Hi Everybody,
I would like to ask you about GMM code in stata. Does GMM need a unit root test before running the model? My model code is : xtabond2 diversificationh L.diversificationh stockfdi realgdpwdi grosscapitalformationwdi cpiinflation secscenrollment, gmm(diversificationh, laglimit(1 4)) iv(stockfdi realgdpwdi grosscapitalformationwdi cpiinflation secscenrollment) robust. Do I need to use all the variables in iv ( ) part as instruments? Or how can I decide for instrumental variables? Is there any test for endonegenity and exonegenity variables for GMM? Or do I follow theory?