Market-wide Events and Time Fixed Effects

Abstract

Accounting for time-varying unobserved heterogeneity poses a fundamental challenge for empirical finance research. We implement and show the relevance of grouped fixed effects (GFE) models in finance settings. GFE models are a general, flexible, robust, and parsimonious class of fixed effects models, which capture a richer set of time-varying unobserved heterogeneities. We extend GFE models to jointly account for time-varying unobserved heterogeneity and simultaneity bias building a two-stage least squares estimator, propose a new Hausman-type specification test to select among fixed effects models, study the finite sample properties of GFE models through simulations, and empirically demonstrate their economic importance.