Page 1 of 1

VECM with flat impulse responses

Posted: Tue Jun 16, 2009 1:48 pm
by Billaudes
Hi everyone,

I am currently trying to run a VECM with data on total bank loans, securities, deposits, and interest rates (in that order). However, my impulse responses look really peculiar, the first three variables show almost no reaction to interest rates. Has anyone experienced this before, is there maybe a common mistake that I am making somewhere.

I am aware that it is going to be hard to help me with that little information, please let me know if you guys have an idea or need more details about my estimation.

Thanks a lot.

Billaudes

Re: VECM with flat impulse responses

Posted: Sun Jun 21, 2009 3:46 am
by peter_h
Hey, I think using impulse response function for the VECM is of no use, as equations entering VECM should be non-stationary, thus having an infinite impulse response. You should use IRF for stationary VAR instead.

Re: VECM with flat impulse responses

Posted: Wed Jul 01, 2009 1:55 pm
by Billaudes
Actually all my variables are non- stationary, or trend stationary. Actually including them in a VAR with first differences, still doesn't give lot of reaction on the IRF's.

Re: VECM with flat impulse responses

Posted: Fri Nov 20, 2015 8:18 pm
by Econodelics
Hi Billaudes , its 2015 and I have the same problem running a VECM (which is a VAR with first differences). Id like to know if you solved your problem and how.
Thank you so much for your response!
@Econodelics