EViews.com

Hello,

I have a system of equations that I am using and there are cross equation restrictions for the system (the same coefficients appear in each equation). I read that the SUR estimation method would be best as I have no IVs for the estimation, the Eviews (8) work file says cross equation restrictions can be used. However, I just read here:

http://forums.eviews.com/viewtopic.php?f=4&t=1862

That "The seemingly unrelated refers to the fact that you have a set of equations with no apparent cross-equation restrictions, but with non-zero off-diagonals"

So I am a little confused. Can we use a SUR estimation if there are cross equation restrictions? If not, would the appropriate approach be to use the GMM estimation?

Thanks!

The discussion there was about Newey-West standard errors which is why the distinction between GMM and SUR was relevant.

The system estimator allows for contemporaneous correlation weights. This is not actually SUR anymore, but rather simply a system of equations, possibly with cross-equation restrictions, that has residuals that are contemporaneouslyl correlated. You will only need GMM if you would also like Newey-West standard errors.