GMM model

[Ruba AB]

Dears,
I am trying to establish GMM analysis for a paper but I am unsure if it is correct. The original OLS reg model is as follows:
FP = Post*CR + CR + Post + Controls

Where,
Dependent var = FB
Indep. var = Post*CR (CR after a specific year, & the year is dummy 0,1 ), Post, CR
Endogenous var = CR
Explanatory var= shall it be the same as independent var??
Control var = SGAt1 RDIt1 CGt1 LnAget1 Sizet1 Levt1 IndSenst1 betat1 TQt1 (t1 = one lagged year)
Instrumental var= MedCR Lnage

I established the following codes and each one gives me different results, I need the overidentification tests (sargan and Hansen tests) which I couldn't obtain using the first code:
1-

gmm (TobinsQ- {xb: RegCRt1 CRt1 Regt1 lambdat1 }- {b0}), instruments(SGAt1 RDIt1 CGt1 LnAget1 Sizet1 Levt1 IndSenst1 betat1 TQt1)

2-

xtabond2 LnTobinsQ l.LnTobinsQ LnCR Reg LnSGA LnRDI LnCG LnAge_w Lnsize LnLev LnTQt1, gmm(l.LnTobinsQ LnRegCR) iv( Med_CRt1 LnAge_w) twostep robust

I read almost all related documents to those commands but I couldn't understand how it should be built correctly. Please advise how it should be built? shall I lag all the variables? the overident results should be always accepting the null to make sure that my model is correct, or it is OK if not?

Thank you in advance

[Sebastian Kripfganz]

This is not a Mata related question. You generally have a higher chance of getting a timely and helpful answer if you post in the right forum, which would be the General forum in this case.

Why do you expect to obtain the same results from the two commands? The models and estimators are vastly different. The set of regressors differs, and so does the set of instruments. The gmm help file contains some examples for fitting dynamic panel models with this command.

On slides 5 and 6 of my 2019 London Stata Conference presentation, I also provide some examples on how to replicate certain estimators with different commands, including gmm:

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

[Ruba AB]

Dear Sebastian, thank you for the reply. I will repost it on the general forum, and I am reading your article. I just want to know which one is the correct model for the given information.