Can you help me write Two-step system GMM command ?

[Ahmed Mahmoudi]

Hello everyone,

to not take from ure time much, im trying to run Two-step system GMM and i dont know how to write the command.

My data :
Independent variable : Y
Explanatory variables : L.Y X1 X2 X3 C1 C2 C3 C4 + year dummies
Endogenous variables : X1 X2 X3
Instrumental variables : IV1 IV2 IV3 L.X1 L.X2 L.X3
Does heteroscedasticity exist : YES

Kindly write me the command

[Jared Greathouse]

https://www.google.com/url?sa=t&sour...6SH_4Q5YaiBqr9

[Sebastian Kripfganz]

The following presentation slides might help:

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

[Ahmed Mahmoudi]

Thanks for the suggestions, they are really helpful.

so if i understand correctly, my command should look like this :

xtabond2 Y L.Y X1 X2 X3 C1 C2 C3 C4 y*, gmm(X1 X2 X3, lag(1 1)) iv(C1 C2 C3 C4 IV1 IV2 IV3) robust twostep

[Sebastian Kripfganz]

You need to add instruments for the time dummies, e.g. iv(y*, eq(level)). You did not specifically include higher-order lags of the dependent variable as instruments for the lagged dependent variable. Unless IV1 IV2 IV3 are good instruments for the lagged dependent variable, your model might suffer from underidentification. If you want to continue using xtabond2, I recommend that your read Roodman's article on "how to do xtabond2" before proceeding.

[Ahmed Mahmoudi]

Thank you for the quick replay.
if i understand correctly my command should be :

xtabond2 Y.L X1 X2 X3 C1 C2 C3 C4 y*, gmm(X1 X2 X3, lag(1 1)) gmm(L.Y, lag(2 3)) iv(C1 C2 C3 C4 IV1 IV2 IV3) iv(y*, eq(level)) robust twostep small