSBs. The KFSB reports that the asset size of MSBs had expanded from 25 trillion won at the end of 2002 to 87 trillion won at its peak at the end of 2010. After then, however, their total asset size diminished very sharply, to approximately 60 trillion won as of the end of 2011.
Korean commercial banks experienced a strong restructuring as a result of the government putting large-scale public funds into insolvent banks from 1997 to 1998, when Korea suffered from a currency crisis. Entering the 2000s, Korean banks also continued to grow larger and larger through mergers. In 2001, the Koreanmegabanks started withWoori Bank, which was formed as a merger by the Peace Bank of Korea and Hanvit, followed by Kookmin (merged with Housing & Commercial Bank) and Shinhan (merged with Cho Hung). Likewise, City Bank Korea took over KorAm in 2004 and First City Bank of Korea was acquired by a British-based bank, Standard Chartered plc., in April 2005. As a result of a series of restructurings in the Korean banking system since the Asian currency crisis, the number of commercial banks (including regional banks) declined from 26 banks in 1997 to 17 banks in 2000, to 13 banks as of 2012. The Korean banking industry as such experienced large-scale restructuring during this short period. Consequently, this has led to a rise in degree of concentration within the Korean banking industry and concerns have ceaselessly been raised over a marked decline in market competition.
5 In Korea, a borrower's credit score assigned by financial institutions ranges from 1 to 10. A higher credit score indicates a weaker status in terms of repayment capacity. Accordingly, borrowers with a credit score of 7 and above have higher possibility of delinquency than those with lower credit scores.
6 http://www.fsb.or.kr/.
258 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
3. Sample and methodology
3.1. Sample
The sample period for the Korean commercial banks starts in January 1999 and ends in December 2011. For the MSBs, however, it starts from 2003, due to data limitations. We obtain financial data on the population of commercial banks and hand-collect the data for the MSBs, most of which are non-listed firms. The databases that we use include the Data Analysis, Retrieval and Transfer System (DART) and the Financial Statistics Information System (FISIS) maintained by the Korean Financial Supervisory Service (FSS), as well as the Economic Statistics System (ECOS) by the Bank of Korea.7 Since the data from the DART, FISIS and ECOS are not sufficient for our analysis, we collect data based on the reports of each MSB. Even though most empirical studies on banking competition use annual financial data, we use quarterly data because this enables us to provide more reliable results.
Table 1 reports the descriptive statistics for our sample of MSBs and commercial banks. Market shares are calculated based on total assets, total loans, household loans, and commercial loans. The average market share in total assets is 8.489% for MSBs and 5.882% for commercial banks. The size of the commercial banks is greater than that of the MSBs. The natural logarithm of total assets for MSBs is, on average, 7.652, while it is 17.404 for commercial banks. The average profit ratio, calculated as net income divided by total revenues, of MSBs, is −5.599%, which reflects the depression that has existed in the MSB industry since the late 2000s, as is described in Section 2. The greater standard deviations of the profit ratio for MSBs compared to that of the commercial banks suggest that MSBs are less stable than are commercial banks. The loan to deposit ratio and commercial loan to home loan ratio for MSBs are, on average, 83.095% and 8.221 times, respectively, while those for commercial banks are 113.466% and 2.259 times, respectively. The average BIS ratio defined as the ratio of the capital to the risk-adjusted assets is 10.851% for the MSBs and 11.862% for the commercial banks.
3.2. Methodology
We estimate the following regression models in order to examine the effect of bank competition on stability.
7 DAR
Stabilityit ¼ α þ β Competition Measuresit þ XN k¼1
γkitXkit þ eit
i is the indexed banks and t is the index quarters. Our measures for Stability and Competition are
where discussed in more details below. X is the set of N control variables, including the bank-specific and market-related variables. We conduct both OLS and panel analysis to estimate the equations. Based on previous literature, we include the bank size (ln(assets)), profitability (Profit ratio), loan–deposit ratio, (Loan to deposit), and commercial-house loan ratio (Commercial to home loan) as control variables in our regression models. We also include the fluctuation of CD rates (CD volatility) in order to control for the effect of market situations.
Following Laeven and Levine (2009), Boyd et al. (2006), and many others, we use the Z-score as a measure of bank stability. The Z-score is defined as the average of the ROAs and capital ratios divided by the standard deviation of the ROAs.
Z−score ¼ ROAþ Capital Ratio � �
σROA
We calculate the Z-score by using the ROAs and capital ratios for the prior 4 quarters, ln(Z4), as well as the prior 8 quarters, ln(Z8). The Z-score measures how distant a bank is from insolvency, inversely
T: http://dart.fss.or.kr, FISIS: http://efisis,fss,or,kr, ECOS: http://ecos.bok.or.kr/.
Table 1 The sample period starts in January 1999 for commercial banks and in January 2003 for MSBs, and ends in December 2011. We obtain quarterly financial data from the Data Analysis, Retrieval and Transfer System (DART) and Financial Statistics Information System (FISIS) maintained by the Korean Financial Supervisory Service (FSS), Economic Statistics System (ECOS) maintained by the Bank of Korea, and the financial reports by MSBs. Market share is calculated based on total assets, total loans, household loans, and commercial loans. Variable definitions are provided in Appendix A.
N Mean Median St dev
Panel A. MSBs Market share (%) of
Total assets 5890 8.489 4.743 9.991 Total loans 3456 0.955 0.489 1.481 Household loans 3456 0.955 0.541 1.600 Commercial loans 3456 4.557 2.062 6.586
ln(assets) 5890 7.652 7.524 1.089 Profit ratio (%) 3432 −5.599 4.632 76.620 Loan to deposit (%) 5888 83.095 84.623 19.487 Commercial to home loan (x) 3456 8.221 3.564 14.635 BIS (%) 3117 10.851 9.820 16.615
Panel B. commercial banks Market share (%) of
Total assets 884 5.882 4.300 5.110 Total loans 880 5.909 3.416 5.887 Household loans 874 5.950 2.545 8.360 Commercial loans 880 8.156 4.945 7.934
ln(assets) 884 17.404 17.723 1.322 Profit ratio (%) 820 3.225 4.213 13.923 Loan to deposit (%) 833 113.466 97.737 57.881 Commercial to home loan (x) 874 2.259 1.973 1.876 BIS (%) 866 11.862 11.655 3.868
259J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
proportional to the probability of bank defaults. That is, a greater Z-score represents a lower bankruptcy risk.
Our primary measure of bank competition is the Boone index.8 Boone (2008a, 2008b) estimates the level of bank competition by investigating the relationship between bank performance and efficiency, which is measured as marginal costs. According to Boone, when a market becomes more competitive, efficient firms are more greatly rewarded and inefficient firms are more harshly punished, compared to the situation in a less competitive market. Hence, Boone calculates the level of competition by estimating the elasticity of a firm' performance, in terms of its market shares, with respect to its marginal costs, as follows:
8 Giv conduc banking market Hirschm assume small n (Baumo Panzar type of
ln MSitð Þ ¼ α̂ þ β̂ln MCitð Þ
MS denotes the market shares in the loans or total assets of bank i andMC is the marginal costs. The
where estimated coefficient β is interpreted as the profit elasticity, or the Boone index, which is negative, i.e. banks with greater marginal costs lose market share. Since competition enhances this negative relationship, the greater is the bank competition, the more negative is the Boone index.
en that competition cannot be measured directly, indirect measurement techniques have been divided into structure– t–performance and competition–contestability approaches. The structural methods to assess the competitive level in the industry are based on the structure–conduct–performance assumption, which predicts that the number of banks and their shares determine competitive behavior. The conventional measures include the concentration ratio and Herfindahl– an Index. On the contrary, the nonstructural approach, which is based on the competition–contestability theory, does not , a priori, that concentrated markets are less competitive. It suggests that, in the absence of an entry barrier, markets where a umber of firms serve can be nevertheless characterized by competitive equilibrium because of potential short-term entrants l, 1982; Baumol et al., 1982; among others). This non-structural approach uses the measures such as the H statistic from and Rosse (1982, 1987) and the Boone (2008a, 2008b) index, which measure bank competitiveness without considering the market structure.
260 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
Since marginal costs cannot be directly observed, empirical studies approximate the marginal costs by calculating the average variable costs (Schaeck and Cihak, 2010) or by using a translog cost function (Leuvensteijn et al., 2011). In this paper, we follow Leuvensteijn et al. and estimate the following cost function:
9 The
ln C=ω3ð Þ ¼ δ0 þ X
jδ jlnyjþ X
j
X kμ jklnyjlnyk þ
X kβkln ωk=ω3ð Þ þ
X j
X kγjklny jln ωk=ω3ð Þ
þ X
j
X kθjkln ω j=ω3
� � ln ωk=ω3ð Þ þ e
C is the bank operating expenses, y is the output including total loans (y1), total securities (y2), and
where non-interest income (y3). w represents the inputs including the prices of labor (w1), of physical capital (w2), and of human capital (w3). By taking the first derivative of the translog cost function with respect to total loans, we obtain the marginal costs of loans as follows:
MC1 ¼ ∂ C
ωg
! ∂y1
¼ C ωg
y1
0BB@ 1CCA
∂ln C ωg
! ∂lny1
Finally, we estimate the Boone index for the entire sample period as well as for the specific quarter so that,
ln MSitð Þ ¼ α̂ þ β̂ln MCitð Þ þ eit ln MSitð Þ ¼ α þ
X2011Q4 t¼1999Q1
βt ln MCitð Þ � Quarter Dummyt þ δtQuarter Dummyt þ εit :
In the first regression, the Boone index for the period of 1999–2011 is obtained. In the second regression, we interact the Boone index with the quarter dummies and control for the quarter effect. The estimation of the Boone index for each quarter is a result of this specification. The market share is mainly based on total assets and total loans, but is also calculated with respect to total house loans and commercial loans. A negative Boone index suggests that an increase in marginal costs decreases market shares and, thus, competitive pressure does in fact exists in the bank industry. In order to make this directly proportional to the level of competition, we employ the opposite of the Boone index (Tabak et al., 2012a). We also use a dummy that takes a value of 1 for a significantly negative Boone index and is otherwise 0, as a proxy for competition.
The conventional measures for market concentration, the concentration ratio (CR, hereafter) and the Herfindahl–Hirschman Index (HHI hereafter), are also employed as supplementary measures for competition. CR is defined as the sum of market shares of the N largest banks in the market.
CRN ¼ XN
i¼1 MarketSharei
Although there is no rule for the determination of the number of N, CR4 and CR8 are the most frequently used methods. The HHI is calculated by squaring the market share of each bank and then summing the squares:
HHI ¼ XK i¼1
MarketShareið Þ2:
The U.S. Department of Justice and the FTC (Federal Trade Commission) consider a market with a value less than 1500 to be a competitive market place, one with 1500 ~ 2500 to be a moderately concentrated market place, and one with 2500 or higher to be a highly concentrated market place.9
market is perfectly competitive if HHI is 100. See http://www.justice.gov/atr/public/guidelines/hmg-2010.html.
261J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
4. Empirical results
4.1. Analysis of market competition
Table 2 presents the results of the Boone indices for both Korean commercial banks and MSBs, as estimated for the entire sample period. In Panels A and B, we calculate the market shares based on the total assets and total loans. Additionally, given that MSBs concentrate their finance on SMEs, the market share is then re-calculated using the total commercial loans for MSBs and the total household loans for commercial banks in Panel C. The table shows that, for all three cases, the coefficients of the Boone indices, as measures for bank competition, are negatively and significantly correlated with the market shares. The results suggest that, during our sample period, competition pressure did in fact exist for both the commercial bank and the MSB industry in Korea.
In order to construct panel data, which enable us to examine how bank competition affects stability, we estimate the Boone indices on a quarter basis. In Table 3, the market shares are calculated based on total assets as well as total loans. The sample periods start in 1999 for commercial banks and in 2003 for MSBs, due to data availability. The results show that, for both commercial banks and MSBs, the distributions of the Boone indices based on total assets and total loans are qualitatively similar. In the commercial bank industry, Boone indices, coefficients of ln(MC), are generally negative and significant in early 2000 in terms of the market share of total assets. With respect to total loans, however, the Boone indices are significant until 2007, consistent with the fact that commercial banks competed fiercely for household loans in the mid-2000s due to the house price boom at this time. In the MSB industry, there existed considerable market pressure during the mid-2000s when the Korean government relaxed loan regulations drastically in order to support microfinance and promote small loan markets for low-income groups and low-rated SMEs. Likewise, real estate project financing loans increased steeply during this period, as a result of the real estate boom in the mid-2000s.
Table 2 Boone index for the whole period. The table reports the results of the following regression specification for MSBs and commercial banks for the full sample period:
ln MSiτð Þ ¼ α þ βln MCiτð Þ þ eiτ
where MS denotes market shares, MC is marginal costs, and the coefficients stand for the Boone index. All tests are based on robust standard errors and the symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively.
MSBs Commercial banks
Coef. t P N |t| Coef. t P N |t|
Panel A. Dependent variable: ln(market share) based on total assets ln(MC) −0.581 −7.20 0.000⁎⁎⁎ −0.348 −7.25 0.000⁎⁎⁎
Intercept 0.010 0.16 0.869 2.080 16.99 0.000⁎⁎⁎
No. of obs 1970 860 F tests 51.82 52.54 R2 0.033 0.064
Panel B. Dependent variable: ln(market share) based on total loans ln(MC) −0.913 −10.57 0.000⁎⁎⁎ −0.432 −9.76 0.000⁎⁎⁎
Intercept −4.377 −64.35 0.000⁎⁎⁎ −2.341 −19.87 0.000⁎⁎⁎
No. of obs 1777 860 F tests 111.78 95.18 R2 0.073 0.104
Panel C. Dependent variable: ln(market share) based on total commercial loans for MSBs and on total household loans for commercial banks
ln(MC) −1.023 −9.36 0.000⁎⁎⁎ −1.005 −12.06 0.000⁎⁎⁎
Intercept −3.144 −36.01 0.000⁎⁎⁎ −1.467 −7.48 0.000⁎⁎⁎
No. of obs 1777 854 F tests 87.55 145.43 R2 0.055 0.223
Table 3 Estimation of quarterly Boone indices. The table reports estimated of quarterly Boone indices from the following regression for MSBs and commercial banks:
ln MSitð Þ ¼ α þ X2011Q4
τ¼1999Q1 βτ ln MCiτð Þ � Quarter Dummyτ þ δτQuarter Dummyτ þ εiτ
whereMS denotes market shares,MC is marginal costs, and β represents Boone indices for each quarter. All tests are based on robust standard errors and the symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively.
Commercial banks MSBs
Total assets Total loans Total assets Total loans
Year Quarter Coef. t-Value Coef. t-Value Coef. t-Value Coef. t-Value
1999 1 −1.464 −3.36⁎⁎⁎ −1.202 −2.91⁎⁎⁎
2 −0.177 −0.49 −0.097 −0.25 3 −0.885 −7.34⁎⁎⁎ −0.816 −6.79⁎⁎⁎
4 −0.232 −0.56 −0.221 −0.67 2000 1 −1.104 −2.09⁎⁎ −1.085 −2.51⁎⁎
2 −1.458 −5.76⁎⁎⁎ −1.379 −5.88⁎⁎⁎
3 −1.000 −10.19⁎⁎⁎ −0.982 −9.91⁎⁎⁎
4 −0.836 −2.90⁎⁎⁎ −0.848 −3.35⁎⁎⁎
2001 1 −0.713 −1.11 −0.800 −1.41 2 −1.270 −2.24⁎⁎ −1.363 −2.84⁎⁎⁎
3 −0.831 −3.09⁎⁎⁎ −0.859 −3.09⁎⁎⁎
4 −1.024 −2.55⁎⁎ −1.196 −3.85⁎⁎⁎
2002 1 −0.832 −1.32 −1.017 −1.92⁎
2 −1.044 −2.50⁎⁎ −1.235 −3.94⁎⁎⁎
3 −1.171 −2.47⁎⁎ −1.342 −3.66⁎⁎⁎
4 −1.132 −3.17⁎⁎⁎ −1.279 −4.69⁎⁎⁎
2003 1 −0.888 −1.93⁎ −1.095 −3.12⁎⁎⁎
2 −1.158 −3.18⁎⁎⁎ −1.331 −5.14⁎⁎⁎
3 −0.881 −1.91⁎ −1.096 −3.29⁎⁎⁎
4 −1.038 −2.54⁎⁎ −1.185 −3.96⁎⁎⁎ −0.151 −0.35 −3.315 −9.01⁎⁎⁎
2004 1 −0.869 −1.31 −1.105 −2.07⁎⁎ −0.275 −0.69 −0.396 −0.94 2 −1.156 −1.96⁎ −1.360 −3.06⁎⁎⁎ −0.568 −1.30 −0.888 −2.03⁎⁎
3 −1.142 −1.96⁎ −1.336 −3.03⁎⁎⁎ 0.410 0.82 0.296 0.51 4 −0.476 −0.84 −0.794 −1.73⁎ −0.969 −2.23⁎⁎ −1.235 −2.69⁎⁎⁎
2005 1 −0.673 −0.97 −0.960 −1.71⁎ −0.422 −1.16 −0.514 −1.30 2 −0.879 −1.28 −1.150 −2.11⁎⁎ −0.807 −2.67⁎⁎⁎ −1.275 −3.71⁎⁎⁎
3 −1.024 −1.52 −1.259 −2.40⁎⁎ −0.814 −3.09⁎⁎⁎ −0.980 −3.57⁎⁎⁎
4 −1.053 −1.69⁎ −1.324 −2.83⁎⁎⁎ −0.981 −3.20⁎⁎⁎ −1.413 −4.04⁎⁎⁎
2006 1 −0.576 −0.88 −0.892 −1.63 −1.127 −3.87⁎⁎⁎ −1.560 −4.85⁎⁎⁎
2 −0.607 −0.99 −0.895 −1.96⁎ −1.207 −4.62⁎⁎⁎ −1.869 −5.10⁎⁎⁎
3 −0.806 −1.13 −1.119 −2.01⁎⁎ −1.047 −3.64⁎⁎⁎ −1.185 −4.75⁎⁎⁎
4 −0.986 −1.35 −1.292 −2.33⁎⁎ −1.560 −3.90⁎⁎⁎ −1.658 −3.84⁎⁎⁎
2007 1 −0.585 −0.94 −0.881 −1.85⁎ −1.409 −3.51⁎⁎⁎ −1.646 −3.82⁎⁎⁎
2 −0.799 −1.35 −1.012 −2.29⁎⁎ −0.855 −2.18⁎⁎ −1.473 −3.34⁎⁎⁎
3 −0.611 −1.19 −0.842 −2.09⁎⁎ −0.330 −1.18 −0.689 −3.14⁎⁎⁎
4 −0.553 −1.01 −0.807 −1.84⁎ −0.022 −0.09 −0.418 −1.60 2008 1 −0.141 −0.38 −0.414 −1.32 2.043 1.95⁎ 1.724 1.90⁎
2 −0.342 −0.81 −0.609 −1.70⁎ 1.548 1.30 1.173 1.11 3 0.000 0.00 −0.288 −1.02 −1.069 −1.28 −0.930 −1.01 4 0.469 1.44 0.162 0.51 1.133 0.81 1.048 0.72
2009 1 0.439 1.69⁎ 0.125 0.47 −0.125 −0.20 −0.127 −0.20 2 −0.651 −1.12 −0.860 −1.60 0.331 0.27 −0.033 −0.03 3 −0.460 −0.67 −0.749 −1.24 0.358 0.87 −0.025 −0.05 4 −0.618 −1.06 −0.826 −1.56 0.008 0.02 −0.499 −1.13
2010 1 −0.273 −0.52 −0.556 −1.11 −0.228 −0.51 −0.632 −1.64 2 −0.138 −0.23 −0.407 −0.71 −0.808 −1.12 −1.312 −2.00⁎⁎
3 −0.751 −1.12 −0.991 −1.64 −0.527 −0.75 −0.707 −0.89 4 −1.438 −1.93⁎ −1.637 −2.50⁎⁎ −0.026 −0.02 −0.311 −0.22
2011 1 −0.111 −0.25 −0.253 −0.63 −0.995 −0.77 −1.437 −1.13 2 −0.185 −0.40 −0.344 −0.82 −1.176 −1.06 −1.748 −1.80⁎
3 0.602 1.40 0.427 1.00 −0.090 −0.08 −0.290 −0.25
262 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
Table 3 (continued)
Commercial banks MSBs
Total assets Total loans Total assets Total loans
Year Quarter Coef. t-Value Coef. t-Value Coef. t-Value Coef. t-Value
4 0.011 0.02 −0.110 −0.19 −0.489 −0.89 −0.972 −1.86⁎
Quarter dummies included Quarter dummies included R2 0.197 0.274 0.092 0.148
263J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
4.2. Competition and stability in the MSB industry
Before conducting regression analysis, we first take an initial look at the correlations between the independent variables. By obtaining the correlation coefficients between these figures, we can become aware of potential multicollinearity. As reported in Table 4, in the MSB sample, all the correlation coefficients are below around 10%. In the case of the commercial banks, however, the largest coefficient is 36.1%, between the size of firm and loan to deposit ratio. In both cases, we can conclude that the correlation coefficients between explanatory variables are rather small and thus muticollinearity is not likely to affect the results of our regressions.
Using the Boone indices and the concentration measures estimated in the previous sections, we examine the effect of competition on stability in the MSB industry. We conduct OLS regressions as a baseline-model as well as panel analysis to control for the time invariant heterogeneity of each bank. Table 5 reports the results of our regressions, where the dependent variables in Panel A and Panel B are the Z-scores using ROAs and the capital ratios for prior four quarters and for prior eight quarters, respectively. The key independent variables are a dummy for the presence of market pressure (Competition) and the opposite of the Boone indices (Inverse of Boone Index). The significant and positive coefficients for Competition and the Inverse of Boone Index in both the OLS and fixed effects models suggest that the probability of MSB defaults is lower in competitive markets, consistent with our hypothesis that the higher interest rates charged by MSBs in a less competitive market would increase the risk-taking behavior of borrowers, most of whom are SMEs possessing greater business and credit risk (Boyd and De Nicolo, 2005). As was previously noted, Martinez-Miera and Repullo (2010) argue that the risk shifting effect overwhelms the interest effect for MSBs when the market is less competitive.
Note that most of the coefficients of our control variables are signed in accordance with our expectations and prior literature. The profit ratio is positively correlated with MSB stability. The loan to
Table 4 Correlation coefficients matrix. The table presents Pearson correlation coefficients matrix for the variables used in our main analysis. Variable definitions are provided in Appendix A.
Competition Inverse of Boone index ln(assets) Profit ratio Loan to deposit
Panel A. MSBs Competition 1 Inverse of Boone index 0.573 1 ln(assets) −0.134 −0.105 1 Profit ratio 0.057 0.037 −0.015 1 Loan to deposit 0.091 0.040 0.006 0.118 1 Commercial to home loan −0.121 −0.088 0.386 −0.034 0.046
Panel B. Commercial banks Competition 1 Inverse of Boone index 0.894 1 ln(assets) −0.145 −0.155 1 Profit ratio −0.203 −0.164 0.075 1 Loan to deposit −0.168 −0.177 0.361 0.076 1 Commercial to home loan −0.049 −0.061 0.185 0.032 0.244
Table 5 Effects of competition on stability in the MSB industry. OLS and panel regressions are estimated for the stability of MSBs. The dependent variable is the Z-score using the ROAs for prior four or eight quarters. Year dummy is included but the coefficients are not reported. The t-statistics based on robust standard errors are reported in brackets below coefficient values. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively. Variable definitions are provided in Appendix A.
Panel A. Analysis of Z-score with the ROAs and capital ratios for prior four quarters
ln(Z4) OLS Fixed effects
Coef. t-Value Coef. t-Value Coef. t-Value Coef. t-Value
Competition 0.4987 [8.88]⁎⁎⁎ 0.5551 [6.87]⁎⁎⁎
Inverse of Boone index 0.2716 [7.92]⁎⁎⁎ 0.2756 [6.18]⁎⁎⁎
ln(assets) −0.0357 [−1.23] −0.0410 [−1.40] 0.0120 [0.08] −0.1399 [−0.97] Profit ratio 0.0042 [2.10]⁎⁎ 0.0043 [2.11]⁎⁎ 0.0035 [1.96]⁎ 0.0036 [1.93]⁎
Loan to deposit 0.0465 [3.92]⁎⁎⁎ 0.0455 [3.72]⁎⁎⁎ 0.0567 [1.59] 0.0544 [1.52] Loan to deposit2 −0.0002 [−3.29]⁎⁎⁎ −0.0002 [−3.03]⁎⁎⁎ −0.0003 [−1.45] −0.0002 [−1.31] Commercial to home loan
0.0019 [1.37] 0.0013 [0.87] −0.0074 [−1.70]⁎ −0.0083 [−1.85]⁎
CD volatility −0.1801 [−2.95]⁎⁎⁎ −0.2921 [−4.98]⁎⁎⁎ −0.1878 [−2.64]⁎⁎⁎ −0.2785 [−3.81]⁎⁎⁎
Intercept −1.7630 [−3.04]⁎⁎⁎ −1.5559 [−2.62]⁎⁎⁎ −2.5086 [−1.25] −1.1059 [−0.55] No. of obs 2137 2137 2137 2137 F/Wald tests 24.95 21.29 14.68 12.38 R2 0.084 0.074 0.072 0.052
Panel B. Analysis of Z-score with the ROAs and capital ratios for prior eight quarters
ln(Z8) OLS Fixed effects
Coef. t-Value Coef. t-Value Coef. t-Value Coef. t-Value
Competition 0.5543 [9.73]⁎⁎⁎ 0.6905 [7.75]⁎⁎⁎
Inverse of Boone index 0.2614 [7.65]⁎⁎⁎ 0.2847 [6.71]⁎⁎⁎
ln(assets) −0.0024 [−0.08] −0.0106 [−0.37] 0.2352 [1.22] −0.0464 [−0.27] Profit ratio 0.0059 [2.17]⁎⁎ 0.0059 [2.15]⁎⁎ 0.0056 [2.06]⁎⁎ 0.0056 [2.05]⁎⁎
Loan to deposit 0.0526 [3.88]⁎⁎⁎ 0.0514 [3.80]⁎⁎⁎ 0.0684 [2.21]⁎⁎ 0.0731 [2.41]⁎⁎
Loan to deposit 2 −0.0002 [−3.16]⁎⁎⁎ −0.0002 [−2.95]⁎⁎⁎ −0.0003 [−1.83]⁎ −0.0003 [−1.88]⁎
Commercial to home loan
0.0002 [0.18] −0.0004 [−0.32] −0.0061 [−1.74]⁎ −0.0075 [−1.92]⁎
CD volatility −0.1574 [−2.51]⁎⁎ −0.2665 [−4.43]⁎⁎⁎ −0.1685 [−2.36]⁎⁎ −0.2626 [−3.61]⁎⁎⁎
Intercept −2.8477 [−4.37]⁎⁎⁎ −2.6309 [−4.05]⁎⁎⁎ −5.6183 [−2.89]⁎⁎⁎ −3.5127 [−1.95]⁎
No. of obs 1764 1764 1764 1764 F/Wald tests 27.87 21.83 16.90 15.14 R2 0.115 0.097 0.090 0.084
264 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
deposit ratio increases the stability of MSBs but the negative effect of the squared loan to deposit ratio suggests that its increasing rate in fact decreases. Market volatility, measured by CD volatility, significantly increases the defaults of MSBs.
In Table 6, we additionally investigate the relationship between the degree of concentration and the stability in the MSB market. According to the structure–conduct–performance view, the number of firms and their market shares determine the competitive level of the industry. The CR and HHI, which were discussed in Section 3.2, are used as proxies of market concentration and, more specifically, CR4 and CR8 represent the concentration ratio of the top four and eight banks, respectively. As in Table 5, we use the Z-scores as the dependent variables and estimate OLS and fixed effects regressions. The results show that the coefficients for the concentration measure are significantly negative, which is consistent with the results in Table 5 given that greater concentration ratios imply less competitive markets. However, we also find that the squared terms of concentration ratios are positively correlated with the Z-scores, which suggests that the relationship between market concentration and MSB stability is in fact non-linear. The coefficients of control variables are signed in a way that is consistent with the results of the previous table.
Table 6 Effects of concentration on stability in the MSB industry. OLS and panel regressions are estimated for the stability of MSBs as a function of the concentration ratio (CR) or Herfindahl–Hirschman index (HHI). The dependent variable is the Z-score using the ROAs and capital ratios for prior four or eight quarters. Year dummy is included but the coefficients are not reported. The t-statistics based on robust standard errors are reported in brackets below coefficient values. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively. Variable definitions are provided in Appendix A.
Panel A. Analysis of Z-score with the ROAs and capital ratios for prior four quarters
ln(Z4) OLS Fixed effects
Coef. Coef. Coef. Coef. Coef. Coef. Coef. Coef.
[t-Value] [t-Value] [t-Value] [t-Value] [t-Value] [t-Value] [t-Value] [t-Value]
CR4 −0.1288⁎⁎⁎ −1.8701⁎⁎ −0.1735⁎⁎⁎ −2.1478⁎⁎
[−7.38] [−2.25] [−4.42] [−2.32] (CR4)2 0.0274⁎⁎ 0.0314⁎⁎
[2.10] [2.15] HHI −0.0122⁎⁎⁎ −0.5760⁎⁎⁎ −0.0160⁎⁎⁎ −0.5849⁎⁎⁎
[−6.22] [−7.15] [−3.39] [−6.38] HHI2 0.0013⁎⁎⁎ 0.0013⁎⁎⁎
[7.01] [6.21] ln(assets) −0.0309 −0.0303 −0.0344 −0.0311 0.1579 0.0401 0.0774 0.0339
[−1.06] [−1.03] [−1.17] [−1.06] [0.84] [0.73] [0.40] [0.60] Profit ratio 0.0041⁎⁎ 0.0041⁎⁎ 0.0043⁎⁎ 0.0040⁎⁎ 0.0033⁎ 0.0037⁎⁎ 0.0036⁎ 0.0036⁎⁎
[2.05] [2.09] [2.12] [2.04] [1.85] [2.07] [1.95] [2.05] Loan to Deposit
0.0458⁎⁎⁎ 0.0456⁎⁎⁎ 0.0460⁎⁎⁎ 0.0451⁎⁎⁎ 0.0543 0.0525⁎ 0.0548 0.0527⁎
[4.07] [4.01] [4.06] [3.75] [1.54] [1.88] [1.54] [1.82] Loan to deposit2
−0.0002⁎⁎⁎ −0.0002⁎⁎⁎ −0.0002⁎⁎⁎ −0.0002⁎⁎⁎ −0.0003 −0.000⁎3 −0.0003 −0.0002 [−3.53] [−3.44] [−3.48] [−3.07] [−1.53] [−1.72] [−1.50] [−1.58]
Commercial to
0.0024⁎ 0.0023 0.0021 0.0019 −0.0063 −0.0034 −0.0069 −0.0041
Home loan [1.72] [1.61] [1.46] [1.38] [−1.41] [−0.96] [−1.52] [−1.22] CD volatility
−0.2215⁎⁎⁎ −0.1797⁎⁎⁎ −0.2502⁎⁎⁎ −0.0681 −0.2407⁎⁎⁎ −0.1824⁎⁎⁎ −0.2610⁎⁎⁎ −0.0748 [−3.61] [−2.80] [−4.12] [−1.04] [−3.35] [−2.66] [−3.61] [−1.12]
Intercept 2.5578⁎⁎⁎ 30.103⁎⁎9 1.1709⁎ 62.8115⁎⁎⁎ 2.3239 33.9917⁎⁎ 0.9291 63.1114⁎⁎⁎
[3.31] [2.29] [1.68] [7.14] [1.18] [2.34] [0.47] [6.42] No. of obs 2137 2137 2137 2137 2137 2137 2137 2137 F/Wald tests
22.93 20.27 20.18 21.95 12.41 92.18 10.89 94.76
R2 0.075 0.076 0.067 0.088 0.055 0.070 0.061 0.081
Panel B. Analysis of Z-score with the ROAs for prior eight quarters
ln(Z8) OLS Fixed effects
Coef. t-Value Coef. t-Value Coef. t-Value Coef. t-Value
CR8 −0.1274 [−7.10]⁎⁎⁎ −0.1818 [−4.40]⁎⁎⁎
HHI −0.0106 [−5.16]⁎⁎⁎ −0.0136 [−2.80]⁎⁎⁎
ln(assets) −0.0018 [−0.06] −0.0089 [−0.31] 0.2777 [1.22] 0.0700 [0.31] Profit ratio 0.0057 [2.13]⁎⁎ 0.0060 [2.18]⁎⁎ 0.0052 [1.95]⁎ 0.0055 [2.03]⁎⁎
Loan to deposit
0.0539 [4.28]⁎⁎⁎ 0.0541 [4.23]⁎⁎⁎ 0.0658 [1.99]⁎ 0.0695 [2.10]⁎⁎
Loan to deposit2
−0.0002 [−3.59]⁎⁎⁎ −0.0002 [−3.49]⁎⁎⁎ −0.0003 [−1.78]⁎ −0.0003 [−1.80]⁎
Commercial to home loan
0.0005 [0.39] 0.0000 [0.03] −0.0054 [−1.45] −0.0065 [−1.59]
CD volatility −0.2159 [−3.48]⁎⁎⁎ −0.2527 [−4.11]⁎⁎⁎ −0.2346 [−3.28]⁎⁎⁎ −0.2609 [−3.59]⁎⁎⁎
Intercept 1.4092 [1.68]⁎ −0.2816 [−0.37] 0.4300 [0.24] −0.9431 [−0.52] No. of obs 1764 1764 1764 1764 F/Wald tests
22.58 18.24 10.17 8.42
R2 0.097 0.083 0.068 0.074
265J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
Table 7 Effects of competition on stability in the commercial bank industry. OLS and panel regressions are estimated for the stability of commercial banks. The dependent variable is the Z-score of commercial banks using the ROAs and capital ratios for prior four or eight quarters. Year dummy is included but the coefficients are not reported. The t/z-statistics based on robust standard errors are reported in brackets below coefficient values. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively. Variable definitions are provided in Appendix A.
Panel A. Analysis of Z-score with the ROAs and capital ratios for prior four quarters
ln(Z4) OLS Random effects
Coef. t-Value Coef. t-Value Coef. z-Value Coef. z-Value
Competition −0.5224 [−5.77]⁎⁎⁎ −0.3989 [−5.17]⁎⁎⁎
Inverse of Boone index −1.0715 [−3.43]⁎⁎⁎ −0.5305 [−2.45]⁎⁎
Inverse of Boone index2 0.5037 [1.79]⁎ 0.1764 [1.20] ln(assets) −0.1554 [−3.92]⁎⁎⁎ −0.1552 [−3.95]⁎⁎⁎ 0.3811 [2.28]⁎⁎ 0.5541 [2.98]⁎⁎⁎
Profit ratio 3.7195 [5.22]⁎⁎⁎ 3.8649 [5.48]⁎⁎⁎ 3.5188 [4.43]⁎⁎⁎ 3.6598 [4.85]⁎⁎⁎
Foreign 0.6560 [5.06]⁎⁎⁎ 0.6556 [5.05]⁎⁎⁎ 0.5303 [1.69]⁎ 0.4775 [1.33] Loan to deposit 0.0367 [8.18]⁎⁎⁎ 0.0352 [7.89]⁎⁎⁎ 0.0329 [2.44]⁎⁎ 0.0296 [2.25]⁎⁎
Loan to deposit2 −0.0001 [−6.91]⁎⁎⁎ −0.0001 [−6.64]⁎⁎⁎ −0.0001 [−2.29]⁎⁎ −0.0001 [−2.18]⁎⁎
Commercial to home loan 0.0004 [1.44] 0.0003 [1.21] 0.0007 [1.10] 0.0007 [1.06] CD volatility −0.1230 [−1.11] −0.2707 [−2.28]⁎⁎ −0.1994 [−1.34] −0.2971 [−1.97]⁎⁎
Intercept 3.6527 [6.44]⁎⁎⁎ 3.8079 [6.78]⁎⁎⁎ −5.3765 [−2.30]⁎⁎ −8.1030 [−3.03]⁎⁎⁎
No. of obs 738 738 738 738 F/Wald tests 32.16 29.52 443.36 473.39 R2 0.240 0.249 0.329 0.350
Panel B. Analysis of Z-score with the ROAs and capital ratios for prior eight quarters
ln(Z8) OLS Random effects
Coef. t-Value Coef. t-Value Coef. z-Value Coef. z-Value
Competition −0.5812 [−6.79]⁎⁎⁎ −0.4712 [−5.14]⁎⁎⁎
Inverse of Boone index −1.6732 [−6.29]⁎⁎⁎ −1.1938 [−5.27]⁎⁎⁎
Inverse of Boone index2 0.9888 [4.09]⁎⁎⁎ 0.6597 [3.74]⁎⁎⁎
ln(assets) −0.1555 [−4.17]⁎⁎⁎ −0.1559 [−4.25]⁎⁎⁎ 0.3199 [1.70]⁎ 0.3393 [1.77]⁎
Profit ratio 1.9727 [2.57]⁎⁎ 2.2290 [2.96]⁎⁎⁎ 1.7940 [2.06]⁎⁎ 2.0102 [2.41]⁎⁎
Foreign 0.5775 [4.65]⁎⁎⁎ 0.5796 [4.80]⁎⁎⁎ 0.4631 [1.48] 0.4515 [1.43] Loan to deposit 0.0303 [7.87]⁎⁎⁎ 0.0280 [7.36]⁎⁎⁎ 0.0279 [2.02]⁎⁎ 0.0250 [1.94]⁎
Loan to deposit2 −0.0001 [−6.41]⁎⁎⁎ −0.0001 [−5.74]⁎⁎⁎ −0.0001 [−1.82]⁎ −0.0001 [−1.74]⁎
Commercial to home loan 0.0001 [0.66] 0.0001 [0.26] 0.0005 [0.81] 0.0005 [0.72] CD volatility −0.2519 [−2.22]⁎⁎ −0.4390 [−3.65]⁎⁎⁎ −0.3248 [−3.98]⁎⁎⁎ −0.4664 [−6.26]⁎⁎⁎
Intercept 3.9713 [7.47]⁎⁎⁎ 4.2033 [8.07]⁎⁎⁎ −4.1093 [−1.57] −4.1711 [−1.54] No. of obs 682 682 682 682 F/Wald tests 30.08 29.83 443.48 713.57 R2 0.223 0.251 0.343 0.359
266 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
4.3. Competition and stability in the commercial bank industry
In this section, we examine whether the relationship between competition and stability is different between MSBs and commercial banks. Table 7 reports the estimates of OLS and panel regressions for the Z-scores of commercial banks as a function of the measures for market competition. In the first regression of Panel A, where the key independent variable is Competition, a dummy for the presence of competitive pressure, the negative and significant coefficient implies that, consistent with the competition–fragility view, the presence of competition significantly increases the probability of bank defaults. In the second regression, the level of competition, measured by Inverse of Boone Index, is negatively correlated with the Z-score, while its squared term has a positive effect. That is, market competition significantly decreases the stability of commercial banks, but the effect is non-linear, consistent with the non-linear U-shaped relationship between competition and stability that is suggested by Tabak et al. (2012a). The results are
267J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
qualitatively similar when panel regressions are run and when the Z-score with the ROAs for prior eight quarters is used in Panel B.10
Most of the coefficients of our control variables are signed as expected, except for those of the size of banks. The table shows the negative effect of ln(assets) on bank stability. Boyd et al. (2006) point out that competition and bank size are endogenously determined and thus one may need to employ instrumental variable estimations in this area. Without endogeneity corrections, they also report the negative effect of bank size. Tabak et al. (2012a) provide a possible explanation for this result in terms of the negative relationship between capital ratio and bank size. They argue that the larger a bank is, the more it benefits from competition, and that a greater capital ratio is advantageous for banks in the less competitivemarkets.
4.4. Additional analysis
Thus far, we have used market shares calculated based on total assets. In this section, we re-estimate the Boone indices using the market shares of total loans, in light of the fact that competitive pressure is more intensified in loan markets compared to deposit markets. We also re-calculate the market shares of commercial loans and household loans and see whether our results change.
In Table 8, we alternatively estimate the Boone indices using market shares based on total loans and re-examine the relationship between competition and stability in the MSB and commercial bank markets. The Z-score with the ROAs and capital ratios for the prior four quarters is used as the dependent variable. The results confirm our previous findings that the relationship between competition and stability varies depending on the type of bank involved. In Panel A, the coefficients for Competition and Inverse of Boone Index are positive and significant, suggesting that, consistent with Tables 5 and 6, competitive pressure enhances the stability of MSBs. Panel B reports that commercial banks are less vulnerable in less competitive markets, while the effect of competition is non-linear. The results are consistent with Table 7.
In Table 9, we re-estimate the Boone indices using market shares based on commercial loans for MSBs and on household loans for commercial banks, so as to reflect that MSBs concentrate their finance on SMEs while commercial banks compete to attract households. We then test whether our previous findings are sensitive to the definition of market shares. Panel A reports the determinants of the stability of MSBs as a function of competition re-estimated by using market shares of commercial loans. In both OLS and panel regressions, the coefficients for Competition and Inverse of Boone Index are still negative and statistically significant, confirming that MSBs tend to be more stable in more competitive markets. In Panel B, we examine the commercial bank industry and test the effects of competition based on its market shares of household loans. Again, the competition dummy is negatively correlated with commercial bank stability. Likewise, the Boon indices have a negative effect, but the relationship between Boon indices and stability is non-linear. Overall, our finding that the relationship between competition and stability varies depending on the characteristics of banks is robust to a variety of measures of competition and stability as well as model specifications.
4.5. Difference-in-differences analysis
Since the mid-2000s in accordance with drastic deregulations by the government in order to support microfinance and promote small loan markets, MSBs have aggressively expanded their microcredit loans and
10 We employ random effects models rather than fixed effects models for our panel analysis in order to estimate the effect of foreign banks, Foreign. Note that fixed effects regressions do not provide estimates for time-invariant variables such as Foreign. We also conduct fixed effects regressions, the results of which, as shown below, are similar to those of Table 7. More detailed results are available upon request.
ln(Z4) Inverse of Boone
Inverse of Boone2
ln(assets) Profit ratio
Loan to deposit
Loan to deposit2
Commercial to home loan
CD volatility
Intercept
Coef. −0.2920 0.2820 −2.0130 3.8959 0.0080 0.0000 0.0005 −0.2897 −31.7132 P 0.06⁎ 0.02⁎⁎ 0.00⁎⁎ 0.00⁎⁎ 0.58 0.22 0.32 0.08⁎ 0.00⁎⁎
No. of Obs: 738, F/Wald tests: 78.67, R2: 0.416. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% levels, respectively.
Table 8 Effects of competition based on total loans. In this table, the Boone index is re-estimated using the market share based on total loans. OLS and panel regressions are estimated for the stability of both MSBs and commercial banks. The dependent variable is the Z-score using the ROAs and capital ratios for prior four quarters. Year dummy is included but the coefficients are not reported. The t/z-statistics based on robust standard errors are reported inbrackets belowcoefficient values. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively. Variable definitions are provided in Appendix A.
Panel A. Competition and stability in the MSB industry
ln(Z4) OLS Fixed effects
Coef. t-Value Coef. t-Value Coef. t-Value Coef. t-Value
Competition 0.1461 [2.84]⁎⁎⁎ 0.1421 [2.49]⁎⁎
Inverse of Boone index 0.1120 [3.59]⁎⁎⁎ 0.1127 [3.65]⁎⁎⁎
ln(assets) −0.0769 [−3.51]⁎⁎⁎ −0.0789 [−3.61]⁎⁎⁎ −0.3346 [−2.62]⁎⁎ −0.3471 [−2.81]⁎⁎⁎
Profit ratio 0.0027 [4.11]⁎⁎⁎ 0.0026 [3.95]⁎⁎⁎ 0.0006 [0.66]⁎⁎ 0.0005 [0.58] Loan to deposit 0.0002 [2.48]⁎⁎ 0.0002 [2.46]⁎⁎ 0.0006 [2.21]⁎⁎ 0.0006 [2.31]⁎⁎
Loan to deposit2 0.0000 [−1.16] 0.0000 [−1.09] 0.0000 [−1.83] 0.0000 [−1.89]⁎
Commercial to home loan
0.0043 [3.08]⁎⁎⁎ 0.0043 [3.09]⁎⁎⁎ −0.0038 [−1.17] −0.0038 [−1.17]
CD volatility −0.2935 [−4.51]⁎⁎⁎ −0.2970 [−4.85]⁎⁎⁎ −0.2961 [−4.16]⁎⁎⁎ −0.2930 [−4.22]⁎⁎⁎
Intercept −0.2260 [−0.52] −0.2219 [−0.51] 0.2526 [0.15] 0.2345 [0.14] No. of obs 2164 2164 2164 2164 F/Wald tests 20.10 20.83 9.52 9.98 R2 0.058 0.063 0.083 0.864
Panel B. Competition and stability in the commercial bank industry
ln(Z4) OLS Random effects
Coef. t-Value Coef. t-Value Coef. z-Value Coef. z-Value
Competition −0.5479 [−5.40]⁎⁎⁎ −0.3068 [−3.00]⁎⁎⁎
Inverse of Boone index −1.4628 [−5.27]⁎⁎⁎ −1.1246 [−4.58]⁎⁎⁎
Inverse of Boone index2
0.8295 [4.10]⁎⁎⁎ 0.6944 [3.93]⁎⁎⁎
ln(assets) −0.1648 [−4.21]⁎⁎⁎ −0.1670 [−4.29]⁎⁎⁎ 0.5067 [2.79]⁎⁎⁎ 0.3463 [3.77]⁎⁎⁎
Profit ratio 4.8229 [6.92]⁎⁎⁎ 4.9438 [7.04]⁎⁎⁎ 4.2733 [5.87]⁎⁎⁎ 4.4288 [7.56]⁎⁎⁎
Foreign 0.6879 [5.23]⁎⁎⁎ 0.6929 [5.29]⁎⁎⁎ 0.5316 [1.53] 0.5862 [1.54] Loan to deposit 0.0388 [8.42]⁎⁎⁎ 0.0387 [8.33]⁎⁎⁎ 0.0349 [2.44]⁎⁎ 0.0373 [7.68]⁎⁎⁎
Loan to deposit2 −0.0001 [−6.88]⁎⁎⁎ −0.0001 [−6.77]⁎⁎⁎ −0.0001 [−2.22]⁎⁎ −0.0001 [−6.87]⁎⁎⁎
Commercial to home loan
0.0003 [0.94] 0.0002 [0.83] 0.0007 [0.89] 0.0006 [1.47]
CD volatility −0.3547 [−2.86]⁎⁎⁎ −0.3380 [−2.74]⁎⁎⁎ −0.3409 [−2.18]⁎⁎ −0.3294 [−3.14]⁎⁎⁎
Intercept 3.8635 [6.70]⁎⁎⁎ 3.9032 [6.80]⁎⁎⁎ −7.6638 [−2.99]⁎⁎⁎ −5.0618 [−3.42]⁎⁎⁎
No. of obs 738 738 738 738 F/Wald tests 30.80 28.12 518.10 287.27 R2 0.234 0.240 0.338 0.326
268 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
real estate project financing loans. This aggressive loan policy by MSBs could expand their earnings during the real estate boom, while it became the main cause of increased risk during and after the global financial crisis from 2007. In this section, we examinewhetherMSBs have been riskier compared to commercial banks aswell as how the magnitude of the effect of competition on stability has changed since 2007.
In order to answer these questions, we employ the Difference-in-Differences (DID) approach to identify the effects of the 2007 financial crisis, using MSBs as a treatment group and commercial banks as a control group. Specifically, we estimate the following panel regressions.
Stabilityit ¼ α þ β1MSBDit � 2007Dit þ β22007Dit þ β3Competition Measuresit þ XN k¼1
γkXk þ eit
Stabilityit ¼ α þ β1MSBDit � 2007Dit � Competition Measuresit þ ρ22007Dit
þρ3Compertition Measuresit þ XN k¼1
φkXk þ uit
Table 9 Effects of competition based on commercial or house loans. In this table, the Boone index is re-estimated using the market share based on commercial loans for MSBs and on household loans for commercial banks. OLS and panel regressions are estimated for the stability of both MSBs and commercial banks. The dependent variable is the Z-score using the ROAs and capital ratios for prior four quarters. Year dummy is included but the coefficients are not reported. The t/z-statistics based on robust standard errors are reported in brackets below coefficient values. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively. The variable definitions are provided in Appendix A.
