On Thursday, the Federal Housing Finance Agency ended a monopoly with a social media post. Director Bill Pulte announced he had directed Fannie Mae and Freddie Mac to approve every mortgage lender to use the VantageScore credit scoring system, effective immediately. "FICO has enjoyed a monopoly. No more."
The order converts a limited pilot into the default. Since May 1, the two government-sponsored enterprises had accepted VantageScore 4.0 only on a trial basis, and Pulte said the rollout had delivered loans through roughly 50 lenders. Now the cap is gone. Every lender that sells a loan to Fannie or Freddie can score the borrower with the model the GSEs spent four decades refusing to touch.
The announcement did what announcements of this kind usually do. Fair Isaac shares fell about 20 percent in early Friday trading. TransUnion dropped 9 percent, Equifax 8 percent, and Experian, which reports in London, shed 4.6 percent. The market priced the end of one company's lock on a mortgage market that prices trillions of dollars of American housing debt.
Pulte paired the order with a claim about why the monopoly lasted: FICO, he said, had raised its price per score by 1,800 percent since 2020. The number comes from the regulator's account and has not been independently verified, but the direction of it is not in dispute among the industry's own trade press. Mortgage credit scoring has been a single-vendor market for so long that the buyer side has no muscle memory of comparison shopping. The statutory foundation for the order is the Credit Score Competition Act of 2018, which instructed the housing finance agencies to open Fannie and Freddie to advanced alternative models. It took seven years, a four-month pilot, and a social media post to complete the job.
The harder question is whether the market priced the right thing.
The pilot already showed what a mandate cannot do
Permission was tested before it was given. When Fannie and Freddie opened the pilot in May, the constraint most observers blamed for FICO's dominance had been removed for a select group. The result is the best data available on what happens when the gate opens.
As of August 31, VantageScore 4.0 had been used in slightly more than 9 percent of the mortgages Fannie and Freddie securitized since the pilot launch, according to HousingWire's reporting on the rollout. Adoption concentrated in two lenders, Rocket Mortgage and United Wholesale Mortgage. Fifty firms were allowed through the gate; two of them accounted for the traffic.
The pilot itself had a history that points in the same direction. Fannie and Freddie announced the limited rollout in April, and that announcement was the payoff of a process that began in July 2025, when Pulte first blessed VantageScore as an acceptable alternative. Each step was permission, and each step was followed by exactly the adoption the market felt like giving. Permission scales instantly. Practice does not.
That pattern matters because it separates two things the Thursday order treats as one. The order removes the last licensing obstacle to VantageScore. It does not, and cannot, remove the reasons lenders were slow to use the score when the obstacle was already gone.
Lenders underwrite against decades of accumulated machinery. Their pricing models, their investor agreements, their staff training, their compliance scripts all assume a FICO score arrives in the file. Replacing that assumption costs money in the quarter it happens and returns money, if at all, across the years that follow. A mandate from Washington removes the regulatory reason to hesitate. It does nothing to the operational one.
The tollbooth moved; it did not disappear
The most overlooked detail of the announcement is the ownership structure of the winner. VantageScore has been jointly owned by Equifax, Experian and TransUnion since its founding in 2006, the same three firms Pulte spent the same week describing as "cartel-like" and accusing of having overcharged American consumers for far too long.
The tension is not a footnote. It is the shape of the new market. A mortgage credit score needs two inputs: a scoring model and the underlying consumer data. FICO supplied the model and paid the bureaus for the data. VantageScore is the bureaus' own model running on their own data. Every dollar that shifts from Fair Isaac to VantageScore moves the margin from one vendor to the vendors Pulte says are the problem.
Friday's market action already told this story in miniature. Fair Isaac lost a fifth of its value because it lost the regulatory guarantee on its mortgage franchise. The bureaus lost far less because they cannot lose this market in any scenario. FICO scores and VantageScores are both built on their files. The tollbooth moved; the road still runs through the same three properties.
The score that sees more people
None of this means the policy change is a wash for borrowers. The inclusion numbers behind VantageScore 4.0 are the strongest argument for the directive, and they are worth taking at face value.
According to the data VantageScore's backers have circulated, the model can score roughly 33 million more American adults than the FICO models currently used in mortgage underwriting, including nearly 5 million people who would qualify for a mortgage under the new model and were invisible to the old one. The model pulls trended data and payment histories from rent and utilities that traditional scoring ignores. A study by Deep Future Analytics, a mortgage analytics firm, estimated a full rollout would generate more than $930 million in market-wide savings in its first year.
