Hi ,everyone,I have no idea how to shut off certain shocks in estimation command.
If I only want to simulate series of endogenous variables using smoother, within only some shocks, which are among the whole set of shocks that have been used to estimation. And during that procedure, the parameters that has been estimated should be the same. In this way the relative importance of certain shocks could be judged.
So what should I do?
Could it be that you are looking for the shock_decomposition command?
Yeah, that is what you have done in replication of the paper ,Jermann, Urban and Vincenzo Quadrini. 2012. "Macroeconomic Effects of Financial Shocks."
as is listed below.
shock_decomposition(parameter_set=posterior_mode) y_obs;
timeline=1984:0.25:2010.25;
y_pos=strmatch(‘y_obs’,M_.endo_names,‘exact’);
xi_pos=strmatch(‘eps_xi’,M_.exo_names,‘exact’);
However, I am confused about maybe there is another potential method to work that out. First do the estimation command and get the estimated parameters, and then use the calib_smoother to get the smoothed variables, with only certain shocks. As in the case of Jermann, Urban and Vincenzo Quadrini. 2012, first do estimation and do calib_smoother with the given parameters but with only the financial shock.
I do not know whether this method is reasonable and whether it will give the same result as by shock_decompositon.
And if possible , could you give me an example paper on the usage of calib_smoother?
Thanks a lot.
The smoother only gives you the best estimates of the states and the shocks. It does not simulate the counterfactual series if only one of these shocks had happened. That is what shock_decomposition does. It first runs the smoother and then computes the respective counterfactual series if there were no other shocks except for one.
An example for the calib_smoother is at github.com/DynareTeam/dynare/blob/master/tests/fs2000/fs2000_calib.mod