Euler Equations and Durable Goods

This paper deals with the estimation of the Euler Equation when durable consumption is taken into account. If durables are not separable in utility from non-durables, estimating the Euler Equation without conditioning on them leads to incorrect inference. I use microdata on non-durable and durable consumption from a US rotating panel, the Consumer Expenditure Survey (CEX). I concentrate on cars (new and used). Apart from housing, they represent the largest share of durable expenditure in the sample. I find an estimate of the intertemporal rate of substitution which is higher than in the case where durable goods are not conditioned on, while the evidence on the excess sensitivity is more mixed.