But does it work? Can you assume that a fund that did well in the past will continue to outperform its peers going forward? Not according to a 2003 study by Morningstar. Funds that ranked in the top quartile in their category based on one-year, three-year, and five-year trailing returns were no more likely to end up there again over the ensuing periods. Except for the one-year time period (one-year trailing returns, one-year ensuing returns), investors had a better chance of choosing a top-performing fund by flipping a coin than by picking one off the top-quartile list. "The general conclusion was that past performance was not generally predictive," says Christine Benz, editor of Morningstar Mutual Funds, the firm's flagship publication. "And somewhat surprisingly, the longer the time period the less predictive it was."

A dozen studies highlighted in "Hot Hands, Cold Hands: Does Past Performance Predict Future Returns?" in the May 2006 issue of the Journal of Financial Planning came up with mixed results. In some cases performance persisted, in other cases it did not. Results largely depended on the time period being studied. Although past performance can be somewhat useful in deciding which funds to avoid -- funds that had sub-par performance one year were more likely to do poorly again the next year -- the author concluded that past performance "should not be the sole or even most important criterion for selecting a mutual fund."

Expenses and manager tenure as predictors of future performance

If mutual fund investors were to isolate one predictive data point, it might be fund costs, says Benz. "If an expense ratio was low relative to its peers five years ago, the odds are superb that it will also be cheap today."

But do low expenses in the past predict anything other than low expenses in the future? What do they say about future performance? It may be natural to assume that funds with low expenses will do better than funds with high expenses because there's less drag on performance. But if that were true choosing mutual funds would be a slam dunk -- just screen for the funds with the lowest expense ratios and be done with it. Vanguard would win every time. But we all know that low-expense funds are not always among the top performers. We also know of funds that have high expense ratios and still deliver superior performance to shareholders. For example, the top performer on Business Week's Mutual Fund Scoreboard (Wasatch Micro Cap, WMICX) has an expense ratio of 2.15% and still managed to give shareholders a 21.5% annualized compound return over ten years. Who wouldn't take that over the 10.2% returned by the Vanguard Emerging Markets Index (VEIEX) even with its low 0.42% expense ratio?

If past performance is unlikely to persist, and if fund expenses (though likely to persist) can't identify future winners, what factors can help you choose good mutual funds for your clients? What about manager tenure? It's long been assumed that the longer a manager has been with a fund the more likely the fund is to perform well going forward. But Benz said no. A study she ran found no correlation between manager tenure and performance, sending that particular data point out the window, along with past performance and expenses as predictors of future performance.

What's going on here? If past performance, expense ratio, and manager tenure can't help you pick a fund, what can? I believe the problem lies with the attempt to isolate a single variable as being significant in fund selection, when it is really the combination of factors and their interplay that exerts the strongest influence. Also, studies that measure funds in the aggregate lose sight of what is going on within each individual fund. One fund's inconsistency cancels out another fund's consistency, creating a result that is statistically insignificant across all funds while ignoring the superiority of the exceptional funds.

Take a look at the most popular actively managed equity funds in the Advisor Perspectives universe -- the Dodge & Cox International Fund (DODFX), the American Europacific Growth Fund (AEGBX), and the Julius Baer International Equity Fund (BJBIX). These are all outstanding funds, ranking in the top 5%, 16%, and 1% in their categories respectively and blowing well past their benchmarks over the past five years. Advisors must be doing something right to have chosen these funds for a significant portion of their clients' assets. It probably wasn't performance, expenses, manager tenure, or any other single factor that led to the choice but rather a combination of factors that led to the conclusion that these funds have a strong likelihood of doing well in the future. Whatever it was, it worked.

While advisors would surely appreciate having some independent study to support their fund selections, there seems to be a dearth of statistically significant studies isolating variables having persistency across time periods, probably due to the nature of the studies themselves. Instead, advisors may have to fall back on such qualitative resources as their own experience, judgment, common sense, and intuition when evaluating individual funds and assessing their suitability for individual clients.

Incidentally, Morningstar's star system does seem to have some persistence. "If there's a fund that was a five-star fund four years ago (when Morningstar revised its rating system), there's a very good chance it will continue to be a five-star fund," says Benz, "meaning it has had a superior risk-adjusted return." Whether you use Morningstar's rating system or some other method to identify top-performing funds going forward, look at a confluence of factors and apply your own judgment based on fund characteristics, your outlook for the economy and the markets, and your understanding of your client's objectives. Numbers can be comforting, but sometimes fuzzy math leads to better decision-making.

Elaine Floyd, CFP® is a financial writer based in Bellingham, Washington.

Display article as PDF for printing.

Would you like to send this article to a friend?


Remember, if you have a question or comment, send it to .