Fair Isaac sells the de facto standard score the entire US lending stack must use, with embedded distribution through the bureaus and visible per-pull price hikes. The valuation is rich, the leverage is meaningful, and VantageScore's GSE entry is real, but the moat is one of the cleanest on the public exchange.

Plain English

FICO sells the credit score that almost every American lender has to look at before giving you a loan. They charge a small fee every time anyone checks a score, and they have been raising that fee, especially for home loans. The fee is so embedded in how lending works that even though competitors exist, almost no one switches. The business prints cash. The catch is that the government just allowed a competing score for mortgages, the company carries a lot of debt, and the stock is expensive. It is a great business at an okay price.

Thesis

Fair Isaac is two businesses welded together: a software platform that does about half the revenue and a Scores segment that throws off most of the cash. The Scores business is one of the most enviable royalty streams in American finance: every time a US lender pulls a tri-merge credit report through Equifax, Experian, or TransUnion, FICO collects a per-score fee with essentially no incremental cost. This is what shows up as a 28% trailing 10-year ROIC and a 191% incremental ROIC over five years — the marginal dollar invested has compounded extraordinarily because the marginal dollar invested was tiny. FCF conversion of 128% confirms it is real cash, not GAAP creativity. Management has been consistent: no dividend, modest M&A, near-100% of free cash flow into buybacks at high multiples — share count is down only ~4% over a decade because they keep buying with the stock at 40-50x. Net debt of 4.1x EBITDA is uncomfortable but serviceable against contracted, recurring royalty cash. At $1,035, the stock trades at 0.77x our base-case IV ($1,344), inside the IV low ($906) only on a sharper drawdown. EV/FCF of 41.6x and a reverse-DCF implied growth rate of 10.5% mean the market is paying for another decade of mortgage-pull price hikes plus Software platform conversion. Owning FICO under $900 gives a real margin of safety; above bull-case IV of $1,478 the math gets uncomfortable. Today is closer to fair than cheap.

Moat

FICO's moat is among the widest in the public market, but it is narrower in one segment than in the other, and that distinction matters.

Intangibles / regulatory entrenchment (WIDE). The FICO Score is the single specified credit risk standard inside Fannie Mae and Freddie Mac mortgage underwriting, inside most credit card and auto lender pricing grids, and inside almost every securitization document written in the last twenty years. This is not branding in the Coca-Cola sense Damodaran describes [1]; it is the much rarer phenomenon of a private number that has become embedded in regulatory and contractual plumbing. To switch off a FICO Score is to renegotiate thousands of underwriting policies, audit committee charters, regulatory filings, and ABS prospectuses. The 2025 FHFA decision permitting VantageScore 4.0 alongside FICO 10T for GSE-eligible mortgages is the first material crack in this moat in a decade, and is the bear case's single best argument — see inversion. But even after that ruling, every existing securitization, every pricing model, and every covenant continues to reference FICO. Damodaran specifically warns [2] that legal monopolies granted by government can be revoked with the stroke of a pen; FICO's is not a legal monopoly but a coordinative one, which is harder to dislodge because no single regulator owns it.

Switching costs (WIDE in B2B Scores, NARROW in Software). The Scores switching cost is exactly the Microsoft-style switching friction Damodaran describes [2]: every downstream system speaks the language of a 300-850 FICO number, and rebuilding pricing tiers, capital adequacy models, and consumer-facing rate sheets to a different scale is enormously expensive. A $10B competitor cannot duplicate this in five years because the competitor has to convince every lender, regulator, and bondholder simultaneously to re-platform — VantageScore has been trying since 2006 with the bureaus' own balance sheets behind it and still has trivial share. The Software segment (FICO Platform, originations, fraud) faces real, if narrower, switching costs once embedded in a bank workflow, but here it competes with SAS, Pega, Experian PowerCurve, and homegrown systems.

Network effects (NARROW). There is a weak two-sided network in that more lenders using FICO scores creates more standardized data points consumers and regulators reference, reinforcing the standard. But this is not a Visa-style network where each new node makes the system mechanically more useful; it is a coordination equilibrium more than a network.

