On August 14, David Viniar, the chief financial officer of Goldman, Sachs & Co., struggled to explain why two hedge funds managed by his firm had suffered severe losses during the first few hectic days of the month. He came up with the rationale that the markets had been hit by a rare "25 standard deviation event." In other words, an occurrence so rare that a person living since the dawn of time would be lucky to have witnessed it. The comment was, of course, far-fetched. History shows that financial crises are as inevitable as death and taxes.
Viniar's comment appears particularly off the mark given the euphoric behavior of the financial markets in the years leading up to this summer's credit crunch. This euphoria was initially induced by a period of low interest rates that followed the collapse of the technology bubble at the turn of the century. That things were carried to an extreme even after the Federal Reserve Board began to raise rates in the summer of 2004 is largely the result of recent changes in the financial system. Credit had became a game of hot potato, with loans passed around and no one taking responsibility for the final outcome. Given this reckless behavior, it was only a matter of time before a crisis appeared.
The credit cycle in history
Recent events in the financial markets have followed a course familiar to students of history. The traditional credit cycle starts with low interest rates. As credit expands, prosperity returns. Over time, lending standards decline. Good bankers have always known that to lend long and borrow short is the fastest way to the bankruptcy court. But when liquidity is abundant, it's easy to roll over short-term loans, and there are fat profits to be made from lending against illiquid collateral. Competition among lenders ensures that those who adhere to conservative practices are pushed aside. Confidence runs high. There is a belief that prosperity will last forever.
But in time prices rise and interest rates follow. It becomes apparent that corners have been cut. "The good times almost always engender much fraud," wrote Walter Bagehot in 1873 in Lombard Street. "All people are most credulous when they are most happy; and when money has been made, when some people are really making it, when most people think they are making it, there is a happy opportunity for ingenious mendacity."
A bank, or some other important financial player, suddenly fails. The careless confidence of the boom is replaced by suspicion and fear. No one is sure whom to trust. Everyone seeks the safety of cash. "As credit by growing makes itself grow," wrote the 19th-century economist Alfred Marshall, "so when distrust has taken the place of confidence, failure and panic breed failure and panic." Credit suddenly contracts and liquidity dries up. The central bank is forced to intervene to shore up the financial system.
Historical precedent doesn't help us predict at what exact date a boom will end, which is perhaps why most people on Wall Street don't bother to read up on the past. Still, there are several leading indicators of a crisis. They include the rapid expansion of credit above its long-term trend; financial innovation and deregulation; greater competition among lenders, and the arrival of new entrants; the evolution of fragile debt structures excessively dependent on liquidity; a growing mismatch between assets and liabilities among financial players and a rise in short-term borrowing; falling risk premiums for loans and investments; and, most important of all, the appearance of an asset price boom.
These leading indicators are not precise. Yet when they are present simultaneously, it is pretty likely that a credit crunch will appear at some stage. Given that all these factors were present in the financial markets in the years running up to this summer's events, a crisis was surely inevitable. The only mystery was when it would arrive.
The role of the authorities
In the past it was unusual for one panic to follow hard upon another. Crises tended to appear roughly once a decade -- ten years being the amount of time evidently needed for people to forget their previous follies and woes. What's remarkable about this summer's crisis is that it appeared only five years after a the severe credit crunch associated with the collapse of Enron Corp. and WorldCom.
So why did it take a mere five years for the current wave of credit to swell to the breaking point? The shortening of the cycle is largely because of the actions of the Federal Reserve Board, whose monetary policy generated a housing bubble shortly after the end of the dot-com mania. On its own a deflating real estate market would probably have been enough to produce a financial crisis of sorts. But that outcome was assured by the irresponsible behavior of just about every participant in a newly constituted credit system that had been endorsed by the authorities for its superior qualities in managing and distributing risk.
Under then-chairman Alan Greenspan the official policy of the Federal Reserve was that speculative bubbles could not be accurately identified before they popped and, therefore, the authorities should do nothing to hinder their growth. It is enough, the thinking went, for the central bank to deal with the bubble's aftermath. Following this line, the Fed waited for the Internet bubble to burst, which duly happened a few months into the new millennium. Subsequently, interest rates were slashed from 6.5 percent at the end of 2000 to 1 percent in June 2003 and remained at that level for 12 months.
After a brief recession, the U.S. economy recovered. Yet this apparently successful response to a deflating bubble had profound consequences. Low interest rates made it attractive for households to borrow. Cheap mortgages fueled a housing boom and enabled Americans to live above their incomes. "We know," Greenspan declared in February 2004, "that increases in home values and borrowing against home equity likely helped cushion the effects of the declining stock market during 2001 and 2002."
