The best credit scoring models change with the times
“Even fine wine can turn to vinegar.”
That’s a quote from my colleague Sarah Davies, who heads research, analytics and product development here at VantageScore Solutions. It’s taken from a presentation she delivered at the first annual Structured Finance Industry Group (SFIG) conference last month.
Sarah reminds us all that changes over time in consumer credit behaviors could render even the best-designed credit scoring models less effective. It’s an important message for consumers as well. Below is a chart demonstrating how the predictive strength of factors in a consumer credit file changes over time.
Clearly much has changed since 2005. In particular, overall “payment history” is 25 percent more important in terms of predictive contribution than pre-recession models, while the predictive power of “recent credit” has fluctuated over time.
The takeaway for lenders? Using a model based on older data means sacrificing predictiveness. Prudent risk managers can address these changes by continually validating their scoring models, and by using models based on the freshest consumer data available.
For consumers, the takeaway is that lenders must adapt and upgrade to newer models, such as the VantageScore 3.0 model, in order to obtain the most accurate credit scores possible—scores that reflect consumers’ true credit profiles.
Speaking of validation and prudent risk management, there is more terrific content in this month’s newsletter. We’ve included an article about a new webcast series we recently launched to help lenders evaluate their credit scoring models and upgrade to models that take greater advantage of advances in decision analytics. We’ve also included an article that combats new myths about credit scoring, and an informative piece on how various degrees of delinquency impact a person’s credit score.
Last, but far from least, I am pleased that Steve Deggendorf, director of business strategy at Fannie Mae, joins us for this month’s “Five Questions With” profile. Steve oversees Fannie Mae’s monthly housing survey, and he answers questions about how consumers use technology to shop for mortgage loans.