A new white paper released by FICO indicates that its FICO Score 10 T model demonstrates significantly greater predictive power for mortgage originations compared to VantageScore 4.0. The analysis detailed in the white paper suggests that the enhanced prediction capabilities of FICO Score 10 T could lead to increased loan approvals and reduced costs for borrowers across the mortgage industry.
The white paper highlights an independent third-party analysis that found VantageScore 4.0 provided minimal predictive improvement over Classic FICO. In contrast, FICO’s data scientists determined that FICO Score 10 T substantially surpasses Classic FICO in identifying potential loan losses. Specifically, FICO Score 10 T’s improvement over Classic FICO was shown to be five times greater than VantageScore 4.0’s improvement, identifying 18% more defaulters in the critical score decile used for mortgage origination, compared to just 3.4% for VantageScore 4.0. The analysis also suggests that VantageScore 4.0’s improvement over Classic FICO is so minor that its superiority could be questioned when accounting for a truncation anomaly described in the white paper.
FICO asserts that the adoption of FICO Score 10 T by market participants, including mortgage insurers and investors, would improve mortgage pricing and lower costs for borrowers due to its superior model performance, benefiting millions of Americans.
Ethan Dornhelm, vice president of Scores and Predictive Analytics at FICO, stated, “I’m proud of my analytics team for developing such a strong credit score for predicting mortgage default. We were confident that FICO Score 10 T would significantly outperform VantageScore 4.0 in predicting mortgage risk, and this analysis shows that conclusively.”
Beyond its predictive accuracy, FICO Score 10 T is noted for its approach to fairness. The white paper points out that VantageScore 4.0 incorporates mortgage-specific variables in an attempt to improve performance, which can penalize individuals who have never owned a home. This approach could lead to lower scores for millions of Americans, including young people, military members, and those from disadvantaged groups, simply because they lack homeownership history. In contrast, FICO Score 10 T incorporates rental data without penalizing individuals for not owning a home.
Jim Wehmann, president of Scores at FICO, commented, “Just as the mortgage industry begins to embrace rent as a meaningful indicator of credit readiness, VantageScore 4.0 penalizes renters with lower scores. VantageScore’s mortgage variables will force lenders to inform disadvantaged groups or members of the military through adverse action notices that they have been rejected for a mortgage because they don’t currently own a home. Imagine telling a potential borrower, ‘you need a mortgage to get a mortgage.’ Not only is that confusing to consumers, it is unfair.”
FICO Score 10 T received validation and approval from the Federal Housing Finance Agency (FHFA) in October 2022 for use by Fannie Mae and Freddie Mac, enabling its application in conforming mortgage loan originations. Implementation of the model is currently underway. Early adopters, representing over $300 billion in annual mortgage originations and $1.5 trillion in mortgage servicing portfolios, have already begun adopting FICO Score 10 T and trading loans based on the model.
