GARCH Eviews calculation of forecast questions

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

phoenix1914
Posts: 7
Joined: Sat Nov 17, 2018 1:48 pm

GARCH Eviews calculation of forecast questions

Postby phoenix1914 » Mon Apr 05, 2021 5:55 am

Hello all

I am working on some project to evaluate the forecast ability of different GARCH models, but before doing that I have some questions on how Eviews estimates some values when it comes to forecast of volatility.

Assume we have some log returns of an index and I estimate a GARCH(1,1) model, with zero mean equation. So the return formula will be:

rt = μ +et where μ=0 and et = square(ht)*ηt , where ηt is i.i.d with mean zero and 1 variance and ht the GARCH specification:
ht = ω + α*e(t-1^2) + β * h(t-1) . So a classic GARCH(1,1) with zero mean.

I have some questions when I use the static and dynamic forecast of Eviews:

1) How is the dynamic forecast's e(t-1) calculated by eviews? By trial and error I found that Eviews uses e(t-1) = square(h(t-1) but shouldn't it be equal to rt instead?

2) How is the static forecast calculated? Here I want to ask, what is Eviews using as actual variance proxy for t-1 for use in the GARCH model?

3) Finally, a general question. Is there a way to obtain the forecasted values of variance as a list? I see that on the graph the values are displayed but not sure how I can have them as a series.

Thank you in advance!

phoenix1914
Posts: 7
Joined: Sat Nov 17, 2018 1:48 pm

Re: GARCH Eviews calculation of forecast questions

Postby phoenix1914 » Mon Apr 05, 2021 8:04 am

Hello all

After investigating a bit more I realized that I was a bit confused between the static and dynamic forecast. The static is using the t-1 period actual residual for the ht calculation, while the dynamic is using the h(t-1) in place of e(t-1) , so the dynamic ht formula is actually ω+(α+β)*h(t-1)

But what I want still to ask is if I can obtain somehow the numerical series of the GARCH forecasted values (either dynamic or static). I want to calculate MAE/RMSE etc for the variance, using the squared returns as proxy of the true volatility, but since Eviews doesn't do that automatically, I need to have the volatility forecast series to do that manually if possible.

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13307
Joined: Tue Sep 16, 2008 5:38 pm

Re: GARCH Eviews calculation of forecast questions

Postby EViews Gareth » Mon Apr 05, 2021 8:20 am

Hit the forecast button and enter a name in the garch box
Follow us on Twitter @IHSEViews


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 29 guests