Modeling conditional variance without ARCH/GARCH terms
Posted: Tue Sep 07, 2010 10:07 am
I face the following problem. I estimate a GARCH(1,1) model with implied volatility (IV) as an extra variance regressor. I want to run some likelihood ratio tests in order to test the significance of the various terms. I can run all the alternative specifications (and get the respective Log-likelihoods) except one: The specification where only IV appears in the conditional variance equation (ARCH:0, GARCH:0, Variance Regressors: IV). When I try to do that, a box appears "must include one ARCH or GARCH term in specification".
I guess this can only be done by writing a program but,unfortunately, my programming skills in EViews are non-existent. So, I would be very happy if someone can help me estimate this model (or, at least, to show me how to write the specification in the command line).
I guess this can only be done by writing a program but,unfortunately, my programming skills in EViews are non-existent. So, I would be very happy if someone can help me estimate this model (or, at least, to show me how to write the specification in the command line).