请教用stata做pooled随机前沿，为什么结果不好，数据和测算过程如下。
frontier lnc lny lnpl lnpc t,cost

Iteration 0:   log likelihood = -29.253253  
Iteration 1:   log likelihood = -29.192028  (not concave)
Iteration 2:   log likelihood = -29.159278  (not concave)
Iteration 3:   log likelihood = -29.147611  (not concave)
Iteration 4:   log likelihood =  -29.14322  (not concave)
Iteration 5:   log likelihood = -29.141573  (not concave)
Iteration 6:   log likelihood = -29.140637  (not concave)
Iteration 7:   log likelihood = -29.140075  (not concave)
Iteration 8:   log likelihood = -29.139784  (not concave)
Iteration 9:   log likelihood = -29.139593  (not concave)
Iteration 10:  log likelihood = -29.139531  (not concave)
Iteration 11:  log likelihood = -29.139492  (not concave)
Iteration 12:  log likelihood = -29.139467  (not concave)
Iteration 13:  log likelihood = -29.139452  (not concave)
Iteration 14:  log likelihood = -29.139443  (not concave)
Iteration 15:  log likelihood = -29.139439  (not concave)
Iteration 16:  log likelihood = -29.139434  (not concave)
Iteration 17:  log likelihood = -29.139433  (not concave)
Iteration 18:  log likelihood = -29.139432  (not concave)
Iteration 19:  log likelihood = -29.139431  (not concave)
Iteration 20:  log likelihood = -29.139431  (not concave)
Iteration 21:  log likelihood =  -29.13943  (not concave)
Iteration 22:  log likelihood =  -29.13943  (not concave)
Iteration 23:  log likelihood =  -29.13943  (not concave)
Iteration 24:  log likelihood =  -29.13943  (not concave)
Iteration 25:  log likelihood =  -29.13943  (not concave)
Iteration 26:  log likelihood = -29.139429  (not concave)
Iteration 27:  log likelihood = -29.139429  (not concave)
Iteration 28:  log likelihood = -29.139429  (not concave)
Iteration 29:  log likelihood = -29.139429  (not concave)
Iteration 30:  log likelihood = -29.139429  (not concave)
Iteration 31:  log likelihood = -29.139429  (not concave)
Iteration 32:  log likelihood = -29.139429  (not concave)
Iteration 33:  log likelihood = -29.139429  (not concave)
Iteration 34:  log likelihood = -29.139429  (not concave)
Iteration 35:  log likelihood = -29.139429  (not concave)
Iteration 36:  log likelihood = -29.139429  (not concave)
Iteration 37:  log likelihood = -29.139429  (not concave)
Iteration 38:  log likelihood = -29.139429  (not concave)
Iteration 39:  log likelihood = -29.139429  (not concave)
Iteration 40:  log likelihood = -29.139429  (not concave)
Iteration 41:  log likelihood = -29.139429  (not concave)
Iteration 42:  log likelihood = -29.139429  

Stoc. frontier normal/half-normal model           Number of obs   =        159
                                                  Wald chi2(4)    =    2252.51
Log likelihood = -29.139429                       Prob > chi2     =     0.0000

------------------------------------------------------------------------------
         lnc |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         lny |   .8796548   .0194147    45.31   0.000     .8416026    .9177069
        lnpl |  -.0691331   .0207013    -3.34   0.001    -.1097069   -.0285593
        lnpc |   .6796639   .0453517    14.99   0.000     .5907761    .7685517
           t |  -.0722736   .0094518    -7.65   0.000    -.0907988   -.0537484
       _cons |   15.14764   2.002343     7.56   0.000     11.22312    19.07216
-------------+----------------------------------------------------------------
    /lnsig2v |  -2.471344   .1122202   -22.02   0.000    -2.691291   -2.251396
    /lnsig2u |  -17.24676   27617.34    -0.00   1.000    -54146.25    54111.75
-------------+----------------------------------------------------------------
     sigma_v |   .2906395   .0163078                      .2603716    .3244259
     sigma_u |   .0001799   2.483511                             0           .
      sigma2 |   .0844713   .0094909                      .0658695    .1030731
      lambda |   .0006188   2.484123                     -4.868173     4.86941
------------------------------------------------------------------------------
Likelihood-ratio test of sigma_u=0: chibar2(01) = 0.00   Prob>=chibar2 = 1.000