Using conditional forecasts to keep monetary policy response the same across specifications

Stochastic simulations
1

I want to compare the IRFs to a TFP shock under two different model specifications. Unsurprisingly, the endogenous response of monetary policy (Taylor Rule) is different in both cases. I would like to do a comparison where the path of monetary policy is the same in both specifications.

I run the first dynare mod and save the R path:

r_a_path = oo_.irfs.R_etfp + R_ss
save rpath_model_a.mat r_a_path

In the second specification dynare mod file, I have the following code at the end:

shocks;
var etfp; stderr 1;

end;

stoch_simul(order=1,irf=40, nograph);

conditional_forecast_paths;

var R;
periods 1:40;
values (rpath);

end;
conditional_forecast(parameter_set = calibration, controlled_varexo = (emon));

When comparing the unconditioned and conditioned forecasts, it looks like I have just executed a monetary policy path AND not the tfp shock. Is there a way to still run the tfp shock while keeping MP path the same as in model specification 1? It seems I need to specify the shock somewhere?

2

What exactly are you trying to do? The conditional forecast path does not use a particular shock but rather creates a shock series for the monetary policy shock that yields a particular interest rate response. If you now put in a TFP shock, the endogenous feedback will change and hence the exogenous shock path for monetary policy. Interpreting this may be challenging.
To answer your question, what you can do is to add the TFP process to the conditional_forecast_paths and etfp to the controlled_varexo . That should give you desired paths for monetary policy and TFP.