In my regression, dependent(y) variable is inj_hour
independent (x) variable is dummy_recessions_worked
Instrument variable(z) is dummy_MBA
(1) Between 2SLS and GMM, how do I know which test is more fits in the regression using my sample?
I'm not sure whether I should GMM or 2SLS for my regression using instrument variable (z), which is dummy_mba
For this, should I check whether it is overidentified or not first? If so, how can I do it?Can I use these control variables in the GMM as well?
(2) Are the codes below right to test gmm with controls?
As in OLS, I used additional 3 groups of control variables of firm_controls, firm_level_controls, executive_controls .
global firm_controls " log_asset_ debt_asset_ market_to_book_ ppe_asset_ sale_asset_ capex_asset_ cash_asset_ cashflow_asset_ div_asset_ roa_"
global firm_level_controls " firm_hour_emp "
global executive_controls "AGE sex "
-------------------
ivreg2 firm_dafw_hour (dummy_co_recession_worked = dummy_mba) $firm_controls $firm_level_controls $executive_controls , endog(dummy_co_recession_worked)
OR
ivreg2 firm_inj_hour $firm_controls $firm_level_controls $executive_controls (dummy_co_recession_worked = dummy_mba), gmm2s
--------------------
(3) If I want to add fixed effects of Year and ID in GMM test, how should I change the code?
(4) How can I interpret the result from endogeneity test (endog(dummy_co_recession_worked))?
(5) How can I export the ivreg2 results to excel as a table using outreg2?
ivreg2 firm_inj_hour $firm_controls $firm_level_controls $executive_controls (dummy_co_recession_worked = dummy_mba), endog(dummy_co_recessi
> on_worked)
IV (2SLS) estimation
--------------------
Estimates efficient for homoskedasticity only
Statistics consistent for homoskedasticity only
Number of obs = 36728
F( 14, 36713) = 315.49
Prob > F = 0.0000
Total (centered) SS = 512501.6218 Centered R2 = 0.0855
Total (uncentered) SS = 1659870.293 Uncentered R2 = 0.7176
Residual SS = 468665.3791 Root MSE = 3.572
-------------------------------------------------------------------------------------------
firm_inj_hour | Coefficient Std. err. z P>|z| [95% conf. interval]
--------------------------+----------------------------------------------------------------
dummy_co_recession_worked | -5.592068 3.096943 -1.81 0.071 -11.66197 .4778298
log_asset_ | -.040547 .0412214 -0.98 0.325 -.1213394 .0402454
debt_asset_ | .7757439 .4659329 1.66 0.096 -.1374678 1.688956
market_to_book_ | .4158894 .0502584 8.28 0.000 .3173848 .514394
ppe_asset_ | .3365047 .2940274 1.14 0.252 -.2397785 .9127879
sale_asset_ | .0158751 .0470417 0.34 0.736 -.0763251 .1080752
capex_asset_ | 3.046417 1.591023 1.91 0.056 -.0719301 6.164764
cash_asset_ | .9886102 .2152142 4.59 0.000 .5667981 1.410422
cashflow_asset_ | .8910776 .5531531 1.61 0.107 -.1930826 1.975238
div_asset_ | -13.08983 1.401671 -9.34 0.000 -15.83706 -10.34261
roa_ | -4.25651 .6629458 -6.42 0.000 -5.55586 -2.95716
firm_hour_emp | -.00299 .0000506 -59.05 0.000 -.0030892 -.0028907
AGE | .0250092 .0043878 5.70 0.000 .0164093 .0336092
sex | -.0519918 .1675637 -0.31 0.756 -.3804106 .276427
_cons | 15.50359 3.295896 4.70 0.000 9.043753 21.96343
-------------------------------------------------------------------------------------------
Underidentification test (Anderson canon. corr. LM statistic): 112.800
Chi-sq(1) P-val = 0.0000
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic): 113.101
Stock-Yogo weak ID test critical values: 10% maximal IV size 16.38
15% maximal IV size 8.96
20% maximal IV size 6.66
25% maximal IV size 5.53
Source: Stock-Yogo (2005). Reproduced by permission.
------------------------------------------------------------------------------
Sargan statistic (overidentification test of all instruments): 0.000
(equation exactly identified)
-endog- option:
Endogeneity test of endogenous regressors: 3.107
Chi-sq(1) P-val = 0.0779
Regressors tested: dummy_co_recession_worked
------------------------------------------------------------------------------
Instrumented: dummy_co_recession_worked
Included instruments: log_asset_ debt_asset_ market_to_book_ ppe_asset_
sale_asset_ capex_asset_ cash_asset_ cashflow_asset_
div_asset_ roa_ firm_hour_emp AGE sex
Excluded instruments: dummy_mba
------------------------------------------------------------------------------
.
end of do-file
.
0 Response to Instrument Variable and endogeneity test interpretations
Post a Comment