I assume you have a lagged dependent var in your model.
Since you are using an GMM-estimator, autocorrelation of order 1 is no
problem but the reason for using GMM in order to instrument the endogenuous
variable. You should look at the p-value of the autocorrelation test of
second order. You have problems if there is autocorrelation of second order.
Refer to the original arellano/bond paper (1991) that is cited in - help
xtabond2. By the way: Maybe xtabond2 suits your needs better than xtabond.
Or refer to Greene, Verbeek or Baltagi.
Niko
-----Urspr�ngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Rafael Terra de
Menezes
Gesendet: Freitag, 13. Januar 2006 13:55
An: [email protected]
Betreff: st: xtabond - test for autocorrelation
Dear Stata users,
I'm working with a panel of N=2610 and T=7. I ran my specification with
xtabond and, according to the results below, the tests for autocorrelation
indicates that I can reject the hypotesis of no average autocovariance in
residuals of order 1. I would like to know what should I do to deal with
this problem. Or is this not a problem at alll?
Thanks
Rafael Terra de Menezes
(Student)
Department of Economics - University of S�o Paulo
Brazil
. xtabond lndo lnro LRF trend, robust
Arellano-Bond dynamic panel-data estimation Number of obs =
13050
Group variable (i): cod Number of groups =
2610
Wald chi2(4) =
2434.80
Time variable (t): ano Obs per group: min =
5
avg =
5
max =
5
One-step results
----------------------------------------------------------------------------
--
| Robust
lndo | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------
-------------+------
lndo |
LD | .080924 .0145315 5.57 0.000 .0524428
.1094052
lnro |
D1 | .7556548 .0155879 48.48 0.000 .725103
.7862066
LRF |
D1 | -.0396747 .0030355 -13.07 0.000 -.0456241
-.0337253
trend |
D1 | .0074261 .0005826 12.75 0.000 .0062843
.0085679
_cons | .0084472 .0007847 10.77 0.000 .0069093
.0099852
----------------------------------------------------------------------------
--
Arellano-Bond test that average autocovariance in residuals of order 1 is 0:
H0: no autocorrelation z = -20.98 Pr > z = 0.0000
Arellano-Bond test that average autocovariance in residuals of order 2 is 0:
H0: no autocorrelation z = 1.67 Pr > z = 0.0951
*
* 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/
