Finance paper assignment

Thegodsstrikeback

A special report on financial risk l February 13th 2010

( The Economist February 13th 2010 A special report on financial risk The gods strike back Also in this section Number crunchers crunched The uses and abuses of mathematical models. Page 3 Cinderella’s moment Risk managers to the fore. Page 6 A matter of principle Why some banks did much better than others. Page 7 When the river runs dry The perils of a sudden evaporation of liquidity.Page 8 Fingers in the dike What regulators should do now. Page 10 Blocking out the sirens’ song Moneymen need saving from themselves. Page 13 )1

Financial risk got ahead of the world’s ability to manage it. Matthew Valencia asks if it can be tamed again

( "T )HE revolutionary idea that defines the boundary between modern

ed by larger balance sheets and greater le verage (borrowing), risk was being capped

Acknowledgments

In addition to those mentioned in the text, the author would like to thank the following for their help in preparing this report: Madelyn Antoncic, Scott Baret, Richard Bookstaber, Kevin Buehler, Jan Brockmeijer, Stephen Cecchetti, Mark Chauvin, John Cochrane, José Corral, Wilson Ervin, Dan Fields, Chris Finger, Bennett Golub, John Hogan, Henry Hu, Simon Johnson, Robert Kaplan, Steven Kaplan, Anil Kashyap, James Lam, Brian

Leach, Robert Le Blanc, Mark Levonian, Tim Long, Blythe Masters, Michael Mendelson, Robert Merton, Jorge Mina, Mary Frances Monroe, Lubos Pastor, Henry Ristuccia,

Brian Robertson, Daniel Sigrist, Pietro Veronesi, Jim Wiener, Paul Wright and Luigi Zingales.

A list of sources is at

Economist.com/specialreports

An audio interview with the author is at

Economist.com/audiovideo

times and the past is the mastery of risk: the notion that the future is more than a whim of the gods and that men and wom en are not passive before nature." So wrote Peter Bernstein in his seminal history of risk, "Against the Gods", published in 1996. And so it seemed, to all but a few Cassan dras, for much of the decade that followed. Finance enjoyed a golden period, with low interest rates, low volatility and high re turns. Risk seemed to have been reduced to a permanently lower level.

This purported new paradigm hinged, in large part, on three closely linked devel opments: the huge growth of derivatives; the decomposition and distribution of credit risk through securitisation; and the formidable combination of mathematics and computing power in risk management that had its roots in academic work of the mid 20th century. It blossomed in the 1990s at firms such as Bankers Trust and JPMorgan, which developed "value at

risk" (VAR), a way for banks to calculate how much they could expect to lose when

things got really rough.

Suddenly it seemed possible for any fi nancial risk to be measured to five decimal places, and for expected returns to be ad justed accordingly. Banks hired hordes of PhD wielding "quants" to fine tune ever more complex risk models. The belief took hold that, even as profits were being boost

by a technological shift.

There was something self serving about this. The more that risk could be cali brated, the greater the opportunity to turn debt into securities that could be sold or held in trading books, with lower capital charges than regular loans. Regulators ac cepted this, arguing that the "great moder ation" had subdued macroeconomic dan gers and that securitisation had chopped up individual firms’ risks into manageable lumps. This faith in the new, technology driven order was reflected in the Basel 2 bank capital rules, which relied heavily on the banks’ internal models.

There were bumps along the way, such as the near collapse of Long Term Capital Management (LTCM), a hedge fund, and the dotcom bust, but each time markets re covered relatively quickly. Banks grew cocky. But that sense of security was de stroyed by the meltdown of 2007 09, which as much as anything was a crisis of

modern metrics based risk management. The idea that markets can be left to police themselves turned out to be the world’s most expensive mistake, requiring $15 tril lion in capital injections and other forms ofsupport. "It has cost a lot to learn howlit tle we really knew," says a senior central banker. Another lesson was that managing risk is as much about judgment as about

numbers. Trying ever harder to capture 1

( A special report on financial risk The Economist February 13th 2010 )2

2 risk in mathematical formulae can be counterproductive if such a degree of ac curacy is intrinsically unattainable.

