Page 1 of 1

GIRF for STR models

Posted: Sun Sep 03, 2017 2:42 pm
by Estefaniame
Good evening

I would like to know if there is a RATS code for doing impulse-response functions with Smooth Transition (nonlinear) models? I found a close answer in a post from 2009 but I would like to know if there has been anything new from that moment on.

Thank you beforehand.

Re: GIRF for STR models

Posted: Sun Sep 03, 2017 3:06 pm
by TomDoan

Re: GIRF for STR models

Posted: Fri Mar 30, 2018 12:54 pm
by Estefaniame
Good evening
I tried to run the GIRF procedure stated in:

See viewtopic.php?f=30&t=2650&p=12277

But I am afraid it only works with STAR models, not with STR ones, is it correct? Otherwise, I am not able to see where the error is.

Thank you beforehand.

Re: GIRF for STR models

Posted: Fri Mar 30, 2018 2:28 pm
by TomDoan
A STAR model is a self-contained model, therefore you can track the response to shocks. A "STR" model isn't. What does an impulse response even mean in that case?