Dear Statalisters,
I am trying to run a GMM model on a pooled OLS data (420 individuals
over 22 years. After running some tests, it seems that I need to
correct for with a Newey-West AR(1) and heteroscedasticity
corrections.
Hence, I tried the following syntax:
ivreg2 $paper (lreppaperg lrecpaperg dbrcapaper = l.lrgdp95x
l.lrgdp95m l.rcapaperx l.lreppaperg), bw(2) gmm
...Stata gives me a couple of lines:
Warning - collinearities detected
Vars dropped: d99
...and then the following error:
Error: estimated covariance matrix of moment conditions not of full
rank;
cannot calculate optimal weighting matrix for GMM estimation.
Possible causes:
estimated covariance matrix of moment conditions not positive
definite
estimated covariance matrix uses too many lags
singleton dummy variable (dummy with 1 one and N-1 zeros or visa-
versa)
r(498);
Has anyone got any suggestions on other ways to correct for serial
correlation and heteroscedasticity when this problem arises?
Many thanks in advance!
Andrea
Andrea Molinari
PhD in Economics student
University of Sussex
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
