Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Lola Arieza <lola.arieza@yahoo.com> |
To | "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu> |
Subject | st: Extabond2 twostep GMM |
Date | Thu, 15 Aug 2013 13:12:58 +0100 |
Dear Statalist, I am currently using extabond2 to estimate a two step system GMM dynamic panel. The regression results are fine with no autocorrelation in second order and the result from sargan test indicates that the instruments are valid. However when i include the corrected standard error (twostep robust), many of the coefficient estimate turns to be not significant, while the results still detect no autocorrelation in second order and the Sargan test shows valid instrument. Thereby, is this mean that my first GMM regresion (without robust) is not valid? Does using the result from the second one (with robust) is the only way in this case? Thanks in advance for any assistance. Best regards, Lola Arieza * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/faqs/resources/statalist-faq/ * http://www.ats.ucla.edu/stat/stata/