Intertemporal Substitution, Risk Aversion and the Euler Equation for Consumption

This paper investigates the empirical performance of intertemporal optimization models that relax the restriction imposed by expected utility that risk aversion and intertemporal substitution are negatively related. The authors estimate a system of rates of return and consumption growth equations, and interpret their results in the light of the expected utility, the ordinal certainty equivalence, and the Kreps and Porteus (1978) models. Their results are based on average cohort data for consumption, and thus should not reflect births or deaths. They suggest that non-expected-utility models afford major efficiency gains in the estimation of the elasticity of intertemporal substitution by providing simple cross-equation restrictions.