Thank you Jean for your corrections and suggestion.
However, the 2 specifications (xtabond and xtabond2,
using the corrected specification you suggested, still
do not seem to yield the same results. Can anyone see
what is going wrong?
.xtabond ROA size if OK==1, lags(1) maxldep(4)
maxlags(4) pre(vara varb varc,lagstruct(0,4))
pre(vard, lagstruct(0,4) endogenous) robust artests(2)
noconstant
. xtabond2 ROA l.ROA size vara varb varc vard if
OK==1, iv(size) gmm(l.ROA vara varb varc l.vard, lag(0
4)) robust noleveleq
(commands and output pasted below)
Thanks,
Janine
XTABOND:
.xtabond ROA size if OK==1, lags(1) maxldep(4)
maxlags(4) pre(vara varb varc,lagstruct(0,4))
pre(vard, lagstruct(0,4) endogenous) robust artests(2)
noconstant
Arellano-Bond dynamic panel-data estimation
Number of obs = 19188
Group variable (i): gvkey
Number of groups = 644
Wald chi2(.) = .
Time variable (t): date
Obs per group: min = 4
avg = 29.79503
max = 43
One-step results
Robust
D.ROA Coef. Std. Err. z P>z [95%
Conf. Interval]
ROA
LD. .0851945 .0163457 5.21 0.000
.0531575 .1172315
vara
D1. -1.03e-06 7.78e-07 -1.32 0.186
-2.55e-06 4.96e-07
varb
D1. -.0123586 .0029563 -4.18 0.000
-.0181528 -.0065644
varc
D1. .0222433 .008245 2.70 0.007
.0060834 .0384032
vard
D1. .0387917 .0102903 3.77 0.000
.0186231 .0589602
size
D1. .0790556 .0227795 3.47 0.001
.0344086 .1237026
Arellano-Bond test that average autocovariance in
residuals of order 1 is 0:
H0: no autocorrelation z = -5.02 Pr > z = 0.0000
Arellano-Bond test that average autocovariance in
residuals of order 2 is 0:
H0: no autocorrelation z = -0.39 Pr > z = 0.6937
--------------------
XTABOND2:
. xtabond2 ROA l.ROA size vara varb varc vard if
OK==1, iv(size) gmm(l.ROA vara varb varc l.vard, lag(0
4)) robust noleveleq
Favoring space over speed. To switch, type or click on
mata: mata set matafavor speed.
Warning: Number of instruments may be large relative
to number of observations.
Suggested rule of thumb: keep number of instruments <=
number of groups.
Arellano-Bond dynamic panel-data estimation, one-step
difference GMM results
Group variable: gvkey
Number of obs = 19188
Time variable : date
Number of groups = 644
Number of instruments = 1055
Obs per group: min = 4
Wald chi2(6) = 161.10
avg = 29.80
Prob > chi2 = 0.000
max = 43
Robust
Coef. Std. Err. z P>z [95% Conf.
Interval]
ROA
L1. -.0851043 .0278936 -3.05 0.002
-.1397748 -.0304338
size .1497114 .0231574 6.46 0.000
.1043238 .195099
vara -9.15e-07 7.05e-07 -1.30 0.194
-2.30e-06 4.66e-07
varb -.0244269 .0030017 -8.14 0.000
-.0303103 -.0185436
varc .0233414 .0068269 3.42 0.001
.0099608 .036722
vard .0608585 .0170534 3.57 0.000
.0274344 .0942826
Hansen test of overid. restrictions: chi2(1140) =
643.47 Prob > chi2 = 1.000
Arellano-Bond test for AR(1) in first differences: z =
-4.25 Pr > z = 0.000
Arellano-Bond test for AR(2) in first differences: z =
-2.44 Pr > z = 0.015
-Janine
--- "Salvati, Jean" <[email protected]> wrote:
> > -----Original Message-----
> > From: [email protected]
> > [mailto:[email protected]] On
> Behalf Of
> > Janine Darfield
> > Sent: Monday, June 05, 2006 11:03 AM
> > To: [email protected]
> > Subject: st: xtabond/xtabond2 equivalent
> specification
> >
> > Hi all,
> >
> > I thought that the following 2 commands should be
> equivalent,
> > but they are not. Can anyone point out where I am
> going
> > wrong, or what an equivalent specification would
> be?
> >
> > .xtabond depvar size if date<120, lags(1)
> maxldep(4)
> > maxlags(4) pre(a b c, lagstruct(0,4)) pre(d,
> > lagstruct(0,4) endogenous) robust artests(2)
> noconstant and
> > .xtabond2 depvar size a b c d if date<120,
> gmm(l.depvar a b c
> > d, lag(0 4)) robust noleveleq
>
> There are at least three differences.
>
> First, d is endogenous in your xtabond command, but
> predetermined in
> your xtabond2 command. If you want xtabond2 to treat
> d as endogenous,
> you should replace gmm(l.depvar a b c d, lag(0 4))
> with gmm(l.depvar a
> b c l.d, lag(0 4)).
>
> Second, you need to include l.depvar in the list of
> regressors in your
> xtabond2 command.
>
> Third, you need to tell xtabond2 that size is
> exogenous.
>
> The resulting xtabond2 command is:
>
> .xtabond2 depvar l.depvar size a b c d if date<120,
> iv(size)
> gmm(l.depvar a b c l.d, lag(0 4)) robust noleveleq
>
> Jean Salvati
>
>
> *
> * 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/
>
___________________________________________________________
Yahoo! Messenger - with free PC-PC calling and photo sharing. http://uk.messenger.yahoo.com
*
* 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/
