st: robust estimation SUREG

[<pierrick.jan@art.admin.ch>]

Dear Stata-List Members, 

I am investigating the determinants of three indicators of corporate performance for a cross-section made of 480 firms. Due to the presence of outliers I performed in a first step a robust regression using the rreg command of Stata (rreg combines two types of robust weighting: iteratively reweighted least squares and biweight estimator). In a second step, for each of the three regressions performed, I predicted the residuals and investigated the correlation between the residuals of each of the three equations. It turned out that the residuals were correlated, which advocates for the use of a SUR model. Unfortunately I could not find any information on how to perform in Stata a robust estimation of the SUR model (i.e. on how to combine the robust approach underlying the rreg command and the SUR approach underlying the sureg command).  

Any suggestion will be greatly appreciated!

Thanks in advance!

Kind regards,  

Pierrick Jan

*
*   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/