Generate forecasts after the estimation is done?

Dear all,

Does anyone know if it’s posisble to generate the forecasts of some endogenous variables by setting the parameters at their posterior means, after the estimation is finished? Should I use “estimation” or “stoch_simul” to do this?

I guess this question is somehow related to the question in this post:

However, Professor Pfeifer only mentioned that Dynare will set parameters to the posterior mode after estimation. Would it be possible to use the posterior means of the parameters?


Whether it’s the posterior mean or mode depends on the type of estimation you are doing, see the manual. That being said, why don’t you use the forecast-option of the estimation-command?

Thanks for your reply Professor Pfeifer.

I used the Bayesian method and obtained the posterior means of the parameters.

The problem is that I only forecasted one variable in the estimation command, but now I need to forecast some other variables. However, as the estimation is extremely slow (40+ hours), I was wondering if there is a way that could let me directly generate forecasts without running the estimation a second time.

I would make a backup of the whole folder and then rerun estimation, but with

load_mh_file, mode_file= ..., mode_compute=0,mh_replic=0

Thanks for your suggestion!