VantageScore claims that Classic FICO is deficient and that its own model is more predictive of defaults and scores more borrowers. Based on our analysis, their methodology is flawed, so much so that the claimed differences between VantageScore 4 and Classic FICO disappear. Given the trillions in mortgage debt, millions of borrowers, and immense taxpayer exposure, it is essential that the debate be grounded in facts rather than marketing.

FICO has anchored mortgage underwriting since the mid-1990s. VantageScore, created in 2006 by the three credit bureaus, spent years trying to break in, pitching itself as better at scoring millions of "credit invisible," a way to break FICO's monopoly, and lowering costs.

READ MORE: Pulte plans 'full scale review' of credit bureaus

That campaign gained traction with Congress's 2017 Credit Score Competition Act, which required FHFA to set up a transparent approval process for new models. In 2022, FHFA approved VantageScore 4.0 alongside FICO 10T. Political momentum grew after the COVID-19 affordability crisis, which fueled calls to expand credit access. Rising FICO fees—from about $0.60 in 2018 to $4.95 in 2025—added industry frustration and political appetite for an alternative and may well have affected FHFA's decision to adopt VantageScore 4.0.

Riding that momentum, VantageScore recently claimed that its 4.0 model predicted "up to 49% more mortgage defaults leading into the COVID-19 pandemic period than Classic FICO." The company described this as a "head-to-head" comparison, but a closer look reveals fundamental flaws.

An AEI Housing Center analysis found two: a methodological mismatch and selection bias. VantageScore compared the average of three bureau scores for its model against the GSE standard (middle of three, or lower of two) for Classic FICO — an apples-to-oranges test. More importantly, its most dramatic findings came from a sample limited to loans with FICO scores of 720 or less, which skewed the results in VantageScore's favor. Run on the entire sample, and Classic FICO comes out ahead.

When tested on a full, unfiltered dataset with consistent methodology, VantageScore's advantage disappears. On two of VantageScore's preferred measures, Classic FICO scores slightly better, with VantageScore having a marginal advantage on the third. For each metric the individual differences are small.

READ MORE: Score modernization taking shape but ROI tough to quantify

Other independent studies confirm this. The Urban Institute found cumulative defaults at the bottom decile differed by just one percentage point, and Milliman concluded that both models have strong predictive power with only marginal differences across score ranges.

VantageScore claims it will reduce costs for consumers, but the numbers tell a different story. A typical tri-merge credit report costs $80–$100, paid to a reseller. To create it, the reseller purchases credit files from each of the three repositories at $13–$25 apiece. Each repository then pays FICO roughly $5 in royalties — about $15 total. In the context of thousands of dollars in closing costs, FICO's share is marginal at best.

The truth is while FICO has long been the sole provider of an accepted mortgage credit score, the three credit repositories — which jointly own VantageScore — are also the exclusive sellers of the underlying credit data. Replacing FICO with their own model doesn't necessarily lower costs for consumers — it simply allows them to capture a bigger slice of the pie.

Perhaps most troubling are VantageScore's claims that more competition will strengthen the market. In reality, designing new Loan Level Pricing Adjustors and rebuilding investor models is a Herculean task, made harder by the fact that FICO and VantageScore don't line up cleanly.

The spread for individual scores can be huge: for example, a borrower with a 720 FICO might register anywhere from 660 to 800 on VantageScore. This dispersion makes it hard to directly map Classic FICO-based LLPAs onto VantageScore 4.0 as VantageScore has proposed. This challenge is compounded by a lack of historical performance data on loans denied on the basis of a Classic FICO, but which would have been approved using a VantageScore 4.0. This leaves investors without a reliable way to calibrate and price risk.

More competition also invites gaming. Because FHFA explicitly allows "score shopping," lenders, realtors, and borrowers could simply choose the more favorable score, leading to more approvals, looser credit, and greater default risk, while leaving vulnerable borrowers with unsustainable debt. Score providers, in turn, may lower their criteria to win more business.

These risks multiply if FHA follows suit: Far more borrowers would qualify – VantageScore estimates 5 million— but with housing supply already constrained, that extra demand would only push home prices higher. This would result in worsening affordability, especially for first-time buyers. And if investors in mortgage-backed securities were to lose confidence in the credit scores, they may demand higher yields, raising costs for all borrowers. Any supposed savings from ending FICO's monopoly would vanish, while taxpayers would be left holding the bag.

One thing is clear: the predictive power of VantageScore 4.0 and Classic FICO is effectively the same. Thus, the real risk for the housing finance system isn't Classic FICO— it's the premature and fraught adoption of a two-score system. Policymakers should proceed cautiously, ensuring changes serve borrowers, lenders, investors, and taxpayers — not just a score provider. If the goal is to save consumers money, perhaps the better question is whether a tri-merge credit report is even necessary today.