Model comparison tests of linear factor models in U.K. stock returns

Abstract

This study uses the Bayesian approach of Barillas and Shanken (2018) and the classical approach of Barillas et al. (2018) to conduct model comparison tests of nine linear factor models in U.K. stock returns. The mean-variance efficiency of each factor model is rejected. The Bayesian and classical approaches to model comparison can give different results. Combining the evidence from the two approaches suggests that the six-factor model of Fama and French (2017) with small spread factors provides the best performance among the set of models considered.