International Banking
youeef
ib_location_14.pdf
International Banking
The Location Decision
2
Learning outcomes
Comprehend the extent to which banking is international
Review advantages and disadvantages of international banks and barriers to international bank expansion
Identify factors that determine a banks’ decision to branch or establish an overseas subsidiary
Use empirical evidence to see the determinants of the demand for bank services Concierge & home cookin’ hypotheses
Use empirical evidence to demonstrate the concepts of and measure the effects of distance and competitor-remoteness
3
Overview
International banking was very common at end 19th century (financing trade/colonies); re-emerges in 1970s (trading blocs, financial innovation); recent historical peak (cross-border M&A)
How do banks expand abroad?
Directly e.g. syndicated lending; via branching and/or subsidiaries “Ring fencing” – insulate the main banking organisation from country risk
Exploit safety-net arrangements in host nation; control agency problems
Relatively low foreign bank penetration in developed nations (around 10%)
Much higher levels in developing nations …
Latin America, 40%
Some Asian & African countries, > 50%
Some central & eastern Europe countries, 90%
Rationale for international banking
Trade barrier theories
If trade costs are high, exporting becomes costly establish foreign
operation … gain market access (horizontal FDI)
Arbitrage and cost of capital
Borrow in cheap currency markets and lend in more expensive
The carry trade (e.g. borrow in Japan and lend in the US)
Ownership advantages and other bank-specific advantages
Alleged superiority of foreign bank processes and management skills
Scale and scope economies of large internationally active banks
Diversification – geography, product
Aim to diversify earnings and lower risk
Location benefits of international financial centres
4
5
Foreign penetration in Europe, 2004
Nation Assets (€bn) %, EEA %, non-EEA % foreign
UK 6,970 26.39 24.87 51.26
Germany 6,584 4.91 1.37 6.28
France 4,415 9.39 1.97 11.36
Ireland 723 36.43 9.02 45.44
Luxembourg 695 89.50 4.59 94.09
EU15 \ UK 21,616 13.34 1.61 14.94
Poland 132 58.79 8.66 67.45
Czech Rep 87 86.86 4.96 91.82
A10 424 60.11 4.95 65.06
Source: adapted from Berger (2007, Table 1, p. 1959)
6
Foreign banks’ comparative advantages
FBs benefit from being able to (geographically) diversify risk Can profit via higher revenues/lower costs
“Follow-your-customer” strategy i.e. offer services to existing MNE clients in countries where MNEs have foreign affiliates
Foreign banks from developed countries may possess skills & expertise, access to technology & capital, well educated labour force, market power over suppliers, ability to make larger loans
Foreign banks from well regulated developed countries may provide credit smoothing and financial stability during crises
7
Foreign banks’
comparative disadvantages
Operating diseconomies of distance from home headquarters Differences in language, culture, economic development
Persuading managers to work overseas, measuring & monitoring their performance
All of the above raise costs for internationally active banks
“Soft” information is more difficult for foreign banks to process when operating at distance i.e. concierge effect Large international banks prefer to produce hard information using
techniques like credit scoring to price risk (lending rates). Reliance on information processing technology, which implies …
Foreign banks may ration credit to informationally opaque market segments e.g. SMEs
Easier to transmit soft information from subsidiary to parent when the distance between host and home is close (geographically, culturally)
8
Government barriers to foreign bank competition
Explicit barriers
Rules and regulations limiting the entry and behaviour of FB or treating FB differently from domestic banks
Trend towards lowering this barrier over time … EU Single Market Programme, Monetary Union, single currency,
further moves towards Banking Union are continuing
India – FB buying shares in local banks limited to 10% of voting rights, with additional capital requirements & permission required for branching expansion
China – traditional restrictions on FB minority ownership, activities & branching. WTO agreement removed this with effect from 2007
9
Government barriers to foreign bank competition
Implicit barriers do not treat foreign banks differently from local banks 1. Rules & regulations governing banks/environment differ from FB home country
Regulation, supervision; accounting/information; payment/settlements; legal system; tax; currency
