Chapter 18. Quant Funds
“Because of the lack of transparency and coordination within the hedge-fund industry and the strong relationship between performance and business viability, competitive pressures will lead managers and prime brokers to increase leverage in an ‘arms race’ for generating better returns.”
Andrew W. Lo of Massachusetts Institute of Technology
Too much money tends to chase good ideas until they get overcrowded, setting up a crash when the crowd goes the other way. When the same investors crowd into many different investments, as quant funds did in 2007, the herd can move in surprising directions. This happened when a credit hedge fund asked its lenders for leniency in June 2007 and sparked a crisis for unrelated quantitative equity funds.
On June 19, 2007, a few days after the bond market revolt, a hedge fund run by Bear Stearns, Wall Street’s fifth biggest broker, asked its creditors for a rescue.1 The High-Grade Structured Credit Strategies Enhanced Leverage fund had raised about $600 million from investors and then borrowed another $6 billion against it. That leverage would have multiplied investors’ returns many times over if the fund’s investments rose in value. Unfortunately, the fund was invested in securities backed by subprime mortgages, and it was obvious that their value was falling.
The fall-out when Bear tried to get out of the mess was a moment of truth for the credit market. It also savagely demonstrated that markets were interconnected by inflicting sweeping losses on investments that had nothing to do with mortgages.
The leverage came from other large Wall Street banks, and Bear wanted them to forego interest payments for a year to relieve the pressure on the fund. This was a lot to ask, and the request was refused. Instead, the banks demanded the securities, mostly packages of subprime debt that had been stuck together and resold known as CDOs, that Bear had pledged to them as collateral for the loans. At their purchase price, the CDOs covered the value of the loans.
But CDOs were not designed to be traded. The idea was to buy and then hold them until all the loans within them had been repaid, and there was barely any secondary market for them. If the creditors wanted their money back, they would have to sell the CDOs in the market. This they tried to do by holding auctions. And so it was that a raft of subprime securities suddenly hit the market on June 20. This revealed a basic truth—nobody was quite sure what the securities were worth, so nobody wanted to buy them. As with dot-com stocks seven years earlier, CDOs had stayed overpriced on paper only because nobody tried to sell them on the market, at least in any volume.
This was credit markets’ “Wile E. Coyote” moment. The character in the old Warner Brothers Roadrunner cartoons would run off the edge of a cliff and keep running in defiance of gravity. Then he would stop and look down. Then, and only then, he would fall, a look of exasperated resignation on his face. In much the same way, markets ran off a cliff on the Shanghai Surprise. The need to maintain the fiction that their subprime investments were worth something kept them running. Rising bond yields made the fiction harder to sustain by forcing up rates and making it harder for borrowers to repay their loans. When the Bear CDOs finally hit the market, the coyote had no choice but to look down. And he could no longer defy gravity.
That meant trouble for banks with similar CDOs on their balance sheet. Cold evidence from the market suggested they were overpriced, but marking down their price meant cutting the very value of the banks’ assets. Given how much banks had borrowed against those CDOs, it might have meant admitting that they were insolvent—which explains why the fictionally high values were maintained for so long.
Bear, the funds’ parent, bailed out a similar, less leveraged fund at a cost of a $1.6 billion, plus a lasting blow to the bank’s credibility. It eventually had to admit that the Enhanced Leverage fund had lost all its value.2 With this, faith in subprime securities was irreparably damaged—and that loss of faith rippled through interconnected world markets to inflict losses in an unlikely place.
The next victims were a group of computer-driven hedge funds whose investments had virtually nothing to do with subprime mortgages. “Market-neutral” or “long/short” hedge funds practice an arcane form of investing. They borrow and then sell stocks in companies that seem overvalued (setting up a profit if their price falls) and balance those bets by putting the money into stocks that they believe to be undervalued.
This is called “market-neutral” because the direction of the market should not matter. If it goes down, the “short” positions will make money, and the “long” positions will lose. If the market rises, it will be the other way around. All that matters is that the managers select their stocks right, and spot which are relatively over- and under-priced. To avoid betting that some sectors will do better than others, many funds pair stocks in the same industry, so a bet against General Motors might be balanced by a bet in favor of Ford.