For now, the hubris of spurious preci sion has given way to humility. It turns out that in financial markets "black swans", or extreme events, occur much more often than the usual probability models suggest. Worse, finance is becoming more fragile: these days blow ups are twice as frequent as they were before the first world war, ac cording to Barry Eichengreen of the Uni versity of California at Berkeley and Mi chael Bordo of Rutgers University. Benoit Mandelbrot, the father of fractal theory and a pioneer in the study of market swings, argues that finance is prone to a "wild" randomness not usually seen in na ture. In markets, "rare big changes can be more significant than the sum of many small changes," he says. If financial mar kets followed the normal bell shaped dis tribution curve, in which meltdowns are very rare, the stockmarket crash of 1987, the interest rate turmoil of 1992 and the 2008 crash would each be expected only once in the lifetime of the universe.

This is changing the way many finan cial firms think about risk, says Greg Case, chief executive of Aon, an insurance bro ker. Before the crisis they were looking at things like pandemics, cyber security and terrorism as possible causes of black swans. Now they are turning to risks from within the system, and how they can be come amplified in combination.

Cheap as chips, and just as bad for you It would, though, be simplistic to blame the crisis solely, or even mainly, on sloppy risk managers or wild eyed quants. Cheap money led to the wholesale underpricing of risk; America ran negative real interest rates in 2002 05, even though consumer price inflation was quiescent. Plenty of economists disagree with the recent asser tion by Ben Bernanke, chairman of the Federal Reserve, that the crisis had more to do with lax regulation of mortgage pro ducts than loose monetary policy.

Equally damaging were policies to pro mote home ownership in America using Fannie Mae and Freddie Mac, the coun try’s two mortgage giants. They led the duo to binge on securities backed by shoddily underwritten loans.

In the absence of strict limits, higher le verage followed naturally from low inter est rates. The debt of America’s financial firms ballooned relative to the overall economy (see chart 1). At the peak of the madness, the median large bank had bor

rowings of 37 times its equity, meaning it could be wiped out by a loss of just 2 3% of its assets. Borrowed money allowed inves tors to fake "alpha", or above market re turns, says Benn Steil of the Council on Foreign Relations.

The agony was compounded by the proliferation of short term debt to support illiquid long term assets, much of it issued beneath the regulatory radar in highly le veraged "shadow" banks, such as struc tured investment vehicles. When markets froze, sponsoring entities, usually banks, felt morally obliged to absorb their losses. "Reputation risk was shown to have a very real financial price," says Doug Roeder of the Office of the Comptroller of the Cur rency, an American regulator.

Everywhere you looked, moreover, in centives were misaligned. Firms deemed "too big to fail" nestled under implicit guar antees. Sensitivity to risk was dulled by the "Greenspan put", a belief that America’s Federal Reserve would ride to the rescue with lower rates and liquidity support if needed. Scrutiny of borrowers was dele gated to rating agencies, who were paid by the debt issuers. Some products were so complex, and the chains from borrower to end investor so long, that thorough due di ligence was impossible. A proper under

standing of a typical collateralised debt ob ligation (CDO), a structured bundle of debt securities, would have required reading 30,000 pages of documentation.

Fees for securitisers were paid largely upfront, increasing the temptation to origi nate, flog and forget. The problems with bankers’ pay went much wider, meaning that it was much better to be an employee than a shareholder (or, eventually, a tax payer picking up the bail out tab). The role of top executives’ pay has been over

( Borrowed time US financial-industry debt as % of GDP 1 120 100 80 60 40 20 0 1978 1988 1998 2008 Source: Federal Reserve )blown. Top brass at Lehman Brothers and American International Group (AIG) suf

fered massive losses when share prices tumbled. A recent study found that banks where chief executives had more of their wealth tied up in the firm performed worse, not better, than those with appar ently less strong incentives. One explana tion is that they took risks they thought were in shareholders’ best interests, but were proved wrong. Motives lower down the chain were more suspect. It was too easy for traders to cash in on short term gains and skirt responsibility for any time bombs they had set ticking.

