Hello everyone,
I have an issue regarding xtabond2 and Hansen statistics and I was wondering if anyone could help me with that.
Here is my code and what Stata provides:
Lagged dependent variable, L.winv2, and two other regressors, wdiv2 and wndf, are endogenous variables. I treat the rest as exogenous. As you can see the Hansen test of overid. restrictions comes out highly significant, which is a problem I assume. Given that AR(2) is insignificant, what could cause this and what would a possible solution be? Is there an issue with the code? I am pretty new to xtabond2 and I would appreciate every comment on this.
Many thanks.
I have an issue regarding xtabond2 and Hansen statistics and I was wondering if anyone could help me with that.
Here is my code and what Stata provides:
Code:
xtabond2 winv2 wndf wdiv2 wcf wmtb wtang cage wsalesg wsize l.stdev l.winv2 y4-y32, gmmstyle(l.winv2 wndf
> wdiv2, lag(2 5) collapse) iv( wmtb wcf wtang l.stdev cage wsalesg wsize y4-y32, eq(level)) iv(wmtb wcf wta
> ng l.stdev cage wsalesg wsize, eq(diff)) twostep small robust
Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm.
Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: companyno Number of obs = 4925
Time variable : datayearfi~l Number of groups = 463
Number of instruments = 59 Obs per group: min = 1
F(39, 462) = 14.51 avg = 10.64
Prob > F = 0.000 max = 30
------------------------------------------------------------------------------
| Corrected
winv2 | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
wndf | .5197724 .0736962 7.05 0.000 .3749511 .6645937
wdiv2 | -.0761525 .0361605 -2.11 0.036 -.1472119 -.005093
wcf | .1376568 .0394813 3.49 0.001 .0600716 .215242
wmtb | .0065393 .0032135 2.03 0.042 .0002244 .0128543
wtang | .031711 .0093283 3.40 0.001 .0133798 .0500421
cage | -.0117218 .0047613 -2.46 0.014 -.0210783 -.0023652
wsalesg | .0277694 .010588 2.62 0.009 .0069628 .048576
wsize | .0029869 .0013757 2.17 0.030 .0002836 .0056903
|
stdev |
L1. | -.1754014 .1576928 -1.11 0.267 -.4852853 .1344825
|
winv2 |
L1. | .1345404 .0392788 3.43 0.001 .0573531 .2117277
|
y4 | -.0670014 .0490034 -1.37 0.172 -.1632986 .0292958
y5 | -.0764702 .0436855 -1.75 0.081 -.1623172 .0093768
y6 | -.0466703 .043778 -1.07 0.287 -.132699 .0393584
y7 | -.0535834 .039934 -1.34 0.180 -.1320582 .0248914
y8 | -.0046665 .0448924 -0.10 0.917 -.0928851 .083552
y9 | -.0310349 .0449769 -0.69 0.491 -.1194196 .0573498
y10 | -.0266905 .0421176 -0.63 0.527 -.1094563 .0560752
y11 | -.047943 .0454478 -1.05 0.292 -.137253 .041367
y12 | -.0457779 .0416776 -1.10 0.273 -.127679 .0361233
y13 | -.0440037 .0414834 -1.06 0.289 -.1255232 .0375157
y14 | -.0183853 .0432421 -0.43 0.671 -.1033609 .0665902
y15 | -.0394861 .0432887 -0.91 0.362 -.1245532 .0455809
y16 | -.0272495 .0436253 -0.62 0.533 -.1129781 .058479
y17 | -.0428482 .0433147 -0.99 0.323 -.1279664 .04227
y18 | -.0159717 .0433348 -0.37 0.713 -.1011294 .0691861
y19 | -.0042772 .0432104 -0.10 0.921 -.0891906 .0806361
y20 | -.0204746 .0426472 -0.48 0.631 -.1042811 .0633319
y21 | -.0016345 .043723 -0.04 0.970 -.087555 .0842861
y22 | -.0181546 .042542 -0.43 0.670 -.1017543 .0654451
y23 | -.0264723 .0429187 -0.62 0.538 -.1108123 .0578677
y24 | -.0299797 .0425649 -0.70 0.482 -.1136245 .0536651
y25 | -.0318005 .0421004 -0.76 0.450 -.1145325 .0509314
y26 | -.0389249 .0422431 -0.92 0.357 -.1219373 .0440875
y27 | -.0237457 .0427874 -0.55 0.579 -.1078278 .0603365
y28 | -.02853 .042558 -0.67 0.503 -.1121612 .0551012
y29 | -.0178813 .0427483 -0.42 0.676 -.1018865 .0661239
y30 | -.0284094 .0426339 -0.67 0.506 -.1121898 .055371
y31 | -.0375211 .0422606 -0.89 0.375 -.120568 .0455257
y32 | -.0400235 .0424093 -0.94 0.346 -.1233625 .0433156
_cons | .0029791 .0440888 0.07 0.946 -.0836603 .0896185
------------------------------------------------------------------------------
Instruments for first differences equation
Standard
D.(wmtb wcf wtang L.stdev cage wsalesg wsize)
GMM-type (missing=0, separate instruments for each period unless collapsed)
L(2/5).(L.winv2 wndf wdiv2) collapsed
Instruments for levels equation
Standard
_cons
wmtb wcf wtang L.stdev cage wsalesg wsize y4 y5 y6 y7 y8 y9 y10 y11 y12
y13 y14 y15 y16 y17 y18 y19 y20 y21 y22 y23 y24 y25 y26 y27 y28 y29 y30
y31 y32
GMM-type (missing=0, separate instruments for each period unless collapsed)
DL.(L.winv2 wndf wdiv2) collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z = -6.77 Pr > z = 0.000
Arellano-Bond test for AR(2) in first differences: z = 0.84 Pr > z = 0.400
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(19) = 110.06 Prob > chi2 = 0.000
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(19) = 32.49 Prob > chi2 = 0.027
(Robust, but can be weakened by many instruments.)
Difference-in-Hansen tests of exogeneity of instrument subsets:
GMM instruments for levels
Hansen test excluding group: chi2(16) = 30.31 Prob > chi2 = 0.016
Difference (null H = exogenous): chi2(3) = 2.18 Prob > chi2 = 0.536
gmm(L.winv2 wndf wdiv2, collapse lag(2 5))
Hansen test excluding group: chi2(4) = 17.36 Prob > chi2 = 0.002
Difference (null H = exogenous): chi2(15) = 15.14 Prob > chi2 = 0.442
iv(wmtb wcf wtang L.stdev cage wsalesg wsize, eq(diff))
Hansen test excluding group: chi2(12) = 14.38 Prob > chi2 = 0.277
Difference (null H = exogenous): chi2(7) = 18.12 Prob > chi2 = 0.011
.
Many thanks.
Comment