How can I compute with Dynare the autocovariance functions of a model (with bands) to compare them with the ones in the data? Thank you.
I suggest that you perform Monte Carlo simulations and use the same filtering method you use when computing the data moments. In the case you have stationarized the model, you should “re-trend” the simulated data before filtering (according to my limited experience this heavily influence the results).
Antti