Empirical evidence on the Euler equation for investment in the US

Research highlights

• Is the typical specification of the Euler equation for investment employed in DSGE models consistent with aggregate macro data? YES, is the answer using state-of-the- art econometric methods that are robust to weak instruments and exploit information in possible structural changes.
• This is a positive message that we add to the literature, as the first paper that checks the consistency of the implied Euler equation for investment with aggregate data as a single equation, rather than through the lens of Bayesian estimation of a full DSGE model.
• How much can we learn from aggregate time series data about the values of the three key investment Euler equation parameters? Unfortunately, the answer to this question is NOT MUCH.
• Those parameters are generally very poorly identified, that is, there is very little information about these parameters in aggregate data, when using single-equation limited-information methods and available instruments.
• This finding contrasts with the results commonly reported in the DSGE literature. Our analysis investigates how DSGE models achieve identification of the key parameters of the investment equation and finds that this comes either from the prior, or from the tight cross-equation restrictions implied by the theoretical structure of a standard medium-scale DSGE model, or from the cross-correlations and joint dynamic behavior of the time-series used in the Bayesian estimation.