Fair Isaac (FICO): Moat Analysis

18 August 2026. First assessment. Written against the FHFA's July 2025 admission of a competing score into GSE-eligible mortgages, and against FICO's figures through the Q3 FY2026 report of 29 July, whose numbers were strong and are read here as lagging confirmation rather than reassurance.

This document asks one question: how strong is the moat, and is it still intact. It contains no price, no position and no target. FICO has fallen a long way from its 2024 high, and that is exactly why the moat must be judged on its own, before any price is looked at.

Verdict

Field	Reading
Strength	Strong, not exceptional. A powerful coordination lock, but one that depends on an external coordinator rather than on the company itself
Condition	Impaired. The lock has been breached at one entry point by regulatory action, but the market has not yet moved
Security	Moderate. An outside actor has reached the mechanism
Pricing authority	Constrained in mortgage, unconstrained elsewhere
Demand anchoring	Level 2. The need to assess whether a borrower will repay is permanent wherever there is lending; it falls with the credit cycle but always returns. What is contested is which score fills that need, not whether the need exists
Verdict	Strong / Impaired / Level 2 demand. Declined at the first two gates, not the third: the demand is durable, but the moat is strong-not-exceptional and already impaired
Class A conditions	1 of 6 triggered (A1, regulatory admission of a competitor). But the competitor is capped at a theoretical ~20% of the channel and runs alongside FICO rather than replacing it
Class B gauges	0 of 6 triggered. Every number still reads clean
Decision	Not eligible for a Layer 2 entry. The demand is durable (level 2), but the name fails the first two gates: the moat is strong rather than exceptional, and it is already impaired. Held on watch if owned, no fresh entry while impaired, because the outcome, whether the lock holds or adoption builds, cannot yet be known

In one line: a business whose moat is a coordination lock rather than a product advantage, whose numbers are the best in its history, and whose mechanism was nonetheless reached last year by the one actor able to reach it.

Why this verdict is uncomfortable, and why that is the point. Every gauge FICO reports is not just clean but exceptional. Scores revenue grew 41% in the most recent quarter, driven by higher mortgage score prices, and non-GAAP operating margin reached 62% against 57% a year earlier. If this file scored on the numbers, it would read intact without hesitation.

It does not score on the numbers, because the numbers measure the moat's output and the damage is upstream of them. What changed is not a number. It is that a regulator admitted a competitor into the one channel where FICO's use was mandated rather than chosen, and that admission is exactly what a damaged mechanism looks like before the damage reaches the income statement.

How strength and condition are judged is in the annex. The short version: strength is judged first, and FICO's strength is real but carries an external dependency the strongest moats do not have, which is why strength reads strong rather than exceptional even before condition is scored.

1. What the company does

Fair Isaac sells two things, and only one of them is a moat.

The Scores segment licenses the FICO Score, a three-digit number predicting how likely a borrower is to repay. It is sold to the three credit bureaus, to lenders, and directly to consumers, and it costs almost nothing to produce once the model exists. This segment is a minority of revenue and the overwhelming majority of the profit, and it is the entire subject of this document.

The Software segment sells analytics and decisioning tools, lately repackaged as the FICO Platform. It grows, it is real, and it competes on ordinary terms against capable rivals. It is not a moat and this file does not treat it as one. Everything below is about the Score.

How the money is actually made

A credit score costs a fraction of a cent to compute. FICO sells it for a few dollars in mortgage and cents elsewhere, and keeps most of that as profit. The question that matters is why anyone pays for a number that a competitor could compute just as accurately, and the answer is not accuracy.

The answer is that everyone else uses it. A lender underwriting a mortgage uses the FICO Score because the loan will be sold to Fannie Mae or Freddie Mac, who required the FICO Score, so the whole chain was priced and documented around it. A bond investor buying a pool of those loans reads their credit quality in FICO terms because that is the language the pool is described in. A regulator sets capital rules that reference it. None of these parties uses FICO because they compared models and FICO won. They use it because everyone else uses it, and switching alone, while everyone else stays, buys nothing but incompatibility.

