做单门槛模型发现单门槛成立,继而做三重门槛模型,发现存在双重门槛,那么回归分析的结果用三重门槛的回归结果还是再做一个双重门槛模型,用双重门槛的结果呢?
我做出来的三重门槛模型显示存在双重门槛,但是做双重门槛时,又显示只存在单门槛了,这到底该看哪一个的结果啊?就很愁啊,各位老师同学,求助求助求帮忙看看呀
xthreg ROA 流动比率 LEV 大股东持股比例 上市年限 outbd, rx( F ) qx( ln总资产 )
> thnum(3) trim(0.01 0.01 0.05) grid(400) bs(300 300 300)
Threshold estimator (level = 95):
-----------------------------------------------------
model | Threshold Lower Upper
-----------+-----------------------------------------
Th-1 | 21.6691 21.6386 21.6788
Th-21 | 21.7012 21.6967 21.7050
Th-22 | 21.7254 21.6634 21.7312
Th-3 | 20.3273 . .
-----------------------------------------------------
Threshold effect test (bootstrap = 300 300 300):
-------------------------------------------------------------------------------
Threshold | RSS MSE Fstat Prob Crit10 Crit5 Crit1
-----------+-------------------------------------------------------------------
Single | 3.19e+04 33.4619 35.82 0.0100 12.8533 18.3119 31.9598
Double | 3.03e+04 31.7708 50.78 0.0067 14.4608 18.5174 33.9000
Triple | 2.99e+04 31.3548 12.66 0.2433 17.0081 21.3248 31.5064
-------------------------------------------------------------------------------
Fixed-effects (within) regression Number of obs = 960
Group variable: co Number of groups = 160
R-sq: within = 0.1735 Obs per group: min = 6
between = 0.1666 avg = 6.0
overall = 0.1452 max = 6
F(9,791) = 18.45
corr(u_i, Xb) = -0.4105 Prob > F = 0.0000
------------------------------------------------------------------------------
ROA | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
流动比率 | -.0769054 .0961273 -0.80 0.424 -.2656002 .1117893
LEV | -.1886741 .0242716 -7.77 0.000 -.2363184 -.1410298
大股东持~例 | .0831966 .0337386 2.47 0.014 .0169689 .1494243
上市年限 | .1517099 .124748 1.22 0.224 -.0931665 .3965862
outbd | -3.60574 1.215842 -2.97 0.003 -5.9924 -1.219081
|
_cat#c.F |
0 | -29.80297 6.845059 -4.35 0.000 -43.2396 -16.36634
1 | -6.698993 3.088673 -2.17 0.030 -12.76196 -.6360275
2 | 83.57929 11.25001 7.43 0.000 61.4959 105.6627
3 | 3.312492 1.09183 3.03 0.002 1.169265 5.455718
|
_cons | 7.431972 2.485268 2.99 0.003 2.553472 12.31047
-------------+----------------------------------------------------------------
sigma_u | 4.1600332
sigma_e | 6.147416
rho | .31410062 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(159, 791) = 2.11 Prob > F = 0.0000
xthreg ROA 流动比率 LEV 大股东持股比例 上市年限 outbd, rx( F ) qx( ln总资产 )
> thnum(2) trim(0.01 0.05) grid(400) bs(300 300)
Threshold estimator (level = 95):
-----------------------------------------------------
model | Threshold Lower Upper
-----------+-----------------------------------------
Th-1 | 21.6691 21.6386 21.6788
Th-21 | 21.6691 21.6316 21.6788
Th-22 | 20.3273 . .
-----------------------------------------------------
Threshold effect test (bootstrap = 300 300):
-------------------------------------------------------------------------------
Threshold | RSS MSE Fstat Prob Crit10 Crit5 Crit1
-----------+-------------------------------------------------------------------
Single | 3.19e+04 33.4619 35.82 0.0000 13.0184 19.4930 24.3843
Double | 3.16e+04 33.1546 8.84 0.2233 11.6645 14.3510 20.7395
-------------------------------------------------------------------------------
Fixed-effects (within) regression Number of obs = 960
Group variable: co Number of groups = 160
R-sq: within = 0.1247 Obs per group: min = 6
between = 0.1546 avg = 6.0
overall = 0.1082 max = 6
F(8,792) = 14.10
corr(u_i, Xb) = -0.4858 Prob > F = 0.0000
------------------------------------------------------------------------------
ROA | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
流动比率 | -.0723744 .098971 -0.73 0.465 -.2666509 .121902
LEV | -.193835 .0249347 -7.77 0.000 -.2427808 -.1448891
大股东持~例 | .0865873 .0347147 2.49 0.013 .0184437 .154731
上市年限 | .2057072 .1280104 1.61 0.108 -.0455727 .456987
outbd | -3.701498 1.250424 -2.96 0.003 -6.156036 -1.24696
|
_cat#c.F |
0 | -31.02053 7.047515 -4.40 0.000 -44.85455 -17.18651
1 | -10.9646 3.391107 -3.23 0.001 -17.62122 -4.307977
2 | 3.668621 1.114021 3.29 0.001 1.481838 5.855405
|
_cons | 6.927754 2.563226 2.70 0.007 1.896235 11.95927
-------------+----------------------------------------------------------------
sigma_u | 4.3584477
sigma_e | 6.3224648
rho | .32213306 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(159, 792) = 2.02 Prob > F = 0.0000