We confront prominent asset pricing models with the classical out-of-sample cross-sectional test of Fama and MacBeth (1973). For all models, we uncover three main findings: (i) the intercept coefficients are economically large and highly statistically significant; (ii) the cross-sectional factor...