That is the mechanism. It is not a better product. It is a standard that coordinates a whole market, and its value comes entirely from the fact that everyone is on it together.

Why the price could rise so far

Because the mechanism is a lock rather than a preference, FICO has been able to raise the mortgage score price repeatedly without losing volume. For most of its history a mortgage pull cost a few cents. In recent years FICO moved that to a few dollars, and in 2025 introduced tiers around a roughly $5 to $10 wholesale royalty, still a rounding error against total mortgage closing costs. Volume did not respond, because a lender cannot decline to pay for the score the buyer of the loan requires. Pricing power this complete is the visible signature of a coordination lock, and it is also the thing that draws the attention of anyone with the power to break the lock.

2. The moat

FICO's moat is a single mechanism with a few reinforcing layers around it. It is worth being precise about what that mechanism is, because the entire assessment turns on it.

Layer	Mechanism	Why it works
Foundational	Coordination lock	Every party uses FICO because every other party does. Value comes from universal adoption, not from the model
Reinforcing	Regulatory and contractual embedding	The score is written into GSE guides, capital rules, securitisation documents and decades of contracts
Reinforcing	The standard as a language	Credit quality across the market is expressed in FICO terms, so alternatives must be translated rather than simply adopted
Reinforcing	Data and brand with consumers	"Your FICO Score" is a recognised consumer object, which deepens the default
Optional	The Software segment	A real business, competes normally, not part of the moat

Foundational: the coordination lock

This is the whole moat, and it needs to be distinguished sharply from the thing it is usually mistaken for.

A product moat says the product is better and people choose it. It breaks when a competitor builds something better. That is not FICO's moat. VantageScore, built by the three credit bureaus themselves, has existed since 2006 and is, by most technical accounts, a perfectly good model. For nineteen years it made almost no difference to FICO, and not because it was worse. It made no difference because the market was coordinated on FICO, and a good alternative that nobody else has adopted is not an alternative anyone can use.

A coordination lock says something different: the product is the standard, and the standard holds because everyone is on it. Its strength does not come from being better and is not undone by a rival being as good. It comes from the cost of moving while everyone else stays, and that cost is enormous, because the party who switches alone loses compatibility with the entire market that did not.

Three things follow from that, and the third is the one that matters most for this file.

The lock produces exceptional pricing power, because a locked-in user cannot walk away over price. FICO has raised the mortgage price many times over and lost no volume.

The lock is extremely durable against competitors, because building a better model does not touch it. Capability was never the binding constraint. VantageScore proved that over nineteen years.

And the lock has exactly one way to fail, which is the mirror image of how it holds. It holds because everyone is coordinated on it. It fails if a party with the power to move the whole market at once chooses to. No single lender can break it. No competitor can break it by being better. Only a coordinator can, by admitting an alternative for everyone simultaneously, so that switching no longer means switching alone.

The asymmetry that defines it, and where it differs from a great moat

Hold FICO's mechanism against the strongest kind of moat, the sort where the company controls its own fate. In that kind, the only party who can damage the moat is the company itself, by losing its own discipline. There is no outside actor to worry about, which is what makes the security unusually high.

FICO is not that kind, and this is the single most important fact in the file. The coordinator of FICO's market is not FICO. It is the Federal Housing Finance Agency, through the mortgage giants it oversees, because they set the requirement that coordinated everyone onto the FICO Score in the first place. The party that built the lock is the party that can open it, and that party sits outside the company and outside anything the company controls.

This is why the security rating is moderate rather than high, and it was moderate before anything actually happened. A moat whose coordinator is external is a moat with a door in it, whoever happens to be holding the key.

Reinforcing: regulatory and contractual embedding

The lock is deepened by how thoroughly the FICO Score is written into the plumbing. GSE selling guides required it. Bank capital rules reference it. Securitisation documents describe pools in its terms. Decades of lending contracts and systems assume it. Each of these is a place the standard is nailed down, and together they are why even an admitted competitor does not displace FICO overnight: the embedding has to be unpicked one document and one system at a time.