Asymmetries wreaked havoc in the vast over the counter derivatives market, too, where even large dealing firms lacked the information to determine the conse quences of others failing. Losses on con tracts linked to Lehman turned out to be modest, but nobody knew that when it collapsed in September 2008, causing pan ic. Likewise, it was hard to gauge the expo sures to "tail" risks built up by sellers of

swaps on CDOs such as AIG and bond in surers. These were essentially put options,

with limited upside and a low but real probability of catastrophic losses.

Another factor in the build up of exces sive risk was what Andy Haldane, head of financial stability at the Bank of England, has described as "disaster myopia". Like drivers who slow down after seeing a crash but soon speed up again, investors exercise greater caution after a disaster, but these days it takes less than a decade to make them reckless again. Not having seen a debt market crash since 1998, investors piled into ever riskier securities in 2003 07 to maintain yield at a time of low interest rates. Risk management models rein forced this myopia by relying too heavily on recent data samples with a narrow dis tribution of outcomes, especially in sub prime mortgages.

A further hazard was summed up by the assertion in 2007 by Chuck Prince, then Citigroup’s boss, that "as long as the music is playing, you’ve got to get up and dance." Performance is usually judged relative to rivals or to an industry benchmark, en couraging banks to mimic each other’s risk taking, even if in the long run it bene fits no one. In mortgages, bad lenders drove out good ones, keeping up with ag gressive competitors for fear of losing mar ket share. A few held back, but it was not easy: when JPMorgan sacrificed five per centage points of return on equity in the short run, it was lambasted by share holders who wanted it to "catch up" with zippier looking rivals.

An overarching worry is that the com 1

( The Economist February 13th 2010 A special report on financial risk )3

2 plexity of today’s global financial network makes occasional catastrophic failure in evitable. For example, the market for credit derivatives galloped far ahead of its sup porting infrastructure. Only now are seri ous moves being made to push these con tracts through central clearing houses which ensure that trades are properly col lateralised and guarantee their completion if one party defaults.

Network overload

The push to allocate capital ever more effi ciently over the past 20 years created what Till Guldimann, the father of VAR and vice chairman of SunGard, a technology firm, calls "capitalism on steroids". Banks got to depend on the modelling of prices in esoteric markets to gauge risks and became

adept at gaming the rules. As a result, capi tal was not being spread around as effi ciently as everyone believed.

Big banks had also grown increasingly interdependent through the boom in de rivatives, computer driven equities trad ing and so on. Another bond was cross ownership: at the start of the crisis, finan cial firms held big dollops of each other’s common and hybrid equity. Such tight coupling of components increases the danger of "non linear" outcomes, where a small change has a big impact. "Financial markets are not only vulnerable to black swans but have become the perfect breed ing ground for them," says Mr Guldimann. In such a network a firm’s troubles can have an exaggerated effect on the per ceived riskiness of its trading partners. When Lehman’s credit default spreads

( In lockstep Weighted average cumulative total returns, % 2 200 Large complex financial institutions 150 Banks Insurers 100 50 + 0 – 50 2000 01 02 03 04 05 06 07 08 09 Source: “Banking on the State” by Andrew Haldane and Piergiorgio Alessandri; Bank for International Settlements Hedge funds )

rose to distressed levels, AIG’s jumped by twice what would have been expected on its own, according to the International Monetary Fund.

Mr Haldane has suggested that these knife edge dynamics were caused not only by complexity but also-paradoxically-by homogeneity. Banks, insurers, hedge funds and others bought smorgasbords of debt securities to try to reduce risk through di versification, but the ingredients were sim ilar: leveraged loans, American mortgages and the like. From the individual firm’s perspective this looked sensible. But for the system as a whole it put everyone’s eggs in the same few baskets, as reflected in their returns (see chart 2).

Efforts are now under way to deal with these risks. The Financial Stability Board, an international group of regulators, is try ing to co ordinate global reforms in areas

such as capital, liquidity and mechanisms for rescuing or dismantling troubled banks. Its biggest challenge will be to make the system more resilient to the failure of giants. There are deep divisions over how to set about this, with some favouring tougher capital requirements, others break ups, still others-including Ameri ca-a combination of remedies.

