Hi, I am trying to estimate the covariance matrix of an unbalanced SUR system and I want to make the covariance matrix robust for heteroskedasticity and serial correlation of unknown kind. The system framework together with the unbalanced panel makes it quite difficult for me, in that standard correction techniques like Newey and West cannot be applied like that. Can someone give we some good, accessible documentation about how I can estimate the covariance matrix. The matrix algebra must really be good explained!
Thanks in advance! Ron . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