This layer cuts both ways, and honesty requires stating both. It slows any competitor down enormously, which is why the current impairment has not transmitted to the numbers. But it is also the layer the regulator acts through, because the same guides that embed FICO are the guides the FHFA can amend. The depth that protects the lock is the depth the coordinator edits.

Reinforcing: the standard as a language

Credit quality across the US market is spoken in FICO. A 740 means something to everyone in the chain without further explanation. A competing score, even an equally good one, has to be translated into that shared language before anyone can act on it, and the lender pilots now underway show exactly this friction: one large lender applies a fixed points adjustment to VantageScore results to keep them comparable to the FICO numbers everyone already understands. That translation cost is a real part of the moat, and it buys FICO time even where a competitor is now admitted.

Reinforcing: consumer data and brand

"Your FICO Score" is a thing consumers know and check, sold to them directly. This is the softest of the reinforcing layers and the least load-bearing, but it deepens the default: a score consumers recognise and ask for is a score lenders have one more reason to keep using. It does nothing against a regulator and is noted as reinforcing rather than protective.

Optional: the Software segment

The FICO Platform is a genuine and growing business, and in FY2026 it grew well, with platform ARR up sharply. It is explicitly not part of the moat. It competes against other analytics and decisioning vendors on ordinary terms, with no coordination lock and no standard-setting position. It is recorded here so that its growth is never read as evidence about the Score. A reader watching total revenue could be comforted by Software strength while the Score mechanism erodes underneath, which is exactly the confusion this file exists to prevent.

Governance, as it bears on the moat

Governance is a normal US public-company structure, and this is itself a finding. Tight founder or family control is not the relevant defence here, and its absence is not a flaw, because FICO's failure mode is not internal indiscipline. No governance structure, however tight, protects a company against its market's coordinator changing the rules. Governance provides no guard against the one thing that actually threatens this moat, and saying so plainly is more useful than scoring it as though it did.

Evidence of strength: what nineteen years of VantageScore proves

The best evidence that this moat is a coordination lock and not a product advantage is that a good competitor spent nineteen years unable to make a dent in it. VantageScore was built by the bureaus, was technically competitive, was priced aggressively, and still went almost nowhere in mortgage. That is not the record of a company winning on quality every year. It is the record of a market that was coordinated onto one standard and stayed there regardless of the alternative's merits.

This evidence is double-edged, and the file has to read both edges. It proves the lock was extraordinarily strong. It also proves that the lock never depended on capability, so the arrival of the same competitor through a regulatory door, rather than through the market, is a completely different event from the nineteen years that preceded it. The thing that changed in 2025 is not that VantageScore got better. It is that the coordinator opened the door.

Evidence of strength: pricing power

FICO's pricing record is the clearest quantitative proof the lock has been real. A mortgage score that cost cents now costs several dollars, raised in steps over a few years, absorbed each time with no volume loss. Outside mortgage, in the auto and card and consumer channels where no regulator mandated the score, FICO also holds pricing power, and this matters for the verdict: the impairment is specific to the mortgage channel, where the coordinator acted, and does not currently extend to the channels the coordinator does not control. Pricing authority is therefore constrained in one channel and unconstrained in the others, which is why the verdict splits them rather than blending them into one figure.

Evidence of strength: the financial fingerprint

The numbers are the fingerprint of a coordination lock operating at full strength, and they are worth stating precisely because the argument of this file is that they are lagging indicators.

Q3 FY2026 total revenue was $674m, up 26% year on year. Scores revenue was $459m, up 41%, with B2B Scores up 49% driven mostly by higher mortgage origination unit prices, so mortgage now carries the bulk of the segment's growth. Non-GAAP operating margin reached 62%, up from 57% a year earlier, and non-GAAP EPS was $12.18, up 42%. Free cash flow was $370m in the quarter and $961m over the trailing year, and the company bought back $1.96bn of stock at an average of $1,149 a share. The full-year guidance was raised. On the standard, everything an investor could ask for is present.

One caution governs all of it, and it is the central point of this document. These figures measure the moat's output, not its mechanism. The mortgage price increases landing before the competitor's admission has had time to bite is not evidence that the admission did not happen. A coordination lock that has just been breached at one entry point looks, in the quarter of the breach, exactly like a coordination lock in perfect health, because the embedding that protects it also delays the transmission. The strength of the numbers is not evidence against the impairment. It is what the impairment looks like before it transmits.

