Equipped with non-traditional, “alternative” data, plus “time-series” data, plus more computing power, lenders are getting a better look at how individual consumer payment behavior changes over time, with an eye toward predicting future behavior.
“If you look at point-in-time data, you’re looking at a snapshot ― information that’s very static,” said Karl Stabler, vice president of risk and analytics at Flagship Credit Acceptance, in a recent Auto Finance News webinar, “Taking Action: Strategies to Manage Shifting Credit Behavior, ” sponsored by Equifax Inc. . A free recording of the webinar is posted here.
In the webinar, Stabler said two different consumers could look alike at one point in time, with the same credit score and the same utilization of their available credit.
A closer look though, might show that for the last six months, one made only the minimum monthly payment on credit card debt, while the other always paid additional amounts. To reach the same point, one might have run up balances on several different cards, while the other started out with a higher balance but managed to pay it down over time, he said.
“One has gone from 10% to 40% of their credit limit, while the other went from 80% to 40%. If you only look at one point in time in the credit data, you only see [they’re both] 40%,” Stabler said.
The use of alternative data, such as rental payments, public records, income verification, utility and phone bills, and even club memberships can help identify high-risk credit profiles, he said. Conversely, alternative data can help assign a score to previously “unscoreable” customers.
“Eighty percent could be scored. And of those, 60% had lower to moderate risk profiles,” Stabler said. “They are not monolithic.”
For more content like this, check out the upcoming Auto Finance Summit, October 21-23 at Wynn Las Vegas. Visit www.AutoFinanceSummit.com for more information.Like This Post