At the time, Greenspan denied that rapidly rising home prices were a cause for concern, claiming, rather implausibly in light of countless past real estate booms and busts, that "houses aren't as prone to bubbles as stocks because high transaction costs and a seller's need of shelter are significant impediments to speculative trading." The housing bulls pointed out that the average price of a U.S. home had never declined and never would (this claim wasn't strictly true, but it served its purpose). Besides, they reasoned, if there were ever any trouble in the housing market, the Federal Reserve would surely lower rates.
The debacle in the subprime mortgage market, which sparked the crisis on Wall Street, has brought a widespread change of opinion. Median U.S. home prices are set to decline this year, according to the National Association of Realtors. The Fed's easy money policy in the wake of the dot-com collapse is now widely held responsible for the housing bubble. Edward Gramlich, a Fed governor from 1997 to 2005, recently told the Wall Street Journal that he failed to realize at the time that low rates were making it so easy to issue subprime loans. The Journal's editorial board, formerly a staunch supporter, admits that "the Greenspan Fed is one reason for the current mortgage mess."
Greenspan's policy also fostered an appetite for risk-taking in the financial markets. In the fall of 1998, the prospective collapse of the overleveraged hedge fund Long-Term Capital Management put Wall Street in a deep funk. But the Fed came to the rescue, arranging a bailout for LTCM and twice cutting interest rates. The stock market soon rebounded, and the "Greenspan put" was born. Wall Street learned that panics were likely to be short and relatively painless.
This "put" was exercised again after the failure of Enron. Between 2002 and 2004, short-term interest rates were kept below the rate of inflation. By the summer of 2004, the spread between short- and long-term interest rates was at its highest level in half a century, making it extremely profitable to borrow in the money markets and acquire longer-dated securities. The so-called carry trade became a license for hedge funds, banks and others to print money. Debt leverage, as measured by net borrowing by primary dealers in the repo markets, soared after 2002. The carry trade also provided an incentive for lenders to borrow short and lend long, producing the mismatch of assets and liabilities that generally precedes a financial crisis.
Many expected the carry trade to end after interest rates rose and the yield curve flattened towards the end of 2005. But that didn't happen. It still paid to borrow and acquire riskier debt securities, such as junk bonds and leverage loans, despite the fact their premiums were near record lows. As the "dash to trash" continued, bond market leverage crept upwards.
The Federal Reserve doesn't just set monetary policy. It is also the nation's top bank regulator. In recent years the credit system has changed profoundly. Once, banks played the key role in the credit system, lending and keeping loans on their balance sheets until they were repaid. Now most loans are bundled together, turned into securities and sold in the secondary markets to be bought by hedge funds and others.
Despite the LTCM panic in 1998, Greenspan used his influence to fend off regulation of this new securitized credit system. Hedge funds, he argued, helped to improve the distribution and pricing of risk, producing a more efficient allocation of capital and making the financial system more stable. His views were reiterated by Timothy Geithner, president of the Federal Reserve Bank of New York in May 2005. "By spreading risk more broadly, providing opportunities to manage and hedge risk and making it possible to trade and price credit risk, credit market innovation should help make markets both more efficient and more resilient," he said.
Yet is credit really better priced and distributed than ever before? The activities of lenders and borrowers in the years leading up to this summer's crisis suggest otherwise.
The modern credit system and its flaws
The history of banking has not been without blemish, hence the continual cycle of crises over hundreds of years. Bankers have always moved in shoals, favoring supposedly safe borrowers -- whether third-world countries in the 1970s or owners of Japanese real estate in the following decade -- only to discover their errors later. As John Maynard Keynes wrote in the 1930s, "A sound banker, alas, is not one who foresees danger and avoids it, but one who, when he is ruined, is ruined in a conventional and orthodox way along with his fellows, so that no one can really blame him."
Despite its flaws, there were certain advantages to traditional banking. Bankers knew their borrowers. They had long-standing relationships that, in theory at least, were based on trust between both parties. Because his employer retained loans until they were repaid, the banker had a career interest in seeing that loans were prudently made. Today participants focus on the fees that they will immediately earn. Incentives have encouraged myopic behavior.
Consider how this works in the mortgage market. A borrower gets an appraisal on a property. In this competitive market, the appraiser with the most generous valuation is likely to land the job. The loan applicant then goes to a mortgage broker, who is paid according to how many loans he or she originates. This creates an incentive to cut corners, possibly exaggerating the borrower's income. The loan application is handed to a bank, which earns a fee from packaging the mortgages together and a profit from selling them to Wall Street. The bonds are sent to the credit agencies, which stamp them with an investment-grade rating in exchange for another fee. Wall Street firms get their revenue from distributing mortgage loans to their clients.