Panel A. Competition based on commercial loans and stability of MSBs
ln(Z4) OLS Fixed effects
Coef. t-Value Coef. t-Value Coef. t-Value Coef. t-Value
Competition 0.2122 [4.63]⁎⁎⁎ 0.2200 [3.75]⁎⁎⁎
Inverse of Boone index 0.0974 [4.06]⁎⁎⁎ 0.0975 [3.20]⁎⁎⁎
ln(assets) −0.0723 [−3.31]⁎⁎⁎ −0.0755 [−3.46]⁎⁎⁎ −0.2748 [−2.19]⁎⁎ −0.3083 [−2.43]⁎⁎
Profit ratio 0.0027 [4.09]⁎⁎⁎ 0.0027 [4.18]⁎⁎⁎ 0.0006 [0.66] 0.0007 [0.71] Loan to deposit 0.0002 [2.56]⁎⁎ 0.0002 [2.45]⁎⁎ 0.0006 [2.22]⁎⁎ 0.0006 [2.23]⁎
Loan to deposit2 0.0000 [−1.29] 0.0000 [−1.13] 0.0000 [−1.87]⁎ 0.0000 [−1.85] Commercial to home loan
0.0046 [3.31]⁎⁎⁎ 0.0045 [3.19]⁎⁎⁎ −0.0034 [−1.06] −0.0036 [−1.13]
CD volatility −0.2854 [−4.77]⁎⁎⁎ −0.3413 [−6.02]⁎⁎⁎ −0.2986 [−4.11]⁎⁎⁎ −0.3484 [−4.83]⁎⁎⁎
Intercept −0.2881 [−0.66] −0.1637 [−0.38] −0.2078 [−0.12] 0.1397 [0.08] No. of obs 2164 2164 2164 2164 F/Wald tests 22.13 21.38 10.76 10.03 R2 0.067 0.062 0.090 0.876
Panel B. Competition based on house loans and stability of commercial banks
ln(Z4) OLS Random effects
Coef. t-Value Coef. t-Value Coef. z-Value Coef. z-Value
Competition −0.5402 [−4.76]⁎⁎⁎ −0.2509 [−1.86]⁎
Inverse of Boone index −1.2158 [−8.42]⁎⁎⁎ −1.0331 [−8.14]⁎⁎⁎
Inverse of Boone index2
0.4295 [8.94]⁎⁎⁎ 0.3784 [9.60]⁎⁎⁎
ln(assets) −0.1617 [−4.11]⁎⁎⁎ −0.1618 [−4.25]⁎⁎⁎ 0.6043 [3.13]⁎⁎⁎ 0.1956 [1.29] Profit ratio 4.9084 [7.01]⁎⁎⁎ 3.7364 [5.37]⁎⁎⁎ 4.2732 [6.06]⁎⁎⁎ 3.3183 [4.34]⁎⁎⁎
Foreign 0.6918 [5.16]⁎⁎⁎ 0.6694 [5.34]⁎⁎⁎ 0.5124 [1.37] 0.6069 [2.29] Loan to deposit 0.0396 [8.42]⁎⁎⁎ 0.0349 [7.61]⁎⁎⁎ 0.0347 [2.39]⁎⁎ 0.0350 [2.40]⁎⁎
Loan to deposit2 −0.0001 [−6.79]⁎⁎⁎ −0.0001 [−5.80]⁎⁎⁎ −0.0001 [−2.19]⁎⁎ −0.0001 [−2.01]⁎⁎
Commercial to home loan
0.0003 [0.94] 0.0000 [0.06] 0.0007 [0.96] 0.0004 [0.56]
CD volatility −0.1936 [−1.59] −0.0552 [−0.48] −0.2527 [−1.67]⁎ −0.1115 [−0.70] Intercept 3.7491 [6.54]⁎⁎⁎ 3.9344 [7.14]⁎⁎⁎ −9.3874 [−3.39]⁎⁎⁎ −2.3821 [−1.17] No. of obs 738 738 738 738 F/Wald tests 29.59 34.38 419.01 1069.54 R2 0.233 0.277 0.347 0.354
269J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
where MSBD is a dummy variable equal to 1 for MSBs (treatment group) and 0 for commercial banks (control group) and 2007D is a dummy variable equal to 1 if an observation occurs after 2007 and 0, otherwise.
In the first regression, the panel estimate of the interaction term MSBD∙2007D, cβ1 is called the DID estimator which captures the mean difference in the stability between MSBs and commercial banks after 2007. Likewise, cρ1 in the second regression represents the difference in the effect of competition on stability between pre and after 2007 and between MSBs and commercial banks.11
Table 10 presents the estimation results of the DID model. The first regression reveals that the stability of MSBs experienced a significant decline after 2007. Specifically, the average stability of MSBs decreased
11 cρ1 can be called the difference-in-difference-in-differences (DDD) estimator (See Long et al. (2010) and a lecture note by Jeffrey Wooldridge entitled “What's New in Econometrics? Difference-in-Differences Estimation” at the NBER (National Bureau of Economic Research) Summer Institute 2007 (http://www.nber.org/WNE/Slides7-31-07/slides_10_diffindiffs.pdf). The estimator captures the time change in the average effect of competition for MSBs by netting out the change in the mean effect for commercial banks.
Table 10 Difference-in-differences approach for the stability of MSBs and commercial banks. The following panel regressions are estimated in this table:
Stabilityit ¼ α þ β1MSBDit � 2007Dit þ βz2007DitCompetition Measuresit þ XN k¼1
γkXk þ eit
Stabilityit ¼ α þ ρ1MSBDit � 2007Dit � Competition Measuresit þ ρz2007Dit þ ρzCompetition Measuresit þ XN k¼1
φkXk þ uit :
MSBD is a dummy variable equal to 1 for MSBs and 0 for commercial banks and 2007D is a dummy variable equal to 1 if the observation occurs after 2007 and 0, otherwise. The dependent variable is the Z-score using the ROAs and capital ratios for prior four quarters. The t-statistics based on robust standard errors are reported in brackets below coefficient values. The symbols ***, **, and * represent statistical significance at the 1%, 5%, and 10% level, respectively. Variable definitions are provided in Appendix A.
ln (Z4) Coef. t-Value Coef. t-Value Coef. t-Value
MSBD∙2007D∙competition 0.5141 [5.85]⁎⁎⁎
MSBD∙2007D∙inverse of Boone index 0.4416 [7.07]⁎⁎⁎
MSBD∙2007D −1.3411 [−16.50]⁎⁎⁎ −1.4952 [−17.60]⁎⁎⁎ −1.4813 [−17.98]⁎⁎⁎
2007D 0.4737 [6.05]⁎⁎⁎ 0.3127 [3.79]⁎⁎⁎ 0.2630 [3.20]⁎⁎⁎
Competition −0.1036 [−2.32]⁎⁎ −0.3455 [−5.69]⁎⁎⁎
Inverse of Boone index −0.4111 [−7.98]⁎⁎⁎
ln(assets) 0.4761 [6.69]⁎⁎⁎ 0.5502 [7.65]⁎⁎⁎ 0.5223 [7.61]⁎⁎⁎
Profit ratio 0.0002 [0.39] 0.0002 [0.24] 0.0010 [0.16] Deposit to loan 0.0007 [4.18]⁎⁎⁎ 0.0006 [4.06]⁎⁎⁎ 0.0007 [4.16]⁎⁎⁎
Deposit to loan2 0.0000 [−3.24]⁎⁎⁎ 0.0000 [−3.20]⁎⁎⁎ 0.0000 [−3.23]⁎⁎⁎
Commercial to home loan −0.0006 [−1.89]⁎ −0.0006 [−1.89]⁎ −0.0006 [−2.01]⁎⁎
CD volatility −0.3612 [−7.78]⁎⁎⁎ −0.3218 [−6.89]⁎⁎⁎ −0.3542 [−7.76]⁎⁎⁎
Intercept −5.5825 [−6.15]⁎⁎⁎ −6.0271 [−6.65]⁎⁎⁎ −5.6586 [−6.20]⁎⁎⁎
No. of obs 2902 2902 2902 F/Wald tests 53.88 52.5 47.88 R2 0.413 0.499 55.210
270 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
by 1.3411 compared to commercial banks after deregulation by the government and the global financial crisis. In the second and third regressions, we interact the competition measures with MSBD∙2007D. As discussed in Section 3.2, we use two proxies for competition; the opposite of the Boone index and Competition, a dummy variable which takes a value of 1 for a significantly negative Boone index and is otherwise 0. The table shows that the magnitude of the effect of competition on stability for MSBs is significantly greater after 2007 than that for commercial banks. When the market is competitive, the magnitude of an increase in the stability of MSBs after 2007 is 0.5141 greater than that of commercial banks. Overall, the results are consistent with our previous findings that MSBs with weak corporate governance tend to be more stable in competitive markets and this evidence is even stronger after the global financial crisis.
5. Conclusion
There are two alternative hypotheses that relate bank competition to stability. The conventional competition–fragility theory suggests that higher competition in financial industries causes financial institutions to lose their market power, leading to a decrease in profitability. In order to recover from financial losses, financial institutions are thus more likely to invest in riskier portfolios. Consequently, this risk-taking behavior will undermine the financial institutions' stability. On the other hand, competition– stability theory suggests that competition has a potentially positive effect on the stability of financial institutions. Accordingly, Boyd and De Nicolo (2005) show that banks with a greater market share in the loan market, which experience lower competition, tend to impose higher interest rates on their loans. A greater interest rate being charged by banks in a less competitive market may increase the risk-taking behavior of borrowing firms. Boyd and De Nicolo consequently argue that since the risk is ultimately shifted from borrowers to banks in this case, the default probability of banks increases in the riskiness of bank loans.
271J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
A recent study by Martinez-Miera and Repullo (2010) forms partial support for Boyd and De Nicolo's (2005) risk-shifting effect in which the greater interest rates that exist in less competitive markets raise the risk of loans and thus make bank bankruptcy more likely. However, this work additionally takes into account the fact that greater interest rates also improve bank profitability, which is known as the interest effect, and suggests that there exists a U-shaped relationship between competition and the risk of bank failure. Thus far, empirical research has produced mixed results on the issue of the influence of bank competition on stability.
The purpose of this paper has been to investigate the relationship between competition and stability by using a sample comprising two different types of banks: MSBs and commercial banks in Korea. By doing so, we attempt to fill a gap in the current literature by examining whether differences in terms of governance structure, business models and regulatory treatments cause banks to interact with industry competition differently. We use quarterly panel data for MSBs and commercial banks from 1999, the end of Asian financial crisis, to 2011. We follow Boone (2008a, 2008b) who measures competitiveness using operating efficiency, which is known as the Boone index. The conventional concentration ratio and the Herfindahl–Hirschman Index are also used as supplementary measures for bank competition. We employ the Z-score as a measure of bank stability, which can capture how distant a certain bank is from insolvency. Finally, we perform pooling regressions as well as panel analysis to examine the relationship between competitive levels and stability for both MSBs and commercial banks. Section 3 described our measures and methodology in more detail.
Our empirical results generally support the hypothesis that the effect of bank competition on stability differs depending on the characteristics of banks involved. The results of multiple regressions show that bank competition has a significant and positive effect on the stability of MSBs with weak corporate governance. On the contrary, competition pressure significantly reduces the stability of commercial banks but, consistent with Tabak et al. (2012a), this relationship turns out to be non-linear. These results are robust across alternative proxies for competition and stability as well as various model specifications. Therefore, we can conclude that for commercial banks at least, higher competition creates a trade-off between the risk-shifting effect and interest effect, but the risk shifting effect overwhelms the interest effect for MSBs when the market is less competitive. In addition, our difference-in-differences analysis shows that the positive effect of competition on the stability of MSBs became even greater after deregulations by the Korean government and after the global financial crisis.
The main contribution of the paper is that this is one of the first works to provide evidence on the relationship between bank competition and stability, as conditional on different bank characteristics. Our new evidence that the influence of bank competition on stability differs depending on the characteristics of banks involved provides important implications on competition policy in the banking industry. According to our empirical results, competition policy should be applied differently to commercial banks and MSBs. In addition, our paper is the first paper to analyze Korean MSBs by using hand-collected data.
Appendix A. Variable descriptions
Name Descriptions
ln(Z) Natural logarithm of the Z-score, described in Section 3.2. ln(Z4) and ln(Z8) are calculated using the ROAs and capital ratios for prior 4 quarters and prior 8 quarters, respectively.
Competition A dummy which takes a value of 1 if Boone index is significantly negative, i.e., if competitive pressure exists, and 0, otherwise.
Inverse of Boone index The opposite of Boone index, described in Section 3.2. Inverse of Boone index2 Squared inverse of Boone index. CR Concentration ratio. CRN is defined as the market share held by top N banks.
CR4 and CR8 represent the concentration ratio by top four and eight banks, respectively. HHI Herfindahl–Hirschman Index, calculated by squaring the market share of each
bank and then summing the squares. ln(assets) Natural logarithm of total assets of banks. Profit ratio Total net income to total revenues ratio.
(continued on next page)
Appendix A (continued)
Name Descriptions
Foreign A dummy equal to 1 for a foreign bank and 0, otherwise. Loan to deposit Total loans to total deposits ratio. Loan to deposit2 Squared loan to deposit. Commercial to home loan Commercial loans to household loans ratio. CD volatility Monthly volatility of CD rates for prior one year. MSBD A dummy variable equal to 1 for MSBs and 0 for commercial banks. 2007D A dummy variable equal to 1 if the observation occurs after 2007 and otherwise 0.
272 J.Q. Jeon, K.K. Lim / Pacific-Basin Finance Journal 25 (2013) 253–272
References
Allen, F., Gale, D., 2000. Comparing Financial Systems. MIT Press, Cambridge, Massachussetts. Baumol, W.J., 1982. Contestable markets: an uprising in the theory of industry structure. Am. Econ. Rev. 72, 1–15. Baumol, W.J., Panzar, J.C., Willig, R.D., 1982. Contestable Markets and the Theory of Industry Structure. Harcourt Brace Jovanovich,
New York. Beck, T., De Jonghe, O., Schepens, G., 2013. Bank competition and stability: cross-country heterogeneity. J. Financ. Intermed. 22,
218–244. Berger, A., Rosen, R., Udell, G., 2007. Does market size structure affect competition? The case of small business lending. J. Bank.
Financ. 31, 11–33. Berger, A.N., Klapper, L.F., Turk-Ariss, R., 2009. Bank competition and financial stability. J. Financ. Serv. Res. 35, 99–118. Boone, J., 2008a. A new way to measure competition. Econ. J. 118, 1245–1261. Boone, J., 2008b. Competition: theoretical parameterization and empirical measure. J. Inst. Theor. Econ. 164, 587–611. Boyd, J.H., De Nicolo, G., 2005. The theory of bank risk taking and competition revisited. J. Financ. 60 (3), 1329–1343. Boyd, J.H., De Nicolo, G., Jalal, A.M., 2006. Bank risk taking and competition revisited: new theory and evidence. IMF Working papers
06/297. Cyree, K.B., Spurlin, W.P., 2012. The effects of big-bank presence on the profit efficiency of small banks in rural markets. J. Bank.
Financ. 36, 2593–2603. Erel, I., 2011. The effect of bank mergers on loan prices: evidence from the United States. Rev. Financ. Stud. 24, 1068–1101. Gropp, R., Hakenes, H., Schnabel, I., 2011. Competition, risk-shifting, and public bail-out policies. Rev. Financ. Stud. 24, 2084–2120. Hakenes, H., Schnabel, I., 2011. Capital regulation, bank competition, and financial stability. Econ. Lett. 256–258. Hellmann, T.F., Murdock, K.C., Stiglitz, J.E., 2000. Liberalization, moral hazard in banking, and prudential regulation: are capital
requirements enough? Am. Econ. Rev. 90, 147–165. International Monetary Fund, 2001. International Capital Market. Keeley, M.C., 1990. Deposit insurance, risk, and market power in banking. Am. Econ. Rev. 80, 1183–1200. Korea Financial Supervisory Service, 2012. FSS Annual Report. Laeven, L., Levine, R., 2009. Bank governance, regulation and risk taking. J. Financ. Econ. 93, 259–275. Leuvensteijn, M.V., Bikker, J.A., Rixtel, A.V., Sorensen, C.K., 2011. A new approach to measuring competition in the loan markets of the
Euro area. Appl. Econ. 43, 3155–3167. Liu, H., Wilson, J., 2011. Bank type, competition and stability in Japanese banking. University of St Andrews Working papers. Long, S., Yemane, A., Stockley, K., 2010. Disentangling the effects of health reform in Massachusetts: how important are the special
provisions for young adults? Am. Econ. Rev. 100, 297–302. Martinez-Miera, D., Repullo, R., 2010. Does competition reduce the risk of bank failure? Rev. Financ. Stud. 23, 3638–3664. Panzar, J.C., Rosse, I.N., 1982. Structure, conduct and comparative statistics. Bell Laboratories Economic Discussion Paper No.248. Panzar, J.C., Rosse, J.N., 1987. Testing for “monopoly” equilibrium. J. Ind. Econ. 35, 443–456. Park, K., Pennacchi, G., 2009. Harming depositors and helping borrowers: the disparate impact of bank consolidation. Rev. Financ.
Stud. 21, 1–40. Schaeck, K., Cihak, M., 2010. Banking competition and capital ratios. Eur. Financ. Manag. 18, 836–866. Tabak, B.M., Fazio, D.M., Cajueiro, D.O., 2012a. The relationship between banking market competition and risk-taking: do size and
capitalization matter? J. Bank. Financ. 36, 3366–3381. Tabak, B.M., Fazio, D.M., Cajueiro, D.O., 2012b. Systemically important banks and financial stability: the case of Latin America. J. Bank.
Financ. 37, 3855–3866.
- Bank competition and financial stability: A comparison of commercial banks and mutual savings banks in Korea
- 1. Introduction
- 2. Literature review
- 2.1. Bank competition and stability
- 2.2. Mutual savings banks and commercial banks in Korea
- 3. Sample and methodology
- 3.1. Sample
- 3.2. Methodology
- 4. Empirical results
- 4.1. Analysis of market competition
- 4.2. Competition and stability in the MSB industry
- 4.3. Competition and stability in the commercial bank industry
- 4.4. Additional analysis
- 4.5. Difference-in-differences analysis
- 5. Conclusion
- Appendix A. Variable descriptions
- References
Bank-competition-and-financial-stability-in-Asia-Pacific_2014_Journal-of-Banking-Finance.pdf
Journal of Banking & Finance 38 (2014) 64–77
Contents lists available at ScienceDirect
Journal of Banking & Finance
journal homepage: www.elsevier .com/locate / jbf
Bank competition and financial stability in Asia Pacific
0378-4266/$ - see front matter � 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jbankfin.2013.09.012
⇑ Corresponding author. Tel.: +44 1248382170. E-mail addresses: [email protected] (Xiaoqing (Maggie) Fu), [email protected]
(Yongjia (Rebecca) Lin), [email protected] (P. Molyneux). 1 The authors are grateful for funding from the University of Macau. We highly
appreciate the comments from the BOFIT Research Seminar, Australasian Finance & Banking Conference, Auckland Finance Meeting, and FMA Asian Conference Doctoral Student Consortium, especially those from Franklin Allen, Iikka Korhonen, Minghua Liu, Ronald Masulis and Yukihiro Yasuda. All errors are our responsibility.
2 Beck (2008) and Carletti (2008, 2010) provide excellent surveys of the l 3 Soedarmon et al. (2011) and Liu et al. (2012) estimate the competition
nexus for banks in 12 Asian economies and four South East Asian c respectively. In addition, a small number of cross-country empirical studie several Asia Pacific economies into their large sample sets in testing this rela See, for example, Beck et al. (2006a), Boyd et al. (2006), Evrensel (2008), Ber (2009), Schaeck et al. (2009), Behr et al. (2010), Turk Ariss (2010), and Angi (2012).
Xiaoqing (Maggie) Fu a, Yongjia (Rebecca) Lin a, Philip Molyneux b,⇑,1
a Faculty of Business Administration, University of Macau, Taipa, Macau b Bangor Business School, Bangor University, UK
a r t i c l e i n f o
Article history: Received 21 May 2013 Accepted 20 September 2013 Available online 1 October 2013
JEL classification: G21 G28
Keywords: Bank competition Financial stability Regulation Banks in Asia Pacific
a b s t r a c t
Analysis of the tradeoff between competition and financial stability has been at the center of academic and policy debate for over two decades and especially since the 2007–2008 global financial crises. Here we use information on 14 Asia Pacific economies from 2003 to 2010 to investigate the influence of bank competition, concentration, regulation and national institutions on individual bank fragility as measured by the probability of bankruptcy and the bank’s Z-score. The results suggest that greater concentration fosters financial fragility and that lower pricing power also induces bank risk exposure after controlling for a variety of macroeconomic, bank-specific, regulatory and institutional factors. In terms of regulations and institutions, the results show that tougher entry restrictions may benefit bank stability, whereas stronger deposit insurance schemes are associated with greater bank fragility.
� 2013 Elsevier B.V. All rights reserved.
1. Introduction
The impact of bank competition on financial stability has been a focus of academic and policy debate over the last two decades and particularly since the 2007–2008 global financial crises (Beck, 2008; Carletti, 2008; Careletti, 2010; Acharya and Richardson, 2009; Beck et al., 2010; OECD, 2011). Under the traditional compe- tition-fragility view, banks cannot earn monopoly rents in compet- itive markets and this results in lower profits, capital ratios and charter values. This makes banks less able to withstand demand- or supply-side shocks and encourages excessive risk-taking (Marcus, 1984; Keeley, 1990). Alternatively, the competition- stability view suggests that competition leads to greater stability. A less competitive banking market may lead to more risk-taking if the big banks are deemed too important to fail and as such obtain implicit (or explicit) subsidies via government safety nets (Mish- kin, 1999). In addition, banks with more market power tend to charge higher loan rates, which may induce borrowers to assume greater risk leading to greater default. In competitive banking
markets loan rates are lower, Too-Big-To-Fail issues and safety net subsidies are smaller, and this results in a positive link between bank competition and stability (Boyd and De Nicoló, 2005). It could also be the case, as noted by Martinez-Miera and Repullo (2010) that bank competition and stability are linked in a non-linear manner, and in a similar vein Berger et al. (2009) argue that com- petition and concentration may coexist and can simultaneously induce stability or fragility.
As noted above, recent studies on the causes of the credit crunch have highlighted deregulation and excessive competition as factors that led to financial sector meltdowns in the US and the UK (Llewellyn, 2007; Brunnermeier, 2009; Milne, 2009; OECD, 2011). Moreover, it is of interest to assess whether the relationship between banking competition and financial stability has been af- fected after the outbreak of the recent financial crisis. While a sub- stantial literature has emerged addressing this critical issue,2 to our knowledge, the problem has been inadequately covered for banks operating across the Asia Pacific region.3 Against this backdrop our paper investigates the impact of bank competition on financial
iterature. -stability ountries, s include tionship. ger et al. ner et al.
Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77 65
stability for 14 Asia Pacific economies over the period from 2003 to 2010 and extends the previous empirical literature in several respects.4
First, previous studies have focused on using Z-scores or evi- dence of a real bank crisis as measures of banking sector risk/sta- bility. Here we extend the analysis by employing the probability of bankruptcy as an indicator of individual bank fragility.5 A real banking crisis can be an accurate indicator of banking sector stabil- ity, but its significance may be distorted for the following reasons: (1) banking crises are defined and announced differently across countries; (2) regulators may be less inclined to report bank insol- vencies because they may imply regulatory failure; and finally (3) regulators are reluctant to announce the failures of banks that play a key role within the system because they wish to avoid contagion effects (Uhde and Heimeshoff, 2009). The probability of bankruptcy, computed using the Black and Scholes (1973) and Merton (1974) contingent claims approaches provide a more appealing alternative. Compared to the use of accounting-based models (e.g., Z-score), this market-based measure of stability has the following advantages: (1) in efficient markets, stock prices reflect all available information; (2) market variables are unlikely to be influenced by firm’s accounting policies; and (3) market prices reflect future expected cash flows and thus should be more appropriate for use for prediction purposes.
Second, according to the structure-conduct-performance propo- sition, competition and concentration are inversely related; a more concentrated market will feature a lower degree of competition. However, criticisms of this view have led to a shift away from the presumption that structure is the most important determinant of the level of competition. Instead, proponents of what is now known as the New Industrial Organization (NIO) literature, such as Schmalensee (1982), argue that the strategies (conduct) of indi- vidual firms are equally, if not more, important than concentration, in explaining competitive conditions. Also, the related emergence of the theory of contestability (Baumol, 1982; Baumol et al., 1982) has spawned a variety of non-structural indicators of com- petition aimed at identifying firm conduct.6 In our study we include both structural and non-structural measures of competition to examine the concentration, competition and stability nexus in Asia Pacific banking.7
Thirdly, we incorporate both regulatory and institutional envi- ronmental factors in our models and also highlight the impact of the global turmoil on individual risk exposure in the region. Fol- lowing Berger et al. (2009), we adopt an instrumental variable technique with a Generalized Method of Moments (GMM) estima- tor to address potential endogeneity problems between bank com- petition and risk. We also include a series of sensitivity analyses using different model specifications.
4 See, for example, De Nicoló et al. (2003), Beck et al. (2006a), Boyd et al. (2006), Yeyati and Micco (2007), Berger et al. (2009), Schaeck and Cihak (2008), Schaeck et al. (2009), Uhde and Heimeshoff (2009), Behr et al. (2010), Turk Ariss (2010), Agoraki et al. (2011), Soedarmon et al. (2011), and Liu et al. (2012).
5 The Z-score is also used in this study to determine the robustness of our results. 6 These include measures of competition between oligopolists such as Iwata (1974)
and those that test for competitive behavior in contestable markets, Bresnahan (1982, 1989), Lau (1982) and Panzar and Rosse (1987). These indicators have been developed from (static) theory of the firm models under equilibrium conditions and mainly use some form of price mark-up over a competitive benchmark, such as price over marginal cost for the Lerner index and price over marginal revenue for the Bresnahan (1982) measure. The main exception is the Panzar and Rosse (1987) indicator that measures the relationship between changes in factor input prices and revenues earned by firms. See also Koetter et al. (2012) for recent studies using adjusted-Lerner indices to measure market power in banking.
7 The structural approach focuses on market structure measures such as market shares, concentration ratios for the largest sets of firms, and a Hirschman–Herfindahl index. Structural indicators measure actual market shares but do not allow inferences regarding the competitive behavior of banks. Non-structural measures are used to quantify bank pricing behavior. They include the Lerner index and the Panzar Rosse H-statistic (Berger et al., 2004).
Overall our results suggest that greater concentration fosters financial fragility, whereas lower pricing power also induces bank risk exposure after controlling for macroeconomic, bank-specific, regulatory and institutional factors. This finding supports the neutral view of the competition-stability relationship. It also implies that some banks in the region are able to attain greater discretion in price-setting to boost profits and reduce insolvency risk through channels other than increased concentration, such as product differentiation. Furthermore, there is evidence that larger banks are more likely to fail than their smaller counterparts. In addition, our results indicate that tougher entry restrictions may benefit bank stability, whereas stronger deposit insurance schemes appear to create greater bank fragility.
The remainder of the paper is organized as follows. Section 2 provides a review of the literature on competition and stability in banking. Section 3 introduces the econometric methodology. Section 4 describes the data used in the econometric tests. Sec- tion 5 presents the empirical results and Section 6 are the conclusions.
2. Literature review
Under the traditional competition-fragility hypothesis, com- petitive and/or less concentrated banking systems are more frag- ile. The ‘‘charter/franchise value’’ of banking, as modeled by Marcus (1984), and Keeley (1990), suggests that competition drives banks to undertake risk-taking strategies due to the con- traction of the latter’s franchise value. These models show that a higher charter or franchise value arising from increased market power may deter excessive risk-taking by the bank’s manage- ment. Because higher franchise value results in greater opportu- nity costs during bankruptcy, bank managers and shareholders may become more reluctant to engage in risky activities improv- ing bank asset quality.
Diamond (1984), Ramakrishnan and Thakor (1984), Boyd and Prescott (1986), Williamson (1986), and others show that more concentrated banking systems are composed of larger banks and that larger banks can capitalize on economies of scale and scope and better diversify their portfolios. Smith (1984) argues that banking relationships may endure for longer periods in less competitive environments if the information on the probability distribution of depositors’ liquidity needs is private. Hence, greater concentration and less competition could reduce liability risk and lead to greater stability in banking. Boot and Green- baum (1993) and Allen and Gale (2000, 2004) suggest that in a more competitive environment, banks earn less informational rent from their relationships with borrowers, which reduces their incentives to properly screen borrowers and increases the risk of fragility.
Competition can impact stability through contagion. Using a model of financial contagion in the interbank market Allen and Gale (2000) suggest that under perfect competition, all banks are price takers and none have an incentive to provide liquidity to troubled banks. As a result, troubled banks eventually fail with negative repercussions for the entire sector. Similarly, Saez and Shi (2004) argue that banks can cooperate, act strategically and help other banks to cope with temporary liquidity shortages in a market characterized by imperfect competition. Allen and Gale (2000) also find that a concentrated banking system with a small number of large institutions is more stable because banks are easier to monitor, less burdened by supervision, and therefore more resilient to shocks. Boot and Thakor (2000) suggest that larger banks tend to engage in ‘‘credit reputation/rating’’ because making fewer high-quality credit investments can increase the return of individual investments and thereby encourage financial
66 Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77
soundness. Additionally, larger banks are assumed to enjoy com- parative advantages related to the provision of credit monitoring services.
Allen and Gale (2004) claim that financial crises are more likely to occur in less concentrated banking systems due to the absence of powerful providers of financial products that could reap benefits from the high profits that thus serve as a buffer against asset qual- ity deterioration. Similarly, Boyd et al. (2004) state that the pres- ence larger (monopolistic) banks in concentrated banking systems might enhance profits and thus reduce financial fragility by providing higher ‘‘capital buffers’’ that protect these systems against external macroeconomic and liquidity shocks.
A different argument among proponents of the competition-fra- gility hypothesis is that deposit insurance schemes can reduce fra- gility by preventing bank runs but also introduce moral hazard by providing incentives to banks to engage in riskier activities. Thus, in more competitive environments, more generous deposit insur- ance may undermine bank stability (Diamond and Dybvig, 1983; Matutes and Vives, 1996). In addition, Hellmann et al. (2000) sug- gest that deposit interest rate ceilings are still necessary to prevent banks from taking excessive risk in competitive markets, although minimum capital requirements can boost the charter value.
Under the alternative competition-stability hypothesis, more competitive and/or less concentrated banking systems are more stable. The ‘‘too big to fail’’ doctrine (Mishkin, 1999, 2006; Barth et al., 2012b) indicates that policymakers are more concerned about bank failures when the number of banks in a concentrated banking system is low. Thus, these large banks are often more likely to receive public guarantees or subsidies, which may gener- ate a moral hazard problem, encourage risk-taking behavior and intensify financial fragility (Kane, 2010; Rosenblum, 2011). More- over, contagion risk may increase in a concentrated banking sys- tem with larger banks.
Caminal and Matutes (2002) claim that lower competition can result in reduced credit rationing and larger loans, ultimately increasing the probability of bank failure. Boyd and De Nicoló (2005) argue that concentrated banking systems allow banks to charge higher loan rates, which may encourage borrowers to as- sume greater risk. Consequently, the volume of non-performing loans may increase, resulting in a higher probability of bank failure. However, Martinez-Miera and Repullo (2010) suggests that higher loan rates also produce higher interest revenues for banks. This dy- namic might generate a U-shaped relationship between bank com- petition and stability.
Beck et al. (2006a,b) suggest that bank size is positively corre- lated with organizational complexity; for example, monitoring a large bank is more difficult than monitoring a small bank. Accord- ingly, as firm size increases, transparency may decrease as a result of expansion across multiple geographic markets and business lines and the use of sophisticated financial instruments that facil- itate the establishment of complex corporate organizations. These developments may reduce managerial efficiency and internal cor- porate control and may increase operational risk. Increasing orga- nizational complexity can render both market discipline and regulatory action less effective in preventing excessive risk expo- sure (Cetorelli et al., 2007).
However, as indicated in Berger et al. (2009), the two strands of the literature do not necessarily produce opposing predictions regarding the relationship between bank competition and financial stability. The aforementioned authors argue that bank risks may not increase even if market power encourages riskier asset portfo- lios because banks may protect their charter values by using other methods to offset the greater risk exposure. Such methods may in- clude increasing equity capital, reducing interest rate risk, and sell- ing credit derivatives. As noted earlier, market structure measures may not be good measures of competition and this (to some ex-
tent) has been confirmed by Berger et al. (2004) and Beck (2008) who show that banking industry concentration can influence sta- bility through channels other than competition.
A substantial empirical literature has emerged testing for con- centration, competition and banking stability relationships across countries. Yeyati and Micco (2007), for instance, use a sample of commercial banks from eight Latin American countries over the period 1993–2002 and find a positive link between bank risk (as measured by the Z-score) and competition (as captured by the Panzar and Rosse 1987, H-statistic), whereas the coefficient for bank concentration is not significant. This result lends support to the competition-fragility paradigm. Schaeck and Cihak (2008) analyze the relationship between bank competition and soundness using a sample of more than 3600 banks from ten European countries and more than 8900 US banks for the period from 1995 to 2005. They suggest that competition as measured by the Boone indicator increases bank soundness by increasing efficiency and that more concentrated banking markets benefit from financial stability. Using data from 31 systemic banking crises in 45 coun- tries for the period from 1980 to 2005, Schaeck et al. (2009) show that competition (as captured by the Panzar Rosse H-statistic) re- duces the likelihood of a crisis and increases the time to crisis, even after they control for banking system concentration, which is neg- atively related to financial fragility.
In a similar study, Berger et al. (2009) use a sample of 8235 banks from 23 industrial countries over 1999–2005 and find that banks with market power (measured using the Lerner index) have less overall risk exposure, as captured by their Z-scores. These find- ings support the traditional competition-fragility view. On the other hand, they show that bank-level market power also results in riskier loan portfolios, as indicated by non-performing loan ra- tios. Berger et al. (2009) argue that banks can protect their charter value from higher loan risk by holding more equity capital. More recently, Anginer et al. (2012) examine the relationship between competition according to the Lerner index and systemic stability as captured by default risk under Merton’s (1974) contingent claim pricing framework. Using a sample of 1872 publicly traded banks from 63 countries between 1997 and 2009, they find a positive relationship between competition and systemic stability (and the results remain the same even when they conduct a robustness check using bank asset concentration as an alternative proxy for bank competition).
Liu et al. (2012) introduce a variety of bank-specific risk indica- tors (the ratio of loan-loss provisions to total loans, loan-loss re- serves to total loans, after-tax ROA volatility, and the natural logarithm of the Z-index) to investigate similar relationships for banks operating in South East Asia (Indonesia, Malaysia, the Philip- pines and Vietnam) between 1998 and 2008. They find that com- petition measured using the Panzar Rosse H-statistic is inversely and significantly related to most risk indicators except the natural logarithm of the Z-index, which suggests that competition does not erode bank stability. The researchers also find that concentration is negatively associated with bank risk, whereas regulatory restric- tions positively influence bank fragility.
Overall, cross-country evidence yields mixed results regarding the relationship between bank concentration, competition, and stability. Meanwhile, the findings do confirm that concentration and competition can coexist and may influence financial stability through different channels.
3. Methodology
We test whether bank concentration and competition influence bank stability employing bank-level data from 14 Asia Pacific econ- omies. To address potential endogeneity issues associated with
Table 1 Variable definitions and sources.
Variable Definition Data sources
Dependent variables Probability of bankruptcy The bank-level probability of bankruptcy based on method of Bharath and
Shumway (2008) Bankscope, Datastream
Z-score The bank-level Z-score; a larger value means less overall bank risk and higher bank stability
Bankscope
Independent variables CR3 A country-level structural indicator of bank concentration, measured by the
concentration of assets held by the three largest banks in each country, with higher value indicating greater market concentration
World Bank database on financial development structure and Bankscope
LERNER A bank-level non-structural indicator of bank competition, measured by the Lerner index using fixed-effects method, with higher values indicating less competition in the banking sector
Bankscope
E-LERNER A bank-level non-structural indicator of bank competition, measured by the efficiency-adjusted Lerner index using a stochastic frontier analysis approach, with higher values indicating less competition in the banking sector
Bankscope
SIZE The natural logarithm of total assets in thousands of USD BankScope LLP The ratio of loan loss provisions to total assets BankScope NIM Bank’s net interest income as a share of its interest-bearing (total earning) assets BankScope Entry restrictions Ratio of entry applications denied to applications received from domestic and
foreign banks World Bank Survey of Bank Regulation and Supervision (for details see Barth et al., 2008, 2012a,b)
Capital requirements Minimum regulatory capital-to-assets ratio per country World Bank Survey of Bank Regulation and Supervision (for details see Barth et al., 2008, 2012a,b)
Deposit insurance A dummy variable that takes a value of one if the country has deposit insurance, and zero otherwise
World Bank Survey of Bank Regulation and Supervision (for details see Barth et al., 2008, 2012a,b)
RGDP Rate of real GDP growth rate World Economic Outlook Database, IMF CRISIS A dummy variable that takes a value of one for the years 2008–2009, and zero
otherwise Compiled by the authors
Instrumental variables Activity restrictions Index measure that indicates whether bank activities in the
securities, insurance and real estate markets, ownership and control of non-financial firms are unrestricted, permitted, restricted or prohibited
World Bank Survey of Bank Regulation and Supervision (for details see Barth et al., 2008, 2012a,b)
The aggregate indicator ranges from 1 to 4. A higher value indicates greater activity restrictions arising from legal requirements
Financial freedom The indicator of the openness of the banking system is a composite index of whether government interference exists in the financial sector, such as regulation, financial products, allocation of credit, whether foreign banks are free to operate. Higher values indicate fewer restrictions on banking freedoms
Heritage Foundation (2010)
Property rights The Heritage Foundation property rights protection index. A higher value signifies weaker protection
Heritage Foundation (2010)
Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77 67
measures of market power, we use an instrumental variable technique with a GMM estimator.8 Our panel data model has the following general form:
Bank Risk¼ f ðConcentration;Competition;Bank Controls; Regulatory and Institutional Controls;Macro ControlsÞ
ð1Þ
Notes on our dependent, explanatory and instrumental variables as well as data sources are presented in Table 1.
3.1. Market-based risk measure
Black and Scholes’s (1973) and Merton’s (1974) Distance to De- fault model is used to estimate the insolvency risk of listed banks. The model has been widely used in empirical research.9 However, there is only one paper employing this model in comparing the per- formance of market-based and accounting-based bankruptcy predic- tion models (Agarwal and Taffler, 2008). The Distance to Default
8 GMM is more efficient than 2SLS because it accounts for heteroskedasticity (Hall, 2005).
9 For example, see Hillegeist et al. (2004), Vassalou and Xing (2004), Gropp et al. (2004, 2006), Akhigbe et al. (2007), Chan-Lau and Sy (2007), Duffie et al. (2007), Bharath and Shumway (2008), and Campbell et al. (2008).
model views equity as a call option on the assets of a firm, with a strike price equal to the face value of the liabilities at time T when the liabilities mature. At time T, equity holders exercise their option and pay off the debt holders if the value of the firm’s assets is greater than the face value of its liabilities. Otherwise, if the value of the as- sets is insufficient to fully repay the firm’s debts, the call option be- comes worthless, and equity holders let it expire. In this scenario, the firm files for bankruptcy, and ownership is assumed to be trans- ferred to the debt holders at no cost, whereas the payoff for equity holders is zero. Estimates for the probability of bankruptcy are given by McDonald (2002). They are modified for dividends, and they re- flect the fact that the stream of dividends paid by the firm accrues to the equity holders:
P ¼ N � ln VA
D
� � þ u� d� r2
A 2
� �� � T
rA
ffiffiffi T p
0 B@
1 CA ð2Þ
where P is the probability of bankruptcy, N( ) is the cumulative nor- mal density function, VA is the value of assets, D is the face value of debts proxied by total liabilities, u is the expected return, d is the dividend rate estimated as total dividends/(total liabilities + market value of equity), rA is the standard deviation of assets (asset volatil- ity), and T is the time to expiration (taken to be 1-year).
68 Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77
VA, u and rA are non-observable. This study uses the following method outlined by Bharath and Shumway (2008):
VA ¼ VE þ D ð3Þ
rA ¼ VE
VA rE þ
D VA
rD ð4Þ
rD ¼ 0:05þ 0:25 � rE: ð5Þ
u ¼ ri;t�1: ð6Þ
where VE is the market value of common equity, VA is the total value of assets, D is the face value of debts proxied by total liabilities, rA is the standard deviation of assets (asset volatility), rE is the standard deviation of daily stock returns multiplied by the square root of the average number of trading days in the year (set at 252 trading days), u is the expected return, and ri,t�1 is the bank’s stock returns over the previous year.10
3.2. Accounting-based risk measure
For our accounting based risk measure we use the Z-score which is widely used in the literature as a stability indicator (see, for instance, Boyd and Runkle, 1993; Lepetit et al., 2008; Laeven and Levine, 2009; Čihák and Hesse, 2010). Using accounting infor- mation on asset returns, its volatility and leverage, the Z-score is calculated as follows:
Zit ¼ ROAit þ Eit=TAit
rROAit ð7Þ
where ROA is the return on assets, E/TA is the equity to total assets ratio, and rROA is the standard deviation of return on assets.
The Z-score is inversely related to the probability of a bank’s insolvency. A bank becomes insolvent when its asset value drops below its debt and the Z-score shows the number of standard devi- ations that a bank’s return has to fall below its expected value to deplete equity and make the bank insolvent.
3.3. Concentration and competition measures
First, based on the structural approach, the degree of market concentration is used. Market concentration is measured as the ra- tio of the assets of the three largest banks to the total assets of the banking system in the country in question (CR3). Second, a non- structural indicator, the Lerner index (LERNER), is used to measure the degree of competition. This indicator has been widely used in recent bank research.11 The Lerner index captures the capacity of price power by calculating the difference between price and mar- ginal cost as a percentage of price.12 The degree of competition is gi- ven by the range 0 < Lerner index < 1. In the case of perfect
10 Hillegeist et al. (2004) use this approach to assess the probability of bankruptcy and they note that ‘‘since expected returns cannot be negative, we set the expected growth rate equal to the risk-free rate in these cases (p. 10)’’. Bharath and Shumway (2008) also use the risk-free rate to replace the expected return on assets as a robustness check (p. 1348). In our sample, the risk-free rates range between 0.20 and 5, whereas the expected returns are negative during the crisis period. Therefore, we follow these two studies and replace the expected return with the risk-free rate when the former is negative.
11 For example, see Claessens and Laeven (2004), Maudos and Fernández de Guevara (2004), Fernández de Guevara et al. (2005), Berger et al. (2009), and Maudos and Solís (2009).
12 The H-statistic, developed by Panzar and Rosse (1987), is an alternative tool for inferring the degree of competition in the banking industry. It is computed from reduced form revenue equations, and it measures the sum of the elasticities of a bank’s revenue with respect to the bank’s input prices (Claessens and Laeven, 2004). A critical feature of the Panzar Rosse H-statistic is that the test must be undertaken in long-run equilibrium.
competition, the Lerner index = 0; under a pure monopoly, the Lern- er index = 1. A Lerner index < 0 implies pricing below the marginal cost and could result, for example, from non-optimal bank behavior. Algebraically, the Lerner index is calculated as follows:
Lernerit ¼ ðPTAit �MCTAit
Þ=PTAit ð8Þ
where PTAit is the price of total assets proxied by the ratio of total revenues (interest and non-interest income) to total assets for bank i at time t, and MCTAit is the marginal cost of total assets for bank i at time t.
Following Fernández de Guevara et al. (2005) and Carbó-Valverde et al. (2009), we can calculate the output price (PTAit ) as the ratio of total revenues (interest and non-interest income) to total assets. Given the limited information on prices for loans and deposits,13 we use a single indicator of banking activity, namely total assets as a measure of bank output, as suggested by Shaffer (1993) and Berg and Kim (1994). Assuming that the heterogeneous flow of goods and services supplied by a bank is proportional to its total assets, the output price includes both inter- est income and non-interest income. Following Hasan and Marton (2003), Soedarmon et al. (2011), Sun and Chang (2011) and Jiang et al. (2013) we use a two input cost function specification that tends to be used in emerging market bank efficiency studies (due to data availability issues) to estimate marginal costs. We also cross check with a three-input cost function specification and also follow Koetter et al. (2008, and 2012) and estimate the efficiency-adjusted Lerner index using a stochastic frontier analysis approach for another robustness test.14
3.4. Other control variables
Following Schaeck and Cihak (2008), Laeven and Levine (2009) and Uhde and Heimeshoff (2009), we also include a range of bank- specific variables. A bank’s asset size (SIZE) is defined as the loga- rithm of its total assets. The ratio of loan-loss provisions to total as- sets (LLP) is used to measure output quality and the way in which managers invest in high risk assets. The net interest margin (NIM) is employed to track the profitability of a bank’s investing and lending activities.
Beck et al. (2006a) argue that there are two reasons why cross- country differences in bank regulatory policies and national insti- tutions should be considered in assessing the relationship between bank competition and financial stability. First, this approach pro- vides a simple robustness test for the competition-stability rela- tionship. Second, it presents additional information about the links between bank regulations, national institutions, and financial stability. Hence, following previous studies (Beck et al. 2006a; Lae- ven and Levine 2009; Delis et al. 2011, and Goddard et al. 2011), we also control for bank regulations and institutional environ- ments in investigating the effects of concentration and competition on bank stability.
Deposit insurance is a dummy variable that takes a value of one if a country has explicit deposit insurance and a value of zero otherwise.15 Credible deposit insurance can enhance financial stabil- ity by decreasing the likelihood of depositor runs. Conversely, if the capital positions and risk-taking of insured institutions are not super- vised carefully, insurers tend to accrue loss exposures that under- mine bank stability over the long-run. Capital requirement indicates the minimum capital requirement (capital-to-assets ratio) per coun-
13 Loan revenue data do not separate earned income from fixed income investments, and the financial costs of deposits are included with those of other liability products.
14 Appendix A presents the translog cost function used to estimate bank marginal cost.
15 In our sample, five countries (Australia, China, Sri Lanka, Pakistan, and Thailand) do not have deposit insurance.
Table 2 Descriptive statistics.
