Making Credit Scoring “Plug & Play”

Lenders must adopt new credit score models in order to reach the true credit eligible universe, says model developer VantageScore Solutions. One way for lenders to bring new data and scoring methods alongside their in-house models is to adopt a “Plug & Play” approach. But when should it be adopted and how can FIs fully benefit?

Take one recession, add a credit scoring model created a few decades ago and you get lots of lenders demanding a new credit scoring model that can address the new post-recessionary consumer. Inspired, VantageScore Solutions debuted the latest iteration of its scoring model.

Known as VantageScore 3.0, the model was built on 45 million anonymized consumer credit files, and the company has said it uses data points that are more granular than information used in earlier models. VantageScore maintains that the VantageScore 3.0 model helps shed light on the creditworthiness of as many as 30 million-35 million individuals in the U.S. who would otherwise be overlooked, including millions of minorities who are all but ignored by other scoring models.

But the question on lenders’ minds is how to take advantage of these new innovations.

In a white paper, VantageScore said lending institutions can benefit from adoption of its credit model into their own platforms. This is known as model conversion, and among the most easily adopted and quickest methods of that conversion can take place within a “Plug & Play” option. In a “Plug & Play” undertaking, the lender would typically see results within three to six months, with a few caveats in place that must be kept in mind as the model conversion process plays out.

When implementing a “Plug & Play” strategy, managers must examine risk levels and changes in populations being measured. The “Plug & Play” model may be most immediately useful when lenders look to have a “single score cut-off” as a decision criteria as to whether or not to lend to an individual. The white paper describes in detail how to leverage this relatively simple process to rollout a new scoring model.

According to VantageScore, there are data sources that can complement or an implementation, ranging from product performance charts to probability of default maps generated through FDIC guidelines.

To use one example, lenders can arrange data by a “probability of default” (or PD), via product performance charts. Alternatively, there could be further granularity as companies use, say the FDIC PD maps, which assess higher risk loans at banks with more than $10 billion in assets.

In reality, there is just one central question that must be answered for successfully converting a strategy to use a new credit score. What is the value of the new score (NewScore) that represents the same default rate or population volume designated by the previous score (OldScore)? All conversion processes revolve around answering this question and essentially follow the same steps.

 

[gview file=”http://www.pymnts.com/wp-content/uploads/2015/07/VS-Model-Conversion-PlugPlay-WP-7-20152.pdf”]