FICO Fell 16.7% After Regulators Opened Its Mortgage Monopoly

On Thursday night, after the U.S. market closed, Bill Pulte, director of the Federal Housing Finance Agency, posted a few sentences on X. By Friday's close FICO had fallen from $1,118.93 to $932.26 — down 16.7% on the day, and 21% at the intraday low. The company's market capitalization ended at $20.13 billion, about $4 billion lighter than the day before.

The same session, SPY fell 0.4% and QQQ 0.02%. August nonfarm payrolls had come in at 162,000 with unemployment steady at 4.1%, and the market spent the day arguing about whether that raised the odds of a rate hike. So what happened to FICO on Friday was not a market move. It was one company being repriced on one official's post.

The sentence read:

"EFFECTIVE IMMEDIATELY, I'm instructing Fannie and Freddie to approve ALL lenders to use VantageScore"

The reason came in the same thread: "FICO has enjoyed a monopoly. No more."

The third of those figures is the heart of the matter. Between May 1 and August 31, VantageScore 4.0 was the sole credit score on more than 9% of the mortgages Fannie Mae and Freddie Mac securitized. It did that with 50 lenders. On Thursday night the cap came off.

What FICO Sells

FICO sells a credit scoring model. Banks and mortgage lenders pull an applicant's payment history from three credit bureaus — Equifax, Experian and TransUnion — and run it through FICO's model, which returns a number. Every loan sold to Fannie Mae or Freddie Mac had to carry that number. There was no alternative.

Running a model one more time costs close to nothing. That is why the scoring business is extraordinarily profitable: FICO's Scores segment reported $475 million of revenue in fiscal Q2 2026 at a 91% operating margin. Segment revenue rose 60% year over year and mortgage originations revenue rose 127%. Eligible origination volume in the same quarter was $495 billion — and that volume had not grown.

When revenue doubles while volume stands still, the only thing carrying the increase is price.

The Price Ladder

Price did climb. According to a tally by the Community Home Lenders of America, FICO's base charge for a tri-merge credit report went from $1.80 in late 2022 to $30 in 2026 — more than fifteenfold in four years. The all-in credit report cost per loan, bureau markups included, went from roughly $50 in 2022 to roughly $540 in 2026.

That climb drew political attention. In March, Senator Josh Hawley asked the FTC to examine FICO's pricing. His letter put it this way:

"Rather than competing on price, FICO has leveraged this market position to impose a pattern of extraordinary price increases."

Pulte's own claim was blunter: he said the price per score had risen 1,800% since 2020. The verifiable series is CHLA's, and it points to roughly 1,500%. The two run in the same direction without being the same number.

FICO's April Move

FICO did not sit still. On January 1, 2026 it began distributing scores directly to tri-merge resellers, bypassing the bureaus, and offered lenders two structures: a flat $10 per score, or $4.95 per score plus $33 per borrower on funded loans. In April it went further, cutting the upfront fee on its 10T score from $4.95 to $0.99 and raising the closing fee to $65.

Headlines read that as capitulation. It was not. What changed was the gate at which the toll is collected.

The arithmetic says the 75 applications that never close were always low-yield work for FICO. Moving the fee to the closing shrank the number the company was politically exposed on — price per score — while growing the revenue.

There was a second effect, and at the time it worked in FICO's favor. Once the upfront cost is $2.97, adding a VantageScore pull alongside FICO costs a lender almost nothing. Comparison gets cheap, and FICO becomes immune to it. On the July earnings call, CEO Will Lansing said volumes had not been lost — lenders were pulling both scores.

The Other Face of the Same Design

The price of that immunity is that revenue piles up at a single point.

FHFA's own implementation document defines that point. It states that the tri-merge requirement does not change; what changes is which score the lender delivers to the Enterprise. A lender reports either Classic FICO or VantageScore 4.0 on a given loan — not both. The choice is loan by loan, and it is binary.

Combine that binary choice with FICO's new fee structure and the result is this: if a lender qualifies a borrower on VantageScore and never pulls FICO on that loan, FICO's revenue from it does not shrink, it goes to zero. Under the old model, revenue was spread across hundreds of pulls, and partial substitution — pull both, deliver one — still paid FICO. Under the new model, 85% of the money sits in a single line item collected at closing.

The stock is 53% below its 52-week high of $1,998.01 and down 43% year to date. At Friday's close FICO's market capitalization was $20.13 billion; Equifax, one of the bureaus that resells its score, was worth $20.80 billion. The company that makes the score is now smaller than the company that distributes it — enough of a picture to ask what the valuation multiple rests on.

A price chart shows the path, not the cause. The cause is which category the following sentence falls into.

Statement or Instrument

Headlines said Washington ended FICO's monopoly. Three questions have to be asked separately.

What exactly was said? An instruction. The verb is imperative: instructing. Not a rule, not a statute, not a court order. The venue was not the Federal Register either — it was a social media post.

Is there an instrument that makes it binding? Partly. FHFA has been the conservator of Fannie Mae and Freddie Mac since 2008, and a conservator's instruction binds the Enterprises. The legal ground was already laid: the Credit Score Competition Act of 2018 and FHFA's approval of VantageScore 4.0. So no new Federal Register notice is needed. What is needed is a Selling Guide update from each Enterprise — and FHFA's own document says they are "completing the final steps." Neither FHFA nor the Enterprises had commented by Friday evening.

Does the instrument's scope actually cover the event? No, and this is the instructive part. FHFA can order the Enterprises to accept a score. It cannot order any lender to use one. The instruction removes a barrier; it does not perform the substitution. The 9% reached in four months shows what happens when the barrier comes down for 50 lenders. Whether the same holds for thousands is a separate and still unanswered question.

The Other Side

That third row is Friday's second story. The same day, Pulte turned on the bureaus:

"Equifax, Experian, and TransUnion have been overcharging Americans for far too long."

He added that FHFA is seriously considering a move to two-bureau reports, or bi-merge. Equifax fell 6.37% to $177.05 and TransUnion fell 5.93% to $79.88. Given that both companies co-own VantageScore, the reaction looks backwards. It is not: the share VantageScore wins brings the bureaus marginal revenue, while cutting the three-bureau requirement to two erases a third of the reports outright.

The index side of the same day, for comparison:

What Is Left

Until the Selling Guide update is published, the instruction has no operative effect on any lender. When it is published, it will still compel no one; it will only open the option. The number to watch is where VantageScore's sole-score share goes from 9% — because FICO's revenue now depends on which side of that line a loan falls.

The second thing to watch is bi-merge. If the three-bureau requirement drops to two, the company that suffers is not FICO but the bureaus — and the larger part of the $540 bill sits with them.

This piece draws on FHFA's implementation FAQ and news releases, VantageScore's September 4 announcement, FICO's fiscal 2026 second- and third-quarter presentations, pricing reporting by HousingWire and National Mortgage News, American Banker's and Scotsman Guide's accounts of Pulte's posts, the Community Home Lenders of America cost tally, and Senator Hawley's March letter. Price and market-capitalization figures are as of the September 4, 2026 close. Pulte's claim of an 1,800% price increase could not be independently verified; the verifiable series cited here is CHLA's.