Hello,
I am building a system GMM model, using GDP growth and inflation as instruments. However, it seems that my xtabond2 command results in too many instruments, which leads to overidentification. Below is my output. I would greatly appreciate any ideas about what may be wrong and how I could improve my approach as overall I was quite confident in my specification. Hence, I'm not sure how I could reduce the number of instruments.
Any input would be dearly appreciated!
Best,
Ulla
I am building a system GMM model, using GDP growth and inflation as instruments. However, it seems that my xtabond2 command results in too many instruments, which leads to overidentification. Below is my output. I would greatly appreciate any ideas about what may be wrong and how I could improve my approach as overall I was quite confident in my specification. Hence, I'm not sure how I could reduce the number of instruments.
Any input would be dearly appreciated! Code:
. xtabond2 LOAN_GROWTH l.LOAN_GROWTH NSFR_POST SIZE LLR DEPOSITS ROAA INFLATION GDPGROWTH CAPITAL LIQ i
> .year, gmm(l.LOAN_GROWTH, lag(1 3)) gmm(GDPGROWTH INFLATION) iv(INFLATION GDPGROWTH) gmm(l.(LLR DEPOS
> ITS CAPITAL SIZE ROAA LIQ)) robust twostep
Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm.
2014b.year dropped due to collinearity
2018.year dropped due to collinearity
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: id1 Number of obs = 2955
Time variable : year Number of groups = 605
Number of instruments = 190 Obs per group: min = 1
Wald chi2(15) = 1650.93 avg = 4.88
Prob > chi2 = 0.000 max = 6
------------------------------------------------------------------------------
| Corrected
LOAN_GROWTH | Coefficient std. err. z P>|z| [95% conf. interval]
-------------+----------------------------------------------------------------
LOAN_GROWTH |
L1. | -.0202411 .0119092 -1.70 0.089 -.0435828 .0031005
|
NSFR_POST | 12.75757 2.393688 5.33 0.000 8.066024 17.44911
SIZE | .3409934 .1950072 1.75 0.080 -.0412136 .7232004
LLR | -1.767063 1.058481 -1.67 0.095 -3.841648 .3075228
DEPOSITS | .4577923 .1134956 4.03 0.000 .235345 .6802397
ROAA | .8439068 1.948948 0.43 0.665 -2.975962 4.663775
INFLATION | 1.743606 1.16189 1.50 0.133 -.5336569 4.020868
GDPGROWTH | 2.966035 .5796437 5.12 0.000 1.829954 4.102116
CAPITAL | -.2811744 .3665101 -0.77 0.443 -.999521 .4371721
LIQ | .2000795 .0938573 2.13 0.033 .0161226 .3840364
|
year |
2015 | 16.16414 1.431909 11.29 0.000 13.35765 18.97063
2016 | 11.04342 1.079505 10.23 0.000 8.927631 13.15921
2017 | 12.83542 1.113706 11.52 0.000 10.65259 15.01824
2019 | 16.38724 .9603001 17.06 0.000 14.50509 18.2694
2020 | 29.67164 2.965166 10.01 0.000 23.86002 35.48326
|
_cons | -56.87381 14.17832 -4.01 0.000 -84.66281 -29.08481
------------------------------------------------------------------------------
Instruments for first differences equation
Standard
D.(INFLATION GDPGROWTH)
GMM-type (missing=0, separate instruments for each period unless collapsed)
L(1/6).(L.LLR L.DEPOSITS L.CAPITAL L.SIZE L.ROAA L.LIQ)
L(1/6).(GDPGROWTH INFLATION)
L(1/3).L.LOAN_GROWTH
Instruments for levels equation
Standard
INFLATION GDPGROWTH
_cons
GMM-type (missing=0, separate instruments for each period unless collapsed)
D.(L.LLR L.DEPOSITS L.CAPITAL L.SIZE L.ROAA L.LIQ)
D.(GDPGROWTH INFLATION)
D.L.LOAN_GROWTH
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z = -5.71 Pr > z = 0.000
Arellano-Bond test for AR(2) in first differences: z = 0.27 Pr > z = 0.784
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(174) = 552.30 Prob > chi2 = 0.000
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(174) = 245.35 Prob > chi2 = 0.000
(Robust, but weakened by many instruments.)
Difference-in-Hansen tests of exogeneity of instrument subsets:
GMM instruments for levels
Hansen test excluding group: chi2(127) = 164.93 Prob > chi2 = 0.013
Difference (null H = exogenous): chi2(47) = 80.43 Prob > chi2 = 0.002
gmm(L.LOAN_GROWTH, lag(1 3))
Hansen test excluding group: chi2(157) = 235.07 Prob > chi2 = 0.000
Difference (null H = exogenous): chi2(17) = 10.29 Prob > chi2 = 0.891
gmm(GDPGROWTH INFLATION, lag(1 .))
Hansen test excluding group: chi2(124) = 184.78 Prob > chi2 = 0.000
Difference (null H = exogenous): chi2(50) = 60.57 Prob > chi2 = 0.145
gmm(L.LLR L.DEPOSITS L.CAPITAL L.SIZE L.ROAA L.LIQ, lag(1 .))
Hansen test excluding group: chi2(54) = 107.12 Prob > chi2 = 0.000
Difference (null H = exogenous): chi2(120) = 138.24 Prob > chi2 = 0.122
iv(INFLATION GDPGROWTH)
Hansen test excluding group: chi2(172) = 232.20 Prob > chi2 = 0.002
Difference (null H = exogenous): chi2(2) = 13.15 Prob > chi2 = 0.001
.
end of do-file
Ulla
Comment