Fundamental Risk Sources and Pricing Factors

Neoclassical theory suggests that stocks exposed to common pricing factors must face common production risks. We estimate firm-level productivity shocks and decompose them into six aggregate risk components via asymptotic principal component analysis. We find that fundamental risks drive 13 of 15 prevailing pricing factors. First, we show that the fundamental shocks capture most factors proposed in the Fama-French sixfactor model (Fama and French, 2018), the q-factor model (Hou et al., 2015), except the expected investment growth factor in the q5 model (Hou et al., 2018). Second, we find that fundamental shocks explain most mispricing and behavioral factors (Stambaugh and Yuan, 2017; Daniel et al., 2018), except the post-earnings-announcement-drift factor. Third, we identify an important fundamental risk, the first principal component of productivity shocks, is missed in all of these empirical factor models. We interpret this missing factor as the labor risk. Overall, the productivity-based model performs at least as well as the prevailing factor models.