The gods strike back A special report on financial risk l February 13th 2010

RISk.indd 1 2/2/10 13:08:02 The Economist February 13th 2010 A special report on †nancial risk 1

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

HE revolutionary idea that de†nes ed by larger balance sheets and greater le• ŒTthe boundary between modern verage (borrowing), risk was being capped times and the past is the mastery of risk: by a technological shift. the notion that the future is more than a There was something self•serving whim of the gods and that men and wom• about this. The more that risk could be cali• en are not passive before nature. So wrote brated, the greater the opportunity to turn Peter Bernstein in his seminal history of debt into securities that could be sold or risk, ŒAgainst the Gods, published in 1996. held in trading books, with lower capital And so it seemed, to all but a few Cassan• charges than regular loans. Regulators ac• dras, for much of the decade that followed. cepted this, arguing that the Œgreat moder• Finance enjoyed a golden period, with low ation had subdued macroeconomic dan• interest rates, low volatility and high re• gers and that securitisation had chopped turns. Risk seemed to have been reduced up individual †rms’ risks into manageable to a permanently lower level. lumps. This faith in the new, technology• This purported new paradigm hinged, driven order was re‡ected in the Basel 2 in large part, on three closely linked devel• bank•capital rules, which relied heavily on opments: the huge growth of derivatives; the banks’ internal models. the decomposition and distribution of There were bumps along the way, such Acknowledgments credit risk through securitisation; and the as the near•collapse of Long•Term Capital In addition to those mentioned in the text, the author formidable combination of mathematics Management (LTCM), a hedge fund, and would like to thank the following for their help in and computing power in risk management the dotcom bust, but each time markets re• preparing this report: Madelyn Antoncic, Scott Baret, Richard Bookstaber, Kevin Buehler, Jan Brockmeijer, that had its roots in academic work of the covered relatively quickly. Banks grew Stephen Cecchetti, Mark Chauvin, John Cochrane, José mid•20th century. It blossomed in the cocky. But that sense of security was de• Corral, Wilson Ervin, Dan Fields, Chris Finger, Bennett 1990s at †rms such as Bankers Trust and stroyed by the meltdown of 2007•09, Golub, John Hogan, Henry Hu, Simon Johnson, Robert Kaplan, Steven Kaplan, Anil Kashyap, James Lam, Brian JPMorgan, which developed Œvalue•at• which as much as anything was a crisis of Leach, Robert Le Blanc, Mark Levonian, Tim Long, Blythe risk (VAR), a way for banks to calculate modern metrics•based risk management. Masters, Michael Mendelson, Robert Merton, Jorge Mina, how much they could expect to lose when The idea that markets can be left to police Mary Frances Monroe, Lubos Pastor, Henry Ristuccia, Brian Robertson, Daniel Sigrist, Pietro Veronesi, Jim things got really rough. themselves turned out to be the world’s Wiener, Paul Wright and Luigi Zingales. Suddenly it seemed possible for any †• most expensive mistake, requiring $15 tril• nancial risk to be measured to †ve decimal lion in capital injections and other forms A list of sources is at places, and for expected returns to be ad• of support. ŒIt has cost a lot to learn how lit• Economist.com/specialreports justed accordingly. Banks hired hordes of tle we really knew, says a senior central PhD•wielding Œquants to †ne•tune ever banker. Another lesson was that managing An audio interview with the author is at more complex risk models. The belief took risk is as much about judgment as about Economist.com/audiovideo hold that, even as pro†ts were being boost• numbers. Trying ever harder to capture 1 2 A special report on †nancial risk The Economist February 13th 2010

2 risk in mathematical formulae can be rowings of 37 times its equity, meaning it fered massive losses when share prices counterproductive if such a degree of ac• could be wiped out by a loss of just 2•3% of tumbled. A recent study found that banks curacy is intrinsically unattainable. its assets. Borrowed money allowed inves• where chief executives had more of their For now, the hubris of spurious preci• tors to fake Œalpha, or above•market re• wealth tied up in the †rm performed sion has given way to humility. It turns out turns, says Benn Steil of the Council on worse, not better, than those with appar• that in †nancial markets Œblack swans, or Foreign Relations. ently less strong incentives. One explana• extreme events, occur much more often The agony was compounded by the tion is that they took risks they thought than the usual probability models suggest. proliferation of short•term debt to support were in shareholders’ best interests, but Worse, †nance is becoming more fragile: illiquid long•term assets, much of it issued were proved wrong. Motives lower down these days blow•ups are twice as frequent beneath the regulatory radar in highly le• the chain were more suspect. It was too as they were before the †rst world war, ac• veraged Œshadow banks, such as struc• easy for traders to cash in on short•term cording to Barry Eichengreen of the Uni• tured investment vehicles. When markets gains and skirt responsibility for any time• versity of California at Berkeley and Mi• froze, sponsoring entities, usually banks, bombs they had set ticking. chael Bordo of Rutgers University. Benoit felt morally obliged to absorb their losses. Asymmetries wreaked havoc in the Mandelbrot, the father of fractal theory ŒReputation risk was shown to have a very vast over•the•counter derivatives market, and a pioneer in the study of market real †nancial price, says Doug Roeder of too, where even large dealing †rms lacked swings, argues that †nance is prone to a the Oˆce of the Comptroller of the Cur• the information to determine the conse• Œwild randomness not usually seen in na• rency, an American regulator. quences of others failing. Losses on con• ture. In markets, Œrare big changes can be Everywhere you looked, moreover, in• tracts linked to Lehman turned out to be more signi†cant than the sum of many centives were misaligned. Firms deemed modest, but nobody knew that when it small changes, he says. If †nancial mar• Œtoo big to fail nestled under implicit guar• collapsed in September 2008, causing pan• kets followed the normal bell•shaped dis• antees. Sensitivity to risk was dulled by the ic. Likewise, it was hard to gauge the expo• tribution curve, in which meltdowns are ŒGreenspan put, a belief that America’s sures to Œtail risks built up by sellers of very rare, the stockmarket crash of 1987, the Federal Reserve would ride to the rescue swaps on CDOs such as AIG and bond in• interest•rate turmoil of 1992 and the 2008 with lower rates and liquidity support if surers. These were essentially put options, crash would each be expected only once in needed. Scrutiny of borrowers was dele• with limited upside and a low but real the lifetime of the universe. gated to rating agencies, who were paid by probability of catastrophic losses. This is changing the way many †nan• the debt•issuers. Some products were so Another factor in the build•up of exces• cial †rms think about risk, says Greg Case, complex, and the chains from borrower to sive risk was what Andy Haldane, head of chief executive of Aon, an insurance bro• end•investor so long, that thorough due di• †nancial stability at the Bank of England, ker. Before the crisis they were looking at ligence was impossible. A proper under• has described as Œdisaster myopia. Like things like pandemics, cyber•security and standing of a typical collateralised debt ob• drivers who slow down after seeing a terrorism as possible causes of black ligation (CDO), a structured bundle of debt crash but soon speed up again, investors swans. Now they are turning to risks from securities, would have required reading exercise greater caution after a disaster, but within the system, and how they can be• 30,000 pages of documentation. these days it takes less than a decade to come ampli†ed in combination. Fees for securitisers were paid largely make them reckless again. Not having seen upfront, increasing the temptation to origi• a debt•market crash since 1998, investors Cheap as chips, and just as bad for you nate, ‡og and forget. The problems with piled into ever riskier securities in 2003•07 It would, though, be simplistic to blame bankers’ pay went much wider, meaning to maintain yield at a time of low interest the crisis solely, or even mainly, on sloppy that it was much better to be an employee rates. Risk•management models rein• risk managers or wild•eyed quants. Cheap than a shareholder (or, eventually, a tax• forced this myopia by relying too heavily money led to the wholesale underpricing payer picking up the bail•out tab). The role on recent data samples with a narrow dis• of risk; America ran negative real interest of top executives’ pay has been over• tribution of outcomes, especially in sub• rates in 2002•05, even though consumer• blown. Top brass at Lehman Brothers and prime mortgages. price in‡ation was quiescent. Plenty of American International Group (AIG) suf• A further hazard was summed up by economists disagree with the recent asser• the assertion in 2007 by Chuck Prince, then tion by Ben Bernanke, chairman of the Citigroup’s boss, that Œas long as the music Federal Reserve, that the crisis had more to Borrowed time 1 is playing, you’ve got to get up and dance. do with lax regulation of mortgage pro• US financial-industry debt as % of GDP Performance is usually judged relative to ducts than loose monetary policy. rivals or to an industry benchmark, en• Equally damaging were policies to pro• 120 couraging banks to mimic each other’s mote home ownership in America using 100 risk•taking, even if in the long run it bene• Fannie Mae and Freddie Mac, the coun• †ts no one. In mortgages, bad lenders try’s two mortgage giants. They led the duo 80 drove out good ones, keeping up with ag• to binge on securities backed by shoddily 60 gressive competitors for fear of losing mar• underwritten loans. ket share. A few held back, but it was not In the absence of strict limits, higher le• 40 easy: when JPMorgan sacri†ced †ve per• verage followed naturally from low inter• 20 centage points of return on equity in the est rates. The debt of America’s †nancial short run, it was lambasted by share• †rms ballooned relative to the overall 0 holders who wanted it to Œcatch up with 1978 1988 1998 2008 economy (see chart 1). At the peak of the zippier•looking rivals. Source: Federal Reserve madness, the median large bank had bor• An overarching worry is that the com•1 The Economist February 13th 2010 A special report on †nancial risk 3

