Logit Regression Results
Dep. Variable: default No. Observations: 6623
Model: Logit Df Residuals: 6592
Method: MLE Df Model: 30
Date: Mon, 14 Feb 2022 Pseudo R-squ.: 0.7014
Time: 13:44:59 Log-Likelihood: -246.83
converged: False LL-Null: -826.49
Covariance Type: nonrobust LLR p-value: 1.021e-224
coef std err z P>|z| [0.025 0.975]
const -7.3147 2.666 -2.744 0.006 -12.540 -2.089
最新评级_AA -0.1559 0.710 -0.219 0.826 -1.548 1.237
最新评级_B 3.1253 0.347 8.998 0.000 2.445 3.806
最新评级_C 6.7123 0.430 15.622 0.000 5.870 7.554
企业性质_国有企业 -3.4180 0.733 -4.660 0.000 -4.856 -1.980
企业性质_外资企业 -2.9746 0.937 -3.175 0.001 -4.811 -1.138
企业性质_民营企业 -1.7211 0.706 -2.439 0.015 -3.104 -0.338
企业性质_集体企业 -2.8521 1.117 -2.554 0.011 -5.041 -0.663
是否上市_是 -0.5207 0.387 -1.346 0.178 -1.279 0.238
最新发债日_2014 0.0876 0.806 0.109 0.913 -1.493 1.668
最新发债日_2015 0.7734 0.808 0.957 0.338 -0.810 2.357
最新发债日_2016 2.3629 0.945 2.502 0.012 0.512 4.214
最新发债日_2017 2.1768 0.891 2.443 0.015 0.431 3.923
最新发债日_2018 2.3469 0.726 3.233 0.001 0.924 3.770
最新发债日_2019 2.4301 0.928 2.618 0.009 0.610 4.250
最新发债日_2020 2.5645 1.091 2.351 0.019 0.427 4.702
最新发债日_2021 0.1101 1.247 0.088 0.930 -2.335 2.555
最新发债日_2022 -15.2080 2480.879 -0.006 0.995 -4877.642 4847.226
政府支出/GDP 2.0418 4.677 0.437 0.662 -7.124 11.208
SHIBOR 0.5866 0.400 1.465 0.143 -0.198 1.371
波动率 0.0671 0.033 2.002 0.045 0.001 0.133
流动性 -1.482e-10 1.58e-10 -0.938 0.348 -4.58e-10 1.61e-10
房地产政策 2.9309 0.977 3.000 0.003 1.016 4.845
持有基金占比 -0.1531 0.135 -1.131 0.258 -0.418 0.112
大股东持股比例 -0.0036 0.005 -0.671 0.503 -0.014 0.007
主营业务收入(万元) 5.105e-08 2.66e-08 1.918 0.055 -1.12e-09 1.03e-07
应付账款 -3.556e-12 6.32e-12 -0.563 0.574 -1.59e-11 8.83e-12
标准券 -1.0401 1.294 -0.804 0.422 -3.577 1.496
净资产(万元) -4.853e-08 5.58e-08 -0.871 0.384 -1.58e-07 6.07e-08
现金短债比 -0.4579 0.406 -1.129 0.259 -1.253 0.337
Z -0.0895 0.045 -2.001 0.045 -0.177 -0.002


Possibly complete quasi-separation: A fraction 0.12 of observations can be
perfectly predicted. This might indicate that there is complete
quasi-separation. In this case some parameters will not be identified.