Re: AW: st: panel-specific autocorrelation with unbalanced panels

[Marck Bulter <177316mb@student.eur.nl>]

Jessica �lschl�ger wrote:
> Delia and Nicola,
>
> thank you very much for your suggestions. 
> I've already had tried both but I don't see how to test for panel-specific
> autocorrelation with xtserial or abar.
> It may be done with xttest1 but this works for balanced panels only.
>
> Jessica
>  
> ----------------------------------------------------------------
> Dipl. oec. Jessica �lschl�ger
>  
> Forschungsstelle Europ�ische Integration
> Universit�t Hohenheim
>
> -----Urspr�ngliche Nachricht-----
> Von: owner-statalist@hsphsun2.harvard.edu
> [mailto:owner-statalist@hsphsun2.harvard.edu] Im Auftrag von Delia Ionascu
> Gesendet: Montag, 10. Dezember 2007 20:00
> An: statalist@hsphsun2.harvard.edu
> Betreff: Re: st: panel-specific autocorrelation with unbalanced panels
>
> Another option is to use 
> abar  (you could download it from 
> http://ideas.repec.org/c/boc/bocode/s437501.html)
>
> abar performs the Arellano-Bond (1991) test for autocorrelation. 
> abar runs after regress, ivreg, ivreg2, and ivreg2, gmm; also after 
> newey and newey2
>
> Delia
>
>
> ----- Original Message -----
> From: nicola.baldini2@unibo.it
> Date: Monday, December 10, 2007 6:51 pm
> Subject: Re: st: panel-specific autocorrelation with unbalanced panels
> To: statalist@hsphsun2.harvard.edu
> Cc: jessica.oelschlaeger@uni-hohenheim.de
>
>   
> > Does -xtserial- from sj3-2 help?
> > Nicola
> > At 02.33 08/12/2007 -0500, =?iso-8859-1?Q?Jessica_=D6lschl=E4ger?= 
> > wrote:>Dear Statalisters,
> >     
> > > I want to test for panel-sepcific autocorrelation in my panel 
> > >       
> > data set.
> >     
> > > Unfortunately xttest1 doesn?t work as my data is unbalanced.
> > > Is there a way to fix this problem in Stata?
> > >
> > > Any comments would be greatly appreciated.
> > >
> > > Jessica
> > >       
> > *
> > *   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/
>   
Dear Jessica,

If you regress each cross-section i using,

forvalues i=1/108 {
display "Regress pstrmon for CUSIP" `i'
quietly ivreg pstrmon price mat age coup pstrmonprev pstrprev intr ivol 
compl (precmon = precmonprev) if cusip == `i'
estimates store model `i'
more
}

Then you should be able to test for autocor. etc, just like a normal 
time series regression for each i. Although you should change the 
estimate line a bit, but you get picture. Give it a try.

regards,
Marck Bulter




*
*   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/