Dear all,
I’m trying to estimate a linear model using Bayesian estimation in Dynare, and I’m having some problems with the shock decomposition exercise. According to my understanding, the shock_decomposition command does historical shock decomposition to the smoothed variables X_{t|T}, what if I want to do shock decomposition with the updated variables X_{t|t}? Can Dynare help me do this? Thank you very much!
@rattoma Do you have functionality like this? If not, it would be straightforward to start at the steady state and simulate the model with one filtered shock at a time.