A test for the distributional comparison of simulated and historical data

In this note, we apply the testing methodology of Corradi and Swanson [Journal of Econometrics, 2003a], which is based on a distributional generalization of write [Econometrica 68 (2000) 1097] reality check in order to examine the “distributional fit” of alternative real business cycle models. Parameter calibration and estimation error is found to matter much less than distributional assumptions, when some simple illustrative macroeconomic models are considered.