I suppose GLS clustered by firms may help most. However, please read Petersen, 2006. Estimating Standard Errors in Finance Panel Data Sets: Comparing Approaches, Working paper.
http://www.kellogg.northwestern.edu/faculty/petersen/htm/working.htm
Nicola
P.S. I'll NOT receive/read any email but the Digest.
At 02.33 20/09/2008 -0400, "Steven Proud" wrote:
>Dear all,
>
>I was wondering if anyone could help me.
>
>I am trying to estimate a model in a panel, with time t and cross sectional
>variable s
>
>F(st)=F(s)+e(st)+u(st)
>where e(st)=p*e(st-1)+a(st)
>where u(st) and a(st) are random error terms, and e(st) is a persistent
>error of AR(1)
>and F(s) is some underlying fixed effect.
>
>Obviously if there wasn't a persistent term, then it would be easy to
>estimate as simply a fixed effect model, and similarly, if there wasn't a
>random shock term, I could simply estimate it using xtregar. However, I'm
>slightly confused as to how to go about estimating it as it is. I
>considered mixed models, but couldn't see how I could use these to estimate
>the model.
>
>Any help would be very gratefully received.
>
>Thanks in advance,
>
>Steven Proud
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
