[Sujata Luiss]

Hello All,

Is there any way by means of which I can do weak instruments checking for multiple endogenous regressors...like in my case....as stock-yogo is giving me some values like shown below and not sure how to interpret these results
Stock-Yogo weak ID F test critical values for single endogenous regressor:
5% maximal IV relative bias 16.85
10% maximal IV relative bias 10.27
20% maximal IV relative bias 6.71
30% maximal IV relative bias 5.34
10% maximal IV size 24.58
15% maximal IV size 13.96
20% maximal IV size 10.26
25% maximal IV size 8.31

Your reply is highly appreciated.

[Carlo Lazzaro]

Sujata:
you report neither the F-test, nor the minimum eigenvalue statistic that you should have obtained from something like:

estat firststage

Hence, it is impossible to help you out.

[Sujata Luiss]

Yes..I am sorry for not doing this. Here are my results:
Minimum eigenvalue statistic = 6.48742

Critical Values # of endogenous regressors: 2
Ho: Instruments are weak # of excluded instruments: 4
---------------------------------------------------------------------
| 5% 10% 20% 30%
2SLS relative bias | 11.04 7.56 5.57 4.73
-----------------------------------+---------------------------------
| 10% 15% 20% 25%
2SLS Size of nominal 5% Wald test | 16.87 9.93 7.54 6.28
LIML Size of nominal 5% Wald test | 4.72 3.39 2.99 2.79
---------------------------------------------------------------------

[Carlo Lazzaro]

Sujata:
you can not reject the null of weak instruments for most of the 2SLS nominal 5% Wald test relative biases.

[Sujata Luiss]

Thank you so very much for this answer..can I presume that I will be able to reject null if my eigenvalue is more than 10 or something greater than 16?

[Sujata Luiss]

In relation to the last question, if I am getting something like the one given below, can I reject null?
Minimum eigenvalue statistic = 11.7795

Critical Values # of endogenous regressors: 2
Ho: Instruments are weak # of excluded instruments: 4
---------------------------------------------------------------------
| 5% 10% 20% 30%
2SLS relative bias | 11.04 7.56 5.57 4.73
-----------------------------------+---------------------------------
| 10% 15% 20% 25%
2SLS Size of nominal 5% Wald test | 16.87 9.93 7.54 6.28
LIML Size of nominal 5% Wald test | 4.72 3.39 2.99 2.79
---------------------------------------------------------------------

[Sujata Luiss]

Can you pls respond Carlo Lazzaro Prof...I am sorry but my submission deadline is near and I am stuck at this point

[Sujata Luiss]

I also tried weakiv test and below is the result: not sure whether we need to look at ARchi2(2) value and can safely reject null
----------------------------------------
Test | Statistic p-value
------+---------------------------------
CLR | stat(.) = 13.25 0.0000
K | chi2(2) = 10.74 0.0047
J | chi2(2) = 2.03 0.3616
K-J | <n.a.> 0.0058
AR | chi2(4) = 12.77 0.0124
------+---------------------------------
Wald | chi2(2) = 10.86 0.0044
----------------------------------------

[Sujata Luiss]

Sorry the results with weakiv is as below:
----------------------------------------
Test | Statistic p-value
------+---------------------------------
CLR | stat(.) = 9.22 0.0081
K | chi2(2) = 7.28 0.0262
J | chi2(2) = 10.26 0.0059
K-J | <n.a.> 0.0295
AR | chi2(4) = 17.54 0.0015
------+---------------------------------
Wald | chi2(2) = 7.27 0.0264
----------------------------------------

[Carlo Lazzaro]

Sujata:
I do not get why minimum eigenvalue statistic changed.
That said, with a minimum eigenvalue statistic=11.7795 you can reject the null of weak instruments at the 2SLS nominal 10% Wald test relative bias.
I'm not famliar with the community-contributed module (as you're kindly requested tomention it as per FAQ) -weakiv-.

[Sujata Luiss]

Thank you so very much...no the eigen value statistic did not change but I made a mistake in selecting one of the variables. However, since I cannot reject null at 5%, would this result be acceptable? I mean if I am not wrong, I think it depends on the level of bias we would like to accept...but would be curious to know what the research fraternity thinks?

[Sujata Luiss]

Also, I feel that I did something wrong with the weakiv command..so I feel it will be good for me to go with estat firststage...however, it will be great if I can get any idea from you regarding the accepting level for rejecting null

[Carlo Lazzaro]

Sujata:
most of your previous queries depends on the traditions in force in your research field.

[Sujata Luiss]

I really can't thank you enough Carlo Lazzaro Prof and Joro Kolev Prof for your helpful comments and suggestions....it worked for me.