先认真阅读一些相关文献。面板数据集的截面纬度应大于时间纬度（N〉T）。估计后可从残差自相关检验（要求差分残差服从一阶负相关，无二阶和更高阶的相关）和sargen/hansen检验来检验工具变量的有效性，sargan是用于同方差，hansen检验更为一般可适用于异方差。gauss程序dpd98或stata程序xtabond2是估计GMM的现成程序，但gauss程序容易出错，还是xtabond2好用一些。
xi: xtabond2 d.x dl(1/1).x dl(1/1).y, gmm( x y, lag(2 8)c )
> iv(i.year,equation(level)) noc small robust h(1)
i.year _Iyear_1980-2006 (naturally coded; _Iyear_1980 omitted)
Favoring space over speed. To switch, type or click on mata: mata set matafav
> or speed, perm.
Warning: Number of instruments may be large relative to number of observation
> s.
Warning: Two-step estimated covariance matrix of moments is singular.
Using a generalized inverse to calculate robust weighting matrix for Hansen
> test.
Difference-in-Sargan statistics may be negative.
Dynamic panel-data estimation, one-step system GMM
-----------------------------------------------------------------------------
> -
Group variable: id Number of obs = 95
> 0
Time variable : year Number of groups = 3
> 8
Number of instruments = 41 Obs per group: min = 2
> 5
F(2, 38) = 412.71 avg = 25.0
> 0
Prob > F = 0.000 max = 2
> 5
-----------------------------------------------------------------------------
> -
| Robust
D.x | Coef. Std. Err. t P>|t| [95% Conf. Interval
> ]
-------------+---------------------------------------------------------------
> -
x |
LD. | .1555062 .0066141 23.51 0.000 .1421166 .168895
> 8
y |
LD. | .034008 .0368939 0.92 0.362 -.0406797 .108695
> 8
-----------------------------------------------------------------------------
> -
Instruments for first differences equation
GMM-type (missing=0, separate instruments for each period unless collapsed)
L(2/8).(x y) collapsed
Instruments for levels equation
Standard
_Iyear_1981 _Iyear_1982 _Iyear_1983 _Iyear_1984 _Iyear_1985 _Iyear_1986
_Iyear_1987 _Iyear_1988 _Iyear_1989 _Iyear_1990 _Iyear_1991 _Iyear_1992
_Iyear_1993 _Iyear_1994 _Iyear_1995 _Iyear_1996 _Iyear_1997 _Iyear_1998
_Iyear_1999 _Iyear_2000 _Iyear_2001 _Iyear_2002 _Iyear_2003 _Iyear_2004
_Iyear_2005 _Iyear_2006
GMM-type (missing=0, separate instruments for each period unless collapsed)
DL.(x y) collapsed
-----------------------------------------------------------------------------
> -
Arellano-Bond test for AR(1) in first differences: z = -1.08 Pr > z = 0.28
> 1
Arellano-Bond test for AR(2) in first differences: z = 0.91 Pr > z = 0.36
> 1
-----------------------------------------------------------------------------
> -
Sargan test of overid. restrictions: chi2(39) = 310.42 Prob > chi2 = 0.00
> 0
(Not robust, but not weakened by many instruments.)
Hansen test of overid. restrictions: chi2(39) = 37.82 Prob > chi2 = 0.52
> 4
(Robust, but can be weakened by many instruments.)
Difference-in-Hansen tests of exogeneity of instrument subsets:
GMM instruments for levels
Hansen test excluding group: chi2(37) = 37.64 Prob > chi2 = 0.44
> 0
Difference (null H = exogenous): chi2(2) = 0.18 Prob > chi2 = 0.91
> 2
gmm(x y, collapse lag(2 8))
Hansen test excluding group: chi2(23) = 22.73 Prob > chi2 = 0.47
> 6
Difference (null H = exogenous): chi2(16) = 15.09 Prob > chi2 = 0.51
> 8
iv(_Iyear_1981 _Iyear_1982 _Iyear_1983 _Iyear_1984 _Iyear_1985 _Iyear_1986
> _Iyear_1987 _Iyear_1988 _Iyear_1989 _Iyear_1990 _Iyear_1991 _Iyear_1992 _Iy
> ear_1993 _Iyear_1994 _Iyear_1995 _Iyear_1996 _Iyear_1997 _Iyear_1998 _Iyear
> _1999 _Iyear_2000 _Iyear_2001 _Iyear_2002 _Iyear_2003 _Iyear_2004 _Iyear_20
> 05 _Iyear_2006, eq(level))
Hansen test excluding group: chi2(14) = 17.87 Prob > chi2 = 0.21
> 3
Difference (null H = exogenous): chi2(25) = 19.95 Prob > chi2 = 0.74
> 9
加好友,备注cda
京公网安备 11010802022788号
论坛法律顾问：王进律师
知识产权保护声明
免责及隐私声明
发帖
雷达卡
，拜托各位大侠