Variable Listed banks Listed and non-listed banks
Obs. Mean Std. dev. Min Max Obs. Mean Std. dev. Min Max
Probability of bankruptcy 1500 0.18 0.18 0.54 0 Z-score 1500 40.86 32.69 �2.69 196.64 4069 39.78 47.14 �40.28 681.92 CR3 1500 0.44 0.11 0.26 0.99 4069 0.46 0.13 0.26 0.99 Conventional Lerner index (LERNER) 1500 0.31 0.14 �1.26 0.68 4069 0.32 0.18 �2.75 0.82 Efficiency-adjusted Lerner index (E-LERNER) 1500 0.26 0.14 �1.32 0.65 4069 0.27 0.19 �2.79 0.81 Bank size (SIZE) 1500 16.22 1.60 10.11 21.05 4069 15.59 2 1.78 21.4 Loan loss provision% (LLP) 1500 1.80 4.27 0 149 4069 1.69 3.82 0 149 Net interest margin% (NIM) 1500 2.79 1.56 �1.49 11.04 4069 2.94 2.69 �60.57 39.36 Real GDP growth% (RGDP) 1500 3.96 3.88 �6.29 14.47 4069 5.37 4.23 �6.29 14.47 Global financial crisis (CRISIS) 1500 0.28 0.45 0 1 4069 0.27 0.44 0 1 Entry restrictions 1500 0.13 0.25 0 0.92 4069 0.09 0.22 0 0.92 Capital requirements% 1500 8.42 0.69 8 10 4069 8.32 0.63 8 10 Deposit insurance 1500 0.86 0.35 0 1 4069 0.75 0.43 0 1 Activity restrictions 1500 10.84 1.97 4 16 4069 11.28 2.59 4 16 Financial freedom 1500 44.85 14.43 30 90 4069 44 16.59 30 90 Property rights 1500 57.56 18.84 20 90 4069 52.45 22.03 20 90
The probability of bankruptcy is a market-based bank-level measure of financial fragility that is calculated using the method developed by Bharath and Shumway (2008). The Z-score is an accounting-based bank-level indicator of financial stability. The conventional Lerner index (LERNER) is a bank-level indicator of bank competition that is calculated as the difference between price and marginal cost as a percentage of price using fixed effect regression. The efficiency-adjusted Lerner index (ELERNER) is a bank-level efficiency-adjusted indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using a stochastic frontier analysis approach. CR3 is a country-level structural indicator of bank concentration calculated as the fraction of assets held by the three largest banks in each country. SIZE is the natural logarithm of total assets in thousands of USD. LLP is the ratio of loan loss provisions to total assets. NIM is the ratio of net interest income to interest-bearing (total earning) assets. RGDP is the rate of real GDP growth. Entry restrictions is the ratio of entry applications denied to applications received from domestic and foreign banks. Global financial crisis is a dummy variable that takes a value of one for the years 2008–2009 and zero otherwise. Activity restrictions is an aggregate index measure that indicate whether bank activities in the securities, insurance and real estate markets and the ownership and control of non-financial firms are unrestricted, permitted, restricted or prohibited. The capital requirement is the minimum regulatory capital-to-assets ratio per country. Financial freedom is an indicator of the openness of the banking system; it functions as a composite index of government interference in the financial sector, including regulations on financial products, allocation of credit, whether foreign banks are free to operate and other factors. Deposit insurance is a dummy variable that takes a value of one if the country has deposit insurance and zero otherwise. Property rights are measured using the Heritage Foundation property rights protection index.
Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77 69
try, which is interpreted as another entry barrier indicator. In addi- tion, greater equity capital encourages prudent behavior. Hence, greater capital requirements are expected to indicate a more stable banking market. The variable entry restrictions is the ratio of the num- ber of banking licence applications denied to the number of applica- tions received from domestic and foreign entities. The effect of this control variable on bank stability is expected to be ambiguous be- cause restricted entry may reduce competitive pressure and thereby increase domestic bank profits, but it may also induce market ineffi- ciencies. The rate of real GDP growth (RGDP) is used as a proxy for the fluctuations in economic activity. CRISIS is a dummy variable that takes a value of one for the years 2008–2009 and zero otherwise.
To deal with the potential presence of endogeneity and heter- oskedasticity, following Berger et al. (2009), we employ a GMM pa- nel data estimator using activity restrictions, financial freedom, and property rights as instruments. Activity restrictions are a key deter- minant of the scope of a bank’s ability to provide fee-paying ser- vices. This measure reflects the level of regulatory restrictiveness for bank participation in securities market, insurance activities, real estate activities, and the ownership of non-financial firms. Financial freedom is an indicator of the openness of a financial sys- tem. This measure indicates the extent of government involvement in the financial sector, considering regulation, financial products, and the allocation of credit; the freedom of foreign banks to oper- ate; and the degree of regulation of financial market activities. Fi- nally, the protection of property rights is an important pre-requisite for a well-functioning financial system. A higher value of the Her- itage Foundation property rights protection index signifies weaker protection of property rights.
16 Please refer to Table 1 for details.
4. Data
The sample data focus on commercial banks in 14 Asia Pacific economies over 2003 and 2010. Financial information and stock market information, converted to US dollars, are obtained from the
Bankscope database by Bureau van Dijk and are supplemented by information from Datastream. Banking sector concentration ratios are obtained from the updated version of the World Bank database on financial development structures and supplemented by the Bankscope database; real GDP growth data are taken from the World Economic Outlook by the International Monetary Fund (IMF); and information on regulations and the institutional environ- ment come from several sources, including the World Bank database on ‘‘Bank Regulation and Supervision’’ (developed by Barth et al., 2001 and updated by Barth et al., 2006, 2008 and Barth et al., 2012a) and the 2010 index of Economic Freedom, which was pub- lished by The Wall Street Journal and The Heritage Foundation.16
After excluding banks with (1) missing, negative or zero values for the cost function needed to calculate the Lerner index, (2) miss- ing values for loan loss provisions, and (3) missing Z-score values, we obtain a final sample that includes unbalanced panel data for 14 Asia Pacific economies, with 4069 observations (see Appendix B). The subsample for listed banks includes 1500 observations (see Appendix C). All of the data are deflated by their correspond- ing year CPIs to the 2003 price level to control for inflation effects. Table 2 presents the descriptive statistics for all variables used in the study. All bank-level variables are averaged by bank for the period from 2003 to 2010, and the country-level variables are aver- aged by country for the same study period. Comparing listed banks with non-listed banks in the sample, Table 2 shows that on average listed sample banks enjoy a higher Z-score, lower loan loss provi- sion ratio, and larger in size, whereas non-listed banks have a high- er Lerner index and rely more on interest income. Moreover, markets with listed banks are less concentrated, subject to less activity restrictions, have a higher capital requirement ratio, and enjoy more financial freedom and better property rights protec- tion. Listed banks join the deposit insurance scheme in the markets with more entry restrictions.
Table 3 Concentration, Competition, and Stability measures.
Listed banks Listed and non-listed banks
Obs. CR3 LERNER E-LERNER Prob. of bankruptcy Obs. CR3 LERNER ELERNER Z-score
Panel A: mean by year 2003 153 0.4748 0.3191 0.2712 0.2004 423 0.4767 0.3154 0.2724 39.2417 2004 167 0.4305 0.3351 0.2885 0.0940 460 0.4596 0.3363 0.2943 42.0883 2005 185 0.4181 0.3273 0.2807 0.0834 519 0.4593 0.3173 0.2750 40.778 2006 188 0.4175 0.3116 0.2637 0.0799 549 0.4624 0.3076 0.2645 40.2366 2007 197 0.4297 0.3028 0.2529 0.1862 565 0.4644 0.3142 0.2695 39.448 2008 203 0.4472 0.2585 0.2056 0.2561 552 0.4661 0.2826 0.2351 37.3431 2009 211 0.4483 0.3151 0.2651 0.3369 534 0.4655 0.3206 0.2740 39.9518 2010 196 0.4550 0.3336 0.2828 0.1523 467 0.4647 0.3545 0.3072 39.4407
Panel B: mean by country Australia 48 0.6827 0.2954 0.2291 0.0820 111 0.6577 0.3206 0.2740 44.5649 China 35 0.5191 0.4343 0.3811 0.1295 700 0.5387 0.3914 0.3518 40.6025 Hong Kong 32 0.7064 0.4268 0.3823 0.0447 193 0.6971 0.3683 0.3281 41.2276 India 226 0.3409 0.3093 0.2598 0.1330 447 0.3389 0.3106 0.2663 42.8772 Indonesia 134 0.4562 0.2653 0.2270 0.1265 409 0.4583 0.2991 0.2661 47.6483 Japan 597 0.4089 0.3091 0.2538 0.2616 988 0.4077 0.3074 0.2521 39.9749 Korea 31 0.5057 0.3486 0.3009 0.1872 126 0.5033 0.3380 0.2866 28.4216 Malaysia 24 0.4563 0.4315 0.3842 0.0362 191 0.4571 0.3945 0.3547 47.0183 Pakistan 98 0.4376 0.2671 0.2316 0.0936 174 0.4404 0.2129 0.1766 17.5765 Philippines 78 0.5088 0.3175 0.2769 0.1145 184 0.4953 0.2448 0.2070 40.8976 Singapore 16 0.9156 0.4889 0.4410 0.0670 67 0.9145 0.3315 0.2856 62.3927 Sri Lanka 55 0.6165 0.2669 0.2365 0.1409 81 0.6171 0.2147 0.1857 34.9779 Taiwan 62 0.2719 0.2753 0.2236 0.1974 253 0.2712 0.3126 0.2646 28.714 Thailand 64 0.4550 0.3622 0.3147 0.1167 145 0.4539 0.2520 0.2070 34.2761
The probability of bankruptcy is a market-based bank-level measure of financial fragility that is calculated using the method developed by Bharath and Shumway (2008). The Z-score is an accounting-based bank-level indicator of financial stability. LERNER is a bank-level indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using fixed effect regression. ELERNER is a bank-level efficiency-adjusted indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using a stochastic frontier analysis approach. CR3 is a country-level structural indicator of bank concentration calculated as the fraction of assets held by the three largest banks in each country.
17 Yeyati and Micco (2007) use a sample of commercial banks from eight Latin American countries over the period 1993–2002 and find a positive link between bank risk (as measured by the Z-score) and competition (as captured by the Panzar and Rosse, 1987, H-statistic), whereas the coefficient of bank concentration is insignificant.
18 Using data from 31 systemic banking crises in 45 countries for the period from 1980 to 2005, Schaeck et al. (2009) show that competition reduces the likelihood of a crisis and increases the time to crisis and concentration is positively related to financial fragility. Anginer et al. (2012) use a sample of 1,872 publicly traded banks from 63 countries between 1997 and 2009 and find a positive relationship between competition and systemic stability and a negative relationship between concentration and systemic stability.
70 Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77
Table 3 presents a summary of our concentration, competition and bank stability measures from 2003 to 2010 for 14 Asia Pacific countries by year (panel A) and by country (panel B). The pattern derived from the sample of listed banks is quite similar to that from the whole sample. Thus, we focus on the sample of listed banks. Based on the market measure of bank stability, bank risk in- creased overall from 2007 to 2009. The results imply that bank per- formance was most affected over 2009, a finding also confirmed by the IMF (2009). Bank risk decreased dramatically in 2010, which implies that this region was initially hit hard by the global crisis but has rapidly rebounded. Comparing bank risk by country using the market-based measure indicates that on average, banks operat- ing in Malaysia, Hong Kong, and Singapore are exposed to lower risk than those in other Asia Pacific economies. Meanwhile, Japa- nese, Taiwanese, and Korean banks are the most fragile.
When the findings regarding market concentration and compe- tition are compared by year, the structural and non-structural measures reveal different trends. The trend for the Lerner index (non-structural measure) is descending between 2005 and 2008 suggesting a decrease in pricing power, whereas industry concen- tration (structural measure) increases over the same period. The Lerner index exhibits varying degrees of market power across countries. Singapore has the highest efficiency-adjusted Lerner in- dex value (0.44), whereas Taiwan has the lowest value (0.22). Con- centration also varies across countries. The results suggest that concentration of assets held by the three largest banks in Singapore is 91.6%, indicating that the system is dominated by these banks. However, concentration in Taiwan is relatively low at 27.2%.
5. Empirical results
Table 4 presents the main results that indicate the impact of bank concentration and competition on financial stability. Two different risk exposure indicators are used as the dependent variables that proxy for financial stability: the probability of bankruptcy for listed banks (specifications 1–4) and the Z-score for both listed and non-
listed banks (specifications 5–8). We use the First Stage F-test and the Hansen’s J test to test for the relevance and validity of the instru- ments of the degree of market power, respectively. The Second Stage F-test is also used to test for goodness of fit for all regression models. The results support the use of the GMM panel data estimator.
Based on market measures, Table 4 indicates the significantly negative correlation for the Lerner index used in regression (1), suggesting that increases in the degree of bank pricing power are positively related to individual bank stability in Asia Pacific. Mean- while, the coefficient of bank concentration is significantly posi- tive, indicating that banks in more concentrated markets face greater risk. The robustness of the results is verified using regres- sion specifications (2)–(4). The findings lend support to the neutral view of the competition-stability nexus as both the competition- stability and competition-fragility views can be simultaneously va- lid. In this case, excessive concentration and lower pricing power simultaneously lead to bank fragility.
Our findings vary from those of most previous studies which fo- cus on banks operating in a specific geographic region such as Latin America,17 or a broader area.18 However, the results findings (we be- lieve) are not surprising for banks operating in Asia Pacific. On the one hand, most countries in this region (developing countries in par- ticular) have adopted ‘‘finance for growth’’ policies for a long period. The protected, larger banks in these concentrated banking systems
Table 4 Concentration, competition, and financial stability.
Dependent variable: prob. of bankruptcy Dependent variable: Z-score
(1) (2) (3) (4) (5) (6) (7) (8)
LERNER �1.3250** �1.0569** �1.4570*** �1.5704*** 53.4755*** 50.0600*** 50.5653*** 57.0264***
(0.5237) (0.5258) (0.5175) (0.5559) (18.1775) (18.4541) (17.0571) (20.4786) CR3 2.2529*** 2.5413*** 2.2277*** 2.1215*** �46.3050*** �49.0417*** �46.6299*** �45.7716***
(0.5459) (0.5498) (0.5612) (0.5727) (9.0731) (9.5586) (8.8796) (9.4445) SIZE 0.0700*** 0.0623** 0.0641** 0.0671** �3.0907*** �3.1619*** �3.0507*** �3.0621***
(0.0271) (0.0269) (0.0286) (0.0282) (0.9571) (0.9242) (0.9380) (0.9675) LLP �0.5129 �0.2709 �0.9785 �0.7478 �23.1895 �28.1637 �27.5442 �20.3329
(0.8826) (0.8516) (0.9875) (0.9410) (33.3662) (33.6761) (34.3624) (33.4386) NIM 0.0260 0.0195 0.0277 0.0296 0.2372 0.2588 0.2575* 0.2111
(0.0255) (0.0239) (0.0266) (0.0275) (0.1593) (0.1590) (0.1539) (0.1730) RGDP �0.0040 �0.0015 �0.0038 �0.0052 �0.0475 �0.0707 �0.0409 �0.0456
(0.0031) (0.0031) (0.0032) (0.0033) (0.0826) (0.0813) (0.0807) (0.0841) CRISIS 0.0642** 0.0822*** 0.0622** 0.0480* 1.1194 0.9282 1.0695 1.2774
(0.0250) (0.0252) (0.0257) (0.0272) (0.6965) (0.7283) (0.6760) (0.8057) Entry restrictions �0.1791*** 3.8920
(0.0623) (3.1427) Capital Requirement 0.0587 1.3472
(0.0435) (1.0743) Deposit Insurance 0.0618** �0.9649
(0.0309) (1.5376) First Stage F-test (LERNER) 8.25*** 8.01*** 8.72*** 7.78*** 10.36*** 9.91*** 11.23*** 9.68***
First Stage F-test (CR3) 117.24*** 134.54*** 120.29*** 108.27*** 234.57*** 226.27*** 236.17*** 233.67***
Hansen’s J v2 0.859 0.126 1.064 2.619 0.871 0.503 0.849 1.169 (P-value) (0.3541) (0.7226) (0.3023) (0.1056) (0.3508) (0.4783) (0.3569) (0.2797) Second Stage F-test 66.85*** 69.60*** 54.38*** 51.07*** 16.51*** 15.49*** 15.21*** 14.14***
No. of observations 1320 1320 1320 1320 3299 3299 3299 3299
Results from GMM panel data estimations to explain the impacts of bank concentration and competition on financial stability. The first dependent variable (specifications 1– 4) is the probability of bankruptcy, which is a market-based bank-level measure of financial fragility that is calculated using the method developed by Bharath and Shumway (2008). The second dependent variable (specifications 4–8) is Z-score, which is an accounting-based bank-level indicator of financial soundness. LERNER is a bank-level indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using the stochastic frontier analysis approach. CR3 is a country-level structural indicator of bank concentration calculated as the fraction of assets held by the three largest banks in each country. SIZE is the natural logarithm of total assets in thousands of USD. NIM is the ratio of net interest income to interest-bearing (total earning) assets. LLP is the ratio of loan loss provisions to total assets. RGDP is the rate of real GDP growth. Crisis is a dummy variable that takes a value of one for the years 2008–2009 and zero otherwise. Deposit insurance is a dummy variable that takes a value of one if the country has deposit insurance and zero otherwise. Capital requirement is the minimum regulatory capital-to-assets ratio for each country. Entry restrictions is the ratio of entry applications denied to applications received from domestic and foreign banks. The instrumental variables include activity restrictions, financial freedom, and property rights. *** Indicate significance at the 1% levels, respectively. Robust standard errors are in parentheses. ** Indicate significance at the 5% levels, respectively. Robust standard errors are in parentheses. * Indicate significance at the 10% levels, respectively. Robust standard errors are in parentheses.
Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77 71
channel resources to ‘‘priority sectors’’. Their borrowers become ‘‘too large to fail’’, and hence, banks lose their incentive to develop an appropriate credit culture and may find themselves faced with rela- tively high levels of non-performing loans. In addition, banks are the most important source of public savings in the majority of Asia Paci- fic economies, which also makes them ‘‘too-big or too-systemically- important-to-fail’’ possibly leading to moral hazard problems (Sheng, 2009).
On the other hand, according to Elzinga and Mills (2011), the Lerner index is a ‘‘better indicator of a firm’s price-setting discre- tion than its ability to sustain monopoly prices’’ (p. 1). Thus, the re- sults may imply that banks in this region are able to obtain greater discretion in terms of price-setting to boost their profits and re- duce their insolvency risk through channels other than increased concentration (product differentiation).19 In other words, greater
19 For example, as indicated in a survey report provided by the IDC Financial Insights Asia/Pacific division, banks across the Asia Pacific region are considering the uniquely Asian opportunities for sustainable growth that have been generated by governments identifying new priority industries such as aerospace and defense in Singapore, green technology in China, and high technology in Taiwan and China. Banks in this region have identified two strategic technology initiatives that they can use to expand their reach and profitability – risk management and channel efficiency. The focus on risk management has mainly been generated by the growing availability and sophistica- tion of analytics technologies, whereas the emphasis on channel efficiency stems from the vast expansion of mobility across the region. As a result, there are a growing number of innovative strategic IT projects that drive business differentiation in Asia Pacific banks (IDC, 2012).
pricing power enhances the ability of banks to generate higher ‘‘cap- ital buffers’’ to protect them against external macroeconomic and liquidity shocks.
Among other control variables, the significantly positive coefficient for bank size suggests that larger banks face greater risk. Laeven and Levine (2009) also find the same result. The crisis dummy is positively and significantly related to bank risk, which implies that banks are more fragile during financial turmoil. In considering regulatory and institutional environments, we find that entry restrictions are significantly and negatively associated with the probability of bankruptcy, which suggests that a lower level of competitive pressure induces greater fragility for listed banks. This result is consistent with the empirical findings of Uhde and Heimeshoff (2009), who find that restricted market entry is likely to enhance bank stability in Western European banking. Deposit insurance is significantly associated with a higher proba- bility of bankruptcy, supporting the moral hazard argument regarding excessive risk-taking when a financial safety net is available. Again, the result is similar to the findings of Laeven and Levine (2009).
Table 4 also examines the impact of bank concentration and competition on the soundness of both listed and non-listed banks using the Z-score as a proxy for financial stability. The results of regressions (5)–(8) show that the Lerner index is positively and significantly related to the Z-score, whereas the coefficient of concentration is significantly negative. The finding confirms that lower pricing power and excessive concentration may simulta-
72 Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77
neously lead to bank fragility. Meanwhile, the coefficient on bank size is significantly negative, which is also consistent with our pre- vious finding that larger banks face greater risk.
We undertake a variety of robustness tests on our main models. First, following Koetter et al. (2008, 2012) and Turk Ariss (2010), we use the efficiency-adjusted Lerner index to replace the conven- tional Lerner index as a measure of banking market competition. Our main results are similar (see Appendix D). Second, following Berger et al. (2009), we also use a quadratic term for the Lerner index (namely, LERNER2) to capture a possible non-linear relation- ship between competition and stability. The coefficient of the qua- dratic term is significantly negative for the probability of bankruptcy model and positive for the Z-score model. Based on the inflection points calculated, the results remain unchanged and are reported in Appendix E. Third, we use Tobit regression models to estimate the competition-stability nexus for listed banks, because the probability of bankruptcy is between zero and one. The main results are maintained (see Appendix F). Fourth, we employ the Lerner index estimated using a three-input cost function specification replacing the one estimated using the two- input specification. Overall, the key findings remain unchanged (see Appendix G).
6. Conclusions
This study investigates the competition-stability nexus using cross-country data from 14 Asia Pacific countries for the period from 2003 to 2010. Both market-based and accounting-based risk measures are employed to measure individual bank fragility for the first time. Meanwhile, both concentration and competition indicators are included in the models to determine their impacts on bank stability. The initial results show a substantial shift in the average risk exposure of banks over the entire sample period, accompanied by gradual increases in concentration and competi- tion. The main results not only highlight the significant negative association between the Lerner index and individual bank risk but also illustrate the significant positive relationship between the concentration ratio and bank fragility. In other words, the find- ings provide support for the neutral view of the competition- stability nexus, indicating that the competition-stability and com- petition-fragility theories can simultaneously apply to Asia Pacific banking markets. The results also confirm that bank concentration is an insufficient measure of bank competitiveness. Overall our findings hold when we control for an array of bank-specific, macroeconomic, regulatory and institutional factors.
In addition, our analyses indicate that smaller bank size may improve financial soundness. In terms of regulations and
Table B1 Number of both listed and non-listed banks in sample. Source: BankScope (Bureau Van Di
2003 2004 2005 2006
Australia 8 9 13 17 China 46 54 71 98 Hong Kong 11 25 30 29 India 58 57 56 58 Indonesia 48 51 56 55 Japan 130 129 126 124 Korea 17 17 19 18 Malaysia 22 23 25 24 Pakistan 16 17 22 24 Philippines 14 22 28 27 Singapore 4 6 9 10 Sri Lanka 9 10 10 11 Taiwan 24 24 36 35 Thailand 16 16 18 19
Total 423 460 519 549
institutions, the results show that tougher entry restrictions may enhance bank stability, whereas stronger deposit insurance schemes negatively influence financial soundness. Unsurprisingly, banks are found to be more fragile during the recent financial crisis.
The findings highlight several important issues for policymak- ers in Asia Pacific economies. First, to prevent excessive concentra- tion, regulators should adopt a more cautious approach to evaluating and approving merger and acquisitions at the national level. Policymakers should also seek to reduce policy lending by encouraging banks to develop stronger independent credit cul- tures. Second, to improve the efficiency of resource allocation within an economy, regulators should encourage financial innova- tion among banks based on the premise of effective risk manage- ment, which also enables banks to become more stable via product innovation. Third, a certain level of entry restriction is needed for both domestic and foreign entrants to maintain finan- cial soundness. This suggests there should be greater scrutiny of foreign banks that seek to make acquisitions in Asia Pacific coun- tries. Finally, deposit insurance schemes appear to foster moral hazard and risk shifting behavior so any policy moves to increase coverage should be treated with caution as this could have the unintended consequence of boosting risk as opposed to promoting stability.
Appendix A. Translog cost function for estimating bank marginal cost
To derive MCTAit , the following translog cost function is esti-
mated while capturing bank specificities using bank fixed effects:
ln TCit ¼ a0 þ X2
j¼1
a1 ln wj it þ
1 2
X2
j¼1
X2
k¼1
ajk ln wk it þ b1 ln TAit
þ 1 2
b2ðlnTAitÞ2 þ X2
j¼1
b2j ln TAitlnwj it þ c1tT
1 2 c2tT
2
þ X2
j¼1
c3tT ln wj it þ c4tT ln TAit þ ei ð9Þ
MCTAit ¼ @TCit
@TAit ¼ b1 þ b2 ln TAit þ
X2
j¼1
b2j ln wj it þ c4tT
! TCit
TAit ð10Þ
where TCi is the bank’s total costs, TAi is the total assets, wi is the price of the factors of production, defined as follows: w1 is the price of purchased funds: interest expenses/total deposits and short-term funding, w2 is the price of labor and physical capital: non-interest
jk).
2007 2008 2009 2010 Total
19 18 14 13 111 117 110 107 97 700
28 26 24 20 193 58 56 54 50 447 56 55 52 36 409
127 127 122 103 988 16 15 15 9 126 24 24 25 24 191 25 26 23 21 174 23 24 24 22 184 10 9 10 9 67
9 9 11 12 81 33 33 34 34 253 20 20 19 17 145
565 552 534 467 4069
Table C1 Number of listed banks in sample. Source: BankScope (Bureau Van Dijk).
2003 2004 2005 2006 2007 2008 2009 2010 Total
Australia 6 6 6 6 6 6 6 6 48 China 2 2 3 3 4 5 9 7 35 Hong Kong 4 4 4 4 4 4 4 4 32 India 14 22 26 27 31 35 36 35 226 Indonesia 12 14 17 18 18 20 20 15 134 Japan 74 74 75 75 76 75 77 71 597 Korea 4 4 4 4 4 4 4 3 31 Malaysia 3 3 3 3 3 3 3 3 24 Pakistan 7 8 11 12 15 15 15 15 98 Philippines 8 10 10 10 10 10 10 10 78 Singapore 2 2 2 2 2 2 2 2 16 Sri Lanka 5 6 7 7 7 7 8 8 55 Taiwan 4 4 9 9 9 9 9 9 62 Thailand 8 8 8 8 8 8 8 8 64
Total 153 167 185 188 197 203 211 196 1500
Table D1 Concentration, competition, and financial stability (using efficiency-adjusted Lerner index).
Dependent variable: prob. of bankruptcy Dependent variable: Z-score
(1) (2) (3) (4) (5) (6) (7) (8)
E-LERNER �1.2413** �0.9908** �1.3632*** �1.4688*** 50.3293*** 47.1489*** 47.6084*** 53.4184***
(0.4886) (0.4899) (0.4824) (0.5183) (17.0441) (17.2802) (16.0116) (19.1364) CR3 2.2323*** 2.5235*** 2.2064*** 2.1004*** �45.6700*** �48.4773*** �46.0288*** �45.1490***
(0.5494) (0.5531) (0.5641) (0.5761) (9.1422) (9.6322) (8.9477) (9.5238) SIZE 0.0610** 0.0551** 0.0543** 0.0564** �2.6723*** �2.7750*** �2.6554*** �2.6184***
(0.0252) (0.0249) (0.0266) (0.0263) (0.9133) (0.8851) (0.9002) (0.9121) LLP �0.4634 �0.2298 �0.9220 �0.6894 �24.1559 �29.0915 �28.4276 �21.5510
(0.8652) (0.8345) (0.9635) (0.9221) (33.1960) (33.4751) (34.1855) (33.2407) NIM 0.0246 0.0184 0.0262 0.0280 0.2416 0.2626* 0.2615* 0.2178
(0.0249) (0.0233) (0.0260) (0.0268) (0.1571) (0.1570) (0.1521) (0.1702) RGDP �0.0041 �0.0016 �0.0039 �0.0053 �0.0466 �0.0703 �0.0401 �0.0448
(0.0031) (0.0031) (0.0032) (0.0033) (0.0824) (0.0812) (0.0806) (0.0838) CRISIS 0.0638** 0.0820*** 0.0618** 0.0476* 1.1244 0.9327 1.0746 1.2706
(0.0250) (0.0252) (0.0257) (0.0272) (0.6966) (0.7267) (0.6764) (0.8029) Entry restrictions �0.1804*** 3.9543
(0.0618) (3.0849) Capital requirement 0.0582 1.3380
(0.0428) (1.0507) Deposit insurance 0.0619** �0.8995
(0.0308) (1.5083) First Stage F-test (ELERNER) 8.52*** 8.29*** 9.02*** 8.05*** 10.78*** 10.34*** 11.66*** 10.15***
First Stage F-test (CR3) 117.24*** 134.54*** 120.29*** 108.27*** 234.57*** 226.27*** 236.17*** 233.67***
Hansen’s J v2 0.856 0.120 1.061 2.631 0.864 0.492 0.843 1.140 (P-value) (0.3547) (0.7291) (0.3029) (0.1048) (0.3526) (0.4828) (0.3586) (0.2857) Second Stage F-test 66.52*** 69.41*** 54.15*** 50.75*** 16.59*** 15.52*** 15.26*** 14.26***
No. of observations 1320 1320 1320 1320 3299 3299 3299 3299
Results from GMM panel data estimations to explain the impacts of bank concentration and competition on financial stability. The first dependent variable (specifications 1– 4) is the probability of bankruptcy, which is a market-based bank-level measure of financial fragility that is calculated using the method developed by Bharath and Shumway (2008). The second dependent variable (specifications 4–8) is Z-score, which is an accounting-based bank-level indicator of financial soundness. ELERNER is a bank-level efficiency-adjusted indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using the stochastic frontier analysis approach. CR3 is a country-level structural indicator of bank concentration calculated as the fraction of assets held by the three largest banks in each country. SIZE is the natural logarithm of total assets in thousands of USD. NIM is the ratio of net interest income to interest-bearing (total earning) assets. LLP is the ratio of loan loss provisions to total assets. RGDP is the rate of real GDP growth. Crisis is a dummy variable that takes a value of one for the years 2008–2009 and zero otherwise. Deposit insurance is a dummy variable that takes a value of one if the country has deposit insurance and zero otherwise. Capital requirement is the minimum regulatory capital-to-assets ratio for each country. Entry restrictions is the ratio of entry applications denied to applications received from domestic and foreign banks. The instrumental variables include activity restrictions, financial freedom, and property rights. *** Indicate significance at the 1% levels, respectively. Robust standard errors are in parentheses. ** Indicate significance at the 5% levels, respectively. Robust standard errors are in parentheses. * Indicate significance at the 10% levels, respectively. Robust standard errors are in parentheses.
Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77 73
expenses/fixed assets,20 T is the time trend that captures the influ- ence of technological changes that lead to shifts in the cost function over time, and e is the error term.
Following Hasan and Marton (2003), Soedarmon et al. (2011), Sun and Chang (2011) and Jiang et al. (2013) we use a two input cost function specification (we also re-estimate the translog cost
20 Because of the lack of labor data, non-interest expenses are used as a proxy for labor and physical capital costs.
function with three inputs – purchased funds, labor and physical capital – as a further robustness test to investigate market power and risk issues, the sample size falls but results are in-line with the two input specification, (see Appendix G). As usual,21 symmetry restrictions apply to this function (i.e. ajk = akj). Meanwhile, the total cost and input price terms are normalized by w2. This imposes linear
21 See Claessens and Laeven (2004), Maudos and Fernández de Guevara (2004), Fernández de Guevara et al. (2005), Berger et al. (2009), and Maudos and Solís (2009).
Table E1 Test for non-linear relationship.
Dependent variable: probability of bankruptcy Dependent variable: Z-score
(1) (2) (3) (4)
LERNER �1.8018** 51.6104 (0.8726) (37.2177)
LERNER2 �4.3906* 143.7211***
(2.3836) (46.0858) ELERNER �2.2149*** 62.2878*
(0.7124) (33.1187) ELERNER2 �4.0066* 140.1059***
(2.2827) (47.1578) SIZE 0.1088*** 0.0771* �2.0614 �1.3844
(0.0402) (0.0419) (1.2560) (1.1248) LLP 1.1718 0.8852 �60.9575 �60.9301
(2.2720) (2.2487) (54.9509) (55.7692) NIM 0.0202 0.0236 �0.0187 �0.0437
(0.0600) (0.0600) (0.2700) (0.2759) RGDP �0.0031 �0.0031 �0.1090 �0.1212
(0.0061) (0.0064) (0.1386) (0.1430) CRISIS 0.0167 0.0145 2.3794** 2.4364**
(0.0449) (0.0465) (1.1293) (1.1467) Inflection point �0.205 �0.275 �0.180 �0.222 First Stage F-test (LERNER/ELERNER/CR3) 8.25*** 8.52*** 10.78*** 10.36***
First Stage F-test (LERNER2/ELERNER2/CR32) 2.35* 1.70 7.53*** 9.99***
Hansen’s J v2 0.044 0.151 0.058 0.137 (P-value) (0.8340) (0.6980) (0.8097) (0.7108) Second Stage F-test 22.25*** 20.92*** 6.76*** 6.37***
No. of observations 1320 1320 3299 3299
Results from GMM panel data estimations to explain the impacts of bank concentration and competition on financial stability. The first dependent variable (specifications 1– 2) is the probability of bankruptcy, which is a market-based bank-level measure of financial fragility that is calculated using the method developed by Bharath and Shumway (2008). The second dependent variable (specifications 3–4) is Z-score, which is an accounting-based bank-level indicator of financial soundness. LERNER is a bank-level indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using the stochastic frontier analysis approach. ELERNER is a bank-level efficiency-adjusted indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using the stochastic frontier analysis approach. CR3 is a country-level structural indicator of bank concentration calculated as the fraction of assets held by the three largest banks in each country. SIZE is the natural logarithm of total assets in thousands of USD. NIM is the ratio of net interest income to interest-bearing (total earning) assets. LLP is the ratio of loan loss provisions to total assets. RGDP is the rate of real GDP growth. Crisis is a dummy variable that takes a value of one for the years 2008–2009 and zero otherwise. *** Indicate significance at the 1% levels, respectively. Robust standard errors are in parentheses. ** Indicate significance at the 5% levels, respectively. Robust standard errors are in parentheses. * Indicate significance at the 10% levels, respectively. Robust standard errors are in parentheses.
Table F1 Concentration, competition, and probability of bankruptcy (Tobit regression).
(1) (2)
LERNER �0.1880***
(0.0331) ELERNER �0.1799***
(0.0310) CR3 0.4596*** 0.4588***
(0.0748) (0.0748) SIZE 0.0025 0.0015
(0.0037) (0.0036) LLP 0.1758* 0.1773**
(0.0898) (0.0896) NIM �0.0090** �0.0090**
(0.0041) (0.0041) RGDP �0.0028 �0.0029
(0.0021) (0.0021) CRISIS 0.1454*** 0.1451***
(0.0122) (0.0122) Country effect yes yes Wald test 1051.70*** 1055.12***
No. of observations 1500 1500
This table presents the results of Tobit regressions. The dependent variable is the probability of bankruptcy, which is a market-based bank-level measure of financial fragility calculated using the method developed by Bharath and Shumway (2008). LERNER is a bank-level indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using fixed effect regression. ELERNER is a bank-level efficiency-adjusted indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using a stochastic frontier analysis approach. CR3 is a country-level structural indicator of bank concentration calculated as the fraction of assets held by the three largest banks in each country. SIZE is the natural logarithm of total assets in thousands of USD. LLP is the ratio of loan loss provisions to total assets. NIM is the ratio of net interest income to interest-bearing (total earning) assets. RGDP is the rate of real GDP growth. Crisis is a dummy variable that takes a value of one for the years 2008–2009 and zero otherwise. *** Indicate significance at the 1% levels, respectively. Robust standard errors are in parentheses. ** Indicate significance at the 5% levels, respectively. Robust standard errors are in parentheses. * Indicate significance at the 10% levels, respectively. Robust standard errors are in parentheses.
74 Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77
Table G1 Concentration, competition, and financial stability – LERNER estimated using a 3-input specification (robustness check).
Dependent variable: Prob. of bankruptcy Dependent variable: Z-score
(1) (2) (3) (4)
LERNER �1.5303** 51.8100*
(0.6683) (29.8397) ELERNER �1.5496** 55.6842*
(0.7050) (33.7837) CR3 �0.2389 �0.2819 �99.7069** �105.8960**
(0.9480) (0.9521) (40.3922) (45.1951) SIZE 0.0059 0.0054 �3.7235* �3.9748*
(0.0198) (0.0199) (1.9838) (2.2412) LLP �1.6558 �1.6724 105.5617* 114.6456
(1.5140) (1.5641) (62.6924) (71.6090) NIM 0.0415 0.0406 0.0438 0.0029
(0.0435) (0.0448) (0.3333) (0.3784) CRISIS 0.1262*** 0.1260*** 2.0772 2.2418
(0.0195) (0.0198) (1.3357) (1.5081)
First Stage F-test 9.98*** 9.48*** 9.07*** 8.77***
First Stage F-test (CR3) 12.68*** 12.68*** 40.15*** 40.15***
Hansen’s J v2 1.897 2.266 1.193 0.699 Second Stage F-test 22.41*** 22.43*** 3.41*** 2.92***
No. of observations 786 786 2120 2120
Using the three-input specification to derive LERNER measures the number of observations reduces from 3299 to 2120 for listed and non-listed banks, and from 1320 to 786 for listed banks. We have to drop RGDP in our GMM model to avoid multicollineary problems, because the correlation coefficient between RGDP and CRISIS is �0.4548 for listed banks and �0.4222 for listed and non-listed banks. Results from GMM panel data estimations explain the impact of bank concentration and competition on financial stability. The first dependent variable (specifications 1–2) is the probability of bankruptcy, which is a market-based bank-level measure of financial fragility calculated using the method developed by Bharath and Shumway (2008). The second dependent variable (specifications 3–4) is Z-score, which is an accounting-based bank-level indicator of financial soundness. LERNER is a bank-level indicator of bank competition calculated as the difference between price and marginal cost as a percentage of price using the stochastic frontier analysis approach. In this case the LERNER is calculated using a three input specification (purchased funds, labor and physical capital). CR3 is a country- level structural indicator of bank concentration calculated as the fraction of assets held by the three largest banks in each country. SIZE is the natural logarithm of total assets in thousands of USD. NIM is the ratio of net interest income to interest-bearing (total earning) assets. LLP is the ratio of loan loss provisions to total assets. Crisis is a dummy variable that takes a value of one for the years 2008–2009 and zero otherwise. The instrumental variables include activity restrictions, financial freedom, and property rights. Overall, the key findings remain unchanged. *** Indicate significance at the 1% levels, respectively. Robust standard errors are in parentheses. ** Indicate significance at the 5% levels, respectively. Robust standard errors are in parentheses. * Indicate significance at the 10% levels, respectively. Robust standard errors are in parentheses.
Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77 75
homogeneity to ensure that the cost minimizing bundle does not change if all of the input prices are multiplied by the same positive scalar. Thus, only changes in the ratios of the input prices affect the allocation of inputs. Following Lozano-Vivas and Pasiouras (2010), we also include ln(equity) in the efficiency model to control for the effect of risk. We then use the system GMM model to test the link between market power and financial stability. The key results are reported in Appendix G and remain the same.
In addition, following Koetter et al. (2008), we estimate Eq. (10) using a stochastic cost frontier approach and calculate marginal costs (MCSFA
TAit ).
Appendix B
See Table B1.
Appendix C
See Table C1.
Appendix D
See Table D1.
Appendix E
See Table E1.
Appendix F
See Table F1.
Appendix G
See Table G1.
References
Acharya, V., Richardson, M. (Eds.), 2009. Restoring Financial Stability: How to Repair a Failed System? John Wiley and Sons, New York.
Agarwal, V., Taffler, R., 2008. Comparing the performance of market-based and accounting-based bankruptcy prediction models. Journal of Banking and Finance 32, 1541–1551.
Agoraki, M.K., Delis, M.D., Pasiouras, F., 2011. Regulation, competition and bank risk taking in transition countries. Journal of Financial Stability 7, 38–48.
Akhigbe, A., Madura, J., Martin, A.D., 2007. Effect of Fed policy actions on the default likelihood of commercial banks. Journal of Financial Research 30, 147–162.
Allen, F., Gale, D., 2000. Financial contagion. Journal of Political Economy 108, 1–33. Allen, F., Gale, D., 2004. Competition and financial stability. Journal of Money,
Credit, and Banking 36, 453–480. Anginer, D., Demirguc-Kunt, A., Zhu, M., 2012. How Does Bank Competition Affect
Systemic Stability? Policy Research Working Paper No. 5981, World Bank. Barth, J.R., Caprio Jr., G., Levine, R., 2001. The regulation and supervision of bank
around the world: a new database. In: Litan, R.E., Herring, R. (Eds.), Integrating Emerging Market Countries into the Global Financial System. Brookings- Wharton Papers in Financial Services. Brooking Institution Press, pp. 183–240.
Barth, J.R., Caprio Jr., G., Levine, R., 2006. Rethinking Bank Regulation. Cambridge University Press, Till angels govern.
Barth, J.R., Caprio Jr., G., Levine, R., 2008. Bank Regulations are Changing: But for Better or Worse? Policy Research Working Paper 4646, World Bank.
Barth, J.R., Caprio Jr., G., Levine, R., 2012a. The Evolution and Impact of Bank Regulations. Policy Research Working Paper 6288, World Bank.
Barth, J.R., Prabha, A., Swagel, P., 2012b. Just How Big is the Too Big to Fail Problem? Working Paper#12-06, Financial Institutions Center.
Baumol, W., 1982. An uprising in the theory of industry structure. The American Economic Review 72, 1–15.
76 Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77
Baumol, W., Panzar, J.C., Willig, R.D., 1982. Contestable Markets and the Theory of Industry Structure. Harcourt Brace Jovanovic, New York.
Beck, T., Demirguc-Kunt, A., Levine, R., 2006a. Bank concentration, competition, and crises: first results. Journal of Banking and Finance 30, 1581–1603.
Beck, T., Demirguc-Kunt, A., Levine, R., 2006b. Bank concentration and fragility: impact and mechanics. In: Carey, M., Stulz, R. (Eds.), The Risks of Financial Institutions. University of Chicago Press, Chicago.
Beck, T., 2008. Bank Competition and Financial Stability: Friends or Foes? Policy Research Working Paper No. 4656, World Bank.
Beck, T., Coyle, D., Dewatripoint, M., Freixas, X., Seabright, P., 2010. Bailing out the Banks: Reconciling Stability and Competition. Centre for Economic Policy Research, London.
Behr, P., Schmidt, R.H., Xie, R., 2010. Market structure, capital regulation, and bank risk taking. Journal of Financial Services Research 37, 131–158.
Berg, S.A., Kim, M., 1994. Oligopolistic interdependence and the structure of production in banking: an empirical evaluation. Journal of Money Credit and Banking 26, 309–322.
Berger, A., Demirguc-Kunt, A., Levine, R., Haubrich, J., 2004. Bank concentration and competition: an evolution in the making. Journal of Money, Credit and Banking 36, 433–453.
Berger, A., Klapper, L., Turk-Ariss, R., 2009. Bank competition and financial stability. Journal of Financial Services Research 35, 99–118.
Bharath, T.S., Shumway, T., 2008. Forecasting default with the Merton distance to default model. Review of Financial Studies 21, 1339–1369.
Black, F., Scholes, M., 1973. The pricing of options and corporate liabilities. Journal of Political Economy 81, 637–654.
Boot, A., Greenbaum, S., 1993. Bank regulation, reputation and rents: theory and policy implications. In: Mayer, C., Vives, X. (Eds.), Capital Markets and Financial Intermediation. Cambridge University Press, Cambridge, MA, pp. 262–285.
Boot, A.W.A., Thakor, A., 2000. Can relationship lending survive competition? Journal of Finance 55, 679–713.
Boyd, J.H., De Nicoló, G., 2005. The theory of bank risk-taking and competition revisited. Journal of Finance 60, 1329–1343.
Boyd, J.H., De Nicolo, G., Jalal, A.M., 2006. Bank Risk Taking and Competition Revisited: New Theory and Evidence. IMF Working Paper, WP/06/297.
Boyd, J.H., De Nicolo, G., Smith, B.D., 2004. Crises in competitive versus monopolistic banking systems. Journal of Money, Credit and Banking 36, 487–506.
Boyd, J.H., Prescott, E.C., 1986. Financial intermediary-coalitions. Journal of Economic Theory 38, 211–232.
Boyd, J.H., Runkle, D.E., 1993. Size and performance of banking firms. Journal of Monetary Economics 31, 47–67.
Bresnahan, T.F., 1982. The oligopoly solution concept is identified. Economic Letters 10, 87–92.
Bresnahan, T.F., 1989. Empirical studies in industries with market power. In: Schmelensee, R., Willig, R. (Eds.), Handbook of Industrial Organisation. North Holland, New York.
Brunnermeier, M.K., 2009. Deciphering the liquidity and credit crunch 2007–08. Journal of Economic Perspectives 23, 77–100.
Caminal, R., Matutes, C., 2002. Market power and bank failures. International Journal of Industrial Organisation 20, 1341–1361.
Campbell, J.Y., Hilscher, J., Szilagyi, J., 2008. In search of distress risk. Journal of Finance 63, 2899–2939.
Carbó-Valverde, S., Humphrey, D., Maudos, J., Molyneux, P., 2009. Cross-country comparisons of competition and pricing power in European banking. Journal of International Money and Finance 28, 115–134.
Careletti, E., 2010. Competition, Concentration and Stability in the Banking Sector. IstEin Working Paper 2010-10. <http://www.istein.org/images/attachments/ paper_8.pdf>.
Carletti, E., 2008. Competition and regulation in banking. In: Boot, A.W.A., Thakor, A. (Eds.), Handbook of Financial Intermediation and Banking. Elsevier, Amsterdam.
Cetorelli, N., Hirtle, B., Morgan, D., Peristiani, S., Santos, J., 2007. Trends in financial market concentration and their implications for market stability. Federal Reserve Bank of New York Economic Policy Review, 33–51.
Chan-Lau, J.A., Sy, A.N.R., 2007. Distance to default in banking: a bridge too far? Journal of Banking and Regulation 9, 14–24.
Čihák, M., Hesse, H., 2010. Islamic banks and financial stability: an empirical analysis. Journal of Financial Services Research 38, 95–113.
Claessens, S., Laeven, L., 2004. What drives bank competition? Some international evidence. Journal of Money, Credit, and Banking 36, 563–583.
Delis, M., Molyneux, P., Pasiouras, F., 2011. Regulation and productivity growth in banking: evidence from transition economies. Journal of Money, Credit and Banking 43, 735–764.
Diamond, D., 1984. Financial intermediation and delegated monitoring. Review of Economic Studies 51, 393–414.
Diamond, D.W., Dybvig, P.H., 1983. Bank runs, deposit insurance, and liquidity. Journal of Political Economy 91, 401–419.
Duffie, D., Saita, L., Wang, K., 2007. Multi-period corporate failure prediction with stochastic covariates. Journal of Financial Economics 83, 635–665.
Elzinga, K.G., Mills, D.E., 2011. The Lerner index of monopoly power: origins and uses. American Economic Review 101, 558–564.
Evrensel, A.Y., 2008. Banking crisis and financial structure: a survival-time analysis. International Review of Economics and Finance 17, 589–602.
Fernández de Guevara, J., Maudos, J., Perez, F., 2005. Market power in European banking sectors. Journal of Financial Services Research 27, 109–137.
Goddard, J., Liu, H., Molyneux, P., Wilson, J., 2011. The persistence of bank profit. Journal of Banking and Finance 35, 2881–2890.
Gropp, R., Vesala, J., Vulpes, G., 2004. Market indicators, bank fragility, and indirect market discipline. Federal Reserve Bank of New York Economic Policy Review 10, 53–62.
Gropp, R., Vesala, J., Vulpes, G., 2006. Equity and bond market signals as leading indicators of bank fragility. Journal of Money, Credit, and Banking 38, 399–428.
Hall, A.R., 2005. Generalized Method of Moments. Oxford University Press, Oxford. Hasan, I., Marton, K., 2003. Development and efficiency of a banking sector in a
transitional economy: Hungarian experience. Journal of Banking and Finance 27, 249–2271.
Hellmann, T., Murdock, K., Stiglitz, J., 2000. Liberalization, moral hazard in banking, and prudential regulation: are capital requirements enough? American Economic Review 90, 147–165.
Heritage Foundation, 2010. Index of Economic Freedom, 2010, Methodology for the 10 Economic Freedoms.
Hillegeist, S.A., Keating, E.K., Cram, D.P., Lundstedt, K.G., 2004. Assessing the probability of bankruptcy. Review of Accounting Studies 9, 5–34.
IDC, 2012. 72% of Asia/Pacific Banks Expect Higher Profitability in 2012, Reports IDC Financial Insights, IDC Press Release. <http://www.idc.com/ getdoc.jsp?containerId=prMY23339812> (23.02.12).
International Monetary Fund, 2009. Global Financial Stability Report, April. Iwata, G., 1974. Measurement of conjectural variations in oligopoly. Econometrica
42, 947–966. Jiang, C., Yao, S., Feng, G., 2013. Bank ownership, privatization, and performance:
evidence from a transition country. Journal of Banking and Finance 37, 3364– 3372.
