Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • GMM estimation with sample mean condition

    I am reading a paper and try to implement its methodology in Stata. The paper says the following:

    Equations (1)-(3) depend on only seven parameters. At the same time, we have seven sample means that can be calculated from the data. The expected values of these sample moments are a function of seven unknown parameters. Therefore, they provide a basis for a GMM estimator that is just-identified.
    I understand the idea but not sure how to implement this in Stata. Stata's GMM estimator seems to use the condition E(Zu) = 0, where Z are instruments and u are regression residuals. I am not sure how to transform the sample moments are a function of unknown parameters to the E(Zu)=0 condition.
Working...
X