Asset pricing models with and without consumption data: An empirical evaluation

Abstract

This paper evaluates the ability of the empirical model of asset pricing of Campbell (1993a,b) to explain the time-series and cross-sectional variation of expected returns of portfolios of stocks. In Campbell's model, an alternative risk-return relationship is derived by substituting consumption out of the linearized first-order condition of the representative agent. We compare this methodology to models that use actual consumption data, such as the model of Epstein and Zin, 1989, 1991, and the standard consumption-based CAPM. Although we find that Campbell's model fits the data slightly better than models which explicitly price consumption risk, and provides reasonable estimates of the representative agent's preference parameters, the parameter restrictions of the Campbell model, as well as its over-identifying orthogonality conditions, are generally rejected. The parameter restrictions of the Campbell model, and the over-identifying conditions, are marginally not rejected when the empirical model is augmented to account for the "size effect".

Keywords: G12

JEL Classification: Consumption, Asset Pricing Models, Size Effect