# Fair Isaac Corporation (FICO) — InvestMoat Analysis

_Last analyzed: July 31, 2026_
_Asset class: equity · Canonical page: https://investmoat.com/stocks/fico_

## Scores

| Dimension | Score (0–100) |
| --- | --- |
| Moat durability | 85 |
| Growth trajectory | 79 |
| Valuation | 76 |
| **Composite** | **82** |
| **Recommendation** | **Strong Buy** |

Scores are computed deterministically from this asset’s data by the InvestMoat formula (see https://investmoat.com/llms.txt for methodology). Scores are not directly comparable across asset classes.

## Key stats

- **Ticker:** FICO
- **Market Cap:** ~$26B

## Moat

FICO remains the dominant credit scoring standard in the U.S., but two things have changed the character of the moat since the last review. First, the regulatory architecture is fully dismantled: FHFA approved VantageScore 4.0 for GSE loans (July 2025), Fannie Mae eliminated minimum credit score requirements entirely (November 2025), and VA, FHA and USDA all now permit 24 months of verified rent and utility history as primary creditworthiness evidence. Second — and this is the new fact — FICO stopped defending the price. In April 2026 it re-cut the FICO Score 10T mortgage direct licence from $4.95 per score plus $33 per funded loan to $0.99 per score plus $65 per funded loan, matching VantageScore's headline price after TransUnion and the other bureaus moved to 99¢. The 10× price gap that framed last quarter's analysis no longer exists in either direction. What that buys FICO is price parity at the pull — the cost argument for building a VantageScore pipeline is gone — and what it costs is the per-pull royalty model itself: mortgage economics now sit in a success fee collected only when a loan closes, which makes the most profitable franchise cyclical in a way it has never been. FICO still commands roughly 90% of B2B score pulls, and management says it has seen no volume loss to lenders adding VantageScore. But on the Q3 call the CEO described the emerging market as one of 'score shopping', where lenders pull more than one score per file — the first management acknowledgment that the reference standard is becoming bilateral.

### The Toll Collector Gave Up the Toll

FICO's moat has three layers. The first has now absorbed both a regulatory breach and a self-inflicted price reset; the second and third remain intact but are no longer compounding as cleanly as the ARR headline suggests:

- **GSE Mandate — Exclusivity Gone, and the Price Defence Withdrawn:** On July 8, 2025, FHFA Director Bill Pulte approved VantageScore 4.0 as a lender-choice alternative to Classic FICO, ending FICO's exclusive GSE mandate after nearly three decades. The breach has since widened to every major federal mortgage channel: Fannie Mae eliminated its minimum credit score requirement in November 2025, stating it would make risk analysis 'agnostic of third-party credit scores'; VA and FHA now permit 24 months of verified on-time rent and utility payments as a primary creditworthiness indicator; USDA participates in the same framework. Senator Hawley opened a Senate investigation on March 24, 2026 and referred FICO to the FTC, citing the doubling of the per-score price from $4.95 to $10.00 for 2026 — a change he estimated would add roughly $500M to industry credit-score costs. FICO's answer in April 2026 was to reprice rather than litigate: FICO Score 10T under the mortgage direct licence moved to $0.99 per score plus a $65 fee collected only on funded loans. That defuses much of the pricing complaint and removes the lender's financial incentive to integrate an alternative, but it also concedes the $10 price point that produced the +97% YoY mortgage origination revenue reported in Q3 FY2026. The pricing lever that drove the last four quarters has been spent, and the FY2027 comparison laps it.
- **Transaction Embedding — Intact, but Now Levered to Closings:** FICO's Mortgage Direct Licence programme (launched October 1, 2025) licences scores directly to tri-merge resellers, bypassing the bureau markup that historically doubled the end cost to lenders. The embedding itself is undamaged — FICO is still in the transaction layer of essentially every consumer credit decision in the U.S., and Q3 FY2026 mortgage origination revenue grew 97% YoY to 71% of B2B and 62% of total Scores revenue. But the April 2026 repricing changed what FICO earns on. Under $0.99 + $65, the pull is close to free and the economics arrive at funding, so mortgage revenue now tracks closings rather than inquiries. Management said originations and volumes have run ahead of the original FY2026 guide, which is why the guide went up — but that is a rate-and-affordability call, not a royalty. A franchise that used to monetise shopping activity now monetises completed transactions, and the concentration is extreme: a single line item is 62% of the segment that carries the company.
- **FICO Platform — Real, and Still Being Eaten by the Legacy Book:** The FICO Platform is a cloud decision-management system with $413M ARR growing 62% YoY as of Q3 FY2026 — it passed non-platform ARR for the first time this quarter and is now 51% of software ARR. Platform net retention is 148%. Banks that migrate origination, account management and collections onto it face multi-year re-implementation costs to leave, entirely independent of which score they pull. Trailing-twelve-month ACV bookings reached $128M, up 39%. The qualification is that none of this is showing up in reported software growth yet: Software segment revenue grew 2% YoY to $215.3M in Q3, because non-platform ARR fell 17% to $403M (net retention 82%), on-premises revenue declined 16% and professional services declined 24%. Blended software net retention is 109%. The legacy runoff is now a smaller share of the base each quarter, so the drag mechanically fades — but for the moment the 'second engine independent of GSE politics' is contributing roughly nothing to consolidated growth.

