请教诸位，
用ivreg2 .......... (posi2 = L1.posi2 L2.posi2) y*, gmm2s robust cluster(ctry)

即，加上时间虚拟变量y*后，Hansen J statistic检验无法进行，呈现下面的结果：
请教一下是什么原因呢？
NB: Critical values are for Cragg-Donald F statistic and i.i.d. errors.
------------------------------------------------------------------------------
Warning: estimated covariance matrix of moment conditions not of full rank.
         overidentification statistic not reported, and standard errors and
         model tests should be interpreted with caution.
Possible causes:
         number of clusters insufficient to calculate robust covariance matrix
         singleton dummy variable (dummy with one 1 and N-1 0s or vice versa)
partial option may address problem.
------------------------------------------------------------------------------
Instrumented:         posi2
Included instruments: hc cpckna lncostlbre ctfp lnfdis institution year yr3 yr4
                      yr5 yr6 yr7 yr8 yr9 yr10 yr11 yr12
Excluded instruments: L.posi2 L2.posi2
Dropped collinear:    yr2 yr13 yr14
------------------------------------------------------------------------------