mgarch with tgarch

[Max Overkott]

Hi there,

is there any way to implement asymmetric garch volatility (tgarch) in a multivariate garch model?
The "common" stata command does unfortunately only allow for arch and garch effects.

There was a post in the old statalist (http://www.stata.com/statalist/archi.../msg01380.html) but sadly no answer to it.

Thanks very much in advance!
Max

[Jorge Eduardo Perez Perez]

I am not an expert on this, but I think a CCC multivariate garch can be estimated in two steps, as described here.

http://vlab.stern.nyu.edu/doc/13?topic=mdls

If so, then maybe you can include the threshold arch terms in the univariate garch models of the first step, then use the standardized residuals from these models for the second step.

[Nav]

Hi

I am also looking into implementing asymmetric garch volatility into a multivariate model (DCC) to try and replicate the works of Capiello et al. (2006) in their AG-DCC GARCH model.

Is the above the best way to go about this on Stata, through manipulating the first step? I have also looked thoroughly and find nothing else, other than on other software I am unfamiliar with.

Any responses would be greatly appreciated,

Navdeep