Are Mutual Fund Managers Good Gamblers?

67 Pages Posted: 17 Feb 2017 Last revised: 27 Apr 2018

Roberto Stein

University of Nebraska at Lincoln - Department of Finance

Date Written: April 18, 2018


I investigate the skill of mutual fund managers by focusing on their investments in 'Lottery Stocks'. Kumar (2009) characterizes these as stocks with low price, high idiosyncratic risk and high idiosyncratic skewness, and shows that these securities on average underperform. I find that Lottery Stocks held by mutual fund managers strongly outperform the market, by about 6% per year, while Lottery Stocks which fund managers avoid underperform others, by close to 12%. A hedged portfolio which buys Lottery Stocks held by mutual funds, and sells those they ignore, averages an annual return of 22%, with an alpha of similar magnitude. I introduce the 'Lottery Score', the percentage of fund equity assets invested in these stocks. I find that Lottery Funds, those that invest in Lottery Stocks, outperform their Non-Lottery competitors by 1.2% per year. For these funds, the strategy of loading on Lottery Stocks is persistent in time, as is the benefit derived from it. I show that the Lottery Score is a good predictor of fund performance, even after controlling for a number of other measures of skill. Finally, I show that the cross-sectional variation in the Lottery Score can be explained by the intensity of buy-side competition, as funds that have fewer competitors can extract higher returns from their private information. The evidence is consistent with fund managers possessing superior stock-picking skills.

Keywords: Mutual funds, portfolio holdings, fund performance, fund manager skill

JEL Classification: G11, G23

Suggested Citation

Stein, Roberto, Are Mutual Fund Managers Good Gamblers? (April 18, 2018). Available at SSRN: or

Roberto Stein (Contact Author)

University of Nebraska at Lincoln - Department of Finance ( email )

Lincoln, NE 68588-0490
United States

Register to save articles to
your library


Paper statistics

Abstract Views