2 plexity of today’s global †nancial network such as capital, liquidity and mechanisms makes occasional catastrophic failure in• In lockstep 2 for rescuing or dismantling troubled evitable. For example, the market for credit Weighted average cumulative total returns, % banks. Its biggest challenge will be to make derivatives galloped far ahead of its sup• the system more resilient to the failure of 200 porting infrastructure. Only now are seri• Large complex giants. There are deep divisions over how ous moves being made to push these con• financial to set about this, with some favouring 150 tracts through central clearing•houses institutions tougher capital requirements, others Banks which ensure that trades are properly col• 100 break•ups, still others‹including Ameri• lateralised and guarantee their completion Insurers ca‹a combination of remedies. Hedge funds if one party defaults. 50 In January President Barack Obama + shocked big banks by proposing a tax on Network overload 0 their liabilities and a plan to cap their size, The push to allocate capital ever more eˆ• – ban Œproprietary trading and limit their ciently over the past 20 years created what 50 involvement in hedge funds and private Till Guldimann, the father of VAR and 2000 01 02 03 04 05 06 07 08 09 equity. The proposals still need congressio• vice•chairman of SunGard, a technology Source: “Banking on the State” by Andrew Haldane and nal approval. They were seen as energising Piergiorgio Alessandri; Bank for International Settlements †rm, calls Œcapitalism on steroids. Banks the debate about how to tackle dangerous• got to depend on the modelling of prices in ly large †rms, though the reaction in Eu• esoteric markets to gauge risks and became rose to distressed levels, AIG’s jumped by rope was mixed. adept at gaming the rules. As a result, capi• twice what would have been expected on Regulators are also inching towards a tal was not being spread around as eˆ• its own, according to the International more Œsystemic approach to risk. The old ciently as everyone believed. Monetary Fund. supervisory framework assumed that if Big banks had also grown increasingly Mr Haldane has suggested that these the 100 largest banks were individually interdependent through the boom in de• knife•edge dynamics were caused not only safe, then the system was too. But the crisis rivatives, computer•driven equities trad• by complexity but also‹paradoxically‹by showed that even well•managed †rms, ing and so on. Another bond was cross• homogeneity. Banks, insurers, hedge funds acting prudently in a downturn, can un• ownership: at the start of the crisis, †nan• and others bought smorgasbords of debt dermine the strength of all. cial †rms held big dollops of each other’s securities to try to reduce risk through di• The banks themselves will have to †nd common and hybrid equity. Such tight versi†cation, but the ingredients were sim• a middle ground in risk management, coupling of components increases the ilar: leveraged loans, American mortgages somewhere between gut feeling and num• danger of Œnon•linear outcomes, where a and the like. From the individual †rm’s ber fetishism. Much of the progress made small change has a big impact. ŒFinancial perspective this looked sensible. But for in quantitative †nance was real enough, markets are not only vulnerable to black the system as a whole it put everyone’s but a †rm that does not understand the swans but have become the perfect breed• eggs in the same few baskets, as re‡ected in ‡aws in its models is destined for trouble. ing ground for them, says Mr Guldimann. their returns (see chart 2). This special report will argue that rules In such a network a †rm’s troubles can E orts are now under way to deal with will have to be both tightened and better have an exaggerated e ect on the per• these risks. The Financial Stability Board, enforced to avoid future crises‹but that all ceived riskiness of its trading partners. an international group of regulators, is try• the reforms in the world will never guaran• When Lehman’s credit•default spreads ing to co•ordinate global reforms in areas tee total safety. 7 Number•crunchers crunched

The uses and abuses of mathematical models

T PUT noses out of joint, but it changed late the value of derivatives, for helping to These eggheads are now in the dock, Imarkets for good. In the mid•1970s a few legitimise a market that had been derided along with their probabilistic models. In progressive occupants of Chicago’s op• as a gambling den. America a congressional panel is investi• tions pits started trading with the aid of Thanks to Black•Scholes, options pric• gating the models’ role in the crash. Wired, sheets of theoretical prices derived from a ing no longer had to rely on educated a publication that can hardly be accused of model and sold by an economist called guesses. Derivatives trading got a huge technophobia, has described default•prob• Fisher Black. Rivals, used to relying on their boost and quants poured into the industry. ability models as Œthe formula that killed wits, were unimpressed. One model• By 2005 they accounted for 5% of all †• Wall Street. Long•standing critics of risk• based trader complained of having his pa• nance jobs, against 1.2% in 1980, says Thom• modelling, such as Nassim Nicholas Taleb, pers snatched away and being told to as Philippon of New York University‹and author of ŒThe Black Swan, and Paul Wil• Œtrade like a man. But the strings of num• probably a much higher proportion of pay. mott, a mathematician turned †nancial bers caught on, and soon derivatives ex• By 2007 †nance was attracting a quarter of educator, are now hailed as seers. Models changes hailed the Black•Scholes model, all graduates from the California Institute Œincreased risk exposure instead of limit• which used share and bond prices to calcu• of Technology. ing it, says Mr Taleb. ŒThey can be worse 1 4 A special report on †nancial risk The Economist February 13th 2010

2 than nothing, the equivalent of a danger• the more exotic types of security, turned the equivalent of someone who cannot ous operation on a patient who would risk into full•blown uncertainty (and thus swim feeling con†dent of crossing a river stand a better chance if left untreated. extreme volatility). having been told that it is, on average, four Not all models were useless. Those for For some, the crisis has shattered faith feet deep, says Jaidev Iyer of the Global As• interest rates and foreign exchange per• in the precision of models and their inputs. sociation of Risk Professionals. formed roughly as they were meant to. They failed Keynes’s test that it is better to Regulators are encouraging banks to However, in debt markets they failed ab• be roughly right than exactly wrong. One look beyond VAR. One way is to use Co• jectly to take account of low•probability number coming under renewed scrutiny is VAR (Conditional VAR), a measure that but high•impact events such as the gut• Œvalue•at•risk (VAR), used by banks to aims to capture spillover e ects in trou• wrenching fall in house prices. measure the risk of loss in a portfolio of †• bled markets, such as losses due to the dis• The models went particularly awry nancial assets, and by regulators to calcu• tress of others. This greatly increases some when clusters of mortgage•backed securi• late banks’ capital bu ers. Invented by egg• banks’ value at risk. Banks are developing ties were further packaged into collateral• heads at JPMorgan in the late 1980s, VAR their own enhancements. Morgan Stanley, ised debt obligations (CDOs). In traditional has grown steadily in popularity. It is the for instance, uses Œstress VAR, which fac• products such as corporate debt, rating subject of more than 200 books. What tors in very tight liquidity constraints. agencies employ basic credit analysis and makes it so appealing is that its complex Like its peers, Morgan Stanley is also re• judgment. CDOs were so complex that formulae distil the range of potential daily viewing its stress testing, which is used to they had to be assessed using specially de• pro†ts or losses into a single dollar †gure. consider extreme situations. The worst sce• signed models, which had various faults. nario envisaged by the †rm turned out to Each CDO is a unique mix of assets, but the Only so far with VAR be less than half as bad as what actually assumptions about future defaults and Frustratingly, banks introduce their own happened in the markets. JPMorgan mortgage rates were not closely tailored to quirks into VAR calculations, making com• Chase’s debt•market stress tests foresaw a that mix, nor did they factor in the tenden• parison diˆcult. For example, Morgan 40% increase in corporate spreads, but cy of assets to move together in a crisis. Stanley’s VAR for the †rst quarter of 2009 high•yield spreads in 2007•09 increased The problem was exacerbated by the by its own reckoning was $115m, but using many times over. Others fell similarly credit raters’ incentive to accommodate Goldman Sachs’s method it would have short. Most banks’ tests were based on his• the issuers who paid them. Most †nancial been $158m. The bigger problem, though, is torical crises, but this assumes that the fu• †rms happily relied on the models, even that VAR works only for liquid securities ture will be similar to the past. ŒA repeat of though the expected return on AAA•rated over short periods in Œnormal markets, any speci†c market event, such as 1987 or tranches was suspiciously high for such and it does not cover catastrophic out• 1998, is unlikely to be the way that a future apparently safe securities. At some banks, comes. If you have $30m of two•week 1% crisis will unfold, says Ken deRegt, Mor• risk managers who questioned the rating VAR, for instance, that means there is a 99% gan Stanley’s chief risk oˆcer. agencies’ models were given short shrift. chance that you will not lose more than Faced with either random (and there• Moody’s and Standard & Poor’s were as• that amount over the next fortnight. But fore not very believable) scenarios or sim• sumed to know best. For people paid ac• there may be a huge and unacknowledged plistic models that neglect fat•tail risks, cording to that year’s revenue, this was un• threat lurking in that 1% tail. many †nd themselves in a Œno•man’s• derstandable. ŒA lifetime of wealth was So chief executives would be foolish to land between the two, says Andrew Free• only one model away, sneers an Ameri• rely solely, or even primarily, on VAR to man of Deloitte (and formerly a journalist can regulator. manage risk. Yet many managers and at The Economist). Nevertheless, he views Moreover, heavy use of models may boards continue to pay close attention to it scenario planning as a useful tool. A †rm have changed the markets they were sup• without fully understanding the caveats‹ that had thought about, say, the mutation posed to map, thus undermining the valid• of default risk into liquidity risk would ity of their own predictions, says Donald have had a head start over its competitors MacKenzie, an economic sociologist at the Never mind the quality 3 in 2008, even if it had not predicted pre• University of Edinburgh. This feedback CDOs of subprime-mortgage-backed securities cisely how this would happen. process is known as counter•performativ• Issued in 2005-07, % To some, stress testing will always seem ity and had been noted before, for instance maddeningly fuzzy. ŒIt has so far been seen Estimated Actual with Black•Scholes. With CDOs the mod• 3-year default default as the acupuncture•and•herbal•remedies els’ popularity boosted demand, which rate rate corner of risk management, though per• lowered the quality of the asset•backed se• AAA 0.001 0.10 ceptions are changing, says Riccardo Reb• curities that formed the pools’ raw materi• AA+ 0.01 1.68 onato of Royal Bank of Scotland, who is al and widened the gap between expected writing a book on the subject. It is not AA 0.04 8.16 and actual defaults (see chart 3). meant to be a predictive tool but a means A related problem was the similarity of AA- 0.05 12.03 of considering possible outcomes to allow risk models. Banks thought they were div• A+ 0.06 20.96 †rms to react more nimbly to unexpected ersi†ed, only to †nd that many others held A 0.09 29.21 developments, he argues. Hedge funds are comparable positions, based on similar A- 0.12 36.65 better at this than banks. Some had models that had been built to comply with BBB+ 0.34 48.73 thought about the possibility of a large the Basel 2 standards, and everyone was BBB 0.49 56.10 broker•dealer going bust. At least one, trying to unwind the same positions at the BBB- 0.88 66.67 AQR, had asked its lawyers to grill the same time. The breakdown of the models, fund’s prime brokers about the fate of its Source: Donald MacKenzie, University of Edinburgh which had been the only basis for pricing assets in the event of their demise. 1 The Economist February 13th 2010 A special report on †nancial risk 5

