Diebold-Yilmaz index

For questions about EViews Add-ins available from the EViews Add-ins webpage. Note each add-in available on our webpage will have its own individual thread.

Moderators: EViews Gareth, EViews Moderator, EViews Esther

dakila
Posts: 425
Joined: Tue Nov 24, 2015 4:57 pm

Re: Diebold-Yilmaz index

Postby dakila » Thu May 07, 2020 6:13 pm

No. it does not affect the results. If you want to label variables for the table results, use this option.

Ning
Posts: 1
Joined: Sat May 09, 2020 8:27 am

Re: Diebold-Yilmaz index

Postby Ning » Sat May 09, 2020 8:09 pm

Hello, when I choose "Generalized" to run the index I get an error that says "Dot, diagonal, and solve require a vector argument", but I can get result when I choose "Cholesky". I wonder what does that mean because I want to use generalized variance decomposition. Could you please help me?

Haoren
Posts: 1
Joined: Thu Jul 02, 2020 3:51 pm

Re: Diebold-Yilmaz index

Postby Haoren » Mon Jul 06, 2020 10:15 am

Could anybody help me? How to calculate the directional spillovers, net spillovers and net pairwise spillovers? My endogeneous variables are sp500 r_10y djubscom usdx, number of lags is 4, generalized, and I choose Roll. I choose save pairwise spillovers. It says Error 14 in encrypted program. All of those ps are NA.
Attachments
dy2012 (1).xlsx
(183.86 KiB) Downloaded 241 times

Stataman
Posts: 3
Joined: Thu Jun 25, 2020 11:58 am

Re: Diebold-Yilmaz index

Postby Stataman » Thu Jul 16, 2020 2:56 am

Hi, I am new in this forum.

I am working on my dissertation and I came across with this methodology (Diebold-Yilmaz index) that i found specially interesting.

Maybe it is a silly question but I would like to know if we must work with STATIONARY data (making for instance the 1º difference and testing unit roots) whenever we are going to build the index. Or the calculations for the spillovers work also for the raw data, and there is no need for transformation?

Many Thanks

Kind Regards

Stataman
Posts: 3
Joined: Thu Jun 25, 2020 11:58 am

Re: Diebold-Yilmaz index

Postby Stataman » Mon Jul 20, 2020 5:22 am

Hi,

I am performing the Diebold-Yilmaz spillover index using the Add-In.

The results are consistent for the TOTAL spillover idex. However, the output does not generate the Net splillover index.

One could just substract the two matrix output (To matrix - From Matrix) to obtain the NET SPILLOVER, however when I open those matrix I realize that there are not dates inside, that complicates the interpretation.

Furthermore, the matrix output does not contain the names of the endogeous variables, and it assign new variables names.

I would really apreciate your help.

Many thanks

Kind Regards

dakila
Posts: 425
Joined: Tue Nov 24, 2015 4:57 pm

Re: Diebold-Yilmaz index

Postby dakila » Mon Jul 20, 2020 6:02 pm

Hi,

You are right. You already did it manually. That is only possible way to get the net spillover index. I will update it sometimes later.

a.rachmasyaputri
Posts: 5
Joined: Sun Nov 29, 2020 8:03 am

Re: Diebold-Yilmaz index

Postby a.rachmasyaputri » Mon Feb 08, 2021 9:20 pm

dakila wrote:Hi,

You are right. You already did it manually. That is only possible way to get the net spillover index. I will update it sometimes later.


Hi dakila, have you done to get net spillover index?

a.rachmasyaputri
Posts: 5
Joined: Sun Nov 29, 2020 8:03 am

Re: Diebold-Yilmaz index

Postby a.rachmasyaputri » Mon Feb 08, 2021 9:22 pm

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


Anybody can help answer this question, thank you in advance

a.rachmasyaputri
Posts: 5
Joined: Sun Nov 29, 2020 8:03 am

Re: Diebold-Yilmaz index

Postby a.rachmasyaputri » Thu Feb 11, 2021 10:00 pm

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.


Hai Dakila, I use eviews 11 to calculate dyindex. But I can't save the pairwise spillover, the information is : error 14 encrypted. How to solve the problem? thanks

gary
Posts: 1
Joined: Tue Oct 05, 2021 3:10 am

Re: Diebold-Yilmaz index

Postby gary » Tue Oct 05, 2021 6:33 pm

Hello! I have a problem about the settings of Rolling Windows and Number of lags. I found that when setting 100 as Rolling Window and 2 as Number of lags, the first spillover index in the SPILLINDEX series is the result based on the first 102 obs, and when setting 1 as the Number of lags, the result is based on the first 101 obs. I am wondering if this calculation setting is reasonable because in other ways we can use the 3rd to 100th obs as the dependent variables when setting 2 as Number of lags by holding the number of calculated obs the same as Rolling Windows setting(=100). I don't know which method is more appropriate and I didn't find explanation in Diebold and Yilmaz's paper(2009). Thank you very much!


Return to “Add-in Support”

Who is online

Users browsing this forum: No registered users and 2 guests