Need help on a GMM estimation

Hello to all,

I am trying to do a GMM estimation (with xtabond2) on a little sample of about 250 observation, but I cannot get a good result (in terms of t-test) of the observed variable. How can I improve these results?

The estimation was made considering other 12 variables in addition to the one which I'm interested. I tried also to insert "small" on xtabond2 command, but the result doesn't changes.

Do someone has an idea?

Thanks :wave: