Interpretation of under identification weak identification and over identification tests in ivreg2

[Asad Rind]

Hello everyone,

I have some results and it has become very difficult for me to interpret them.

my following ivreg equation is provide some results and i think they contradict with each other.

ivreg2 art_w roa_w lev_w fix_w ct_w sg_w size_w gm_w sa_w cs_w p_roa p_lev p_fix p_ct p_sg p_size p_gm p_sa p_cs i.year ( p_art= p_idiosyncratic), cluster(gvkey)

the results are
[QUOTE][Underidentification test (Kleibergen-Paap rk LM statistic): 10.108
Chi-sq(1) P-val = 0.0015
------------------------------------------------------------------------------
Weak identification test (Cragg-Donald Wald F statistic): 75.032
(Kleibergen-Paap rk Wald F statistic): 26.082
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.
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Hansen J statistic (overidentification test of all instruments): 0.000
(equation exactly identified)
/QUOTE]

what i understand is that both underidentification and weak identification tests conclude that my instrument is neither underidentified nor a weak instrument. However, Hensen J statistic is suggesting that my model is not endogeneous and i can use simple ols regression instead of instrument!?

Am i correctly interpreting these results.?

[Simon Schillebeeckx]

I don't think this interpretation is correct. Been trying to understand the ivreg2 output as well and my understanding now is that the Hansen J statistic in this case does not provide you with any information because your model is exactly identified. This seems to happen when the number of endogenous variables = number instruments. If that's the case, the Hansen J statistic does not tell you anything about the validity of your instrument.
I am not sure about this though

[Asad Rind]

Thanks Simon Schillebeeckx for your response. Yes you are right. Hensen-J/Sargan does not tell us about the validity of the instrument when the equation is exactly identified (i.e. the number of endogeneous variables= number of instruments). I saw this in following article.

http://onlinelibrary.wiley.com/doi/1...12044/abstract