Kane, E.J., 2010. Redefining and Containing Systemic Risk. Working paper, Boston College.
Keeley, M., 1990. Deposit insurance, risk, and market power in banking. American Economic Review 80, 1183–1200.
Koetter, M., Kolari, J., Spierdijk, L., 2008. Efficient Competition? Testing the ‘‘quiet life’’ of US Banks with Adjusted Lerner Indices. Working Paper, Groningen University.
Koetter, M., Kolari, J., Spierdijk, L., 2012. Enjoying the quiet life under deregulation? evidence from adjusted Lerner indices for US Banks. Review of Economics and Statistics 94, 462–480.
Laeven, L., Levine, R., 2009. Bank governance, regulation and risk taking. Journal of Financial Economics 93, 259–275.
Lau, L., 1982. On identifying the degree of competitiveness from industry price and output data. Economics Letters 10, 93–99.
Lepetit, L., Nys, E., Rous, P., Tarazi, A., 2008. Bank income structure and risk: an empirical analysis of European banks. Journal of Banking and Finance 32, 1452– 1467.
Liu, H., Molyneux, P., Nguyen, Linh.H., 2012. Competition and risk in South East Asian commercial banking. Applied Economics 44, 3627–3644.
Llewellyn, D.T., 2007. The Northern Rock crisis: a multi-dimensional problem waiting to happen. Journal of Financial Regulation and Compliance 16, 35– 58.
Lozano-Vivas, A., Pasiouras, F., 2010. The impact of non-traditional activities on the estimation of bank efficiency: International evidence. Journal of Banking and Finance 34, 1436–1449.
Marcus, A.J., 1984. Deregulation and bank financial policy. Journal of Banking and Finance 8, 557–565.
Martinez-Miera, D., Repullo, R., 2010. Does competition reduce the risk of bank failure? Review of Financial Studies 23, 3638–3664.
Matutes, C., Vives, X., 1996. Competition for deposits, fragility and insurance. Journal of Financial Intermediation 5, 184–216.
Maudos, J., Fernández de Guevara, J., 2004. Factors explaining the interest margin in the banking sectors of the European Union. Journal of Banking and Finance 28, 2259–2281.
Maudos, J., Solís, L., 2009. The determinants of net interest income in the Mexican banking system: an integrated model. Journal of Banking and Finance 33, 1920– 1931.
McDonald, R., 2002. Derivative Markets. Addison Wesley, Boston, MA. Merton, R.C., 1974. On the pricing of corporate debt: the risk structure of interest
rates. Journal of Finance 29, 449–470. Milne, A., 2009. The Fall of the House of Credit. Cambridge University Press,
Cambridge. Mishkin, F.S., 1999. Financial consolidation: dangers and opportunities. Journal of
Banking and Finance 23, 675–691. Mishkin, F.S., 2006. How big a problem is too big to fail? Journal of Economic
Literature 44, 988–1004. OECD, 2011. Bank Competition and Financial Stability. <http://www.oecd.org>. Panzar, J.C., Rosse, J.N., 1987. Testing for monopoly equilibrium. Journal of Industrial
Economics 35, 443–456. Ramakrishnan, R., Thakor, A., 1984. Information reliability and a theory of financial
intermediation. Review of Economic Studies 51, 415–432. Rosenblum, H., 2011. Choosing the Road to Prosperity: Why we must End Too Big to
Fail—Now, Federal Reserve Bank of Dallas, Annual Report, 3–23. Saez, L., Shi, X., 2004. Liquidity pools, risk sharing and financial contagion. Journal of
Financial Services Research 25, 5–23. Schaeck, K., Cihak, M., 2008. How does Competition Affect Efficiency and Soundness
in Banking? New Empirical Evidence. Working Paper No. 932, European Central Bank.
Schaeck, K., Cihak, M., Wolfe, S., 2009. Are competitive banking systems more stable? Journal of Money, Credit and Banking 41, 711–734.
Xiaoqing (Maggie) Fu et al. / Journal of Banking & Finance 38 (2014) 64–77 77
Schmalensee, R., 1982. Antitrust and the new industrial economics. American Economic Review 72, 24–28.
Shaffer, S., 1993. A test of competition in Canadian banking. Journal of Money, Credit and Banking 25, 49–61.
Sheng, A., 2009. From Asian to Global Financial Crisis: an Asian Regulator’s View of Unfettered Finance in the 1990s and 2000s. Cambridge University Press, Cambridge.
Smith, B., 1984. Private information, deposit interest rates, and the ‘stability’ of the banking system. Journal of Monetary Economics 14, 293–317.
Soedarmon, W., Machrouh, F., Tarazi, A., 2011. Bank market power, economic growth and financial stability: evidence from Asian banks. Journal of Asian Economics 22, 460–470.
Sun, L., Chang, T.-P., 2011. A comprehensive analysis of the effects of risk measures on bank efficiency: evidence from emerging Asian countries. Journal of Banking and Finance 35, 1727–1735.
Turk Ariss, R., 2010. On the implications of market power in banking: evidence from developing countries. Journal of Banking and Finance 34, 765–775.
Uhde, A., Heimeshoff, U., 2009. Consolidation in banking and financial stability in Europe: empirical evidence. Journal of Banking and Finance 33, 1299– 1311.
Vassalou, M., Xing, Y., 2004. Default risk in equity returns. Journal of Finance 59, 831–868.
Williamson, S., 1986. Costly monitoring financial intermediation, and equilibrium credit rationing. Journal of Monetary Economics 18, 159–179.
Yeyati, E.L., Micco, A., 2007. Concentration and foreign penetration in Latin American banking sectors: impact on competition and risk. Journal of Banking and Finance 31, 1633–1647.
- Bank competition and financial stability in Asia Pacific
- 1 Introduction
- 2 Literature review
- 3 Methodology
- 3.1 Market-based risk measure
- 3.2 Accounting-based risk measure
- 3.3 Concentration and competition measures
- 3.4 Other control variables
- 4 Data
- 5 Empirical results
- 6 Conclusions
- Appendix A Translog cost function for estimating bank marginal cost
- Appendix B
- Appendix C
- Appendix D
- Appendix E
- Appendix F
- Appendix G
- References
Capital-regulation-bank-competition-and-financial-stability_2011_Economics-Letters.pdf
Economics Letters 113 (2011) 256–258
Contents lists available at SciVerse ScienceDirect
Economics Letters
journal homepage: www.elsevier.com/locate/ecolet
Capital regulation, bank competition, and financial stability Hendrik Hakenes a,b,∗, Isabel Schnabel b,c,d,1 a University of Bonn, Adenauerallee 24-42, 53113 Bonn, Germany b Max Planck Institute for Research on Collective Goods, Bonn, Germany c Gutenberg School of Management and Economics, Johannes Gutenberg University Mainz, 55099 Mainz, Germany d Centre for Economic Policy Research, London, United Kingdom
a r t i c l e i n f o
Article history: Received 18 November 2008 Received in revised form 2 July 2011 Accepted 14 July 2011 Available online 29 July 2011
JEL classification: G21 G28 D43
Keywords: Bank competition Capital regulation Risk-shifting Banking stability
a b s t r a c t
We analyze capital requirements if banks compete for loans and deposits. Banks and firms are subject to a risk-shifting problem. The ambiguous effect of competition on banks’ risk-taking translates into an ambiguous effect of capital requirements on financial stability.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
It is a widely held view that there is a trade-off between competition and stability in banking. The argument goes that competition erodes banks’ profit margins and charter values, which increases risk-taking incentives (Keeley, 1990; Allen and Gale, 2004). In an important paper, Boyd and De Nicolò (2005) have shown that this trade-off is not robust to the introduction of loan market competition. In their model, higher competition induces banks to lower loan rates, which mitigates the borrowers’ moral hazard problem, and hence risk-taking. Under this view, competition increases banking stability.
The diverging results are driven by the way that banks’ risk- taking ismodeled. In the first type ofmodels, banks solve a portfolio problem: they hold a portfolio of projects and choose the riskiness of these projects; given limited liability and deposit insurance, banks are subject to a risk-shifting problem. In the second type of
∗ Corresponding author at: University of Bonn, Adenauerallee 24-42, 53113Bonn, Germany. Tel.: +49 228 73 9225; fax: +49 228 73 5048.
E-mail addresses: [email protected] (H. Hakenes), [email protected] (I. Schnabel). 1 Tel.: +49 6131 39 24191; fax: +49 6131 39 25588.
0165-1765/$ – see front matter© 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.econlet.2011.07.008
models, banks solve an optimal contracting problem: they extend loans to entrepreneurs who determine the risk of their projects. Here the entrepreneurs are subject to a risk-shifting problem.
Our paper analyzes what these results imply for the effective- ness of bank capital regulation. For this purpose, we develop a model that encompasses both approaches. The optimal contract- ing problem looks exactly as in the paper by Boyd and De Nicolò (2005). We modify that model by adding a portfolio problem, al- lowing banks to choose the correlation of their loans. In such a setup, the relationship between banking competition and stabil- ity is ambiguous.2 We then introduce costly bank equity and capi- tal regulation and study the impact of capital requirements on the risk of individual loans, a bank’s correlation, and its probability of default.
Our model shows that capital regulation may destabilize the banking sector through its effect on banking competition. Stricter capital requirements attenuate competition for loans, implying higher loan rates, and hence higher risk-taking by firms. Therefore,
2 Martínez-Miera and Repullo (2010) present yet another channel. As in Boyd and De Nicolò (2005), bank competition reduces the moral hazard problem of entrepreneurs, but it also erodes banks’ capital buffers, leading to a U-shaped relationship between competition and stability.
H. Hakenes, I. Schnabel / Economics Letters 113 (2011) 256–258 257
the risk of single loans increases. Stricter capital requirementsmay also induce banks to choose a higher correlation of loans. Overall, these two effects may translate into an increase in the banks’ probability of default.
2. Model setup
Our setup follows the model by Boyd and De Nicolò (2005). Consider an economy with three dates, 0, 1, and 2. There are three types of agents: entrepreneurs, depositors, and banks. All agents are risk neutral. Entrepreneurs. There is a continuum of entrepreneurs who have no own resources, but have access to risky projects of fixed size, normalized to 1. Projects have constant returns to scale and yield, per invested unit, S with probability p(S), and 0 otherwise. p(S) satisfies p(0) = 1, p′(S) < 0, and p′′(S) ≤ 0 for all S ∈ [0, S̄]. The entrepreneurs’ choice of S (date 1) is unobservable by the bank and depositors. At date 2, the bank observes only whether the project has been successful. By assumption, financing contracts are simple debt contracts. The aggregate demand for loans L̄ is represented by a downward-sloping inverse demand curve rL(L̄), satisfying rL(0) > 0, r ′
L(L̄) < 0, and r ′′
L (L̄) ≤ 0. Depositors. The aggregate supply of deposits D̄ is represented by an upward-sloping inverse supply curve rD(D̄), satisfying rD(0) < rL(0), rD(0) ≥ 0, r ′
D(D̄) > 0, and r ′′
D(D̄) ≥ 0. Deposits are insured at a flat premium α. Banks. There are N banks. Each bank j extends loans Lj that are financed by deposits Dj and inside equity Ej, hence Lj = Ej +
Dj. Aggregate deposits in the banking sector are equal to D̄ =∑N j=1 Dj, aggregate loans are L̄ =
∑N j=1 Lj. Each bank is run by
a single owner-manager who provides the equity; the owners’ opportunity costs of capital are rE > p(0) rL(0), such that equity finance is expensive.3 Equity finance would be inefficient in the absence of moral hazard, but it helps to mitigate excessive risk- taking by banks. Assume that a regulator imposes a minimum capital requirement β , i. e., Ej ≥ β Lj. The introduction of capital requirements is the firstmajor deviation from the setup introduced by Boyd and De Nicolò (2005).
Banks compete for deposits and loans in a Cournot fashion. When setting loan volumes, banks take into account the best responses of entrepreneurs. The secondmajor innovation concerns banks’ asset side: We assume that a bank can influence the correlation of its loans, ρj ∈ [0, 1], which is unobservable by outsiders. More specifically, we assume that a bank’s projects have some ‘‘natural’’ correlation ρ0. The bank can increase or decrease the correlation at a (non-monetary) cost Cj, which is proportional to the size of its portfolio, i. e., Cj = C(Lj, ρj) = Lj c(ρj). The cost function c(ρj) satisfies the following conditions: c(ρ0) = 0 and c ′(ρ0) = 0 for some ρ0 ∈ [0, 1], and c ′′(ρj) > 0 for all ρj ∈ [0, 1]. Hence, the cost function is strictly convex with a minimum at ρ0; any deviation from the ‘‘natural’’ correlation is costly.
We make the following simplifying assumption about the correlation structure: For a given choice of ρ, all projects are perfectly correlated with probability ρ, and perfectly uncorrelated with probability 1−ρ. In both cases, the expected portfolio payoff of the projects is equal to p(S) S. But the probability of default is equal to 1 − p(S) in the first case, and (due to the law of large numbers) zero in the second case. Hence, by raising ρ, a bank increases its default probability. ρ has a natural interpretation as the correlation between any two projects in the portfolio: with probability ρ, the correlation between any two projects is 1, with probability 1 − ρ, the correlation is 0. As a consequence, the ex- ante correlation between any two projects (and, hence, loans) is ρ ·1+ (1−ρ) ·0 = ρ. Fig. 1 shows the time structure of the game.
3 This assumption is also used by Hellmann et al. (2000), Repullo (2004), and Repullo and Suarez (2004).
Fig. 1. Time structure.
3. Equilibrium
We solve the model by backward induction. We concentrate on symmetric equilibria and drop the index j when there is no danger of confusion. At date 1, the entrepreneurs choose project risk S to maximize expected profits for a given loan rate rL. An entrepreneur’s expected return is p(S) (S − rL). The first-order condition is p′(S) (S − rL) + p(S) = 0. Compared with the first best, S is too high. An increase in rL induces a further increase in risk.
Before the entrepreneurs’ choice of S, banks choose the loan correlation ρj for given deposit, loan, and equity volumes. The expected profit of bank j is
Πj = ρj p(S)[rL(L̄) Lj − (rD(D̄) + α)Dj]
+ (1 − ρj)[p(S) rL(L̄) Lj − (rD(D̄) + α)Dj]
− rE Ej − Lj c(ρj)
= p(S) rL(L̄) Lj − [ρj p(S) + (1 − ρj)] (rD(D̄) + α)Dj
− rE Ej − Lj c(ρj). (1)
We make use of the fact that all firms choose the same S in equi- librium. The first-order condition with respect to ρj is
∂Πj
∂ρj = (1 − p(S)) (rD(D̄) + α)Dj − Lj c ′(ρj) = 0, (2)
where S depends on rL(L̄) because banks anticipate the en- trepreneurs’ risk choices. The first-best solution has Lj c ′(ρj) = 0, implying that ρj = ρ0. Since c ′′(ρj) > 0, ρj is higher than ρ0 for any p(S) < 1. Banks overspecialize due to limited liability and deposit insurance.
At the beginning of date 0, banks choose the profit-maximizing volumes of deposits, equity, and loans. Due to the balance sheet identity, Lj = Dj + Ej; hence, only two variables can be chosen independently. Since equity is expensive, a profit-maximizing bank takes nomore equity than is required by regulation, Ej = β Lj and Dj = (1 − β) Lj. A bank’s optimization problem becomes
max Lj
{Lj[p(S)rL(L̄) − (1 − β)[ρjp(S) + (1 − ρj)](rD((1 − β)L̄)
+α) − βrE − c(ρj)]} s. t. S + p(S) p′(S)
= rL(L̄)
and (1 − p(S)) (rD((1 − β)L̄) + α) (1 − β) = c ′(ρj). (3)
4. Capital regulation and bank risk
Now we analyze how capital regulation affects banks’ risk- taking. We consider the effect of capital regulation on the riskiness S of a single bank loan, a bank’s correlation ρ, and a bank’s probability of default. We will see that capital regulation affects banks’ risk not only by aligning the interests of banks and their creditors, but also through competition.
258 H. Hakenes, I. Schnabel / Economics Letters 113 (2011) 256–258
Consider an increase in the capital requirement β . This raises capital costs, which induces banks to choose lower deposit and loan volumes. The decrease in the aggregate loan volume L̄ translates into an increase in the loan rate rL(L̄) and into higher risk-taking by entrepreneurs. Hence, a tighter capital regulation increases the risk of individual loans because it attenuates the competition for loans and exacerbates the entrepreneurs’ moral hazard problem.
Proposition 1. Stricter capital requirements increase the risk of individual loans by raising the entrepreneurs’ risk-taking, dS/dβ > 0.
The effect on overall bank risk depends also on a bank’s correlation. A bank’s portfolio choice is governed by (2), which can be written as c ′(ρ) = (1 − β) Φ , where Φ = [rD((1 − β) L̄) +
α][1 − p(S(L̄))] > 0. Φ depends on β through rD and L̄. Totally differentiating (2), we derive the effect of capital regulation on a bank’s correlation:
dρ dβ
= 1
c ′′(ρ)
[ −Φ + (1 − β)
∂Φ
∂ L̄ ∂ L̄ ∂β
+ ∂Φ
∂β
] , (4)
where ∂ L̄/∂β < 0, ∂Φ/∂ L̄ = (1 − β) r ′
D(D̄) (1 − p(S)) − (rD(D̄) +
α) p′(S) S ′(rL) r ′
L(L̄), and ∂Φ/∂β = −(1 − p(S)) r ′
D(D̄) L̄ < 0. The effect of capital regulation on a bank’s correlation is am-
biguous: First, a stricter regulation forces the banker to hold a higher equity share. As a consequence, the banker has a higher stake in the bank and chooses a lower correlation ρj. Second, an increase in β decreases the deposit rate through its negative effect on the aggregate deposit volume D̄. As a result, the bank’s margin increases, which also induces the bank to choose a lower ρj. This is the standard ‘‘charter value’’ effect found in the literature on the tradeoff between competition and stability. The third channel goes in the opposite direction. An increase in β raises the loan rate and thereby entrepreneurs’ risk-taking (Proposition 1), which trans- lates into a higher default probability of the bank’s loans, 1− p(S). This makes gambling more attractive for the bank because default occurs more frequently; ρj increases. In this sense, the bank’s and the entrepreneurs’ risk-taking are complementary to each other. Proposition 2 gives the condition under which the third channel overcompensates the other two channels.
Proposition 2. Stricter capital requirements increase a bank’s port- folio correlation, dρ/dβ > 0, if and only if
(1 − β)
∂Φ
∂ L̄ ∂ L̄ ∂β
+ ∂Φ
∂β
> Φ. (5)
Finally, we discuss how capital regulation affects a bank’s default probability, PD = ρ (1 − p(S)). Taking the derivative with respect to β yields
d PD dβ
= dρ dβ
(1 − p(S)) − ρ p′(S) dS dβ
, (6)
leading to the following proposition.
Proposition 3. Stricter capital requirements increase a bank’s prob- ability of default, d PD/dβ > 0, if and only if
dρ dβ
> ρ p′(S)
1 − p(S) dS dβ
. (7)
Proposition 3 shows that tighter capital regulation may render a bank more risky. From Proposition 1, we know that a higher β makes every loan in a bank’s portfolio more risky. From Proposition 2, we know that a stricter capital regulation may also induce the bank to choose a more highly correlated portfolio. In such situations, it is obvious that the bank’s default probability must increase. Even if a higher β makes the bank take a less correlated portfolio, the effect on the entrepreneurs’ risk-taking may dominate, leading to a higher default probability of the bank.
5. Conclusion
Our paper has shown that capital regulation may destabilize the banking sector through its effect on banking competition. The ambiguous effect of competition on banks’ risk-taking translates into an ambiguous effect of capital regulation. A stabilizing effect of capital regulation tends to obtain in those situations where competition has a destabilizing effect (i. e., the ‘‘charter value effect’’ dominates), and vice versa. Given the prominence of capital requirements in today’s regulation, the empirical relationship between competition and stability should be of great interest to policy makers. The existing empirical evidence on the presumed trade-off between competition and stability in banking is rather mixed. In the light of our model, this suggests that capital regulation may not be suited in all circumstances to prevent excessive risk-taking in banking.
Acknowledgments
We thank John Boyd, Gianni De Nicolò, Jens Grunert, Robert Hauswald, Martin Hellwig, Eric Maskin (the editor), and an anony- mous referee for helpful suggestions.We also thank participants of the EFA in Ljubljana, the ESEM in Budapest, the Tor Vergata Confer- ence in Rome, the GEABA in Tübingen, as well as seminar partici- pants at theMPI in Bonn and the SFB/TR 15 in Berlin for comments. Financial support fromDeutsche Forschungsgemeinschaft is grate- fully acknowledged.
References
Allen, F., Gale, D., 2004. Competition and financial stability. J. Money, Credit, Banking 36, 453–480.
Boyd, J.H., De Nicolò, G., 2005. The theory of bank risk taking and competition revisited. J. Finance 60, 1329–1343.
Hellmann, T., Murdock, K.C., Stiglitz, J.E., 2000. Liberalization, moral hazard in banking, and prudential regulation: are capital requirements enough? Amer. Econ. Rev. 90, 147–165.
Keeley, M.C., 1990. Deposit insurance, risk and market power in banking. Amer. Econ. Rev. 80, 1183–1200.
Martínez-Miera, D., Repullo, R., 2010. Does competition reduce the risk of bank failure? Rev. Finan. Stud. 23, 3638–3664.
Repullo, R., 2004. Capital requirements, market power, and risk-taking in banking. J. Finan. Intermediation 13, 156–182.
Repullo, R., Suarez, J., 2004. Loan pricing under basel capital requirements. J. Finan. Intermediation 13, 496–521.
- Capital regulation, bank competition, and financial stability
- Introduction
- Model setup
- Equilibrium
- Capital regulation and bank risk
- Conclusion
- Acknowledgments
- References
Competition-and-financial-stability-in-European-cooperative-banks_2014_Journal-of-International-Money-and-Finance.pdf
Journal of International Money and Finance 45 (2014) 1–16
Contents lists available at ScienceDirect
Journal of International Money and Finance
journal homepage: www.elsevier .com/locate/ j imf
Review
Competition and financial stability in European cooperative banks
Franco Fiordelisi a,b,1, Davide Salvatore Mare c,*
a Faculty of Economics, University of Rome III, Italy bBangor Business School, Bangor University, UK cBusiness School, The University of Edinburgh, 29 Buccleuch Place, Edinburgh EH8 9JS, UK
JEL classification: C23 G21
Keywords: Bank soundness Cooperative banks Competition Financial stability
* Corresponding author. Tel.: þ44 (0)131 651 50 E-mail addresses: [email protected] (F. Fio
1 Faculty of Economics, Via S. D’Amico 77, 00145
http://dx.doi.org/10.1016/j.jimonfin.2014.02.008 0261-5606/� 2014 Elsevier Ltd. All rights reserved
a b s t r a c t
Cooperative banks are a driving force for socially committed business at the local level, accounting for around one fifth of the European Union (EU) bank deposits and loans. Despite their importance, little is known about the relationship between bank stability and competition for these small credit institutions. Does competition affect the stability of cooperative banks? Does the financial stability of banks increase/decrease when competition is higher? We assess the dynamic relationship between competition and bank soundness (both in the short and long run) among Eu- ropean cooperative banks between 1998 and 2009. We obtain three main results. First, we provide evidence in line with the competition-stability view proposed by Boyd and De Nicolò (2005). Bank market power negatively “Granger-causes” banks’ soundness, meaning that there is a positive relationship between competition and stability. Second, we find that this fundamental relationship does not change during the 2007–2009 financial crisis. Third, we show that increased homogeneity in the cooper- ative banking sector positively affects bank soundness. Our find- ings have important policy implications for designing and implementing regulations that enhance the overall stability of the financial system and in particular of the cooperative banking sector.
� 2014 Elsevier Ltd. All rights reserved.
77; fax: þ44 (0)131 651 3197. rdelisi), [email protected] (D.S. Mare). Rome, Italy. Tel.: þ39 065 733 5672; fax: þ39 065 733 5797.
.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–162
1. Introduction
Recent regulatory developments toward a more integrated European banking market point to the establishment of a European Banking Union.2 In response to the 2007–2009 financial turmoil, the new European banking union reforms3 introduce the issue of the supervision over highly heterogeneous types of banks. Moreover, international efforts have been coordinated toward the development of a new regulatory framework to control systemic risks (e.g., the Basel Committee’s framework on global systematically important banks). Nevertheless, small banks are different from international com- mercial banks and are important for local economic development. This is not recognised in the existing literature, as little is known about the relationship between bank stability and competition among small credit institutions, particularly given the variety of local structures.
Cooperative banks4 are a driving force for socially committed business at a local level, accounting for around one fifth of the European Union (EU) bank deposits and loans. As stated by the European As- sociation of Co-operative Banks (EACB, 2012, p. 4) “in the unique context of the global financial crisis, this sector demonstrated its robustness and resilience, as well as its ability to act as a key driver for the real economy.” Specifically, in 2012, the EU had 4000 cooperative banks with 72,000 branches, more than 850,000 employees, 56 million members, 217 million clients, 3932 billion Euro in deposits, 4034 billion Euro of loans, and 6951 billion Euro in total assets. Cooperative banks are significantly different from other types of credit institutions in three important aspects: ownership, control, and benefits.5
Specifically, owners of cooperative banks are often also their customers (usually referred to as “members”). Membership is not transferable, is limited to individual equity shares, and is redeemable only at a nominal value. As a consequence, members cannot accumulate votes by purchasing shares on the market. In addition, cooperative banks are characterized by the one-member one-vote principle regardless of the amount of capital owned. EACB (2012) notes that “their unique stakeholders structure brings about efficiency and sound governance: members control the co-operative by exerting checks and balances at each level of the business. This minimizes the organization’s risk, identifies credit- worthiness and gives immediate response to customers’ needs”. As such, cooperative banks usually raise capital through newmemberships or retained profits (often a fixedminimum percentage amount established by law). In regard to benefits, cooperative banks aim to maximize members’ value by of- fering products and services alongwith the distribution of profits. It follows that profit-maximization is not the sole business objective of these credit institutions given their risk profile and business model (e.g., focus on retail banking). In addition, cooperative banks are mainly local or regionally based6 with strong links to the communities they serve.
A number of papers have analysed small credit institutions by focussing on performance (Goddard et al., 2008a; Kontolaimou and Tsekouras, 2010), diversification (Goddard et al., 2008b; Lepetit et al., 2008; Mercieca et al., 2007; McKillop and Wilson, 2011), risk of failure (Fiordelisi and Mare, 2013), and ownership structure (Gorton and Schmid, 1999). A debated issue is whether cooperative banks are more stable than commercial banks. During the recent financial crisis, cooperative banks performed
2 The new framework involves regulatory arrangements designed to mitigate longer-term financial crises through the centralized delivery of EU-wide rules. It engenders a Single Supervisory Mechanism led by the European Central Bank, a Single Resolution Mechanism, a Single Resolution Fund and a Single Rulebook (i.e., unified regulatory framework).
3 The new reforms comprise the Single Supervisory Mechanism (conferring specific supervision tasks on the European Central Bank with the objective to strengthen the Economic and Monetary Union) voted by European Parliament Plenary on September 12, 2013 and the Single Resolution Mechanism. On July 10, 2013 the European Commission proposed a new text and the European Council agreed the general approach on Single Resolution Mechanism on December 19, 2013.
4 Cooperative banks belong to the broader category of financial cooperatives, which also includes other credit institutions such as credit unions, banks set up by other cooperatives (such as The Co-operative Bank in the UK), and building societies.
5 Note that the cooperative credit sector in Europe is not entirely uniform in terms of legal framework, size, and organization. Nevertheless, the distinctive features described differentiate cooperative banks from other types of credit institutions.
6 An exception is the large cooperative networks such as Dutch Rabobank Group or French Crédit Agricole Group. In this case, the local and regional cooperatives are members of a central institution (called APEX) that centralises some services and processes to reap the benefits from economies of scale and scope.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–16 3
better than commercial banks, as discussed by Jose Manuel Barroso (President of the European Commission) in 2011: “Co-operative businesses that have stayed faithful to co-operative values and principles and the co-operative banks which rely onmembers’ funds and are controlled by local people have generally been able to resist the crisis very well.”7 Barroso’s statement is consistent with various papers (Hesse and Cihák, 2007; Ayadi et al., 2010) that provide empirical evidence that cooperative banks are more stable than commercial banks because they have a great deal of soft information (which is hard to collect) on the creditworthiness of members/customers and are therefore less likely tomake lending mistakes. Furthermore, size appears to be positively related to systemic risk (Vallascas and Keasey, 2012; De Jonghe, 2010), and the majority of cooperative banks are small, rural credit in- stitutions. Conversely, several studies suggest that cooperative banks are more fragile than commercial banks (Goodhart, 2004; Brunner et al., 2004; Fonteyne, 2007) and have higher default rates. For instance, Fiordelisi and Mare (2013) document that the default rate of Italian cooperative banks was four times higher than that of commercial banks in the period before the financial crisis (1997–2006). We believe supervisory behaviour could be the key to reconciling these opposing views. Specifically, cooperative banks are likely to have less volatile earnings than commercial banks, but in periods of financial stability, supervisors are more inclined to shut down distressed cooperative banks than distressed commercial banks, consistent with a Too-Big-To-Fail policy.
Rather than entering the debate about whether cooperative banks aremore fragile than commercial banks,8 this study of the financial stability of cooperative banks investigates three key issues. First, cooperative banks are different from commercial banks, and their stability is influenced by different factors.9 Recent studies have focused on productive performance related to technological development (Kontolaimou and Tsekouras, 2010), performance and risk factors associated with ownership structure (Iannotta et al., 2007), and cost efficiency and financial structure (Girardone et al., 2009). Second, competition is likely to be one of the key factors influencing bank stability, but its influence is probably different for commercial and cooperative banks. Third, it is necessary to account for banking super- visors’ behaviour to assess the link between competition and risk. Surprisingly, whilst there is a substantial literature investigating the link between competition and bank stability among commercial banks, no studies have specifically analysed cooperative banks.
Does competition affect the stability of cooperative banks? Does bank soundness increase/decrease when competition is higher? This study empirically addresses these questions. By analysing a large sample of cooperative banks in the EU between 1998 and 2009, we obtain three main results. First, we show that market power is negatively related to individual bank stability, meaning there is a positive relationship (both in the short and long run) between competition and stability in line with the competition-stability view proposed by Boyd and De Nicolò (2005). Second, we show that the 2007 financial crisis does not change the direction of the relation between competition and stability. Third, we find a statistically significant relationship between the Herding measure and bank stability. This result is particularly interesting for policy makers because the level of industry homogeneity has a positive influence on bank stability.
We estimate competition using the Lerner Index of Monopoly Power, recently used in a variety of studies (Maudos and de Guevara, 2007; Turk Ariss, 2010; Radi�c et al., 2012; among many others). We use both pooled ordinary least squares regressions and fixed-effects panel regressions to control for spurious relationships and country-specific effects in our tests of whether changes in competition predict variations in bank risk measures. We also analyse the impact that multifarious factors have on the competition–risk relationship, such as herding behaviour, the financial crisis, concentration in the loan and deposit markets, and bank-level fundamentals.
The remainder of the paper is structured as follows. Section 2 summarizes the literature and the research hypotheses. Section 3 presents the data and variables employed in the analysis. In Section 4,
7 Source: European Association of Co-operative Banks (2012), http://www.eacb.eu/en/cooperative_banks/what_they_say_ about_us.html.
8 This paper does not aim to discuss cooperative bank fragility. Rey and Tirole (2007), Beck et al. (2009), Hesse and Cihák (2007), and Fonteyne (2007) are useful sources from a theoretical, empirical, and policy perspective, respectively.
9 See Boonstra and Mooij (2012) for a detailed explanation.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–164
we discuss the empirical approach. Section 5 summarises the results from the estimations and the robustness checks. Section 6 concludes.
2. Literature review and research hypotheses
We empirically assess whether an increase in competition predicts higher instability among cooperative banks. Many studies have examined this relationship within the commercial banking sector from both theoretical and empirical standpoints.
From a theoretical perspective, there are two concurrent views. The competition-fragility view (see, among others, Marcus, 1984; Keeley, 1990; Allen and Gale, 2004; Beck et al., 2006; Matsuoka, 2013) argues that higher competition leads to more risk in banking and the erosion of bank charter value. In contrast, other papers suggest that higher competition could transform the nature of banking and induce banks to become more relationship-oriented (Boot and Thakor, 2000). As such, the competition-stability view (Boyd and De Nicolò, 2005; De Nicolò and Lucchetta, 2009) challenges the negative effects of concentration. Under this theory, the considerablemarket power of only a few banks will cause them to raise the interest rates on loans, which will induce adverse selection (risky projects are financed) and moral hazard (risk shifting), with a negative impact on the stability of the banking system.
A recent stream of empirical studies has tried to measure the effects of competition and market power on stability. Several works have tested the relation between banking market structure and risk by focussing on credit risk (Hakenes and Schnabel, 2010; Fiordelisi et al., 2011), interest rate risk (Delis and Kouretas, 2011), or the broader default risk (Repullo, 2004; Schaeck et al., 2009; Berger et al., 2009; Jiménez et al., 2010; Turk Ariss, 2010) and have found mixed evidence. For instance, Boyd et al. (2006) and De Nicoló and Loukoianova (2007) show that financial instability increases in less competitive markets, but Jiménez et al. (2010) find opposite evidence (i.e., risk decreases as bank market power increases). Schaeck et al. (2009) analyse banks operating in 45 nations over 1980–2005 and find that more competitive and more concentrated banking systems are less likely to experience a systemic crisis and have a longer time to crisis. Berger et al. (2009) study a large sample of banks in 23 developed countries and observe that even if an increase in bank market power leads to riskier portfolios, the effect on stability could be offset by a greater franchise value. In an attempt to reconcile the mixed empirical evidence, Beck et al. (2013) show that greater competition is generally associated with a larger impact on banks’ risk-taking activities in countries with more stringent activity restrictions, more herding in revenue structure, less concentrated banking markets, and more generous deposit insurance.
While the extant literature focuses on commercial banking, we find few studies that examine cooperative banking. Cooperative banks are key to the EU economy; hence, it is important to inves- tigate the competition-stability link among these credit institutions. EU policy makers seem to agree, pointing out the importance of these differences when designing new regulations. For example, Michel Barnier, the EU commissioner responsible for the internal market and services, stated in 2011 that “we are totally faithful to Basel’s spirit, letter and level of ambition. But you cannot apply rules to 8200 banks as you would to 20 banks. That is why we take into account the specificities of the European banking sector, with its mutual or co-operative banks and its bank and insurance groups.”10
Only a few papers are loosely related to the research questions we address in this study. The most relevant research is by Liu et al. (2012), who investigate the link between competition and stability among regional banks (which include cooperative banks) in 11 European countries between 2000 and 2008.Without explicitly focussing on cooperative banks, which are substantially different from savings banks, the authors find a positive link between competition and bank stability and show that coop- erative banks have a positive marginal effect on bank stability. Hesse and Cihák (2007) analyse the stability of cooperative, commercial, and savings banks (measured using the Z-score) by estimating a linear regression model with dummy variables capturing different bank types. Without taking into
10 Source: European Association of Co-operative Banks (2012), http://www.eacb.eu/en/cooperative_banks/what_they_say_ about_us.html.
Table 1 Summary statistics.
Variable Symbol Obs Mean Std dev Min Max
Output price P 17,074 0.0571 0.0320 0.0003 2.5915 Marginal cost MC 17,074 0.0271 0.0236 0.0000 1.7527 Funding-adjusted MC FA_MC 17,074 0.0286 0.0275 0.0000 2.0872 Lerner index LER 17,074 0.5228 0.1265 �2.5818 1.0000 Funding-adjusted LER FA_LER 17,074 0.4952 0.1494 �2.5301 1.0000 Herding measure HERD 17,074 0.0672 0.0271 0.0123 0.2386 Concentration loans HHI LOANS 17,074 0.0029 0.0073 0.0001 0.0504 Concentration deposits HHI DEPOSITS 17,074 0.0070 0.0156 0.0001 0.0911 Z-score Z 17,074 17.2572 8.0582 0.1823 158.9526 Z-rob Z_Rob 17,074 17.2514 7.8394 1.4571 162.9922
This table presents the descriptive statistics of our sample of cooperative banks in the European banking system between 1998 and 2009 for the main variables used in the model. It is at first surprising that both measures of market power (i.e., Lerner Index and funding-adjusted Lerner Index) are negative for some observations, though for 97 and 132 observations only (0.6% and 0.8% of the sample respectively). We argue that this could be the case when cooperative banks start operations and bear high fixed costs (e.g., for fixed assets).
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–16 5
account competition in the banking industry, the authors conclude that cooperative banks are more stable than commercial banks.
Our paper contributes to the existing literature in several ways. First, we focus on cooperative banks, which allows us to use a homogenous data set rather than relying on dichotomous variables to control for differences across different types of banks. Second, we estimate stability and competition at the individual bank level using Z-scores and the Lerner Index, both of which have been used in recent studies (Boyd et al., 2006; De Nicoló and Loukoianova, 2007; Berger et al., 2009). Third, we account for the potential impact of regulatory intervention on the relation between competition and bank risk. Whilst the Too-Big-To-Fail or Too-Important-To-Fail views do not apply to cooperative banks, we recognize that cooperative bank closure policies may suffer from an implicit “Too-Many-To-Fail” problem, as suggested by Acharya and Yorulmazer (2007). Specifically, when the number of bank failures is large, the regulator finds it ex-post optimal to bail out some or all distressed institutions, triggering incentives to herd ex-ante and increasing the risk that many banks may concurrently fail together ex-post. Similar to Beck et al. (2013), we investigate the assumption that competitionwill have a stronger impact on bank stability inmore homogeneous banking system (where herding behaviour is more likely).
3. Data sources and variables
Bank financial statements are taken from the Bureau van Dijk Bankscope database. We restrict our analysis to banks from the five largest cooperative banking sectors in Europe–Austria, France, Germany, Italy, and Spain – over the period of 1998–2009. In 2010, cooperative banks in these five countries accounted for 85% of total assets held by all EU cooperative banks.
To avoid duplication, we consider consolidated data where possible and unconsolidated data otherwise. We also omit banks for which relevant information is not available (e.g., total assets and total costs). After data cleaning, our final sample consists of 17,074 observations for 2529 cooperative banks in Austria, France, Germany, Italy, and Spain (accounting for 4%, 6%, 60%, 28%, and 2% of the observations, respectively). Table 1 reports the sample summary statistics.
Additional information on total market deposits and total market loans for each country comes from the European Central Bank. Data include the total loans and total deposits for all monetary and financial institutions toward unspecified sectors.11
11 We are aware that, as per definition of the European Central Bank, monetary financial institutions includes central banks, resident credit institutions, and other resident financial institutions (e.g., money market funds). Nevertheless, we believe this is a convenient proxy for the amount outstanding of loans and deposits markets in each country.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–166
A comprehensive set of variables is considered in the analysis to control for the effect of other factors on the relationship between competition and risk. We include variables such as heterogeneity, market concentration, and the 2007–2009 financial crisis that can directly affect the relationship be- tween stability and competition. We also include factors that could explain bank financial soundness, such as bank-level fundamentals and environmental determinants. Below, we first describe the main variables of interest in our analysis – the Lerner Index and bank stability – and then the other variables we include in the estimation.
3.1. Measuring competition: the Lerner Index
Following recent studies (Maudos and de Guevara, 2007; Casu and Girardone, 2009; Turk Ariss, 2010; among many others), we directly estimate competition using the Lerner Index of Monopoly Power (LER) as ameasure of cooperative bankmarket power. This indicator, which represents the extent to which market power allows firms to fix a price above the marginal cost, is calculated as follows:
LERi;t ¼ Pi;t �MCi;t Pi;t
; (1)
where Pi,t is the price of the output of bank i at year t, andMCi,t is the marginal cost. Higher index values imply greater market power. The price of output Q is calculated as total revenues (interest plus non- interest income) divided by total assets. In line with recent papers (Berger et al., 2009; Beck et al., 2013), we estimate the conventional marginal cost using a translog cost function with three inputs, one single output, and a time trend. The final specification is as follows:
ln TCi;t ¼ a0 þ a1 ln Q þ a2 2 ln Q2 þ
X3 j¼1
bj ln Pj þ 1 2
X3 j¼1
X3 k¼1
djk ln Pj ln Pk
þ 1 2
X3 k¼1
gj ln Q ln Pj þ s1t þ s2 2 t2 þ s3t � ln Q þ
X3 k¼1
jjt ln Pj þ εit ;
(2)
where TCi,t is the total costs (i.e., the sum of personnel expenses, other administrative expenses, and other operating expenses), and Q is the cooperative banks’ single output proxied by total assets. P1, P2, and P3 are the prices of the inputs used in the production process: P1 is the price of labour (i.e., personnel expenses over total assets); P2 is the price of physical capital (i.e., other administrative ex- penses plus other operating expenses over total fixed assets); and P3 is the price of borrowed funds (i.e., interest expenses over the sum of total deposits and money market funds). t is a time trend capturing the dynamics of the cost function over time, and a, b, d, g, s, and j are coefficients to be estimated. εit is a two-component error term computed as follows:
εit ¼ uit þ vit ; (3)
where vit is a two-sided error term, and uit is a one-sided disturbance term representing inefficiency.12
From Equation (2), the marginal costs can be derived as follows:
MCi;t ¼ TCi;t Qi;t
2 4ba1 þ ba2 ln Q þ
X3 j¼1
bgj ln Pj þ bs3t 3 5; (4)
12 vit is assumed to be independently and identically normally distributed with a mean of zero and a variance of s2 v, and
independent of uit ¼ {ui exp[�n (t – T)]}, where uit is a one-sided error term capturing the effects of inefficiency and assumed to be half-normally distributed with a mean of zero and a variance of s2u . n is an unknown parameter to be estimated that captures the effect of inefficiency change over time. We apply the common restrictions of standard symmetry and homogeneity of degree one in prices to the translog functional form (see Appendix A for the specification of the restrictions in the case of two inputs).
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–16 7
MCi,t obtained from Equation (4) is then substituted into Equation (2) to calculate the Lerner Index for bank i at time t, thereby giving us the dynamic change in market power across banks over time. We also calculate a different specification of MC using the funding-adjusted Lerner Index suggested by Maudos and de Guevara (2007) and Turk Ariss (2010). Specifically,MCi,t is derived from the estimation of the cost function that omits funding costs as one of the inputs. This procedure enables us to account for market power that may have previously been exercised in the deposit market. Specifically, by excluding funding costs, we obtain a clean proxy of pricing power that is not affected by market power that had previously originated in the deposit market while banks raise funds. Appendix A summarises the main steps to compute the funding-adjusted Lerner Index.
3.2. Measuring bank stability
We proxy individual bank stability using the Z-score, which has been used extensively in the banking literature (e.g., Boyd et al., 2006; Iannotta et al., 2007; Laeven and Levine, 2009). This measure is computed as the sum of the capital-asset ratio (CAR) and the return on assets (ROA) divided by the standard deviation of the return on assets (sROA). The underlying idea is to capture the number of standard deviations by which returns have to diminish in order to deplete the equity of a bank. Various approaches have been proposed to construct time-varying Z-score measures (for a review, see Lepetit and Strobel, 2013). We follow the methodology used by Hesse and Cihák (2007) to obtain a time- varying measure of individual bank stability: we estimate the Z-score employing sroa calculated using a cross-sectional technique and combine this with current period values of CARt and ROAt for each individual bank. The Z-score is then computed as follows:
Z � scorei;t ¼ ROAi;t þ CARi;t s � ROAc;t
� ; (5)
where ROAi,t is the return on assets for bank i in current period t. CARi,t denotes the capital-asset ratio for bank i in current period t. s(ROAc,t) is computed as the standard deviation of return on assets within each individual country c in current period t. The Z-score provides a measure of bank soundness. Higher values imply a higher degree of solvency, and therefore it gives a direct measure of bank sta- bility. In our analysis, we consider the natural logarithm of the Z-score to smooth out higher values within the distribution.
As it is estimated in (5), the Z-score allows us to have a time-varying measure of bank stability that does not suffer from endogeneity problems.13 Nevertheless, since ROAi,t and s(ROAc,t) are drawn from different distributions, this could create an inconsistency problem. To overcome this issue and getmore robust results, we estimate the Z-score similarly to Yeyati and Micco (2007) by using the following equation:
Z Robi;t ¼ m � ROAc;t
�þ CARi;t s � ROAc;t
� ; (6)
where m(ROAc,t) is the mean of the return on assets within each individual country c in current period t. Once more, we take the natural logarithm to smooth out extreme values in the Z-score distribution.
We also calculate two alternative measures of bank risk-taking to get more robust results and avert the presence of a spurious correlation, given that net operating profits are considered both in the computation of the Z-score and the Lerner Index (Beck et al., 2013). Since cooperative banks are mostly oriented toward lending activities (e.g., loans are 58% of total assets)14 rather than non-traditional intermediation activities, we are confident that their stability is strictly related to the quality of their loan portfolio. As such, we use the ratio of loan-loss provisions to total loans and loan-loss provisions to net interest margin to test for this relationship.
13 We would like to thank an anonymous referee for raising this issue and for suggesting several approaches to solving the problem. 14 Source: EACB (2012).
Table 2 Variables definition.
Variables Symbol Definition and calculation method
Z-score Z The ratio synthesizes a measure of overall banking risk. Similarly to Hesse and Cihák (2007), it is computed as the sum of the current period t return on assets (ROA) and the equity ratio (equity over total assets) divided by the standard deviation of ROA computed within each individual country (c) in year t.
Z-Rob Z_Rob Alternative stability measure computed as the sum of return on assets (ROA) calculated at the individual country (c) level in year t and the equity ratio divided by the standard deviation of ROA computed at the individual country (c) level in year t.
lerner index LER The Lerner Index represents the extent to which market power allows the bank to fix a price (P) above its marginal cost (MC).
Funding-adjusted LER FA_LER Lerner Index adjusted for the market power previously originated in the deposit market while raising funds.
Output price P Following recent studies (Berger et al., 2009; Turk Ariss, 2010) and assuming that banks produce an heterogeneous flow of services that is proportional to their dimension, we use banks’ total assets as a proxy of their overall activity (Angelini and Cetorelli, 2003), and we estimate average price as total revenues (interest and non-interest income) on total assets.
Marginal costs MC Marginal cost of the product as described in Section 3.2. Funding-adjusted MC FA_MC Marginal cost of the product as described in Appendix A. Herding measure HERD This is a measure of banking industry heterogeneity
computed, as in Beck et al. (2013), as the within-country standard deviation of the percentage non-interest income (with respect to total assets) per year (t) and per country (c).
Financial crisis FINCIR_LER Interaction terms computed as the product of the Lerner Indexes and a dummy variable for the 2007–2009 financial crisis. The categorical variable takes a value of 1 in 2007– 2009, and 0 otherwise. Both Lerner Indexes are used to build two different interaction terms.
Concentration loans HHI LOANS Concentration Index (Herfindahl–Hirschman Index) calculated as the sum of the squares of the market shares (considering loans) of each bank (i) in a specific country (c) in a determined year (t). We consider one observation per year (t) per country (c) (i.e., 60 values).
Concentration deposits HHI DEPOSIT Concentration Index (Herfindahl–Hirschman Index) calculated as the sum of the squares of the market shares (considering deposits) of each bank (i) in a specific country (c) in a determined year (t). We consider one observation per year (t) per country (c) (i.e., 60 values).
Herd Lerner HERD_LER Mixed measure that combines the banks with the highest tendency to herd (i.e., lowest third of the distribution of HERD) with market monopoly power.
Concentration Lerner loans HHIL_LER Mixed measure that combines banks’ concentration index in the loan market with market monopoly power.
Concentration Lerner deposits HHID_LER Mixed measure that combines banks’ concentration index in the deposit market with market monopoly power.
Bank-level controls Bank controls Liquidity ratio built as cash and due from other banks to total assets. Credit risk ratio built as loan loss provision to interest income margin. Credit orientation built as total loans to total assets.
Time-country dummies Time-countries Interaction terms computed as the product of year dummies and country dummies (i.e., 44 variables).
