Olga,
When you have too many parameters so that you have a flat or non-concave
likelihood surface, you have parameterized. In this case, some of the
parameters may be redundant and/or others may be non-significant when
that happens.
- You're welcome,
Robert A. Yaffee, Ph.D.
Research Professor
Shirley M. Ehrenkranz
School of Social Work
New York University
home address:
Apt 19-W
2100 Linwood Ave.
Fort Lee, NJ
07024-3171
Phone: 201-242-3824
Fax: 201-242-3825
[email protected]
----- Original Message -----
From: Olga Gorbachev Melloni <[email protected]>
Date: Wednesday, August 30, 2006 2:34 pm
Subject: Re: st: testing for optimal lag in garch
> What do you mean by "with the reduction of the arch (1/3)
> garch(1/3) maximum
> length"? Do you mean I should be looking at the loglikelyhood at
> the
> convergence? I know that t-stats are significant at arch(1/5), but
> how can I
> be sure I am not overparametrizing?
>
> thank you,
>
>
>
> >From: Robert A Yaffee <[email protected]>
> >Reply-To: [email protected]
> >To: [email protected]
> >Subject: Re: st: testing for optimal lag in garch
> >Date: Wed, 30 Aug 2006 14:28:56 -0400
> >
> >Olga,
> > Testing several models can be done sequentially
> >with the reduction of the arch(1/3) garch(1/3) maximum lag length.
> But
> >there is no step down procedure. Also, with such nonlinear models,
> >one needs to be careful about overparameterizing
> >them, lest your model does not converge.
> > - Bob
> >
> >
> >Robert A. Yaffee, Ph.D.
> >Research Professor
> >Shirley M. Ehrenkranz
> >School of Social Work
> >New York University
> >
> >home address:
> >Apt 19-W
> >2100 Linwood Ave.
> >Fort Lee, NJ
> >07024-3171
> >Phone: 201-242-3824
> >Fax: 201-242-3825
> >[email protected]
> >
> >----- Original Message -----
> >From: Olga Gorbachev Melloni <[email protected]>
> >Date: Wednesday, August 30, 2006 2:07 pm
> >Subject: st: testing for optimal lag in garch
> >
> > > hi,
> > >
> > > I was wondering whether there is a test for an optimal
> structure of
> > > Garch(p,q) process in stata, and if so, what is it? I need to test
> > > several
> > > models and choose the one that fits the data best, i.e. Arch(1)
> vs.> > Garch(1,1) vs. Arch (1/2)...
> > >
> > >
> > > thank you
> > >
> > >
> > > *
> > > * For searches and help try:
> > > * http://www.stata.com/support/faqs/res/findit.html
> > > * http://www.stata.com/support/statalist/faq
> > > * http://www.ats.ucla.edu/stat/stata/
> > >
> >*
> >* For searches and help try:
> >* http://www.stata.com/support/faqs/res/findit.html
> >* http://www.stata.com/support/statalist/faq
> >* http://www.ats.ucla.edu/stat/stata/
>
>
> *
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
