Overid

[ALKEBSEE RADWAN]

Hello dear analysts
I used to use overidentifying test by running the following

ivregress gmm fraud_dummy (CEO_pay_log= local_pay_instrumental industry_pay_instrumental) num_of_CEO_degree_expert num_of_CEO_financial_expertise Fem_indep av_id_age av_id_shareholding board_size ceo_duality board_ind soe big4 roa btm1 eps1 loss_total i.year i.Sic
estat overid

it was working with me well
i got the following result

  Tests of overidentifying restrictions:
  Sargan (score) chi2(1) =  1.60561  (p = 0.2051)
  Basmann chi2(1)        =   1.5971  (p = 0.2063)

then suddenly when i run same model i get the following finding

Test of overidentifying restriction:

  Hansen's J chi2(1) = 8.5e-16 (p = 1.0000)

please some explanation
why it became like that.

[William Lisowski]

The documentation for estat overid found the output of in help ivregress postestimation tells us

estat overid performs tests of overidentifying restrictions. If the 2SLS estimator was used, Sargan's (1958) and Basmann's (1960) chi-squared tests are reported, as is Wooldridge's (1995) robust score test; if the LIML estimator was used, Anderson and Rubin's (1950) chi-squared test and Basmann's F test are reported; and if the GMM estimator was used, Hansen's (1982) J statistic chi-squared test is reported. A statistically significant test statistic always indicates that the instruments may not be valid.

So presumably your first example was run, not with ivregress gmm, but instead with ivregress 2sls .

[ALKEBSEE RADWAN]

The documentation for estat overid found the output of in help ivregress postestimation tells us



So presumably your first example was run, not with ivregress gmm, but instead with ivregress 2sls .

thank you so much
yes you are right
i just tried with 2sls, it works well

and kindly in case i used it with GMM as i did
how can i interpret the finding that i got above

  Test of overidentifying restriction:    Hansen's J chi2(1) = 8.5e-16 (p = 1.0000)

[William Lisowski]

The results of your test have a p-value of 1.0000 so the test is not significant, since 1.0000 > 0.05 or 0.01 or whatever your favorite significance level is. So the test does not indicate that the instruments may not be valid. Or put another way, the test gives no reason to worry about your instrumental variables.

[ALKEBSEE RADWAN]

The results of your test have a p-value of 1.0000 so the test is not significant, since 1.0000 > 0.05 or 0.01 or whatever your favorite significance level is. So the test does not indicate that the instruments may not be valid. Or put another way, the test gives no reason to worry about your instrumental variables.

i got it
thanks a lot