EViews.com

Hi I am new to Eviews. I was trying to estimate a Garch (1,1) using variance regressors. My goal is to demonstrate that the creation of an ETF generates an impact on the variance of the underlying stocks. So I estimate a garch (1,1) using the return of a stock as dependent variable an d the return of an index as the independent variable. Then as variance regressors I use a dummy that gets the value of one for all the dates after the inception of the ETF and 0 otherwise.
My problem is that I use a multiplicative dummy with the arch and garch terms I don’t know how to estimate those series in eviews (the serie garch*dummy).
My question is, is there a way to save the garch term of a regression as a separate series in order to create garch*dummy manually?
Here is the model:
(H_t) = c+d(D_t)+f(e^2)_(t-1)+ g(D_t)(e^2)_(t-1)+(iH_(t-1))+j(D_(t) )(H_(t-1))

UV_before=c/(1-(a+b))

UV_after=(c+d)/(1-(a+b+g+j))

Where Dt is the dummy variable
Thank you

It is a custom specification. Such extensions can be estimated using the logl object, which requires defining the log likelihood contribution of each observation in your sample as a function of unknown parameters. Please search the forum and/or sample programs in your EViews folder for examples of maximum likelihood estimation of garch models.