Questions about GMM

[Salma Raboudi]

Hello,
I'm a beginner in economitrics and statistics and I'm working on system GMM, sometimes I found that all variables are significants, and sometimes not. I want to learn how can I choose the right Instrument variables.
Is it right to choose the same exogenous variables as instrumental variables ?
Thanks a lot

[Sebastian Kripfganz]

Selecting instruments based on the significance of the results is usually not an acceptable approach. That is basically p-hacking. Exogenous variables can be their own instruments (assuming that they are exogenous with respect to both the unobserved fixed effects and the idiosyncratic error component).

You might find my 2019 London Stata Conference presentation useful:

• Kripfganz, S. (2019). Generalized method of moments estimation of linear dynamic panel data models. Proceedings of the 2019 London Stata Conference.

[Salma Raboudi]

thank you