We estimate a nonlinear VAR model to study the real effects of monetary policy shocks in regimes characterized by high vs. low macroeconomic uncertainty. We find unexpected monetary policy moves to exert a substantially milder impact in presence of high uncertainty. We then exploit the set of impulse responses coming from the nonlinear VAR framework to estimate a medium-scale new-Keynesian DSGE model with a minimum-distance approach. The DSGE model is shown to be able to replicate the VAR evidence in both regimes thanks to different estimates of some crucial structural parameters. In particular, we identify a steeper new-Keynesian Phillips curve as the key factor behind the DSGE model’s ability to replicate the milder macroeconomic responses to a monetary policy shock estimated with our VAR in presence of high uncertainty. A version of the model featuring firm-specific capital is shown to be associated to estimates of the price frequency which are in line with some recent evidence based on micro data.
monetary policy shocks uncertainty Threshold VAR medium scale DSGE framework minimum-distance estimation