Alternative explanations

A moat claim is only worth anything if the competing explanations fit the data worse.

It is simply the best model. The nineteen-year VantageScore record rejects this directly. If FICO won on quality, a competitive model priced below it would have taken share over nineteen years, and it did not. The moat is coordination, not accuracy, which is precisely why a regulatory door matters more than any model comparison.

The recent growth proves the moat is intact. This is the explanation the file most wants to resist, because it is the most tempting and the most wrong. The growth proves the lock was intact through the period the revenue was earned. It says nothing about the mechanism going forward, because the event that threatens the mechanism works on a lag the embedding imposes. Reading current strength as evidence of future durability is the exact error this framework is built to avoid.

The impairment is priced, so it must be real. The price is not evidence here and this file does not use it. The share price has fallen a long way, but the price is a Layer 2 input and the moat is judged without it. The impairment is established from the regulatory action and the mechanism, not from the drawdown.

The preferred explanation is that FICO's moat is a coordination lock built by an external coordinator, exceptionally strong against competitors and against price pressure, and vulnerable to exactly one thing: that coordinator opening the market to an alternative. That is the account consistent with the nineteen-year VantageScore record, the extraordinary pricing power, the current strength of the numbers, and the 2025 regulatory action all at once.

3. What could break it

3a. Who can break it

The register follows from one question: who takes the decision that ends this moat, and would their action be visible in the numbers?

Mechanism	Actor	Visible in the numbers?
Coordination lock	The FHFA and the GSEs	No. It arrives as a rule change, not a result
Contractual embedding	Lenders and the GSEs, over years	Only late, as the embedding is unpicked
Pricing in mortgage	The coordinator, by enabling substitution	Late, after switching actually occurs
Pricing outside mortgage	Competitors, if a lock ever loosens there	Would show as slowing Scores growth
The standard as language	Erodes only as the market re-learns to speak another	Very late

Every row that matters says no, or late. That is the defining feature of this moat and the reason it needs a register built on the regulator's actions rather than on FICO's numbers. The actor who can break it sits outside the company, acts through rules rather than through the market, and their action shows up in the income statement years after it is taken, if at all. A clean quarter is not evidence the mechanism is safe. It is evidence the lag has not yet elapsed.

Note the contrast worth carrying: in a company-controlled moat the dangerous actor is inside the business and their loss of discipline would eventually surface in the accounts. Here the actor is a regulator, their decision surfaces in a Federal Register notice rather than a revenue line, and no amount of financial monitoring would have caught it. The condition that caught it is a condition that watches the coordinator.

3b. Class A: mechanism conditions

These describe events. Each has an actor and a date, and none of them can fire because of a cyclical downturn in lending volumes. A trigger here is a rejection of the moat's full strength. The reasoning behind the class split is in the annex.

#	Condition	Actor	Observable event	Reading now
A1	A competing score is admitted into the mortgage channel for the whole market	FHFA / GSEs	A rule or announcement permitting an alternative in GSE-eligible loans	TRIGGERED. VantageScore 4.0 admitted July 2025; interim phase live through 2026. This is the impairment
A2	The score requirement is removed entirely, not merely opened to a competitor	FHFA / GSEs	A guide change dropping the mandated-score requirement	Clean. The requirement persists; a second score was added, not the requirement removed
A3	A competitor achieves real mortgage adoption, not merely admission	Lenders, with their own money	Meaningful share of GSE-delivered loans scored on the alternative	Clean so far. Adoption is early and piloted; most lenders still use FICO. This is the condition to watch
A4	The lock loosens outside mortgage, in auto, card or consumer	Competitors or a second regulator	A second channel opening to an admitted alternative	Clean. No comparable action in the non-mortgage channels
A5	Pricing is capped or reviewed by a regulator	Legislature or regulator	A price control or formal review of score pricing	Clean, but proximate. The pricing has drawn political attention alongside the competition push
A6	The score loses standing in securitisation and capital rules	Regulators, ratings agencies	Pool documentation or capital rules ceasing to reference FICO as standard	Clean. The embedding here is deep and unchanged so far