Cost advantage (WIDE within Scores). Once the FICO model is built and validated, the marginal cost of producing a score from a credit file is essentially zero. The bureaus do the data work and pay FICO a royalty. This is structurally better even than GEICO's scale-driven cost moat described by Buffett [4]: GEICO has to actually deliver a service; FICO ships a number. The Scores segment operating margin reflects this — north of 80% — and is the dominant reason 10-year average ROIC is 28%.

Pricing power (WIDE). FICO has raised mortgage tri-merge wholesale prices materially each year since 2018 — from a few cents per pull to several dollars per pull on certain product tiers — without losing volume. The customers (lenders, bureaus) cannot say no because the alternative is operational chaos. This is the cleanest pricing-power test there is: price up, volume flat, no customer churn. Buffett's GEICO description [4] of widening 'the price advantage we offer customers' is the exact opposite of FICO's playbook, which is to widen the price and pocket it.

$10B / 5-year stress test. A $10B challenger funding VantageScore could plausibly capture 10-25% of new mortgage originations within five years if the GSEs aggressively level the playing field, but cannot dislodge the installed base across cards, auto, and ABS. Securitizations live a decade; pricing grids change slowly. Erosion is real but slow.

Erosion risks. (1) GSE bi-score regime — credible. (2) Direct-to-bureau competing scores eating B2C — already happening but small. (3) AI/ML challengers offering bespoke scoring to the largest lenders, who are sophisticated enough to build their own — slow but real.

Moat verdict: WIDE.

Management

Will Lansing has run FICO since 2012, and his capital allocation playbook over that period has been so consistent it can be summarized in one sentence: take 100% of free cash flow, plus modest incremental debt, and buy back stock — at almost any price.

Reinvestment (B+). Internal R&D and Software platform investment have been material and disciplined. The pivot from a sprawl of point-solution decisioning products to FICO Platform has produced visible ARR growth in Software, and Scores has continued to ship versioned upgrades (FICO 9, 10, 10T, Resilience Index). The 191% 5-year incremental ROIC is the single most flattering number on the scorecard, and almost certainly overstates organic reinvestment quality because it includes the price-hike effect on essentially zero incremental capital — i.e., FICO is not really 'reinvesting' to earn that return; it is harvesting pricing in a fixed-cost business. Still, the platform investments have been legitimate and the Scores R&D has not been starved.

Acquisitions (B). FICO has been a remarkably restrained acquirer relative to peers. There has been no transformative deal, no roll-up, no equity-funded empire-build. This is exactly the discipline Buffett praises in the GEICO and MidAmerican stories [3][4]. The flip side: management has not used the strong currency to buy adjacent franchises that could extend the moat into a post-FICO-score future.

Debt (C). Net debt to EBITDA at 4.1x is meaningfully levered for a software-and-royalty business that the market values at 41x EV/FCF. The debt was not taken to fund a strategic acquisition or capacity expansion — it was taken to fund buybacks. In a stable interest-rate environment, against contracted recurring revenue, this is acceptable. In a regime where mortgage origination volume halves (2022-2023 was a rehearsal) and a refinancing wall hits, this leverage will look much less benign. Interest coverage was not provided in the scorecard, which is a flag worth investigating before sizing.

Buybacks (C+). This is where the analysis gets uncomfortable. Share count is down only 4.0% over ten years despite essentially 100% of free cash going to buybacks plus incremental debt. Why? Because management has been buying at progressively higher multiples — including $400, $700, $900, and now $1,035 territory — against a 10-year average P/E of 56.9 and a current TTM P/E of 44.3. The buybacks have been more about offsetting equity comp and absorbing capital than about opportunistic value capture. Average P/IV on buybacks has almost certainly been north of 0.85, meaning management is paying a thin or negative spread to intrinsic value. Buffett would not approve of this cadence; he would let cash build and wait. FICO instead pays out at any price, which is structurally fine when the business compounds at 20%+ but turns into value destruction during any multiple compression.

Dividends (n/a). No dividend, which is the right call given the buyback machine and reinvestment opportunities.

