Estimate AR-EGARCH with 1;7 ARCH, 1;7 GARCH

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

moeflon
Posts: 1
Joined: Fri Apr 10, 2015 2:11 am

Estimate AR-EGARCH with 1;7 ARCH, 1;7 GARCH

Postby moeflon » Fri Apr 10, 2015 2:22 am

Hello everyone,

For a study on estimating and forecasting Value-at-Risk, I developed several models.
Unfortunately on my fourth model I am stuck..

I aim to estimate an AR-EGARCH-N model, thus with normal distributed error terms and also AR-EGARCH-t, thus with t-distributed ones.
My mean equation and variance equation can be found in the image that I uploaded.

The mean equation is a simple AR(7) model with intercept, and the variance equation consists of an ARCH(1) and ARCH(7) term, a GARCH(1) and GARCH(7) term and an asymmetric term.
Now: my issue is that eviews cannot estimate this model with the standard equation screen, selecting egarch, etc. because I cannot select lag 1 and lag 7...

I tried to do this in the logl window, but I'm stuck

Can anyone help me or give any advice?

Thank

ps: my logl code attempt

@logl LL3

//MEAN EQUATION

res = log_returns - c(1) - c(2)*log_returns_1 - c(3)*log_returns_2 -
c(4)*log_returns_3 - c(5)*log_returns_4 - c(6)*log_returns_5 -
c(7)*log_returns_6 - c(8)*log_returns_7

//VARIANCE EQUATION

LOG(var) =
c(9)
+ c(10)*(RESID(-1)/@SQRT(GARCH(-1))
+ c(11)*LOG(GARCH(-1))
+ c(12)*ABS(RESID(-1)/@SQRT(GARCH(-1)))
+ c(13)*(RESID(-7)/@SQRT(GARCH(-7))
+ c(14)*LOG(GARCH(-7))


//DISTRIBUTION TYPE ERROR TERMS -normal

logl3 = log(@dnorm(res/@SQRT(var))) - log(var)/2
Attachments
eviews.jpg
model
eviews.jpg (33.45 KiB) Viewed 1790 times

Return to “Estimation”

Who is online

Users browsing this forum: Cscottmcwall-ea and 27 guests