Alternatively, these loans can be stuffed into yet another security, such as a collateralized debt obligation. This provides another fee bonanza for investment banks, rating agencies and the managers of securitized loans. A similar process is at work in the market for leveraged loans, which finance corporate buyouts. Banks earn fees of about 2 percent from financing buyouts but seek to get these loans off their books as fast as possible, often by packaging them into structured securities. In short, the aim of the modern bank is to be "credit neutral" -- in other words, to maximize fee income and protect itself as much as possible against adverse outcomes.
Nor is it clear that the buyers of Wall Street's securitized loans have a much greater interest in the eventual outcome. Hedge fund managers generally take a 20 percent cut of any performance gains, paid annually. The skill of hedge fund managers is generally measured by comparing returns with the volatility of portfolios. That makes investing in complex debt securities very attractive. These loans tend to produce large payoffs with little volatility during good times. One of the Bear, Stearns & Co. hedge funds boasted 51 months of consecutive positive returns before it blew up this summer.
The trouble is that this investment strategy can completely wipe out investors' funds when the loans sour, hedges fail and liquidity dries up, forcing the leveraged hedge fund to sell assets into a declining market. Still, owing to the compensation structure, extreme competition and the need to justify exorbitant fees by producing large returns, there is an overwhelming incentive to place such bets.
In a recent paper, "Cracking the Market Code," Henry Maxey of London-based fund manager Ruffer reveals the fundamental problem with the securitized credit system. The system suffers from a fallacy of composition, Maxey writes: Each participant behaves rationally with regard to his or her own interest, but the outcome is irrational for all. Hence underwriting standards declined dramatically during the housing boom and continued declining into 2006 despite evidence that the bubble was beginning to deflate and delinquencies on subprime loans were picking up. And excesses appeared in the leveraged buyout world earlier this year, as covenants on loans were torn up, payment-in-kind, or "toggle," notes proliferated, and ratios of debt to cash flow continued rising.
Modeling credit risk
The skewed incentives of participants in the credit markets are no secret. But there's a more fundamental problem with securitized credit. Bagehot defined credit as "the disposition of one man to trust another." In the modern credit system the activity of "trusting" has been handed over to machines, or rather quantitative risk models.
When people acquire a bundle of credits packaged in a complex security, such as a CDO, they often have no understanding of the underlying risks. Instead, they rely on the security's credit rating. The world of securitized credit is inconceivable without the ratings provided by a handful of firms. The subprime debacle has dented confidence in these arrangements. Slow to downgrade subprime loans, the rating agencies are now coming under fire. These firms face a potential conflict of interest; they are supposed to represent the interests of investors but are paid by bond issuers. A large chunk of their fees now come from rating securitized debt. Some, including Ohio Attorney General Marc Dann, claim that the rating agencies have worked too closely with the investment banks, providing inherently risky loans with an investment-grade stamp. They have turned pigs' ears into silk purses. The rating agencies deny such charges. They also deny responsibility for their errors, claiming that ratings are an expression of opinion without legal liability.
But even if one could resolve the rating agencies' conflict of interest and their exercise of power without responsibility, there would still be a problem with the models they employ. For all their mathematical sophistication, risk models are inherently flawed. They assume that the future in large measure will resemble the past and seek to extract probabilities of loss from historical data. The trouble is that credit data doesn't go back far enough to draw robust probabilistic inferences of future losses.
And even if there were more data, there would still be a problem. Financial outcomes don't fall in normal distributions, such as those produced by throwing dice. Rather, they are influenced by complex feedback effects. For instance, when Bear Stearns announced in June severe losses on its subprime investments, it provoked a sell-off in other risky assets, which in turn caused liquidity in the credit markets to evaporate, bringing a whole new set of problems. Although risk modeling has become more sophisticated since LTCM, models will always remain incapable of handling market dislocations. Hence, Goldman's CFO complaining about the "25 standard deviation event."
Applying the principles of physics to finance won't work, says Emanuel Derman, author of My Life as a Quant: Reflections on Physics and Finance and a former Goldman Sachs managing director, "because the universe is time invariant, while people's behavior isn't." Take the risk models used by the rating agencies to assess subprime mortgages. These models assumed that home prices would continue rising, just as they had in the past. They also assumed that correlations of defaults among the various mortgage loans would remain constant.
Common sense would suggest that these assumptions were increasingly questionable once a nationwide bubble had appeared in the housing market and underwriting standards had declined. Unfortunately, the rating agencies' models lacked such common sense. "Credit," wrote Bagehot, "is an opinion generated by circumstances and varying with those circumstances." Credit opinions formed by mathematical models will never be able to grasp the varying circumstances of borrowers' and lenders' behavior in a constantly changing world.