2. Actions of government prevent FB entry & expansion in favour of local banks Delaying, denying FB M&A; national “champions” i.e. creating larger local banks via
M&A lowers risk of FB M&A
3. Government ownership of local banks and provision of subsidies to them … Evidence shows state-owned banks are relatively inefficient but they may “crowd
out” private-owned banks because of subsidised lending at below market rates to preferred economic sectors, regions, new entrepreneurs
Subsidies may mean state-owned banks skimp on monitoring their borrowers, and enforcing the repayment of loans
Typically, FB lend to large, corporate customers FB is less crowded out c.f. private, local banks; but, if state ownership is high, both types of bank may be crowded out
10
Foreign & state-owned market shares; by region
Source: Berger (2007, Fig 1, p. 1966)
11
Foreign & state-owned market shares; by country (%)
High (FO/SO) Foreign State Low (FO/SO) Foreign State
Transition Eastern Europe
Czech 90% 4% Russia 9% 36%
Romania 47% 54% Hungary 89% 9%
Latin America
Mexico 83% 0% El Salvador 12% 4%
Brazil 30% 32% Peru 42% 0%
Asia
Jordan 64% 0% Kuwait 0% 0%
China 2% 73% Japan 7% 0%
Africa
Lesotho 100% 0% Sudan 4% 12%
Egypt 13% 65% Botswana 100% 0%
Source: adapted from Berger (2007, Table 2, p. 1968)
12
Determining internationalisation
Economic integration … Internationalisation is correlated with the level of integration
between the home country of an international bank and the host country of the branch &/or subsidiary
Institutional characteristics … Affects the probability of foreign bank entry e.g. limits on bank
scope; regulations against foreign banks; national champions
Profit opportunities … Size (larger banks more likely to operate internationally)
Home country features (highly developed financial markets comparative advantage incentives to expand overseas)
13
Internationalisation – results I
Characteristics of bank & country of origin …
Larger and more diversified banks go abroad
Banks from more profitable home nations go
Economic integration between countries …
Internationalisation with bilateral trade
Bank expansion negatively related to distance
Expansion to countries sharing language and culture
14
Internationalisation – results II
Institutional characteristics …
Implicit/explicit barriers play a crucial role
Heavily regulated and more concentrated banking systems are less accessible for international banks
Nations with stronger legal systems, & offshore financial centres, are more likely to attract international banks
Profit opportunities …
The growth and development of host nations is important
IBs attracted by relative inefficiencies in host banking systems e.g. higher costs & potential to earn abnormal profits
15
Dimensions of globalisation
Berger et al (2003) use survey to identify which banks do 2000 affiliates of large MNEs operating in Europe choose for their cash management services
1. Host nation bank – headquartered in the same nation in which an MNE affiliate operates
“Concierge effect” – the foreign affiliate of a MNE may prefer to use a host nation bank for concierge services i.e. to secure the concierge’s knowledge of local market conditions (regulations/culture) in the host nation
2. Home nation bank – HQ in same nation as the MNE HQ “Home cookin’ effect” – a home bank knows the market conditions in the
affiliates home nation; home bank may serve MNE HQ yielding relationship benefits to MNE as a whole
3. Third nation bank – HQ in neither home or host Neither effect dominates & bank choice reflects other factors
16
Dimensions of globalisation
Bank reach – refers to geographic scope & size of chosen bank
Global bank – operates across nations, world’s largest institutions Offers broadest range of financial services; expert in many foreign
markets; able to transact large deals; may offer superior stability due to risk diversification &/or implicit government protection against closure
Local bank – operates in a single nation Focuses on establishing close relationships with affiliates; may have
specific local market knowledge
Regional bank – operates between the two extremes Other affiliates may prefer a trade-off between the above banks
17
Firm choice of bank nationality
and reach, by host nation (%)
Bank nationality (%) Bank reach (%)
Nation Host Home Third Global Regional Local
UK 52.2 29.0 18.8 25.9 71.9 2.2
France 76.7 12.1 11.2 66.8 17.0 16.1
Germany 73.3 14.2 12.5 40.0 49.2 10.8
Sweden 85.3 9.2 5.5 11.0 79.8 9.2
Greece 40.0 20.0 40.0 45.0 32.5 22.5
Czech R. 28.6 28.6 42.9 42.9 28.6 28.6
Poland 28.3 26.7 45.0 50.0 21.7 28.3
Hungary 18.8 43.8 37.5 40.6 40.6 18.8
Source: adapted from Berger et al (2003, Table 2, p. 392)
18
What determines firm choice?
