More Reasons to Trust VantageScore
VantageScore 4.0, our fourth-generation tri-bureau credit scoring model, sets a new standard for predictive performance and modeling innovation. With the utilization of trended credit data, the application of machine learning techniques and the anticipation of new consumer data protection plans, we have developed a credit scoring model that benefits both lenders and borrowers.
- Refreshed Data Set – VantageScore was developed using 2014-2016 data to account for the latest credit products and trends in consumer behavior.
- More Consistently Aligned Credit Scores – A patented hallmark of VantageScore is the ability to deploy the same model independently at Equifax, Experian and TransUnion.
- Predictive Performance Lift Across All Credit Industries – Our new model outperforms previous models in all major credit categories, with significant predictive lift overall in loan originations and a 5.3 percent lift in mortgage originations, despite the current low-default risk credit cycle.
- More Consistent Scoring – Our single, uniform model with patented characteristic leveling treats pieces of data from each credit reporting company (CRC) the same. That means less variation in scores pulled simultaneously from multiple CRCs. And the variations that do appear, can be chalked up to differences in credit-file content, rather than the scoring process itself.
- Greater Stability – VantageScore is developed using an exceptionally wide breadth of consumer credit-file data, allowing the model to retain its predictive power for longer periods of time, even through spans of economic volatility.
- Scores Approximately 40 million More Consumers Than Conventional Models – Scores assigned to this previously unscoreable population are highly predictive, yielding an incremental predictive performance lift among dormant consumers and a strong gini coefficient among the expanded universe of consumers.
- Yearly Model Performance Assessments – To ensure the VantageScore model continues to deliver highly predictive and consistent credit scores, our analytics team tests its performance yearly. Unlike developers of other credit scoring models, we publish the results of those annual model performance assessments for all to see.
A demographic breakdown of the 40 million more potential customers who were previously “unscoreable” with traditional credit scoring models:
|2018 Unscoreables - All Scores||2018 Unscoreables - All Scores 620 and above|
|Total||40 million||10.06 million|
|Black and Hispanic||12.2 million||2.4 million|
|Asian/Pacific Islander||1.6 million||<1 million|
|White||25.7 million||7 million|