We investigate the role of extreme positive payoffs in the distribution of monthly fund returns in investors’ mutual fund preferences. We document a positive and significant relationship between the maximum style-adjusted monthly return (MAX) and future fund flows. The relationship is robust to controlling for average performance, volatility, skewness, and various other fund characteristics. Our findings are consistent with the notion that fund investors overweight the probability of high payoff states in the past return distribution. We further show that MAX is not a useful predictor of future performance and that an increase in a fund’s visibility does not explain our findings.

We examine the performance of mutual funds whose managers simultaneously manage portfolios with performance-based incentive fees for three account types: mutual funds, hedge funds, and separate accounts. Importantly, our data set is free of selection bias because it is hand-collected from mandatory SEC filings. We find that only funds whose managers also manage hedge funds significantly underperform peer mutual funds. Moreover, underperformance begins only after fund managers begin to manage a hedge fund. We find that managerial incentives and opportunities for cross-subsidization explain variation in underperformance across funds, supporting the conflicts of interest hypothesis in the debate on “side-by-side management.”