Dear all,

I have been trying to replicate the paper Global Capital Fows and Financing Constraints by Ann E. Harrison, Inessa Love, Margaret S. McMillan to learn about system GMM. And I am stucked by the stata code. In xtabond2, there are two types of instruments, gmmstyle and ivstyle. From my comprehension, endogenous instruments go to gmmstyle, while exogeneous instruments go to ivstyle.
The regression in the paper is attached below, and it use "t-1 and t-2 lags of all of the regressors as instruments" (I quote the exact words from the paper). I guess all the lagged variables are endogeneous, so I put all of them in the gmmstyle. But I am not confident with my understanding.Am I doing right? Could you help me with it? Thank you so much!!

Here is my code:
[/code] xtabond2 IK F.IK L.IK SK L.cash cashxFDI FDI i.quarter, gmm(L.(IK SK cashK cashKxFDI FDI), lag(1 2)) nodiffsargan robust orthogonal small [/code]
Array Array