Idiosyncratic Consumption Risk and the Cross-Section of Asset Returns

This paper investigates the importance of idiosyncratic consumption risk for the cross- sectional variation in asset returns. We find that besides the rate of aggregate con- sumption growth, the cross-sectional variance of consumption growth is also a priced factor. This suggests that consumers are not fully insured against idiosyncratic con- sumption risk, and that asset returns reflect their attempts to reduce their exposure to this risk. The resulting two-factor consumption-based asset pricing model signifi- cantly outperforms the CAPM, and its performance compares favorably with that of the Fama-French three-factor model. THE VALIDITY OF THE WORKHORSE CONSUMPTION-BASED representative-agent model rests upon a number of assumptions, including the presence of markets that allow agents to perfectly insure themselves against idiosyncratic consumption risk. The only relevant pricing factor in the resulting standard multiperiod asset pricing model without frictions is the growth rate of aggregate consump- tion. The consumption-based asset pricing model (CCAPM) that reflects this allocation, however, is not able to match important aspects of the distribution of historical asset returns, such as the risk premium on the market portfolio. Its performance in a cross-sectional context is weak; as a result this model has not succeeded in replacing alternative cross-sectional models such as the capital asset pricing model (CAPM) as the standard in the profession.1 However, factors other than aggregate consumption growth become rele- vant to price assets if the assumptions underlying the standard representa- tive agent model do not hold. For example, if agents cannot perfectly insure