Firms choose bank nationality and then reach
65.5% of firms select a host nation bank (concierge)
34.5% choose home & third nation banks (> 60% select global banks)
Decisions are a function of variables capturing …
Host country characteristics
Differences between home and host countries
Corporate attributes
19
Results on bank nationality
Host country characteristics … Firms operating in former Soviet countries less likely to use host
nation banks, & more likely to use home or third nation banks i.e. some banking sectors too under developed to offer concierge services
Differences between home & host countries … As distance between home & host firms more likely to use home or
third nation banks Firms more likely to use third nation banks when home & host do not
share common language, & when size difference between home & host banking sectors is large
Removing US-owned affiliates changes language effect concierge
Corporate attributes … As MNEs become larger & more pan-European, they are more likely to
use home country banks less demand for concierge services or firms adopting a follow the customer strategy
20
Results on bank reach
Conditioned on choosing a host nation bank … Firms are more likely to use regional banks (re local) as firm size rises
Firms are less likely to use global banks as the number of countries a firm operates in increases
Conditioned on choosing a home nation bank … Firms are more likely to use global banks as firm size and the number
of countries it operates in increases
Conditioned on choosing a third nation bank … Little variation in reach is explained by variables
Concept of competitor remoteness
We know bank location decisions reflect
Expected growth on host country and local bank inefficiencies
Costs and regulations; bilateral trade, FDI flows, investments
Institutional, cultural closeness as well as geographic proximity
None of the above considers the competition of other investors as a factor to explain the location decision
When deciding where to locate, a bank should consider the attractiveness of the new market to other competitor banks; if competition is high, it might be better to invest elsewhere
This implies location is driven by bilateral distance and competitor-remoteness 21
Some summary information
Over 1995-2009, 560 foreign bank entries took place with these banks offering a broad range of products & services
Foreign bank penetration (share of numbers of FBs) in local banking systems has grown from 20% to 34%
But, the variance of bank ownership remains high
Some local banking sectors are 100% foreign-owned
As at 2009, foreign banks active in 126 out of 137 countries
After exclusions, 1,199 FBs from 75 homes in 110 hosts
Total possible home-host combinations is 8,185 pairs
On average, 11 FBs from 6 different homes present in a host
Luxembourg is the largest host hosting 73 FBs
UK is the host with FBs from the most home countries (22)
On average, a home country has its banks invest in 9 hosts
US has the most FB investments spread among the most hosts (126 FBs in 58 countries) 22
Host country coverage
Host
country
Number of
FBs
% Share FB,
Number
% Share FB,
Assets
No. home
countries in
Armenia 11 73 64 6
Brazil 54 38 22 16
China 19 14 2 7
Greece 4 24 10 3
Hong Kong 23 66 89 8
India 7 11 5 4
Indonesia 25 42 13 13
Malaysia 12 30 13 8
UAE 4 21 2 4
UK 49 56 19 22
23
Source: adapted from Claessens and van Horen (2014)
Claessens & van Horen, 2014
Use a gravity model common in the trade literature
FPij = α + β · Dij + γ · CRj + δ · Xij + εij
FPij = foreign bank presence from home country i in host country j in 2009
Dij = bilateral distance between home and host country
CRj = competitor remoteness of host country j
Xij = control variables e.g. home & host characteristics
A negative coefficient on β would confirm importance of distance while a positive coefficient on γ would confirm hypothesis on competitor-remoteness
24
Variables
Constructing competitor-remoteness indicator
𝐶𝑅𝑗 = 𝐷𝑚𝑗/𝑌𝑚𝑚≠𝑖
Dmj = bilateral distance between host country j and each of the potential home countries m other than i
Ym = GDP of respective home country (weighting factor)
A bank is foreign-owned if foreigners hold 50% of shares
Alternative distance measures include distance between home and host regulatory quality (0 – 5, mean = 3.63)
Dummy variable if home and host share same legal system
30% of country pairs share same legal system
Differences in scope, access, quality of credit information
Ranges from 0-6; mean distance = 2.25; large variance
Dummy variable if home and host share same language
8% of country pairs share a common language
25
Distance, competitor remoteness, and the location decision of foreign banks
26
TABLE 2 (1) (2) (3) (4) Distance −0.115*** −0.055*** −0.036*** −0.037*** (0.000) (0.000) (0.000) (0.000) Comp. remoteness 0.096*** 0.050*** 0.035** 0.027** (0.000) (0.002) (0.018) (0.040) GDP host 0.021*** −0.030*** −0.030*** −0.030 (0.000) (0.000) (0.000) (0.000) GDP home 0.060*** 0.016*** −0.022*** (0.000) (0.000) (0.000) Trade 0.047*** 0.043*** 0.042*** (0.000) (0.000) (0.000) GDP/CAP host −0.020*** −0.019 (0.000) (0.000) Foreign banks host 0.067*** 0.063*** (0.000) (0.000) GDP/CAP home −0.001 (0.753) Foreign banks home 0.056*** (0.000) Home country no no no yes Number of observations 8,185 7,503 7,503 7,503
Source: Claessens and van Horen (2014)
Interpretation
Coefficients show marginal effects on the presence of banks from home country i in host country j in 2009
Distance and competitor-remoteness affect location
If a host country is located further away it is less likely to experience foreign investment from a particular home
But, if the host is more remote, it is more likely to receive
Findings hold even after controlling for
Bilateral trade (though trade does capture some effects of distance and remoteness on banking FDI)
GDP per capita in home and host; number of foreign banks present in the host; number of foreign investments by banks from home country; home country fixed effects
27
Some implications …
Both the absolute distance between home & host country & the host’s remoteness drive the location decision
Distance and remoteness in language also important
These factors matter more for non-OECD banks, for investments in non-OECD countries, when the scale of outward & inward banking FDI is relatively small, & when foreign banks hold the majority of assets in a host country
Results show there is more scope for foreign banks, particularly non-OECD, to expand regionally rather than globally, making very diversified forms of banking less likely and/or less profitable
28
