Estimation of the moments in dynare

[agustindiaz90]

I run the tutorial available in the dynare website (http://www.dynare.org/documentation-and ... t/tutorial) and the standard deviation of the level version (example1) is different from the standar desviation of the log version (example2) but the steady state results are the same. I am working with a model similar to the small open economy of Schmitt-Grohe and Uribe, when I run the models in level or in log, the steady state and correlations results are the same but the standard deviation are different. I want to know if someone know the reason of this and which variance is more appropriate?

Thanks
Agustín

[jpfeifer]

You are looking at different objects here. The steady state refers to the levels of the variables. But the standard deviation refers to the level in the baseline and to the log- level (percentages) in the loglinear version. Due to constant returns to scale the levels in our models usually have no interpretation. Thus, the loglinear version in percentages is usually what we look at

[agustindiaz90]

Dynare works great. Was just me that didnt understand the output.

Regards Agustín