Example of VECM with BQ restrictions and Error Bands

[macro_man]

All,

Are there any examples of a VECM that uses Blanchard-Quah long-run restrictions and calculates error bands for the impulse response functions? Thanks.

[TomDoan]

KPSW4.RPF (from King, Plosser, Stock and Watson AER 1991) does Monte Carlo integration on a VECM with long-run restrictions. BQ itself isn't that interesting since if you have a two-variable VECM, if one variable has a long-run 0 restriction, the other has to as well.

[macro_man]

Thanks. Can you direct me to a resource that explains why this is so. My models have 6-8 variables, but the issue I assume still applies.

[TomDoan]

There are a couple of forum threads which discuss this:

http://www.estima.com/forum/viewtopic.php?f=4&t=730
http://www.estima.com/forum/viewtopic.php?f=4&t=263

[macro_man]

Tom,

I followed the ShortandLong example with VECM. Is there an example for the shortandlong where standard error bands are generated? Thanks.

[TomDoan]

Yes. Bjornland-Leitemo http://www.estima.com/forum/viewtopic.php?f=8&t=1118

[carljames]

I was actually having a hard time to determine the who permanete shock originates from, but it seems that you can't, huh?!
Honestly, it's a little confusing

[TomDoan]

I was actually having a hard time to determine the who permanete shock originates from, but it seems that you can't, huh?!
Honestly, it's a little confusing

That's correct. The BQ restriction actually identifies the transitory shock---the permanent shock is whatever is orthogonal to that. You would need a more structural model than a simple VECM to identify the source of the permanent shock.