EViews.com

Thanks so much for your feedback. I realize it is for garch in mean, I also did the same without the mean garch effect.
I adjusted the coding and made the iterations 1000. Thanks for pointing that out!

On your last comment - consider this paper by Christopher, et al and note on page 1076. Do they use the BV Garch Bekk variances (h11, h22 and h12) to estimate the DCC?
And would that be comparable with the explicit specification of the DCC technique? I note your point, but are the authors wrong in their use of it?

Thanks for your kind help!

No, it is not wrong of course. Different approach, different results. That's all...

Right, but is it safe to say that we can then use either the DCC's of multivariate GARCH models or the DCC of Engle (where the correlation is handled explicitly) to study dynamic conditional correlations between assets? Thanks again for your help!

Yes, either one of these methods can be used to estimate DCC.

Thanks so much for your feedback. I realize it is for garch in mean, I also did the same without the mean garch effect.
I adjusted the coding and made the iterations 1000. Thanks for pointing that out!

On your last comment - consider this paper by Christopher, et al and note on page 1076. Do they use the BV Garch Bekk variances (h11, h22 and h12) to estimate the DCC?
And would that be comparable with the explicit specification of the DCC technique? I note your point, but are the authors wrong in their use of it?

Thanks for your kind help!

Dear Nick, is it possible for you to share the code without the mean garch effects? Thank you

Hi everyone.

Hopefully, you'll find the EViews codes provided in the attached master thesis useful. The codes are written to estimate the general form BEKK-GARCH where A and B are unrestricted. Thus, you could test for volatility spillover

http://www.diva-portal.org/smash/get/di ... text01.pdf

Best of luck to you all