Return Contributions of Factors and Characteristics in an Integrated Asset Pricing Approach

Abstract

We extend the standard one-period expected utility model based on exponential preferences with additional preference variables for asset characteristics, which are able to capture cross-sectional return differences. Contrary to the literature we treat characteristics as indicator variables to indicate whether a stock exhibits a certain characteristic or not. These indicator variables can be interpreted as coloured markers of stock characteristics. We derive simple empirically testable relationships, from which we dissect portfolio excess returns into return contributions coming from the covariance with common factors (market excess return, HML, SMB) and characteristics. Especially we find the robust results that the characteristics distress and high volatility carry significant and large absolute return contributions. Moreover, the characteristic return contributions exhibit strong return seasonalities and changes over the sample period.