I just did an estimation of a mod file in which I wanted to obtain the variance decomposition. I then included moments_varendo as an option in the estimation command. Two issues are associated with what I observe instead. I notice that for all variables for which I need variance decomposition, the same values are reported. But I don't believe this is true. Also the manual reference does not conform with the structure of the results reported in the fields to oo_.PosteriorTheoreticalMoments.
Am I missing something here? Can someone please explain what I am not paying attention to?