Luck Bias and Skill Performance of Individual Mutual Funds

Abstract

This is the first study measuring luck bias and skill performance of individual funds. We show that luck and bad luck balance within the cross-section, such that average alpha is virtually unbiased. Many individual fund alphas, however, are significantly biased by luck resulting in significant rank changes after our correction. In performance persistence tests, we show that our skill alpha predicts performance and discriminates better between future outperformers and underperformers. As a result, hypothetical skill-High-minus-Low portfolios outperform standard-High-minus-Low portfolios by up to 0.9% p. a. over holding periods up to 4 years. This \skill premium is also statistically significant. Moreover, our results confirm very recent developments in mutual fund research regarding time-variation in management skill, individual manager skill and the relation between skill and management activity. This is important not only for academics but also for investors as our methodology is easy to implement and based on readily available information.