Dear all,

I need your help regarding endogeneity.

I am running an OLS model like the following:

Dechow_DACC = a0+ a1 EARNINGS BEFORE DACC +a2 LAG DACC + a3 change in revenues + a4 change in other revenues + a5 change in financial expences + a6 other financial items + a7 SIZE + a8 LEVERAGE + year dummies.

++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++
Number of obs = 454
F( 15, 94) = 7.18
Prob > F = 0.0000
R-squared = 0.2767
Root MSE = .06621

(Std. Err. adjusted for 95 clusters in ID_prefecture)

Robust
Dechow_DACC Coef. Std. Err. t P>t [95% Conf. Interval]

EARNINGS_BEFORE_DACC | -.2260247 .0759327 -2.98 0.004 -.3767909 -.0752585
LAG_DACC | -.006433 .0444744 -0.14 0.885 -.0947378 .0818719
change_in_revenues | -.1850096 .1219579 -1.52 0.133 -.4271599 .0571408
change_in_other_revenues | -.0437916 .0248281 -1.76 0.081 -.0930884 .0055052
change_in_financial_expences | -.1495044 .0380577 -3.93 0.000 -.225069 -.0739399
other_financial_items | .0019309 .0940782 0.02 0.984 -.1848637 .1887255
size | .0007769 .003582 0.22 0.829 -.0063352 .0078891
leverage | .0742486 .0753726 0.99 0.327 -.0754054 .2239025

year
2012 | .0602698 .0201734 2.99 0.004 .020215 .1003246
2013 | .0362745 .0187222 1.94 0.056 -.0008988 .0734479
2014 | .0461442 .0175783 2.63 0.010 .0112421 .0810464
2015 | .0763397 .0216758 3.52 0.001 .0333018 .1193776
2016 | .0633845 .0190233 3.33 0.001 .0256132 .1011558
2017 | .0613977 .0207133 2.96 0.004 .020271 .1025244

_cons -.0700081 .0661986 -1.06 0.293 -.2014469 .0614307

++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++

if I use the ivreg command then my model is the following:

ivregress 2sls Dechow_DACC LAG_DACC change_in_revenues change_in_other_revenues change_in_financial_expences other_financial_items size leverage i.year (EARNINGS_BEFORE_DACC = dummy_for_EBDA ), vce (cluster ID_prefecture)


Instrumental variables (2SLS) regression Number of obs = 453
Wald chi2(15) = 102.42
Prob > chi2 = 0.0000
R-squared = 0.1787
Root MSE = .06935

(Std. Err. adjusted for 95 clusters in ID_prefecture)
---------------------------------------------------------------------------------
Robust
Dechow_DACC Coef. Std. Err. z P>|z| [95% Conf. Interval]
----------------+----------------------------------------------------------------
EARNINGS_BEFORE_DACC | -.3533502 .0630489 -5.60 0.000 -.4769237 -.2297767
LAG_DACC | .0003085 .0491827 0.01 0.995 -.0960878 .0967047
change_in_revenues | -.3713585 .2528131 -1.47 0.142 -.866863 .124146
change_in_other_revenues | -.0493394 .0223761 -2.21 0.027 -.0931957 -.005483
change_in_financial_expences | -.0731608 .0351097 -2.08 0.037 -.1419745 -.0043471
other_financial_items | .0688041 .0903238 0.76 0.446 -.1082273 .2458354
size | -.001597 .0035644 -0.45 0.654 -.0085831 .0053891
leverage | .1668156 .062739 2.66 0.008 .0438495 .2897818


2012 | -.0631525 .0218966 -2.88 0.004 -.1060691 -.020236
2013 | -.0063217 .01383 -0.46 0.648 -.0334279 .0207846
2014 | -.0194567 .0092224 -2.11 0.035 -.0375322 -.0013812
2015 | -.0228222 .0097566 -2.34 0.019 -.0419448 -.0036996
2016 | -.0029112 .0094139 -0.31 0.757 -.021362 .0155397
2017 | .0043437 .0050881 0.85 0.393 -.0056289 .0143163

_cons | .0406694 .0649287 0.63 0.531 -.0865885 .1679273
---------------------------------------------------------------------------------
Instrumented: EARNINGS BEFORE DACC
Instruments: LAG_DACC change_in_revenues change_in_other_revenues change_in_financial_expences other_financial_items size leverage 2012.year 2013.year 2014.year
2015.year 2016.year 2017.year dummy_for_EBDA

. estat endog

Tests of endogeneity
Ho: variables are exogenous

Robust regression F(1,94) = 7.05062 (p = 0.0093) <---my model has an endogeneity problem
(Adjusted for 95 clusters in ID_prefecture)
************************************************** ************************************************** ************************************************** *****

My questions are the following:
1) How can I check if there is an endogeneity problem in my model? Before starting the procedure of ivreg....
2) How can I check if I have found the correct endogenous variable? Or if there is another one?
3) In order to support that I have fixed an endogeneity problem what exactly do I need to do/present?
4) Can I define 1 endogenous variable and 1 instrumental variable only?

Thank you in advance
Best regards
Kleon