Only a few years ago, using a computer program to pick stocks was the ultimate in investing sophistication. Any portfolio manager who used a stock screen to, say, identify large-cap companies with low price-earnings ratios could justifiably claim to be a "quant" -- shorthand for investors who use quantitative research and trading strategies.
Now that most professional investors use a stock screen to narrow their choices, being a true quant means a lot more. It requires using computer models based on complex algorithms to implement arcane trading strategies that typically require a PhD to understand.
And it may mean taking humans out of the stock-buying decision altogether, as John Montgomery, founder of Houston-based Bridgeway Funds, has done with his mutual-fund company. "There are a dozen different ways emotion gets in the way and systematically causes people to make the wrong decision at just the wrong time," says Montgomery, who calls his company "a pure quant shop."
"SOMETHING OF A FAD."
Quantitative strategies have proliferated recently, thanks to the fact that they were about the only thing that worked during the three-year bear market that ended last March. "If you're on an uphill growth path, similar to the heyday of the dot-com era, fundamental analysis always wins out," says Sang Lee, manager of the securities and investments group at Celent Communications, a Boston-based financial services technology research firm. "Quant shops do better when markets are flat or unstable."
Today, more than 300 investment firms use quantitative research or trading techniques (up from fewer than 100 some 15 years ago), estimates Lee in a September report that reviews developments in quantitative research and trading. Those firms have more than $4 trillion in assets under management, or about 17% of the U.S. market. Lee expects quantitative research and trading activity to grow an additional 40% in the next three to four years, even though his report calls this trend "something of a fad."
The quant world breaks down roughly into two camps for which quantitative trading has both strengths and drawbacks. In one camp are the portfolio managers who use computer models to pick stocks and other investments to buy. These models are mostly used by mutual-fund families like Bridgway, ICON, or Hennessy, and their models range widely in complexity and success.
The biggest problem with quant strategies in managing mutual funds is that the best ones don't last long: Models that work in one market environment tend to fail in the next. To run a quant fund "requires constant reconsideration of the model," says Jeffrey Ptak, an analyst at fund research firm Morningstar.
The most successful quant mutual funds today are those that have several models working in conjunction with each other, known as "multivariant" or "multifactor" funds. The idea is that they can switch quickly between sectors and investment styles to take advantage of a shifting market.
Montgomery uses five models in Bridgeway Aggressive Investor (BRAGX ), which has a stunning 27% five-year average annual return. The different models (one finds stocks with strong momentum, another identifies stocks that are cheap) are designed to offset each other, which should reduce volatility in the fund. "If each model is doing its part," says Montgomery, "when there's one style in favor that model should be kicking in and something else lagging."
The other main quant camp is in the hedge-fund world, where selling stocks short (a way to bet that their price will fall), piling on complex derivatives, and borrowing lots of money (so you can put more cash behind each bet and therefore magnify returns) are permitted. (Federal regulations prohibit ordinary mutual funds from engaging in such risky strategies, but hedge funds, which only wealthy people have enough money to participate in, are almost entirely unregulated.)
"Quant methods is where a lot of the sophistication in investing is going," says Salomon Konig, president of GP Funds, which distributes hedge funds made up of other hedge funds (known as "funds of funds").
Quantitative hedge funds typically use esoteric trading strategies, often employing derivatives, options, and futures to achieve returns that are entirely divorced from stock market returns. Many of the strategies are designed to take advantage of tiny inefficiencies in the market. A hedge-fund manager might design an arbitrage play to take advantage of a slight, temporary difference in price between a stock index and the underlying securities that make up the index.
Some of these hedge funds have been wildly successful, including those run by Jim Simons of Rennaissance Technologies, who manages the Medallian Fund, and Steve Cohan, of SAC Capital Advisors (see BW Cover Story, 7/21/03, "The Most Powerful Trader on Wall Street You've Never Heard Of"). But when the use of leverage to magnify returns goes wrong, as it did in the famed case of Long Term Capital Management in 1998, the losses can be disastrous.
Another foible of quantitative hedge funds is that once traders discover an inefficiency in the market to exploit, their own trading activity eventually eliminates the opportunity. "Many of the market imperfections can only be revealed by computers and quantitative methods," explains Konig. "Once some of those imperfections are taken out of the market, you need more quantitative methods and more automated tools to get at arbitrage opportunities."
Despite their complexity and apparent sophistication, most quantitative strategies probably don't add much value for investors. Nassim Nicholas Taleb is an experienced options trader, hedge-fund manager, math professor -- and author of the book Fooled by Randomness. He believes that all successful investors, including Warren Buffett, are mainly just lucky.
Nonetheless, a case can be made that taking emotion out of the investing process can help boost returns and reduce risk. "I can't tell you how big a deal I think that is," says Montgomery. Of course, if you produced returns like his, you might feel the same way.
By Amey Stone in New York