I know how to do it in simulation, but I don’t know how to do it in estimation.
Currently, I put no restrictions on the correlation matrix on the shocks. The outcome shows strong correlation between shocks. Is there some way in Dynare that we can limit the correlation between shocks when doing estimation ?
I’m sorry, maybe I am not getting your point right. But if you want to limit the correlation between shocks, i.e., making it small, why don’t you try using strict priors on the correlation coefficient? You can define a normal distribution with mean zero and a really small standard deviation.
Hope this can help.