I use Maximum likelihood estimation function in Dynare 3.064. I use csminwel algorithm. Dynare produces several figures called “smoothed shocks”. What do these figures tell me? What sort of “smoothed shocks” should I get for good estimation results? Do these shocks have to have a zero mean.
Smooth shocks are a reconstruction of the values of unobserved shocks over the sample, using all the information contained in the sample of observation. It is computed via the Kalman smoother.
Contrarily to OLS regression, there is no built in requirement that the sum of the shocks be zero. However the assumption of the model is that they are zero mean. If you obtain smoothed shocks that are systematically away from zero, there is indeed a problem in your model, either missing constants or a mismatch between the meaning of the variables in your model and in your data.
How to do counterfactual simulations with smoothed shocks?
Suppose my oo_ structure contains oo_.SmoothedShocks.e and oo_.SmoothedShocks.u
suppose I have the model which has created these shocks, e and u. Which command will simulate the endogenous variables? I understood that I have to save the shocks into a file (in columns, in the same order as listed in declarations). How these shocks should be described? As exogenous variables?
Dynare comes with a function “simult_” which can be used to compute simulations. See the header of that function for details usage. You can give any series of shocks you want, either the smoothed shocks (you should get back the estimated data), or any counterfactual.