Modelling conditional variance in eviews using 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

naz
Posts: 3
Joined: Tue Jul 31, 2012 7:41 pm

Modelling conditional variance in eviews using GARCH

Postby naz » Tue Jul 31, 2012 7:56 pm

Hi,

Am stuck in the middle of my analysis with the following problem. Am actually estimating a model using GARCH with the presence of conditional variance.. The problem is, I do not know how to do so in EViews. I'll be grateful if I can obtain help on this forum.

The equation of the conditional variance is :

var = c + B0 (E t-1)^2 + B1 (VAR t-1)^2 + B2 (Dummy)((E t-1)^2)

with the dummy variable=1 if the lagged error E t-1 is negative and equal to zero otherwise.

note: E t is the error term of the mean equation

trubador
Did you use forum search?
Posts: 1518
Joined: Thu Nov 20, 2008 12:04 pm

Re: Modelling conditional variance in eviews using GARCH

Postby trubador » Wed Aug 01, 2012 12:07 am

This looks like a TARCH model. The command line estimation can be done as follows:

Code: Select all

equation myeq.arch(1,1,thrsh=1) y c

Here, y is your dependent variable of the mean equation not the conditional variance. You can also use the Estimation dialog box, but do not forget to change the value of "Threshold order".

naz
Posts: 3
Joined: Tue Jul 31, 2012 7:41 pm

Re: Modelling conditional variance in eviews using GARCH

Postby naz » Wed Aug 01, 2012 1:02 pm

thanks trubador, but actually i cannot insert the following mean equation

I am trying to estimate a garch model whose mean equation includes the variance equation three times. And also, two of the variances are multiplied by another independent variable. I don't know how to estimate this in eviews.

The mean equation is:

Rt = C+ B0(variance) - B1 (Rt-1) - B2(Rt-1)(Variance) - B3 (Rt-1)(dummy variable)(variance) + error_term

and the conditional variance is modelled by:

variance_t = C + X1 (error_term_t-1 ^2) + X2 (variance_t-1)

Please help.

trubador
Did you use forum search?
Posts: 1518
Joined: Thu Nov 20, 2008 12:04 pm

Re: Modelling conditional variance in eviews using GARCH

Postby trubador » Wed Aug 01, 2012 11:19 pm

This is a very specific form of GARCH model. In such cases, you have to build the model from scratch. You can use EViews' LogL object for that purpose. You will find plenty of examples if you search the forum.

wangwz
Posts: 16
Joined: Sat Feb 20, 2016 3:45 am

Re: Modelling conditional variance in eviews using GARCH

Postby wangwz » Fri Sep 15, 2017 1:32 am

has it beeb solved, please?


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 30 guests