I am estimating a model with Financial frictions. The problem is that i get some strange results. For example the Historical Variance decomposition.
What I do is to take logs, detrend the variables (with the Matlab command detrend) and use the differences in the measurement equation.
My original data for y (not in differences) and the data generated by the model for this variable y (the observable is dy) has the same shape but differently scaled. Is there a problem with this? Can my estimations be wrong? Am I doing something wrong with the data?
Moreover, the estimated Calvo parameters for wage and price are really low in comparison with other similar models, 0.34 and 0.25. The risk aversion coefficient of the CRRA utility function is close to 0 (0.05).
Thank you very much in advance!