Good Luck, Bad Luck. Can Mutual Funds Really Pick Stocks?

Abstract

The literature examining mutual fund performance is vast and researchers have studied the performance of professional money managers extensively, but there is still an ongoing debate whether or not mutual fund managers have stock picking skills. This article investigates the profitability of institutional trades using a unique dataset comprising the transactions executed by a sample of mutual funds during the years 2002-2007. We implement an innovative bootstrap technique to examine the non-normal distribution of our characteristics-based performance measure across the trades of mutual funds to precisely separate luck from skill. Moreover, we apply a new nonlinear technique based on cluster analysis to construct matching benchmark portfolios. Results indicate that, controlling for luck, a negligible minority of managers pick stocks well enough to beat the market. The underperformance of managers can mainly be attributed to ‘bad luck’, and not to the absence of genuine stock picking skills. Finally, funds trading in illiquid stocks with higher book-to-market ratio and smaller size perform better.

Keywords: Mutual funds, Transaction data, Performance evaluation, Bootstrap, Stock picking skills

JEL Classification: G11, G17, G23