International Factor Models
76 Pages Posted: 11 Feb 2020 Last revised: 5 May 2020
Date Written: February 6, 2020
We compare prominent factor models from Behavioral finance (Daniel, Hirshleifer, and Sun, 2020; Stambaugh and Yuan, 2017) and Neoclassical finance (Sharpe, 1964 and Lintner, 1965; Fama and French, 1993, 2015, 2018; Hou et al., 2020; Hou, Xue, and Zhang, 2015) on the four regions North America, Europe, Asia Pacific and Japan based on the maximum squared Sharpe ratio. Relying on the pairwise comparison tests introduced by Barillas et al. (2019), we find that the Hou et al. (2020) five-factor model dominates all other models besides the Daniel, Hirshleifer, and Sun (2020) three-factor model in North America. In Europe, the Daniel, Hirshleifer, and Sun (2020) model, the Stambaugh and Yuan (2017) model and the Fama and French (2018) model share the first rank. In Asia Pacific, the Stambaugh and Yuan (2017) model dominates all other models besides the Daniel, Hirshleifer, and Sun (2020) model. In Japan, the Stambaugh and Yuan (2017) model has the highest in-sample maximum squared Sharpe ratio, but only the difference to the CAPM is significant. These results are supported by means of multiple model comparison tests and a bootstrap simulation.
Keywords: Asset pricing, Factor models, International stock markets
JEL Classification: G12, G15
Suggested Citation: Suggested Citation