2 Some of the blame lies with bank regu• sure to derivatives counterparties. lators, who were just as blind to the dan• To be fair, totting up counterparty risk is gers ahead as the †rms they oversaw. not easy. For each trading partner the cal• Sometimes even more so: after the rescue culations can involve many di erent types of Bear Stearns in March 2008 but before of contract and hundreds of legal entities. Lehman’s collapse, Morgan Stanley was re• But banks will have to learn fast: under portedly told by supervisors at the Federal new international proposals, they will for Reserve that its doomsday scenario was the †rst time face capital charges on the too bearish. creditworthiness of swap counterparties. The regulators have since become The banks with the most dysfunctional tougher. In America, for instance, banks systems are generally those, such as Citi• have been told to run stress tests with sce• group, that have been through multiple narios that include a huge leap in interest marriages and ended up with dozens of rates. A supervisors’ report last October Œlegacy systems that cannot easily com• †ngered some banks for Œwindow•dress• municate with each other. That may ex• ing their tests. Oˆcials are now asking for plain why some Citi units continued to Œreverse stress testing, in which a †rm pile into subprime mortgages even as oth• imagines it has failed and works back• ers pulled back. wards to determine which vulnerabilities In the depths of the crisis some banks caused the hypothetical collapse. Britain were unaware that di erent business units has made this mandatory. Bankers are di• were marking the same assets at di erent vided over its usefulness. prices. The industry is working to sort this out. Banks are coming under pressure to Slicing the Emmental appoint chief data oˆcers who can police These changes point towards greater use of the integrity of the numbers, separate from judgment and less reliance on numbers in chief information oˆcers who concen• future. But it would be unfair to tar all mod• trate on system design and output. els with the same brush. The CDO †asco Some worry that the good work will be was an egregious and relatively rare case cast aside. As markets recover, the biggest of an instrument getting way ahead of the able but is not suˆcient. temptation will be to abandon or scale ability to map it mathematically. Models Not surprisingly, more investors are back IT projects, allowing product devel• were Œan accessory to the crime, not the now willing to give up some upside for the opment to get ahead of the supporting perpetrator, says Michael Mauboussin of promise of protection against catastrophic technology infrastructure, just as it did in Legg Mason, a money manager. losses. Pimco’s clients are paying up to 1% the last boom. As for VAR, it may be hopeless at signal• of the value of managed assets for the The way forward is not to reject high• ling rare severe losses, but the process by hedging‹even though, as the recent crisis tech †nance but to be honest about its limi• which it is produced adds enormously to showed, there is a risk that insurers will tations, says Emanuel Derman, a professor the understanding of everyday risk, which not be able to pay out. Lisa Goldberg of at New York’s Columbia University and a can be just as deadly as tail risk, says Aaron MSCI Barra reports keen interest in the an• former quant at Goldman Sachs. Models Brown, a risk manager at AQR. Craig Bro• alytics †rm’s extreme•risk model from should be seen as metaphors that can en• derick, chief risk oˆcer at Goldman Sachs, hedge funds, investment banks and pen• lighten but do not describe the world per• sees it as one of several measures which, sion plans. fectly. Messrs Derman and Wilmott have although of limited use individually, to• In some areas the need may be for more drawn up a modeller’s Hippocratic oath gether can provide a helpful picture. Like a computing power, not less. Financial †rms which pledges, among other things: ŒI will slice of Swiss cheese, each number has already spend more than any other indus• remember that I didn’t make the world, holes, but put several of them together and try on information technology (IT): some and it doesn’t satisfy my equations, and ŒI you get something solid. $500 billion in 2009, according to Gartner, will never sacri†ce reality for elegance Modelling is not going away; indeed, a consultancy. Yet the quality of informa• without explaining why I have done so. number•crunchers who are devising new tion †ltering through to senior managers is Often the problem is not complex †nance ways to protect investors from outlying fat• often inadequate. but the people who practise it, says Mr Wil• tail risks are gaining in‡uence. Pimco, for A report by bank supervisors last Octo• mott. Because of their love of puzzles, instance, o ers fat•tail hedging pro• ber pointed to poor risk Œaggregation: quants lean towards technically brilliant grammes for mutual•fund clients, using many large banks simply do not have the rather than sensible solutions and tend to cocktails of options and other instru• systems to present an up•to•date picture of over•engineer: ŒYou may need a plumber ments. These are built on speci†c risk fac• their †rm•wide links to borrowers and but you get a professor of ‡uid dynamics. tors rather than on the broader and in• trading partners. Two•thirds of the banks One way to deal with that problem is to creasingly ‡uid division of assets between surveyed said they were only Œpartially self•insure. JPMorgan Chase holds $3 bil• equities, currencies, commodities and so able (in other words, unable) to aggregate lion of Œmodel•uncertainty reserves to on. The relationships between asset class• their credit risks. The Federal Reserve, lead• cover mishaps caused by quants who have es Œhave become less stable, says Mo• ing stress tests on American banks last been too clever by half. If you can make hamed El•Erian, Pimco’s chief executive. spring, was shocked to †nd that some of provisions for bad loans, why not bad ŒAsset•class diversi†cation remains desir• them needed days to calculate their expo• maths too? 7 6 A special report on †nancial risk The Economist February 13th 2010