This table reports the name, symbol and definition of the variables employed in the analysis. The source of data is Bureau van Dijk Bankscope. The market concentration of loans and deposits is computed using data from both Bankscope and the European Central Bank.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–168
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–16 9
3.3. Other variables
We compute a herding measure and loan market concentration to control for the effects of other factors on the relationship between stability and competition. The herding measure, as in Beck et al. (2013), is the within-country standard deviation per year of non-interest income (i.e., fees and com- missions) as a share of total assets. It takes into consideration the possible incentives for banks to increase their risk-taking following an increase in competition. If the regulator finds it ex-post optimal to bail out a large number of banks when the number of bank failures is high (Acharya and Yorulmazer, 2007; Brown and Dinç, 2011), cooperative banks are more likely to expand their operations outside their core business in response to an increase in competition. The higher the value of this indicator, the lower is the herding behaviour in the cooperative banking sector. We also compute a combined measure using the interaction between the herding indicator and the Lerner Index. The Herd-Lerner is estimated as the product of a dummy variable and the Lerner Index. The dummy takes a value of one if the banking sector in a country is in the lowest third of the herding measure distribution (i.e., more homogeneous sources of revenues), and zero otherwise.
The Herfindahl–Hirschman Index (HHI) conveys information on the market concentration of loans and deposits. The index is computed annually at the country level because cooperative banks operate both at the regional level (e.g., big cooperatives in Germany and Austria) and the local level (e.g., small rural Italian cooperatives) and in our database it is not available detailed information on the geographical scope of the operations of each credit institution. The higher the value of HHI, the higher is the concentration of the market. We also calculate combined measures using both HHI and the Lerner Index. The HHI-Lerner for loans and the HHI-Lerner for deposits are computed as the product of a dummy and the two Lerner Index measures. The two dummies take a value of one if the banking sector in a country is in the highest third of the HHI distribution (i.e., more concentrated markets), and zero otherwise.
The 2007–2009 financial crisis could have affected the fundamental relationship holding between bank soundness and competition among cooperative banks. On the one hand, small rural cooperative banks might have been shielded from the market jitters in the interbank lending market. On the other hand, the economic crisis that followed might have had a profound impact on the fundamental eco- nomic relationship between cooperative banks and their members. We investigate this effect via an interaction term computed using the Lerner Index and a dummy variable for the 2007–2009 financial crisis.
We also include in the analysis two sets of control variables to limit the problem of spurious re- lationships. First, we consider bank-level fundamentals to account for liquidity risk, credit risk, and asset composition. Second, as in Beck et al. (2013), we control for the dynamic changes in country conditions using an interaction term between year dummies and country dummies (Table 2).
4. Empirical approach
To investigate the relationship between bank competition (measured using the Lerner Index) and stability (measured via the Z-score), we first employ the Granger causality technique. This approach has the advantage of permitting us to test unique time-ordered and signed relationships among pairs of variables.15 Although Granger causality tests have several limitations,16 this approach has been widely used in the economics (e.g., Jaeger and Paserman, 2008; Assenmacher-Wesche and Gerlach, 2008) and banking (e.g., Fiordelisi et al., 2011; Fiordelisi and Molyneux, 2010; Casu and Girardone, 2009; Williams, 2004) literature to analyse intertemporal relationships. Specifically, to disentangle
15 Granger’s (1969, p. 428) notion of causality states that “. yt is causing xt if we are better able to predict xt using all available information than if the information apart from yt had been used.” Granger’s suggestion to regress xt on its own lags and a set of lagged yt has become a standard procedure. If lagged yt provides a statistically significant explanation of xt, then yt “Granger causes” xt. 16 For instance, Granger-testing does not prove economic causation between two variables, but it identifies gross statistical associations.
Table 3 The link between bank stability and competition in cooperative banks: The Granger causality test.
Dependent variable: ln (Z-score)
(1) (2) (3) (4) (5) (6)
Zt�1 0.685*** (0.012)
0.679*** (0.012)
0.633*** (0.012)
0.630*** (0.012)
0.766*** (0.026)
0.767*** (0.026)
Zt�2 �0.014 (0.013)
�0.001 (0.013)
0.019 (0.012)
0.028** (0.012)
0.114*** (0.024)
0.112*** (0.024)
LERt�1 �0.008 (0.056)
0.033 (0.051)
�0.135*** (0.030)
LERt�2 �0.304*** (0.067)
�0.241*** (0.060)
0.083*** (0.025)
FA_LERt�1 0.168*** (0.057)
0.170*** (0.051)
�0.146*** (0.031)
FA_LERt�2 �0.370*** (0.064)
�0.281*** (0.056)
0.089*** (0.027)
HERDt�1 �2.106*** (0.111)
�2.152*** (0.111)
�5.325*** (0.469)
�5.359*** (0.470)
HHI LOANSt�1 9.701*** (1.318)
8.979*** (1.322)
18.301 (13.244)
19.331 (13.165)
HHI DEPOSITSt�1 �3.078*** (0.547)
�3.043*** (0.544)
�9.455 (6.557)
�10.054 (6.516)
Constant 1.149*** (0.035)
1.066*** (0.033)
1.286*** (0.034)
1.222*** (0.031)
0.872*** (0.059)
0.869*** (0.058)
LER (Total) �0.313*** (0.038)
�0.207*** (0.037)
�0.052*** (0.020)
FA_LER (Total) �0.203*** (0.031)
�0.111*** (0.031)
�0.058*** (0.020)
Observations 11,426 11,426 11,426 11,426 11,426 11,426 R-squared 0.382 0.383 0.404 0.404 0.863 0.863 TIME-COUNTRY NO NO NO NO YES YES BANK CONTROLS NO NO NO NO YES YES
Robust standard errors in parentheses: ***p < 0.01, **p < 0.05, *p < 0.1. This table reports the results of the Granger causality test performed using two different specifications of the Lerner Index. In columns (1), (3), and (5) we report the results for the traditional Lerner Index; columns (2), (4), and (6) summarise the results for the funding-adjusted measure. Notice that because we are using two lags in the variables of interest, we lose two time periods for each individual. Therefore, the number of observations drops from 17,074 to 11,426.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–1610
the intertemporal relationships between competition and stability, we first analyse the one-sided historical association between competition and bank stability via a simple causality regression using two lags in the variables of interest, with the Z-score as the dependent variable and the Lerner Index as the independent variable. The main quantity of interest is the sum of the coefficients of the two Z-score lags, computed as in Hesse and Cihák (2007). We assess the long-run effect of competition over financial stability by testing for the restriction that the sum of all lagged coefficients of the Z-score is zero: a rejection of the restriction suggests evidence that x has a long-run effect on y.17 To get more robust results, we use two specifications for the Lerner Index (traditional and funded-adjusted), andwe control for other exogenous factors that may affect the relationship, namely the homogeneity in bank investment activities (herding measure), the concentration in the deposits and loans markets, the dynamic trend of environmental conditions at the country level, and bank-level fundamentals.
Second, in line with previous studies (for instance, Beck et al., 2013), we analyse the economic causality using both pooled ordinary least squares and panel fixed-effects techniques. We specify the following relationship:
17 Following Casu and Girardone (2009), we also assess the Granger causality as the joint test of the null hypothesis that the two lags are equal to zero. If the probability is less than 10%, then the null hypothesis that x Granger-causes y is rejected at the 10% significance level. Results are available from the authors upon request.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–16 11
Zi;t ¼ conþ � u0 þ u1Xi;t�1
� � LERi;t�1 þ gXi;t�1 þ zKi;t�1 þ εi;t ; (7)
where the i subscript denotes the cross-sectional dimension across banks, and t denotes the time dimension. Zi,t is the Z-score for bank i at time t. LERi,t-1 is the Lerner Index for bank i lagged one year to minimize any simultaneity problems. Xi,t-1 is the one-year lagged factors that we expect to influence the relationship between competition and stability (i.e., herding measure, Herfindahl–Hirschman Index for deposits and loans, and the 2007–2009 financial crisis). Ki,t are the control variables (as detailed in Section 3.3), and u0 is the coefficient that summarises the strength and the sign of the relationship between bank soundness and competition. u1 is a vector of coefficients that captures the combined effects of competition with factors that influence the relationship between stability and competition. g and z are vectors of coefficients to be estimated, and εi,t indicates robust standard errors clustered at the individual bank level. The pooled ordinary least squares and the panel fixed-effects models allow us to further investigate the economic causality running from competition to bank stability.
5. Results
We first analyse the competition-stability nexus using Granger causality. The aim is to determine whether changes in market power precede financial troubles for European cooperative banks. We begin with a simple two-variable model that relates the two measures of market power (traditional Lerner and funding-adjusted Lerner) to bank soundness, computed as in Hesse and Cihák (2007). Our focus is on the significance and magnitude of the sum of the coefficients of the two lags in the two market power measures. The results in columns (1) and (2) of Table 3 indicate a negative association between market power and bank stability, meaning that a decrease in competition anticipates an increase in bank instability.18 This finding is not robust for specification bias and could also be related to spurious causality. To curb these potential issues, two additional sets of variables are included in the analysis. First, we consider factors that might affect the relationship such as the heterogeneity in banks’ revenues or the concentration in the loan and deposit markets. Second, we control for the bank-level fundamentals and for the dynamic changes in the environment where the banks operate. The results, reported in columns (3) to (6) of Table 3, are consistent with the previous estimation showing a negative statistically significant relationship between banks’ market power and bank stability.19 Note that the long-run effect is lessened by the spurious factors. In addition, the higher R2 in the regressions with the full set of control variables suggests, as noted in Beck et al. (2013), that environmental con- ditions might have a great impact on the cross-country variation in the relationship between bank market power and bank soundness.
We also analyse the relationship between competition and risk in a panel data setting.We introduce a new element in the analysis by investigating whether the 2007–2009 financial crisis has affected the European cooperative banks and the nexus under study. In addition, we run separate regressions with the twomeasures of market concentration, the herdingmeasure, and all the control variables together. Furthermore, to mitigate the omitted bias problem, in each regression we add bank-level information to control for other potential variables that drive individual bank stability. All the explanatory variables are lagged one year to minimise simultaneity problems.
In all the regressions, we find that bank market power is negatively related to bank stability, meaning that there is a positive relationship between competition and stability: when competition is low (i.e., market power is high), stability is low. This in turn provides evidence in favour of the competition-stability view proposed by Boyd and De Nicolò (2005). Moreover, to investigate if the
18 As noted in Berger (1995), it is possible to have a different sign in the lag coefficients without losing generality in the results. The only potential problem pertains to the serial correlation with lags more than two periods past. Nevertheless, we assume that two lags are sufficient to capture the long-run effect of competition on financial stability. 19 We also investigate a potential feedback relationship running from stability to competition. Results, available from the authors upon request, show that stability is not Granger-causing competition while controlling for spurious causality as the sum of the lagged Z-score is not significant at the 10% level.
Table 4 The link between bank stability and competition in cooperative banks.
Dependent variable
(1) (2) (3) (4) (5)
lnZ lnZ lnZ lnZ lnZ
LERt�1 �0.175*** (0.054) �0.230*** (0.056) �0.147** (0.062) �0.102* (0.058) �0.119*** (0.020) FINCIR_LER 0.573*** (0.022) �0.433*** (0.087) �0.277*** (0.029) HERDt�1 �2.374*** (0.160) �0.436*** (0.138) �1.034*** (0.231) �0.469* (0.249) �0.802*** (0.162) HERD_LERt�1 0.445*** (0.019) 0.124 (0.084) 0.321*** (0.024) HHI LOANSt�1 31.642*** (2.398) 23.003*** (3.655) 113.537*** (11.281) 110.473*** (13.827) 104.407*** (4.406) HHI DEPOSITSt�1 �10.406*** (1.324) �3.518* (2.035) �53.415*** (6.043) �52.106*** (7.654) �50.575*** (2.335) HHIL_LERt�1 �7.042 (6.332) 62.180*** (17.076) 17.724*** (4.755) HHID_LERt�1 �4.266 (3.432) �32.049*** (11.232) �7.354*** (2.411) Constant 2.844*** (0.050) 2.694*** (0.052) 2.607*** (0.081) 2.618*** (0.081) 3.687*** (0.050) Observations 14,070 14,070 14,070 14,070 14,070 R-squared 0.070 0.199 0.506 0.510 0.791 Number of clusters
2395 2395 2395 2395 2395
BANK FE NO NO NO NO YES TIME-COUNTRY NO NO YES YES YES BANK CONTROLS YES YES YES YES YES
Robust standard errors in parentheses: ***p < 0.01, **p < 0.05, *p < 0.1. This table summarises the results of the estimations. Coefficients in columns (1) and (2) are estimated using pooled ordinary least square regressions, whereas in the remaining regressions, a panel-fixed effects technique is used. In regressions (1–5), we control for the bank-level determinants described in Table 2. Additionally, we include individual bank fixed effects (BANK FE) and country-year dummies to account for the dynamic changes in the environmental conditions (TIME-COUNTRY). In re- gressions (1–5), the dependent variable is the natural logarithm of the Z-score built using a method similar to Hesse and Cihák (2007). Notice that because we are using one lag in the variables of interest, we lose one time period for each individual; therefore, the number of observations is reduced from 17,074 to 14,070.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–1612
results are driven by individual country conditions, we also run separate regressions at the country level and obtain qualitatively the same results.20 Our finding is consistent with the results of various studies on commercial banking (Beck et al., 2006; Schaeck et al., 2009, among the others) showing that European cooperative banks become more risky in less competitive markets. This evidence offers useful insights for policy makers undertaking the current redesign of the supervisory approach to European banks.
Turning to the effect of the financial crisis, we obtain contrasting results depending on the speci- fication. On the one hand, if we do not consider time-varying country effects (Table 4, column 2), higher market power during the period seems to induce higher stability. On the other hand, the coefficient on the financial crisis signals higher instability whenwe consider the full set of controls (Table 4, columns 4 and 5). Nevertheless, and for the aim of our analysis, the most important element is that the crisis seems not to have affected the fundamental relationship between competition and stability in coop- erative banking.
We also account for cooperative bank closure policies and an implicit Too-Many-To-Fail problem (Acharya and Yorulmazer, 2007). Specifically, we investigate the assumption that competitionwill have a stronger impact on bank stability in more homogeneous banking systems (where herding behaviour is more likely). As such, we introduce the herding measure and a combined measure obtained by interacting the Lerner Index with a dummy capturing the bank herding behaviour. As reported in Table 4, in all regressions the herding measure is negatively related to bank stability, meaning that coop- erative banks tend to become more stable in more homogenous banking markets. Moreover, the positive sign in the combined measure reinforces this presumption. These findings are particularly pertinent to policy makers. First, we show that the level of homogeneity influences the stability of the cooperative banking system. Second, we find support for the view that the expansion of cooperative
20 Results from this analysis are not reported in the paper and are available from the authors upon request.
Table 5 Test of robustness using alternative dependent variables.
Dependent variable
(1) (2) (3) (4) (5)
lnZ lnZ lnZ lnZ lnZ
LERt�1 �0.295*** (0.050) �0.172*** (0.049) �0.100*** (0.016) �0.004*** (0.001) �0.101** (0.043) FINCIR_LER 0.565*** (0.021) �0.401*** (0.080) �0.240*** (0.024) �0.003*** (0.000) �0.039*** (0.008) HERDt�1 �0.415*** (0.126) �0.545*** (0.198) �0.866*** (0.133) �0.016*** (0.003) �0.537*** (0.059) HERD_LERt�1 0.437*** (0.016) 0.089 (0.079) 0.282*** (0.019) 0.003*** (0.000) 0.061*** (0.010) HHI loanst�1 22.367*** (3.008) 111.866*** (12.696) 104.561*** (3.622) �0.103* (0.060) �2.954*** (1.128) HHI depositst�1 �3.355** (1.566) �52.559*** (6.676) �49.669*** (1.919) �0.011 (0.037) 0.530 (0.583) HHIL_LERt�1 �7.904 (5.638) 54.752*** (13.806) 14.883*** (3.909) 0.126 (0.087) 0.061 (1.823) HHID_LERt�1 �3.202 (2.773) �27.953*** (8.659) �7.447*** (1.982) �0.057* (0.029) �0.079 (0.912) Constant 2.754*** (0.048) 2.678*** (0.069) 3.648*** (0.041) 0.011*** (0.001) 0.249*** (0.020) Observations 14,070 14,070 14,070 14,070 14,070 R-squared 0.221 0.569 0.848 0.0281 0.0259 Number of clusters
2395 2395 2395 2395 2395
BANK FE NO NO YES NO NO TIME-COUNTRY NO YES YES NO NO BANK CONTROLS YES YES YES NO NO
Robust standard errors in parentheses: ***p < 0.01, **p < 0.05, *p < 0.1. This table summarises the results of the estimations using alternative dependent variables. Coefficients in columns (1), (4), and (5) are estimated using pooled ordinary least square regressions, whereas in the remaining regressions, a panel-fixed effects technique is used. In regressions (1–3), we control for the bank-level determinants described in Table 2. Additionally, in column (3) we include individual bank fixed effects (BANK FE) and in columns (2) and (3) country-year dummies to account for the dynamic changes in the environmental conditions (TIME-COUNTRY). In regressions (1–3) we employ as the dependent variable the natural logarithm of the Z-score built using a method similar to Yeyati and Micco (2007). In columns (4) and (5) we employ two different dependent variables to check for the robustness of our results. Namely, we use the ratio of loan-loss provisions to total loans (LLPTL) and the variable loan-loss provisions to net interest margin (LLPNIC). Notice that because we are using one lag in the variables of interest, we lose one time period for each individual; therefore, the number of observations is reduced from 17,074 to 14,070.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–16 13
banks into non-traditional business lines (i.e., non-interest income activities) could lead to higher insolvency risk.21 Although policymakers should carefully evaluate how diversification could affect the safety and soundness of the overall banking system, more homogenous cooperative banking markets seem to be more stable. The herding measure and the combined measure are strongly related to bank stability: the coefficients are statistically significant at least at the 10% level, and the signs of the co- efficients are stable in the majority of the regressions.
To further enrich our analysis, we consider the effect of concentration in the loan and deposits markets. The one-year lagged Herfindahl–Hirschman Index for loans is positively related to the Z-score, suggesting that bank stability is higher in more concentrated markets. Similarly, when it is statistically significant, the interaction term with the Lerner Index is positively and strongly related to bank sta- bility, meaning that the higher the market power associated with more concentrated markets, the more financially sound banks are. On the deposits side, we observe opposite results. Concentration in the deposits market is negatively related to individual bank soundness; we obtain the same result when we consider the combined effect of market power and market structure.
We use different specifications of the dependent variable to get more robust results. As reported in Table 5 in columns (1–3), we first employ as the dependent variable the natural logarithm of the Z- score built using a method similar to Yeyati and Micco (2007). The results remain qualitatively the same and this further reinforces our previous findings. In addition, the relationship between risk- taking and market power is also confirmed using credit risk measures (Table 5, columns 4 and 5).
21 See Mercieca et al. (2007) for the negative implications of diversification in the case of small European banks and Goddard et al. (2008a, b) for evidence from the U.S. market.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–1614
6. Conclusions
Cooperative banks are a driving force for socially committed business at the local level, accounting for around one fifth of the European banking system. Despite their importance, few if any studies have assessed the relationship between competition and financial stability in the cooperative banking sector. Our paper empirically fills this void examining a large sample of cooperative banks in the Eu- ropean Union between 1998 and 2009.
We show that bank market power negatively Granger-causes banks’ stability, meaning that there is a positive relationship between competition and soundness: when competition is low (i.e., market power is high), stability is low. The positive link between competition and stability is observed both in the short and long run. We also provide empirical evidence that bank soundness is higher in more homogenous markets where the herding behaviour is stronger. Cooperative bank closure policies may suffer from an implicit Too-Many-To-Fail problem, as suggested by Acharya and Yorulmazer (2007), when the banking system is more competitive because herding behav- iour is more likely. This result is particularly interesting for policy makers because it suggests that the level of industry homogeneity influences the stability of the cooperative banking system. We also show that the financial crisis of 2007 has not changed the relationship between competition and stability.
We consider a combination of measures to account for herding behaviour in the case of high mo- nopoly power (combination of herding and Lerner Index) and for concentration within the loan and deposit markets (combination of concentration of the loan market and the Lerner Index). We do not find evidence that the sign of the relationship between stability and market power in cooperative banking is affected by the introduction of these variables into the analysis.
Acknowledgements
The authors wish to thank Alessandro Carretta, Miguel Duran, Iftekhar Hasan, Phil Molyneux, and Ornella Ricci, who kindly provided comments on earlier versions of this paper, Karen DeVivo for professional and timely proofreading. The participants at the 2nd International Conference of the Financial Engineering and Banking Society (FEBS) and the 5th International Conference of the Inter- national Finance and Banking Society (IFABS) provided constructive comments. The authors are particularly grateful to Kees Koedijk (the editor) and to an anonymous referee for their constructive suggestions that have significantly improved the work. All remaining errors and omissions rest with the authors.
Appendix A
As a robustness test, we estimate an alternative measure of the marginal cost in the Lerner Index formula following some recent papers (Maudos and de Guevara, 2007; Turk Ariss, 2010). We compute the marginal cost using a translog cost function for each country separately with two inputs, one single output, and a time trend. The specification is as follows:
ln TCi;t ¼ a0 þ a1 ln Q þ a2 2 ln Q2 þ
X2 j¼1
bj ln Pj þ 1 2
X2 j¼1
X2 k¼1
djk ln Pj ln Pk
þ 1 2
X2 j¼1
gj ln Q ln Pj þ s1t þ s2 2 t2 þ s3t � ln Q þ
X2 j¼1
jjt ln Pj þ εit ;
(A.1)
where TCi,t is total costs (the sum of personnel expenses, other administrative expenses, and other operating expenses), and Q is the cooperative banks’ single output proxied by total assets. P1 and P2 are respectively the price of labour and the price of physical capital. t is a time trend capturing the dy- namics of the cost function over time. a, b, d, g, s, and j are coefficients to be estimated, and εit is a two- component error term. As in the estimation of equation (2), we apply the common restrictions of
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–16 15
standard symmetry and homogeneity of degree one in input prices defined as:P2 j¼1bj ¼ 1;
P2 j¼1gj ¼ 0; and ck˛f1;2g :
P2 j¼1dj;k ¼ 0.
From equation (A.1), the marginal costs are derived as follows:
MCi;t ¼ TCi;t Qi;t
2 4ba1 þ ba2 ln Q þ
X2 j¼1
bgj ln Pj þ bs3t 3 5; (A.2)
Using the funding-adjusted Lerner Index, we curb the potential issue of considering the market power that had previously originated in the deposit market while raising funds.
References
Acharya, V.V., Yorulmazer, T., 2007. Too many to fail: an analysis of time-inconsistency in bank closure policies. J. Financ. Intermediat. 16 (1), 1–31.
Allen, F., Gale, D., 2004. Competition and financial stability. J. Money, Credit Bank 36 (3), 453–480. Angelini, P., Cetorelli, N., 2003. The effects of regulatory reform on competition in the banking industry. J. Money, Credit Bank 35
(5), 663–684. Assenmacher-Wesche, K., Gerlach, S., 2008. Interpreting euro area inflation at high and low frequencies. Eur. Econ. Rev. 52 (6),
964–986. Ayadi, R., Llewellyn, D.T., Schmidt, R., Arbak, E., De Groen, W., 2010. Investigating Diversity in the Banking Sector in Europe: Key
Developments, Performance and Role of Cooperative Banks. Centre for European Policy Studies, Brussels. Beck, T., De Jonghe, O., Schepens, G., 2013. Bank competition and stability: cross-country heterogeneity. J. Financ. Intermediat.
22 (2), 218–244. Beck, T., Demirgüç-Kunt, A., Levine, R., 2006. Bank concentration, competition, and crises: first results. J. Bank. Financ. 30 (5),
1581–1603. Beck, T., Hesse, H., Kick, T., VonWesternhagen, N., 2009. Bank Ownership and Stability: Evidence from Germany. In: Bundesbank
Working Paper Series. The Deutsche Bundesbank, Frankfurt. Berger, A.N., 1995. The relationship between capital and earnings in banking. J. Money, Credit Bank 27 (2), 432–456. Berger, A.N., Klapper, L.F., Turk Ariss, R., 2009. Bank competition and financial stability. J. Financ. Serv. Res. 35 (2), 99–118. Boonstra, W.W., Mooij, J., 2012. Raiffeisen’s Footprint. The Cooperative Way of Banking. VU University Press, Amsterdam. Boot, A.W.A., Thakor, A.V., 2000. Can relationship banking survive competition? J. Financ. 55 (2), 679–713. Boyd, J.H., De Nicolò, G., 2005. The theory of bank risk taking and competition revisited. J. Financ. 60 (3), 1329–1343. Boyd, J.H., De Nicolò, G., Al Jalal, A., 2006. Bank Risk-taking and Competition Revisited: New Theory and New Evidence. IMF
Working Papers 297. International Monetary Fund, Washington. D.C. Brown, C.O., Dinç, I.S., 2011. Too many to fail? Evidence of regulatory forbearance when the banking sector is weak. Rev. Financ.
Stud. 24 (4), 1378–1405. Brunner, A., Decressin, J., Hardy, D., Kudela, B., 2004. Germany’s Three-Pillar Banking System–Cross-country Perspectives in
Europe. IMF Occasional Paper 233. International Monetary Fund, Washington. D.C. Casu, B., Girardone, C., 2009. Testing the relationship between competition and efficiency in banking: a panel data analysis.
Econ. Lett. 105 (1), 134–137. Delis, M.D., Kouretas, G.P., 2011. Interest rates and bank risk-taking. J. Bank. Financ. 35 (4), 840–855. De Jonghe, O., 2010. Back to the basics in banking? A micro-analysis of banking system stability. J. Financ. Intermediat. 19 (3),
387–417. De Nicoló, G., Loukoianova, E., 2007. Bank Ownership, Market Structure and Risk. IMF Working Paper 215. International
Monetary Fund, Washington, D.C. De Nicolò, G., Lucchetta, M., 2009. Financial Intermediation, Competition, and Risk: a General Equilibrium Exposition. IMF
Working Paper 105. International Monetary Fund, Washington, D.C. European Association of Co-operative Banks, 2012. Annual Report for 2011. http://www.eurocoopbanks.coop. Fiordelisi, F., Mare, D.S., 2013. Probability of default and efficiency in cooperative banking. J. Intern. Financ. Mark. Inst. Money 26,
30–45. Fiordelisi, F., Marques-Ibanez, D., Molyneux, P., 2011. Efficiency and risk-taking in European banking. J. Bank. Financ. 35 (5),
1315–1326. Fiordelisi, F., Molyneux, P., 2010. Total factor productivity and shareholder return in banking. Omega 38 (5), 241–253. Fonteyne, W., 2007. Cooperative Banks in Europe: Policy Issues. IMF Working Paper 159. International Monetary Fund,
Washington, D.C. Girardone, C., Nankervis, J.C., Velentza, E., 2009. Efficiency, ownership and financial structure in European banking. Manag.
Financ. 35 (3), 227–245. Goddard, J., McKillop, D., Wilson, J.O., 2008a. What drives the performance of cooperative financial institutions? Evidence for US
credit unions. Appl. Financ. Econ. 18 (11), 879–893. Goddard, J., McKillop, D., Wilson, J.O.S., 2008b. The diversification and financial performance of US credit unions. J. Bank. Financ.
32 (9), 1836–1849. Goodhart, C.A.E., 2004. Some New Directions for Financial Stability. The Per Jacobsson Lecture. Bank for International Settle-
ments, Basel. Gorton, G., Schmid, F., 1999. Corporate governance, ownership dispersion and efficiency: empirical evidence from Austrian
cooperative banking. J. Corp. Financ. 5 (2), 119–140.
F. Fiordelisi, D.S. Mare / Journal of International Money and Finance 45 (2014) 1–1616
Granger, C.W.J., 1969. Investigating causal relations by econometric models and cross-spectral methods. Econometrica 37 (3), 424–438.
Hakenes, H., Schnabel, I., 2010. Credit risk transfer and bank competition. J. Financ. Intermediat. 19 (3), 308–332. Hesse, H., Cihák, M., 2007. Cooperative Banks and Financial Stability. IMF Working Papers 2. International Monetary Fund,
Washington. D.C. Iannotta, G., Nocera, G., Sironi, A., 2007. Ownership structure, risk and performance in the European banking industry. J. Bank.
Financ. 31 (7), 2127–2149. Jaeger, D.A., Paserman, M.D., 2008. The cycle of violence? An empirical analysis of fatalities in the Palestinian-Israeli conflict.
Am. Econ. Rev. 98, 1591–1604. Jimenez, G., Lopez, J.A., Saurina, J., 2010. How Does Competition Impact Bank Risk-taking?. Banco de Espana Working Paper
1005 Bank of Spain, Madrid. Keeley, M.C., 1990. Deposit insurance, risk, and market power in banking. Amer. Econ. Rev., 1183–1200. Kontolaimou, A., Tsekouras, K., 2010. Are cooperatives the weakest link in European banking? A non-parametric metafrontier
approach. J. Bank. Financ. 34 (8), 1946–1957. Laeven, L., Levine, R., 2009. Bank governance, regulation and risk taking. J. Financ. Econ. 93 (2), 259–275. Lepetit, L., Nys, E., Rous, P., Tarazi, A., 2008. Bank income structure and risk: an empirical analysis of European banks. J. Bank.
Financ. 32 (8), 1452–1467. Lepetit, L., Strobel, F., 2013. Bank insolvency risk and time-varying Z-score measures. J. Intern. Financ. Mark. Inst. Money 25, 73–
87. Liu, H., Molyneux, P., Wilson, J.O.S., 2012. Competition and Stability in European Banking: a Regional Analysis. The Manchester
School. Marcus, A.J., 1984. Deregulation and bank financial policy. J. Bank. Financ. 8 (4), 557–565. Matsuoka, T., 2013. Sunspot bank runs in competitive versus monopolistic banking systems. Econ. Lett. 118 (2), 247–249. Maudos, J., Fernandez de Guevara, J., 2007. The cost of market power in banking: social welfare loss vs. cost inefficiency. J. Bank.
Financ. 31 (7), 2103–2125. McKillop, D., Wilson, J.O.S., 2011. Credit unions: a theoretical and empirical Overview. Financ. Mark. Instit. Instr. 20 (3), 79–123. Mercieca, S., Schaeck, K., Wolfe, S., 2007. Small European banks: benefits from diversification? J. Bank. Financ. 31 (7), 1975–1998. Radi�c, N., Fiordelisi, F., Girardone, C., 2012. Efficiency and risk-taking in pre-crisis investment banks. J. Financ. Serv. Res. 41 (1–2),
81–101. Repullo, R., 2004. Capital requirements, market power, and risk-taking in banking. J. Financ. Intermediat. 13 (2), 156–182. Rey, P., Tirole, J., 2007. Financing and access in cooperatives. Intern. J. Ind. Organ. 25 (5), 1061–1088. Schaeck, K., Cihák, M., Wolfe, S., 2009. Are competitive banking systems more stable? J. Money, Credit Bank 41 (4), 711–734. Turk Ariss, R., 2010. On the implications of market power in banking: evidence from developing countries. J. Bank. Financ. 34
(4), 765–775. Vallascas, F., Keasey, K., 2012. Bank resilience to systemic shocks and the stability of banking systems: small is beautiful. J.
Intern. Money & Financ. 31 (6), 1745–1776. Williams, J., 2004. Determining management behaviour in European banking. J. Bank. Financ. 28 (10), 2427–2460. Yeyati, E.L., Micco, A., 2007. Concentration and foreign penetration in Latin American banking sectors: impact on competition
and risk. J. Bank. Financ. 31, 1633–1647.
- Competition and financial stability in European cooperative banks
- 1 Introduction
- 2 Literature review and research hypotheses
- 3 Data sources and variables
- 3.1 Measuring competition: the Lerner Index
- 3.2 Measuring bank stability
- 3.3 Other variables
- 4 Empirical approach
- 5 Results
- 6 Conclusions
- Acknowledgements
- Appendix A
- References
FINANCIAL-STABILITY-COMPETITION-AND-EFFICIENCY-IN-LATIN-AMERICAN-AND-CARIBBEAN-BANKING_2014_Journal-of-Applied-Economics.pdf
* Adnan Kasman (corresponding author): Department of Economics, Faculty of Business, Dokuz Eylul University, Izmir, Turkey; [email protected]. Oscar Carvallo: Economic Research Office, Central Bank of Venezuela Caracas, Venezuela; [email protected]. Valuable comments of two anonymous referees and the editor are greatly appreciated. However, any remaining errors are ours.
Journal of Applied Economics. Vol XVII, No. 2 (November 2014), 301-324
FINANCIAL STABILITY, COMPETITION AND EFFICIENCY IN LATIN AMERICAN AND CARIBBEAN BANKING
AdnAn KAsmAn* Dokuz Eylul University
OscAr cArvAllO
Central Bank of Venezuela
Submitted May 2013; accepted March 2014
Using a sample of 272 commercial banks from fifteen Latin American countries for the period 2001-2008, we estimate cost and revenue efficiency scores, financial stability scores (Z-scores) and competition scores (Lerner indexes and Boone indicators) at the bank level. The Granger causality technique in dynamic panels is used to establish dynamic relationships among these variables. We find evidence that strongly supports the “quite life” hypothesis, while we also find partial support for causality running in the opposite direction. Moreover, the results suggest that more competition is conducive to greater financial stability (when the revenue efficiency score is used). Banks seem to achieve market power through better efficiency, leverage and earning ability. As size and complexity increase, however, agency problems and increasing risk-taking might start gaining momentum, generating inefficiency and fragility.
JEL classification codes: G21, D24, C23 Key words: financial stability, competition, efficiency, Latin American banking
302 Journal of applied economics
I. Introduction
The global financial crisis of 2007 has not only shaken most of financial markets
and institutions, but also key underlying assumptions regarding financial market
mechanisms. During the run-up to the crisis, it became apparent that bubbles
could survive for long periods of time despite the presence of well-informed and
well-financed rational arbitrageurs. Moreover, the “efficient markets hypothesis”
would rule out such a phenomenon (Abreu and Brunnermeier 2003; Brunnermeier
et al. 2009). For a long time, competition was not thought to be an important
determinant of financial stability. In fact, in most countries standard competition
policy was not applied fully to financial markets. Although competition policy
had been substantially strengthened at the national level over the last couple of
decades, regulators in most countries were complacent about market power in
national markets. The crisis also changed that. Particularly, the largest and more
interconnected financial institutions in developed markets have failed.
The Latin American and Caribbean banking systems have shown an apparent
resilience during the current crisis. The region’s financial systems have undergone
intense structural change since the 1990s. In addition, starting in the nineties,
consolidation and restructuring have changed the competitive environment (Levy-
Yeyati and Micco 2007; Yildirim and Philippatos 2006; Carvallo and Kasman
2005; BIS 2007). However, there are concerns regarding the impact of increased
concentration on the level of competition, performance and the financial stability
of the banking systems in the region.
As the experience of the crisis underscores, a better understanding of links
between stability and competition is a must. Moreover, the studies so far have
focused on structural and aggregated measures of competition, concentration,
stability and efficiency, mostly in a static setting. In the related literature, studies
have paid attention to performance (cost or profit efficiencies) or competition
(and/or risk) of banks. In this paper, we integrate these dimensions in search of
mutual dynamic relationships between bank efficiency, competition and financial
stability in the banking sectors within the region. The sweeping changes in the
global regulatory paradigm will affect the way the region regulates its financial
markets. Competition policy will be part of those changes and solid empirical
evidence to guide it is required.
This paper extends the related literature in several ways. First, we estimate
market power at the bank level by computing Lerner indexes for an unbalanced
financial stability, competition and efficiency in banking 303
sample panel of 272 Latin American and Caribbean banks from fifteen countries
in the period 2001-2008. We also estimate cost and revenue efficiency of banks
operating in the region during the period by taking into account legal and economic
environmental variables. Following Fiordelisi et al. (2011) and Casu and Girardone
(2009), this paper applies a relatively new empirical methodology, Granger-
causality tests in dynamic panels. Regressions using pooled OLS and panel data
models might present endogeneity problems, since the lagged dependent variable
is often correlated with the disturbance term. Thus, we present several regression
estimation methods. In doing so, we test two prevailing hypotheses regarding
stability and competition, namely, the “competition-fragility” and “competition-
stability” hypotheses. Lastly, we investigate the dynamic feedbacks between
efficiency and market power. In this regard, two different views are contrasted:
the so called “structure-conduct-performance” and the “efficient structure”
hypotheses.
The rest of the paper is organized as follows. In Section II, we present a review of
the literature regarding market power, financial stability and efficiency in banking.
In Section III, we discuss the methodology and the econometric specification used
for the estimation of competition indexes, efficiency scores, financial stability and
causality tests. The data and empirical results of the estimations are reported in
Section IV. The paper’s concluding remarks are provided in Section V.
II. Related literature, empirical evidence and research hypotheses
Numerous studies have looked at the determinants of market power in banking
(Berger and Hannan 1989; Claessens and Laeven 2004; Maudos and Nagore
2005). For instance, Claessens and Laeven (2004) use the Panzar and Rosse (1987)
methodology to check for the relevance of indicators of countries’ banking structure
and regulation. They find that contestability along with activity restrictions, rather
than concentration, is the main determinant of market power. Foreign entry and
fewer activity restrictions enhance the competitiveness of banking systems. Maudos
and Nagore (2005) construct bank-level measures of market power (Lerner indexes).
They find an inverted U-shaped relationship between bank size and market power.
More efficient banks enjoy greater market power, as banks seem to pass on to
customer their cost advantages. They also find a positive effect of concentration on
market power, in line with the conventional view.
304 Journal of applied economics
The relationship between competition and financial stability has been a controversial issue long before the current crisis started. Both at theoretical and empirical level, the issue remain ambiguous and unresolved, despite a large body of literature. The most traditional view asserts that competition increases risk - taking as it undermines the “charter value” of banks. Alternatively, banks that enjoy market power can extract monopoly rents that enhance charter value. As increased risk-taking endanger that value, banks have less incentive to engage in it (Keeley 1990). Also, larger banks are supposed to diversify better and be more efficient due to the economies of scale and scope (Boyd and Pescott 1986). However, the traditional “too big to fail” doctrine implies increased risk-taking by the largest banks, which enjoy implicit public guarantees (Mishkin 1999).
The controversy has spilled over to more recent contributions. Larger banks in concentrated systems can easily create capital buffers against macroeconomic and liquidity shocks, improving systemic stability (Boyd et al. 2004). In addition to traditional scale and scope diversification, larger banks can spread their operation geographically to reduce risk (Meon and Weill 2005). The “charter value hypothesis” has been revived on the grounds of increased profit buffers (Allen and Gale 2004; Matutes and Vives 2000; Cordella and Levy-Yeyati 2003). Larger banks engaging in “credit rationing” enhance soundness as they pick fewer investments with higher quality (Boot and Thakor 2000). Based on moral hazard arguments, it can be argued that rates charged by banks with market power can induce increased risk-taking by entrepreneurs (Boyd and De Nicolo 2005). Also, the effect of increased scope and scale economies can be counterbalanced by increased managerial inefficiency (“X-inefficiency”) due to increased complexity (Beck et al. 2006). The empirical evidence reflects these conflicting views. Whereas recent evidence support the “competition - fragility” view (Beck et al. 2006), several studies provide evidence supporting the “competition - stability” hypothesis (De Nicolo et al. 2004; Schaeck and Cihak 2007; Schaeck et al. 2006; Uhde and Heimeshoff 2009). Adding to that, the bulk of the empirical literature uses market concentration measures, which might not necessarily reflect effective competition. Accordingly, what is found to be positively correlated with financial
stability is contestability, rather than the actual level of foreign presence or market
concentration (Barth et al. 2004). 1
1 One of the factors that complicate this picture is the complex nature of the relation between concentration and effective competition. Either on a theoretical or empirical basis, it is becoming clear that they measure different dimensions of the competitive landscape (Matutes and Vives 1996).
financial stability, competition and efficiency in banking 305
Williams (2012) examines the relationship between bank efficiency and market
power to test the so called “quiet life” hypothesis in Latin American banking for
the period 1985-2010.2 He uses an efficiency-adjusted Lerner index, proposed by
Koetter et al. (2012), in addition to the conventional one, to measure market power.
He also produces conventional Lerner indexes in deposits and loans markets. Their
results show that deposit markets are more competitive than lending markets.
Chortareas et al. (2011), like Williams (2012), examine the relationship between
bank efficiency and market power for Latin American banks for the period 1997-
2005. In contrast to Williams (2012) and the present study, they use the Herfindahl-
Hirschman Index (HHI) to measure market power. Their results uncover evidence
supporting the efficient structure hypothesis in Latin American countries. 3
Regarding financial fragility, Berger et al. (2009) examine the competition -
fragility relationship by examining different dimensions of risk. Although banks
with higher market power are found to have overall less risk exposure - consistent
with “competition - fragility” view, banks tend to offset loan risk by increasing
capital ratios. In line with this idea, Maudos and Nagore (2005) examine the
determinants of stability as measured by the Z-scores. They find support for the
“competition - fragility” view when using Lerner indexes as measures of market
power. More recent studies examine the issue in a dynamic context. Uhde and
Heimeshoff (2009) compute Z-scores and concentration indexes in a dynamic
panel context to examine the inter-temporal relationship between consolidation
in European banking and financial stability. They find that banks are more prone
to financial fragility in the Eastern European banking markets exhibiting a lower
level of competitive pressure, fewer diversification opportunities and a higher
fraction of government-owned banks.
Claessens et al. (2001) present evidence with regard to foreign entry, asserting
that it increases bank efficiency by reducing margins. Demirguc-Kunt et al. (2004)
2 The “quiet life” hypothesis was first proposed by Hicks (1935). According to this hypothesis, banks with more market power put less effort in pursuing cost efficiency as the pressure to increase efficiency is absent. Hence, instead of taking advantage of their favorable position by cutting costs, they prefer to enjoy a “quite life”. 3 Several papers in the related literature examine the relationship between market power and concentration. For instance, Bikker and Haaf (2002) provide support for the conventional view that concentration reduces competitiveness. Claessens and Laeven (2004), however, find that competition is not negatively related to market concentration. With respect to Latin American and Caribbean banks, Yildirim and Philippatos (2006) find that regardless of increased concentration and foreign entry, competition as a rule has not been affected in the region. Banks seem to be operating under monopolistic competition conditions.
306 Journal of applied economics
find that concentration has a negative effect on efficiency in less developed banking
systems. Schaeck and Cihak (2010) find that competition enhances efficiency.
In this regard, Zarutskie (2009) and Dick and Lehnert (2010) provide evidence
that competition affect banks’ efficiency by improving specialization, screening,
monitoring and lending ability. More recently, Fiordelisi et al. (2011) use Granger-
causality technique and find that lower cost and revenue efficiency cause higher
bank risk. Using the same methodology, Casu and Girardone (2009) investigate
the relationship between competition and efficiency for a sample of European
banks. They find a positive relationship between market power and efficiency, and
a weak relation running from efficiency to competition.
It is also becoming increasingly clear that efficiency is an important factor
with regard to market power and risk. Different efficiency levels will affect equity,
leverage and risk decisions taken by banks, under regulatory pressure or because
of agency considerations (Hughes and Mester 1998). Also, according to the “quiet
life” hypothesis, firms enjoying market power tend to operate inefficiently rather
than to reap all potential rents (Nickell et al. 1997). Theoretically, then, increased
competition should induce firms’ efficiency. Even though empirical studies have
found that deregulation tends to improve efficiency and performance, few studies have explored the simultaneous relationship (Wilson 1994; Claessens and Laeven 2004; Bikker and Spierdijk 2008). The “quiet life” hypothesis is in the tradition of the “Structure-Conduct-Performance” (SCP) paradigm. The SCP hypothesis argues that market power and limits to competition create an environment that affects bank conduct and performance in socially unfavorable ways. A competing view is the “Efficient Structure” (ES) hypothesis, according to which an industry’s structure arises because of superior operating efficiency by particular firms.4 Carvallo and Kasman (2005) and Kasman, et al. (2005) estimate common stochastic cost and profit frontiers for sixteen Latin American and Caribbean banks countries for the period 1995-1999. They find that concentration is significant and positively related to the cost and profit inefficiency. They also find that foreign banks are more cost and profit efficient in twelve of the sixteen countries in their sample.
4 Berger et al. (2004) provide a complete survey of the “structure-conduct-performance” and “efficient structure” literature.
financial stability, competition and efficiency in banking 307
By using Granger-causality techniques, this paper tries to integrate competing views regarding the relationships among financial stability, efficiency and market power. In particular, we investigate the “competition-stability” versus “competition fragility” controversy, together with the “Structure-Conduct-Performance” versus “Efficient Structure” one. Under the ES hypothesis, we can expect a positive dynamic relationship between our measures of efficiency and financial soundness, and market power. However, under the SCP hypothesis, a negative relationship is expected, which is running from market power to efficiency and soundness. Finally, under the “competition-stability” hypothesis, a dynamic negative relation is expected between our measure of market power and financial soundness. Under the alternative “competition-fragility” view, however, a positive sign is to be
expected.
III. Methodology
A. Efficiency estimations
We estimate cost and revenue efficiencies employing the stochastic frontier model of Battese and Coelli (1995), in which the inefficiency term is drawn from a truncated normal distribution. One of the main features of this model is that it allows controlling for environmental differences across countries and analyzes the effects of these variables on estimated efficiency scores. Moreover, this model allows for a firm-specific and time-varying intercept shift in the distribution of the inefficiency term, and this intercept shift is itself a function of the exogenous environmental variables that vary across countries.
Assuming that costs, for bank i at time t, are function of output Q, input prices W, inefficiency u, and random error v, then cost function can be specified as follows:
, (1)
where C denotes costs if and transformed total revenues if
are independently and identically distributed random errors that are
independently distributed of the are independently distributed, such that
is obtained by truncation (at zero) of the normal distribution with mean
, and variance , that is ; is a (1× m) vector of country-
308 Journal of applied economics
specific environmental variables (GDP growth, GDP per capita, deposit density,
population density, the ratio of average equity to total assets, capitalization of
stock market, inflation, M2 for a proxy of financial deepening) that are allowed to
vary over time.
The inefficiency effects, , in (1) can be specified as follows:
, (2)
where is an (m×1) vector of unknown coefficients of the environmental variables;
is defined by the truncation of the normal distribution, such that the point of
truncation is .5 Battese and Coelli (1995) show that when (1) is assumed, the
cost efficiency (or the alternative revenue efficiency) for an individual banking
firm can be defined as follows: 6
, (3)
In modeling the cost (or revenue) function, we adopt a translog functional form since it does not require too many restrictive assumptions about the nature of the technology. A one output and three input cost (or revenue) function is specified
as follows:
5 The model without country-specific environmental variables has some limitations (see for example Dietsch and Lozano-Vivas 2000; Lozano-Vivas et al. 2001; Kasman and Yildirim 2006). The main limitation is that the model is based on the assumption that cross-country efficiency differences are mainly attributable to managerial decisions within the banks. However, different economic and regulatory environments across countries can also explain the differences. 6 Both cost and revenue efficiency measure departure from optimal behavior, but the standard against which their performance is evaluated is different. In the case of revenue efficiency the standard is the revenue frontier. In the case of cost efficiency, the standard is the cost frontier. The failure to maximize revenue stems from either an inability to technically achieve the efficient output levels or to allocate an appropriate output mix, given output prices and the input vector employed. In the case of cost efficiency, the inability relates to sub optimal input utilization and mix. Thus, the efficiency aspect being measured is different.
(4),
financial stability, competition and efficiency in banking 309
where TC ( TR ) is total costs (total revenues) of the banking firm in a given year. Q is the output (total assets), are the input prices (borrowed funds, labor and capital) and TREND is a time trend included to take into account technical change.7 The and are the inefficiency and the error terms, respectively.8 To ensure that the estimated cost frontier is well-behaved, standard restrictions of linear homogeneity in input prices and symmetry of the second order parameters
are imposed.
