Detrend

Hi all,

After having read a lot of posts about matching data with with the model I have 4 questions that I would like to confirm.

  1. If we have our model in log deviation from steady state form and the observed variable is in levels we should detrend the observed variable.
    Can somebody confirm this?

  2. Second, we should also take logs of the observed variable to match it with the model.
    Can somebody confirm this as well.

  3. Also I would like to know which is the preferred way to demean the data- the prefilter option in estimation or HP-filter.

  4. How does dynare prefilter the data.

Thanks,
Abhishek.