Cinderella’s moment

Risk managers to the fore

N A speech delivered to a banking•indus• Itry conference in Geneva in December 2006, Madelyn Antoncic issued a warning and then o ered some reassurance. With volatility low, corporate credit spreads growing ever tighter and markets all but ig• noring bad news, there was, she said, Œa seemingly overwhelming sense of com• placency. Nevertheless, she insisted that the †rm she served as chief risk oˆcer, Lehman Brothers, was well placed to ride out any turbulence, thanks to a keen awareness of emerging threats and a rock• solid analytical framework. Behind the scenes, all was not well. Ms Antoncic, a respected risk manager with an economics PhD, had expressed unease at the †rm’s heavy exposure to commer• cial property and was being sidelined, bit by bit, by the †rm’s autocratic boss, Dick Fuld. Less than two months after her speech she was pushed aside. Lehman’s story ended particularly bad• ly, but this sort of lapse in risk governance was alarmingly common during the boom. So much for the notion, generally accepted back then, that the quality of banks’ risk regimes had, like car compo• manager or other sceptic, he could adjourn bling to convince markets and regulators nents, converged around a high standard. the meeting, then reconstitute the commit• that they will continue to take risk serious• ŒThe variance turned out to be shocking, tee a week or two later with a more pliable ly once memories of the crisis fade. Some says Jamie Dimon, chief executive of membership that would approve the loan. are involving risk oˆcers in talks about JPMorgan Chase. Another common trick was for a busi• new products and strategic moves. At The banks that fared better, including ness line to keep quiet about a proposal on HSBC, for instance, they have had a bigger his own, relied largely on giving their risk• which it had been working for weeks until role in vetting acquisitions since the bank’s managing roundheads equal status with a couple of hours before the meeting to ap• American retail•banking subsidiary, the risk•taking cavaliers. That was not easy. prove it, so the risk team had no time to bought in 2003, su ered heavy subprime• In happy times, when risk seems low, pow• lodge convincing objections. Exasperated mortgage losses. ŒEveryone should now er shifts from risk managers to traders. roundheads would occasionally resort to see that the risk team needs to be just as in• Sales•driven cultures are the natural order pleading with regulators for help. In the volved on the returns side as on the risk of things on Wall Street and in the City. Dis• years before the crash the Basel Commit• side, says Maureen Miskovic, chief risk of• couraging transactions was frowned upon, tee of bank supervisors reportedly re• †cer at State Street, an American bank. especially at †rms trying to push their way ceived several requests from risk managers up capital•markets league tables. Risk to scrutinise excessive risk•taking at their Glamming up managers who said no put themselves on institutions that they felt powerless to stop. Ms Miskovic is one of an emerging breed a collision course with the business head Many banks’ failings exposed the tri• of more powerful risk oˆcers. They are and often the chief executive too. umph of form over substance. In recent seen as being on a par with the chief †nan• At some large banks that subsequently years it had become popular to appoint a cial oˆcer, get a say in decisions on pay su ered big losses, such as HBOS and Roy• chief risk oˆcer to signal that the issue was and have the ear of the board, whose al Bank of Scotland (RBS), credit commit• receiving attention. But according to Leo agreement is increasingly needed to re• tees, which vetted requests for big loans, Grepin of McKinsey, Œit was sometimes a move them. Some report directly to a could be formed on an ad hoc basis from a case of management telling him, ‘you tick board committee as well as‹or occasional• pool of eligible members. If the commit• the boxes on risk, and we’ll worry about ly instead of‹to the chief executive. tee’s chairman, typically a business•line generating revenue’. For many, the biggest task is to disman• head, encountered resistance from a risk Since 2007 banks have been scram• tle cumbersome Œsilos, says Ken Chalk of 1 The Economist February 13th 2010 A special report on †nancial risk 7 A matter of principle

Why some banks did much better than others 2 America’s Risk Management Association. Risks were often stu ed into convenient PMORGAN CHASE managed to avoid fore subprime defaults began to explode. but misleading pigeonholes. Banks were Jbig losses largely thanks to the tone set More willing than rivals to take risks, slow to re†ne their approach, even as by its boss, Jamie Dimon. A voracious Goldman is also quicker to hedge them. growing market complexity led some of reader of internal reports, he understands In late 2006 it spent up to $150m‹one• the risks to become interchangeable. †nancial arcana and subjects sta to de• eighth of that quarter’s operating pro†t‹ Take the growth of traded credit pro• tailed questioning. PowerPoint presenta• hedging exposure to AIG. ducts, such as asset•backed securities and tions are discouraged, informal discus• The †rm promotes senior traders to CDOs made up of them. Credit•risk de• sions of what is wrong, or could go risk positions, making clear that such partments thought of them as market risk, wrong, encouraged. These Œsoft princi• moves are a potential stepping stone to because they sat in the trading book. Mar• ples are supplemented by a hard•headed the top. Traders are encouraged to nurture ket•risk teams saw them as credit instru• approach to the allocation of capital. the risk manager in them: Gary Cohn, the ments, since the underlying assets were Though the bank su ered painful losses †rm’s president, rose to the top largely be• loans. This buck•passing proved particu• in leveraged loans, it was not tripped up cause of his skill at hedging Œtail risks. larly costly at UBS, which lost SFr36 billion by CDOs or structured investment vehi• Crucially, Goldman generally does not ($34 billion) on CDOs. Many banks are cles (SIVs), even though it had been in• †re its risk managers after a crisis, allow• now combining their market• and credit• strumental in developing both products. ing them to learn from the experience. Yet risk groups, as HSBC did last year. Nor was it heavily exposed to AIG, an in• despite everything, it still needed govern• For all the new•found authority of risk surance giant that got into trouble. ment help to survive. managers, it can still be hard to attract tal• This was not because it saw disaster By contrast, UBS’s risk culture was aw• ent to their ranks. The job is said to have coming, says Bill Winters, former co•head ful. Its investment bank was free to bet the risk pro†le of a short option position of the †rm’s investment bank, but be• with subsidised funds, since transfers with unlimited downside and limited up• cause it stuck by two basic principles: from the private bank were deeply under• side‹something every good risk manager don’t hold too much of anything, and priced. It confused itself by presenting should avoid. Moreover, it lacks glamour. only keep what you are sure will generate risk in a Œnet and forget format. Trading Persuading a trader to move to risk can be a decent risk•adjusted return. The bank desks would estimate the maximum pos• Œlike asking a trapeze artist to retrain as an jettisoned an SIV and $60 billion of CDO• sible loss on risky assets, hedge it and accountant, says Barrie Wilkinson of Oli• related risks because it saw them as too then record the net risk as minimal, inad• ver Wyman, a consultancy. dicey, at a time when others were still vertently concealing huge tail risks in the keen to snap them up. It also closed 60 gross exposure. And it moved its best trad• A question of culture credit lines for other SIVs and corporate ers to a hedge fund, leaving the B•team to Besides, there is more to establishing a sol• clients when it realised that these could manage the bank’s positions. id risk culture than empowering risk oˆ• be simultaneously drawn down if the Publicly humbled by a frank report on cers. Culture is a slippery concept, but it bank’s credit rating were cut. And it took a its failings, the bank has made a raft of matters. ŒWhatever causes the next crisis, conservative view of risk•mitigation. changes. Risk controllers have been hand• it will be di erent, so you need something Hedging through bond insurers, whose †• ed more power. Oswald Grübel, the chief that can deal with the unexpected. That’s nances grew shaky as the crisis spread, executive, has said that if his newish risk culture, says Colm Kelleher of Morgan was calculated twice: once assuming the chief, Philip Lofts, rejects a transaction he Stanley. One necessary ingredient is a tra• hedge would hold, and again assuming it will never overrule him. If the two dis• dition of asking and repeating questions was worthless. agree, Mr Lofts must inform the board, until a clear answer emerges, suggests Goldman Sachs’s risk management which no longer delegates risk issues to a Clayton Rose, a banker who now teaches stood out too‹unlike the public•relations trio of long•time UBS employees. A new, at Harvard Business School. skills it subsequently displayed. Steered independent risk committee is bristling The tone is set at the top, for better or by its chief †nancial oˆcer, David Viniar, with risk experts. Whether all this worse. At the best•run banks senior †gures the †rm’s traders began reducing their ex• amounts to a Œnew paradigm, as Mr spend as much time fretting over risks as posure to mortgage securities months be• Lofts claims, remains to be seen. they do salivating at opportunities (see box). By contrast, Lehman’s Mr Fuld talked of Œprotecting mother but was drawn to Prince or the board, but to a newly hired ex• loitte found that only seven of 30 large the glister of leveraged deals. Stan O’Neal, ecutive with a background in corporate•go• banks had done so in any detail. Everyone who presided over giant losses at Merrill vernance law, not cutting•edge †nance. agrees that boards have a critical role to Lynch, was more empire•builder than risk Another lesson is that boards matter play in determining risk appetite, but a re• manager. But imperial bosses and sound too. Directors’ lack of engagement or ex• cent report by a group of global regulators risk cultures sometimes go together, as at pertise played a big part in some of the found that many were reluctant to do this. JPMorgan and Banco Santander. worst slip•ups, including Citi’s. The Œso• Boards could also make a better job of A soft•touch boss can be more danger• ciology of big banks’ boards also had policing how (or even whether) banks ad• ous than a domineering one. Under Chuck something to do with it, says Ingo Walter just for risk in allocating capital internally. Prince, who famously learned only in Sep• of New York’s Stern School of Business: as Before the crisis some boards barely tember 2007 that Citigroup was sitting on the members bonded, dissidents felt pres• thought about this, naively assuming that $43 billion of toxic assets, the lunatics were sure to toe the line. procedures for it were well honed. A for• able to take over the asylum. Astonishing• Too few boards de†ned the parameters mer Lehman board member professes ly, the head of risk reported not to Mr of risk oversight. In a survey last year De• himself Œastonished, in retrospect, at how1 8 A special report on †nancial risk The Economist February 13th 2010