The subprime mess was bad news for the market as a whole because it implied that consumers would have less money to spend. But falling prices for mortgage-backed bonds should not affect a bet that Ford would outperform General Motors. Such news might cause the price of both stocks to fall, but that would mean making money on the short position in GM to balance the loss on Ford stock. In any case, the funds invested only in publicly traded stocks, not the kind of illiquid, infrequently traded securities that can lead to mispricing and accidents.
But between August 7 and August 9, some of the most famous names on Wall Street took terrible losses investing this way. The Global Equity Opportunities Fund, flagship of Goldman Sachs, lost slightly more than 30 percent of its value that week. Goldman had to join Bear Stearns in putting up money—$2 billion in this case—to help out one of its hedge funds. They were not alone. Many others suffered similar fates in a week when the stock market as a whole was relatively quiet.
These funds turned out to be exposed to mortgages. Piecing together what happened is difficult, but detective work by Andrew Lo, a hedge fund expert at the Massachusetts Institute of Technology, looked convincing—and convinced many of the hedge fund managers themselves who took losses that week.3 On August 6, he believed, one big long-short fund liquidated its positions, buying back the stocks it had shorted and selling the stocks it held.
Why would it have done this? Many hedge funds are ultimately owned by funds of hedge funds, a popular way for investors to get a stake in the market. A fund of funds that had just suffered a loss on its credit investments or could not get money out of a credit fund might well decide instead to ask for their money back from a long-short fund.
Or the fund that started liquidating may have faced a demand from its lenders. That could have been critical. Long-short strategies do not appear risky, but they also do not make much money—unless borrowed money is thrown in. Returns of 1 or 2 percent per year get much more interesting if they are multiplied ten times by leverage. Lenders alarmed by credit losses needed to call in loans where they could, which might mean demanding money back from a long-short fund.
When Professor Lo attempted to map how market-neutral funds would have done that week, he found they would have lost 6.8 percent in three days—far worse than anything before experienced. That they in fact lost as much as 30 percent can be attributed to leverage. Once that leverage was removed, the bubble that it had inflated had to pop. By getting out, the first fund caused its “short” positions to rise, and its “long” positions to fall. That created losses for the many other funds that, guided by the same models, had made the same bets. The losses made it harder for them to meet the demands from their creditors, so they also exited, inflicting even worse losses on those who were still holding on to those positions. The funds had backed their bets with so much borrowed money that they now drove certain stocks’ prices. In the market vernacular, these were “crowded trades.”
How had this happened? In 2000 or 2001, when hedge funds had less money to play with, they could make such bets and make money even as the market was tumbling. But that had attracted far more money into the funds, and there were no longer enough mispriced shares to go around. With so much money doing the same thing, those mispricings tended to be corrected quickly and profits were harder to come by.
This created exactly the dilemma that faced Fidelity’s Magellan fund a decade earlier. They could have responded by closing shop and returning money to investors. But they had an alternative that was not open to a regulated mutual fund, which was to borrow up to the hilt and improve returns that way. This meant pumping unnaturally large sums of money into the same bets. Leverage, a force for keeping the market efficient a few years earlier, now became a vehicle for distorting value through booms and busts.
Many of the investors themselves believed that the crash that befell them was literally impossible. David Viniar, the chief financial officer of Goldman Sachs, went so far as to tell the Financial Times that the market had suffered events so implausible that they were “25 standard deviation moves, several days in a row.”4
In a bell curve, events deviate only from the norm by as much as three standard deviations 1 percent of the time. A 25 standard deviation event is so unusual that it might never happen in the history of the universe. So by saying this, Mr. Viniar was either saying that these events had not in fact happened—or that the statistical models that had produced the estimated standard deviations in the first place were faulty.
While flawed, none of this was illegal. Prosecutors tried to convict the Bear Stearns funds’ managers of fraud and failed. They were found not guilty by a jury in Manhattan in November, 2009.5 Quant funds survived and suffered less than other investors during the disasters of 2008. But the damage had been done. And now, mortgages were about to infect the money market itself.
In Summary
• When creditors of a troubled Bear Stearns hedge fund auctioned its mortgage-backed securities, it became clear that nobody knew what they were worth. That led to withdrawal of cheap leverage from the whole system.
• Cheap leverage allows hedge funds to boost returns from strategies that are minimally profitable and should otherwise be abandoned. Instead, their trades get “overcrowded,” leading to crashes when leverage leaves.
• Quantitative models brought many investors into the same investments, but they did not model the effects that such crowding would have.