Communication (B). Investor materials are clear, segment reporting is clean, and management does not over-promise on Software ARR. They are also somewhat tone-deaf about the political risk of mortgage price hikes — they have publicly defended price increases on consumer-facing scores in ways that have drawn explicit Senate and FHFA attention.

Capital allocator: B.

Industry

FICO sits in a one-of-one industry: standardized consumer credit scoring sold to regulated lenders. Porter's Five Forces:

Threat of new entrants (LOW for Scores, MEDIUM for Software). New scoring entrants face the chicken-and-egg problem of needing simultaneous lender, bureau, regulator, and securitization-market adoption. VantageScore — funded jointly by Equifax, Experian, and TransUnion since 2006, the three companies with the most leverage to push an alternative — still has minor share outside auto and B2C. That is the strongest possible evidence of how high the entry barrier is. In Software/decisioning, entry is much easier; FICO competes with SAS, Pega, Experian, and increasingly cloud-native ML vendors and customer-built systems.

Bargaining power of buyers (LOW in Scores, MEDIUM in Software). Lender buyers of FICO Scores have essentially no leverage on the Scores side because the alternative — switching to VantageScore — requires re-platforming pricing, regulatory, and securitization systems. The largest lenders (JPM, BAC, COF, the GSEs) have nominally tried to push back on price hikes and have succeeded only in slowing them, not stopping them. On the Software side, large bank buyers absolutely have power and exercise it — RFP cycles are competitive and FICO loses deals.

Bargaining power of suppliers (LOW). FICO's primary 'supplier' is the credit bureaus, who provide the underlying credit file data. The relationship is symbiotic — bureaus need FICO's brand to maximize the value of credit pulls — but not adversarial. Bureau royalties are a cost of revenue, not a strategic chokepoint. Recent disclosures show FICO has been raising prices to bureaus, not the other way around.

Threat of substitutes (MEDIUM and rising). This is the most important force to track. Substitutes include: VantageScore (now GSE-blessed for mortgages), bespoke lender-built ML models, alternative-data scores (LexisNexis, ZestFinance, etc.), and open banking-driven cash-flow underwriting. Each has captured slivers, but none has captured a regulated-mortgage market until 2025. The substitute threat is not zero anymore — it has stepped up.

Industry rivalry (LOW in Scores, HIGH in Software). Scores is essentially a monopoly with a ceremonial competitor. Software is a knife fight.

Value pool location and trajectory. The value pool sits with FICO and to a much lesser extent with the bureaus and VantageScore. Lenders capture economic value from using the scores, not from owning them. The pool is growing because (a) FICO is raising per-pull pricing faster than mortgage volumes shrink, (b) Software platform consolidation is genuine secular tailwind, (c) the global expansion into 40+ countries is a real but slow optionality.

Trajectory. Stable-to-improving for 5 years; uncertain past year 7 as the GSE bi-score decision plays out and as ML-native lenders mature. The unique business — high-margin standardized score business with regulatory entrenchment — has no real comparable on the public exchange. Closest analogues (S&P Global ratings, Moody's) face their own regulatory pressure but do not face a sponsored direct competitor.

Industry Verdict: Excellent.

Inversion

I am short FICO. Here is why this stock could be cut in half.

1. The single event that kills this: a fast GSE migration to VantageScore plus an FHFA price cap. The November 2024 FHFA decision allowing FICO 10T and VantageScore 4.0 in GSE mortgage underwriting was framed by the bull case as a non-event because lenders will move slowly. That framing is wrong. It is a one-way ratchet. The moment a single major mortgage originator (Rocket, UWM, JPM) chooses VantageScore for new originations to capture lower per-pull cost, the rest of the industry follows in 18-36 months because no large lender wants to be on the wrong side of a regulatory and cost trend. FICO loses 30-50% of mortgage Scores volume in three years. Worse, the FHFA has already publicly criticized FICO's mortgage price hikes by name, and a Democratic administration plus a populist Republican administration have both shown willingness to price-cap administered standards. A direct cap on per-pull mortgage scoring fees — say at 50 cents — would erase the $4-5 per pull pricing FICO has built into its 2026 plan and crush Scores segment EBITDA.

