The question about modifying existing scoring models is “how much do you want to tweak to bring in more customers?”. Moreover, if you are asking that question, you must project “should established customers carry the burden of increased write-offs?”
There is plenty of emotion around the topic. The claims suggest that lending is a social responsibility, not a business function.
More than half of Americans are effectively shut out of the financial system because they have a credit score that is considered subprime.
Why do we continue to think it is acceptable to turn a blind eye to over half of our country, instead of rethinking decades-old processes?
There is an unlikely comparison of rental responsibilities to homeowner mortgages. Certainly, a renter on Sutton Place in NYC would bear a larger burden that a homeowner in middle America, where one might pay $6,000 in rent and the other $2,000 in their mortgage,but one is building their landlord’s equity and the other builds their own..
Yet there are still instances where some rent, mobile phone and utility payments histories are excluded from scores, even though these can be some of the largest and most frequent payments a person will make today.
When you start comparing the attributes of paying an electric bill to keeping your unsecured debt current, is that a realistic comparison?
For half of America, the fact that this information is not reported is not an issue — they pay these regular bills via credit card anyway, meaning these expenses are already reported to the bureaus and impact their scores.
But what happens for the other half of Americans who pay in cash or might not have a credit card? It is no surprise that the exclusion of readily-available data from today’s traditional credit score calculations is a direct contributor to financial exclusion today — the reason half of Americans are shut out.
The results were compelling. The research found that 85% of people would have a higher credit score if short-term loan repayment data were included in credit reports. In fact, 15% would go from having subprime scores to near prime scores.
US revolving debt is up to record levels, with an increase of $200 billion. Credit losses are up, as anticipated in a growing market. The industry should test scoring models but must be extremely careful. Models that promise to bring in previously “un-found” customers will increase risk. Established scoring models help manage risk and ensure consistency in a risk environment.
The industry must be sensitive to the unbanked population but that does not mean we do not already have fair lending practices in place. If we start including short-term loan repayment data as suggested, then perhaps the models should also consider why the customer went to a short-term lender, presumably a pay-day lender, in the first place. In this instance the additional data suggest more of a household risk than a solid credit attribute.
Overview by Brian Riley, Director, Credit Advisory Service at Mercator Advisory Group
Read the quoted story here