Bayesian approaches to the estimation of DSGE models are becoming increasingly popular. Prior knowledge is normally formalized either be information concerning deep parameters’ values (‘microprior’) or some macroeconomic indicator, e.g. moments of observable variables (‘macroprior’). In this paper we introduce a non parametric prior which is elicited from impulse response functions. Results show that using either a microprior or a macroprior can lead to different posterior estimates. We probe into the details of our result, showing that model misspecification is to blame for that.
Bayesian estimation DSGE Models impulse response function Prior Distribution