2 some of the risks in the company’s proper• vices backgrounds. Mr Pozen suggests as• determined or heavily in‡uenced by the ty investments were brushed aside when sembling a small cadre of †nancially ‡u• managers of the trading desks they over• assessing expected returns. The survivors ent Œsuper•directors who would meet saw, or their bonus linked to the desks’ per• are still struggling to create the sort of more often‹say, two or three days a month formance, says Richard Apostolik, who joined•up approach to risk adjustment rather than an average of six days a year, as heads the Global Association of Risk Pro• that is common at large hedge funds, ad• now‹and may serve on only one other fessionals (GARP). Boards need to elimi• mits one Wall Street executive. board to ensure they take the job seriously. nate such con‡icts of interest. That sounds sensible, but the case for Meanwhile risk teams are being beefed Board games another suggested reform‹creating inde• up. Morgan Stanley, for instance, is increas• Robert Pozen, head of MFS Investment pendent risk committees at board level‹is ing its complement to 450, nearly double Management, an American asset manager, less clear. At some banks risk issues are the number it had in 2008. The GARP saw thinks bank boards would be more e ec• handled perfectly well by the audit com• a 70% increase in risk•manager certi†ca• tive with fewer but more committed mem• mittee or the full board. Nor is there a clear tions last year. Risk is the busiest area for †• bers. Cutting their size to 4•8, rather than link between the frequency of risk•related nancial recruiters, says Tim Holt of Hei• the 10•18 typical now, would foster more meetings and a bank’s performance. At drick & Struggles, a †rm of headhunters. personal responsibility. More †nancial• Spain’s Santander the relevant committee When boards are looking for a new chief services expertise would help too. After met 102 times in 2008. Those of other executive, they increasingly want some• the passage of the Sarbanes•Oxley act in banks that emerged relatively unscathed, one who has been head of risk as well as 2002 banks hired more independent direc• such as JPMorgan and Credit Suisse, con• chief †nancial oˆcer, which used to be the tors, many of whom lacked relevant expe• vened much less often. standard requirement, reckons Mike rience. The former spymaster on Citi’s Moreover, some of the most important Woodrow of Risk Talent Associates, an• board and the theatrical impresario on risk•related decisions of the next few years other headhunting †rm. Lehman’s may have been happy to ask will come from another corner: the com• The big question is whether this inter• questions, but were they the right ones? pensation committee. It is not just invest• est in controlling risk will †zzle out as econ• Under regulatory pressure, banks such ment bankers and top executives whose omies recover. Experience suggests that it as Citi and Bank of America have hired pay structures need to be rethought. In the will. Bankers say this time is di erent‹but more directors with strong †nancial•ser• past, risk managers’ pay was commonly they always do. 7 When the river runs dry

The perils of a sudden evaporation of liquidity

TAMPEDING crowds can generate pres• just as fatal as insolvency. Many of those vestments spilled quickly into interbank Ssures of up to 4,500 Newtons per square clobbered in the crisis‹including Bear loan markets, commercial paper, prime metre, enough to bend steel barriers. Rush• Stearns, Northern Rock and AIG‹were brokerage, securities lending (lending es for the exit in †nancial markets can be struck down by a sudden lack of cash or shares to short•sellers) and so on. just as damaging. Investors crowd into funding sources, not because they ran out As con†dence ebbed, mortgage•backed trades to get the highest risk•adjusted re• of capital. securities could no longer be used so easily turn in the same way that everyone wants Yet liquidity risk has been neglected. as collateral in repurchase or Œrepo agree• tickets for the best concert. When someone Over the past decade international regula• ments, in which †nancial †rms borrow shouts Œ†re, their ‡ight creates an Œendog• tors have paid more attention to capital. short•term from investors with excess enous risk of being trampled by falling Banks ran liquidity stress tests and drew cash, such as money•market funds. This prices, margin calls and vanishing capi• up contingency funding plans, but often was a big problem because securities †rms tal‹a Œnegative externality that adds to half•heartedly. With markets awash with had become heavily reliant on this market, overall risk, says Lasse Heje Pedersen of cash and hedge funds, private•equity †rms tripling their repo borrowing in the †ve New York University. and sovereign•wealth funds all keen to in• years to 2008. Bear Stearns had $98 billion This played out dramatically in 2008. vest in assets, there seemed little prospect on its books, compared with $72 billion of Liquidity instantly drained from securities of a liquidity crisis. Academics such as Mr long•term debt. †rms as clients abandoned anything with Pedersen, Lubos Pastor at Chicago’s Booth Even the most liquid markets were af• a whi of risk. In three days in March Bear School of Business and others were doing fected. In August 2007 a wave of selling of Stearns saw its pool of cash and liquid as• solid work on liquidity shocks, but practi• blue•chip shares, forced by the need to cov• sets shrink by nearly 90%. After the col• tioners barely noticed. er losses on debt securities elsewhere, lapse of Lehman Brothers, Morgan Stanley What makes liquidity so important is caused sudden drops of up to 30% for had $43 billion of withdrawals in a single its binary quality: one moment it is there in some computer•driven strategies popular day, mostly from hedge funds. abundance, the next it is gone. This time its with hedge funds. Bob McDowall of Tower Group, a con• evaporation was particularly abrupt be• Liquidity comes in two closely connect• sultancy, explains that liquidity poses Œthe cause markets had become so joined up. ed forms: asset liquidity, or the ability to most emotional of risks. Its loss can prove The panic to get out of levered mortgage in• sell holdings easily at a decent price; and 1 The Economist February 13th 2010 A special report on †nancial risk 9

2 funding liquidity, or the capacity to raise †• nance and roll over old debts when need• ed, without facing punitive Œhaircuts on collateral posted to back this borrowing. The years of excess saw a vast increase in the funding of long•term assets with short•term (and thus cheaper) debt. Short• term borrowing has a good side: the threat of lenders refusing to roll over can be a source of discipline. Once they expect losses, though, a run becomes inevitable: they rush for repayment to beat the crowd, setting o a panic that might hurt them even more. ŒFinancial crises are almost al• ways and everywhere about short•term debt, says Douglas Diamond of the Booth School of Business. Banks are founded on this Œmaturity mismatch of long• and short•term debt, AIG did not allow for the risk that the in• global liquidity standard for international• but they have deposit insurance which re• surer would have to post more collateral ly active banks. Tougher requirements duces the likelihood of runs. However, this against credit•default swaps if these fell in would reverse a decades•long decline in time much of the mismatched borrowing value or its rating was cut. banks’ liquidity cushions. took place in the uninsured Œshadow Now that the horse has bolted, †nan• The new regime, which could be adopt• banking network of investment banks, cial †rms are rushing to close the door, for ed as early as 2012, has two components: a structured o •balance•sheet vehicles and instance by adding to liquidity bu ers (see Œcoverage ratio, designed to ensure that the like. It was supported by seemingly in• chart 4). British banks’ holdings of sterling banks have a big enough pool of high• genious structures. Auction•rate securities, liquid assets are at their highest for a de• quality, liquid assets to weather an Œacute for instance, allowed the funding of stodgy cade. Capital•markets †rms are courting stress scenario lasting for one month (in• municipal bonds to be rolled over month• deposits and shunning ‡ighty wholesale cluding such inconveniences as a sharp ly, with the interest rate reset each time. funding. Deposits, equity and long•term ratings downgrade and a wave of collater• The past two years are littered with sto• debt now make up almost two•thirds of al calls); and a Œnet stable funding ratio, ries of schools and hospitals that came a Morgan Stanley’s balance•sheet liabilities, aimed at promoting longer•term †nancing cropper after dramatically shortening the compared with around 40% at the end of of assets and thus limiting maturity mis• tenure of their funding, assuming that the 2007. Spending on liquidity•management matches. This will require a certain level of savings in interest costs, small as they systems is rising sharply, with specialists funding to be for a year or more. were, far outweighed the risk of market sei• Œalmost able to name their price, says one It remains to be seen how closely na• zure. Securities †rms became equally com• banker. ŒCollateral management has be• tional authorities follow the script. Some placent as they watched asset values rise, come a buzzword. seem intent on going even further. In Swit• boosting the value of their holdings as col• zerland, UBS and Credit Suisse face a tri• lateral for repos. Commercial banks in• Message from Basel pling of the amount of cash and equiva• creased their reliance on wholesale fund• Regulators, too, are trying to make up for lents they need to hold, to 45% of deposits. ing and on †ckle Œnon•core deposits, such lost time. In a †rst attempt to put numbers Britain will require all domestic entities to as those bought from brokers. on a nebulous concept, in December the have enough liquidity to stand alone, un• Regulation did nothing to discourage Basel Committee of central banks and su• supported by their parent or other parts of this, treating banks that funded them• pervisors from 27 countries proposed a the group. Also controversial is the compo• selves with deposits and those borrowing sition of the proposed liquidity cushions. overnight in wholesale markets exactly Some countries want to restrict these to the same. Markets viewed the second cate• Filling the pool 4 government debt, deposits with central gory as more eˆcient. Northern Rock, US banks’ cash assets, $trn banks and the like. The Basel proposals al• which funded its mortgages largely in capi• low high•grade corporate bonds too. tal markets, had a higher stockmarket rat• 1.25 Banks have counter•attacked, arguing HSBC ing than , which relied more on con• 1.00 that Œtrapping liquidity in subsidiaries ventional deposits. The prevailing view would reduce their room for manoeuvre in was that risk was inherent in the asset, not 0.75 a crisis and that the bu er rules are too re• the manner in which it was †nanced. strictive; some, unsurprisingly, have called At the same time †nancial †rms built 0.50 for bank debt to be eligible. Under the Brit• up a host of liquidity obligations, not all of ish rules, up to 8% of banks’ assets could be which they fully understood. Banks were 0.25 tied up in cash and gilts (British govern• expected to support o •balance•sheet enti• ment bonds) that they are forced to hold, ties if clients wanted out; Citigroup had to 0 reckons Simon Hills of the British Bankers 197375 80 85 90 95 2000 05 09 take back $58 billion of short•term securi• Association, which could have Œa huge im• Sources: Federal Reserve; Goldman Sachs ties from structured vehicles it sponsored. pact on business models. That, some ar•1 10 A special report on †nancial risk The Economist February 13th 2010

