I am new to STATA and would like to receive a valuable feedback.
I did run a GMM using xtabond2 and I used Lagged dependent variable and Lagged independent variable. The code is

xtabond2 BL BL_L1 RO RO_L1 Div TANG R N L Lo De Size Age Growth GD Inf Year1 Year2 Year3 Year4 Year5 Year6 Year7 Year8,gmm(BL_L1, collapse) iv( RO RO_L1 Div TANG R N L Lo De Size Age Growth GD Inf Year1 Year2 Year3 Year4 Year5 Year6 Year7 Year8, equation(level)) nodiffsargan twostep robust small


The issue is the lagged independent variable. It gives me an opposite sign (if the main independent variable is positive and significant then the lagged independent variable is negative and significant). Although AR1 and AR2 as well as Hansen test and its P-Value all are fine. I just wanna make sure is it correct to have a contradict results between main independent and its lagged variable?



Thanks in advance