On A1 and A3, the distinction that carries the whole verdict. A1 has fired: a competitor has been admitted for the whole market at once, which is the precise event a coordination lock cannot survive indefinitely, because it removes the "switching alone" cost that held everyone in place. But admission is not adoption. A3 has not fired, because a lender switching still faces the translation cost, the systems cost and the untested-model risk, and most have not moved. The nineteen-year VantageScore record is the reason to take A1 seriously despite A3 being clean: the only thing that ever protected FICO was that nobody could switch together, and A1 is the removal of exactly that protection. But it is also the reason not to over-read A1: the same embedding that VantageScore could not overcome for nineteen years does not evaporate because a door opened. The honest reading is a mechanism impaired at the entry point, with the impairment not yet transmitted, and A3 as the gauge of whether it will.

One piece of company evidence sharpens A3, and it has to be weighed rather than taken at face value. FICO's management argues the admitted competitor is additive rather than substitutive: under the GSEs' "Lender's Choice" policy a lender pulls both scores and uses whichever gives the borrower a better rate, so to game the system a lender needs both, which keeps FICO in every file rather than displacing it. On that logic the theoretical ceiling for the competitor's share is around 20%, not the whole channel. If that holds, A3 can never fire in the strong form the condition imagines, because "real adoption" is capped well below the level at which the market coordinates away from FICO. This is management's own framing and cannot simply be trusted, since it is exactly what an incumbent would say, but it is testable: B4 measures whether the competitor's share climbs toward that 20% ceiling or past it. Past it would mean the additive story is wrong and the substitution is real. Short of it, the impairment stays what this file calls it, an opened door the market has mostly declined to walk through. FICO is meanwhile pushing its own newer score, 10T, now adopted by 70 lenders covering roughly 55% of top-50 originator volume, which is the incumbent deepening the embedding while the door stands open. This file does not rest its reading on that ceiling. The impairment is judged from the regulatory action and the low observed adoption, both facts; the 20% is management's explanation for why adoption may stay low, carried as a claim to test, not as a premise.

Calibration. A1 is the one condition in this file with a clean, dated, unambiguous trigger, which is rare and valuable. A3 is well-defined but its threshold is a matter of judgement: how much adoption is "real" is not a number anyone can set confidently in advance, and it should be read as a direction to watch closely rather than a precise line. A5 and A6 are reasoned rather than observed, because there is no precedent for a score of this standing being price-capped or stripped from capital rules. The file does not claim precision it has not earned on those.

3c. Class B: gauges

These are measurements. A number has two causes, the mechanism and the lending cycle, and the number alone cannot tell you which moved. A gauge never rejects the moat on its own; it obliges the attribution test in 3e.

Every gauge here currently reads clean, and the whole point of the file is that this is expected, not reassuring.

#	Gauge	What it isolates	Expected direction if the mechanism erodes	Reading now
B1	Mortgage score volume on FICO vs total GSE originations	Is the lock still capturing the whole channel?	FICO's share of scored loans falls while total holds	Clean. FICO still scores the overwhelming majority
B2	Mortgage price realisation	Is pricing power in the locked channel holding?	Realised price per score falls as lenders gain a cheaper option	Clean. 2025 increases landing; revenue up
B3	Non-mortgage Scores growth	Is the lock intact where no regulator acted?	Auto, card, consumer growth slows	Clean. Broad Scores strength
B4	VantageScore mortgage adoption rate	The direct measure of A3	Rises from near zero	Near zero, and capped near a theoretical 20% on management's read. The one to watch; past 20% would break the additive story
B5	Scores revenue vs Software revenue mix	Is the profit engine still the Score?	Score revenue share falls	Clean. Score still carries the profit
B6	Pricing commentary from lenders and the GSEs	Early signal before volumes move	Public pressure to switch, or a GSE nudge toward the alternative	Watch. The savings case is being made publicly

Why B2 reads the opposite of impairment right now. Mortgage score revenue jumped on higher unit prices. On a naive read that is the healthiest possible signal. In this file it is close to meaningless as evidence about the mechanism, because it is the 2025 price increases landing on a channel the competitor has not yet penetrated. A coordination lock breached at the entry point but not yet at scale produces record revenue in the same window as the breach. B2 will only become informative when, and if, adoption under A3 actually forces price realisation down. Until then it measures the past strength of the lock, not its future.

