st: Estimating Simultaneous Equations Models with GMM 3SLS estimator

[Helene Ehrhart <Helene.Ehrhart@u-clermont1.fr>]

Hello,

I am estimating a two-equation linear simultaneous equations model  
where the 'exogenous' variable in one equation might be correlated  
with the error in the other equation.
To estimate the model I used traditional 3SLS with the reg3 command in  
Stata. However, I read in Woolridge' Manual (Econometric Analysis of  
Cross Section and Panel Data, chapter 9.6) that
"the traditional 3SLS estimator is generally valid only when  
instruments are uncorrelated with all errors". It is the GMM 3SLS  
estimator that must rather be used to produce consistent estimations.

In the reg3 command there is no GMM option, so I am wondering which  
command I should use to estimate my simultaneous equation model with  
the GMM 3SLS estimator.

Thank you.

Helene Ehrhart
CERDI, France.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/