B. Competition measures estimations
The Lerner index
The Lerner index is a good candidate for a measure of market power that varies at the bank level rather than a concentration proxy at the country level. It can be defined as the difference between the marginal price and marginal cost divided by the marginal price, as follows: 9
, (5)
where is proxied by the ratio of total revenues (interest and non-interest revenues) to total assets. The marginal cost, , is derive from the translog cost function defined in (4). Marginal cost is obtained as follows:
7 The price of labor is calculated as the ratio between personnel expenses and total assets. The price of capital is given by operating costs net of personnel expenses over fixed assets. Finally, the price of funds is calculated by dividing total interest expenses by total purchased funds. Both financial and operating costs are included in the estimation of the cost function. Moreover, we estimate the alternative revenue efficiency scores. In this approach, banks take input and output quantities as given and these measures were selected since prices are often inaccurately measured in banking. The revenue inefficiency scores are estimated using the translog functional model specified in (4). Hence, the bank output and input definitions used is the same as to estimate the cost inefficiency scores. 8 The composite error term is for the revenue function. 9 Koetter et al. (2012) propose an efficiency-adjusted Lerner index and point out that the conventional approach of computing the Lerner index assumes both profit efficiency and cost efficiency. Hence, the estimated price-cost margins could be biased in measuring the true extent of market power.
(4)
310 Journal of applied economics
The Boone indicator
The second measure of competition used in this paper is the Boone indicator.
Boone’s model is based on the idea that competition enhances the performance
of efficient banks and weakens the less efficient ones. This effect is stronger the
higher the competition in the market is. To support this quite intuitive market
characteristic, Boone (2001, 2008) develops a broad set of theoretical models and
proves that more efficient banks (i.e., banks with lower marginal costs) gain higher
market shares. The Boone indicator is estimated by using the following empirical
model:
, (7)
where ms and mc denote the market shares and marginal costs in the loans market,
respectively. In this paper, we also measure the evolution of competition. Hence,
we include time dummies, D, to control factors common to all banks in the market
and specific to each year. The coefficient denotes the Boone indicator. Market
shares increase for banks with lower marginal costs (i.e, ). Hence, an increase
in competition raises the market share of a more efficient bank relative to a less
efficient one. A larger negative value of is an indication of more competitive
conditions in the banking market. However, positive values of are also possible,
implying that the higher a bank’s marginal costs, the more market share it will
earn. In the case of positive , either the market has an extreme level of collusion
or the banks are competing on quality.
C. Financial stability estimation
The Z-score is one of the most commonly used indicators of financial stability,
which measures the distance from insolvency and is calculated as follows:
, (8)
financial stability, competition and efficiency in banking 311
where ROA is the return on assets, EQ/TA denotes the equity to asset ratio and
is the standard deviation of return on assets in the period of time analyzed.
The Z-score increases with profitability and solvency and decreases as the standard
deviation of the return increases. A higher Z-score implies a lower probability of
insolvency (failure), providing a more direct measure of soundness compared other
measures of risk. Using (8) we can generate Z-scores, as an indicator of financial
stability, for each bank and year. Although the expression in the denominator is
constant during the sample period, the expression in the numerator varies every
year.10
D. Testing the relationship among competition, performance and financial stability
We examine the link among competition, performance and financial stability in
the Latin American banking markets in a dynamic Granger-causality framework,
formally specified in (9), (10), and (11) as follows:
, (9)
, (10)
, (11)
where competition is either the Lerner index or the Boone indicator, performance
is given by cost efficiency scores or revenue efficiency scores, and stability by the
10 Due to the limited number of years in our sample data, we could not calculate the standard deviation of the returns using a three-year (or a four-year) rolling time window to allow for variation in the denominator of (8).
312 Journal of applied economics
Z-scores. , and are the random error terms. Equation (9) tests
whether performance (cost or revenue efficiency) and stability (Z-scores) changes
temporally lead to variations in competition (Lerner index or Boone indicator).
Equation (10) examines whether changes in competition and stability temporally
lead to variations in bank performance. Finally, equation (11) tests whether changes
in competition and performance temporally lead to variation in bank stability.
The Granger causality test results are sensitive to the choice of lag length. In
estimating (9) - (11) with OLS, the optimal lag length is determined based on the
Schwarz Information Criterion. The optimal lag length is two, based on this criterion.
As for the system GMM specification, following Casu and Girardone (2009) and
Fiordelisi et al. (2011), we use two lags and estimate an AR(2) process as in the OLS
case. Hence, Granger causality assesses whether the coefficients on the two lags are
jointly statistically different from zero. The “long-run effect” of each variable is also
tested using the restriction that the sum of the lags of each determinant variable is
zero; a rejection of which signifies evidence of a long run effect. 11
IV. Data and empirical results
The sample in this study includes the commercial banks from 15 Latin American and Caribbean countries over the period 2001-2008. The countries included (number of banks in parentheses) are Argentina (50), Bolivia (8), Brazil (59), Colombia (12), Costa Rica (13), Dominican Republic (17), Ecuador (19), El Salvador (7), Honduras (12), Jamaica (5), Panama (19), Paraguay (11), Peru (10), Trinidad and Tobago (6) and Venezuela (24). Bank level data for all countries in the sample were obtained from the Bankscope database and macroeconomic variables were obtained from the World Development Indicators and International Financial Statistics of the IMF. After reviewing the data for reporting errors, inconsistencies,
11 The introduction of a lagged dependent variable among the right hand side variables in (9)-(11) creates an endogeneity problem since the lagged dependent variable is correlated with the disturbance, . To solve this problem, Arellano and Bond (1991) developed a difference GMM estimator for the coefficients in above mentioned equations where the lagged levels of the regressors are the instruments for the equation in first differences. However, Arellano and Bover (1995) and Blundell and Bond (1998) suggest to difference the instruments instead of the regressors in order to make them exogenous to the fixed effects. This leads from the difference GMM to the system GMM estimator, which is a joint estimation of the equation in levels and in first differences. Hence, we use the two-step system GMM estimators with Windmeijer (2005) corrected standard error, along with the OLS and Fixed Effects estimators, to conduct our analysis.
financial stability, competition and efficiency in banking 313
missing values and outliers, an unbalanced panel of 1828 observations is used, which includes 272 commercial banks over the period 2001-2008. 12
To identify the common frontier for the estimation of cost and revenue effi- ciency scores, we chose several geographic, market structure, and financial depth variables which explain the peculiar features of each country’s banking sector. Av- erages of these variables are reported in Table 1. As in Dietsch and Lozano-Vivas (2000), these variables are categorized in three groups. The first group includes measures of density of population, income per capita, and density of demand for each country. The second group includes a concentration ratio and the average capital ratio. A final group includes a proxy for the accessibility of banking servic- es and other environmental variables that are relevant to determine bank efficiency.
Table 1. Average values of environmental variables (2001-2008)
POP INC ($) DEMAND ($) HHI AEQ (%) INT (%) GDPG (%)
MONEY (%)
INF (%)
Argentina 13.87 8021.27 12715.50 894.89 12.12 8.66 4.45 27.54 10.08
Bolivia 8.36 1079.09 2830.71 1698.53 10.00 5.11 3.90 51.11 5.33
Brazil 21.61 3986.86 30817.25 1006.49 9.31 16.86 3.65 49.88 7.12
Colombia 37.51 2786.39 28048.63 1505.24 12.30 7.30 4.31 19.77 6.12
Costa Rica 83.80 4529.47 184342.13 1689.66 13.46 25.40 4.96 22.12 11.28
Dominican Republic 194.18 3062.38 105870.12 1533.70 12.12 16.99 5.29 21.86 15.20
Ecuador 45.79 1544.36 27296.63 1890.99 11.16 4.55 5.28 22.14 9.62
El Salvador 287.55 2429.11 217390.13 2520.27 11.59 4.84 2.81 40.79 4.03
Honduras 56.35 1280.80 21625.194 1503.55 0.124 7.80 5.02 44.92 8.24
Jamaica 182.33 3711.11 342063.13 2116.44 11.93 11.40 1.67 44.32 11.65
Panama 39.12 4510.14 294822.11 1138.51 11.37 3.44 6.64 78.80 2.51
Paraguay 14.38 1375.93 3895.01 1294.22 11.19 7.51 3.72 32.30 8.88
Peru 21.51 2370.54 29824.75 2750.77 10.18 5.11 5.93 28.98 2.41
Trinidad and Tobago 256.59 8895.44 1859157.87 2463.57 13.86 4.32 7.58 38.55 6.54
Venezuela 28.89 4989.43 40994.50 1019.47 16.55 10.93 4.77 21.20 20.95
Note: INC = Income per capita (constant 2000 US$); POP = Density of population; DEMAND = Density of demand (deposits per km2) ; INF = Inflation rate; HHI = Concentration (Herfindahl Index); INT = Money market rate; AEQ = Average capital ratios; GDPG = GDP Growth; MONEY= Money / GDP. Sources: Bankscope IBCA, World Development Indicators; International Financial Statistics, own calculations.
12 For reviewing outliers in the sample, we used two criteria. First, equity should always be positive. Second, all variables should not increase or decrease between two periods dramatically. Banks that fail to meet these criteria for a given year were dropped from the sample.
314 Journal of applied economics
Some descriptive statistics regarding the variables used in the estimations
are displayed in Table 2. As seen in Table 2, the trend of the Lerner index is
upward, suggesting that the market power of Latin American banks has increased
in average during the sample period. The yearly average of the Lerner index ranges
between 0.066 (in 2001) and 0.145 (in 2008). To estimate the Boone indicators we
regress marginal costs, which are obtained from a traslog cost function specified
in (4), on the market share. The coefficient of market share in (7) is the Boone
indicator. The Boone indicators are mostly statistically significant and fluctuate
between 0.046 and -0.879 over the sample period. As seen in Table 2, there was
a small variation in the degree of competition in the region particularly between
2003 and 2008. The estimates of cost efficiency and revenue efficiency scores are
obtained from the stochastic cost function defined as in (4).
The estimated average cost efficiency and revenue efficiency for 15 Latin
American and Caribbean countries over the sample period are 0.834 and 0.763,
respectively. The average estimated cost efficiency scores do not fluctuate widely
during the sample period, reaching the minimum in 2002 (80.7%) and the
maximum in 2008 (85.5%). The average estimated revenue efficiency scores also
do not fluctuate greatly over the sample, ranging from 73.2% to 78.9%.
The last column of Table 2 reports the Z-scores, which combines in one single
indicator the banks’ profitability, equity to total asset ratio and return volatility.
The Z-score is considered as an indicator of financial stability. A higher (lower)
Z-score indicates a lower (higher) probability of insolvency risk. As seen in the
table, the Z-score ratio displays wide variation over the sample period, ranging
from 14.642 to 19.081.
Country specific mean values of the Lerner index, Boone indicators, efficiency
scores and Z-scores are reported in Table 3. The mean values of the five variables by
country show a wide range of variation. When we take year 2008 as the reference
year, the difference between the country with highest market power (Dominican
Republic, with a Lerner index of 0.043) and the country with lowest market power
(Trinidad and Tobago, with a Lerner index of 0.308) is 1 to 7, showing a great
range of variation. It is important to note that countries where Lerner indexes have
increased during the sample period coexist with the countries where the indexes
have decreased. Likewise, the Boone indicators show a wide range of variation.
The highest value of the coefficient (-0.798) in 2008 is observed in Venezuela,
implying that the banking sector in Venezuela is relatively more competitive than
those in other countries in the region. The coefficient takes positive values in
financial stability, competition and efficiency in banking 315
Colombia, Costa Rica, Honduras, Panama, and Trinidad and Tobago in 2008,
implying that these countries have less competitive banking sectors. As mentioned
above, in the case of a positive , either the market has an extreme level of collusion
or the banks are competing on quality. As seen in Table 3, competition levels
decreased between 2002 and 2008 in most of the sampled countries.
Table 2. Mean values of Lerner indices, Boone indicators, efficiency scores and Z-scores
Year Lerner Boone Cost Efficiency Revenue Efficiency Z-score
2001 0.066 (0.256)
-0.879 (1.133)
0.837 (0.072)
0.789 (0.167)
14.642 (13.688)
2002 0.009 (0.585)
0.046 (0.445)
0.807 (0.119)
0.756 (0.189)
19.081 (17.638)
2003 0.043 (0.816)
-0.157 (0.675)
0.823 (0.115)
0.750 (0.181)
19.010 (17.778)
2004 0.083 (0.485)
-0.144 (0.599)
0.826 (0.111)
0.763 (0.171)
18.804 (17.250)
2005 0.064 (0.892)
-0.222 (0.618)
0.834 (0.109)
0.764 (0.181)
18.912 (17.129)
2006 0.121 (0.743)
-0.194 (0.522)
0.842 (0.103)
0.782 (0.169)
18.297 (16.516)
2007 0.149 (0.259)
-0.209 (0.600)
0.847 (0.101)
0.788 (0.189)
17.821 (17.162)
2008 0.145 (0.173)
-0.201 (0.563)
0.855 (0.083)
0.732 (0.189)
17.459 (16.177)
Overall 0.090 (0.609)
-0.245 (0.645)
0.834 (0.106)
0.763 (0.181)
18.353 (16.977)
Note: Figures in parentheses are the standard deviations
As for the Z-score, the difference is almost the same as in the market power
case, with minimum and maximum values in 2008 of 6.123 (Argentina) and 44.138
(Trinidad and Tobago), respectively. As seen in the table, the Z-score has increased
during the sample period in about half of the sample countries.
Table 3 also shows that efficiency results display a wide range of cost and
revenue efficiency scores across countries. All the banking systems display
significant levels of cost (revenue) inefficiency ranging from 28.3% (71.3%) to
9.4% (7.1%).
316 Journal of applied economics
Tables 4 and 5 show regressions results from AR(2) models with pooled
OLS, with panel fixed effects model and with the system GMM estimator. 13
Table 4 incorporates cost efficiency into the equations, whereas Table 5 utilizes
the revenue efficiency estimates instead. At the bottom of each table, we report
specification test results for the GMM estimations.14 According to these tests, all
GMM equations are properly specified. Regarding market power, the Granger-causality results in the first panel of
Table 4 indicate that financial stability positively Granger-cause market power, suggesting that sounder banks are able to develop future market power. The results prevail when the revenue measure of efficiency is used, for the OLS and FE estimation techniques, but not for the SYS-GMM as shown in Table 5. Financial stability is not only related to banks exposure to risk, but also with their ability to cope with it, through capitalization or earning capacity, as reflected in the Z-scores. Hence, banks in the position to handle risk better and improve profitability can be able to gain market power at the expense of those less able to do so. Also, Table 4 shows that for two estimation techniques evidence is provided of the fact that banks with greater cost efficiency are able to gain future market power, as found in Maudos and Nagore (2005). Taken together, both pieces of evidence lend partial support for the “efficient structure” hypothesis.15
Regarding efficiency, however, the second panels of Tables 4 and 5 provide evidence that the Lerner indexes negatively Granger-cause cost and revenue efficiency. Hence, the results support the view that more competition is conducive to greater efficiency. Alternatively, firms enjoying greater market power tend to be less efficient. In the case of the specification incorporating cost efficiency the results are significant for all equations, where the same is true for the GMM and OLS estimations of the revenue efficiency specifications. Overall, these results provide evidence supporting the “quiet life” hypothesis, as discussed above.
13 In accordance to the Hausman test, the random effects model was rejected. 14 The Sargan test is a test on whether the instruments are uncorrelated with the error tem. Moreover, the Arellano- Bond test results also require significant AR(1) serial correlation and lack of AR(2) serial correlation. 15 We also used the Boone indicator as a measure of competition in the regressions. The results indicate that financial stability positively Granger-cause competition, suggesting that sounder banks are operating in a less competitive environment given that lower values of the Boone indicator signify more competition. The results also show that efficiency negatively Granger-causes competition, suggesting that bank efficiency increases in more competitive banking sectors. For the sake of flow and size of the paper, the results are not reported but available upon request from the authors.
financial stability, competition and efficiency in banking 317
Ta bl
e 3.
M ea
n va
lu es
o f L
er ne
r i nd
ic es
, Z -s
co re
s, a
nd e
ffi ci
en cy
s co
re s
ac ro
ss s
am pl
ed c
ou nt
rie s
Le rn
er Bo
on e
Z- sc
or e
Re ve
nu e
Ef fic
ie nc
y Co
st E
ffi ci
en cy
20 02
20 05
20 08
20 02
20 05
20 08
20 02
20 05
20 08
20 02
20 05
20 08
20 02
20 05
20 08
Ar ge
nt in
a -0
.5 64
0. 05
4 0.
17 2
-0 .5
26 -0
.7 50
-0 .6
27 9.
83 6
6. 52
6 6.
12 3
0. 97
9 0.
92 9
0. 92
9 0.
80 0
0. 88
6 0.
90 1
Bo liv
ia -0
.0 51
0. 11
8 0.
24 8
-0 .1
95 -0
.4 06
-0 .0
54 10
.9 4
13 .3
04 15
.4 49
0. 42
8 0.
59 3
0. 38
5 0.
63 5
0. 72
0 0.
78 4
Br az
il 0.
14 8
0. 20
4 0.
11 5
-0 .1
63 -0
.2 58
-0 .2
76 15
.9 36
17 .5
84 15
.7 67
0. 80
1 0.
85 7
0. 79
3 0.
85 7
0. 83
8 0.
87 9
Co lo
m bi
a 0.
02 2
0. 20
0 0.
13 8
-0 .0
33 0.
01 0
0. 17
1 11
.8 52
16 .7
01 15
.8 78
0. 79
3 0.
84 8
0. 86
2 0.
77 6
0. 76
8 0.
71 7
Co st
a Ri
ca 0.
11 3
0. 19
1 0.
14 3
1. 15
5 -0
.0 16
0. 47
1 33
.1 58
31 .3
84 28
.8 36
0. 82
6 0.
79 5
0. 78
4 0.
87 3
0. 89
5 0.
89 1
Do m
in ic
an R
ep ub
lic -0
.0 17
0. 06
5 0.
04 3
-0 .6
68 0.
57 4
-0 .1
31 14
.2 73
13 .7
97 12
.9 41
0. 73
8 0.
70 3
0. 65
9 0.
85 8
0. 84
8 0.
86 3
Ec ua
do r
0. 10
5 0.
08 5
0. 08
8 -0
.1 16
0. 16
4 -0
.0 51
32 .2
37 29
.0 39
27 .1
28 0.
78 5
0. 79
8 0.
74 9
0. 69
9 0.
67 4
0. 74
5
El S
al va
do r
0. 10
2 0.
14 4
0. 08
4 -0
.2 31
-1 .9
61 -1
.1 64
30 .2
83 31
.7 88
32 .1
72 0.
28 8
0. 34
5 0.
36 2
0. 89
9 0.
88 4
0. 87
4
Ho nd
ur as
0. 26
7 0.
10 6
0. 18
6 0.
27 7
0. 02
6 0.
28 8
27 .8
39 27
.2 48
29 .7
07 0.
35 7
0. 29
2 0.
28 7
0. 88
9 0.
84 3
0. 83
2
Ja m
ai ca
0. 18
2 0.
22 6
0. 25
2 0.
32 7
-0 .7
09 -1
.3 82
21 .9
15 32
.7 75
30 .8
98 0.
75 0
0. 87
8 0.
65 2
0. 90
6 0.
90 8
0. 91
2
Pa na
m a
0. 15
7 0.
21 4
0. 17
5 0.
19 6
0. 55
3 0.
37 2
35 .6
04 33
.4 38
31 .2
36 0.
39 7
0. 47
2 0.
42 4
0. 84
8 0.
88 6
0. 87
7
Pa ra
gu ay
0. 26
0 0.
05 5
0. 08
6 -0
.1 49
-0 .0
58 0.
23 3
8. 61
7 15
.1 76
19 .2
64 0.
62 4
0. 57
6 0.
64 5
0. 45
7 0.
71 9
0. 76
7
Pe ru
0. 12
9 0.
23 9
0. 25
6 0.
44 6
-0 .1
62 -0
.2 83
18 .5
58 18
.3 84
17 .0
03 0.
78 8
0. 85
8 0.
86 0
0. 78
5 0.
77 5
0. 81
1
Tr in
id ad
a nd
To ba
go 0.
21 3
0. 28
0 0.
30 8
-0 .0
47 0.
14 9
0. 21
6 51
.6 42
49 .4
71 44
.1 38
0. 79
4 0.
89 9
0. 53
9 0.
92 2
0. 92
3 0.
93 6
Ve ne
zu el
a 0.
22 9
0. 27
3 0.
18 7
0. 41
7 -0
.4 87
-0 .7
98 13
.0 45
8. 35
1 6.
65 6
0. 83
9 0.
82 5
0. 83
6 0.
80 5
0. 85
9 0.
86 0
318 Journal of applied economics
Table 4. Estimation Results: Granger causality between market power, risk and cost efficiency
Dependent variable: Lerner Dependent variable: c-eff Dependent variable: Z-score
OLS FE SYS-GMM OLS FE SYS-GMM OLS FE SYS-GMM
Intercept -0.545* (0.143)
-0.826* (0.032)
-0.468** (0.615)
0.206** (0.013)
0.657* (0.032)
0.375* (0.064)
0.137*** (0.143)
2.127** (0.176)
0.854*** (0.480)
1−tLerner 0.045 (0.029)
-0.431* (0.029)
0.011 (0.028)
-0.011* (0.002)
-0.010* (0.003)
-0.008** * (0.005)
-0.049* (0.016)
-0.041** (0.016)
-0.029 (0.024)
2−tLerner 0.083* (0.025)
-0.295* (0.026)
0.047 (0.029)
-0.000 (0.002)
-0.004** (0.002)
-0.000 (0.002)
-0.024*** (0.014)
-0.044* (0.014)
-0.011 (0.011)
1−− teffc 0.564** (0.256)
0.731* (0.278)
0.365 (0.446)
0.686* (0.024)
0.305* (0.028)
0.572* (0.067)
-0.294** (0.141)
0.396* (0.153)
-0.690** (0.096)
2−− teffc 0.115 (0.245)
0.144 (0.254)
0.024 (0.315)
0.081* (0.023)
-0.094* (0.025)
0.008 (0.053)
0.271** (0.135)
0.123 (0.140)
0.096 (0.232)
1−− tscoreZ 0.076*** (0.042)
0.104** (0.050)
0.101** (0.047)
-0.003 (0.004)
-0.001 (0.005)
-0.003 (0.006)
0.701* (0.023)
0.234* (0.027)
0.669* (0.062)
2−− tscoreZ -0.041 (0.042)
0.012 (0.047)
0.003 (0.064)
0.002 (0.004)
0.006 (0.004)
-0.001 (0.005)
0.248* (0.023)
0.014 (0.026)
0.194* (0.050)
M1(p-value) NA NA 0.030* NA NA 0.003* NA NA 0.005*
M2 (p-value) NA NA 0.468 NA NA 0.223 NA NA 0.622 Sargan/Hansen (p-value) NA NA 0.293 NA NA 0.217 NA NA 0.360
Diff- Sargan/Hansen (p-value)
NA NA 0.479 NA NA 0.416 NA NA 0.795
∑ Lerner
-0.011* (0.002)
-0.011* (0.001)
-0.008*** (0.057)
0.074* (0.000)
-0.094* (0.000)
-0.04 (0.167)
∑ − effc
0.679* (0.000)
0.875* (0.008)
0.389 (0.459)
-0.023 (0.801)
-0.271 (0.132)
-0.594 (0.180)
∑ − scoreZ 0.035** (0.417)
0.116*** (0.074)
0.104** (0.213)
0.001 (0.512)
0.004 (0.531)
-0.004 (0.138)
Granger Causality:
Lerner (p-value) 0.000* 0.002* 0.052*** 0.001* 0.002* 0.381
effc − (p-value) 0.000* 0.015** 0.708 0.097*** 0.035** 0.106
Z-score (p-value) 0.050** 0.098*** 0.090** 0.697 0.500 0.316
Notes: OLS: ordinary least squares: FE: fixed effects: SYS-GMM: system GMM. The Arellano and Bond dynamic panel system GMM estimations (Stata xtabond2 command) with Windmeijer (2005) corrected standard error (reported in parentheses) are used. *, **, and *** denote significance level at 1%, 5% and 10%, respectively. The variables ∑ Lerner , ∑ − effc
, and ∑ − scoreZ are the estimated coefficients for the test that the sum of lagged terms of Lerner, cost efficiency, and
risk, respectively, which show long-run impact. The Granger causality test is used to examine the null hypothesis that x doesn’t Granger-cause y. The Sargan/Hansen is a test of the over-identifying restrictions for the GMM estimators. M1 and M2 are tests for the first-order and second-order serial correlation.
financial stability, competition and efficiency in banking 319
Table 5. Estimation Results: Granger causality between market power, risk and revenue efficiency
Dependent variable: Lerner Dependent variable: r-eff Dependent variable: Z-score
OLS FE SYS-GMM OLS FE SYS-GMM OLS FE SYS-GMM
Intercept -0.090 (0.102)
-0.296 (0.416)
0.141 (0.264)
0.044* (0.009)
0.890* (0.041)
0.097* (0.028)
0.254* (0.056)
2.251* (0.228)
0.455 (0.298)
1−tLerner 0.060** (0.028)
-0.423* (0.029)
0.030*** (0.016)
-0.001 (0.002)
-0.001 (0.003)
-0.001 (0.001)
-0.051* (0.016)
-0.044* (0.016)
-0.043*** (0.025)
2−tLerner 0.084* (0.025)
-0.300* (0.026)
0.055** (0.027)
-0.003 (0.002)
-0.003** (0.002)
-0.004** (0.001)
-0.019 (0.014)
-0.038* (0.014)
-0.014 (0.009)
1−− teffr 0.030 (0.380)
0.393 (0.422)
-0.117 (0.183)
0.758* (0.035)
0.128* (0.042)
0.728* (0.049)
-0.163 (0.209)
-0.698* (0.153)
-0.319** (0.147)
2−− teffr 0.080 (0.383)
-0.140 (0.405)
-0.099 (0.283)
0.199* (0.036)
-0.284* (0.040)
0.151* (0.053)
0.013 (0.210)
0.255 (0.221)
0.039 (0.171)
1−− tscoreZ 0.082*** (0.042)
0.108** (0.047)
0.077 (0.049)
0.000 (0.004)
-0.003 (0.005)
0.000 (0.003)
0.688* (0.023)
0.225* (0.027)
0.682* (0.062)
2−− tscoreZ -0.038 (0.042)
0.008 (0.047)
-0.020 (0.052)
-0.005 (0.004)
-0.001 (0.005)
-0.003** (0.001)
0.251* (0.023)
0.019 (0.026)
0.224* (0.045)
M1(p-value) NA NA 0.030* NA NA 0.000* NA NA 0.007*
M2 (p-value) NA NA 0.477 NA NA 0.491 NA NA 0.597
Sargan/Hansen (p-value)
NA NA 0.184 NA NA 0.000* NA NA 0.196
Diff-Sargan/Hansen (p-value)
NA NA 0.221 NA NA 0.530 NA NA 0.418
∑ Lerner
-0.004 (0.247)
-0.004 (0.382)
-0.005*** (0.087)
-0.070** (0.000)
-0.082* (0.000)
-0.057*** (0.077)
∑ − effr
0.110 (0.266)
0.254 (0.610)
-0.216 (268)
-0.151* (0.006)
-0.443 (0.104)
-0.281 (0.113)
∑ − scoreZ
0.044** (0.017)
0.118*** (0.072)
0.057 (0.291)
-0.005* (0.005)
-0.003 (0.591)
-0.002 (0.433)
Granger Causality:
Lerner (p-value) 0.385 0.456 0.056*** 0.001* 0.003* 0.077***
effr − (p-value) 0.538 0.645 0.401 0.019** 0.010* 0.042**
Z-score (p-value) 0.021** 0.089*** 0.282 0.016** 0.831 0.129
Notes: OLS: ordinary least squares: FE: fixed effects: SYS-GMM: system GMM. The Arellano and Bond dynamic panel system GMM estimations (Stata xtabond2 command) with Windmeijer (2005) corrected standard error (reported in parentheses) are used. *, **, and *** denote significance level at 1%, 5% and 10%, respectively. The variables ∑ Lerner , ∑ − effc ,
and ∑ − scoreZ are the estimated coefficients for the test that the sum of lagged terms of Lerner, cost efficiency, and risk,
respectively, which show long-run impact. The Granger causality test is used to examine the null hypothesis that x doesn’t Granger-cause y. The Sargan/Hansen is a test of the over-identifying restrictions for the GMM estimators. M1 and M2 are tests for the first-order and second-order serial correlation.
320 Journal of applied economics
As for the financial stability, the third panels of Tables 4 and 5 provides evidence for Granger causality. In Table 5, which uses revenue efficiency, we show that Lerner indexes Granger-cause financial stability, both significantly and negatively, for all estimation techniques. More competition, as represented by lower Lerner indexes, causes more stability. This result strongly supports the “competition-stability” view, as discussed in Mishkin (1999), Boyd and De Nicolo (2006), De Nicolo et al. (2004), Schaeck and Cihak (2007), Schaeck et al. (2006) and, Uhde and Heimeshoff (2009). The result is strengthened in the first two equations of the corresponding panel in Table 4. Although, there is some evidence that competition Granger causes financial stability when cost efficiency is included, this evidence is less robust. Accordingly, our results should be qualified as stating that, when considering the dynamic relationship between competition, stability and revenue efficiency, we find that more competition leads to more financial stability. Our indicator of financial stability, the Z-scores, has three components: profitability, capital and the volatility of returns. Hence, we presume that one possible explanation of the results in Table 5, as opposed to Table 4, is that the revenue efficiency component of financial stability is being better controlled in this setting.
V. Conclusions
In this study, we used a sample of 272 commercial banks from fifteen Latin American countries for the period 2001-2008. We estimated cost and revenue efficiency scores, financial stability scores (Z-scores) and competition scores (Lerner indexes and Boone indicators) at bank level. This allows us to use Granger causality techniques in dynamic panels in order to establish dynamic relationships among the variables.
The results support the view that competition is conducive to greater financial stability, as posed by the “competition-stability” hypothesis, when revenue efficiency is included in the specification. At the same time, we detect complex dynamic processes: sounder banks tend to reach higher market power, lending support to the “efficient structure” hypothesis. The complexity of the dynamics is also underlined, as we also found support for the “quiet life” hypothesis. Market power, as reflected in greater Lerner indexes, seems to be conducive to greater efficiency, both in the cost and revenue sides.
Banks seem to achieve market power through better efficiency, leverage and earning ability. As size and complexity increase, however, we hypothesize that agency problems and increasing risk-taking may start to gain momentum, generating inefficiency and fragility. A competitive environment, however, can prove to be crucial to restraint these tendencies.
financial stability, competition and efficiency in banking 321
As a corollary, we regard that contrary to the conventional view, competition policy has prudential implications. Also, efficiency proves to be relevant for an effective regulation. In particular, the regulatory review of potential stake-holders agency problems and internal governance of banks are relevant. This is particularly important for firms with larger size, complexity and systemic importance. Achieving an appropriate combination of these three dimensions — competition, prudential aspects and governance — is an important key to achieving stronger financial systems.
References
Abreu, Dilip, and Markus M. Brunnermeier (2003). Bubbles and crashes. Econometrica 71: 173-204. Allen, Franklin, and Douglas Gale (2004). Competition and financial stability.
Journal of Money, Credit and Banking 36: 453-480. Arellano, Manuel, and Steve R. Bond (1991). Some tests of specification for panel
data: Monte Carlo evidence and an application to employment equations. Review of Economic Studies 58: 277-297.
Arellano, Manuel, and Olympia Bover (1995). Another look at the instrumental variables estimation of error components models. Journal of Econometrics 68: 29-51.
Barth, James R., Gerard Caprio Jr., and Ross Levine (2004). Bank regulation and supervision: What works best? Journal of Financial Intermediation 13: 205- 248.
Battese, George E., and Timothy J. Coelli (1995). A Model for Technical Inefficiency Effects in a Stochastic Frontier Production Function for Panel Data. Empirical Economics 20: 325-332.
Beck, Thorsten, Asli Demirgüc-Kunt, and Ross Levine (2006). Bank concentration, competition, and crises: First results. Journal of Banking and Finance 30: 1581-1603.
Berger Allen N., Leora F. Klapper, and Rima Turk-Ariss (2009) Bank Competition and financial stability, Journal of Financial Services Research 35: 99-118.
Berger, Allen N., and Timothy H. Hannan (1989). The price-concentration relationship in banking. Review of economics and Statistics 71: 291-299.
Bikker, Jacob A., and Katharina Haaf (2002). Competition, concentration and their relationship: An empirical analysis of the banking industry. Journal of Banking and Finance 26: 2191-2214.
Bikker, Jacob A., and Laura Spierdijk (2008). How Banking Changed Over Time. Working Paper 167, DNB.
322 Journal of applied economics
BIS – Bank for Internacional Payments (2007). Evolving banking systems in Latin
America and the Caribbean: challenges and implications for monetary policy
and financial stability. Economic and Monetary Department.
Blundell, Richard, and Steve R. Bond (1998). Initial conditions and moment
restrictions in dynamic panel data models. Journal of Econometrics 87: 115-
143.
Boone, Jan (2001). Intensity of competition and the incentive to innovate.
International Journal of Industrial Organization 19: 705-26.
Boone, Jan (2008). A new way to measure competition. Economic Journal 118:
1245-61.
Boot, Arnoud W., and Anjan V. Thakor (2000). Can relationship lending survive
competition? Journal of Finance 55: 679-713.
Boyd, John H., and Gianni De Nicolo (2005). The theory of bank risk-taking and
competition revisited. The Journal of Finance 60: 1329-1343.
Boyd, John H., Gianni De Nicolo, and Bruce D. Smith (2004). Crises in competitive
versus monopolistic banking systems, Journal of Money, Credit and Banking
36: 487-506.
Boyd, John H., and Edward C. Prescott (1986). Financial intermediary-coalitions.
Journal of Economic Theory 38: 211-232.
Brunnermeier Markus K., Andrew Crocket, Charles Goodhart, Avinash D. Persaud,
and Hyun Shin (2009). The Fundamental Principles of Financial Regulation.
Geneva Reports on the World Economy 11, International Center For Monetary
And Banking Studies, Geneva.
Carvallo, Oscar, and Adnan Kasman (2005). Cost efficiency in the Latin American
and Caribbean banking systems. Journal of International Financial Markets,
Institutions and Money 15: 55-72.
Casu, Barbara, and Claudia Girardone, C. (2009). Testing the relationship between
competition and efficiency in banking: a panel data analysis. Economics Letters
105: 134-137.
Chortareas, George, Jesus G. Garza-García, and Claudia Girardone, C. (2012).
Banking sector performance in Latin America: Market power versus efficiency.
Review of Development Economics 15: 307-325.
Claessens Stijn, Asli Demirgüc-Kunt, and Harry Huizinga (2001). How does
foreign entry affect domestic banking markets? Journal of Banking and
Finance 25: 891-911.
Claessens, Stijn, and Luc Laeven (2004). What drives bank competition? Some
international evidence. Journal of Money, Credit and Banking 36: 563-583.
financial stability, competition and efficiency in banking 323
Cordella, Tito, and Eduardo Levy-Yeyati (2003). Bank bailouts: moral hazard vs. value effect. Journal of Financial Intermediation 12: 300-330.
De Nicolo, Gianni, Philip Bartholomew, Jahanara Zaman, and Mary Zephirin (2004). Bank consolidation, internalization, and conglomerization. Working Paper 03/158, IMF.
Demirguc-Kunt, Asli, Luc Laeven, and Ross Levine (2004). Regulations, market structure, institutions, and the cost of financial intermediation. Journal of Money, Credit and Banking 36: 593-622.
Dick, Astrid, and Andreas Lehnert (2010). Personal bankruptcy and credit market competition. Journal of Finance 65: 655-686.
Dietsch, Michel, and Ana Lozano-Vivas (2000) How the environment determines banking efficiency: A comparison between French and Spanish industries. Journal of Banking and Finance 24: 985-1004.
Fiordelisi, Franco, David Marques-Ibanez, and Philip Molyneux (2011). Efficiency and risk in European banking. Journal of Banking and Finance 35: 1315-1326.
Hicks, John (1935). Annual survey of economic theory: the theory of monopoly. Econometrica 3: 256-63.
Hughes, Joseph P, and Loretta J. Mester (1998). Bank capitalization and cost: evidence of scale economies in risk management and signaling. The Review of Economics and Statistics 80: 314-325.
Kasman, Adnan, and Canan Yildirim (2006). Cost and profit efficiencies in transition banking: the case of new EU members. Applied Economics 38: 1079-1090.
Kasman, Adnan, Saadet Kasman, and Oscar Carvallo (2005). Efficiency and foreign ownership in banking: an international comparison. Discussion Paper 05/03, Department of Economics, Faculty of Business, Dokuz Eylul University.
Keeley Michael (1990). Deposit insurance, risk and market power in banking. American Economic Review 80: 1183-1200.
Koetter, Michael, James W. Kolari, and Laura Spierdijk (2012). Enjoying the quiet life under deregulation? Evidence from adjusted Lerner indices for U.S. banks. Review of Economics and Statistics 94: 462-480.
Levy-Yeyati, Eduardo, and Alejandro Micco (2007). Concentration and foreign penetration in Latin American banking sectors: Impact on competition and risk. Journal of Banking and Finance 31: 1633-1647.
Lozano-Vivas, Ana, Jesus T. Pastor, and Iftekhar Hasan (2001). European bank performance beyond country borders: what really matters? European Finance Review 5: 141-165.
Matutes, Carmen, and Xavier Vives (1996). Competition for deposits, fragility, and insurance. Journal of Financial Intermediation 5: 184-216.
324 Journal of applied economics
Matutes, Carmen, and Xavier Vives (2000). Imperfect competition, risk taking,
and regulation in banking. European Economic Review 44: 1-34.
Maudos, Joaquin, and Amparo Nagore (2005). Explaining market power
differences in banking: a cross-country study. WP-E, Instituto Valenciano de
Investigaciones Económicas.
Méon, Pierre G., and Laurent Weill (2005). Can mergers in Europe help banks
hedge against macroeconomic risk? Applied Financial Economics 15: 315-
326.
Mishkin, Frederic S. (1999). Financial consolidation: dangers and opportunities.
Journal of Banking and Finance 23: 675-691.
Nickell, Stephen J., Daphne Nicolitsas, and Neil Dryden (1997). What makes
firms perform well. European Economic Review 41: 783-796.
Panzar, John C., and James N. Rosse (1987). Testing for “monopoly” equilibrium.
The Journal of Industrial Economics 35: 443-456.
Schaeck, Klaus, and Martin Cihak (2010). Competition, efficiency, and soundness
in banking: an industrial organization perspective. Discussion Paper 20,
European Banking Center.
Schaeck, Klaus, Martin Cihak, and Simon Wolfe (2006). Competition,
concentration and bank soundness: New evidence from the micro-level.
Working Paper 06/143, IMF.
Schaeck, Klaus, and Martin Cihak (2007). Banking competition and capital ratios.
Working Paper 07/216, IMF.
Uhde, Andre, and Ulrish Heimeshoff (2009). Consolidation in banking and
financial stability in Europe: Empirical evidence. Journal of Banking and
Finance 33: 1299-1311.
Windmeijer, Frank (2008). A finite sample correction for the variance of linear
efficient two-step GMM estimators. Journal of Econometrics 126: 25-51.
Wilson, Paul W. (1994). Market-Specific Effects of Rail Deregulation. Journal of
Industrial Economics 42: 1-22.
Williams, Jonathan (2012). Efficiency and market power in Latin American
banking. Journal of Financial Stability 8: 263-276.
Yildirim, Semih H., and George C. Philippatos (2006). Restructuring, consolidation
and competition in Latin American banking markets. Journal of Banking and
Finance 31: 629-639.
Zarutskie, Rebecca (2009). Competition and specialization in credit markets. Paper
presented at the 45th Bank Market Structure Conference. Federal Reserve Bank
of Chicago.
How-does-competition-affect-bank-risk-taking-_2013_Journal-of-Financial-Stability.pdf
H
G a
b
a
A R R A A
J G L
K B F L C F
1
c u a v a a l o u t t “ a
(
a a
1 h
Journal of Financial Stability 9 (2013) 185– 195
Contents lists available at SciVerse ScienceDirect
Journal of Financial Stability
journal homepage: www.elsevier.com/locate/jfstabil
ow does competition affect bank risk-taking?
abriel Jiméneza, Jose A. Lopezb,∗, Jesús Saurinaa
Banco de España, Spain Federal Reserve Bank of San Francisco, United States
r t i c l e i n f o
rticle history: eceived 3 October 2012 eceived in revised form 5 January 2013 ccepted 21 February 2013 vailable online 14 March 2013
EL classification: 21 11
a b s t r a c t
A common assumption in the academic literature and in the supervision of banking systems is that fran- chise value plays a key role in limiting bank risk-taking. As market power is the primary source of franchise value, reduced competition in banking markets has been seen as promoting banking stability. A recent paper by Martínez-Miera and Repullo (MMR, 2010) shows that a nonlinear relationship theoretically exists between bank competition and risk-taking in the loan market. We test this hypothesis using data from the Spanish banking system. After controlling for macroeconomic conditions and bank character- istics, we find support for this nonlinear relationship using standard measures of market concentration in both the loan and deposit markets. When direct measures of market power, such as Lerner indices,
eywords: ank competition ranchise value erner index redit risk inancial stability
are used, the empirical results are more supportive of the original franchise value hypothesis, but only in the loan market. Overall, the results highlight the empirical relevance of the MMR model, even though further analysis across other banking markets is needed.
Published by Elsevier B.V.
B a b r i b c s v d a t k
. Introduction
A standard principle of banking supervision is that increased ompetition among banks could threaten the solvency of partic- lar institutions and hamper the stability of the banking system t an aggregate level. Such competition could erode the franchise alue of a bank and encourage it to pursue riskier policies in an ttempt to maintain its former profits.1 Examples of riskier policies re taking on more credit risk in the loan portfolio, lowering capital evels, or both. These riskier policies should increase the probability f higher non-performing loan ratios and lead to more bank fail- res. In contrast, restrained competition should encourage banks o protect their higher franchise values by pursuing safer policies hat contribute to the stability of the entire banking system. This
franchise value” paradigm has been supported both theoretically nd empirically over time in the banking literature.
∗ Corresponding author. Tel.: +1 415 977 3894. E-mail addresses: [email protected] (G. Jiménez), [email protected]
J.A. Lopez), [email protected] (J. Saurina). 1 The extensive theoretical literature on this topic was started by Keeley (1990)
nd is summarized in Section 2 of this paper. Carletti and Hartmann (2003) as well s Carletti (2008) survey the literature on financial stability and competition.
e D p b
e i a l H m
572-3089/$ – see front matter. Published by Elsevier B.V. ttp://dx.doi.org/10.1016/j.jfs.2013.02.004
A debate regarding this paradigm was initiated by the work of oyd and De Nicoló (BDN, 2005). In their model, less competition mong banks could result in higher interest rates being charged on usiness loans, which might raise the credit risk of borrowers as a esult of moral hazard issues, as in Stiglitz and Weiss (1981). The ncreased default risk could lead to more problem loans and greater ank instability. The authors argue that this “loan market channel” ould eliminate the trade-off between competition and financial tability implied by the “deposit channel” implied by the franchise alue paradigm; that is, the economic rents that banks earn from epositors provide the only incentives to carry out conservative sset side policies. Their proposed “risk-shifting” paradigm argues hat increased competition across both the loan and deposit mar- ets could lower loan rates, decrease borrower credit risk, and nhance financial stability. In fact, Boyd et al. (2006) as well as e Nicoló and Loukoianova (2007) provide empirical evidence of a ositive relationship between banking market concentration and ank risk-taking.
More recently, Martínez-Miera and Repullo (MMR, 2010) xtend the BDN model by allowing for imperfect correlation across ndividual firms’ default probabilities. Their model also identifies
risk-shifting effect that accounts for fewer firm defaults when oan rates decrease in a more competitive banking environment. owever, since imperfect correlation between firms is now per- itted, there is also a “margin” effect that reduces the interest
1 inanci
p t b t e c b d c r m
t e m m S c b i t m i g u m e a b c r
s i a b m W t p r w t i n s r d
t a a t l a b i
2
2
a
b r b t t
( d s c v a f w i t f I t b c
S v e v e s i t H a b g t t i m r t
B o m h r h r p m a a r
m fi e t tion leads to lower loan rates, lower firm default probabilities, and improved bank risk measures. However, the lower rates should also reduce all firms’ interest payments and thus overall bank revenues,
86 G. Jiménez et al. / Journal of F
ayments from performing loans and thus bank revenues. These wo effects work in opposite directions, so that the net effect on ank risk-taking and financial stability is unclear. In their model, he risk-shifting effect is shown to be dominated by the margin ffect in competitive banking environments, such that increased ompetition increases bank failure risk. In a more concentrated anking market, the model suggests that the risk-shifting effect ominates and thus bank failure risk declines with increased ompetition. Overall, the authors show that there is a U-shaped elationship in their model between bank competition, which is easured by the number of banks, and the risk of bank failure. The objective of this paper is to examine empirically whether
he relationship between bank competition and risk-taking is lin- ar, as suggested by both the franchise value and risk-shifting odels (although with opposite signs), or U-shaped as in the MMR odel. We examine this relationship within the context of the
panish banking system. While some papers have used cross- ountry data to examine this relationship, we focus on a single anking system to ensure comparability across both dependent and
ndependent variables. Our analysis of the Spanish banking sys- em permits us to use detailed databases to construct consistent
arket concentration variables, such as Herfindahl–Hirschmann ndexes and the number of banks operating in a market. We also enerate Lerner indexes as alternative measures of market power sing the Banco de España’s interest rate database that contains onthly information about the marginal interest rates charged by
ach bank for several banking products, such as commercial loans nd deposits. Similarly, for our independent variable measure of ank risk, we use the Banco de España’s credit register to obtain onsistent estimates of banks’ commercial non-performing loan atios (NPL), which are an empirical measure of bank risk.
Our empirical results for the Spanish banking market provide upport for the relationships proposed in the MMR model. That s, after controlling for macroeconomic conditions and bank char- cteristics, we find evidence of a nonlinear relationship between anking market competition and bank risk-taking using standard arket concentration measures for both loan and deposit markets. hen Lerner indices are used as measures of bank competition,
he results do not suggest a nonlinear relationship, but do sup- ort the franchise value paradigm directly in the loan market. This esult may be due to the fact that the MMR model is not framed ith respect to such concentration variables. Importantly, while
he empirical relationship between banking market concentration n the Spanish deposit market and bank risk-taking with respect to on-performing loans was found to be nonlinear, the coefficients uggest that the relationship is concave as opposed to the convex elationship found in the loan market, both in theory and in our ata. Further analysis of this deposit market result is necessary.
In summary, we find supportive evidence of a nonlinear rela- ionship between bank market concentration and bank risk-taking, lthough the relationship does not hold across all banking markets nd concentration variables. The paper is organized as follows. Sec- ion 2 contains a brief discussion of the theoretical and empirical iterature on the topic. In Section 3, we present our databases, vari- bles and methodology used to empirically examine the trade-off etween competition and bank risk. Section 4 presents our empir-
cal results, and Section 5 concludes.
. Literature review
.1. Theoretical literature
The “franchise value” paradigm for bank risk-taking, both with nd without government regulation, is well established in the c
al Stability 9 (2013) 185– 195
anking literature. Simply stated, the idea is that banks limit their isk-taking in order to protect the quasi-monopoly rents granted y their government charters. Increased competition would erode hese rents and the value of the charters, which would likely lead o greater bank risk-taking and greater financial instability.
One of the earliest papers in this literature was by Marcus 1984), who used a one-period model to show that franchise value eclines as a bank engages in riskier policies. Chan et al. (1986) howed that increased competition erodes the surplus that banks an earn by identifying high-quality borrowers. The reduction in alue leads banks to reduce their screening of potential borrowers nd, thus, overall portfolio credit quality declines. Keeley (1990), ollowing Furlong and Keeley (1989), used a state preference model ith two periods to show explicitly that a decline in franchise value
ncreases bank risk-taking. Besanko and Thakor (1993) showed hat increased competition erodes informational rents originated rom relationship banking and leads to greater risk-taking by banks. n a context of asymmetric information, Marquez (2002) showed hat an increase in the number of banks in a market disperses the orrower-specific information and results in both higher funding osts and greater access to credit for low-quality borrowers.
