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 denes ed by larger balance sheets and greater le• Tthe 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, Against 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 great 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 reected 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 value•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. It 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 quants 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 prots 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 alpha, 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 black 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 shadow 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 Reputation 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 Oce of the Comptroller of the Cur• the information to determine the conse• wild randomness not usually seen in na• rency, an American regulator. quences of others failing. Losses on con• ture. In markets, rare big changes can be Everywhere you looked, moreover, in• tracts linked to Lehman turned out to be more signicant than the sum of many centives were misaligned. Firms deemed modest, but nobody knew that when it small changes, he says. If nancial mar• too 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 Greenspan put, a belief that America’s sures to tail 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 disaster 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 amplied 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).