**Moat verdict:** FICO is still a net AI beneficiary in Platform — AI-driven decisioning, fraud detection and alternative-data integration all compound Platform ARR, which grew 62% to $413M and passed the legacy book for the first time in Q3 FY2026. The AI-resilient moats hold: the 70-year proprietary dataset grows more valuable as models require longitudinal training data, transaction embedding survives the pricing change, and system-of-record status does not update at AI speed. What changed this quarter is not AI, it is economics. FICO answered the VantageScore price war by matching it — $0.99 per pull plus $65 per funded loan since April 2026 — which protects volume and defuses the Hawley/FTC pricing complaint, but converts the most profitable franchise in the company from a royalty on shopping activity into a success fee on closings, and forfeits the per-pull price as a growth lever. Management's own description of 'score shopping' is the reason networkEffects moves from strong to intact: the reference standard is going bilateral, even though no volume has been lost yet. The durable read is that FICO is now a mid-teens compounder with a strong but no longer exclusive standard, a genuine software engine still masked by legacy runoff, and a mortgage line that has traded pricing power for price parity.

## Growth

Q3 FY2026 (reported July 29, 2026) was a good quarter that the market read as a bad one. Revenue was $674.2M, +26% YoY, roughly $5M below consensus — the first top-line miss in several quarters. Non-GAAP EPS of $12.18 (+42%) beat by $0.21 and GAAP EPS of $10.45 rose 41% on net income of $237M. Scores grew 41% to $458.9M, with B2B +49% and B2C +5%; mortgage origination revenue grew 97% and now accounts for 71% of B2B and 62% of Scores. Software grew 2% to $215.3M: Platform ARR $413M (+62%) overtook non-platform ARR ($403M, −17%) for the first time, but blended software ARR growth was only 10% and blended net retention 109%. Free cash flow was $370M in the quarter and $961M over the trailing four quarters (+28%). FY2026 guidance was raised to $2.53B revenue (+27% on FY2025's $1.991B), $850M GAAP net income and $979M non-GAAP net income, or $42.43 non-GAAP EPS — a raise, but below the ~$43.09 the Street carried. The implied Q4 of roughly $652M is +26% YoY, so the guide itself is not the problem. The problem is FY2027: the mortgage price increase that produced the +97% comp was reversed in April 2026 to $0.99 per pull plus $65 per funded loan, and management has not yet named the pricing lever that replaces it — only that it is examining auto, card and personal loans.

- **Revenue CAGR estimate:** 13–17%
- **Primary type:** both
- **Margin trend:** expanding
- **Key risk (moderate):** Mortgage now carries the company — 62% of Scores revenue — and it is monetised at $0.99 per pull with the economics deferred to a $65 fee collected only when a loan funds. Two unmaterialised things break it. If 'score shopping' converts from parallel pulls into displacement, with VantageScore taking the primary-pull slot on more than 10% of conforming files by FY2028, FICO loses volume it has so far retained. If originations stay depressed on rates and affordability, the funded-loan fee simply does not scale, and there is no per-pull price left to raise — the $10.00 price point has already been given back, and the Hawley/FTC file makes re-raising it politically expensive. Either path takes Scores growth from +41% toward single digits with the pricing lever spent.
- **Drivers:**
  - B2B Scores — Mortgage Direct Licence — +49% YoY in Q3 FY2026, down from +72% in Q2; mortgage origination revenue +97% YoY but now 62% of all Scores revenue, and priced at $0.99/pull + $65/funded loan since April 2026 versus the $10.00 that produced the comp (decelerating)
  - FICO Platform (Decision Management SaaS) — Platform ARR $413M (+62% YoY), 51% of software ARR and past non-platform for the first time; Platform NRR 148%; TTM ACV bookings $128M (+39%). Nets to Software segment revenue +2% and blended software ARR +10% after non-platform ARR −17% and NRR 82% (stable)
  - FICO Score 10T Adoption — Fannie Mae and Freddie Mac released FICO Score 10T performance datasets covering April 2013–September 2025 on July 1, 2026, unblocking lender validation of trended and rental data; the $0.99 upfront price is explicitly structured to drive 10T conversion. Adopters represented $377B in annual originations and $1.6T in eligible servicing volume as of Q2 FY2026 (accelerating)
- **Score derivation:** Base 80 on a 15% midpoint — decayed from the measured FY2026 +27% because that rate is a price increase FICO has since reversed, and the replacement lever (auto, card, personal loan pricing) is stated but unquantified. + 0 trajectory (one accelerating driver, one stable, one decelerating: B2B Scores went +72% in Q2 to +49% in Q3, and the Platform engine nets to +10% blended software ARR) + 4 margin (expanding: Q3 GAAP net income +30% on revenue +26%; FY2026 non-GAAP margin ~38.7% vs ~36.9% in FY2025) − 5 key risk (moderate) = 79. Unchanged in total from the June review, but rebuilt: the pricing war and the regulatory dismantling have now largely happened, so they are charged in the base and in the driver trends rather than sitting in keyRisk at high severity, where they were being counted against an estimate that had not yet absorbed them.

