An Efficient Fixed-Effects Estimator for Corporate Finance
73 Pages Posted: 22 May 2018 Last revised: 13 Apr 2020
Date Written: April 10, 2020
The use of panel data in corporate finance is ubiquitous to estimate the impact of managers' and/or shareholders’ choices on firm value. We evaluate the properties of four existing and widely used estimators (pooled OLS, random-effects, first-difference, and fixed-effects), and find them to be lacking consistency, efficiency, or both. Consequently, we introduce the new consistent efficient fixed-effects (EFE) estimator. When examining the relationship between CEO performance-pay sensitivity and firm value, we find the EFE estimator to be most appropriate. All estimators are presented as GMM estimators, allowing us to straightforwardly design and conduct tests for model misspecification, including endogeneity.
Suggested Citation: Suggested Citation