Using a dynamic optimization model with an infinite horizon, uárez (1994) showed a trade-off between market power and sol- ency. If the market power of the bank decreases, the incentive to ngage in riskier policies increases significantly. As the franchise alue of the bank is a component of bankruptcy costs, it should ncourage the bank to carry out prudent policies that increase the olvency of the bank.2 Matutes and Vives (1996, 2000) showed n a framework of imperfect competition (i.e., product differentia- ion) that higher market power reduces a bank’s default probability. ellmann et al. (2000) showed in a dynamic model of moral haz- rd that competition can have a negative impact on prudent bank ehavior. Capital requirements are not sufficient to reduce the ambling incentives in the system, and deposit rate controls need o be added as an additional regulatory instrument. Building on hat, Repullo (2004) used a dynamic model of imperfect bank- ng competition to show that more competition (i.e., lower bank
argins) leads to more risk-taking in the absence of regulation, isk-based capital requirements were found to effectively control he risk-shifting incentives in that model.
As an interesting alternative to the franchise value paradigm, oyd and De Nicoló (BDN, 2005) developed a model, modifying ne presented by Allen and Gale (2000), where an increase in bank arket power both in the loan and deposit markets translates into
igher loan rates charged to borrowers. In a moral hazard envi- onment as per Stiglitz and Weiss (1981), entrepreneurs facing igher interest rates on their loans would choose to increase the isk of their investment projects, a practice that would lead to more roblem loans and a higher bankruptcy risk for banks. They find a onotonic declining relationship between competition (measured
s the number of banks lending in a market) and bank risk; that is, s the number of banks and competition increases, the level of bank isk would decline.
Martínez-Miera and Repullo (MMR, 2010) extend the BDN odel by introducing imperfect correlation across borrowing
rms. Under this assumption, two potentially countervailing ffects of bank competition are introduced. As in the BDN model, he “risk-shifting” effect captures the result that more competi-
2 Chan et al. (1986) also consider the franchise value a component of the private ost of bankruptcy.
inanci
w T M ( i s s c t a T v o
2
o m o p u k m a b s m k c t r l p
k a a p t s o t c d A a v t G s G n
p a
m t
t p d m
r M i h l r l i d t l n s
b c h e p s s a i r i s C m o a t p c i c
e u a o a s a c f a c L o S s 3 s
G. Jiménez et al. / Journal of F
hich should lead to potentially greater bank risk and bank failures. his effect is defined as the “margin” effect by the authors. In the MR model, a U-shaped relationship between bank competition
measured as the number of banks) and the risk of bank failure s found to represent the net effect of these two forces. The risk- hifting effect is shown to dominate in very concentrated markets, uch that increased entry improves bank risk measures. In already ompetitive markets, the margin effect dominates such that fur- her entry worsens bank risk. Thus, the lowest degrees of bank risk re obtained in loan markets with moderate levels of competition. he authors importantly found that the results hold whether the ariable of interest is loan supply or pricing, thus expanding the set f circumstances under which the model applies.
.2. Empirical literature
The empirical literature that we address in this paper focuses n the relationship between measures of competition in banking arkets and bank risk. The extant studies use different measures
f bank competition, which often highlight deposit market com- etition. Keeley (1990) measured the degree of bank competition sing Tobin’s q, which is defined as the ratio of a bank’s equity mar- et valuation to its book value. First, he showed that liberalization easures eroded Tobin’s q, controlling for macroeconomic vari-
bles and bank characteristics. Second, he related two measures of ank risk to his measure of market power finding that: (1) a bank’s olvency ratio, defined as the market value of capital divided by the arket value of assets, had a positive relationship (i.e., higher mar-
et power was correlated with greater solvency) and (2) funding osts via large certificates of deposit (CDs) had a negative rela- ionship (i.e., as market power declined, the perceived bankruptcy isk of large banks increased and so did the cost of their uninsured arge CDs). On aggregate, these results support the franchise value aradigm.
Demsetz et al. (1996) showed that U.S. banks with greater mar- et power also have the largest solvency ratios and a lower level of sset risk. Saunders and Wilson (1996), for a sample of U.S. bank nd a period of a century, found support for Keeley’s results in the eriod from 1973 to 1992.3 For a sample of publicly traded U.S. hrifts, Brewer and Saidenberg (1996) found a negative relation- hip between franchise value and risk measured as the volatility f their stock prices. Hellmann et al. (2000) expressed the view hat Japanese financial-market liberalization in the 1990s increased ompetition and reduced the profitability and franchise value of omestic banks, which, jointly with other factors, lead to the East sian financial crisis and a weaker financial system in Japan. Salas nd Saurina (2003) replicated Keeley’s work for Spain, finding a ery significant and robust relationship between Tobin’s q and he solvency and non-performing loan ratios of Spanish banks.4
reater market power was found to be correlated with higher bank olvency ratios and lower credit risk losses. For Italy, Bofondi and obbi (2004) found that a bank’s loan default rate increases as the umber of banks in a market increases.
In contrast, Jayaratne and Strahan (1998) showed that bank erformance, measured using return on assets, return on equity, nd several indicators of credit quality, improved significantly after
3 In fact, Rhoades and Rutz (1982) had already found, using a quite different ethodology, that banks with higher market power (measured using a concen-
ration index) were more risk-averse. 4 The paper contains a detailed overview of regulatory changes in Spain during
he last three decades as well as some description of the institutional setting. In articular, Spain has a pre-funded deposit insurance system based on flat rates on eposits, which is independent of bank risk level. The analysis presented here differs arkedly both in terms of the variables and methodology used.
o t n z m t c
e
al Stability 9 (2013) 185– 195 187
estrictions on banks’ geographic expansion were lifted in the U.S. oreover, loan losses decreased sharply after statewide branch-
ng was permitted. Thus, an increase in competition seems to have ad the opposite effect of the franchise value paradigm. Neverthe-
ess, Dick (2006) provides evidence of a positive and significant elationship between banking deregulation and increases in loan osses. Hannan and Prager (1998) showed that liberalization of nterstate branching and operations increased competition in the eposit market and reduced profitability, ceteris paribus Moreover, he literature focusing on new bank entrants finds that increases in oan market competition may increase loan losses due to the win- er’s curse arising from larger degrees of asymmetric information; ee Shaffer (1998).
In the above-mentioned studies, differences in the degree of ank competition were either cross-sectional or caused by key hanges in regulation within one country. Several other studies ave examined this relationship in a cross-country setting. Beck t al. (2006) examine banking data for 69 countries over a 20 year eriod, and they found that more concentrated national banking ystems are subject to a lower probability of systemic banking cri- is and hence are more stable. However, they cast doubts on the ppropriateness of the share of assets of the three largest banks n the banking system of each country (i.e., their C3 measure) and elated measures as proxies for competitiveness in a national bank- ng system. Claessens and Laeven (2004) showed a positive and ignificant relationship between bank concentration measured as 5 and the H-statistic, a measure of the intensity of competition in a arket developed by Panzar and Rosse (1987). Robustness analyses
f this result showed that the relationship between concentration nd the H-statistic could also be insignificant, and they concluded hat bank concentration is not a good summary of the bank com- etitive environment.5 Also using the H-statistic as the measure of ompetitiveness, Levy Yeyati and Micco (2007) found an increase n bank risk as bank competition increased in eight Latin American ountries.
In contrast, Boyd et al. (2006) provided cross-country empirical vidence supporting the risk-shifting model using several meas- res of bank risk – namely a z-score based on bank returns on ssets (ROA), its dispersion measured as �(ROA), and the ratio f equity to total assets – and bank competition measured using
Herfindahl–Hirschmann index (HHI). They examined two data amples: a cross section of around 2500 small, rural banks oper- ting within the U.S. and a panel of about 2700 banks from 134 ountries, excluding Western countries. In both samples, they ound a negative and significant relationship between their z-score nd the HHI; thus, more concentrated banking markets are asso- iated with greater risk of bank failures. Moreover, De Nicoló and oukoianova (2007) found that this result is stronger when bank wnership is taken into account. Also in a cross-country setting, chaek et al. (2006) found that more competitive national banking ystems are less prone to systemic crises based on their analysis of 8 countries over the period from 1980 to 2003 again using the H tatistic.
Berger et al. (2009) found more nuanced results in their study f banking systems across 23 developing countries. In that study, hey examine how several measures of bank-level risk, such as on-performing loans and book-value failure probabilities (i.e., -scores), are affected by a bank-level Lerner index, which is a
ark-up of prices over marginal costs and is generated using a
ranslog cost function. Importantly, the authors include squared ompetition measures in their empirical work to account for
5 A survey of the literature on bank concentration and competition is in Berger t al. (2004).
1 inanci
t r t “ m w r c
3
3
a r t r w c a fi b i B R i D t m u g
c m a p s L i d t o t t p t c l a
e g t m t w i t o
(
o r r a c m t
W i w a w s a t a i e
p s m w m i t B n p b a
m w a m s
r b a s t a i c
f r p t i risk premium applied to the borrower. The Lerner index for the whole loan portfolio is on average positive, but again quite small; banks earn only a 5% margin, on average, on all of their lending.
88 G. Jiménez et al. / Journal of F
he possible nonlinearities suggested by the MMR model. Their esults provide some support for the theoretical MMR result in hat they found support for both the “competition-fragility” and competition-stability” hypotheses. They found that banks with ore market power typically have lower overall risk measures, hich supports the “competition-fragility” hypothesis, as well as
iskier loan portfolios, a result that is partially offset by their higher apital ratios.
. Data and model description
.1. Data
In this paper, we use different measures of bank market power nd risk-taking to test whether the franchise value paradigm, the isk-shifting paradigm, or both (as per the MMR model) apply to he Spanish banking system. Our dependent variable measure of isk-taking is a bank’s commercial non-performing loan (NPL) ratio, hich is an ex-post measure of credit risk. We focus on commer-
ial credit risk for two reasons. First, the BDN and MMR models re based importantly on the borrowing behavior of commercial rms, and second, credit risk is the primary driver of risk for most anks, although other risks obviously exist. The NPL ratios for Span-
sh banks are obtained from the credit register maintained by the anco de España, which is known as the Central de Información de iesgos (CIR). The CIR contains information on any loan, includ-
ng mortgages and consumer loans, above a minimum threshold of 6000 granted by any bank operating in Spain. Therefore, it con- ains a full census of commercial loans granted in Spain. We have
onthly information starting in 1984, but for practical reasons, we se only annual data from the month of December without loss of enerality.6
As discussed previously, various measures of the degree of bank ompetition have been used in the literature. We use three standard easures in our analysis – C5, HHI, and the number of banks oper-
ting in each market, which is defined as one of the fifty Spanish rovinces – as proxies for market power. Note that the last mea- ure is the one explicitly used in the MMR model. We also construct erner indexes for a variety of banking products. The Lerner index s a commonly used measure of market power that captures the egree to which a firm can increase their marginal price beyond heir marginal cost. This computation requires a proper estimate f the marginal cost of the product. For our analysis, we take advan- age of a database maintained by the Banco de España that records he marginal interest rate each bank charges on an array of banking roducts each month over the period from 1988 through 2003 (see he Appendix for details of these calculations). For our analysis, we alculate Lerner indexes for commercial bank receivables, credit ines, and all loans, including mortgages and consumer loans. We lso compute Lerner indexes for deposits.
Given that our dependent variable is the level of credit risk at ach bank and that the Spanish credit market is segmented geo- raphically into 50 provinces, the concentration measures reflect he degree of concentration each bank faces in each of the regional
arkets where it operates. We construct an aggregate concentra- ion measure for each bank using a weighted average, where the
eights are the market share of commercial loans each bank holds
n each province. If a bank only operates in one province, it faces he concentration indicators of that province; whereas if a bank perates nationwide, it has a nationwide weighted index for each
6 A more detailed description of the CIR database can be found in Jiménez et al. 2006) and Jiménez et al. (2009).
t
t
fi o o
al Stability 9 (2013) 185– 195
f the concentration measures.7 Again, the concentration variables efer to the commercial loan market to be consistent with our other isk and competition measures. Finally, in our analysis, we also use
database of bank accounting data to control for individual bank haracteristics, such as return on assets (ROA). We focus on com- ercial and savings banks, which provide 95% of the credit market
o firms.8
Table 1 presents the descriptive statistics for the variables used. e have 1262 bank-year observations for 107 commercial and sav-
ngs banks over the 14-year sample.9 The average NPL ratio is 4.4% ith a large degree of dispersion across banks; ranging from 0% to
bove 38%. There is significant variation over time in this variable ith the median NPL ratio at around 2% at the beginning of the
ample period, rising to a median value of 7% in 1993, and falling to round 1% in more recent years. These time dynamics are related o the Spanish business cycle, which experienced a deep recession round 1993 and two expansion periods before and after 1993. Real nterest rates declined steadily during the period, as the Spanish conomy converged to that of the euro zone countries.
Next, we summarize our various concentration measures as roxies for the degree of bank competition. While there are a rea- onably large number of banks operating in each provincial credit arket, there is a high degree of dispersion, ranging from provinces ith 22 banks to 148 banks in certain years. We do not have ore detailed geographical market breakdowns but, in general, it
s easy to see a significant correlation between the population of he province and the number of banks operating there. Madrid and arcelona, by far the most populated provinces, have a much higher umber of banks. The correlation coefficient for the logged province opulation and the log of the number of banks in the province is sta- le at 0.88 in 1990 and 0.85 in 2000. Across provinces, we observe
variety of patterns regarding the number of banks. For commercial loans, the market share of the first five com-
ercial lenders in each province, denoted as C5, is relatively high ith an average of 58%, ranging from 40% to 74%. Across provinces
nd time, there are occasional jumps in the C5 index as large banks erge. Regarding mergers, we have treated banks merged as two
eparate entities before the merger and as a new one after it. The HHI for commercial loans has an average of around 8, which
oughly implies 12 banks of equal size per market. Since this num- er is well below the average number of 76 banks per province,
large number of banks in each market must have a tiny market hare with only one or a few branches in the province. This fact fur- her points toward the need for careful use of the number of banks s an empirical proxy for competition in a market, even though it s the one often used in theoretical models. The loan HHI shows no lear cross-sectional pattern across provinces.
With respect to our Lerner index measures, the average index or receivables is positive, although relatively small; margins on eceivables are only about 15% of the rates charged once the risk remium has been accounted for. For credit lines, the index is nega- ive on average and zero at the median, suggesting that the median nterest rate on credit lines only covers the funding cost plus the
7 Our robustness tests indicate that this aggregation procedure does not affect he qualitative results of our analysis.
8 Credit cooperatives and specialized lenders are excluded because of the lack of he required data on interest rates.
9 This is the final number of observations used to run the regressions, after taking rst differences and allowing for lags in the instruments. The original number of bservations was 1632. The data is uniformly distributed across the entire 16 years f the sample period from 1988 to 2003.
G. Jiménez et al. / Journal of Financial Stability 9 (2013) 185– 195 189
Table 1 Descriptive statistics for bank-year observations.
Variables Mean S.D. Median Minimum Maximum
NPLit 4.44 4.93 2.66 0.00 38.02 GDPGt 2.92 1.56 2.76 −1.03 5.04 SIZEit 0.70 1.27 0.28 0.00 9.32 LOAN RATIOit 25.41 12.55 23.00 0.08 90.14 ROAit 0.66 1.19 0.72 −16.19 11.08 Number of banksit 75.93 24.77 73.00 22.00 148.00 C5 loansit 57.73 6.60 58.44 40.00 74.25 Her loans firmsit 8.22 1.86 8.09 4.14 15.02 Lerner receivablesit 0.15 0.39 0.19 −7.96 0.64 Lerner credit linesit −0.10 0.50 0.00 −6.09 0.70 Lerner loansit 0.05 0.53 0.11 −12.27 0.52 C5 depositsit 68.00 5.61 67.35 53.70 84.64 Her depositsit 16.77 3.67 16.33 7.58 28.57 Lerner depositsit 0.35 0.11 0.36 −0.49 0.68
NPLit is the commercial non-performing loan ratio of bank i at time t; GDPGt is the real GDP growth rate of the Spanish economy at time t; SIZEit is the market share of bank i at time t in terms of total loans; LOAN RATIOit measures the specialization of firm i at time t in the non-financial sector through the ratio of loans to firms over total loans; ROAit is the return on assets of bank i at time t; Number of banksit is the number of banks that has the representative province for bank i at time t, calculated as the weighted average (by total loans) over all the provinces where the bank grants loans (the other concentration and competition measures are obtained in the same way); C5 denotes the share of the 5 largest banks in the representative province for bank i at time t; Herit is the Herfindahl index of concentration for the representative province of bank i at time t, calculated in each province as the sum of banks’ squared market shares in loans granted in the province; Lernerit is the Lerner index of bank i in year t defined for p cost o T m wh 1
a d t v m s i c a
3
c
R
w a fi m b
l
T a u o a i
r
p c i m a r
t d b l f r t l v e
p a i i l p p i a m
i s l t p w
roduct l of the asset side as (Rl − R)/Rl , where R is the credit risk adjusted marginal he time period analyzed spans from 1988 to 2003. We have 1632 observations fro 07 unique banks (commercial and savings banks).
Table 1 also shows that commercial and savings banks have an verage annual ROA of 0.66% for the period analyzed, with a high egree of heterogeneity. In the sample, we measure bank size using he share of total CIR loans that the bank originated. The average alue is 0.7%, which is relatively small, but the range goes up to a aximum value of 9.3%; thus, we have also heterogeneity in bank
izes. Finally, there is a significant difference in degrees of special- zation in the commercial lending (LOAN RATIO) as some banks oncentrate on commercial lending (as high as 90%), while others lmost do not operate in that market segment.
.2. Model description
To examine the various hypotheses regarding the effect of bank ompetition on bank risk, we estimate the general regression:
ISKit = f (COMPETITION INDEXit, BUSINESS CYCLEt,
BANK CONTROL VARIABLESit), (1)
here the i subscript refers to a bank and the t subscript refers to sample year. The model sets the relationship between the speci- ed bank risk measure and the specified bank market competition easure, controlling for bank characteristics and the state of the
usiness cycle. The actual model specification we examine is:
n (
NPLit
100 − NPLit
) = ̨ + ̌ ln
( NPLit−1
100 − NPLit−1
) + ı1STRUCTUREit
+ ı2STRUCTURE2 it + �1GDPGt + �2GDPGt−1 + �1ROAit + �2SIZEit
+�3LOAN RATIOit + �i + εit . (2)
he dependent bank risk variable is the log-odds transformation of bank’s NPL ratio, which changes the variable’s support from the nit interval to the real number line. There is a significant degree f persistence in the transformed NPL variable, as indicated by the verage value of the first-order autocorrelation of 0.68. Hence, we
nclude the lagged dependent variable as an explanatory variable.10
We control for business cycle conditions by introducing the cur- ent and lagged values of the annual real GDP growth rate, since
10 See Salas and Saurina (2002) for a more detailed discussion.
t i t t s d
f product l for bank j granted in year t, while it is defined as (R − Rl)/R for deposits. ich, after taking first differences and instrumenting remain 1262 corresponding to
roblem loans develop in line with the business cycle. We also ontrol for the profitability of the bank, its size, and its special- zation in commercial lending using its contemporaneous ROA, its
arket share in terms of CIR total loans (SIZEit), and its percent- ge of total assets that represent commercial loans (LOAN RATIOit), espectively.
Our primary variables of interest are related to the structure of he banking market and the degree of bank market competition, enoted STRUCTUREit. For the loan market, we use the number of anks, C5, HHI as well as our Lerner indexes for receivables, credit
ines and all loans. For the deposit market, we use the Lerner index or total deposits. We include the squared STRUCTUREit term in our egressions to address the hypothesis within the MMR model that he relationship between the number of banks and bank risk is not inear. We include the bank fixed effect �i to control for unobser- able bank characteristics constant over time, and εit is a random rror that has a normal distribution.
In our model specification, positive values for ı1 and ı2 would rovide evidence in support of the risk-shifting paradigm; that is, s market power increases (and competition decreases), bank risk- ness as measured by NPL ratios would also increase. In contrast, f these parameters are negative, increased market power would ead to less bank risk, which is supportive of the franchise value aradigm. If ı1 is negative and ı2 is positive, the results would sup- ort the U-shaped pattern proposed in the MMR model. Note that
f ı1 is positive and ı2 is negative, the results would still support nonlinear pattern, although the model implications would not atch the MMR model directly. Regarding the other explanatory variables, we expect a signif-
cant positive coefficient for the lagged dependent variable and a ignificant negative effect for the GDPG variables, since problem oans should increase in bad times. We do not have clear expec- ations for the bank characteristics. In general, there should be a ositive, long-term relationship between risk and return, but banks ith high NPL ratios might experience significant losses in a par-
icular year. The specialization of a bank should be indicative of mproved monitoring and screening of borrowers, but, at the same
ime, specialized banks might be willing to take more risks. Finally, here is no general support for a certain relationship between the ize of the bank and its risk level. A larger bank benefits from risk iversification but, at the same time, bank managers could take
1 inancial Stability 9 (2013) 185– 195
a t
l b o d t fi t i k t u w a
4
4
a b s L t m t s r
w i s l c s t r k c l b
r t f a m a m a e k c A a m
l
1 −0 .4
58 **
* 1
−0 .0
95 **
* −0
.0 10
1 −0
.3 77
** *
0. 13
5** *
0. 08
8** *
1 −0
.2 16
** *
0. 09
0** *
−0 .0
25
0. 06
1** 1
0. 04
0
−0 .0
69 **
0. 17
9** *
−0 .0
62 **
−0 .3
55 **
* 1
−0 .2
16 **
* 0.
25 2**
* −0
.0 95
** *
0. 16
2** *
0. 22
7** *
−0 .6
49 **
* 1
−0 .1
06 **
* 0.
16 5**
* −0
.1 32
** *
0. 11
5** *
0. 20
6** *
−0 .6
30 **
* 0.
85 3**
* 1
−0 .1
89 **
* 0.
03 3
0. 01
8 0.
18 0**
* 0.
15 8**
* −0
.1 67
** *
0. 01
3
0. 04
6
1 −0
.5 57
** *
0. 26
9** *
−0 .0
33
0. 12
5** *
0. 24
0** *
−0 .2
51 **
* 0.
23 7**
* 0.
18 9**
* 0.
14 2**
* 1
−0 .4
65 **
* 0.
30 3**
* −0
.0 19
0. 13
5** *
0. 24
0** *
−0 .2
99 **
* 0.
21 3**
* 0.
18 6**
* 0.
55 1**
* 0.
71 9**
* 1
−0 .1
77 **
* 0.
15 5**
* −0
.1 67
** *
0. 11
1** *
0. 19
6** *
−0 .6
76 **
* 0.
82 0**
* 0.
77 3**
* 0.
10 1**
* 0.
27 6**
* 0.
28 8**
* 1
−0 .1
09 **
* 0.
08 2**
* −0
.1 55
** *
0. 07
6** *
0. 12
8** *
−0 .4
28 **
* 0.
53 7**
* 0.
59 1**
* 0.
08 3**
* 0.
19 9**
* 0.
20 2**
* 0.
83 7**
* 1
−0 .0
89 **
* 0.
08 6**
* −0
.0 90
** *
−0 .1
81 **
* 0.
18 0**
* −0
.1 85
** *
0. 17
1** *
0. 14
5** *
−0 .0
05
0. 18
3** *
0. 12
3** *
0. 16
0** *
0. 15
5** *
1
n -p
er fo
rm in
g
lo an
ra ti
o
of
ba n
k
i a t
ti m
e
t;
G D
PG t
is
th e
re al
G D
P
gr ow
th
ra te
of
th e
Sp an
is h
ec on
om y
at
ti m
e
t;
SI ZE
it is
th e
m ar
ke t
sh ar
e
of
ba n
k
i a t
ti m
e
t
in
te rm
s
of
to ta
l l oa
n s;
LO A
N
R A
TI O
it
n
of
fi rm
i a t t
im e
t i n
th e
n on
-fi n
an ci
al
se ct
or
th ro
u gh
th e
ra ti
o
of
lo an
s
to
fi rm
s
ov er
to ta
l l oa
n s;
R O
A it
is
th e
re tu
rn
on
as se
ts
of
ba n
k
i a t t
im e
t;
N u
m be
r
of
ba n
ks it
is
th e
n u
m be
r
of
ba n
ks
th at
h as
th e
r
ba n
k
i a t
ti m
e
t,
ca lc
u la
te d
as
th e
w ei
gh te
d
av er
ag e
(b y
to ta
l l oa
n s)
ov er
al l t
h e
p ro
vi n
ce s
w h
er e
th e
ba n
k
gr an
ts
lo an
s
(t h
e
ot h
er
co n
ce n
tr at
io n
an d
co m
p et
it io
n
m ea
su re
s
ar e
ob ta
in ed
in
th e
sa m
e
of
th e
5
la rg
es t
ba n
ks
in
th e
re p
re se
n ta
ti ve
p ro
vi n
ce
fo r
ba n
k
i a t
ti m
e
t;
H er
it is
th e
H er
fi n
d ah
l i n
d ex
of
co n
ce n
tr at
io n
fo r
th e
re p
re se
n ta
ti ve
p ro
vi n
ce
of
ba n
k
i a t
ti m
e
t,
ca lc
u la
te d
in
ea ch
p ro
vi n
ce ed
m ar
ke t
sh ar
es
in
lo an
s
gr an
te d
in
th e
p ro
vi n
ce ;
Le rn
er it
is
th e
Le rn
er
in d
ex
of
ba n
k
i i n
ye ar
t
d efi
n ed
fo r
p ro
d u
ct
l o f t
h e
as se
t
si d
e
as
(R l −
R )/
R l,
w h
er e
R
is
th e
cr ed
it
ri sk
ad ju
st ed
m ar
gi n
al
co st
te d
in
ye ar
t,
w h
il e
it
is
d efi
n ed
as
(R
−
R l)
/R
fo r
d ep
os it
s. ve
l. ve
l.
90 G. Jiménez et al. / Journal of F
dvantage of that in order to push further along the risk profile of he bank.11
It is possible that unobservable bank characteristics are corre- ated with the bank NPL ratios; for example, the risk aversion of ank managers and/or shareholders. In this case, an OLS estimation f model (2) would produce biased parameters due to the lagged ependent variable. To address these estimation problems, we use he Arellano and Bond (1991) procedure to estimate the model in rst-difference form using GMM estimation techniques. We thus reat bank characteristics as endogenous and use their second lag to nstrument for them. We also consider the concentration and mar- et power measures as potentially endogenous and instrument for hem with the second lag. The validity of these instruments is tested sing the standard Hansen test. Since we take first differences, e should observe first-order autocorrelation and no second-order
utocorrelation in the residuals.12
. Empirical results
.1. Correlations
Table 2 presents the pairwise correlations between the vari- bles. We find a negative relationship between all our measures of ank market power and bank’s commercial NPL ratios, our mea- ure of bank risk. The correlations for the different loan market erner indexes range from −0.56 to −0.20, and the correlation for he deposit market Lerner index is −0.09. Both the C5 and HHI
easures for both markets show a negative, although low correla- ion, with ex-post credit risk. Therefore, simple correlation analysis uggests a negative relationship between market power and bank isk, supporting the franchise value paradigm.
As expected, commercial NPL ratios are correlated negatively ith the business cycle. Specialization in commercial lending
s correlated with lower NPL ratios, probably due to enhanced creening and monitoring of borrowers. We find that current prob- em loans have a negative impact on current profitability. The orrelation between size of the bank and risk in business loans eems weak. Bank profitability seems to be inversely related to he number of banks operating in each local market and positively elated to the standard concentration measures as well as mar- et power indicators. However, the absolute value of correlation oefficients is, in general, low and in the range of [0.16, 0.24] for oans and is 0.18 for deposits. The correlation with the number of anks is actually −0.36.
Among concentration measures, there is a strong negative cor- elation between the number of banks operating in a market and he C5 and HHI measures for both loan and deposit markets, ranging rom −0.67 to −0.42. The C5 and HHI measures for both markets re highly correlated (around 0.85) with each other. Across the two arkets, the correlations based on these concentration measures
re also high at +0.82 for the C5 measure and +0.59 for the HHI easure. Therefore, C5 and HHI would seem to be interchange-
ble as concentration proxies. However, a very different picture merges for the Lerner measures of market power. Within mar- ets, the correlations between the Lerner measures and the two oncentration measures drop sharply to between +0.15 and +0.21.
cross the markets, the correlations between the Lerner indexes re quite low at +0.12, suggesting that loan and deposit markets ight behave differently.
11 See, for instance, Hughes et al. (1996) for this last result. 12 Note that we also estimated the model using just two lags as well as all available ags as instruments, but the overall qualitative results were unchanged. Ta
b le
2 C
or re
la ti
on
co ef
fi ci
en ts
.
V ar
ia bl
es
N PL
it
G D
PG t
SI ZE
it
LO A
N
R A
TI O
it
R O
A it
N u
m be
r
of
ba n
ks jt
C 5
lo an
s j t
H er
lo an
s
fi rm
s j t
Le rn
er re
ce iv
ab le
s j t
Le rn
er
cr ed
it
li n
es jt
Le rn
er lo
an s j
t
C 5
d ep
os it
s j t
H er
d ep
os it
s j t
Le rn
er
d ep
os it
s j t
N PL
it is
th e
co m
m er
ci al
n o
m ea
su re
s
th e
sp ec
ia li
za ti
o re
p re
se n
ta ti
ve
p ro
vi n
ce
fo w
ay );
C 5
d en
ot es
th e
sh ar
e as
th e
su m
of
ba n
ks ’ s
qu ar
of
p ro
d u
ct
l f or
ba n
k
j g ra
n **
Si gn
ifi ca
n t
at
th e
5%
le **
* Si
gn ifi
ca n
t
at
th e
1%
le
G .
Jim énez
et al.
/ Journal
of Financial
Stability 9 (2013) 185– 195
191
Table 3 Loan market ln
( NPLit
100−NPLit
) = ˛ + ˇ ln
( NPLit−1
100−NPLit−1
) + ı1STRUCTUREit + ı2STRUCTURE2
it + �1GDPGt + �2GDPGt−1 + �1ROAit + �2SIZEit + �3LOAN RATIOit + �i + εit .
STRUCTUREit ln(# banks) C5 loans Her loans firms Lerner receivables Lerner credit lines Lerner loans
Coefficient t-Statistic Coefficient t-Statistic Coefficient t-Statistic Coefficient t-Statistic Coefficient t-Statistic Coefficient t-Statistic
ln(NPLit−1/(100 − NPLit−1)) 0.527 7.57*** 0.516 7.33*** 0.557 8.05*** 0.537 9.32*** 0.484 7.19*** 0.546 9.45***
GDPGt −0.143 −11.21*** −0.155 −11.92*** −0.144 −11.74*** −0.140 −12.54*** −0.122 −10.89*** −0.129 −10.74***
GDPGt−1 −0.030 −1.98** −0.025 −1.64* −0.026 −1.85* −0.046 −3.53*** −0.027 −2.20** −0.045 −3.85***
STRUCTUREit −5.842 −0.76 −0.059 −0.72 −0.395 −2.3** −0.636 −3.23*** −1.333 −5.07*** −0.936 −5.24***
STRUCTURE2 it 1.627 0.87 0.000 0.57 0.023 2.37** −0.065 −3.21*** −0.291 −2.46** −0.076 −5.24***
SIZEit −0.688 −2.49** −0.512 −2.54** −0.625 −3.14*** −0.519 −3.09*** −0.497 −3.17*** −0.453 −3.27***
LOAN RATIOit −0.032 −4.35*** −0.033 −3.79*** −0.027 −2.55** −0.026 −3.09*** −0.013 −1.63 −0.017 −1.95*
ROAit 0.008 0.19 0.015 0.38 −0.006 −0.16 −0.108 −0.86 −0.013 −0.22 0.075 1.27
No. observations 1262 1262 1262 1155 1155 1155 F test (p-value) 0.000 0.000 0.000 0.000 0.000 0.000 Test 1st order serial correlation
(m1)/p-value −3.86 0.00 −5.12 0.00 −4.15 0.00 −4.63 0.00 −4.34 0.00 −4.34 0.00
Test 2nd order serial correlation (m2)/p-value
−1.54 0.12 −1.63 0.10 −1.49 0.14 −1.39 0.16 −1.22 0.22 −1.36 0.17
Hansen test (p-value) 0.35 0.49 0.19 0.41 0.22 0.29 Bank fixed effects, �i Yes Yes Yes Yes Yes Yes
NPLit is the commercial non-performing loan ratio of bank i at time t; GDPGt is the real GDP growth rate of the Spanish economy at time t; SIZEit is the market share of bank i at time t in terms of total loans; LOAN RATIOit
measures the specialization of firm i at time t in the non-financial sector through the ratio of loans to firms over total loans; ROAit is the return on assets of bank i at time t; Number of banksit is the number of banks that has the representative province for bank i at time t, calculated as the weighted average (by total loans) over all the provinces where the bank grants loans (the other concentration and competition measures are obtained in the same way); C5 denotes the share of the 5 largest banks in the representative province for bank i at time t; Herit is the Herfindahl index of concentration for the representative province of bank i at time t, calculated in each province as the sum of banks’ squared market shares in loans granted in the province; Lernerit is the Lerner index of bank i in year t defined for product l of the asset side as (Rl − R)/Rl , where R is the credit risk adjusted marginal cost of product l for bank j granted in year t, while it is defined as (R − Rl)/R for deposits. The time period analyzed spans from 1988 to 2003. We have 1632 observations from which, after taking first differences and instrumenting remain 1262 corresponding to 107 unique banks. Standard errors (SE) of estimated coefficients consistent to any pattern of heteroskedasticity within banks.
* Statistically significant at 10%. ** Statistically significant at 5%.
*** Statistically significant at 1%.
192 G. Jiménez et al. / Journal of Financial Stability 9 (2013) 185– 195
Fig. 1. Empirical relationship between competition measures in the loan market and bank risk-taking. The x-axes in the graphs below correspond to the values of the alternative loan market measures of competition used in our analysis. These axes are ordered from less to more competition, which in most cases causes the numerical indexes to be reversed. The y-axes correspond to bank risk-taking measured as nonperforming loan ratios. The six competition measures presented here are the number of banks in panel A; C5 in panel B; the Herfindahl index in panel C; the Lerner index for receivables in panel D; the Lerner index for credit lines in panel E; and the Lerner index for total loans in panel F. All charts are computed using the correspondent Table 3 estimation results for values of the competition variable between its observed 1% and 99% percentiles. Number of banksit is the number of banks that has the representative province for bank i at time t, calculated as the weighted average (by total loans) over all the provinces where the bank grants loans (the other concentration and competition measures are obtained in the same way); C5 denotes the share of the 5 largest b ex of c p ce; Le ( ranted
i i u b
4
b f t s t a a
c t g l t T a b
i a i
s l t a t t i i
s h t m o b f I ture terms are again of opposite signs, but are now both significant at the 5% level. In fact, their implied U-shaped pattern is strongly supportive of the MMR model, as shown in Fig. 1C.
13 The x-axis values in Figs. 1 and 2 are based on the observed percentiles of the variables’ empirical distributions, starting with the first percentile and ending with the 99th percentile.
14 Note that in the remaining panels of Figs. 1 and 2, concentration decreases along
anks in the representative province for bank i at time t; Herit is the Herfindahl ind rovince as the sum of banks’ squared market shares in loans granted in the provin Rl − R)/Rl , where R is the credit risk adjusted marginal cost of product l for bank j g
Finally, it should be noticed that, in general, market power s procyclical; that is, macroeconomic improvements seem to ncrease market power in the Spanish market. Concentration meas- res are also positively correlated with the business cycle indicator, ut with low values.
.2. Regression results
Table 3 presents the estimation results for our baseline model ased on the Spanish loan market. The table’s six columns dif- er only by the market structure measure used. The validity of he instruments for our specification is satisfactory in all cases, as hown by the Hansen test. Moreover, as expected since we estimate he model in first differences, there is significant first-order serial utocorrelation in the residuals, but no significant second-order utocorrelation.
In all six regressions, the lagged endogenous variable is signifi- ant at the 1% level with a parameter value around 0.5, confirming he persistence shown in the NPL ratios. The contemporaneous GDP rowth rate is negative and significant at the 1% level, while the agged GDP growth rate is always negative but only significant in he last four columns based on the Herfindahl and Lerner measures. he parameters for these lagged variables are, in absolute terms, lways less than half of the contemporaneous ones, indicating that usiness cycle changes quickly influence firms’ problem loans.
For the bank characteristics, larger banks have lower NPL ratios n all six regressions. Thus, it seems that portfolio diversification nd possibly better managerial ability at larger banks play a role n mitigating credit risk within Spain. We find that the more
t a b m c
oncentration for the representative province of bank i at time t, calculated in each rnerit is the Lerner index of bank i in year t defined for product l of the asset side as
in year t, while it is defined as (R − Rl)/R for deposits.
pecialized a bank is in commercial lending, the lower its problem oan ratio in that sector. This result is statistically significant for he first, second and fourth regressions at the 1% level; significant t 5% level for the third one; significant only at the 10% level for he sixth; and insignificant for the fifth. These results suggest hat specialization improves the screening and monitoring abil- ties of banks. Finally, ROA as a measure of bank profitability is nsignificant in the six regressions.
Regarding the structure variables, the first column of Table 3 hows that the number of banks operating in a market does not ave a statistically significant effect on bank risk-taking, even hough the signs on the coefficients are supportive of the MMR
odel. Fig. 1A shows that as competition increases (e.g. the number f banks operating in the market grows), risk taking first declines ut then increases past a certain point.13 A similar outcome holds or the C5 concentration measure in the second column of Table 3.14
n the third column based on the HHI measure for loans, the struc-
he x-axis, and correspondingly, competition increases. For example, in Fig. 1B, s we move along the x-axis, the C5 measure of market share for the five largest anks decreases, which corresponds to increased competition. Similarly, in Fig. 1C, ovement along the x-axis corresponds to declining HHI values and increasing
ompetition.
G. Jiménez et al. / Journal of Financial Stability 9 (2013) 185– 195 193
Table 4 Deposit market ln
( NPLit
100−NPLit
) = ̨ + ̌ ln
( NPLit−1
100−NPLit−1
) + ı1STRUCTUREit + ı2STRUCTURE2
it + �1GDPGt + �2GDPGt−1 + �1ROAit + �2SIZEit + �3LOAN RATIOit + �i + εit .
STRUCTUREit C5 deposits Her deposits Lerner deposits
Coefficient t-Statistic Coefficient t-Statistic Coefficient t-Statistic
ln(NPLit−1/(100 − NPLit−1)) 0.512 6.98*** 0.501 6.52*** 0.548 8.64***
GDPGt −0.135 −8.78*** −0.135 −11.18*** −0.152 −14.04***
GDPGt−1 −0.035 −1.94* −0.034 −2.04** −0.002 −0.13 STRUCTUREit 0.500 2.27** 0.231 3.1*** 0.100 0.07 STRUCTURE2
it −0.004 −2.32** −0.007 −3.85*** −1.618 −0.89 SIZEit −0.719 −3.30*** −0.598 −3.07*** −0.666 −3.68***
LOAN RATIOit −0.023 −1.86* −0.035 −3.48*** −0.039 −4.60***
ROAit −0.012 −0.27 −0.001 −0.02 −0.062 −0.57
No. observations 1262 1262 1155 F test (p-value) 0.000 0.000 0.000 Test 1st order serial correlation (m1)/p-value −5.04 0.00 −4.92 0.00 −4.31 0.00 Test 2nd order serial correlation (m2)/p-value −1.40 0.16 −1.51 0.13 −1.13 0.26 Hansen test (p-value) 0.25 0.47 0.51 Bank fixed effects, �i Yes Yes Yes
NPLit is the commercial non-performing loan ratio of bank i at time t; GDPGt is the real GDP growth rate of the Spanish economy at time t; SIZEit is the market share of bank i at time t in terms of total loans; LOAN RATIOit measures the specialization of firm i at time t in the non-financial sector through the ratio of loans to firms over total loans; ROAit is the return on assets of bank i at time t; Number of banksit is the number of banks that has the representative province for bank i at time t, calculated as the weighted average (by total loans) over all the provinces where the bank grants loans (the other concentration and competition measures are obtained in the same way); C5 denotes the share of the 5 largest banks in the representative province for bank i at time t; Herit is the Herfindahl index of concentration for the representative province of bank i at time t, calculated in each province as the sum of banks’ squared market shares in loans granted in the province; Lernerit is the Lerner index of bank i in year t defined for product l of the asset side as (Rl − R)/Rl , where R is the credit risk adjusted marginal cost of product l for bank j granted in year t, while it is defined as (R − Rl)/R for deposits. The time period analyzed spans from 1988 to 2003. We have 1632 observations from which, after taking first differences and instrumenting remain 1262 corresponding to 107 unique banks. Standard errors (SE) of estimated coefficients consistent to any pattern of heteroskedasticity within banks.
* Statistically significant at 10%.
m a – c fi v t t
b m a l
m k d i f t
o u r
F a n a A t t p
** Statistically significant at 5%. *** Statistically significant at 1%.
In the last three columns based on the Lerner market power easures, the ı1 and ı2 estimates are both negative and significant
t the 1% level in all cases. An increase in market power for loans measured as an increase in the Lerner indexes for receivables, redit lines, or total loans – produces a decline in the risk pro- les of the banks in our sample. These results support the franchise alue paradigm instead of the U-shaped relationship proposed in he MMR model, at least for these Lerner index-based measures in he Spanish banking sector.
What if the source of market power arises not from bank assets
ut from their liability structures (or market power in the deposit arkets? For instance, more market power over deposits might
llow banks to be more aggressive in the loan market and thus end to riskier borrowers with the short term objective of increasing
i t F t
ig. 2. Empirical relationship between competition measures in the deposit market and lternative loan market measures of competition used in our analysis. These axes and all o umerical indexes to be reversed. t and the y-axes correspond to bank risk-taking measu re C5 in panel A; the Herfindahl index in panel B; and the Lerner index for deposits in pa ll charts are computed using the correspondent Table 4 estimation results for values of
he share of the 5 largest banks in the representative province for bank i at time t; Herit is ime t, calculated in each province as the sum of banks’ squared market shares in loans
roduct l of the asset side as (Rl − R)/Rl , where R is the credit risk adjusted marginal cost o
arket share. From a theoretical standpoint, it is unclear how mar- et power over deposits might affect loan underwriting and pricing ecisions. In fact, as mentioned before, correlation between Lerner
ndexes in loan and deposit markets is positive but very low. There- ore, both markets could be separated, contributing to the value of he bank independently.
Table 4 presents our empirical analysis of these questions using ur baseline specification, but with deposit market structure meas- res. The regression results in the first two columns show that the elationship is not linear, fitting the non-linear pattern proposed
n the MMR model. For both, C5 and HHI measures of concentra- ion, the two estimated parameters are significant at the 5% level. ig. 2A and B shows that from a starting point of high concentra- ion in some provincial markets (i.e., on the left-hand side of the
bank risk-taking. The x-axes in the graphs below correspond to the values of the f them are ordered from less to more competition, which in most cases causes the
red as nonperforming loan ratios. The three competition measures presented here nel C.
the competition variable between its observed 1% and 99% percentiles. C5 denotes the Herfindahl index of concentration for the representative province of bank i at
granted in the province; Lernerit is the Lerner index of bank i in year t defined for f product l for bank j granted in year t, while it is defined as (R − Rl)/R for deposits.
1 inanci
x t a r m t v p t t w s
r t f a a t t r b t m w
5
i v a a e s a h N s i a t i p v
e b c m fi i fi t e n t t i c e c r m
e r c i a r b c r m i o fi b
a b r f i
A
e R t t I e w A G g J V h
A
t r t e p
t p i u charged. Failure to take the risk premium into account would result in significant biases in measuring bank market power.17 If the inter- est rate on a loan is denoted as R1, the Lerner index (or gross profit
15 A more detailed description of this database can be found in Martín et al. (2007). Significant changes in the database prevent us from extending the time period consistently after 2003.
16 Under conventional assumptions, Tirole (1988) shows that the Lerner index
94 G. Jiménez et al. / Journal of F
-axis), as we trace a reduction in concentration by moving along he axis, competition increases as does risk. However, soon after, s concentration is further reduced, the level of NPL and thus of isk declines. It could be the case that competition in the deposit arket leads banks to take on more risk. As this market power in
he deposit market decreases, banks should become more conser- ative in their lending and conscious that their favorable funding ositions are disappearing. The third column of Table 4 presents he Lerner index results, which are not statistically significant; i.e., he results suggest that market power in deposits is uncorrelated ith risk. However, the signs of the coefficients are appropriate and
upportive of a nonlinear relationship. Overall, our empirical results are supportive of the nonlinear
elationship between bank market competition and bank risk- aking, as proposed in the MMR model. The results are strongest or the loan market using traditional concentration variables such s the Herfindahl index, whereas the results using Lerner indexes re more supportive of the franchise value hypothesis, for the ime period and banks analyzed in the paper. In deposit markets, raditional concentration measures are supportive of a nonlinear elationship between competition and risk taking in the asset side, ut the relationship is concave. This unexpected result suggests hat further theoretical work is needed to better frame banking
arket competition, when it affects both sides of the balance sheet ith risk-taking behavior just on the asset side.
. Conclusions
In the academic literature and in the actual supervision of bank- ng systems worldwide, the dominant paradigm is that franchise alue plays a key role in limiting the riskiness of individual banks nd hence of banking systems more broadly. That is, bank man- gement and shareholders will typically limit or reduce their risk xposure to preserve the bank’s franchise value. The underlying ource of franchise value is typically assumed to be market power, nd reduced competition (or, equivalently, market concentration) as been considered to promote banking stability. Boyd and De icoló (BDN, 2005) present an alternative view through a the risk-
hifting paradigm, which argues that market concentration could mpact bank stability in different ways, depending on the net effect cross deposit and loan markets. Specifically, the authors suggest hat concentration in the loan market could lead to increased lend- ng rates that both raise the borrowers’ debt loads and default robabilities as well as their incentive to engage in riskier projects ia moral hazard.
More recently, Martínez-Miera and Repullo (MMR, 2010) xtend the BDN model to allow for a nonlinear relationship etween competition and bank risk-taking, such that both the fran- hise value and risk-shifting paradigms could be possible. Their odel also identifies a risk-shifting effect that accounts for fewer
rm defaults when loan rates decrease in a more competitive bank- ng environment. However, since imperfect correlation between rms is now permitted, there is also a “margin” effect that reduces he interest payments from performing loans and thus bank rev- nues. These two effects work in opposite directions, so that the et effect on bank risk-taking and financial stability is unclear. In heir model, the risk-shifting effect is shown to be dominated by he margin effect in competitive banking environments, such that ncreased competition increases bank failure risk. In a more con- entrated banking market, the model suggests that the risk-shifting
ffect dominates and thus bank failure risk declines with increased ompetition. Overall, the authors show that there is a U-shaped elationship in their model between bank competition, which is easured by the number of banks, and the risk of bank failure.
c a i
i
al Stability 9 (2013) 185– 195
Using unique datasets covering the Spanish banking system, we xplicitly examine the relationship between bank competition and isk. Our dependent variable is a bank’s ratio of non-performing ommercial loans (NPL), which is the variable addressed directly n the MMR model. After controlling for macroeconomic conditions nd bank characteristics, we find clear support for this nonlinear elationship using standard measures of market concentration in oth loan and deposit markets. While the relationship between ompetition and risk taking is convex in the loan market, the elationship is unexpectedly concave when examining the deposit arket. When direct measures of market power, such as Lerner
ndices, are used, the empirical results are more supportive of the riginal franchise value hypothesis, but only in the loan market. We nd that the number of banks, which is the measure highlighted in oth models, has little effect on NPL ratios.
