I have seen that many members were looking for a way to group large number of shocks into a small categories to make their historical decomposition and forecast error variance decomposition graphs sensible ( including me)
I think I found a way. The attached file has the codes that helps to deal with this problem. I am very new to Matlab and Dynare. The code is very simple but does the job. But, the graphs should be produced in excel after copying relevant data. All details are given in the file.This is tailor made. So, if someone wants to use this code they have to adjust it to suit to their work. The excel template plots the graphs. This is only for historical decomposition. But, this can be easily extended for forecast error variance decomposition.
Any comment or correction is welcome. I still haven’t used in my paper.
template_shock_decomposition.m (3.22 KB)