The Lucas critique revisited assessing the stability of empirical Euler equations for investment

Lucas (1976) argued that the parameters of traditional macroeconometric models depended crucially on agentsʹ expectations and were unlikely to remain stable in a changing economic environment. In response, econometric modeling has focused on the estimation of rational expectations models that have an explicit structural interpretation — Euler equations in particular. Thus, a natural, though little acknowledged, criterion for judging the success of empirical Euler equations is the stability of their ‘deep’, structural parameters. Examining split-sample tests over multiple breakpoints as well as the sequence of subsample model estimates, we find considerable instability in the estimated parameters of a standard Euler equation of investment.