The distinction of variance in ARCH / GARCH

[Seeking_Knowledge]

Hello folks

I want to know whether there is only conditional variance applied in ARCH /GARCH

What happens to the unconditional variance?

And why this distinction

[nishantvats12]

Hi,

If you closely look at expansion of ARCH it happens to be Auto-regressive Conditional Heteroskedasticity, i.e., this model is used to treat for the problem of conditional heteroskedasticity in the model. This arises from the fact that the conditional variance is a function of time whereas the unconditional variance is constant over time and hence the major cause of heteroskedasticity. So you need not consider unconditional variance as it is constant over time. This topic is explained in detail by Enders in his book "Applied Econometric Time Series". Also the original paper of Engle and Granger in this regard might be helpful in clearing your doubts.

Best of luck!!

Regards
Nishant Vats

[Seeking_Knowledge]

thank you Nishant