Skip to main content

Table 13 Endogeneity check: regressions for the three risk classes

From: Fintech platforms: Lax or careful borrowers’ screening?

  (1) A-B (2) C-D (3) E-F-G
Verification process 0.0408*** (0.0123) 0.0353** (0.0141) 0.0394*** (0.00796)
Term 0.601*** (0.0148) 0.549*** (0.0167) 0.532*** (0.00814)
Interest rate 18.60*** (0.271) 3.236*** (0.269) 7.754*** (0.166)
ln (annual income) − 0.140*** (0.0103) − 0.240*** (0.0147) − 0.164*** (0.00776)
Loan purpose: debt consolidation 0.0391** (0.0175) 0.120*** (0.0204) 0.108*** (0.0116)
Loan purpose: credit card 0.0493*** (0.0188) 0.0514* (0.0286) 0.102*** (0.0139)
Loan purpose: home improvement 0.0822*** (0.0255) 0.0596* (0.0347) 0.118*** (0.0190)
Loan purpose: small business 0.662*** (0.0524) 0.283*** (0.0452) 0.370*** (0.0333)
Home mortaged − 0.340*** (0.0113) − 0.349*** (0.0155) − 0.345*** (0.00839)
Home owned − 0.127*** (0.0173) − 0.215*** (0.0241) − 0.197*** (0.0129)
year 0.350*** (0.0103) 0.193*** (0.0190) 0.272*** (0.00762)
3-digit zip Yes Yes Yes
Constant − 3.078*** (0.120) 0.579*** (0.167) − 1.276*** (0.0887)
Observations 429,448 87,932 394,827
  1. This table reports the results of logistic regression for all risk-classes. The dependent variable is the probability of default in all columns. In the first column the regression results for the lowest risk-class (A-B), in the second column the medium risk-class (C-D) and in the third column the highest risk-class (E-F-G) are shown. Standard errors in parentheses
  2. *, ** and *** denote 1%, 5% and 10% levels of significant, respectively