Should the Holding Period Matter for the Intertemporal Consumption-Basedcapm?

Abstract

Empirical studies of the restrictions implied by the intertemporal capital asset pricing model across different asset markets have found conflicting evidence. In general, restrictions from this model have been rejected over short holding periods, but not over longer holding periods such as a quarter. This paper asks whether an auxiliary assumption implicit in these tests could be responsible for the observed pattern of rejections. The auxiliary assumption requires that covariances of returns with consumption move in constant proportion over time. The paper first describes how this condition may break down within the context of a general equilibrium pricing relationship. Then, the condition is tested empirically using data on foreign exchange, bonds, and equity returns. Interestingly, the pattern of consumption covariances in foreign exchange and bonds indeed match the pattern of rejection in the intertemporal asset pricing relationship.