“A NEW study has found that as many as 20 percent of investors may be able to regularly pick stocks that beat the market,” writes Mark Hulbert in Sunday’s New York Times.

“Most previous studies have emphasized the inability of the average investor to outperform stocks in general, though some researchers have conceded that some people can do so over a given period — sometimes, for many years. Still, researchers have generally assumed that the percentage of investors who can do this consistently is very small.”

“The new study, however, has found that assumption to be wrong. It was conducted by three finance professors — Joshua D. Coval of the Harvard Business School, David A. Hirshleifer of Ohio State University and Tyler G. Shumway of the University of Michigan — and has been circulating since January as an academic working paper (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=364000).”

“The professors reached their conclusion after examining a database containing the complete trading histories of more than 110,000 individual accounts at a national discount brokerage firm. (The professors would not disclose the name of the firm; they did say that the names of the account holders were not included in the database.) The database covered trading from the beginning of 1990 through November 1996.”

“But don’t some investors beat the market through luck alone? To deal with that possibility, the professors conducted a series of statistical tests. They were less interested in investors whose cumulative performance beat the market, because chance can play a large role over such a period. Instead, they looked for those whose stock picks repeatedly surpassed the market average.”

“The study focused on nearly 17,000 accounts in which investors had bought at least 25 stocks — an amount large enough to give the researchers confidence that any outperformance was not a result of mere chance. The findings were also adjusted for risk, to ensure that investors had not beaten the market simply by taking on greater risk in a period when stocks generally were rising. After considering these factors, the professors found that about one-fifth of these frequent traders had genuine stock-picking ability. In that group, the average stock gained 44 percent a year, annualized, over this period, versus 14.5 percent for the Wilshire 5000.”

“The professors saw little evidence of stock-picking ability among most of the other investors. Though their average stock slightly beat the market, the professors found that the margin of outperformance was not enough to pay commissions and bid-asked spreads. In other words, a large majority of investors would still be better off investing in an index fund.”