Theoretical moments with and w/o HP filter

I am running a fairly standard DSGE model using Dynare 4.0.1 with the goal of matching the moments of the model to the data.
I get the following results:
Theoretical moments run without HP filter are the same as empirical moments with HP filter.
Theoretical moments with HP filter are quite different.
Could you explain why this is, and which set of moments should be matched to the data?
Thank you very much!

Are you estimating the model of simulating it?


I am simulating the model using the stoch_simul command.

Please upgrade to the latest Dynare (4.1.2 as of today) since we may have fixed bugs related to the HP filter since the 4.0.1 version.

In particular, if I recall correctly, empirical moments with HP filter were not implemented in Dynare 4.0.1 (so it may explain your strange results).

Thank you, S├ębastien!