On Fri, 26 Aug 2022, ajgalindo@gmail.com wrote:

Dear all. Its there a way to estimate bayesian VARs with sign restrictions in GRETL? Thanks in advance for any help. Best regards, Arturo.

Yes. It's in the SVAR addon. See secion 8 of the pdf doc.

Let me add to that that currently the VAR itself (the reduced form, so basically the autoregressive coefficients) cannot be estimated with actual shrinkage in the SVAR addon. (No real Bayesian prior imposed on those coefficients there.) That doesn't have to be a problem, I just want to make it transparent.

Some tools to estimate BVARs (in the reduced-form sense, not structural) are in the works by some members of the gretl team, but my guess is that it won't happen before the winter (in the Northern Hemisphere).

cheers

sven