A Dynamic Test of Conditional Asset Pricing Models
42 Pages Posted: 1 Mar 2014 Last revised: 23 Nov 2016
Date Written: March 22, 2015
I use Bayesian tools to develop a dynamic testing methodology for conditional factor pricing models, in which time-varying betas, idiosyncratic risks, and factors risk premia are jointly estimated in a single step. Based on this framework, I test over fifty years of post-war monthly data some of the most common factor pricing models on size, book-to-market, and momentum deciles portfolios, both in the time series and in the cross section. The empirical results show that, a conditional specification of the recent five-factor model of Fama and French (2015) outperforms a set of theory-based competing linear pricing models along several dimensions.
Keywords: Factor Pricing Models, Savage-Dickey Density Ratio, Stochastic Betas, Bayesian Econometrics
JEL Classification: G12, E44, C11
Suggested Citation: Suggested Citation