R help - Feb 2013 - GMM for dynamic mdels: what if never passes Sargan test?

Hi! I am looking for some insight with this situation: what to do or how to
analyze when our models fitted with pgmm never pass Sargant test?

With my current dataset, I've been fitting different models and with all
possible combinations of lagged instruments, with all possible lag order
combinations, but no model passes Sargan test. I can not give up gmm here
as I have autocorrelation and only dynamic specification gets rid of this
issue. I also tried using system-GMM but without success.

I would like to ask you for any suggestion on what to do in this situation.
Any readings, plm function or so on will be very welcome.

Thanks in advance

	[[alternative HTML version deleted]]