B4 and B6 are the live ones. Everything else is a lagging confirmation of a lock that was strong through the reporting period. B4 measures whether the admitted competitor is actually being used, and B6 listens for the pressure that precedes it. Those two are where the erosion, if it comes, shows up first.

3d. Comparator sets

Both sets are fixed here, in advance, so that a later reading cannot pick the comparison that suits the conclusion.

Macro peers, which share the lending cycle and separate a mortgage-volume slowdown from a mechanism problem: mortgage originators and title insurers whose fortunes track origination volumes. If FICO's mortgage revenue falls while these fall too, it is the cycle; if FICO's falls while they hold, it is the mechanism.

Mechanism peer, which tests the lock directly: VantageScore. This is the one that matters, and its absence from a naive analysis is exactly the FICO-specific trap. A file that watched only competitors "winning on features" would never have flagged VantageScore, because for nineteen years it won nothing on features and still became the vehicle for the impairment. The mechanism peer is not the company that builds a better model. It is the company positioned to walk through a regulatory door if one opens, and that company must be named in advance and watched regardless of how its product compares.

3e. The attribution test

Run this every time a Class B gauge moves. All three must come back clean for the cause to be recorded as cyclical rather than mechanical.

1. Is there a nameable external cause with a date? For a mortgage-revenue fall, is it a rate-driven origination slump, or is it substitution to VantageScore?

2. Do the macro peers move with it? If originators and title insurers fall together, the cause is the lending cycle. If FICO falls alone, it is the mechanism.

3. Is the mechanism side unchanged? Is FICO's share of scored loans holding, or is it losing the channel to the admitted competitor?

   
   

The escalation rule is different from a fixed counter of quarters. A cyclical attribution holds only while the named cause is present and verifiable. When mortgage volumes recover and FICO's mortgage revenue does not, or when the peers recover and FICO does not, the cause is no longer the cycle and the matter escalates to a Class A judgement regardless of the calendar.

One caution specific to FICO. The lag is the whole problem here. Because the embedding delays transmission, a genuinely eroding mechanism can pass the attribution test for several years, since the numbers keep clearing while adoption builds quietly underneath. That is why A1 is already treated as an impairment on the event alone, without waiting for the gauges to confirm it. Waiting for B-gauge confirmation on a moat with this much lag would mean recognising the damage only after it is irreversible.

How this document is revised. On any FHFA or GSE action touching credit-score policy, whatever the calendar. On any disclosure bearing on mortgage adoption of a competing score. On the quarterly and annual reports. And on anything unforeseen where the question of whether to revise even arises. The scheduled next revision is the Q4 FY2026 report, but a regulatory action would pull it forward, because for this moat the regulator's calendar matters more than the company's.

4. What cannot be seen

Two things carry real weight and have no clean, timely signal. Listing them stops "one condition triggered, the rest clean" from being read as "mostly fine."

The speed of adoption once it starts. A coordination lock does not erode linearly. It holds, holds, holds, and then, if enough of the market moves that switching no longer means switching alone, it can go quickly, because the same "everyone else is on it" logic that held it together starts working in reverse. There is no gauge that reliably gives warning of the inflection, because the inflection is a change in everyone's expectation of what everyone else will do. A3 and B4 are the best available watch, but they measure adoption after it starts, not the tipping point.

The political direction of travel. The 2025 admission was framed as competition and consumer savings, and that framing has political momentum behind it. Whether it extends further, to removing the requirement entirely, to price caps, or to the non-mortgage channels, is a political question with no observable trigger until it happens. It is carried as a standing uncertainty, not a monitorable variable.

