Millions of people around the world invest in actively managed mutual funds, believing fund managers can find enough hot stocks to supercharge results. But others scoff at the idea, pointing to studies showing the average managed fund trails its benchmark index over time.

While some managers do exceptionally well year after year, it is not clear whether the reason is skill or luck. With enough people flipping coins, some will flip a string of heads, the index advocates say. They prefer passive management, in which a fund simply buys the stocks in its benchmark index and holds them for the long term, minimizing the costs and fees that chew into returns at managed funds.

Most studies of active versus passive managing have focused on long-term results, looking at how a class of managed funds performs relative to its benchmark over periods of one to ten years. A managed fund containing stocks in large companies, for example, is typically compared to the Standard & Poor’s 500 index, which consists of the 500 largest U.S. stocks.

But this approach fails to consider the element of risk, says Jessica A. Wachter, a finance professor at Wharton. “The question is, how do you properly risk-adjust?” she asks. Since a managed fund’s holdings are not the same as those in its benchmark index, the two portfolios’ level of risk is not the same. If the managed fund were to hold more small-company stocks than the index, for instance, it would be expected to beat the index when those risky small stocks did exceptionally well, and to trail when they did poorly. Comparing the two funds’ results over any given period does not reveal whether the successful active manager really had an eye for good stocks or simply stumbled upon a group that happened to do well over that period.

This dilemma is known as the “joint-hypothesis problem,” since there is more than one possible explanation for the research results. “What we’re trying to do is get around that,” says Wachter, one of the co-authors of the recent study, Can Mutual Fund Managers Pick Stocks? Evidence from Their Trades Prior to Earnings Announcements. Her colleagues in the project were Malcolm Baker of the Harvard Business School and Lubomir Litov and Jeffrey Wurgler of the NYU Stern School of Business.

Instead of looking at long-term results, the researchers focused on the three days surrounding companies’ quarterly earnings announcements, and they looked only at the stocks that active fund managers bought and sold prior to the announcement. Hence, the research zeroes in on managers’ decisions just before companies reveal information that can drive stock prices up or down. The data involved 6.3 million observations from the second quarter of 1980 through the third quarter of 2002.

The results showed that fund managers do have some ability to determine which stocks will do well, and which will fare poorly. “Consistent with skilled trading, we find that, on average, stocks that funds buy earn significantly higher returns at subsequent earnings announcements than stocks that they sell,” the researchers write. Still, the amount of extra return related to the stock-picking ability identified “is not that big,” Wachter says.

Results around the quarterly earnings announcements were multiplied by four to gauge the effect of this stock-picking ability over an entire year. Looking at the net effect of adding to some holdings prior to earnings announcements and trimming others, the study found an annualized gain of 47 basis points, or 0.47%, due to manager’s decisions.

Similarly, the study looked at stocks that were bought for the first time prior to announcements, along with those that were eliminated from the portfolios before announcements. This found an annualized gain of 58 basis points. In other words, managers were able to identify stocks likely to do well or poorly.

The study looked at stock performance over just a few days around announcements — 12 trading days per year — while managers buy and sell throughout the year. Hence, the effect of manager skill may be greater than the figures suggest, though this study did not measure that effect, Wachter notes. Managed funds that did exceptionally well tended to have a growth-oriented rather than value-oriented style of stock picking, and they tended to be large funds with high portfolio turnover. They also used incentive fees to motivate managers.

The study did not examine the reasons for these results, according to Wachter. She explains, though, that incentives may indeed motivate managers to do better, and that some funds may grow large because their managers have done well. As for the long-running debate about whether managed funds serve investors better or worse than indexers, it remains unresolved.

Active managers may be able to spot good stocks. But because they must support teams of stock pickers and pay higher costs for active trading, managed funds charge fees that are typically more than five times those of indexers. What’s gained through stock picking may be lost to costs. “Are managed funds earning back their fees?” Wachter asks. “That, we can’t answer.”

Can mutual fund managers pick stocks? Evidence from their trades prior to earnings announcements