"Past performance is no guarantee of future performance." That's a standard warning that mutual funds put in their advertising to attract new shareholders.
The caution has validity in that there is no guarantee that a fund's past good performance will be repeated in the future. But there is growing academic evidence that past performance has on average at least some relevance to future performance.
The latest research, for example, indicates that fund managers from more selective undergraduate universities get a modestly higher return - about 1 percentage point a year - for their shareholders than managers who attended less selective schools.
Choosing a hypothetical example, a manager from Princeton University would be expected to outperform a manager from the University of Florida, note economists Judith Chevalier of the University of Chicago and Glenn Ellison of the Massachusetts Institute of Technology, in Cambridge.
Ms. Chevalier suspects she and her coauthor of a National Bureau of Economic Research paper have found a new hole or anomaly in what academics term the "efficient market hypothesis." This thesis says stock prices correctly reflect all known relevant information. It means that technical analysis of past stock price patterns is useless in predicting future prices. It says investors incorporate new public information into stock prices so rapidly that even fundamental analysis of stocks - examination of real values for a company - is not likely to be fruitful. (Investors with inside information on a corporate stock may have an advantage, though one that is usually illegal to use.)
The efficient market hypothesis is not a favorite among many money managers. They are paid handsomely to bring expert knowledge and research to selecting stocks. But on average, the stock portfolios put together by both mutual fund and pension fund managers do worse over time than comparable averages, such as the Standard & Poor's 500 index, notes Eugene Fama, a University of Chicago economist.
Early in the New Year, many publications (including this one) print special sections on mutual funds that often highlight top-performing managers and include performance tables.
But academics have found that a professional manager who has performed well in one period is just as likely to underperform the market in the next. Superior investment managers may exist, but they are extremely rare. "Probably no other hypothesis in either economics or finance has been more extensively tested," writes Burton Malkiel, a Princeton University economist.
Professor Fama notes that there are roughly 2,000 fund managers in the nation. "Some are going to look good," he says. "But the question is, 'Is that more than chance?'" Laws of chance hold that in a group of money managers as large as this, a few are bound to perform well for a number of years.
Fama does describe Peter Lynch, the famed former manager of Fidelity's Magellan Fund, as "pretty good." But he notes that Mr. Lynch has moved on, and the weak performance of his successors is one proof that such success "can't be passed on by an infusion of some kind to another manager."
In 1994, though, William Goetzmann, now at Yale University, and Roger Ibbotson, president of Ibbotson Associates, a Chicago firm providing various services to financial institutions, looked at the performance of 728 funds from 1976 through 1988. Their industry-pleasing finding was that fund money manager winners on average remain winners. So past returns and relative rankings of funds are useful in predicting returns and rankings.
But a year later, Mark Carhart, of the University of Southern California, Los Angeles, a former student of Fama's, reviewed performance data on all mutual funds since 1962. He found that fund performance does persist, but just for poor performers.
So what mutual fund advice would Fama give a relative? "Go for the low expense, low turnover, low management-fee funds. Stay away from active managers," he says. It is high expense ratios that put many funds in the poor-performance category.
The Chevalier-Ellison paper, which looks at performance of fund managers (not just funds, which change managers) may provide a further investment clue. Managers from schools with high Scholastic Assessment Test scores do better on average at picking stocks, even if adjusted for risk. So do younger managers. But why in either case is unclear. Maybe they are smarter.