Hedge Fund Activism Skill

Abstract

Hedge fund activism generates persistent performance, but heterogeneity in performance suggests that some hedge fund activists are more skilled than others. We use a Markov Chain Monte Carlo Bayesian estimation algorithm to isolate a time-invariant activist-specific skill component from cumulative abnormal returns. We find considerable differences in this skill component of cumulative abnormal announcement returns of up to 13 - 20 percentage points between the top and bottom skill quintile of hedge fund activists. Out-of-sample tests confirm that our skill estimates are informative about future performance. Differences in skill are also evident in hedge fund activists’ campaign characteristics. The most skilled activists are associated with higher target firm takeover premiums, improved long-term target performance, and more versatile use of campaign tactics.

Keywords: Hedge fund activism, performance persistence, Bayesian estimation, skill

JEL Classification: G13, G23, G34, C11