Mixing long-run and short-run restrictions in SVAR

[cath_xx]

Hi

I am trying to estimate a SVAR like the one in Smets and Tsatsaronis (1997) and Gali (1992). The identification scheme involves a mixture of long-run and short-run restrictions. The EViews Manual says I cannot mix them but then I found this post http://forums.eviews.com/viewtopic.php?f=4&t=970 which confused me.

So is there ANY way to do it in EViews? If not, does anyone know any other package that would do it? Or do I have to resort to MATLAB (I guess?) or something else?

I use EViews 7 btw.

Many thanks!!

[rihara]

I have the same problem.
ANYONE?

[EViews Gareth]

As stated in the manual, EViews doesn't support mixing long-run and short-run restrictions.

The information in that post is meant to suggest that updating to the latest version will fix the "Internal error 500" problem. It will not allow you to mix short run and long run restrictions, though.

[donihue]

FYI, RATS can estimate an SVAR with both SR and LR restrictions
Regards
Donihue