Hi, this is my first message here in the forum, and I advance I’m completely new on dynare (and dubiously experienced in matlab).

I came with a question, which I’m not sure it’s already replied… anyway, 35 pages are too much, so I find it faster posting it here.

If I wanted to estimate the parameters of a , let’s say, NK model, having the both expressions non linearized and linearized, and I got too the varobs y (a time serie of GDP)… should I get that serie trough a filter (as the HP filter) and use the cyclical part as proxie for the deviations from the S.S.?, or should I just use the raw serie and then insert the non linearized model?

I hope the question is clear enough… in any case, thanks in advance!

You have to map your stationary model to the data, which describes deviation from a steady state or a BGP. This can either be done by pseudo-“stationarizing”/detreding your data by HP-filtering them. This is often the easiest way to obtain deviations from steady state.

Alternatively, you can map your stationary model to the raw data by means of an observation equation as increasingly common. For this procedure, see the following posts

[Simple model. Unkown error!)

[Measurement equations and steady state growth)

I’m sincerely grateful for the quick response, thanks.

I think I’ll use the cyclical component arising from detrending with the HP, it’s the easiest way, and suits fine with my purposes (not that high)

So, to work again!

It went fine, thanks. I eventually solved all the problems that arose trying to estimate my first simple NK model (with the estimation options, that one with the nondetrended data)… gosh, I got problems even with the datafile format! now that the basis is working is time to deepen within my thesis…