Determinants of mutual fund underperformance: A Bayesian stochastic frontier approach

The purpose of this paper is to identify ex ante fund statistics that can be related to future performance of European equity funds. In an efficient market setting, actively managed portfolios cannot outperform a passive benchmark strategy. However, purely by chance, some funds outperform their benchmark ex post, making the identification of performance determinants a difficult task. To alleviate this problem, we decompose the return deviation from its expected return into a noise component and an efficiency term, which is 100% if the fund exhibits no underperformance. The decomposition is based on the Bayesian frontier approach. We find evidence that fund efficiency is positively related to fund size and historical performance, the latter being solely due to the poorly performing funds. We fail to find a link between fund age and performance. (C) 2002 Elsevier B.V. All rights reserved.