Two structural limits are worth stating plainly. A5 and A6 have no precedent to calibrate against, so their thresholds are reasoned rather than observed. And this is a moat whose single point of failure is external and has already been touched once, which is a materially different situation from a moat whose failure mode is internal and untriggered.

5. Assumptions

#	Assumption	Status
1	Lenders continue to need a standardised, market-wide credit score	High confidence. The need for a common language is not in question; which score fills it is
2	The FICO Score remains the standard the market coordinates on	The central uncertainty, now under active challenge in the mortgage channel
3	The embedding continues to slow any competitor's adoption	High confidence near term, declining over the years as the interim phase matures
4	The non-mortgage channels stay locked, with no regulator acting there	Moderate to high confidence. No action so far, but the mortgage precedent exists
5	FICO retains pricing power where the lock holds	High confidence outside mortgage, now qualified inside it
6	No further regulatory action removes the requirement or caps the price	Moderate. The direction of political travel is the risk

6. Basis of this assessment

This is the first Layer 1 written on FICO, so there is no prior tier to move from. It records the starting position that future revisions will read against.

The moat is judged a coordination lock, strong against competitors and price pressure but dependent on an external coordinator. Strength reads strong rather than exceptional on that dependency alone, before condition is scored. Condition A1 is triggered: the FHFA admitted VantageScore 4.0 into GSE-eligible mortgages in July 2025, with the interim phase live through 2026. That is an impairment of the foundational layer at one entry point, which sets condition to impaired, security to moderate, and pricing authority to constrained in mortgage while unconstrained elsewhere. The verdict is Strong / Impaired.

Every Class B gauge reads clean, and this is recorded as expected lag rather than reassurance: the embedding that protects the lock also delays the transmission of the damage, so strong current numbers are exactly what an impaired-but-not-broken coordination lock produces. A3 (real adoption) and B4 (measured adoption rate) are the live watch, and the trigger that would move condition from impaired to broken.

On the third gate, demand anchoring reads level 2: the need to judge a borrower's creditworthiness is permanent wherever lending exists, so it falls with the credit cycle but always returns. FICO passes that gate. It does not matter for the decision, because the name is already declined at the first two gates, strong not exceptional and impaired, but it is recorded so the file carries all three axes.

Future revisions are dated and appended below.

Annex: how this assessment is made

These are the rules the document is written under, kept separate so the file above stays about FICO and the rules cannot quietly change to suit a conclusion.

Judging strength

Strength is settled before condition, because the condition axis only measures whether a moat is undamaged, not whether there was much of a moat to begin with. Three questions, all about how hard the moat is to attack rather than how well the company is trading. Is there a substitute the market could actually move to? Could a competitor replicate the position? Has it been attacked, and what happened?

FICO's answers are unusual and have to be read carefully. For nineteen years there was a substitute the market could not move to, a competitor that replicated the product and got nowhere, and an attack that failed completely. On that record the strength was exceptional. But the strength rested entirely on an external coordinator keeping the market together, which means it was always exceptional-with-a-dependency rather than exceptional outright. That dependency is why FICO reads strong rather than exceptional even before condition is scored, and it is why the 2025 event lands so decisively: the one dependency the moat had was the one that gave way.

Exceptional means all three answers come back clean and the strength is self-controlled. Strong means one is soft or the strength depends on an outside party. FICO is strong rather than exceptional on that definition, even before condition is scored, because the coordinator is external.

The Layer 2 gate requires three things together: exceptional strength, intact condition, and demand anchored at level 1 or level 2. FICO fails the first two independently, so it is declined regardless of the third. The demand axis is judged and recorded anyway, at level 2, because the framework carries all three axes on every name even when the earlier gates have already settled the decision.

The verdict

The verdict is two judgements held apart, because collapsing them into a single grade hides the thing that matters most. One axis is strength: how deep was the moat to begin with, scored exceptional, strong, or ordinary. The other is condition: is it still whole, scored intact, impaired, or broken.

Keeping them separate is deliberate. A once-exceptional moat that has been damaged and a mediocre moat that happens to be undisturbed can look identical under a single letter, and they are not the same asset. The first can recover its strength if the damage heals; the second has no depth to return to. Strength says how much there was; condition says how much is left.