2. Why the moat is narrower than bulls think. Bulls confuse 'embedded in contracts' with 'cannot be replaced.' But every individual contract that references FICO is renegotiable; the friction is finite, not infinite. The bureau JV that owns VantageScore — Equifax, Experian, TransUnion — is the single most powerful coalition that exists in American consumer credit. They have been quietly distributing VantageScore at much lower prices for nearly two decades, building installed base in card prescreens, auto, and B2C. The mortgage carve-out was the last domino, and once mortgage moves the rest follows at compounding speed. The Software segment, which bulls cite as the diversifier, competes against SAS, Pega, Experian PowerCurve, and increasingly cloud-native ML platforms — there is no moat there, just a crowded mid-tier enterprise software business that should trade at 6-8x revenue, not the 12x+ implied by the consolidated multiple.

3. Why management is worse than it appears. Lansing has run a textbook value-extraction playbook: raise prices on a regulatory-protected good, lever up to fund buybacks, repurchase shares at any multiple. The 4% net share count reduction over ten years against essentially 100% FCF + incremental debt deployed to buybacks is not capital allocation excellence — it is paying full freight for the company's own equity and offsetting compensation dilution. The leverage (4.1x net debt / EBITDA) was not taken to invest; it was taken to financialize. When the price-hike algorithm slows or reverses under regulatory pressure, the leverage stops looking conservative. Worse, management has shown zero strategic acquisition discipline of the 'use a strong currency to buy a future' kind — there is no obvious post-FICO-score adjacent franchise being built. They are harvesting, not investing.

4. What bulls are extrapolating that won't hold. The bull deck assumes (a) mortgage per-pull pricing keeps growing 15-25% per year for another five years; (b) Software ARR compounds at 25%+; (c) the multiple stays at 40-50x EV/FCF. All three are unlikely simultaneously. Mortgage pricing will mean-revert under FHFA scrutiny. Software ARR has decelerated in recent quarters — the most recent prints are closer to 15% than 25%. The multiple is the most fragile: at TTM P/E of 44.3 versus a 10-year average of 56.9, the market is already compressing the multiple. The reverse-DCF implied growth rate of 10.5% is not crazy on its own, but it bakes in continuation of exactly the conditions that the FHFA decision threatens.

5. Valuation trap and regime change. EV/FCF at 41.6x against 4.1x leverage means small fundamental disappointments produce outsized equity drawdowns. Owner earnings of ~$657M against an enterprise value of ~$28B leaves no cushion for either margin compression or volume loss. A re-rating from 41x to a Moody's-like 25x EV/FCF — entirely consistent with how the market has historically priced regulatory-cracked franchises — combined with a 15% cut to forward owner earnings (very plausible under bi-score pressure) yields equity value of roughly $400 per share. Even a less dramatic 30x multiple on flat owner earnings yields $480. The IV-low of $906 in the scorecard does not contemplate a regime change; it contemplates a normal multiple cycle. A regime change is the bear case.

If I am right, the stock could be worth $450 within three years.

Lollapalooza Bias Check

Several biases are actively pulling on me as I write this.

Authority bias / canon worship. I started this analysis already half-believing FICO is a Buffett-grade compounder because the qualitative story — toll bridge, monopoly score, embedded in regulation — pattern-matches to Moody's, S&P, and Visa, all of which Buffett has owned or praised. The danger is that I am pricing the franchise off the narrative match rather than the numerical match. The numerical match argues for caution: Moody's traded at 25-35x post-Dodd-Frank, not 41x, and Visa's multiple expanded with growth FICO is not delivering at the same rate. I have to actively discount the authority pull.

Anchoring to the IV range. The scorecard hands me an IV-base of $1,344 and a current price of $1,035, which produces a clean 0.77 P/IV ratio that anchors me toward 'modest discount, decent buy.' But the IV range was widened by the scorer's own note that maintenance capex is uncertain by more than 50%, which means the real IV distribution is much wider than the printed range. I should be treating the printed IV-low of $906 as the upper end of a more conservative range, not the lower end of a normal range. The anchor is making me think the stock is closer to fair value than it probably is.

