Can a computer pick stocks better than an active manager?
By now you've probably heard that a majority of mutual fund managers fail to beat broad market indexes. But if most managers have a poor stock-picking record, can computers succeed where they fail? Computers now can beat grandmasters at chess, so shouldn't a manager be able to harness their incomparable powers of calculation to outperform even the best human stock-pickers? Fund managers who employ so-called quantitative strategies would answer "yes" to these questions. These managers use sophisticated computer models to pick stocks and bonds that the models predict will have market-beating returns.
Quant funds have a lot of intuitive appeal. First, they strip away at least some of the human bias that trips up active managers, who often buy into market trends at precisely the wrong time and overlook real values. Also, quant models allow a manager to sift through thousands of securities and pick out ones that have the characteristics, or factors, he thinks are predictive of high future returns. What takes a model minutes to compute would take a team of analysts weeks--time for market conditions to change and discovered opportunities to disappear. Moreover, managers who have faith in their quant models are unlikely to deviate from their strategies, thereby avoiding the inconsistency that hurts many active managers.
Before proceeding, we should note that the line between quant funds and traditional actively managed offerings is blurry. Many, if not most, nonquant funds use computer models to narrow down their investment universes, so that analysts can focus their research on a manageable number of stocks. Quant funds are not immune to human error. After all, people build the quant models, and, after they're built, the models are rarely static. Most successful quant managers are always looking to maintain their edge versus the market by seeking out new factors to replace ones that have lost their predictive power. For that reason, quant funds can, and often do, have many of the faults of their human creators.
Still, a quant fund can be a good addition to a portfolio. (Consider placing it in a tax-sheltered account, though, as these funds often trade a lot.) And in selecting one, all the usual rules apply for choosing a fund. Investors should look for ones that follow disciplined strategies, have successful track records in a variety of market conditions, and charge low fees. A low expense ratio is particularly important, as quant funds tend to resemble their benchmarks, holding a sizable number of stocks and keeping sector weightings close to those of the benchmark. That controls risk, but it also limits gains, so low fees are essential if the fund is to edge past its benchmark.
Here are descriptions of three of our favorite quant funds:
Vanguard U.S. Value
Bridgeway Large-Cap Growth
Janus Adviser INTECH Risk-Managed Growth
Many of our other favorite quant funds are closed to new investors. Although we excluded them from our list for that reason, they still deserve honorable mentions. These are the aforementioned Bridgeway Aggressive Investors 1