i have unbalanced long heterogenous financial panel data with few endogenous variables and no time in variant variables, is producing in significant results in "system gmm model" via xtabond2 (one step) command in stata. How to handle it? How to decide about the inclusion of lags in model?
*However if tired two step gmm resulting in omitted variables*
Can I use xtregar depvar indepvari
Instead of xtabond 2 ??
As it is producing significant results
please guide
*However if tired two step gmm resulting in omitted variables*
Can I use xtregar depvar indepvari
Instead of xtabond 2 ??
As it is producing significant results
please guide