Recency / ignoring the FHFA decision. The FHFA bi-score decision is recent, ongoing, and its commercial implications are still being negotiated. Markets have not fully priced it. I have a tendency to assume 'the market knows' and price the news as resolved, when in fact this is the kind of slow-motion regulatory event that takes 24-36 months to fully unfold. I am underweighting it.

Confirmation bias from the moat narrative. The moat is genuinely wide, and that fact makes me want to wave away the leverage and the mediocre buyback discipline. I have to remind myself: the canon excerpts about GEICO [4][5] describe a company whose management was widening the customer price advantage. FICO's management is doing the opposite — widening the price and pocketing it. That is a different kind of business than the canon idealizes, and the 'wide moat' reflex should not paper over the difference.

Commitment bias from running the pipeline. Having spent 5,000 words writing about why this is a great franchise, I am psychologically committed to a bullish recommendation. The inversion exercise is the structural correction — I gave it real weight and a $450 target, not a token bear case. The commitment bias would otherwise push me to 'Buy' when the integrated picture is closer to 'Hold' or 'Trim-to-watch.'

Incentive bias I do not have but should note. I am not paid on this call, but most published FICO research is from sell-side analysts whose firms underwrite FICO debt and equity, and their price targets cluster in the $1,300-1,600 range. I should not anchor on their consensus.

10-Year Outlook

Same fundamental business model in 2036? Mostly yes. The Scores segment will still sell standardized credit scores to lenders and bureaus, though the share of revenue from mortgage tri-merge pulls will likely be lower and the share from credit card, auto, and international scoring higher. The Software segment will be larger as a share of revenue and will look more like a typical decisioning-platform business (modular, cloud-native, ARR-priced). The blended business will be less of a pure royalty story.

Customer base larger? Yes. Lender count grows slowly with the economy, but international expansion is real — FICO is in 40+ countries with meaningful traction in Mexico, Brazil, Turkey, and parts of Asia. Software customer count grows faster than Scores customer count.

Profit per customer higher? Probably yes in Scores (price hikes continue at slower rates), maybe in Software (depends on platform consolidation success), but the blended number could be lower if Software grows faster than Scores because Software margins are structurally lower.

Moat wider? No, narrower. The bi-score regime is the single biggest fact and it will erode mortgage Scores share over a decade. The Software moat is real but narrow. Net: moat at 80th percentile in 2036 versus 95th percentile today.

Single biggest threat? Coordinated regulatory pricing pressure on consumer-facing scoring fees, combined with VantageScore mortgage uptake in the 25-50% range. Secondary: AI-native bespoke scoring at the largest lenders, who eventually internalize what they buy from FICO.

The math 10 years out. If owner earnings compound at 8-10% annually (reasonable mid-case given Software offsetting Scores deceleration), 2036 owner earnings are roughly $1.4-1.7B. At a Moody's-rhyme multiple of 25-30x, that is an enterprise value of $35-50B versus today's ~$28B. Equity returns of 4-7% annually plus a couple points of buyback yield on a starting $1,035 entry. Acceptable but not exceptional; below the buyer's required return for a 4x-leveraged single-product franchise.

The business will still be here. The compounding rate will be lower than the trailing decade. The price you pay matters more than usual.

CONFIDENCE: medium

Position Guidance

• Recommendation: Hold
• Conviction: Medium
• Target buy price: $880 (below IV-low of $906, gives ~25% margin of safety against base IV)
• Target trim price: $1,480 (just above bull-case IV of $1,478)
• Position sizing: If initiating below $880, size to 3-5% of portfolio. Maximum 7% even on conviction add. The 4.1x leverage and the active regulatory overhang argue against a larger position regardless of price. Never let it grow above 8% of portfolio through appreciation — trim back into bull-case IV.
• Watch items: (1) FHFA quarterly disclosures on VantageScore mortgage uptake, (2) FICO mortgage pull pricing announcements, (3) Software ARR growth rate sustaining above 18%, (4) net debt / EBITDA trending toward 3x, (5) any sign management slows buybacks at elevated multiples and lets cash build.