Clarity required on Weighting matrix
Hello
I am performing a 2step GMM estimation (General Method of Moments) and I require assistance on how to specify the initial weighting matrix. I am using the software program Stata I am using allows me to enter it in manually. Step 1 of a twostep GMM makes use of an identity matrix to compute a preliminary GMM estimate. This estimator will be consistent but not efficient. Step 2 is then to take the estimate we received from step 1 and plug it into a weighting matrix. This matrix converges in probability and therefore if we compute using this weighting matrix, the estimator will be asymptotically efficient. My problem is going from step 1 to step 2. I have the step 2 weighting matrix formula but I do not know how to put it in to use. Can anyone help me?
Rgds
Tyrone
