Hedge Fund Performance under Misspecified Models

Abstract

We develop a new approach for evaluating the entire alpha distribution across hedge funds. Our methodology allows for performance comparisons between models that are misspecified – a common feature given the numerous factors that drive hedge fund returns. Empirically, we find that the standard models used in previous work are subject to misspecification as they perform like the CAPM and produce large and positive hedge funds alphas. We then construct a new model based on alternative trading strategies (including variance, carry, and time-series momentum) and document a large and statistically significant decrease in performance. Whereas this new model remains misspecified, it identifies several key factors that drive hedge fund returns.

Keywords: Hedge funds, performance, model misspecification, large panel

JEL Classification: G11, G12, C14, C33, C58