Application Scorecard For Auto Loan & Subsequent Analysis

Radix Analytics Pvt Ltd > Pendinng > Application Scorecard For Auto Loan & Subsequent Analysis

Issues & Objectives

Benefits

Discriminatory power of the multiple scorecards was found to be higher than the single scorecard
Scorecard was implemented and is operational for 4 years
Better understanding of payment on KPI
Recommendation on underwriting rule - same characteristics considered in model as well as by underwriters thus causing double penalty
Clear understanding of coverage and potential for scorecards built from both

Project information

Techniques

Risk Analytics

Client

Car Finance Company

Industries

Financial Services

Location

UK

Challenges

Methodology for existing scorecard not well organized
Credit history, product information, loan terms all data scattered
Performance available for 8% TTD among 55% approval
Payment behaviour KPI was not structured
Underwriting rule was not well documented
Bureau data for same time period for like for like comparison

Solution

Data processing and analysis was done with R
Customer classification to obtain homogeneous segments (Decision tree and cluster analysis)
Multiple scorecards all scaled to have same odds
Scorecard as a linear function for easy integration with LOS
Cohort-wise analysis of payment behavior of arrear and upto date customers as well as various loan instruments
Analysis of characteristics to obtain the ones possibly affecting underwriter decisions - made recommendations
Statistical comparison of bureau data as well as quick scorecards using both

What we build, how it performs - Explore our work!

Our Case Studies

Would love to hear your thoughts!