With a lot of trepidation I am starting a topic that has been addressed many times in Satatalist (Perhaps I should have posting in an existing one?)
I have gone through the very helpful posts (including the ones mentioned below) and they have helped but I have still not been able to solve my issue
https://www.statalist.org/forums/for...tests-xtabond2)
https://www.statalist.org/forums/for...difference-gmm

I am estimating the system GMM using xtabond2. My dep variable is yield (WHEAT_YIELD ) which I regress on own price (PRICE), substitutes price (LRM_PR LCPEA_PR), Input price (LNPK_PR LMWAGE_IPO), road (LROAD), irrigation (irrw), literacy (LLIT_RU) rainfall (SZRABIRF) and temp (RABIHW). My T is 46 and N is 311.
The code I run is the following:

xtabond2 WHEAT_YIELD p1-p46 L.WHEAT_YIELD L2.WHEAT_YIELD L.LWHEAT_PR L.LRM_PR L.LCPEA_PR LNPK_PR LMWAGE_IPO LROAD irrw LLIT_RU SZRABIRF RABIHW , gmm(L.WHEAT_YIELD L.LWHEAT_PR L.LRM_PR L.LCPEA_PR LNPK_PR LMWAGE_IPO LROAD L.irrw LLIT_RU, eq(diff) laglimits(3 8)collapse) gmm(L.WHEAT_YIELD L.LWHEAT_PR L.LRM_PR L.LCPEA_PR LNPK_PR LMWAGE_IPO LROAD L.irrw LLIT_RU , eq(level)laglimits(4 6)) iv(SZRABIRF RABIHW, equation(level)) twostep robust small
My issue is the value of my AR(2) test as I cannot reject the null hypothesis of no autocorrelation. As I understood the previous posts suggested to include lags of dep variable, include time dummies and try deeper lags (Sebastian Kripfganz and David Roodman) however i still cant reject it. also the value of the Hansen test is unacceptable (1.00). I understand that there is mispecification in the model but cant see how to rectify it.

Any help would be appreciated

Array