On Wall Street, there is a rage against the machine.
Hedge funds with computer-driven or quantitative investment strategies have been recording significant losses this month.
The managers of these funds are the products of the trading desks of the big investment banks, like Goldman Sachs and Morgan Stanley, both of which have investment operations that use computer models.
The cross-fertilization has raised fears among some analysts that it is not only the hedge funds that are being hit, but the trading desks at the banks as well.
“These guys all know each other, and they all have the same strategies,” said Ernest P. Chan, a quantitative trading consultant who has done computer-driven research at Morgan Stanley and Credit Suisse. “They came from the same schools, and they get together for drinks after work.”
As the quantitative system has come to underpin the investment approaches of some of the largest hedge funds, its use has grown sharply.
Moreover, bankers and investors say, the strategies employed tend to be not only duplicable but broadly followed — the result being a packlike tendency that has helped increase market volatility and, for some hedge funds, has led to losses in the last month.
Wild swings in stock prices have become the norm as fears about the mortgage securities market have expanded into the broader markets. Last week, the Dow Jones industrial average was sharply higher on Monday and Wednesday, only to drop 387 points on Thursday, eventually ending the week about where it began.
A common thread has often been a rise or fall in prices late in the day, a pattern that many analysts attribute to computer models, which are driving a much larger volume of the trading.
Mr. Chan said this predilection for lemming-style buying or selling from investors using similar computer models could turn what would normally be a market setback into a wider contagion.
“If all the models say buy, who is going to say sell? There is just not enough money on the other side,” he said.
The problems of these quantitative funds mirror those of the hedge fund industry as a whole — many funds have seen sharp declines in the last couple of months as the credit markets have dried up. Some quantitative funds could potentially have their worst year on record.
Despite the large sums of money involved, ranging from $250 billion to $500 billion, according to industry estimates, the club of quantitative investors is a small, exclusive one that bridges the trading desks of investment banks and some of the country’s largest hedge funds.
One might call it six degrees of quantitative investing.
Clifford S. Asness, who has a Ph.D. in finance from the University of Chicago, is the founder of AQR Capital Management, a quantitative hedge fund that, according to investors, has had a 13 percent loss so far this month.
Mr. Asness is also a founder of Goldman Sachs’s troubled Global Alpha fund, which controls about $9 billion. The Alpha fund has suffered an 11 percent reversal this month, giving it a decline for the year that is approaching 30 percent, sparking speculation that Goldman would liquidate the fund. Goldman calls the speculation “categorically untrue.”
On a smaller scale, Tykhe Capital, another hedge fund that uses quantitative techniques, was down 19 percent in August. The founders of Tykhe are from D.E. Shaw & Company, the giant hedge fund that manages $35 billion via a broad reliance on quantitative, as well as other, strategies and whose founder, David E. Shaw, who has a Ph.D. from Stanford, originally came from Morgan Stanley.
Hedge funds as a whole have grown exponentially and now manage about $1.7 trillion, more than double the amount five years ago.
In one respect the swoon of these computer-reliant funds is the result of managers, who are faced with a deluge of investor money seeking accelerated returns, using their models to make higher risk market bets by following day-to-day trends. It is an approach that seems to run contrary to the original philosophy underlying a quantitative approach, called statistical arbitrage.
Narrowly defined, statistical arbitrage involves a fairly straightforward investment strategy, like the rapid-fire buying of one stock and the selling short of another so as to use the computer’s speed to identify and make money from even the most minute price discrepancies. Such a strategy will generally provide liquidity to the market by buying stocks on the way down and selling them short on the way up. In so doing, it provides a dose of calming, computerized sang-froid to markets in the grip of panic or euphoria.
But such strategies rarely promise high returns, so quantitative investors have broadened their computer models to include strategies for investing in more risky areas like mortgage-backed securities, derivatives and commodities.
“You can build a computer model for anything that is tradable,” Mr. Chan said. To some extent, that explains the outbreak of losses in these funds.
With many of these new assets being highly illiquid and with the funds themselves having used considerable amounts of borrowed money to enhance their returns, losses have been magnified as worried investors have demanded to pull their money out.
In a letter to investors this week, James H. Simons, the founder of Renaissance Technologies, the most highly regarded of the quantitative funds, gave voice to what he described as “unusual” market conditions. Mr. Simons, who received a Ph.D. in mathematics from the University of California, Berkeley, acknowledged what a difficult month August had been, with his RIEF down close to 9 percent for the month.
For an investor who reportedly earned $1.6 billion last year and whose flagship Medallion fund had an average annual return of over 30 percent since 1988, it was a surprising reversal.
“We cannot predict the duration of the current environment,” Mr. Simons wrote. “But usually such behavior causes first pain and then opportunity. Our basic plan is to stay the course.”
