Countercyclical Stockholders' Consumption Risk and Tests of Conditional CCAPM

Abstract

Heterogeneous-agents asset pricing theories imply that stockholders' consumption has the first-order effect on equity premium. Motivated by these theories, we evaluate the performance of the conditional CCAPM in explaining time-variation in market returns and cross-sectional variation in portfolio returns. At the market level, we show that the conditional stockholders' consumption risk has strong predictive power for market returns with 39% in-sample and 19% out-of-sample R-squared for the three-year horizon, outperforming a broad set of alternative predictors. At the portfolio-level, stockholders' consumption risk explains 40% of the cross-sectional average returns. Stockholders' consumption risk also partially explains the value, size, profitability, investment, and long-term reversal premia. We provide an explanation for why stockholders' consumption risk reverses the findings in the literature using aggregate consumption risk: stockholders' consumption risk varies in the opposite direction to aggregate consumption risk, but in the same direction with the equity premium and value premium. This article also demonstrates that time-variation in both the price and amount of risk should be considered in testing the CCAPM.