Return Dispersion and the Cross-Section of Stock Returns
86 Pages Posted: 29 Jun 2018 Last revised: 23 Jun 2019
Date Written: June 19, 2019
This paper shows that return dispersion (RD) is a significant market risk factor that helps to explain the cross-section of stock returns. We propose that stocks can have positive and negative sensitivity to RD movements over time. To capture asymmetric RD effects, we augment the CAPM market factor with RD in a novel two-component probabilistic mixture model. Each component of the mixture model is a two-factor regression model that captures both the market effect and the RD effect with a specific sign (positive or negative). We employ an expectation-maximization (EM) algorithm to estimate the parameters of the mixture model. Our so-called ZCAPM takes into account beta risk associated with the market factor and zeta risk related to the return dispersion factor. Out-of-sample cross-sectional tests of U.S. stock portfolios in the period 1965 to 2015 yield highly significant estimates for the market price of zeta risk. Our results for zeta risk dominate those for popular multi-factors, which are less significant than RD across different test assets and sample periods. Moreover, estimates of the market premium for zeta risk are economically substantial. These and other empirical tests suggest that zeta risk associated with return dispersion in our ZCAPM is a salient asset pricing factor.
Keywords: Asset Pricing, Cross-sectional Stock Returns, EM Regression, Return Dispersion
JEL Classification: G12, C20
Suggested Citation: Suggested Citation