连老师：您好！我在运用系统GMM分析全国分省资本形成与金融发展动态关系时，运用一步法时的结果如下：
xtabond2 gdzb L.gdzb L(0/1).( deploa  lndp  ) , gmm(L.gdzb,lag(2 5)) iv(L(0/1).
  ( deploa  lndp  ) )

Favoring space over speed. To switch, type or click on mata: mata set matafavor s
> peed, perm.
Warning: Number of instruments may be large relative to number of observations.
Dynamic panel-data estimation, one-step system GMM
Group variable: code                            Number of obs      =       900
Time variable : year                            Number of groups   =        30
Number of instruments = 139                     Obs per group: min =        30
Wald chi2(5)  =  26540.87                                      avg =     30.00
Prob > chi2   =     0.000                                      max =        30
gdzb       Coef.   Std. Err.      z    P>z     [95% Conf. Interval]
gdzb
L1.    1.043036   .0078442   132.97   0.000     1.027661     1.05841
deploa
--.    .1061822   .0417231     2.54   0.011     .0244064    .1879581
L1.   -.1191385   .0418687    -2.85   0.004    -.2011997   -.0370773
lndp
--.     .115945   .0451242     2.57   0.010     .0275032    .2043869
L1.    -.126873   .0443236    -2.86   0.004    -.2137457   -.0400002
_cons    .0012758   .1066421     0.01   0.990    -.2077388    .2102905

Arellano-Bond test for AR(1) in first differences: z = -10.11  Pr > z =  0.000
Arellano-Bond test for AR(2) in first differences: z =  -2.08  Pr > z =  0.037
Sargan test of overid. restrictions: chi2(133)  = 434.70  Prob > chi2 =  0.000
(Not robust, but not weakened by many instruments.)
Difference-in-Sargan tests of exogeneity of instrument subsets:
GMM instruments for levels
Sargan test excluding group:     chi2(105)  = 369.84  Prob > chi2 =  0.000
Difference (null H = exogenous): chi2(28)   =  64.86  Prob > chi2 =  0.000
ivstyle(L(0/1).( deploa lndp ))
Sargan test excluding group:     chi2(129)  = 419.95  Prob > chi2 =  0.000
Difference (null H = exogenous): chi2(4)    =  14.75  Prob > chi2 =  0.005
.
以上结果中的AR(2)和Sargan 检验均不理想，但我在xtabond2语句后加 twostep，其结果较为理想。
xtabond2 gdzb L.gdzb L(0/1).( deploa  lndp  ) , gmm(L.gdzb,lag(2 5)) iv(L(0/1).
> ( deploa  lndp  ) )  twostep
Favoring space over speed. To switch, type or click on mata: mata set matafavor s
> peed, perm.
Warning: Number of instruments may be large relative to number of observations.
Warning: Two-step estimated covariance matrix of moments is singular.
  Using a generalized inverse to calculate optimal weighting matrix for two-step
> estimation.
  Difference-in-Sargan statistics may be negative.
Dynamic panel-data estimation, two-step system GMM
------------------------------------------------------------------------------
Group variable: code                            Number of obs      =       900
Time variable : year                            Number of groups   =        30
Number of instruments = 139                     Obs per group: min =        30
Wald chi2(5)  = 137124.14                                      avg =     30.00
Prob > chi2   =     0.000                                      max =        30
------------------------------------------------------------------------------
        gdzb |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
        gdzb |
         L1. |   1.038108     .01015   102.28   0.000     1.018214    1.058001
      deploa |
         --. |   .1009442   .0239268     4.22   0.000     .0540485    .1478399
         L1. |  -.1123348   .0164816    -6.82   0.000    -.1446382   -.0800314
        lndp |
         --. |   .1112644   .0177955     6.25   0.000     .0763858    .1461429
         L1. |  -.1267078   .0195746    -6.47   0.000    -.1650733   -.0883423
       _cons |   .0324825   .1346606     0.24   0.809    -.2314475    .2964125
------------------------------------------------------------------------------
Warning: Uncorrected two-step standard errors are unreliable.
Arellano-Bond test for AR(1) in first differences: z =  -4.02  Pr > z =  0.000
Arellano-Bond test for AR(2) in first differences: z =  -1.58  Pr > z =  0.115
Sargan test of overid. restrictions: chi2(133)  = 434.70  Prob > chi2 =  0.000
  (Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(133)  =  29.29  Prob > chi2 =  1.000
  (Robust, but weakened by many instruments.)
Difference-in-Hansen tests of exogeneity of instrument subsets:
  GMM instruments for levels
    Hansen test excluding group:     chi2(105)  =  29.29  Prob > chi2 =  1.000
    Difference (null H = exogenous): chi2(28)   =   0.00  Prob > chi2 =  1.000
  ivstyle(L(0/1).( deploa lndp ))
    Hansen test excluding group:     chi2(129)  =  29.79  Prob > chi2 =  1.000
    Difference (null H = exogenous): chi2(4)    =  -0.51  Prob > chi2 =  1.000

以上一步法结果和两步伐结果出现明显差别，我究竟以哪一个结果为准？理由如何解释？特请教连老师！！
多谢！！