I want to use the moments (variance) of variables from a DSGE model to compare different policies. My doubt is if could I use the differences respect to steady state?, or alternatively, first differences?
The guiding principle is to treat your empirical data and the model-generated data the same. You can e.g. use first differences or the HP-filter to compare moments of stationary variables (here the two-sidedness does not matter)
1 Like
Thanks a lot dear Jpfeifer.