Date:         Sun, 31 Jan 2010 18:45:26 -0800
Reply-To:     shiling99@YAHOO.COM
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Shiling Zhang <shiling99@YAHOO.COM>
Organization: http://groups.google.com
Subject:      Re: replicate PROC AUTOREG using PROC REG
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

On Jan 30, 1:45 pm, Nguyen An Nguyen <nguyenanngu...@gmail.com> wrote:
> Hi,
> I wonder if it is possible to replicate a PROC AUTOREG model by
> running multiple linear regressions (such as Proc Reg). The reason for
> doing this is that I need to manipulate data manually (such as weight/
> reweight observations, adjust for outliers and create different
> regimes) that AUTOREG would not accommodate.
> My base-line model is essentially:
>    Proc Autoreg data = a;
>       model y = ylag1 x y dummy_z / garch(1,1);
>
> Any advice is much appreciated!

I believe that you cannot bypass it. The garch model is done under
likelihood estimation which is not available in REG.

But It can be don in nlmixed and nlp.