[Retracted] Evaluating Investors’ Recognition Abilities for Risk and Profit in Online Loan Markets Using Nonlinear Models and Financial Big Data

Table 3

Multifactor estimated results of heteroskedasticity Probit model of default.

(a)

Variables	Explained variable: whether default
		-statistics and value
	0.0002	1.56 (0.119)
	-0.0019	-8.29 (≤0.001)
	0.0021	5.41 (≤0.001)
	0.0002	6.61 (≤0.001)
	0.0001	2.58 (0.010)
	-0.0008	-2.32 (0.020)
	0.0002	0.27 (0.785)
	-0.0030	-18.22 (≤0.001)
	0.0177	26.60 (≤0.001)
	-0.0014	-3.82 (≤0.001)
Housing ()	-0.0016	-2.81 (0.005)
	-0.0013	-1.99 (0.047)
Testing whether existing heteroskedasticity
Null hypothesis	Lr-statistics and value	Conclusions
Homoskedasticity	8971.15 (≤0.001)	Rejecting null hypothesis, namely, existing heteroskedasticity

(b)

Variance equation
Variables	Coefficient estimated values	-statistics and value
	0.0323	4.1500 (≤0.001)
	0.0326	2.3100 (0.0210)
	-0.0805	-3.6600 (≤0.001)
	-0.0335	-14.3000 (≤0.001)
	-0.0071	-2.5300 (0.0110)
	0.0109	0.5000 (0.6180)
	-0.0046	-0.1100 (0.9130)
	0.0413	10.8300 (≤0.001)
	0.5852	30.8900 (≤0.001)
	0.0908	4.0400 (≤0.001)
	0.1730	4.4800 (≤0.001)
	0.0603	1.4900 (0.1350)

Note: means that the corresponding coefficient is significant at 1% significance level, and means that the corresponding coefficient is significant at 5% significance level.