I wanted to tweak the CEE 2005 model and started by replicating it. I haven’t found a decent code for this elsewhere. The code is attached; I hope some of you find it useful. Note that I include the ‘dyngraph’ command, which calls a program I found elsewhere on this forum that cleans up all the files Dynare creates in the working directory.
CEE 2005 Working Paper Version with Appendix.pdf (687 KB)
CEE20054.mod (7.5 KB)
Did u or anyone try to estimate Gamma (the third group of model parameters) using the method of distance minimization as CEE 2005 did? Any idea on coding? Thanks
I haven’t, actually. But the way I would do it would be to search over a grid, repeatedly calling Dynare.
Hmm, it sounds complicated. I will eventually try it. But for now, my colleagues are still discussing about why we believe that the VAR’s irfs are superior and we have to map our DSGE’s irf to those of VAR. I think it is better than having nothing when you can’t calibrate or use any stylized fact, am I right?
Since you appear to have shifted the timing on capital, shouldn’t:
You appear to be right. I tried changing the code to use lagged capital and get an error. However I find that when I treat capital utilized as if it is not preset, but the capital stock as preset, all runs as it should. It makes sense to me that the capital stock would be fixed in period, whereas utilized capital could be chosen in the period. But I don’t know that this legit.