Jan,
I believe you can. The basic idea to use -xtgls- with individual intercepts and
slope variables and with cross-sectional correlation (as in SUR) and
panel-specific AR(1) (as in -prais-).
Here is an example and comparison of -xtgls, p(c)- to -sureg-
webuse grunfeld, clear
tab com, gen(D)
foreach var1 of varlist D* {
foreach var2 of varlist mvalue kstock {
gen `var1'`var2' = `var1'*`var2'
}
}
aorder
xtgls invest D*, nocons p(c) corr(psar1)
*
*Compare to xtgls, p(c) to SUR
*
xtgls invest D*, nocons p(c)
keep invest mval kstock year com
reshape wide invest mval kstock, i(year) j(com)
forv i=1/10 {
local rhs "`rhs' ( invest`i' mvalue`i' kstock`i') "
}
sureg `rhs'
I hope this helps,
Scott
----- Original Message -----
From: "Jan Pettersson" <[email protected]>
To: <[email protected]>
Sent: Friday, February 27, 2004 8:39 AM
Subject: st: Serial Correlation in system estimator
> Dear all,
>
> Does anyone know any (preferably easy) way to correct for serial correlation
> (such as newey or prais) when a system is estimated?
> I am running SUREG.
>
> Thanks in advance,
> Jan
>
*
* 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/
