Good day all,
I'm running a dynamic panel regression with results shown below. I'm wondering if my results are good if the p-value of my Hansen test of overid. restrictions is greater than 0.05 but the instruments are greater than the number of groups. Unfortunately, increasing the number of groups is unlikely and limiting the lags results to a Hansen test p-value of < 0.05.
Thank you all!
Results:
. xtabond2 logNPLR l.logNPLR l(0/1).(BO PU) GDP_G INF FX l(0/1).(EA ROE LA LOANS) yr*, gmm(l.(l
> ogNPLR BO PU), laglimits(2 4)) iv(GDP_G INF FX l(0/1).(EA ROE LA LOANS) yr*) robust nolevel s
> mall
Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm.
Warning: Number of instruments may be large relative to number of observations.
Warning: Two-step estimated covariance matrix of moments is singular.
Using a generalized inverse to calculate robust weighting matrix for Hansen test.
Difference-in-Sargan/Hansen statistics may be negative.
Dynamic panel-data estimation, one-step difference GMM
------------------------------------------------------------------------------
Group variable: BANKID Number of obs = 559
Time variable : YEAR Number of groups = 43
Number of instruments = 57 Obs per group: min = 13
F(32, 43) = 80.94 avg = 13.00
Prob > F = 0.000 max = 13
------------------------------------------------------------------------------
| Robust
logNPLR | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
logNPLR |
L1. | .6504 .0849371 7.66 0.000 .479108 .821692
|
BO |
--. | -.0420058 .0192128 -2.19 0.034 -.0807521 -.0032595
L1. | .0270374 .0096854 2.79 0.008 .0075049 .0465698
|
PU |
--. | -.0364955 .0183151 -1.99 0.053 -.0734314 .0004404
L1. | .0160504 .0079666 2.01 0.050 -.0000158 .0321167
|
GDP_G | -.0359647 .015819 -2.27 0.028 -.0678667 -.0040628
INF | .0297406 .00948 3.14 0.003 .0106224 .0488588
FX | -.0022935 .0009664 -2.37 0.022 -.0042425 -.0003446
|
EA |
--. | .032746 .025654 1.28 0.209 -.0189902 .0844821
L1. | -.0206623 .0285424 -0.72 0.473 -.0782235 .0368989
|
ROE |
--. | -.0044808 .0022454 -2.00 0.052 -.0090091 .0000475
L1. | -.0002462 .0011608 -0.21 0.833 -.0025872 .0020948
|
LA |
--. | -.0053512 .0053306 -1.00 0.321 -.0161014 .005399
L1. | .0055195 .003064 1.80 0.079 -.0006596 .0116985
|
LOANS |
--. | -.0012222 .0022696 -0.54 0.593 -.0057993 .0033549
L1. | .0017062 .0008099 2.11 0.041 .000073 .0033394
|
yr2002 | 0 (omitted)
yr2003 | 0 (omitted)
yr2004 | 0 (omitted)
yr2005 | 0 (omitted)
yr2006 | .1009978 .0523326 1.93 0.060 -.004541 .2065366
yr2007 | .1291074 .0415814 3.10 0.003 .0452505 .2129642
yr2008 | 0 (omitted)
yr2009 | 0 (omitted)
yr2010 | 0 (omitted)
yr2011 | 0 (omitted)
yr2012 | .0991219 .0478417 2.07 0.044 .0026398 .1956039
yr2013 | .0094751 .0227914 0.42 0.680 -.0364881 .0554383
yr2014 | 0 (omitted)
yr2015 | 0 (omitted)
yr2016 | -.0298519 .0340483 -0.88 0.385 -.0985168 .0388131
yr2017 | -.1474086 .0388333 -3.80 0.000 -.2257235 -.0690937
------------------------------------------------------------------------------
Instruments for first differences equation
Standard
D.(GDP_G INF FX EA L.EA ROE L.ROE LA L.LA LOANS L.LOANS yr2002 yr2003
yr2004 yr2005 yr2006 yr2007 yr2008 yr2009 yr2010 yr2011 yr2012 yr2013
yr2014 yr2015 yr2016 yr2017)
GMM-type (missing=0, separate instruments for each period unless collapsed)
L(2/4).(L.logNPLR L.BO L.PU)
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z = -2.57 Pr > z = 0.010
Arellano-Bond test for AR(2) in first differences: z = 1.21 Pr > z = 0.227
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(25) = 153.66 Prob > chi2 = 0.000
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(25) = 26.63 Prob > chi2 = 0.375
(Robust, but weakened by many instruments.)
Difference-in-Hansen tests of exogeneity of instrument subsets:
iv(GDP_G INF FX EA L.EA ROE L.ROE LA L.LA LOANS L.LOANS yr2002 yr2003 yr2004 yr2005 yr2006 yr
> 2007 yr2008 yr2009 yr2010 yr2011 yr2012 yr2013 yr2014 yr2015 yr2016 yr2017)
Hansen test excluding group: chi2(13) = 28.21 Prob > chi2 = 0.008
Difference (null H = exogenous): chi2(12) = -1.58 Prob > chi2 = 1.000
I'm running a dynamic panel regression with results shown below. I'm wondering if my results are good if the p-value of my Hansen test of overid. restrictions is greater than 0.05 but the instruments are greater than the number of groups. Unfortunately, increasing the number of groups is unlikely and limiting the lags results to a Hansen test p-value of < 0.05.