2 gue, is precisely the point of reform. tional investors lend shares from their sales, potential buyers will hold o for a Much can be done to reduce market portfolios to short•sellers for a fee. Some better price, exacerbating fair•value losses. stresses without waiting for these reforms. lenders‹including, notoriously, AIG‹ In future banks will be more alert to In repo lending‹a decades•old practice found they were unable to repay cash col• these dangers. ŒWe were looking at the critical to the smooth functioning of mar• lateral posted by borrowers because they bonds we held, focusing on the credit fun• kets‹the Federal Reserve may soon had invested it in instruments that had damentals. We lost sight of the capital hit toughen collateral requirements and force turned illiquid, such as asset•backed com• from illiquidity and marking to market borrowers to draw up contingency plans mercial paper. Some have doubled the that can seriously hurt you in the mean• in case of a sudden freeze. Banks that clear share of their portfolios that they know time, says Koos Timmermans, chief risk repos will be expected to monitor the size they can sell overnight, to as much as 50%. oˆcer at ING, a large Dutch banking and and quality of big borrowers’ positions Regulators might consider asking them insurance group. ŒWe now know that you more closely. The banks could live with to go further. Bond markets, unlike stock• have to treat the accounting reality as eco• that, but they worry about proposals to markets, revolve around quotes from deal• nomic reality. force secured short•term creditors to take ers. This creates a structural impediment to Another lesson is the Œopportunity val• an automatic loss if a bank fails. the free ‡ow of liquidity in strained times, ue of staying liquid in good times, says Another concern is prime brokerage, argues Ken Froot of Harvard Business Aaron Brown, a risk manager with AQR, a banks’ †nancing of trading by hedge School, because when dealers pull in their hedge fund. In an eˆcient market dollar funds. When the market unravelled, hedge horns they are unable to function properly bills are not left lying around. But in the dis• funds were unable to retrieve collateral as market•makers. He suggests opening up located markets of late 2008 there were that their brokers had Œrehypothecated, access to trade data and competition to lots of bargains to be had for the small mi• or used to fund transactions of their own; quote prices. Some senior †gures at the Fed nority of investors with dry powder. billions of such unsegregated money is like the idea, as do money managers, For some, though, bigger liquidity pro• still trapped in Lehman’s estate, reducing though predictably dealers are resisting. blems may yet lie ahead. Some $5.1 trillion dozens of its former clients to the status of of bank debt rated by Moody’s is due to unsecured general creditors. Brokers suf• Twin realities mature by 2012. This will have to be re†• fered in turn as clients pulled whatever The other brutal lesson of the crisis con• nanced at higher rates. The rates could also funds they could from those they viewed cerns the way liquidity can a ect solvency. be pushed up by an erosion of sovereign as vulnerable. Temporary bans on short• In a world of mark•to•market accounting, a credit quality, given implicit state guaran• selling made things even worse, playing small price movement on a large, illiquid tees of bank liabilities. And, at some point, havoc with some hedge funds’ strategies portfolio can quickly turn into crippling banks face a reduction of cut•price liquid• and leaving them scrambling for cash. Reg• paper losses that eat into capital. Highly ity support from central banks‹o ered in ulators are moving towards imposing lim• rated but hard•to•shift debt instruments return for often dodgy collateral‹which its on rehypothecation. can †nish you o before losses on the un• has buoyed their pro†t margins. Mortgage Early reform could also come to the se• derlying loans have even begun to hurt borrowers on teaser rates are vulnerable to curities•lending market, in which institu• your cash ‡ows. If markets expect †re payment shock. So too are their lenders. 7 Fingers in the dike

What regulators should do now

HE Delta Works are a series of dams, tion. The work is under way, but some bits ing the crisis. After a quarter•century of Tsluices and dikes built in the second are hobbled by a surfeit of architects, oth• ever•increasing †nancial concentration, half of the 20th century to protect the low• ers by a lack of clear plans. the giants of †nance grew even more dom• est•lying parts of the Netherlands from the ¹ Too big to fail. Dealing with Œsystemical• inant in 2008•09 thanks to a series of shot• sea. They are considered one of the seven ly important giants is the thorniest pro• gun takeovers of sickly rivals (see chart 5, wonders of the modern world. The task blem. Having once been cornered into a next page). facing global regulators is to construct the choice between costly rescues and gut• Regulators can tackle the issue either by †nancial equivalent of this protective net• wrenching failures, governments are de• addressing the Œtoo big part (shrinking or work, said Jean•Claude Trichet, president termined to avoid a repeat. When markets erecting †rewalls within giants) or the Œto of the European Central Bank, in an inter• swooned, they were obliged to stand be• fail bit (forcing them to hold more capital view last November. hind the big and the highly connected (as and making it easier to wind down bust This will require success in three con• well as their creditors), but found them• †rms). Until recently the focus was on the nected areas: reducing the threat to stabil• selves ill•equipped. Tim Geithner, Ameri• second of these approaches. But since Pres• ity posed by †rms deemed too big to fail ca’s treasury secretary, said his administra• ident Obama’s unveiling of two initiatives because their demise could destabilise tion had nothing but Œduct tape and string last month‹a tax on the liabilities of big markets; ensuring that banks have bigger to deal with American International banks and the ŒVolcker rule, which pro• cushions against losses; and improving Group (AIG) when it tottered. posed limits on their size and activities‹ system•wide, or macroprudential, regula• The problem has only worsened dur• momentum has been shifting towards 1 The Economist February 13th 2010 A special report on †nancial risk 11

