The Next Buffett?
Bad news for your fund manager: Computers are starting to look like pretty savvy investors.
By Stephen Gandel

(MONEY Magazine) – It's no surprise that George Sauter, the chief investment officer of index fund king Vanguard, argues that most investors are wasting their time and their money when they try to beat the market. So why is Sauter launching funds that will attempt to do precisely that? Because he believes that while man can't outperform, a machine can. "Computers," Sauter says, "can systematically take advantage of all the mistakes we humans make."

Increasingly, Sauter isn't alone in this belief. Charles Schwab now runs nine quantitative, or "quant," funds in which computers make buy and sell decisions based on their constant crunching of hundreds of thousands of numbers.

Of the five Schwab funds that have three-year records, four have whipped at least 75% of their peers. "This strategy has a lot of teeth," says Jeff Mortimer, Schwab's equity investment chief. Vanguard's first computer-run fund, Strategic Equity, has risen 27% on average for the past three years, putting it near the top of the hot midcap category. In April the firm launched a small-cap fund, and Sauter thinks there's room for more. Janus Capital now offers six quant funds. AllianceBernstein and American Century have jumped on board as well.

The popularity of quants stands in sharp contrast to that of the late '90s, when investors lost interest in the funds because they couldn't keep pace in the Internet bubble. In the seesaw market that has followed, however, quant funds have done well. In the past five years, they have returned an average of 6.1% a year, according to fund research firm Lipper, vs. 4.3% for all actively managed U.S. stock funds.

To be sure, that's too short a period to declare the contest between man and machine over. But supporters like Sauter say that computer-managed funds can beat the market with less risk than human-run funds. A recent study by Goldman Sachs Asset Management concluded that while over 15 years a human manager can beat a quant fund in absolute terms, once you adjust for the extra risk the human manager had to take to get that result, the quant fund comes out ahead. "Quant funds are a well-established strategy for institutional investors," says Coral Gables, Fla. financial planner Harold Evensky. "They are becoming a story for individual investors, and for most people they probably make good sense."

Just the Numbers, Ma'am

Both man and machine evaluate investment prospects by comparing financial statements and share prices, looking for stocks that are cheap relative to assets, earnings or some other quantifiable measure. Most human fund managers then visit the management of those companies, try out their products or do whatever it takes so they can make an educated guess about how a business will do in the future. Computers, on the other hand, buy and sell strictly by the criteria they have been programmed to watch. Those criteria range widely. Vanguard compares stocks with others in their industry and buys those with the best growth prospects and lowest valuations. Schwab's computers, which rank 18,000 stocks, focus on those that are increasing earnings faster than expected and are rising in price. Janus, on the other hand, cares more about stock movement. Its computers buy shares of companies that go up and down regularly, and try to make money from that volatility.

So how do quant funds compete without doing hands-on research? For starters, computers can follow far more stocks than can a manager who is out visiting executives. The average quant fund tracks thousands of stocks and owns hundreds. The resulting diversification in their portfolios makes them safer investors than their human rivals, says Joseph Candela, a New York City planner with Ameriprise Financial who is recommending quant funds to clients.

But the real advantage machines have over people is that they're not human. Emotions trip up professional money managers like they do the rest of us. Pros hold on to plummeting stocks, not wanting to admit a mistake. Or they rush in when a stock takes off, often too late. "Computers don't have bad days or biases," says Richard Thaler, a University of Chicago economics professor and principal of Fuller & Thaler, a money-management firm that tries to exploit human error in the market. "A computer won't like a company just because it went to the same college as the CEO." Nor will it latch on to a 1999-style mania.

Finally, unlike human managers, computers don't earn six-figure salaries or require legions of analysts to support them. Vanguard's Sauter says he can run his funds with a third of the people needed by the typical mutual fund. So computer-managed funds should be able to charge lower fees than other funds.

Bugs in the Program

The thing is, many don't. According to Lipper, computer-managed funds charge investors 1.3% of the portfolio in fees a year on average, only a hair less than actively managed funds. And all those data give quant funds itchy trading fingers, driving costs up and returns down.

As the funds get bigger, costs should fall. But the second hurdle they have in trying to outperform is a tougher one: Quant funds have trouble anticipating change. Ron Muhlenkamp, whose eponymous fund is in the MONEY 65, recalls that in the 1970s quant funds failed to understand that the ramp-up in inflation would drive the market. Similarly, even before the Internet bubble, quant funds lagged because they didn't foresee that technology was the next big thing. "What you get from a computer is discipline," he says. "What you don't get is judgment."

Muhlenkamp has a point. The best human investors have a much better chance of significantly outperforming the market than computers do. On the other hand, human managers are also much more likely to lose you a lot of money.

So keep an eye on quant funds. The ones in the box on page 68 have modest expense ratios, and they could help you squeeze a tiny bit more return out of your portfolio without taking on big risk. That's not a Buffett-size gain, but they're not making any more of him. Computers, on the other hand, get better with each new model.

QUANTS BY THE NUMBERS

These computer-managed funds have posted strong returns and don't overcharge for performance.

NOTES: Returns are as of March 28. Three-year returns are annualized. SOURCE: Lipper.

invest

Insider selling is way up. How bad is that? 76

Fund Watch by Penelope Wang 78

Sivy on Stocks by Michael Sivy 82

Market indexes are shown in real time, except for the DJIA, which is delayed by two minutes. All times are ET. Disclaimer LIBOR Warning: Neither BBA Enterprises Limited, nor the BBA LIBOR Contributor Banks, nor Reuters, can be held liable for any irregularity or inaccuracy of BBA LIBOR. Disclaimer. Morningstar: © 2014 Morningstar, Inc. All Rights Reserved. Disclaimer The Dow Jones IndexesSM are proprietary to and distributed by Dow Jones & Company, Inc. and have been licensed for use. All content of the Dow Jones IndexesSM © 2014 is proprietary to Dow Jones & Company, Inc. Chicago Mercantile Association. The market data is the property of Chicago Mercantile Exchange Inc. and its licensors. All rights reserved. FactSet Research Systems Inc. 2014. All rights reserved. Most stock quote data provided by BATS.
Market indexes are shown in real time, except for the DJIA, which is delayed by two minutes. All times are ET. Disclaimer LIBOR Warning: Neither BBA Enterprises Limited, nor the BBA LIBOR Contributor Banks, nor Reuters, can be held liable for any irregularity or inaccuracy of BBA LIBOR. Disclaimer. Morningstar: © 2014 Morningstar, Inc. All Rights Reserved. Disclaimer The Dow Jones IndexesSM are proprietary to and distributed by Dow Jones & Company, Inc. and have been licensed for use. All content of the Dow Jones IndexesSM © 2014 is proprietary to Dow Jones & Company, Inc. Chicago Mercantile Association. The market data is the property of Chicago Mercantile Exchange Inc. and its licensors. All rights reserved. FactSet Research Systems Inc. 2014. All rights reserved. Most stock quote data provided by BATS.