Skip to main content

Table 5 T-test

From: An innovative machine learning workflow to research China’s systemic financial crisis with SHAP value and Shapley regression

Variables

Mean of non-risk period

Standard deviation for non-risk periods

Mean of risk period

Standard deviation for risk periods

T-test and significance

P value

VIF

DidYield

0.04

0.02

0.04

0.02

 − 2.00**

0.048

1.46

Credit

64.12

31.78

80.92

37.05

 − 4.20***

0.000

1.57

CPI-Dif

178.12

188.60

190.20

234.54

 − 0.48

0.633

1.58

Credit-Dif

1.06

3.35

3.42

5.31

 − 4.17***

0.000

1.29

CreditCostGrowth

8.51

19.01

17.52

17.65

 − 4.38***

0.000

1.36

NarrowMoney

21.38

17.12

19.98

15.25

0.76

0.447

1.31

NetExport

 − 0.96

5.06

 − 3.00

7.07

2.69***

0.008

1.25

FinancialConcentration

19.87

10.18

16.72

11.11

2.82***

0.005

1.76

YieldCurveSlope

 − 0.83

1.72

0.21

1.42

 − 5.58***

0.000

1.15

GlobalNetExport

 − 1.48

2.49

 − 1.84

3.21

1.02

0.309

3.28

  1. The null hypothesis of the t-test is that the mean value of the difference between the risk and non-risk periods is 0; ***, **, and * indicate that the null hypothesis is rejected at significance levels of 1%, 5%, and 10%, respectively