In January President Barack Obama shocked big banks by proposing a tax on their liabilities and a plan to cap their size, ban "proprietary" trading and limit their involvement in hedge funds and private equity. The proposals still need congressio nal approval. They were seen as energising the debate about how to tackle dangerous ly large firms, though the reaction in Eu rope was mixed.

Regulators are also inching towards a more "systemic" approach to risk. The old supervisory framework assumed that if the 100 largest banks were individually safe, then the system was too. But the crisis showed that even well managed firms, acting prudently in a downturn, can un dermine the strength of all.

The banks themselves will have to find a middle ground in risk management, somewhere between gut feeling and num ber fetishism. Much of the progress made in quantitative finance was real enough, but a firm that does not understand the flaws in its models is destined for trouble. This special report will argue that rules will have to be both tightened and better enforced to avoid future crises-but that all the reforms in the world will never guaran tee total safety. 7

( Number crunchers crunched The uses and abuses of mathematical models )

( I )T PUT noses out of joint, but it changed markets for good. In the mid 1970s a few progressive occupants of Chicago’s op tions pits started trading with the aid of sheets of theoretical prices derived from a model and sold by an economist called Fisher Black. Rivals, used to relying on their wits, were unimpressed. One model

based trader complained of having his pa pers snatched away and being told to "trade like a man". But the strings of num bers caught on, and soon derivatives ex changes hailed the Black Scholes model, which used share and bond prices to calcu

late the value of derivatives, for helping to legitimise a market that had been derided as a gambling den.

Thanks to Black Scholes, options pric ing no longer had to rely on educated guesses. Derivatives trading got a huge boost and quants poured into the industry. By 2005 they accounted for 5% of all fi nance jobs, against 1.2% in 1980, says Thom as Philippon of New York University-and probably a much higher proportion of pay. By 2007 finance was attracting a quarter of all graduates from the California Institute of Technology.

These eggheads are now in the dock, along with their probabilistic models. In America a congressional panel is investi gating the models’ role in the crash. Wired, a publication that can hardly be accused of technophobia, has described default prob ability models as "the formula that killed Wall Street". Long standing critics of risk modelling, such as Nassim Nicholas Taleb, author of "The Black Swan", and Paul Wil mott, a mathematician turned financial educator, are now hailed as seers. Models "increased risk exposure instead of limit

ing it", says Mr Taleb. "They can be worse1

( A special report on financial risk The Economist February 13th 2010 )4

2 than nothing, the equivalent of a danger ous operation on a patient who would stand a better chance if left untreated."

Not all models were useless. Those for interest rates and foreign exchange per formed roughly as they were meant to. However, in debt markets they failed ab jectly to take account of low probability but high impact events such as the gut wrenching fall in house prices.

The models went particularly awry when clusters of mortgage backed securi ties were further packaged into collateral ised debt obligations (CDOs). In traditional products such as corporate debt, rating agencies employ basic credit analysis and judgment. CDOs were so complex that they had to be assessed using specially de signed models, which had various faults. Each CDO is a unique mix of assets, but the assumptions about future defaults and mortgage rates were not closely tailored to that mix, nor did they factor in the tenden cy of assets to move together in a crisis.

The problem was exacerbated by the credit raters’ incentive to accommodate the issuers who paid them. Most financial firms happily relied on the models, even though the expected return on AAA rated tranches was suspiciously high for such apparently safe securities. At some banks,

risk managers who questioned the rating agencies’ models were given short shrift. Moody’s and Standard & Poor’s were as sumed to know best. For people paid ac cording to that year’s revenue, this was un derstandable. "A lifetime of wealth was only one model away," sneers an Ameri can regulator.

Moreover, heavy use of models may have changed the markets they were sup posed to map, thus undermining the valid ity of their own predictions, says Donald MacKenzie, an economic sociologist at the University of Edinburgh. This feedback process is known as counter performativ

ity and had been noted before, for instance with Black Scholes. With CDOs the mod els’ popularity boosted demand, which lowered the quality of the asset backed se curities that formed the pools’ raw materi al and widened the gap between expected and actual defaults (see chart 3).

