How do you score approximately 40 million more consumers without lowering risk standards? How do modern data and methodologies facilitate more accuracy, predictiveness and inclusiveness in credit scoring? A new white paper discusses the benefits of using the latest trended credit data and leveraging machine learning to identify “unconventional” users of credit. Read “Anything but Conventional: Leveraging New Modeling Techniques and Better Data to Score Tens of Millions More Consumers with More Predictiveness”, and see how it refutes two myths in the marketplace:
- Myth 1: Only recent and repeated credit users can be reliably scored.
- Myth 2: Using modern scoring methodology and data to score consumers creates a “race to the bottom” or lowering of standards.
The results of thorough research and testing showcase the ineffectiveness of using traditional scoring data and methodology to score more unconventional users of credit (e.g., those who are neither recent nor repeated credit users). Specifically, study findings conclude:
- There are tens of millions of consumers, including those with relatively low levels of credit risk, who are not scoreable with conventional models.
- Given the same credit score, there is no statistically significant difference in default outcomes between conventionally scored consumers and newly scoreable consumers, even though different elements of their credit reports are being utilized.
- Newly scoreable consumers do not exhibit different default rates (measured as delinquency of 90 days or more over 24 months) compared to conventionally scored consumers with similar scores.
- Further, across all product categories, how quickly a consumer defaults on a new loan is comparable between newly scoreable consumers and conventionally scored consumers with similar scores.
About 40 million unconventional consumers of credit are considered “invisible” to a traditional credit scoring model.* With the use of the latest trended credit data and leveraging machine learning to score more consumers, VantageScore 4.0 provides a more accurate assessment of credit risk for these previously unscoreable consumers, achieving similar prediction accuracy compared to conventional models when measuring credit behavior over the standard 24-month period.
“There has been much speculation about how usage of the latest approaches in credit score modeling will lead to a ‘race to the bottom,’ but the numbers don’t lie. More than 10.5 billion VantageScore credit scores were used last year with over 2,200 lenders using our tried-and-tested models. This is the ultimate litmus test of success,” said Barrett Burns, CEO and president of VantageScore Solutions, LLC. “More fittingly, the reality of this situation should be characterized as a ‘race to the top’ to whom lenders can trust. And we, at VantageScore, are proud to lead the way.”
For more details on the “Anything but Conventional: Leveraging New Modeling Techniques and Better Data to Score Tens of Millions More Consumers with More Predictiveness” white paper, visit:
* The VantageScore 4.0 model allows lenders to accurately assess approximately 40 million more consumers than conventional models.
Popular Articles
Consumer FAQ: Benefits of Adding Rent and Utility Data to a Credit File
Advantage of Adding Rent and Utility Data whitepaper
Credit with a Conscience fact sheet
Driving Financial Inclusion with Data and Analytics fact sheet