Please Help!

Hi Guys,

I have a quick question. If my model (the files are attached below) is not linearized around the steady state in the mod file, then for the estimation of the parameters, do I have to use the HP filtered data or not?

Thanks a lot in advance.

Best,
paramJN.m (1.61 KB)
djk_shrunk_steadystate.m (3.62 KB)
djk_shrunk.mod (2.72 KB)

The question is wrong. Dynare will always linearize your model around the steady state if you don’t do it. Your model always needs an observation equation. Moreover, you should not use the regular HP-filter as it is a two-sided filter and thus not causal. Use either first differences as in Smets/Wouters or use a one-sided HP-filter.