The value of alternative credit data in separating risk on the margin

by Kevin King

Kevin King

In my last blog, I discussed how the use of alternative credit data has become a mainstream answer to the challenge of identifying creditworthy consumers who are credit invisible. While that is perhaps the most familiar use case for risk managers, there are numerous use cases for alternative data which can have a significant impact for enterprises. One of the most notable is leveraging these insights to help provide a more complete view of credit applicants classified as marginal, by powerfully distinguishing creditworthiness in a population that traditional data may see as homogenous.

Marginal consumers, those consumers often considered the near-prime population, can be the most difficult to make approval and pricing decisions for – they sit right on the cusp of a lender’s credit policy. ID Analytics examined the portfolio of one retail card provider using our alternative score, Credit Optics® Full Spectrum, to demonstrate how enterprises can separate risk within this challenging credit segment.

For this case study the retail card provider defined marginal consumers as falling within a traditional credit score band of 640-660. Traditional bureau scores can struggle to separate risk within such a narrow score band – they tend to view these applicants as having the same degree of creditworthiness. The lender’s credit policy had an approval margin at 650 when using a traditional credit score – applicants above 650 were credit-approved for a basic product, while applicants below this cut-off were declined. When we looked at the performance of consumers in the 640-660 range, we saw that the traditional credit score assessed these consumers as very similar from a risk perspective with little rank ordering observed.

The lender’s current credit policy aimed to set a marginal cut at a 5% incidence rate. Credit Optics Full Spectrum identified 40% of the marginal applicants as falling well above the retail card issuer’s credit policy and qualified for approval (see Figure 1). Notably, Credit Optics Full Spectrum also found 30% of marginal applicants were at least double the acceptable risk in the lender’s policy and should not be approved based on the lender’s policy.









Figure 1. The application of an alternative credit score to separate risk among the retail card issuer’s marginal applicants.

The additional separation of risk using an alternative credit score can help lenders identify marginal consumers whose likelihood of repayment may not be wholly understood by traditional data and scores. Depending on your objectives, this clarity on the margin can help drive increased approval rates, reduced loss rates – or a balance of both.

Are you interested in learning about the value of alternative credit data beyond its use for helping to uncover credit invisibles? Please join me and Senior Research Analyst, Leslie Parrish of Aite Group as we present, The State of Alternative Credit Data, on Tuesday, June 11, from 11:00 am – 12:00 pm PST. This live webcast will cover emerging opportunities presented by alternative data including, market drivers for adoption (i.e., why top lenders call it a competitive advantage), use cases for prescreen and portfolio management and best practices for optimal results across the credit spectrum.



Kevin King is Director of Product Marketing at ID Analytics