Page 3 of 3

Re: Diebold-Yilmaz index

Posted: Tue May 14, 2019 4:40 pm
by dakila
What is your eviews version?

Re: Diebold-Yilmaz index

Posted: Tue May 14, 2019 9:09 pm
by Wiz_Ven
Eviews 11.

Re: Diebold-Yilmaz index

Posted: Tue May 14, 2019 11:30 pm
by dakila
Yes, you are right. It does not work for the version 11. I will fix it soon. But it can work for the version 10 if you have access to it.

Re: Diebold-Yilmaz index

Posted: Tue May 14, 2019 11:59 pm
by Wiz_Ven
I have also Ev 10+ and the same error appears.

Re: Diebold-Yilmaz index

Posted: Thu May 16, 2019 6:15 pm
by dakila
The bug is fixed. Please update the add-in. I hope the moderator will post it soon. But carefully choose the rolling window parameter. It should not exceed the sample size.

Re: Diebold-Yilmaz index

Posted: Fri May 17, 2019 5:31 am
by Wiz_Ven
Thank you very much. As soon as the new version of the code is online I will try to run the code following your instructions.

Re: Diebold-Yilmaz index

Posted: Fri May 17, 2019 10:03 am
by Wiz_Ven
Dear Dakila,
I have tried the updated version. Now it works fine. Thank you very much.

Re: Diebold-Yilmaz index

Posted: Thu Jun 13, 2019 12:20 am
by Asimshaikh
dakila wrote:This thread is about the dyindex add-in that estimates the Diebold-Yilmaz index of spillover using forecast error variance decomposition method of a VAR model.

I am new to this method can anybody help me by telling steps to calculate dyindex?

Re: Diebold-Yilmaz index

Posted: Sat Jun 22, 2019 1:10 am
by Asimshaikh
Helloe everybody
I have calculated spillover index using this add-ins but one thing i am still unable to get is net pairwise connectedness.Can anybody help on this?

Re: Diebold-Yilmaz index

Posted: Sat Jun 22, 2019 1:21 am
by dakila
When estimate the DY index just tick the save pairwise spillovers. After that do it manually get the net pairwise connectedness.

Re: Diebold-Yilmaz index

Posted: Sat Jul 20, 2019 1:36 am
by harpreet_kaur
Hi, i have applied Diebold model in Eview9 to check return spillover among five return series ( daily data, around 3000 observation). I have some queries
1. Endogenous variable should be stationary?

2. On the basis of lag length criteria, appropriate lag is 1 (SC) and 4( AIC, HQ). When in run Diebold model with 4 lag then an error message comes "Near singular matrix'. No such error msg come when i run this model with 1 lag. Should i use lag order 1?

3. After running this Diebold model on return series with lag order 1, spilloverindex file and each PS_ _ file does not show any value. 'NA" only. I don't know why.

4. Last question, if I want to check volatility spillover instead of return spillover then what should be endogenous variable.

Thank You

Re: Diebold-Yilmaz index

Posted: Sun Jul 21, 2019 10:46 pm
by harpreet_kaur
harpreet_kaur wrote:Hi, i have applied Diebold model in Eview9 to check return spillover among five return series ( daily data, around 3000 observation). I have some queries
1. Endogenous variable should be stationary?

2. On the basis of lag length criteria, appropriate lag is 1 (SC) and 4( AIC, HQ). When in run Diebold model with 4 lag then an error message comes "Near singular matrix'. No such error msg come when i run this model with 1 lag. Should i use lag order 1?

3. After running this Diebold model on return series with lag order 1, spilloverindex file and each PS_ _ file does not show any value. 'NA" only. I don't know why.

4. Last question, if I want to check volatility spillover instead of return spillover then what should be endogenous variable.

Thank You


Can anybody please help. I am facing above problems while running dyindex