The search for culprits in the recent credit bubble might continue more or less indefinitely. It would encompass institutional investors so eager to make good portfolio losses at the beginning of this decade that they rushed into hedge funds, ignoring their excessive fees and skewed incentives. Often, these investors were persuaded by the argument that credit constitutes a separate asset class, uncorrelated to traditional investments. Then, there are the banks that engaged in ratings arbitrage, acquiring complex securitized products, such as the 15-times-leveraged Constant Proportion Debt Obligation (recently trading at 70 cents on the dollar), which because of its triple-A rating required less regulatory capital. Even the cautious managers of money market funds couldn't resist the few extra basis points to be earned from buying commercial paper from so-called structured investment vehicles, which turn out to have been leveraged to 20 times and invested in subprime mortgage securities.
Many believe that this summer's crisis was merely the result of a sudden disappearance of liquidity; once that passes, it will be business as normal. They place their trust in Greenspan's heir, confident that Bernanke's recent injections of liquidity into the financial system and a cut in the rate at which the Fed lends to banks will do the trick. The initial positive reaction of the stock market reflects this view.
Yet this view assumes that Greenspan's remedies to past crises still retain their efficacy. It ignores the main consequences of the Greenspan put -- namely the inflation of a technology bubble after the LTCM debacle and of a housing bubble after the tech wreck. The soaring real estate market after 2002 encouraged Americans to live beyond their means. According to the London-based independent economist Andrew Hunt, U.S. household spending has lately been running close to 20 percent above posttax income. Paul Kasriel, chief economist at Northern Trust Co., observes that American households have produced annual deficits on only 13 occasions since 1929. Yet since 1999, they have been in continuous deficit. The vast bulk of consumer borrowing has been mortgage-related.
Now house prices are falling. In the second quarter of this year, the S&P/Case-Shiller U.S. national home price index recorded the largest drop in its 20-year history. Mortgage credit is becoming more expensive and less available. These developments are already reflected by a sharp decline in consumer confidence in August. An increasing number of businesses, in particular retailers such as Home Depot, are feeling the impact of a weakening housing market.
During the credit boom, the earnings of the financial sector rose to roughly 40 percent of the profits of the companies in the Standard & Poor's 500 index. As the leveraged carry trade dwindles, the profitability of banks, hedge funds and other providers of "alternative investment" services will take a hit. Those who argue that the U.S. economy is strong enough to withstand the tremors ignore the fact that record corporate profits are largely a consequence of easy credit, which bolstered consumer demand and inflated financial sector revenues. "Main Street," says Kasriel, "is more dependent on Wall Street than ever before."
If stock market valuations are depressed by falling profits, as seems likely, and real estate prices continue falling, then household balance sheets will suffer. At that point, U.S. consumers may well be prompted to spend less and save more. The virtuous cycle of prosperity created by the credit boom may turn into a vicious cycle of decline. If that were to happen, Bernanke would follow Greenspan's example and slash interest rates. But it's difficult to envisage such a move's inflating another asset bubble large enough to offset the effects of the declining $23 trillion real estate market. The central bank could well find itself pushing on the proverbial piece of string.
Then there's the supply of credit to consider. This summer's events have shaken confidence in the securitized credit system. People no longer trust the ratings on complex debt securities. Hedge funds are likely to find it harder to raise funds now that their risk models have once again been proven faulty. Risk premiums are rising and investors in securitized debt are likely to remain wary for a while to come.
There has been a run on many "virtual banks" -- mortgage REITs, SIVs, CDOs and other structured product vehicles. Many of these blind repositories of credit are fit only for bull market conditions, says Charles Peabody founder of New Yorkbased research firm Portales Partners.
That means the traditional banking system will have to take up the slack. After years of fat profits, banks are currently well capitalized. But they are already tightening lending standards and will probably prove reluctant to further increase their exposure to real estate and leveraged buyouts. Given these circumstances, it's likely that the recent flood of credit will start to ebb.
This summer's convulsions in the financial markets, the sudden implosion of overleveraged hedge funds, the belated downgrades on subprime loans, losses in obscure off-balance-sheet investment vehicles and, above all, the extraordinary decline in underwriting standards for mortgages and corporate loans, suggest there is something deeply "unsound" about our modern credit system. That this system has now been discredited is largely the result of the misguided actions of central bankers and the myopic self-interest of virtually every participant in the credit markets, from the grandest Wall Street banker down to the humblest of mortgage brokers. Common sense dictates there is a cost to financial recklessness. We must now wait to see how great the final reckoning will be.
Edward Chancellor is an editor at breakingviews.com. His "Ponzi Nation," predicting the collapse of the credit bubble, appeared in the February 2007 issue of II.