Thank you all!
Results:
. xtabond2 logNPLR l.logNPLR l(0/1).(BO PU) GDP_G INF FX l(0/1).(EA ROE LA LOANS) yr*, gmm(l.(l
> ogNPLR BO PU), laglimits(2 4)) iv(GDP_G INF FX l(0/1).(EA ROE LA LOANS) yr*) robust nolevel s
> mall
Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, perm.
Warning: Number of instruments may be large relative to number of observations.
Warning: Two-step estimated covariance matrix of moments is singular.
Using a generalized inverse to calculate robust weighting matrix for Hansen test.
Difference-in-Sargan/Hansen statistics may be negative.
Dynamic panel-data estimation, one-step difference GMM
------------------------------------------------------------------------------
Group variable: BANKID Number of obs = 559
Time variable : YEAR Number of groups = 43
Number of instruments = 57 Obs per group: min = 13
F(32, 43) = 80.94 avg = 13.00
Prob > F = 0.000 max = 13
------------------------------------------------------------------------------
| Robust
logNPLR | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
logNPLR |
L1. | .6504 .0849371 7.66 0.000 .479108 .821692
|
BO |
--. | -.0420058 .0192128 -2.19 0.034 -.0807521 -.0032595
L1. | .0270374 .0096854 2.79 0.008 .0075049 .0465698
|
PU |
--. | -.0364955 .0183151 -1.99 0.053 -.0734314 .0004404
L1. | .0160504 .0079666 2.01 0.050 -.0000158 .0321167
|
GDP_G | -.0359647 .015819 -2.27 0.028 -.0678667 -.0040628
INF | .0297406 .00948 3.14 0.003 .0106224 .0488588
FX | -.0022935 .0009664 -2.37 0.022 -.0042425 -.0003446
|
EA |
--. | .032746 .025654 1.28 0.209 -.0189902 .0844821
L1. | -.0206623 .0285424 -0.72 0.473 -.0782235 .0368989
|
ROE |
--. | -.0044808 .0022454 -2.00 0.052 -.0090091 .0000475
L1. | -.0002462 .0011608 -0.21 0.833 -.0025872 .0020948
|
LA |
--. | -.0053512 .0053306 -1.00 0.321 -.0161014 .005399
L1. | .0055195 .003064 1.80 0.079 -.0006596 .0116985
|
LOANS |
--. | -.0012222 .0022696 -0.54 0.593 -.0057993 .0033549
L1. | .0017062 .0008099 2.11 0.041 .000073 .0033394
|
yr2002 | 0 (omitted)
yr2003 | 0 (omitted)
yr2004 | 0 (omitted)
yr2005 | 0 (omitted)
yr2006 | .1009978 .0523326 1.93 0.060 -.004541 .2065366
yr2007 | .1291074 .0415814 3.10 0.003 .0452505 .2129642
yr2008 | 0 (omitted)
yr2009 | 0 (omitted)
yr2010 | 0 (omitted)
yr2011 | 0 (omitted)
yr2012 | .0991219 .0478417 2.07 0.044 .0026398 .1956039
yr2013 | .0094751 .0227914 0.42 0.680 -.0364881 .0554383
yr2014 | 0 (omitted)
yr2015 | 0 (omitted)
yr2016 | -.0298519 .0340483 -0.88 0.385 -.0985168 .0388131
yr2017 | -.1474086 .0388333 -3.80 0.000 -.2257235 -.0690937
------------------------------------------------------------------------------
Instruments for first differences equation
Standard
D.(GDP_G INF FX EA L.EA ROE L.ROE LA L.LA LOANS L.LOANS yr2002 yr2003
yr2004 yr2005 yr2006 yr2007 yr2008 yr2009 yr2010 yr2011 yr2012 yr2013
yr2014 yr2015 yr2016 yr2017)
GMM-type (missing=0, separate instruments for each period unless collapsed)
L(2/4).(L.logNPLR L.BO L.PU)
------------------------------------------------------------------------------
Arellano-Bond test for AR(1) in first differences: z = -2.57 Pr > z = 0.010
Arellano-Bond test for AR(2) in first differences: z = 1.21 Pr > z = 0.227
------------------------------------------------------------------------------
Sargan test of overid. restrictions: chi2(25) = 153.66 Prob > chi2 = 0.000
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(25) = 26.63 Prob > chi2 = 0.375
(Robust, but weakened by many instruments.)
Difference-in-Hansen tests of exogeneity of instrument subsets:
iv(GDP_G INF FX EA L.EA ROE L.ROE LA L.LA LOANS L.LOANS yr2002 yr2003 yr2004 yr2005 yr2006 yr
> 2007 yr2008 yr2009 yr2010 yr2011 yr2012 yr2013 yr2014 yr2015 yr2016 yr2017)
Hansen test excluding group: chi2(13) = 28.21 Prob > chi2 = 0.008
Difference (null H = exogenous): chi2(12) = -1.58 Prob > chi2 = 1.000
