? ============================ cov2stpm =========================
proc cov2stpm eqs ivs imask an2 ;

?  This is as same as cov2step procedure by Dr. Clint Cummins, only
?  changing name of procedure, etc. see
?  http://www.stanford.edu/~clint/tspex/group.htm#G
?  The description below is the description in cov2step procedure
?  made by Dr. Clint Cummins.

?  Procedure to compute the estimated two-step covariance matrix for 
?  GMM estimation (after first step estimation).
?  see Arellano and Bond (REStud 91 for details). 
?  Benchmarked with Steve Bond using DPD and these give estimates that
?  match his.  TSP's default GMM estimator does not match the cov2step
?  results - the default GMM is more like a 3-step estimator since it
?  does 2SLS, 3SLS, and then GMM. 
?  simple version, by Clint Cummins 9/97

?  This procedure is preferable in terms of controlling for
?  missing values.

?  eqs   - equations (number of column of mask matrix, imask)
?  ivs   - a list of instrumental variables
?  imask  - a ninst by neq matrix of zeros and ones that specifies 
?          which instruments are used for which equation. 
?  an2   - the covariance matrix estimate (returned). 

?  supres nob phi gmmovid coef;
  gmm(covoc=own,hetero,maxit=0,nolsqstart,silent,inst=ivs,mask=imask) eqs;
?  nosupres nob phi gmmovid coef;
  mat an2 = @covoc;
endproc cov2stpm ;
? ==================================================================