Use shock decomposition from Bayesian estimation for counterfactual

Estimation ML/Bayesian estimation
1

Hi! I have a question about how I could use the shock decomposition output from the Bayesian estimation. I have a model with 10 shocks and 10 observables. One of the shocks is a govt spending shock. Here is what I would like to do:

  1. Bayesian estimation of the model to get the shock decomposition.
  2. I would like to know what would have happened if there were no govt shocks during the time period. So, can I just remove the govt shock from the govt spending shock from the shock decomposition that Dynare gives me and call it a counterfactual? I understand that the govt shock has to be quite well identified for it to make sense.

Please let me know whether this is the right way to go about it.

2

Yes, the counterfactual would be one with identical parameters and shock values except for all government spending shocks being 0. Thus, you can use the output of oo_.shock_decomposition to extract the contribution of all other shocks.

1 Like