The counterargument is not that these numbers are wrong. It is that a score is an input, not a loan. A borrower who becomes scorable still needs a lender to underwrite the file, an investor to buy the loan, and a guarantor to accept the risk. Inclusion at the scoring stage only becomes inclusion at the closing table if the entire chain reprices itself around the new information. The mandate guarantees that the score can appear in the file. Nothing in the order guarantees the file performs.
The savings estimate deserves the same treatment as any vendor-supplied number in a regulatory fight: it assumes adoption. Nine hundred thirty million dollars in first-year savings materializes only if lenders switch, investors accept, and borrowers actually close at lower cost. The pilot offers no evidence that the full chain moves together, because only two lenders ever pushed volume through it. A score that can see more people changes the pool of scorable applicants. It does not, by itself, change the pool of approved ones. The loan file still has to clear the same underwriting bar. The score is the invitation, not the admission.
Investors price risk; mandates do not
Fair Isaac's response to the directive was short and pointed: its FICO Score 10T, the model it has been positioning for GSE adoption, is "the most predictive credit score available today."
Set aside the self-interest and the claim still describes the real obstacle. Fannie and Freddie do not bear default risk alone. They guarantee mortgage-backed securities that pension funds, insurers and banks hold, and those investors spent three decades learning how FICO-score distributions map to default rates. A new model can be more inclusive and more predictive at the population level, and investors will still discount what they cannot yet validate on their own books. That discount shows up as loan-level price adjustments, and lenders pass it to the borrower or eat it themselves.
Validation is a boring word for an expensive thing. A mortgage investor's pricing engine is a machine built on default-rate histories, and every input in it was tuned against the score distribution the industry has used for decades. Introducing a new distribution means rebuilding the machine or discounting the output. Neither is free, and neither is ordered by Thursday's directive. That is also why the strongest claim Fair Isaac can make is not about law but about mathematics: a model that predicts defaults better should win on price, and FICO argues 10T does. The VantageScore camp answers that a slightly better predictor on a narrow population is worth less than a broadly useful one on a wider population. The pilot did not settle that debate. It only showed who was willing to test it first.
This is why the pilot's 9 percent figure, not the Friday stock move, is the number that matters. The gate was open for four months. The market, unpushed, chose FICO in more than nine of every ten loans.
The next front is already visible. Pulte said the agency is seriously considering bi-merge reporting, pulling credit data from two bureaus instead of three, and studying whether one report could suffice. The Mortgage Bankers Association has supported ending the tri-merge requirement in favor of a single-file approach for strong-credit borrowers. If bi-merge follows, it will do to the bureaus what the VantageScore order did to Fair Isaac, and the tollbooth argument will apply twice over.
The gate is open; the market decides the rest
The FICO mortgage monopoly ended on Thursday as a matter of policy, and the policy reasoning is defensible on its own terms. The 2018 Credit Score Competition Act directed exactly this outcome, the pilot produced real loans, and a borrower market dominated by a single vendor was never a healthy equilibrium.
What ended on Thursday is narrower than the announcement suggested. Fair Isaac lost its legal guarantee. It did not lose the underwriting systems, the investor models, or the decades of validation that made the guarantee valuable in the first place. VantageScore won the right to compete in every loan file, and it did so at the exact moment its owners were being called a cartel by the regulator handing it the market.
The interesting number is not the 20 percent Fair Isaac gave back on Friday. It is what share of next quarter's GSE loans carry a VantageScore model, now that nothing stops them but the market itself. For the next few quarters, the mortgage market will run the experiment the pilot only previewed. The regulator has removed every reason not to switch except the ones that were always the real reasons.
The gate is open. The road runs through the same three tollbooths either way.
Primary sources
- Reuters for the FHFA directive, Pulte's statements, the 50-lender pilot figure, and the market moves in Fair Isaac and the three credit bureaus.
- HousingWire for the pilot adoption figures and the rollout specifics since the May 1 launch.
- American Banker for the "cartel-like" language and the FHFA's bi-merge plans.
- Congress.gov for the 2018 Credit Score Competition Act, the statutory basis for the directive.
- MPA Magazine for the terms of the VantageScore 4.0 rollout to all GSE lenders.