@axltang 感谢你的回复。
关于Hausman test,我后来按照论坛建议又做了遍。
问题3: 不知这个做法及其结果能否采信(按结果是应采用FEM)。
.
hausman fixed ., sigmamore
Note: the rank of the differenced variance matrix (6) does not equal the number of coefficients being tested (10); be sure this is what you expect, or there may be problems computing the test. Examine the output of your estimators for anything unexpected and possibly consider scaling your variables so that the coefficients are on a similar scale.
---- Coefficients ----
| (b) (B) (b-B) sqrt(diag(V_b-V_B))
| fixed . Difference S.E.
-------------+----------------------------------------------------------------
y10 | .6660538 .7591058 -.093052 .0652293
slope | .4768323 .4526078 .0242245 .0136026
pie | .0009431 .0000682 .0008749 .000652
pmi | -.0015364 -.0011519 -.0003846 .0002672
stockreturn | .0067177 .0067018 .0000158 .0001351
cftoliab | -.009605 -.0071677 -.0024373 .0017526
debttoasset | .0288154 .0202052 .0086103 .0076822
roa | .0272551 .0243616 .0028935 .0039506
rating | .0236347 .0227295 .0009052 .0002618
maturity | -.0015142 .0004132 -.0019273 .0013746
------------------------------------------------------------------------------
b = consistent under Ho and Ha; obtained from xtreg
B = inconsistent under Ha, efficient under Ho; obtained from xtreg
Test: Ho: difference in coefficients not systematic
chi2(6) = (b-B)'[(V_b-V_B)^(-1)](b-B)
= 24.02
Prob>chi2 = 0.0005
(V_b-V_B is not positive definite)
问题4 看了你对FEM对dummy variable处理的提示,感觉很有帮助。之前还做过的The fixed effect least-squares dummy variable (LSDV) model, 结果中coefficients跟the fixed-effect within-group estimator (f对应FEM结果)是完全一样,可是R square差异很大。是否可以理解成LSDV和FEM之间对dummy variable(此处是soeornot这个解释变量)处理方式的差异造成的?
LSDV 中新生成的有39个dummy variables (bond2, bond3...bond40)
结果如下(后面有信息省略)
这样看来,所有结果中LSDV是最优的?