Difference in Hansen test

[Anjali Sharma]

Hey these are results from my model on banks profitability. Can someone plz help me with interpretation of difference in hansen test?

xtabond2 roa lroa lnpa lrsb gdp wpi pubdummy, gmm(lroa, lag(2 2)) gmm(lrsb lnpa, lag(2 3) collapse) iv(i.year gdp wpi pubdummy) twostep robus

> t

Favoring space over speed. To switch, type or click on mata: mata set matafavor speed, 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 = 481

Time variable : year Number of groups = 44

Number of instruments = 38 Obs per group: min = 8

Wald chi2(6) = 185.97 avg = 10.93

Prob > chi2 = 0.000 max = 11

Corrected

roa Coef. Std. Err. z P>z [95% Conf. Interval]

lroa .4711673 .1753447 2.69 0.007 .1274979 .8148366

lnpa -.2420303 .0468653 -5.16 0.000 -.3338847 -.1501759

lrsb .0363804 .0224895 1.62 0.106 -.0076981 .0804589

gdp -.0113045 .0099002 -1.14 0.254 -.0307085 .0080995

wpi -.050594 .0125853 -4.02 0.000 -.0752607 -.0259273

pubdummy -.1873037 .0751493 -2.49 0.013 -.3345937 -.0400137

_cons 1.130676 .2324807 4.86 0.000 .6750218 1.586329

Instruments for first differences equation

Standard

D.(2005b.year 2006.year 2007.year 2008.year 2009.year 2010.year 2011.year

2012.year 2013.year 2014.year 2015.year 2016.year gdp wpi pubdummy)

GMM-type (missing=0, separate instruments for each period unless collapsed)

L(2/3).(lrsb lnpa) collapsed

L2.lroa

Instruments for levels equation

Standard

2005b.year 2006.year 2007.year 2008.year 2009.year 2010.year 2011.year

2012.year 2013.year 2014.year 2015.year 2016.year gdp wpi pubdummy

_cons

GMM-type (missing=0, separate instruments for each period unless collapsed)

DL.(lrsb lnpa) collapsed

DL.lroa

Arellano-Bond test for AR(1) in first differences: z = -2.23 Pr > z = 0.026

Arellano-Bond test for AR(2) in first differences: z = -0.03 Pr > z = 0.978

Sargan test of overid. restrictions: chi2(31) = 160.71 Prob > chi2 = 0.000

(Not robust, but not weakened by many instruments.)

Hansen test of overid. restrictions: chi2(31) = 36.77 Prob > chi2 = 0.219

(Robust, but weakened by many instruments.)

Difference-in-Hansen tests of exogeneity of instrument subsets:

GMM instruments for levels

Hansen test excluding group: chi2(20) = 24.09 Prob > chi2 = 0.239

Difference (null H = exogenous): chi2(11) = 12.68 Prob > chi2 = 0.314

gmm(lroa, lag(2 2))

Hansen test excluding group: chi2(13) = 20.52 Prob > chi2 = 0.083

Difference (null H = exogenous): chi2(18) = 16.25 Prob > chi2 = 0.575

gmm(lrsb lnpa, collapse lag(2 3))

Hansen test excluding group: chi2(25) = 33.01 Prob > chi2 = 0.131

Difference (null H = exogenous): chi2(6) = 3.76 Prob > chi2 = 0.709

iv(2005b.year 2006.year 2007.year 2008.year 2009.year 2010.year 2011.year 2012.year 2013.year 2014.year 2015.year 2016.year gdp wpi pubdummy)

Hansen test excluding group: chi2(18) = 27.54 Prob > chi2 = 0.069

Difference (null H = exogenous): chi2(13) = 9.23 Prob > chi2 = 0.755

.

[Phil Bromiley]

Let me suggest you might get a better response with a better formatted question. Look at the FAQ on asking questions - providing Stata code in code delimiters, readable Stata output, and sample data using dataex will increase your chances of a helpful answer. Also, do drop extraneous stuff. While you appear to have provided Stata code and output, it is formatted in a very unusual way.

I'm not an xtabond user so I can't help you substantively. The documentation does address the interpretation of Sargan tests. You might look at
https://www.stata.com/statalist/arch.../msg00870.html
https://www.researchgate.net/post/Sy...r_both_of_them
http://ageconsearch.umn.edu/bitstrea...art_st0030.pdf

[Dario Maimone Ansaldo Patti]

If i am correct, you would like to know the interpretation of the Hansen in difference test. It seems to me that you are using the system gmm estimator. In this case the hansen in difference test tests the null hypothesis that the additional moment conditions implied by the system-gmm are orthogonal to the error term. Said differently, when you use the system-gmm you include more instruments compared to the difference-gmm. The Hansen-in-difference test tests the validity of those additional instruments. The null hypothesis is that the additional instruments are valid.