voured groups, as happened in the govern• ment•orchestrated bankruptcy of General Motors. Another worrying precedent was the generous treatment of troubled banks’ derivatives counterparties in 2008. All counterparty trading exposures, to the ex• tent that they are uncollateralised, should be at the bottom of the capital stack, not at the top. Regrettably, the opposite hap• pened. This prompted a wave of credit•de• fault•swap buying because these contracts were underwritten by the state. ŒToday, too big to fail means too many counter• party exposures to fail, says Peter Fisher of BlackRock, a money manager. ¹ Overhauling capital requirements. In the hope of avoiding having to trigger their resolution regimes in the †rst place, regula• tors will force banks to strengthen their capital bu ers. A number of countries are considering a punitive capital surcharge for the largest †rms. A report from the Bank of England last November suggested va• rious ways of designing this. It could vary by sector, allowing regulators to in‡uence the marginal cost of lending to some of the 2 some combination of the two. regulators to seize and wind down basket• more exuberant parts of the economy. Or it The Volcker plan‹named after Paul cases. The challenge will be to convince could re‡ect the lender’s contribution to Volcker, the former Federal Reserve chair• markets that these measures will not turn systemic risk, based on its size, complexity man who proposed it‹calls for deposit• into life•support machines. Worse, there is and the extent of its connections to other †• takers to be banned from proprietary trad• no international agreement on how to nancial †rms. ing in capital markets and from investing in handle the failure of border•straddling How such a penalty would †t with hedge funds and private equity. The Finan• †rms, nor is one close. That was a huge pro• broader capital reforms is unclear. In De• cial Stability Board (FSB), a Basel•based blem with Lehman Brothers, which had cember the Basel Committee of supervi• body that is spearheading the internation• nearly 3,000 legal entities in dozens of sors and central banks laid out proposed al reform drive, gave it a cautious welcome, countries. And the struggle to retrieve $5.5 revisions to its global bank•capital regime. stressing that such a move would need to billion that a bust Icelandic bank owes These could come into force as early as be combined with tougher capital stan• creditors in Britain and the Netherlands 2012•13. The new standards, dubbed Basel dards and other measures to be e ective. still continues. 3, are less reliant than the last set of reforms The Volcker rule does not seek a full Questions also linger over the treat• on banks’ own risk models. Then the talk separation of commercial banking and in• ment of lenders. America’s plan wants it was of capital Œeˆciency. Now it is all vestment banking. Nor is America pushing both ways, giving regulators discretion to about robustness. With markets already to shrink its behemoths dramatically; for override private creditors but also to subor• demanding that banks hold more equity, a most, the plan would merely limit further dinate the taxpayer’s claims. This fuels reversal of a long trend of falling ratios is growth of non•deposit liabilities (there is concerns about handouts to politically fa• under way (see chart 6, next page). already a 10% cap on national market share Before the crisis banks could get away in deposits). Oˆcials remain queasy about with common equity‹the purest form of dictating size limits. Citigroup’s woes sug• Big banks get bigger 5 capital‹of as little as 2% of risk•weighted gest a †rm can become too big to manage, Top five global banks and hedge funds assets. The new regulatory minimum will but JPMorgan Chase and HSBC are striking Assets as % of industry total not be clear until later this year, but mar• counter•examples. 30 kets now dictate that banks hold four to For all the hue and cry about the †ve times that level. Hybrid instruments‹ 25 Volcker plan, America sees it as supple• Hedge funds part debt, part equity‹will be discouraged menting earlier proposals, not supplanting 20 since these proved bad at absorbing losses. Global banks them. The most important of these is an 15 Regulators are encouraging banks to issue improved Œresolution mechanism for fail• a di erent type of convertible capital: 10 ing giants. Standard bankruptcy arrange• Œcontingent bonds that automatically ments do not work well for †nancial †rms: 5 turn into common shares at times of stress. in the time it takes for a typical case to 0 In another acknowledgment that rely• grind through court, the company’s value 1998 200099 01 02 03 04 05 06 07 08 09 ing too heavily on internal models was a will have evaporated. Source: “Banking on the State” by Andrew Haldane and mistake, the new rules will be supple• Piergiorgio Alessandri; Bank for International Settlements America’s resolution plan would allow mented by a Œleverage ratio. Not weighted1 12 A special report on †nancial risk The Economist February 13th 2010

2 to risk, this measure looks appealingly sim• America’s large banks, having repaid on how systemic regulation might work, ple these days. One aim is to curb gaming their debts to taxpayers, are sure to wage or who should do the regulating. Most of risk•based requirements: European war on higher capital standards. An impact economists see the job falling naturally to banks, which unlike American ones were assessment stretching over several months central banks, because of their closeness to not subject to a leverage ratio, could take will give them ample opportunity to look markets and because of the link between their borrowing to dangerous heights be• for holes‹and to lobby. In Europe, where capital standards and monetary policy cause many of their assets were highly rat• banks were more highly leveraged and through the price of credit. But there are ed securities with low risk weightings. thus face a more wrenching adjustment, political obstacles, particularly in America, One oˆcial likens the new approach to even some supervisors are queasy. where a large and vocal contingent in Con• placing a net under a trapeze artist. ¹ Improving macroprudential regulation. gress accuses the Fed itself of being a threat In an equally big philosophical shift, In the meantime regulators can make pro• to stability, pointing to loose monetary the new measures will lean against Œprocy• gress in other areas, such as overhauling policy as a cause of the housing mania. clicality, or the tendency of rules to exag• day•to•day supervision. In both America International co•ordination is equally gerate both the good and the bad. Banks and Europe they have stepped up compari• tricky. The FSB has singled out 30 of the will be required to accumulate extra capi• sons of pay, lending standards and the like largest banks and insurers for cross•border tal in fat years that can be drawn upon in across big †rms. They are also introducing scrutiny by Œcolleges of supervisors. lean ones. Until now the rules have en• peer review. Within the agency that over• There is, though, a natural limit to co•oper• couraged higher leverage in good times sees Swiss banks, for instance, the lead su• ation. It remains to be seen how well na• and much lower in bad times, adding to pervisors of Credit Suisse and UBS are tional risk regulators work with suprana• distress at just the wrong moment. Securi• now expected to scrutinise each other’s tional bodies such as the European ties regulators contributed to the problem, work. America’s Securities and Exchange Union’s systemic•risk council and the FSB. frowning on boom•time reserve•building Commission, whose failures included neg• Private•sector groups want to have their as possible pro†t•smoothing in disguise. ligible supervision of investment banks say too: the Market Monitoring Group, a The new proposals will encourage Œdy• and the Mado scandal, has set up a new collection of grandees linked to a banking• namic provisioning, which allows banks risk division packed with heavyweight industry group, is already issuing warn• to squirrel away reserves based on expect• thinkers such as Henry Hu, Gregg Berman ings about fresh bubbles emerging. ed losses, not just those already incurred. and Richard Bookstaber. Part of their job Another reason for scepticism is the dif• Addressing procyclicality will also re• will be to scan derivatives markets, hedge †culty of identifying a systemic event. quire tackling issues that straddle capital funds and the like for any emerging threats AIG’s liquidity crunch was thought to rules and accounting standards. Critics of to stability. count as one at the time, hence the o er of fair•value (or Œmark•to•market) account• This stems from a recognition that tradi• an $85 billion emergency loan from the ing, which requires assets to be held at tional oversight needs to go hand in hand Fed. But what exactly was the danger? That market prices (or an approximation), com• with the macroprudential sort that takes markets would be brought to their knees plain that having to mark down assets to account of the collective behaviour of †• by the failure of its derivatives counterpar• the value they would fetch in illiquid mar• nancial †rms, contagion e ects and so on. ties (who were controversially paid o at kets is likely to exacerbate downturns. The ŒFinance is full of clever instruments that par)? Or by trouble at its heavily regulated solution is not to abandon fair value, work as long as the risk is idiosyncratic, but insurance businesses? More than a year which investors like because it is less open can wreak havoc if it becomes systemic, later, no one seems sure. to manipulation than the alternatives. But says Frederic Mishkin of Columbia Uni• Pricking bubbles‹another mooted role there is a case for decoupling capital and versity. Moreover, the crisis showed how for systemic regulators‹is also fraught accounting rules, says Christian Leuz of risk can cross traditional regulatory lines. with danger. Many central bankers consid• the Booth School of Business. This would Pension funds and insurers, previously er it unrealistic to make prevention of as• give bank regulators more discretion to ac• seen as shock•absorbers, were revealed as set•price bubbles a speci†c objective of cept alternative valuation methods yet still potential sources of systemic risk. systemic oversight. But thinking at the Fed allow investors to see the actual or estimat• However, there is no broad agreement has been shifting. Under Alan Greenspan ed market value. its policy had been to stand back, wait for There are lots of potential devils in the the pop and clean up the mess. But Ben details of the proposals. A leverage ratio is Threadbare cushion 6 Bernanke, the current chairman, recently pointless without strict monitoring of as• Banks’ capital ratios, % backed the idea of intervening to take the sets parked o balance•sheets. Contingent air out of bubbles. This could be done capital, meanwhile, could have the oppo• 25 mainly through stronger regulation, he United States site e ect of that intended if the bank’s 20 suggested, though he did not rule out mon• trading partners ‡ee as its ratios near the etary policy as a back•up option. trigger point. There are also worries over 15 Mr Mishkin, a former Fed governor, increases in capital charges for securitisa• draws a contrast between credit•boom tions, exposure to swap counterparties 10 bubbles and irrational exuberance in and the like. These make sense in theory; Britain stockmarkets, such as the dotcom bubble. 5 to treat mortgage•backed securities as al• The †rst is more dangerous, and the case most risk•free was nonsense. But the new 0 for pre•emptive action stronger, he argues, rules swing too far the other way, threaten• 1880 1900 20 40 60 80 2005 because it comes with a cycle of leveraging ing to choke o the recovery of asset• Source: “Banking on the State” by Andrew Haldane and against rising asset values. Piergiorgio Alessandri; Bank for International Settlements backed markets. In retrospect all crashes look inevitable. 1 The Economist February 13th 2010 A special report on †nancial risk 13

