Hedge Fund Activist Skill

Abstract

Using a hierarchical structural model estimated via Bayesian MCMC, we decompose hedge fund activism announcement returns into treatment, stock-picking, and selection effects, identifying a latent activist-specific skill component. We find substantial skill heterogeneity operating primarily through treatment: skill raises returns far more when activists intervene than when they hold passively, and predicts out-of-sample returns only for active campaigns. In mergers, skill does not affect whether targets attract merger attempts but raises the premium conditional on a deal through two channels: treatment skill via bargaining in negotiations, stock-picking skill via target selection. Skill also predicts improved operating and risk-adjusted performance.

Keywords: hedge fund activism, activist skill, Bayesian MCMC, structural estimation, mergers and acquisitions, merger process

JEL Classification: G23, G34, C11, D44