The RATS Software Forum

Dear Tom,

Thanks a lot for the answer.

You suggested that I should change identification from zero restriction to sign restriction in previous posts. I have seen two replication examples of sign restrictions: one is Uhlig (2005), and the second is the Farrant-Peersman. Which one is suitable in my case? Which one do I need to follow?

Best regards,
Elyor

There's also Mountford and Uhlig. And there have been many other papers using the technique. You need to decide which shocks you actually need the model to determine and which you don't. The initial Uhlig paper did just a single shock, Mountford and Uhlig did 3, Farrant and Peersman did a complete 3 or 4 depending upon which VAR they did. And then think about how you would describe that with sign constraints. Read these and some other papers to get some idea about how you map the economics to the sign constraints.

Dear Tom,

I understood. Thank you very much.

Best regards,
Elyor

I understood. Thank you very much!