Dear Statalisters,
I have doubts regarding validity of my GMM estimation, as I am encountering consistent rejection of null hypothesis (mostly for gmmstyle() instruments), for "Hansen test excluding group". Other postestimation tests like AR(2), p-value in Hansen overidentifying restriction tests, are more or less in order. How can I rectify my model mispecification issue without the group of specified instruments? My model tries to estimate bank risk using different bank-specific and macroeconomic variables. Given the persistence of risk and endogeneity issues of bank-specific variables, as documented in literature, I've employed a two-step system GMM estimation. Variables are as follows.
Dependent: StrScore (a measure of banking risk)
Endogenous Regressors: NIM, CRAR, ContLiab, CorpLoan, OpExpOpRev, ROA
Strictly Exogenous Regressors: PCR, Size, Pub_Dummy ("0" if Private, "1" if Public), GDPGr, GsecYld, CMR, EPUInd, CPInfl, ExcUSD and Year dummies
Attached below are the code and results of one of my many trials. Any inputs would be greatly appreciated.
I would urge upon the experts to kindly take a look at my issue and provide their suggestion.
Many thanks and regards
pankaj
I have doubts regarding validity of my GMM estimation, as I am encountering consistent rejection of null hypothesis (mostly for gmmstyle() instruments), for "Hansen test excluding group". Other postestimation tests like AR(2), p-value in Hansen overidentifying restriction tests, are more or less in order. How can I rectify my model mispecification issue without the group of specified instruments? My model tries to estimate bank risk using different bank-specific and macroeconomic variables. Given the persistence of risk and endogeneity issues of bank-specific variables, as documented in literature, I've employed a two-step system GMM estimation. Variables are as follows.
Dependent: StrScore (a measure of banking risk)
Endogenous Regressors: NIM, CRAR, ContLiab, CorpLoan, OpExpOpRev, ROA
Strictly Exogenous Regressors: PCR, Size, Pub_Dummy ("0" if Private, "1" if Public), GDPGr, GsecYld, CMR, EPUInd, CPInfl, ExcUSD and Year dummies
Attached below are the code and results of one of my many trials. Any inputs would be greatly appreciated.
Code:
xtabond2 ln_StrsScore L.ln_StrsScore L.Pub_Dummy L.CRAR L.GNPA L.PCR L.NIM L.CorpLoan L.ContLiab L.OpExpOpRev L.Size
> L.ROA GDPG GsecYld CMR EPUInd CPInfl ExcUSD, gmmstyle(ln_StrsScore, lag(2 4) collapse) gmmstyle(NIM CRAR ContLiab O
> pExpOpRev ROA L.GNPA, lag(2 2) collapse) ivstyle(Year2-Year18 L.Pub_Dummy L.PCR L.Size GDPGr GsecYld CMR EPUInd CPIn
> fl ExcUSD) twostep robust
Favoring speed over space. To switch, type or click on mata: mata set matafavor space, perm.
Warning: Two-step estimated covariance matrix of moments is singular.
Using a generalized inverse to calculate optimal weighting matrix for two-step estimation.
Difference-in-Sargan/Hansen statistics may be negative.
Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: BankID Number of obs = 643
Time variable : Year Number of groups = 39
Number of instruments = 36 Obs per group: min = 14
Wald chi2(17) = 4.42e+06 avg = 16.49
Prob > chi2 = 0.000 max = 17
------------------------------------------------------------------------------
| Corrected
ln_StrsScore | Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
ln_StrsScore |
L1. | .6698829 .1021456 6.56 0.000 .4696812 .8700846
|
Pub_Dummy |
L1. | .0787349 .0409215 1.92 0.054 -.0014698 .1589396
|
CRAR |
L1. | .4747118 .3258154 1.46 0.145 -.1638745 1.113298
|
GNPA |
L1. | -.2653445 .3849624 -0.69 0.491 -1.019857 .4891679
|
PCR |
L1. | .0523373 .0464587 1.13 0.260 -.03872 .1433946
|
NIM |
L1. | -4.540326 3.030449 -1.50 0.134 -10.4799 1.399245
|
CorpLoan |
L1. | -2.885826 1.454889 -1.98 0.047 -5.737356 -.0342955
|
ContLiab |
L1. | .037948 .0254816 1.49 0.136 -.011995 .087891
|
OpExpOpRev |
L1. | 1.468971 .2623539 5.60 0.000 .9547669 1.983175
|
Size |
L1. | .2594649 .1118172 2.32 0.020 .0403073 .4786226
|
ROA |
L1. | 4.266033 2.007027 2.13 0.034 .3323327 8.199733
|
GDPGr | -.6350152 .3527142 -1.80 0.072 -1.326322 .056292
GsecYld | 12.61979 1.724086 7.32 0.000 9.240642 15.99893
CMR | -.957872 1.144962 -0.84 0.403 -3.201956 1.286212
EPUInd | -.0004786 .0004092 -1.17 0.242 -.0012807 .0003234
CPInfl | .5630867 .552654 1.02 0.308 -.5200953 1.646269
ExcUSD | -.0003239 .0017414 -0.19 0.852 -.003737 .0030892
_cons | -.7525114 .8461097 -0.89 0.374 -2.410856 .9058331
------------------------------------------------------------------------------
Instruments for first differences equation
Standard
D.(Year2 Year3 Year4 Year5 Year6 Year7 Year8 Year9 Year10 Year11 Year12
Year13 Year14 Year15 Year16 Year17 Year18 L.Pub_Dummy L.PCR L.Size GDPGr
GsecYld CMR EPUInd CPInfl ExcUSD)
GMM-type (missing=0, separate instruments for each period unless collapsed)
L2.(NIM CRAR ContLiab OpExpOpRev ROA L.GNPA) collapsed
L(2/4).ln_StrsScore collapsed
Instruments for levels equation
Standard
Year2 Year3 Year4 Year5 Year6 Year7 Year8 Year9 Year10 Year11 Year12
Year13 Year14 Year15 Year16 Year17 Year18 L.Pub_Dummy L.PCR L.Size GDPGr
GsecYld CMR EPUInd CPInfl ExcUSD
_cons
GMM-type (missing=0, separate instruments for each period unless collapsed)
DL.(NIM CRAR ContLiab OpExpOpRev ROA L.GNPA) collapsed
DL.ln_StrsScore collapsed
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z = -3.31 Pr > z = 0.001
Arellano-Bond test for AR(2) in first differences: z = -0.96 Pr > z = 0.337
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(18) = 44.48 Prob > chi2 = 0.000
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(18) = 22.02 Prob > chi2 = 0.231
(Robust, but weakened by many instruments.)
Difference-in-Hansen tests of exogeneity of instrument subsets:
GMM instruments for levels
Hansen test excluding group: chi2(11) = 14.66 Prob > chi2 = 0.199
Difference (null H = exogenous): chi2(7) = 7.36 Prob > chi2 = 0.392
gmm(ln_StrsScore, collapse lag(2 4))
Hansen test excluding group: chi2(14) = 21.12 Prob > chi2 = 0.099
Difference (null H = exogenous): chi2(4) = 0.90 Prob > chi2 = 0.925
gmm(NIM CRAR ContLiab OpExpOpRev ROA L.GNPA, collapse lag(2 2))
Hansen test excluding group: chi2(6) = 12.88 Prob > chi2 = 0.045
Difference (null H = exogenous): chi2(12) = 9.14 Prob > chi2 = 0.691
Many thanks and regards
pankaj