A related problem was the similarity of risk models. Banks thought they were div ersified, only to find that many others held comparable positions, based on similar models that had been built to comply with the Basel 2 standards, and everyone was trying to unwind the same positions at the same time. The breakdown of the models, which had been the only basis for pricing

the more exotic types of security, turned risk into full blown uncertainty (and thus extreme volatility).

For some, the crisis has shattered faith in the precision of models and their inputs. They failed Keynes’s test that it is better to be roughly right than exactly wrong. One number coming under renewed scrutiny is

"value at risk" (VAR), used by banks to measure the risk of loss in a portfolio of fi

nancial assets, and by regulators to calcu late banks’ capital buffers. Invented by egg heads at JPMorgan in the late 1980s, VAR has grown steadily in popularity. It is the subject of more than 200 books. What makes it so appealing is that its complex formulae distil the range of potential daily profits or losses into a single dollar figure.

Only so far with VAR

Frustratingly, banks introduce their own quirks into VAR calculations, making com parison difficult. For example, Morgan Stanley’s VAR for the first quarter of 2009 by its own reckoning was $115m, but using

Goldman Sachs’s method it would have been $158m. The bigger problem, though, is that VAR works only for liquid securities over short periods in "normal" markets, and it does not cover catastrophic out comes. If you have $30m of two week 1% VAR, for instance, that means there is a 99% chance that you will not lose more than that amount over the next fortnight. But there may be a huge and unacknowledged threat lurking in that 1% tail.

So chief executives would be foolish to rely solely, or even primarily, on VAR to manage risk. Yet many managers and boards continue to pay close attention to it without fully understanding the caveats-

the equivalent of someone who cannot swim feeling confident of crossing a river having been told that it is, on average, four feet deep, says Jaidev Iyer of the Global As sociation of Risk Professionals.

Regulators are encouraging banks to look beyond VAR. One way is to use Co VAR (Conditional VAR), a measure that aims to capture spillover effects in trou

bled markets, such as losses due to the dis tress of others. This greatly increases some banks’ value at risk. Banks are developing their own enhancements. Morgan Stanley, for instance, uses "stress" VAR, which fac tors in very tight liquidity constraints.

Like its peers, Morgan Stanley is also re viewing its stress testing, which is used to consider extreme situations. The worst sce nario envisaged by the firm turned out to be less than half as bad as what actually happened in the markets. JPMorgan Chase’s debt market stress tests foresaw a 40% increase in corporate spreads, but high yield spreads in 2007 09 increased many times over. Others fell similarly short. Most banks’ tests were based on his torical crises, but this assumes that the fu ture will be similar to the past. "A repeat of any specific market event, such as 1987 or 1998, is unlikely to be the way that a future crisis will unfold," says Ken deRegt, Mor gan Stanley’s chief risk officer.

( Never mind the quality CDOs of subprime-mortgage-backed securities Issued in 2005-07, % 3 AAA Estimated 3-year default rate 0.001 Actual default rate 0.10 A A- BBB+ BBB BBB- 0.09 0.12 0.34 0.49 0.88 29.21 36.65 48.73 56.10 66.67 Source: Donald MacKenzie, University of Edinburgh )Faced with either random (and there fore not very believable) scenarios or sim plistic models that neglect fat tail risks, many find themselves in a "no man’s land" between the two, says Andrew Free man of Deloitte (and formerly a journalist at The Economist). Nevertheless, he views scenario planning as a useful tool. A firm that had thought about, say, the mutation of default risk into liquidity risk would have had a head start over its competitors in 2008, even if it had not predicted pre cisely how this would happen.

( AA+ 0.01 1.68 AA 0.04 8.16 AA- 0.05 12.03 A+ 0.06 20.96 )To some, stress testing will always seem maddeningly fuzzy. "It has so far been seen as the acupuncture and herbal remedies corner of risk management, though per ceptions are changing," says Riccardo Reb onato of Royal Bank of Scotland, who is writing a book on the subject. It is not meant to be a predictive tool but a means of considering possible outcomes to allow firms to react more nimbly to unexpected developments, he argues. Hedge funds are better at this than banks. Some had thought about the possibility of a large broker dealer going bust. At least one,

AQR, had asked its lawyers to grill the fund’s prime brokers about the fate of its

assets in the event of their demise. 1