各位大神,你们好!
我在做GMM估计时,出现了以下问题:
命令:
xtdpdsys ME PC1S PC1SDIR PC1WDIR LIQ ROA ES RA year2-year15 ,lags(3) maxldep(5) endogenous(PC1S,lag(0,1)) endogenous(PC1SDIR,lag(0,1)) endogenous(PC1WDIR,lag(0,1)) twostep
estat abond
estat sargan
. estat abond
Arellano-Bond test for zero autocorrelation in first-differenced errors
+-----------------------+
|Order | z Prob > z|
|------+----------------|
| 1 |-6.5355 0.0000 |
| 2 | .56879 0.5695 |
+-----------------------+
H0: no autocorrelation
. estat sargan
Sargan test of overidentifying restrictions
H0: overidentifying restrictions are valid
chi2(117) = 141.5355
Prob > chi2 = 0.0610
xtdpdsys ME PC1S PC1SDIR PC1WDIR LIQ ROA ES RA year2-year15 ,lags(3) maxldep(5) endogenous(PC1S,lag(0,1)) endogenous(PC1SDIR,lag(0,1)) endogenous(PC1WDIR,lag(0,1)) twostep vce(robust)
其中PC1S为虚拟变量,其他都为连续型变量
> 1)) twostep vce(robust)
note: year2 dropped from div() because of collinearity
note: year15 dropped from div() because of collinearity
note: PC1S dropped because of collinearity
note: PC1SDIR dropped because of collinearity
note: PC1WDIR dropped because of collinearity
note: year2 dropped because of collinearity
note: year3 dropped because of collinearity
note: year4 dropped because of collinearity
System dynamic panel-data estimation Number of obs = 7508
Group variable: stock Number of groups = 1230
Time variable: year
Obs per group: min = 1
avg = 6.104065
max = 12
Number of instruments = 139 Wald chi2(21) = 679.77
Prob > chi2 = 0.0000
Two-step results
------------------------------------------------------------------------------
| WC-Robust
ME | Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
ME |
L1. | .408681 .0429583 9.51 0.000 .3244843 .4928778
L2. | .0099695 .0264797 0.38 0.707 -.0419298 .0618687
L3. | .0609598 .0204079 2.99 0.003 .020961 .1009586
|
PC1S | .0424685 .0150798 2.82 0.005 .0129126 .0720244
PC1SDIR | .016679 .0098087 1.70 0.089 -.0025456 .0359036
PC1WDIR | .0058796 .0049877 1.18 0.238 -.0038961 .0156553
LIQ | -.009085 .019578 -0.46 0.643 -.0474573 .0292872
ROA | -.4542269 .0513951 -8.84 0.000 -.5549594 -.3534944
ES | 1.294591 .2956727 4.38 0.000 .7150829 1.874098
RA | .0341798 .020447 1.67 0.095 -.0058955 .0742551
year5 | -.0066923 .0028612 -2.34 0.019 -.0123002 -.0010845
year6 | -.0065217 .0032774 -1.99 0.047 -.0129453 -.000098
year7 | -.0115055 .0034675 -3.32 0.001 -.0183017 -.0047094
year8 | -.0174932 .0035306 -4.95 0.000 -.0244132 -.0105733
year9 | -.0153215 .0038421 -3.99 0.000 -.0228518 -.0077913
year10 | -.0020231 .0038291 -0.53 0.597 -.009528 .0054818
year11 | -.0088168 .003955 -2.23 0.026 -.0165685 -.001065
year12 | -.0101775 .0040131 -2.54 0.011 -.0180431 -.002312
year13 | -.0047365 .0039604 -1.20 0.232 -.0124988 .0030257
year14 | -.0057678 .0040465 -1.43 0.154 -.0136987 .0021632
year15 | -.0020408 .0042474 -0.48 0.631 -.0103656 .006284
_cons | .0617105 .0177883 3.47 0.001 .026846 .096575
------------------------------------------------------------
请问各位大神,其中虚拟变量PC1S 上面已经drop,为什么下面还有相关系数?这种情况可以忽略,如果不可以忽略,该如何处理呢?