Moderators: EViews Gareth, EViews Moderator, EViews Esther

dakila
Posts: 391
Joined: Tue Nov 24, 2015 4:57 pm

This thread is about the ARW add-in that implements Arias, Rubio-Ramirez and Waggoner algorithm for sign and zero restricted VARs.

pascalle
Posts: 6
Joined: Tue Jul 05, 2016 9:50 am

Hi,
I just run the ARW add-in with the data provided in the add-in file. However, it shows the following warning:
Error 17 in encrypted program, and there is no any output about Figure 6: Importance Sampler Output.
May I know how can I fix this problem?

Thank you
Pascal

dakila
Posts: 391
Joined: Tue Nov 24, 2015 4:57 pm

What is the version of Eviews you use? It only works for the version 11.

pascalle
Posts: 6
Joined: Tue Jul 05, 2016 9:50 am

Dear dakila,
Pascal

pascalle
Posts: 6
Joined: Tue Jul 05, 2016 9:50 am

Dear dakila,
May I know how can I obtain the forecast error of variance decomposition with its long horizon?
Thank you
Pascal

dakila
Posts: 391
Joined: Tue Nov 24, 2015 4:57 pm

Just try to estimate the FEVD with the different long horizon and check the convergence.

pascalle
Posts: 6
Joined: Tue Jul 05, 2016 9:50 am

Dear Dakila,
Pascal

Nynk
Posts: 2
Joined: Sun May 03, 2020 4:50 am

Dear Dakila,

When I try to estimate a zero and sign restricted VAR I always get error 15.
Do you know what might cause the problem?

I use these commands, I have 5 variables with monthly data.
matrix zeromatrix = @zeros(3,5)
zeromatrix(1,2)=1
zeromatrix(2,3)=1
zeromatrix(3,5)=1
matrix signmatrix =@zeros(2,5)
signmatrix(1,1)=1
signmatrix(2,4)=-1
arw(ndraws=1000, nsteps=24, horizons=2) 2 signmatrix zeromatrix @ lta lgdp lcp ciss lmro

dakila
Posts: 391
Joined: Tue Nov 24, 2015 4:57 pm

What is the version?

Nynk
Posts: 2
Joined: Sun May 03, 2020 4:50 am

dakila
Posts: 391
Joined: Tue Nov 24, 2015 4:57 pm

If you don't know the restriction matrix syntax, you can use the dialogbox instead of command line

zero restriction matrix should be like this
matrix zeromatrix = @zeros(1,5)
zeromatrix(1,2)=1
zeromatrix(1,3)=1
zeromatrix(1,5)=1