Page 9 of 9

Re: Threshold Structural VAR

Posted: Mon Sep 17, 2018 7:52 am
by dakila
Sorry, no.

Re: Threshold Structural VAR

Posted: Sun Apr 28, 2019 10:16 am
by Shwan
Hi, is there any recent developed code to obtain confidence intervals for the thsvar impulse responses?
thanks for the add in, but its totally impossible to convince a reviewer to have positive comments about a paper that generates impulse responses without the confidence intervals, its like reporting regression results without inferences! How can we know its statistically significant!

Re: Threshold Structural VAR

Posted: Sun Apr 28, 2019 8:45 pm
by dakila
Unfortunately, no.

Re: Threshold Structural VAR

Posted: Fri Jul 26, 2019 11:39 am
by icamara
How can I see or display the GIRFs graph after estimating the thsvar. I checked in the workfile but couldn't see it.

Best

Re: Threshold Structural VAR

Posted: Wed Oct 30, 2019 6:20 am
by danielamartins97
Dear all,
Since the EViews Add-in regarding the generation of the GIRFs does not include the confidence bands, I would like to know if there is a programming code available to compute the GIRFs as well as the confidence bands. Nevertheless, if you know a code or other way to compute it using R or Matlab, please share.
Thank you so much.

Re: Threshold Structural VAR

Posted: Mon Mar 09, 2020 2:49 am
by refp
Dear community,

While using the THSVAR in Eviews 11, I receive the following error ‘Error 33 in encrypted program’ and it cannot produce the impulse responses

The model I am trying to run contains the following endogenous variables: us_ltrate au_log_ngdp au_log_cpi au_strate au_log_neer
time period: 1993Q1 - 2019Q4
lags: 2
threshold variable: au_buscyc_ngdp

default for all the other settings...what am I doing wrong?

Many thanks in advance for your guidance! :P

Re: Threshold Structural VAR

Posted: Wed Mar 11, 2020 9:00 pm
by dakila
you forget response variable. Also your threshold variable should be endogenous.

Re: Threshold Structural VAR

Posted: Tue Apr 07, 2020 6:00 am
by refp
Thank you very much for your guidance Dakila, and sorry for my late response!

The add-in works fine now…but only for some countries in my workfile. For instance, when I run the thsvar with the following specification I get ‘Error 15’

-endogenous variables: au_growth_ngdp au_growth_cpi au_rate au_growth_neer au_buscyc
-time period: 1974m8-2019m9
-lags: 2
-threshold variable: au_buscyc
-response variable: au_growth_neer
-generalized impulse responses and default for all the other settings

…the very same 'Error 15' appears while using the log-version in the same time period and with the same threshold, any idea why? :)

Thanks again for your time and stay well!