Many of them do so for short periods, of course, but a number of academic studies have generally found that such market-beating funds rarely stay at the top of the rankings for long. For the most part, researchers have concluded that the best-performing hedge funds of one period were rarely the best performers over any significant stretch that followed. That's a dead giveaway that a hedge fund that happens to beat the market over a given time owes its performance more to luck than to genuine investment ability.

But a new study has reached the opposite conclusion. It has found that many hedge funds can outperform their benchmarks consistently. The study, issued by the National Bureau of Economic Research, a nonprofit and nonpartisan research organization based in Cambridge, Mass., has been accepted for presentation this month at the annual meeting of the Western Finance Association, a scholarly organization. It was written by Ravi Jagannathan, a finance professor at Northwestern University; Alexey Malakhov, an assistant professor of finance at the University of Arkansas; and Dmitry Novikov, an associate in the office of equity derivatives strategy at Goldman Sachs in New York. A copy is at nber.org/papers/w12015.

The researchers argue that most previous studies of hedge fund performance were flawed because they failed to correct fully for statistical problems in databases of hedge fund returns. Perhaps the most significant problem, according to Professor Jagannathan, is the one caused by the disappearance of hedge funds from those databases — a problem that he and his co-authors call the "self-selection bias."

Funds vanish from the databases for two primary reasons. Some funds that are particularly poor performers close down and liquidate. And others, particularly good performers, close their doors to new investors and stop reporting their performance to the databases.

Correcting for this problem, of course, is not easy. But by analyzing such funds' performance until they disappeared from the performance databases, the researchers could make educated guesses about what those funds' returns would have been had they not vanished. In general, they found that the funds that closed to new investors because of good performance were more likely to be above-average performers in the period after they closed.

This could explain why so many studies have failed to find evidence of performance persistence. By eliminating from consideration those funds that disappear from databases, such studies may have overlooked funds that provided the strongest such evidence.

The researchers in the new study, after making educated guesses about the returns of disappearing funds, found that many hedge funds were surprisingly persistent in their performance from one period to the next. For every 100 basis points by which a hedge fund beat its benchmark over a given three-year period, the researchers found, it outperformed that benchmark by 57 basis points, on average, over the next three years. (A basis point is one-hundredth of a percentage point.) The researchers estimated a similar degree of underperformance for the worst performers.

The performance advantage lasts far longer for a hedge fund than it does for a mutual fund, Professor Jagannathan said in an interview. On average, he said, a mutual fund tends to stay a top performer for 12 months or less; often, it then becomes a market laggard. In fact, performance persistence among mutual funds is so modest that some researchers say it can be explained by factors having nothing to do with genuine investment ability.

Professor Jagannathan says the results of the new study provide compelling evidence, however, that many hedge fund managers have such skill. That stands to reason, he said, because hedge fund managers typically have the freedom to take longer-term risks, while mutual fund managers have incentives to focus primarily on short-term performance. Undoubtedly, the generous management fees paid by hedge funds also lure many highly skilled managers to the hedge fund world.

To be sure, hedge funds are not for everyone. For starters, one typically has to invest a large amount, often several hundred thousand dollars, and the money is tied up in the fund for a long period. But the new study offers a clear lesson: when money managers are immune from the risk of investor flight after one poor month or quarter, they can often flourish.

Mark Hulbert is editor of The Hulbert Financial Digest, a service of MarketWatch. E-mail: strategy@nytimes.com.