FICO reads Strong on strength, because the lock was powerful but always depended on an external coordinator, and Impaired on condition, because that coordinator has opened one entry point while the market has not yet moved. Security and pricing authority are the inputs behind the condition reading: moderate security, and pricing constrained in one channel. The verdict is Strong / Impaired.

The condition axis has three steps and the bottom one is terminal. Intact means the mechanism is whole. Impaired means an entry point has been opened but the market has not moved. Broken means the market has actually moved, and a broken foundational layer ends the thesis regardless of how strong the moat once was. FICO sits at impaired, not broken, because the door is open but the market has stayed. The distinction is carried by A3: if A3 fires, impaired becomes broken and the name is out.

Judging demand anchoring

Strength and condition together answer whether the company keeps its share of the market. They do not answer whether the market returns after a fall, and that is a separate question that decides whether a drawdown can be waited out at all. A moat can be fully intact while the demand beneath it shrinks permanently, so demand anchoring is judged on its own axis, on two tests: biology, is the underlying demand rooted in a permanent human or physical necessity or drive, and precedent, is there a history of similar demand collapsing and not returning.

The levels run from one to four. Level 1 is anchored in a physical or biological necessity that never stops. Level 2 is anchored in a permanent human drive that can fall cyclically but always returns. Level 3 is anchored in an established infrastructure or habit with a replacement cycle, probably durable but with no biological guarantee. Level 4 is anchored in a moment, a build-out or a specific medium that may never return to its peak. Only level 1 and level 2 are eligible for a Layer 2 entry, because only there does a drawdown reliably reverse.

FICO reads level 2. The demand it ultimately serves is the need to assess whether a borrower will repay, which exists wherever there is lending and returns after every credit downturn. That need is permanent; what is contested is only which score fills it, which is a moat question, not a demand-durability one. So the demand axis is clean, and the name is declined on the moat axes rather than this one.

Why the conditions are split in two

A Class A condition describes something someone did. It has an actor and a date, and it has one cause: somebody decided. A Class B gauge is a number, and a number has two causes, the mechanism and the environment, so it cannot on its own tell you which moved.

For FICO this split is doing more work than it does for most companies, because the lag between the two is unusually long. A1 fired on a dated regulatory event more than a year ago, while every B-gauge still reads clean. If the file scored on the gauges, it would see no problem. It scores on the events, so it sees the problem a regulator created before the market has enacted it. A Class A trigger is treated as an impairment on the event alone. A Class B move only ever obliges investigation, never a verdict by itself.

When a cyclical explanation expires

"It is the mortgage cycle" will be available as an explanation every time origination volumes dip, and it will often be partly true, since FICO's mortgage revenue does move with originations. The rule is that the cyclical attribution holds only while the named cause is present and verifiable. When originations recover and FICO's mortgage revenue does not, or when origination-linked peers recover and FICO does not, the cycle no longer explains it and the matter escalates. Divergence in the recovery is the sharpest signal, because that is where the shared cyclical cause drops away and only the mechanism is left.

Revision

The document is revised whenever something might have changed, and for this moat that means the regulator's calendar takes precedence over the company's. Any FHFA or GSE action on credit-score policy pulls a revision forward regardless of when the last one was. Class A is walked in full every time, because the entire lesson of this name is that the numbers can be at record highs while the mechanism is being altered by someone who never appears in them. The assignment of a condition to Class A or B is fixed before a trigger, never during the revision that reports one.

A note on what this document is and is not. This is written for the author's own investment process and published in that form. It reflects the analysis, assumptions and judgment of the author at the date of writing and nothing more. Nothing here is investment advice, a recommendation tailored to any reader, or an offer to buy or sell any security. Valuation figures, scenarios and price levels are illustrative analyst assumptions unless explicitly sourced to company reports or other primary data. Investing involves risk, including loss of capital, and past performance is no guarantee of future results. Any reader considering an investment should do their own work and consult a qualified adviser who knows their situation. The author may hold, or may come to hold, positions in the securities discussed.