Different reults

Dear Professor Jpfeifer

Recently I am trying to use Dynare to make Bayesian estimation. But the IRF results are different from that of the toy model which I have built without any problems. May I ask is it the matter of the estimation process?


You need to make sure at every step that the results are identical. First, make sure simulation results like IRFs and moments are the same.

Sorry, professor, I didn’t get the point. May I explain my thought again?

Firstly, I have built a toy model, and the IRFs are good, no problems.

While I tried to add Bayesian estimation in it, and the data sets are all stationary. But finally the IRFs are different from that of the toy model.

I wonder whether it is the matter of my Bayesian estimation process.

May I ask do you mean I need to check the Bayesian estimation again?

And check whether the moments in Bayesian estimation process have problems?

TodayI learned DSGE Bayesian estimation for the first time and I am a noob.

Please upload your toy model and your other model for estimation with your data file for more investigation.