Overview
Duke University, Fuqua School of Business. Empirical corporate finance with a focus on Bayesian methods, long-horizon performance, IPOs, share repurchases, and equity issuances.
Key Contributions / Features
- Bayesian predictive inference for long-horizon event studies (Brav 2000, Journal of Finance): SUR framework with lognormal shrinkage priors on residual SDs and equicorrelation structure; predictive distribution of sample mean correctly captures both non-normality and cross-sectional dependence; dominates the Ikenberry-Lakonishok-Vermaelen bootstrap by ~30% in tail coverage; shows FF3F rejected for 1975–1984 IPOs while characteristic-based model holds.
- IPO and SEO long-run underperformance: Brav-Gompers (1997) and Brav-Geczy-Gompers (2000) on whether IPO and SEO abnormal returns are anomalous once proper benchmarks are used.
Related