SVAR and forecast (model)

[fmramos]

Hi all!

I'm trying to estimate a SVAR like Blanchard&Quah. The impulse-response was ok (not so good, but ok). The problem is that i'm trying to solve by model and estimate something like a filtered unemployment (eg. NAIRU). I made the model, but only with the unrestricted VAR. Is it possible to solve for SVAR? Or I have to use something different like Rats or Matlab?

Tks in advance, Fabio

[EViews Gareth]

I believe you cannot make a model from an SVAR, and thus cannot forecast from one :(

[ecshukri]

Hi!

I'm trying to build an SVAR model for a small open economy, ala Kim and Roubini (2000). So far I have succeeded in imposing the short-run restrictions. Could someone please share his/her codes for the long-run restrictions (Blanchard and Quah or other authors) with me? Suffice a simple one will do and it does not need to follow the preceding paper. I just need it as a counter check with my structure and codes.

Thanks
Shukri
Eviews 7.1, build date April 2010.

[EViews Gareth]

I am by no means an SVAR expert, but I think the SVAR dialog should give you an example of a long run restriction (if you scroll down). Something like:

Code: Select all

@LR2(@u1) = 0