Economic Evaluation of Asset Pricing Models Under Predictability

This paper performs an out-of-sample comparison of linear factor asset pricing models from an economic perspective under predictability. I assess the economic value added of several factor models when a Bayesian investor is faced with a portfolio allocation problem whereby each model imposes cross-sectional restrictions on the parameters of a predictive stock return regression. The empirical framework explicitly accounts for investor skepticism about the model, i.e., mispricing uncertainty. Using several US portfolios as test assets, I find that the q5 model of Hou et al. (2020), as well as the behavioral factor models of Stambaugh and Yuan (2017) and Daniel et al. (2020) outperform competing models across investment horizons. At the longest evaluated horizon (one year), a benchmark portfolio built using historical data produces larger portfolio gains than all the factor models, but in the short run (at the one-month horizon), their performance is comparable.