2 Even with the most insidious•looking bub• heavyweights to provide institutional above•average rates. That would shield it bles, though, it is impossible to know at the memory of past crises. For similar reasons, from claims that the next boom is some• time how devastating the pop will be. Andrew Lo, director of MIT’s Laboratory how di erent and should be left to run its Many thought the economic fallout from for Financial Engineering, suggests sepa• course. But it comes at the cost of ‡exibility. the internet crash would be far greater rating regulation from forensics, as hap• All of this suggests that although there than it turned out. The economic cost of pens in the airline industry. America’s Na• is a strong case for a more system•wide ap• prematurely ending a boom can be high. tional Transportation Safety Board is seen proach to oversight, it could do more harm Even so, the worry is that a systemic regu• as independent because its job is to investi• than good if poorly crafted. Meanwhile lator would be biased towards interven• gate crashes, not to set rules after the event. taxpayers will continue to underwrite †• tion, because it would face less criticism for That gives it more moral clout. nancial giants; America’s reforms in their puncturing a non•bubble than for failing to Whatever form it takes, systemic polic• current shape allow the authorities to pull spot a real one. ing would face a problem. During booms, apart those that pose a Œgrave threat, but Alex Pollock of the American Enter• governments are loth to take the punch• also to bail out their creditors if they con• prise Institute (AEI), a think•tank, is con• bowl away, at least until the next election. sider it necessary. cerned that the creation of an oˆcial sys• Nor do they want to be criticised for their The danger is that the very existence of temic regulator would bring false comfort, own contribution to systemic risk. They a systemic regulator creates an illusion of just as the Fed’s founding in 1913 did. Ac• may have become even touchier now that increasing stability even though it does the cording to the then Comptroller of the Cur• they are, as Pimco’s Mohamed El•Erian opposite by strengthening the implicit rency, it had rendered further crises Œmath• puts it, Œmarket players as well as referees. guarantee for the biggest banks‹a Œperma• ematically impossible. Mr Pollock would A way round this would be to introduce nent TARP, as the AEI’s Peter Wallison prefer to see the task go to an independent rules requiring the regulator to step in if, puts it. Too big to fail sometimes seems too advisory body, manned by economic say, credit and asset prices are growing at hard to solve. 7 Blocking out the sirens’ song

Moneymen need saving from themselves

ISK antennae twitch after a crisis. Bank• in preparation for central banks’ with• Rers, regulators and academics, shaken drawal of monetary adrenaline. Parti pris 7 from their complacency, jostle to identify Many will already be doing this as they Global asset-backed-securities issuance the next tempest. Right now gusts are try to show that they have learnt their les• and Moody’s profit per employee blowing from several directions. Many sons. Like the best chess players, bankers Total ABS issuance, Profit per employee, countries’ †scal positions are deteriorating insist that they are now concentrating as $trn $’000 1.25 250 fast after costly interventions to shore up †• hard on avoiding mistakes as on winning. nancial systems and restore growth. There Those that sidestepped the worst mort• 1.00 200 is talk of demanding collateral even on gage•related landmines now top the indus• 0.75 150 deals with formerly unimpeachable sover• try’s new order. Blackrock’s Mr Fisher de• eign entities. Recent terrorist incidents †nes risk as Œdeviation from objective, on 0.50 100 have raised the spectre of external shocks. the upside as well as the downside. If your 0.25 50 Yet at least a fragile sort of optimism has models tell you that a security is safer than surfaced, born of ultra•cheap money and its returns imply, as with CDOs, that might 0 0 relief at having avoided a depression. In just suggest hidden risks. 2000 01 02 03 04 05 06 07 08 Sources: Company accounts; Bloomberg; Benn Steil some markets fresh bubbles may be form• Fancy mathematics will continue to ing. Stockmarkets have rebounded sharp• play a role, to be sure. But †nance is not ly. America’s, though still well o their physics, and markets have an emotional mortgage †asco showed, they are vulner• peaks, are up to 50% overvalued on a his• side. In their struggle with the quants, able to capture by those they police. torical basis. Banks are once again throw• those who would trust their gut instinct Their job will be made easier if new ing money at hedge funds and private•equ• have gained ground. rules tackle incentives for the private sec• ity †rms (though with tougher margin tor to take excessive risk. It is human be• requirements). Issuance of structured•loan Learning to tie knots haviour, more than †nancial instruments, funds, which a few months back looked Governments are taking no chances. that needs changing, says Mr Mauboussin dead, is booming. Investment banks’ pro• Bloomberg counts some 50 bills and other of Legg Mason. Like Odysseus passing the †ts, and bonus pools, are back near pre•cri• serious proposals for †nancial reform in sirens, bankers need to be tied to a mast to sis levels. International regulators have America and Europe. Leaders in America’s stop them from giving in to temptation. been issuing warnings to chief executives Senate hope to pass new rules by March. Pay structures should be better aligned about a return of irrational exuberance. But there are limits to what can be expect• with the timescale of business strategies Banks have been ordered to run stress tests ed from regulators and supervisors. Like that run for a number of years and should involving a sudden jump in interest rates, bankers, they have blind spots. As the not reward Œleveraged beta, unremark•1 14 A special report on †nancial risk The Economist February 13th 2010

o erings, which is bad for consumers. Most would agree, however, that mar• kets need both tighter rules and better en• forcement. The biggest question mark hangs over the fate of those institutions whose collapse would threaten the sys• tem. America’s proposal to cap banks’ size and ban proprietary trading has forti†ed those calling for radical measures to tackle the Œtoo big to fail problem.

The virtues of scepticism By itself, though, the plan does little to back up Barack Obama’s promise to stop such †rms from holding the taxpayer hostage. Proprietary trading and investments are a small part of most big banks’ activities and played only a minor role in the crisis. Nor does the plan cover brokers, insurers or in• dustrial †rms’ †nance arms, all of which had to be bailed out. To persuade markets that the giants no longer enjoy implicit state guarantees, whether they are banks or not, policymakers will need to present a 2 able returns juiced with borrowed money. tive to equity. Phasing this out would dis• cocktail of solutions that also include In securitisation, originators will have to courage the build•up of excessive leverage. tougher capital and liquidity standards, disclose more information about loan But the idea has little political traction. central clearing of derivatives and credible pools and hold a slice of their products. There are, to be sure, risks to rushing re• mechanisms to dismantle †rms whose Some of the worst abuses in securitisa• form. Post•crisis regulation has a long his• losses in a crisis would overwhelm even a tion stemmed from the use of credit rat• tory of unintended consequences, from strengthened safety bu er. ings. Rating agencies systematically under• the pay reforms of the early 1990s (when Together, intelligent regulatory reforms estimated default risk on vast amounts of new limits on the deductibility from cor• and a better understanding of the limita• debt, resulting in pu ed•up prices and a porate tax of executive salaries merely tions of quantitative risk management can surfeit of issuance. Paid by issuers, they shifted the excesses to bonuses) to key help to reduce the damage in‡icted on the had every incentive to award in‡ated rat• parts of the Sarbanes•Oxley act on cor• †nancial system when bubbles burst. But ings and keep the market humming: aver• porate governance. Another danger is the they will never eliminate bad lending or age pay at the agencies rose and fell in tan• pricing of risk by regulation, not markets. excessive exuberance. After every crisis dem with the volume of asset•backed The credit•card act passed in America last bankers and investors tend to forget that it issuance (see chart 7, previous page). year leaves providers with a choice be• is their duty to be sceptical, not optimistic. An obvious way to deal with this tween underpricing for some products, In †nance the gods will always †nd a way would be to eliminate the agencies’ oˆcial which is bad for them, or restricting their to strike back. 7 Œnationally recognised status, opening the business to unfettered competition. O er to readers Raters would then have to persuade inves• Future special reports Reprints of this special report are available at a Managing information February 27th tors of their competence, rather than rely• price of £3.50 plus postage and packing. Germany March 13th ing on a government imprimatur. This, in A minimum order of †ve copies is required. America’s economy April 3rd turn, would force investors themselves to Management innovation in emerging Corporate o er markets April 17th spend more time analysing loans. Oddly, Customisation options on corporate orders of 100 proposed reforms fall far short of this, with Television May 1st or more are available. Please contact us to discuss International banking May 15th no sign of anything tougher on the hori• your requirements. zon. CreditSights, a research †rm, awarded Send all orders to: ratings †rms its ŒHoudini was an Ama• teur award for 2009. The Rights and Syndication Department Nor, alas, is there much appetite to 26 Red Lion Square WC1R 4HQ tackle some of the public policies that con• London Tel +44 (0)20 7576 8148 tributed to the crisis. The non•recourse sta• Fax +44 (0)20 7576 8492 tus of mortgages in large parts of America, e•mail: [email protected] for instance, gives the borrower a highly at• tractive put option: he can, in e ect, sell the For more information and to order special reports Previous special reports and a list of house to the lender at any time for the prin• and reprints online, please visit our website forthcoming ones can be found online cipal outstanding. An even bigger problem Economist.com/rights Economist.com/specialreports is the favourable tax treatment of debt rela•