EViews.com

Hi,

I have not used Eviews before and do not have much experience of econometrics packages in general so i hope someone can help me as I am going mad trying to work out what to do here....and Im sure it will be simple enough!
I am trying to estimate a GJR-GARCH model in Eviews 7 for Brazil which is as standard, apart from also including a term for a lag on the error term of a world index into the conditional variance.
Does anyone know how i cant input this into the conditional variance equation, where and how i need to do it?
I have attached the conditional variance term in a word document as i cannot copy it here.
Also in the mean equation box do i need to add anything else apart from the return series for Brazil.

Thanks!

I may be missing something but don't you just want to add your regressor to the variance equation regressor box?

Hi Glenn, Thanks for getting back,

In the variance term for Brazil(and the other countries involved) I wanted to add a lag of the squared residuals of the conditional mean of the world index to measure the impact of the previous day's foreign news on the Brazilian Index. How would I add this into the variance equation and how would i get the residuals to then save and add them into the country variance equation.

I suppose that this means that I should actually look at the equation :)

Hang on...

Okay, I'm assuming that the world index error is from a different equation. In that case, just estimate the world index equation, save it's residuals using Proc/Make Resids...(say to the series WRES). Then youshould be able to add WRES(-1)^2 to the variance specification Variance regressors edit field.

Unless I'm still missing something...

Thanks again for replying, this sounds about right, Ill give it a go tomorrow and reply in the forum! But this sounds like it should work,

Richard