The main contribution of our paper is to perform a focused nd precise test of the relationship between bank competition and ank risk using data for the Spanish banking market. Our empirical esults provide support for the MMR model, but also point toward urther extensions in the model to take into account competition n deposit markets and its interaction with loan rates.
cknowledgements
The views expressed here are those of the authors and not nec- ssarily those of the Banco de España, the Eurosystem, the Federal eserve Bank of San Francisco or the Federal Reserve System. We hank the editor Iftekhar Hasan and an anonymous referee for heir helpful comments. We also thank seminar participants at the nternational Monetary Fund, the Federal Reserve Board of Gov- rnors, and several conferences for their comments. In particular, e also thank Antonio Antunes, Rima Turk Ariss, Thorston Beck, llen Berger, Lamont Black, Arnoud Boot, John Boyd, Mark Carey, ianni de Nicoló, Olivier De Jonghe, Robert DeYoung, Enrica Detra- iache, Astrid Dick, Mark Flannery, Christopher James, Kevin James, an Pieter Krahnen, Loretta Mester, Rafael Repullo, Nuno Ribeiro, icente Salas, Joao Santos, Miguel Segoviano, and Javier Suárez for elpful comments.
ppendix. Lerner index calculations
For our paper, we take advantage of another database main- ained by the Banco de España that records the marginal interest ate each bank charges on an array of banking each month over he period from 1988 through 2003. That is, for each bank and ach banking product, we have the average interest rate set on that roduct for new transactions.15
The Lerner index is a commonly used measure of market power hat captures the degree to which a firm can increase their marginal rice beyond their marginal cost.16 The computation of the Lerner
ndex requires a proper estimation of the marginal cost of the prod- ct, which for bank loans requires a measure of the risk premium
an be related to measures of welfare losses such as Harberger’s (1964) triangle. In recent paper, Carbó-Valverde et al. (2008) also use the Lerner index to test the mpact of market power on SME lending constraints. 17 We follow Martín et al. (2006) in what follows to properly construct the Lerner ndexes.
inanci
m R I i t h t
t a z c a f c v p b a b b r t L S
b m f d b a c i c
R
A A
B
B
B
B
B
B
B
r
k r a p p t
B
C
C
C
C
C
D
D
D
F
F
H
H
H
H
J
J
J
K
L
M
M
M
M
M
M
M
P
R
G. Jiménez et al. / Journal of F
argin relative to the market price) is defined as (R1 − R)/R1, where is the marginal cost to the bank of acquiring the funds for the loan.
f we introduce the realistic assumption that the marginal operat- ng costs of loans and deposits are either fixed in the very short erm or impossible to calculate separately, we assume that banks ave a lower bound on the marginal cost for their loans equal to he interest rate offered in the interbank market.18
However, banks must introduce a risk premium into their prices o account for credit risk. Let PD be the probability that a loan, with
normalized face value of one, will default over a specified hori- on, and let LGD be the amount of the loan’s value that the bank annot collect in case of default. If the interbank interest rate r is ssumed to be risk-free, the marginal opportunity cost of the loan or a risk-neutral bank will be the interest rate R that satisfies the ondition that the risk-free value of the loan equals the expected alue of the loan, given the PD and LGD parameters. From this sim- le identity, the marginal cost R = (r + PD − LGD)/(1 − PD − LGD) can e derived. For our calculations, the risk-free interest rate r is the nnual average of the daily interbank rate. The bank-specific, not orrower-specific, PD is obtained directly from the CIR; for a given ank and loan product at the end of year t, PD equals the bank’s atio of defaulted loans divided by total outstanding loans of that ype. Since we do not have bank-specific information regarding GD, we use the value 45% set by the Basel Committee of Banking upervisors for its regulatory capital framework.19
For our analysis, we calculate Lerner indexes for commercial anking receivables and credit lines as well as all loans, including ortgages and consumer loans. We also compute Lerner indexes
or deposits by assuming (1) the separability between loan and eposit pricing and (2) that the interbank rate acts as an upper ound for deposit rates. The Lerner index for deposits is calculated s (r − Rd)/r, where Rd is the bank’s offered rate on deposits. We also alculate an average Lerner index for loans and deposits together n order to consider the possibility that loan and deposit markets annot be separated.
eferences
llen, F., Gale, D., 2000. Comparing Financial Systems. MIT Press, Cambridge, MA. rellano, M., Bond, S., 1991. Some tests of specification for panel data: Monte Carlo
evidence and an application to employment equations. Review of Economic Studies 58, 277–297.
eck, T., Demirgüç -Kunt, A., Levine, R., 2006. Bank concentration, competition, and crises: first results. Journal of Banking and Finance 30, 1581–1603.
erger, A., Demirgüç -Kunt, A., Levine, R., Haubrich, J., 2004. Bank concentration and competition: an evolution in the making. Journal of Money, Credit and Banking 36 (June Pt 2 (3)), 433–451.
erger, A.N., Klapper, L.F., Turk-Ariss, R., 2009. Banking structures and financial stability. Journal of Financial Services Research 35, 99–118.
esanko, D., Thakor, A., 1993. Relationship banking, deposit insurance and bank portfolio choice. In: Mayer, C., Vives, X. (Eds.), Capital Markets and Financial Intermediation. Cambridge University Press, Cambridge, UK.
ofondi, M., Gobbi, G., 2004. Bad loans and entry into local credit markets. Temi di Discussione del Servizio Studi #509. Bank of Italy.
oyd, J.H., De Nicoló, G., 2005. The theory of bank risk taking and competition
revisited. Journal of Finance 60, 1329–1343.
oyd, J.H., De Nicoló, G., Al Jalal, A., 2006. Bank risk taking and competition revisited: new theory and new evidence. International Monetary Fund Working Paper 06/297.
18 Freixas and Rochet (1997) review bank pricing models in different competitive egimes and information conditions. 19 We conduct these calculations in order to obtain a measure of bank mar- et power that accounts for the risk premiums included in bank loan. Concerns egarding the potential endogeneity of our Lerner variables might arise, but we ddress them directly in the analysis. First, our GMM estimation methods appro- riately instrument for this variable. Second, the simple correlation between banks’ roduct-specific PDs and commercial NPL ratios is rather low in our sample, and hird, our analysis is robust across banking products.
R
S
S
S
S
S
S
S
T
al Stability 9 (2013) 185– 195 195
rewer III, E., Saidenberg, M.R., 1996. Franchise value, ownership structure, and risk at savings institutions. Federal Reserve Bank of New York. Research Paper 9632.
arbó-Valverde, S., Rodríguez-Fernández, F., Udell, G.F., 2008. Bank market power and SME financing constraints. Manuscript.
arletti, E., 2008. Competition and regulation in banking. In: Boot, A., Thakor, A. (Eds.), Handbook in Financial Intermediation. Elsevier, North Holland, pp. 449–482.
arletti, E., Hartmann, P., 2003. Competition and financial stability. What’s spe- cial about banking? In: Mizen, P. (Ed.), Monetary History, Exchange Rates and Financial Markets: Essays in Honor of Charles Goodhart, vol. 2. Edward Elgar, Cheltenham, UK.
han, Y., Greenbaum, S., Thakor, A., 1986. Information reusability, competition and bank asset quality. Journal of Banking and Finance 10, 243–253.
laessens, S., Laeven, L., 2004. What drives bank competition? Some interna- tional evidence. Journal of Money, Credit and Banking 36 (June Pt 2 (3)), 563–583.
emsetz, R.S., Saidenberg, M.R., Strahan, P.E., 1996. Banks with something to lose: the disciplinary role of franchise value. FRBNY Economic Policy Review October, 1–14.
e Nicoló, G., Loukoianova, E., 2007. Bank ownership, market structure and risk. Monetary Fund Working Paper 07/215.
ick, A., 2006. Nationwide branching and its impact on market structure, quality and bank performance. Journal of Business 79, 567–592.
reixas, X., Rochet, J.C., 1997. Microeconomics of Banking. The MIT Press, Cambridge, USA.
urlong, F.T., Keeley, M.C., 1989. Bank capital regulation and risk taking: a note. Journal of Banking and Finance 13, 883–891.
annan, T.H., Prager, R.A., 1998. The relaxation of entry barriers in the banking industry: an empirical investigation. Journal of Financial Services Research 14, 171–188.
arberger, A.C., 1964. The measurement of waste. American Economic Review 54 (3), 58–76.
ellmann, T.F., Murdock, K.C., Stiglitz, J.E., 2000. Liberalization, moral hazard in banking, and prudential regulation: are capital requirements enough? American Economic Review 90 (1), 147–165.
ughes, J.P., Lang, W., Mester, L.J., Moon, C., 1996. Efficient banking under interstate branching. Journal of Money, Credit and Banking 28 (November (4)), 1045–1071.
ayaratne, J., Strahan, P.E., 1998. Entry restrictions, industry evolution, and dynamic efficiency: evidence from commercial banking. Journal of Law and Economics (April (XLI)), 239–273.
iménez, G., Salas Fumás, V., Saurina, J., 2006. Determinants of collateral. Journal of Financial Economics 81, 255–281.
iménez, G., Lopez, J.A., Saurina, J., 2009. Empirical analysis of corporate credit lines. Review of Financial Studies 22 (December), 5069–5098.
eeley, M.C., 1990. Deposit insurance, risk and market power in banking. American Economic Review 80, 1183–1200.
evy Yeyati, E., Micco, A., 2007. Concentration and foreign penetration in Latin American banking sectors: impact on competition and risk. Journal of Banking and Finance 31, 1633–1647.
arcus, A.J., 1984. Deregulation and bank policy. Journal of Banking and Finance 8, 557–565.
arquez, R., 2002. Competition, adverse selection, and information dispersion in the banking industry. Review of Financial Studies 15, 901–926.
artín, A., Salas, V., Saurina, J., 2007. A test of the Law of One Price in retail banking. Journal of Money 39 (December (8)), 2021–2040.
artín, A., Salas, V., Saurina, J., 2006. Risk premium and market power in credit markets. Economic Letters 93, 450–456.
artínez-Miera, D., Repullo, R., 2010. Does competition reduce the risk of bank failure? Review of Financial Studies 23, 3638–3664.
atutes, C., Vives, X., 1996. Competition for deposits, fragility, and insurance. Jour- nal of Financial Intermediation 5, 184–216.
atutes, C., Vives, X., 2000. Imperfect competition, risk taking, and regulation in banking. European Economic Review 44, 1–34.
anzar, J.C., Rosse, J.N., 1987. Testing for monopoly’ equilibrium. Journal of Industrial Economics 35, 443–456.
hoades, S.A., Rutz, R.D., 1982. Market power and firm risk. A test of the ‘Quiet Life’ hypothesis. Journal of Monetary Economics 9, 73–85.
epullo, R., 2004. Capital requirements, market power, and risk-taking in banking. Journal of Financial Intermediation 13, 156–182.
alas, V., Saurina, J., 2002. Credit risk in two institutional regimes: Spanish commer- cial and savings banks. Journal of Financial Services Research 22 (3), 203–224.
alas, V., Saurina, J., 2003. Deregulation, market power and risk behavior in Spanish banks. European Economic Review 47, 1061–1075.
aunders, A., Wilson, B., 1996. Bank capital structure: charter value and diversifica- tion effects. Working Paper S-96-52. New York University Salomon Center.
chaek, K., Čihák, M., Wolfe, S., 2006. Are more competitive banking systems more stable? IMF Working Paper 06/143.
haffer, S., 1998. The winner’s curse in banking. Journal of Financial Intermediation 7, 359–392.
tiglitz, J., Weiss, A., 1981. Credit rationing with imperfect information. American Economic Review 71, 393–410.
uárez, F.J., 1994. Closure rules, market power and risk-taking in a dynamic model of bank behavior. Discussion Paper 196. LSE, Financial Markets Group.
irole, J., 1988. The Theory of Industrial Organization. MIT Press, Cambridge, MA.
- How does competition affect bank risk-taking?
- 1 Introduction
- 2 Literature review
- 2.1 Theoretical literature
- 2.2 Empirical literature
- 3 Data and model description
- 3.1 Data
- 3.2 Model description
- 4 Empirical results
- 4.1 Correlations
- 4.2 Regression results
- 5 Conclusions
- Acknowledgements
- Appendix Lerner index calculations
- References
Systemic-risk-governance-and-global-financial-stability_2014_Journal-of-Banking-Finance.pdf
176 L. Ellis et al. / Journal of Banking & Finance 45 (2014) 175–181
Another major development in recent years has been a focus on developing macroprudential frameworks to complement the tradi- tional microprudential approach to regulation. There is also a desire on the part of the policy makers and the market participants to see the financial system sharpen its ability to apply effective macroprudential policies without stifling economic growth and innovation. However, the identification and monitoring of systemic risk and optimal macroprudential policies are still in the early stage of development. In search for global financial stability, one area that has received less attention in recent years are the issues related to bank governance. More discussion and analysis of the role of governance within the operation of financial institutions and also the role of governance in contributing to global financial stability would be an essential part of any policy package aiming to address a number of the underlying causes of financial fragility over time. The paper argues that while there have been analyses of banks competition policies, resolution policies, supervisory poli- cies and auditing and valuation policies, less attention has been paid to the role of bank governance and systemic risk, despite a strong link between governance and risk-taking. This is puzzling (Jensen and Meckling, 1976).
Recent empirical evidence suggests this link may be especially strong for financial firms (Beltratti and Stulz, 2009; Ferreira et al., 2012; Laeven and Levine, 2009). Certainly, the global financial cri- sis unearthed bank governance failures operating at multiple lev- els: managers failing to control risk-takers, boards failing to control managers and investors failing to discipline either manag- ers or boards.
The purpose of this paper is to discuss some of the issues related to systemic risk, governance and financial stability. To this end, Section 2 of the paper discusses issues related to systemic risk, the identification of G-SFIS and D-SIBs and the recent policies related to macroprudential policies and the importance of imple- menting globally a number of regulatory policies that have been proposed by various international institutions; Section 3 analyses issues related to measurement of systemic risk. It argues that the diversity within the financial system also supports the fact that a single measure of systemic risk is unlikely to be universally appli- cable. Furthermore, this section discusses how, due to the diversity of the whole financial system, the risks individual financial institu- tions face are also diverse. The section also discusses why one may not assume that the appropriate capitalisation is constant across all risks. Section 4 discusses bank governance, including how this important aspect of global financial stability has been overlooked for vigorous debate and analysis in recent times and the impor- tance of having specific policies to address the current structural weaknesses of bank governance. This section offers solutions regarding how to strengthen bank governance, including the regu- latory capital base of banks, could be increased, the compensation structure of managers could be reformed and efforts could be made by putting resolution regimes which offer the credible prospect of ‘‘bailing-in’’ creditors in the event of stress, into place and Section 5 concludes.
1 This section of the paper benefited from a number of information and issues that can be found in the FSB publications and website.
2. Large financial institutions and propagation of systemic risk
2.1. Identification of systemically important financial institution
According to Financial Stability Board, a financial institution could be defined as a G-SIFI, if its failure or fragility could impact other financial institutions, the wider financial system and the domestic and international economies (FSB, 2011a, 2011b). Given the size and complex nature of these large financial institutions with other institutions, their failure or fragility could negatively impact the overall global financial system. A number of attributes
have been identified by the FSB that could qualify certain financial institutions to be part of G-SIFIS. These attributes include: size, lack of substitutability and interconnectedness. In addition, the Basel Committee for Banking Supervision (Basel Committee) identified cross-jurisdictional activity and complexity as other attributes of financial institutions which could fall into the category of G-SIFIs (see BIS, 2013).
As of November 2013, there are 29 designated G-SIFIs by the FSB. As part of an overall strategy to ensure more stable G-SIFs, these institutions are now required to maintain greater loss absor- bency capabilities under Basel III. As highlighted by the FSB’s var- ious publications, this additional requirement, which applies in addition to the Basel III capital requirements, is intended to reduce the ‘‘cross-border negative externalities’’ of the global financial system and to reduce systemic risk. Some regional and domestic banks could also pose systemic risk to the national and regional financial systems and national economies, (over and above those institutions identified as G-SIFIs). To this end, the Basel Committee has developed a framework and asked local authorities to identify whether a large domestic bank is systemically important from a domestic perspective (domestic systemically important bank (D-SIB). A number of national authorities have already identified their D-SIBs. One should also note that there are also some regional systemically important banks that could be operating in different continents whose operations could pose systemic risk. As part of an overall strategy to increase global financial stability, interna- tional institutions such as the FSB now have a strategy in place to ensure that national and international policy makers work together and ensure that the G-SIFIs and D-SIBs are better super- vised and these institutions, as state above, are required to put aside more capital as part of their operation. This is because the current strategy is to insulate the global financial system from sys- temic shocks on top of the Basel III capital and liquidity require- ments.1 As Moshirian (2012) stated, the Great Depression in the 1930s created incentive for authorities to collect national economic data so policy makers could measure the magnitude of economic downturn in the economy accurately. This attempt in the 1930s led to the emergence of relevant data that are used to generate what is now referred to as Gross Domestic Products (GDP). Similarly the recent global financial crisis has created a new impetus to policy makers and market participants to improve and also share some financial data, particularly when the financial market is becoming increasingly interconnected. The attempt by the FSB, in collaboration with central banks and supervisors, to create a mechanism that could facilitate the international sharing of firm-level data on sys- temically important financial institutions, is yet another step in the process of facilitating globalisation and another attempt to enhance the capacity of the global financial system to be more informed about the overall state of large financial institutions and the nature of their aggregate risk taking and business models.
2.2. Macroprudential policies and a global approach to financial stability
The FSB has been working on issues related to macroprudential policies over the last few years. While a number of proposals have emerged as part of the tools available to enhance the effectiveness of policy recommendations with respect to macroprudential strat- egies, this area of policy development is still in its infancy and requires more work (Arnold et al., 2012). Nevertheless, some good progress has been made in recent times including the work of the FSB itself. At the present time, there are four main areas of
L. Ellis et al. / Journal of Banking & Finance 45 (2014) 175–181 177
development proceeding under the supervision of the FSB. The first is developing measures to identify and monitor systemic risk, including both regulated banking and shadow banking. As the FSB (2011a) stated, despite some progress in collecting new data, the challenge remains to be able to identify tools and instruments that could guide forward policy making and ensure its credibility.
Therefore, to be able to identify and contain systemic risk through credible instruments and tools remains the second area of development. According to the FSB (2011a), there are currently a range of tools used by various countries to address systemic risk, and they fall roughly into three categories. These categories are: (1) tools to address financial stability risks arising from rapid credit expansion; (2) tools to address amplification mechanisms of systemic risk such as leverage and maturity mismatches and (3) tools to limit spillover effects from the failure of SIFIs.
To be able to develop institutional arrangements for macropru- dential regulators at both domestic and regional levels will be the third area of development.
Not surprisingly, the fourth area identified by the FSB is the importance of achieving regional and international cooperation that is essential to address challenges of systemic risk that often become supernational in their dimension. The increasing intercon- nectedness of global finance implies that some of the financial risk may well spill over to other parts of the world. One of the chal- lenges of our generation will remain how to ensure an effective global financial system that has the full support of national author- ities and the private sector, that in turn could minimise the oppor- tunity for regulatory arbitrage. Furthermore, consistency of action by all national authorities could also generate more confidence in the collective implementation of some of the current and future global agreements. Obviously, amongst the key protagonists of the financial market (i.e. policy makers, regulators, researchers and market participants) there should be constant dialogue, con- sultation, and reflection about some of these global agreements.
It is noteworthy that the BIS is also keen to see certain stan- dards globally accepted and implemented. For instance, Caruana (2013a) stated, if we do not control fragmentation tendency and also regulatory arbitrage, we could lose the benefits of globalisa- tion and the pace of process of financial globalisation that has accelerated due to technological changes and the increasing inter- dependence of national economies. Caruana (2013a) also stated that fragmentation could prevent the proper transmission of mon- etary policy, particularly in the Euro-zone area., Furthermore, in addition to a structured supervisory framework (e.g. Peer reviews), greater cooperation is needed between domestic and foreign policy makers, where each are cooperating in order to resolve issues. Fur- thermore, consistency must be achieved across the international stage. Rules and minimum standards should not be different based on the basis of a bank’s nationality. As a result, if some banks are held to a lower standard than the internationally agreed rules, this will result in an uneven playing field.
In this context, as the FSB (2011a) stated that there is a need to establish strong mechanisms to ensure consistent action by coun- tries to contain risk, and to ensure national regulatory frameworks are consistent.
With regard to regional cooperation, it should be noted that regional mechanisms, such as the college of supervisors, may work relatively well for some of the G-SIFs and also regional and domes- tic SIFS in the EU due to the high level of financial integration. However, such regional cooperation and information sharing may not necessarily work in other parts of the world such as Asia, where the political climate is different and the level of national autonomy much higher. One cannot dismiss a number of regional architectures that have been erected to bring the economies of the Asia Pacific region closer. For instance APEC is designed to promote free trade, while ASEAN and ASEAN plus 3 are more focused on free
trade and, economic and financial integration within the member countries of such blocks. Similarly ASEAN plus 6 and the East Asia Summit also have some economic, trade and finance objectives that are ultimately designed to bring a large number of countries economically and financially closer together. Given the importance of Asia and the massive amount of foreign capital flowing to this region and the increasing volume of trade, such cooperation and information sharing amongst policy makers with respect to the operations of some of the G-SIFs and regional and domestic SIFs will be crucial over time.
Regarding the supervision and coordination of G-SIFs and some of the regional and domestic SIFs, Caruana (2013b) has also stated ‘‘this kind of supervision requires not only a new and different kind of expertise but, even more importantly, the capacity and willing- ness to act under significant uncertainty. These include the ability to conduct group-wide consolidated supervision and to challenge banks’ business models, their corporate strategy, governance, risk profiles, ROE targets and capital plans. Also necessary will be the willingness to exercise judgment and to act pre-emptively under conditions of uncertainty’’.
3. Considerations for modelling policies to address systemic risk
3.1. Systemic risk measurement
Post-crisis, a large literature has started to build up, which attempts to incorporate macroprudential policy into canonical macro-models, generally of the micro-founded Dynamic Stochastic General Equilibrium modelling (DSGE) type or variants of these. These papers have therefore had to add some kind of financial sec- tor, so that the policy has something to operate upon. This exercise necessarily requires taking a position on asset and credit dynamics, but these relationships are generally less well understood than the inflation and real-side dynamics that are more relevant for analysis of monetary policy questions.
One of the problems with this line of research is that – unlike inflation – there is little welfare-theoretic basis for using typical measures of asset prices and credit as targets of macroprudential or any other policy. In the monetary policy literature, inflation (or inflation variability) maps into social welfare, either directly because agents are presumed to dislike inflation, or indirectly because price volatility is costly for output. For example, earlier generations of literature on monetary policy assumed an objective function for the policy maker that was a weighted average of infla- tion volatility and output volatility; this formulation, while not exactly corresponding to social welfare, turns out to bear close resemblance to it (Woodford, 2010).
In contrast, asset prices and credit do not enter into welfare functions in a way that suggests that their growth should be resisted by some arm of public policy. To the extent that wealth enters into welfare, one could argue that policies that increase asset prices are desirable. Indeed, some micro-founded models produce the result that under-borrowing, rather than over-borrowing, is the more pertinent issue, and that macroprudential policies could in fact be welfare-reducing (Benigno et al., 2013). More fundamen- tally, in the presence of credit constraints coming from information asymmetries, it is far from clear that ‘normal’ levels of credit are in fact optimal. This observation in no way contradicts the empirical observation that credit booms can be harmful for the subsequent evolution of output, because they can spark a financial crisis (Dell’Ariccia et al., 2012).
Even the internationally agreed prudential rules acknowledge this: rather than mandating a particular level of credit or asset prices or growth rate of these series, the Basel Committee’s guid- ance on application of the countercyclical capital buffer promoted
0
15
30
45
60
0
15
30
45
60
% Share of assets, June 2013
%
Banking groups
Four major Australian banks
G-SIBs
Fig. 1. Trading assets and securities of the largest banking groups (includes derivative assets; sample of 100 banking groups; latest available ratios have been used where June 2013 data are unavailable). Sources: FSB; RBA; SNL Financial; The Banker; banks’ annual and interim reports.
178 L. Ellis et al. / Journal of Banking & Finance 45 (2014) 175–181
(but did not require) the use of a ‘buffer guide’ derived from devi- ation from trend of the ratio of credit to GDP (BCBS, 2010). There are many legitimate reservations about this guide and how it is for- mulated (see, for example Edge and Meisenzahl, 2011 and Box C in APRA and RBA, 2012), but the important point in this context is that it is understood even by the proponents of the buffer guide that there is no welfare-theoretic basis for a fixed level of the ratio of credit to GDP. This ratio can trend over long time horizons, and it can have different sustainable levels in different countries, depend- ing on such factors as the degree of financial development and demographic structure. Even substantial deviations from trend do not necessarily warrant a costly policy to prevent them. Dell’Ariccia et al. (2012) found that around one in three credit booms were followed by a banking crisis, which necessarily implies that around two in three were not.
This disconnect between these variables and welfare-theoretic measures indicates that there is a deeper goal being pursued. What financial stability policies – whether labelled ‘macroprudential’ or otherwise – are intended to do is limit systemic risk. Systemic risk is generally defined as the risk that the financial system might experience a generalised collapse or other form of distress. The rea- son for limiting this risk is to avoid the attendant harm on the real economy. The ultimate target is therefore output. In other words, asset prices and credit growth are useful information variables and possibly even intermediate targets of policy, much as mone- tary aggregates were in an earlier era. They are not, however, ulti- mate targets in same the way as inflation and output are targets of macroeconomic policy.
The above line of argument naturally raises the question of how one might operationalize a policy to limit systemic risk given that a risk is inherently unobservable – only outcomes are observable. Unfortunately, there is no single operational definition of systemic risk and possibly never will be. Bisias et al. (2012) identified 31 dif- ferent measures of systemic risk, emphasising a range of aspects of systemic risk and potential channels to financial distress. Among these are measures that capture deviations from trend in asset prices or credit, much as Dell’Ariccia et al. (2012) do, and measures that stress test the effects of large declines in these and related variables. Other measures focus on the propagation of distress from one financial institution to another, such that an idiosyncratic failure ends up causing a systemic problem. Dimensions of vulner- ability that are emphasised by the various measures include corre- lated exposures, network effects where the default of one party stresses its creditors, illiquidity and leverage.
Indeed, as Bisias et al. (2012) point out, it is probably not desir- able to have a single measure of systemic risk as the focus of policy, as this may result in a ‘Maginot Line’ situation where vulnerabili- ties building in some other part of the financial system are missed. This has obvious implications both for the approach to actual financial stability policy, and how that policy is captured in theo- retical models. In either situation, a single instrument, single target analogue to monetary policy is unlikely to be able to generate use- ful results.
3.2. Systemic risk and diversity
As is well known, the financial system performs a number of functions that are essential to the workings of a modern economy. As well as intermediating between savers and borrowers, the financial sector provides a conduit for payments and enables risk transfer.
These are a diverse set of functions and it should therefore be no surprise that the activities of the financial system give rise to a diverse set of risks. For example, banks and similar entities inter- mediate between borrowers and savers; fund managers also help savers deploy their funds, but without interposing their own
creditworthiness in place of that of borrowers or equity issuers, as is the case for banks. Insurance companies allow customers to transfer diversifiable risks to them, such as property losses from fire, flood and theft, as well as life insurance. Many of the insured risks that the financial system helps to transfer are themselves risks created within the financial system. This is consistent with the principle that financial losses are easier or more attractive to insure because they can be fully compensated by a financial pay- out. (An example of this is that people are more likely to insure their homes against loss from fire or flood than they are to insure their pets’ lives against the same risks: a financial payout may not be seen as a true compensation for such a loss.)
Even within the banking industry, banks serve different func- tions and therefore face different risks. Another way of saying this is that banks’ business models can vary. One way to think about bank business models is as a spectrum with investment banks at one end, commercial or retail banks at the other, and so-called ‘universal’ banks in the middle. Those labels correspond to differ- ent balance sheet structures: investment banks have larger hold- ings of securities and other trading assets, while commercial banks’ assets are mainly loans. As Fig. 1 shows, the importance of trading assets varies greatly within the largest global banking groups, and even within the 29 globally systemically important banks (G-SIBs) identified by the Basel Committee on Banking Supervision and the Financial Stability Board (FSB, 2013a, 2013b, 2013c), shown here as darker bars. While there are a cluster of banks with investment-banking business models in this group of G-SIBs, G-SIBs are represented across the whole range of observed balance sheet structures. In contrast, the four large Australian banks, shown as lighter bars in Fig. 1, for example, are all clearly commercial banks, with relatively little trading business.
Several conclusions can be reasonably drawn from these obser- vations. One is that if even the banking industry, let alone the whole financial system, is relatively diverse, it should be expected that the risks individual institutions face are also diverse. It cannot be assumed that the appropriate capitalisation is constant across all risks. This is a powerful argument against a single leverage ratio as the predominant instrument of prudential regulation. Even though there are concerns about the complexity of the current pru- dential framework, this suggests that risk-weighting is and should be here to stay. While leverage ratios are a useful backstop mea- sure and guard against potential gaming of risk-weights, their appropriate role is as a backstop.
The diversity within the financial system also supports the point made by Bisias et al. (2012) and mentioned earlier, that a single measure of systemic risk is unlikely to be universally
L. Ellis et al. / Journal of Banking & Finance 45 (2014) 175–181 179
applicable, and neither is a single instrument of financial stability policy. It will not be appropriate to conceive of macroprudential policy (or financial stability policy more generally) in the single instrument, single target framework that is generally used to ana- lyse monetary policy. Different aspects of prudential regulation will bear more effectively on particular risks and aspects of sys- temic risk than others. Rather, issues of macroprudential policy may be best thought of as a design problem in calibrating the entire prudential framework. Separating out particular aspects of prudential regulation into specific macroprudential ‘tools’, possibly with different governance from the prudential framework more broadly, may lead to coordination problems.
A more subtle issue arises in considering the niches that banks with different business models serve. Although there has been some debate about the social utility of some financial activity, and attempts such as the Volcker Rule to limit proprietary trading by banks, it is not the case that the optimal share of trading assets in bank balance sheets is zero. Neither is it the case that the opti- mal share of these assets would be the same across all banks. It depends on their business models and the particular risks they take on. Commercial banks are principally involved in intermedia- tion between savers and borrowers from the non-financial sectors. In doing so, though, they may incur risks that may be prudently laid off to other parts of the financial sector that may be able to diversify those risks more effectively. Among these risks are inter- est rate risk and FX risk: even quite plain-vanilla retail banking can involve an array of market risks depending on how diversified the bank’s funding base is and how global the business of the bank’s commercial customers. At least some of the trading assets of banks with more investment-banking business models arise through their interbank exposures built up as counterparties to commercial banks. These counterparty relationships can cross borders. Thus even in a national financial system like Australia’s, where the domestically owned banks are primarily commercial or retail ori- ented, the gamut of financial services can be provided, and risks transferred, because there are other, generally foreign banks, that can serve as counterparties to local banks’ hedging activity.
This implies that, although interconnectedness within the financial system can increase systemic risk, there is a limit below which this interconnectedness cannot prudently fall. In particular, although cross-border contagion is frequently identified as a risk to national financial stability, and inter-bank borrowing often a pre- cursor to financial crisis, a position of financial autarky is not the desirable equilibrium point, either. This has implications for mea- sures of systemic risk and crisis vulnerability, which should per- haps not be seen as linear indicators.
4. Bank governance and systemic risk
4.1. Five planks of effective reform
There are five key planks of any well-defined regulatory regime: entry – that is, competition policy; exit – that is resolution policy: regulation – that is, supervisory policy; accounting – that is, audit- ing and valuation policy; and governance. In the light of the crisis, there have been intense efforts to reform the first four of these pil- lars. For example, new regimes are being put in place for competi- tion (e.g. Parliamentary Commission on Banking Standards, 2013), regulation (e.g. Basel Committee on Banking Supervision, 2010a, 2010b), resolution (e.g. Financial Stability Board, 2011a, 2011b) and accounting (e.g. International Accounting Standards Board, 2013) within banking firms, nationally and internationally.
Yet the fifth pillar – governance – has been largely untouched. That is puzzling. Academic theory has long suggested a strong link between governance and risk-taking (Jensen and Meckling, 1976).
And recent empirical evidence suggests this link may be especially strong for financial firms (Beltratti and Stulz, 2009; Ferreira et al., 2012; Laeven and Levine, 2009). Certainly, the financial crisis unearthed bank governance failures operating at multiple levels: managers failing to control risk-takers, boards failing to control managers and investors failing to discipline either managers or boards.
Governance is also a potentially potent risk-mitigation tool. It has the potential to shape the risk-taking incentives of owners and managers. If reform can act on these incentives at source, it reduces the chances of regulators chasing risk around the system (Haldane, 2012). In other words, governance reform may be among the most effective ways of curbing risk-taking without engender- ing regulatory arbitrage.
4.2. Risk-taking incentives in banking
Although all firms face governance problems, there are three good reasons for believing these may be more acute in banking than in other sectors.
First, public companies have, for well over a century, benefitted from limited liability. This means that payoffs to equity-holders can mimic an out of money call option on its assets, with a strike price given by its debt liabilities (Merton, 1974). This payoff asym- metry shapes risk-taking incentives in potentially important ways. For example, the value of equity can then be boosted by increasing the variability of equity payoffs, either by investing in riskier assets or by leveraging those assets. This shifts risk up the capital struc- ture to the detriment of debt-holders (Jensen and Meckling, 1976). In other words, limited liability gives rise to a principal- agent problem between shareholders and debt-holders.
These incentive problems are likely to be more acute in banking than in other sectors: first, because risk is a deliberate choice var- iable for banks by dint of their choice of assets; and second, because their much greater leverage increases balance sheet risk and thereby the payoff asymmetry. These problems have long been recognised. They are the reason why limited liability came later to banking than other sectors (Crick and Wadsworth, 1936). And they are why double or even treble liability persisted in banking long after unlimited liability had been abolished (Haldane, 2012).
There is a second, well-known principal-agent problem within many firms, namely between shareholders and managers (Jensen and Meckling, 1976). This might arise because managers put their own objectives over those of shareholders. One means of aligning incentives between the two is to remunerate bank managers in equity. That has become a widespread practice during recent years, particularly within the financial sector. To take a striking example, in 2006 the typical bank CEO’s wealth rose by $1 million for every 1% increase in the value of their firm (Fahlenbrach and Stulz, 2011).
Compensating managers in equity is not, however, costless. By aligning managerial incentives with shareholders, the risk-shifting problem is potentially exacerbated. This includes incentives to ‘‘gamble for resurrection’’ when firms are nearing insolvency. Evi- dence during the crisis is revealing here. In 2006, the largest per- centage equity stakes by US bank CEOs were held in Lehman Brothers, Bear Stearns, Merrill Lynch, Morgan Stanley and Country- wide (Fahlenbrach and Stulz, op. cit.). In each case, this may have induced managerial incentives to gamble for resurrection – in each case unsuccessfully. In short, in solving one principal/agent prob- lem (between managers/shareholders), equity-based pay may have worsened another (between shareholders/debt-holders).
A third incentives issue is moral hazard. In principle, if risk is shifted to debt-holders they ought to seek compensation through higher yields. That, in turn, would impose a degree of discipline on shareholder/manager incentives to risk-shift (Haldane, 2012). In practice, evidence of debt-holders having exercised discipline
180 L. Ellis et al. / Journal of Banking & Finance 45 (2014) 175–181
consistently over bank risk-taking is scarce, including in the pre- crisis period (Flannery and Sorescu, 1996; Krishnan et al., 2005). One possible reason is that bank debt-holders’ risk senses tend to be dulled by insurance, either explicit in the case of bank deposi- tors or implicit in the case of bank bondholders. During this and previous crises, both classes of debt were effectively underwritten by the state to protect the wider economy from disorderly bank runs (Haldane, 2010).
But (implicit or explicit) insurance of bank debt has two unfor- tunate side-effects. First, it removes the market disciplining effect of debt which might otherwise damp risk-shifting incentives. Sec- ond, it shifts the burden of risk from debt-holders onto the state, and hence taxpayers, with a corresponding deadweight cost. In other words, a second principal-agent problem emerges – between bank investors and society as a whole (Haldane, 2010). This is not an incentive problem felt as acutely outside banking where the col- lateral costs of firm failure, and hence the probability of state insurance, are lower.
4.3. Reforming risk-taking incentives in banking
If these risk-taking incentives are potent, which crisis experi- ence suggests may have been the case, what might be done to counter them? There are various reform options on the table, which speak to each of the different dimensions of the incentive problem.
First, the regulatory capital base of banks could be increased. This reduces incentives to generate a leverage-induced rise in equity returns (Admati and Hellwig, 2013). Or, put differently, it reduces the amount of risk that will be shifted onto debt-holders. Regulatory reform initiatives, such as Basel III, augment regulatory capital standards and so would tend to reduce the principal-agent problem between shareholders and debt-holders. Nonetheless, there is an open debate about whether levels of capital in the sys- tem, even after Basel III, will be adequate to reshape fundamentally risk-shifting incentives (Admati and Hellwig, op.cit.).
Second, the compensation structure of managers could be reformed. For example, incentives to risk-shift to debt-holders would be diluted or internalised by paying managers in long-term debt, rather than cash or equity (Edmans and Liu, 2011). And incentives to gamble for resurrection could be reduced by having a lengthy period over which rewards are deferred (ex-ante) or clawed-back (ex-post) in the event of risks subsequently material- ising. Some progress has been made in enacting these reforms by regulators since the crisis, under the auspices of the Financial Stability Board (2009, 2012a, 2012b, 2012c). But, as with bank cap- ital, there is an open debate about whether existing international rules go sufficiently far in reshaping risk-taking incentives (Squam Lake Working Group on Financial Regulation, 2010; Parliamentary Commission on Banking Standards, 2013). The extent to which individual countries apply these rules consistently has also been queried (FSB, 2012a, 2012b, 2012c).
Third, efforts could be made to encourage debtor discipline, and hence discourage risk-shifting, by reducing the probability of gov- ernments needing to provide support to banks during crisis. One way this could be done is by putting in place resolution regimes which offer the credible prospect of ‘‘bailing-in’’ creditors in the event of stress (Coeuré, 2013). A number of countries have, or are in the process, of putting in place such regimes, again under the auspices of the Financial Stability Board (2011a, 2012a, 2012b, 2012c). As with the other initiatives, the jury remains out on the practical impact this will have on risk-taking incentives.
There is a final way in which incentives of bank stakeholders, broadly-defined, might be better aligned at source with societal preferences. That would be by reforming the structure of company law – for example, by extending control rights beyond
shareholders. The case for doing so seems especially strong in banking where the imbalance between shareholder control (100% of the balance sheet) and their stake (typically less than 5% of the balance sheet) is so significant. Some reform suggestions have been made in this direction (Parliamentary Commission on Banking Standards, 2013), but so far these have made relatively lit- tle headway.
5. Conclusion
The aim of this paper is to analyse some of the issues with respect to systemic risk, governance and global financial stability. The paper discusses the work of the Financial Stability Board (FSB) with respect to global, systemically important financial insti- tutions (G-SIFs) and domestic systemically important financial institutions (D-SIBs). Given the increasingly interconnected nature of the global financial system, this paper welcomes the attempt by the FSB and central banks and supervisors to facilitate the sharing of firm level data on systemically important financial institutions.
The paper argues that one of the issues, with respect to the glo- bal regulatory framework, is to ensure that certain standards are globally accepted and implemented. This is because, as Caruana (2013a) stated, if we do not control the tendency to fragmentation and also regulatory arbitrage, we could lose the benefits of global- isation and the pace of process of financial globalisation that has accelerated due to technological changes and the increasing inter- dependence of national economies. Furthermore, consistency must be achieved across the international stage. Rules and minimum standards should not be different based on a bank’s nationality. If some banks are held to a lower standard than those internationally agreed upon, this will result in an uneven playing field.
The paper also argues that given the diversity of the whole financial system, it should be expected that the risks individual financial institutions face are also diverse. It cannot be assumed that the appropriate capitalisation is constant across all risks. While leverage ratios are a useful backstop measure and guard against potential gaming of risk-weights, their appropriate role is as a backstop. The diversity within the financial system also sup- ports the fact that a single measure of systemic risk is unlikely to be universally applicable, and neither is a single instrument of financial stability policy.
The paper argues that while there have been analyses of banks competition policies, resolution policies, supervisory policies and auditing and valuation policies, less attention has been paid to the role of bank governance and systemic risk, despite a strong link between governance and risk-taking. The paper offers four solu- tions to strengthen bank governance. First, the regulatory capital base of banks could be increased. This reduces incentives to gener- ate a leverage-induced rise in equity returns. Second, the compen- sation structure of managers could be reformed. For example, incentives to risk-shift to debt-holders would be diluted or interna- lised by paying managers in long-term debt, rather than in cash or equity. Third, efforts could be made to put in place resolution regimes which offer the credible prospect of ‘‘bailing-in’’ creditors in the event of stress. There is a final way in which incentives of bank stakeholders, broadly-defined, might be better aligned at source with societal preferences. This would be by reforming the structure of company law – for example, by extending control rights beyond shareholders.
Disclaimer
The views expressed by Luci Ellis and Andy Haldane in this arti- cle are theirs and not necessarily those of the Reserve Bank of Australia and the Bank of England respectively. Luci Ellis would like to thank Elliott James for his valuable research assistance work.
L. Ellis et al. / Journal of Banking & Finance 45 (2014) 175–181 181
Fariborz Moshirian would like to acknowledge the support of the Australian Research Council (RG124356) for this project. Fariborz Moshirian would like to thank Christopher Wong and Jeffrey Chen for their research assistance work on this project. All errors remain the authors responsibility.
References
Admati, A., Hellwig, M., 2013. The Bankers’ New Clothes: What’s Wrong with Banking and What to Do About it. Princeton University Press.
Arnold, B., Borio, C., Ellis, L., Moshirian, F., 2012. Systemic risk, macroprudential policy frameworks, monitoring financial systems and the evolution of capital adequacy. Journal of Banking and Finance 36, 3125–3132.
Australian Prudential Regulation Authority and Reserve Bank of Australia, 2012. Macroprudential Analysis and Policy in the Australian Financial Stability Framework. September, Australia. <http://www.apra.gov.au/AboutAPRA/ Publications/Documents/2012-09-map-aus-fsf.pdf>, <http://www.rba.gov.au/ fin-stability/resources/2012-09-map-aus-fsf/index.html>.
Bank for International Settlements, 2013. Global Systemically Important Banks: Updated Assessment Methodology and the Higher Loss Absorbency Requirement. BIS, Basel.
Basel Committee on Banking Supervision, 2010a. Basel III: A Global Regulatory Framework for More Resilient Banks and Banking Systems. BCBS, Basel.
Basel Committee on Banking Supervision, 2010b. Guidance for National Authorities Operating the Countercyclical Capital Buffer. BCBS, Basel. <http://www.bis.org/ publ/bcbs187.pdf>.
Beltratti, A., Stulz, R.M., 2009. Why did Some Banks Perform Better during the Credit Crisis? A Cross-country Study of the Impact of Governance and Regulation. NBER Working Paper Series No. 15180.
Benigno, G., Chen, H., Otrok, C., Rebucci, A., Young, E.R., 2013. Financial crises and macro-prudential policies. Journal of International Economics 89, 453–470.
Bisias, D., Flood, M., Lo, A.W., Valavanis, S., 2012. A Survey of Systemic Risk Analytics. Office of Financial Research Working Paper #0001.
Caruana, J., 2013a. Regulatory landscapes: consistent regulatory implementation to keep markets integrated. In: Speech at the 2013 International Monetary Conference, Shanghai, 3 June.
Caruana, J., 2013b. Addressing risks to financial stability. In: Speech to the 49th SEACEN Governors’ Conference and High-level Seminar, Kathmandu, 21 November.
Coeuré, B., 2013. The implications of bail-in rules for bank activity and stability. Speech at ‘Financing the Recovery after the Crisis – The Roles of Bank Profitability, Stability and Regulation. Bocconi University, Milan, 30 September.
Crick, W.F., Wadsworth, J.E., 1936. A Hundred Years of Joint Stock Banking. Hodder & Stoughton.
Dell’Ariccia, G., Igan, D., Laeven, L., Tong, H., 2012. Policies for Macrofinancial Stability: How to Deal with Credit Booms. IMF Staff Discussion Note SDN 12/06.
Edge, R.M., Meisenzahl, R.R., 2011. The unreliability of credit-to-GDP ratio gaps in real-time: implications for countercyclical capital buffers. Federal Reserve Board Finance and Economics Discussion Series, 2011–2037.
Edmans, A., Liu, Q., 2011. Inside debt. Review of Finance 15, 75–102.
Fahlenbrach, R., Stulz, R.M., 2011. Bank CEO incentives and the credit crisis. Journal of Financial Economics 99, 11–26.
Ferreira, D., Kershaw, D., Kirchmaier, T., Schuster, E.-P., 2012. Shareholder Empowerment and Bank Bailouts. London School of Economics Financial Markets Group Discussion Paper No. 714.
Financial Stability Board, 2009. Principles for Sound Compensation Practices. BIS, Basel.
Financial Stability Board, 2011a. Policy Measures to Address Systemically Important Financial Institutions. BIS, Basel.
Financial Stability Board, 2011b. Macroprudential Policy Tools and Frameworks Progress Report to the G20. BIS, Basel.
Financial Stability Board, 2012a. First Progress Report on Compensation Practices. BIS, Basel.
Financial Stability Board, 2012b. Identifying the Effects of Regulatory Reforms on Emerging Market and Developing Economies: A Review of Potential Unintended Consequences. BIS, Basel.
Financial Stability Board, FSB 2012c. Update of Group of Global Systemically Important Banks (G-SIBS). BIS, Basel.
Financial Stability Board, 2013a. 2013 Update of Group of Global Systemically Important Banks (G-SIBs). <http://www.financialstabilityboard.org/ publications/r_131111.pdf>.
Financial Stability Board, 2013b. Global Systemically Important Insurers (G-SIIs) and the Policy Measures that will Apply to Them. BIS, Basel.
Financial Stability Board, 2013c. Thematic Review on Resolution Regimes. BIS, Basel. Flannery, M.J., Sorescu, S.M., 1996. Evidence of bank market discipline in
subordinated debenture yields: 1983–1991. Journal of Finance 51, 1347–1377. Haldane, A.G., 2010. The debt hangover. In: Speech Given at a Professional Liverpool
Dinner, 27 January, Liverpool. Haldane, A.G., 2012. Discussion of Financial instruments, financial reporting, and
financial stability by Christian Laux. Accounting and Business Research 42, 261– 266.
International Accounting Standards Board, 2013. Financial Instruments: Expected Credit Losses. ED/2013/3.
Jensen, M.C., Meckling, W.H., 1976. The theory of the firm: managerial behaviour, agency costs and ownership structure. Journal of Financial Economics 3, 305– 360.
Krishnan, C.N.V., Ritchken, P.H., Thompson, J.B., 2005. Monitoring and controlling bank risk: does risky debt help? Journal of Finance 60, 343–378.
Laeven, L., Levine, R., 2009. Bank governance, regulation and risk taking. Journal of Financial Economics 93, 259–275.
Merton, R.C., 1974. On the pricing of corporate debt: the risk structure of interest rates. The Journal of Finance 29, 449–470.
Moshirian, F., 2011. The global financial crisis and the evolution of markets, institutions and regulation. Journal of Banking and Finance 35, 502–511.
Moshirian, F., 2012. The future and dynamics of global systemically important banks. Journal of Banking and Finance 36, 2675–2679.
Parliamentary Commission on Banking Standards, 2013. Changing Banking for Good – Summary, Conclusions and Recommendations, vol. I. June.
Squam Lake Working Group on Financial Regulation, 2010. Regulation of Executive Compensation in Financial Services. February.
Woodford, M., 2010. Optimal monetary stabilization policy. In: Friedman, B.M., Woodford, M. (Eds.), Handbook of Monetary Economics, vol. 3, pp. 723–828 (Chapter 14).
- Systemic risk, governance and global financial stability
- 1 Introduction
- 2 Large financial institutions and propagation of systemic risk
- 2.1 Identification of systemically important financial institution
- 2.2 Macroprudential policies and a global approach to financial stability
- 3 Considerations for modelling policies to address systemic risk
- 3.1 Systemic risk measurement
- 3.2 Systemic risk and diversity
- 4 Bank governance and systemic risk
- 4.1 Five planks of effective reform
- 4.2 Risk-taking incentives in banking
- 4.3 Reforming risk-taking incentives in banking
- 5 Conclusion
- Disclaimer
- References