## Valuation

FICO fell about 16% on July 30 — its worst session in six years — to roughly $1,145, giving back a July rally that had carried it from the ~$1,150 of the last review to the mid-$1,300s on the July 1 release of Fannie/Freddie 10T performance data. The round trip means the price is close to unchanged since June while the picture underneath it has moved: the mortgage price lever is spent, the Software segment is flat, and the sell-side has abandoned the $2,000+ targets it carried into the print (RBC cut from $2,400 to $1,525, JPMorgan $1,325 to $1,225, BofA $1,550 to $1,400; Wells Fargo raised to $1,450; the 21-analyst average sits near $1,500). At ~27× the raised FY2026 non-GAAP EPS guide of $42.43, FICO trades below S&P Global on forward earnings and at roughly 27× trailing free cash flow of $961M. Against the revised ladder the stock sits ~64% above the bear case ($700) and ~24% below the base ($1,500) — a genuine discount, but a narrower one than the old $1,800 base implied, because that base assumed a re-rate to 40–45× forward, a multiple that belonged to the mandate era.

| Multiple | Value | Note |
| --- | --- | --- |
| Trailing P/E (GAAP) | ~33× | TTM GAAP EPS ~$34.62 (Q4 FY25 $6.42 + Q1 $6.61 + Q2 $11.14 + Q3 $10.45) |
| Forward P/E (FY2026 non-GAAP) | ~27× | FY2026 non-GAAP EPS guide raised to $42.43 |
| PEG Ratio | ~1.3× | fwd P/E ÷ ~20% blended forward EPS CAGR — down from ~30% before the mortgage price reset |
| Price / Sales (NTM) | ~10× | ~$2.7B NTM revenue off the $2.53B FY2026 guide |
| Price / FCF | ~27× | $961M TTM free cash flow (+28% YoY); $370M in Q3 alone |

The multiple did most of the damage. FICO went into the print at roughly 32× forward on a franchise whose pricing power had just been voluntarily reset, and a ~$5M revenue miss plus a guidance raise that landed $0.66 below consensus was enough to break it. At ~27× forward the stock is cheaper than S&P Global (~33×) and MSCI (~40×) and in line with Moody's (~30×) — but the PEG has moved from ~0.9× to ~1.3× because the blended EPS growth rate behind it is no longer 30%. The remaining discount is no longer purely a regulatory-overhang discount, as it was in June; part of it is now an honest reassessment of what a per-funded-loan royalty on a cyclical mortgage market is worth. _(as of July 2026)_

## Price scenarios

### Bear — $700

'Score shopping' converts into displacement, mortgage originations stay depressed so the $65 funded-loan fee does not scale, and the Software segment fails to convert Platform ARR into reported revenue growth.

- VantageScore 4.0 moves from a parallel pull to the primary-pull slot on more than 10% of conforming files by FY2028 — FICO retains the reference role but loses the unit economics on a growing share of the 62% of Scores revenue that mortgage now represents
- Rates and affordability keep originations below historical levels through FY2028. Under $0.99 + $65 the pull is nearly free and the revenue arrives only at closing, so a flat origination market translates directly into flat Scores revenue — with no per-pull price left to raise after the $10.00 point was given back
- Non-platform ARR keeps declining faster than Platform ARR converts to revenue: Software segment growth stays near zero beyond FY2027, blended net retention slips below 105%, and the 'second engine' thesis fails on timing rather than on product
- Multiple compresses to 17–19× on a flat-to-down FY2028 non-GAAP EPS of ~$40 as the growth premium unwinds entirely: ~$700/share, roughly 17× the current $961M free cash flow run-rate

### Base — $1,500

No binding regulatory outcome, FICO keeps ~90% of pulls under price parity, the auto/card/personal-loan pricing ladder partially replaces the mortgage lever, and Software growth re-accelerates as the legacy book finishes running off — a mid-teens compounder at a data-franchise multiple rather than a mandate multiple.

- The Hawley investigation and FTC referral close without a consent decree — largely defused by FICO's own April 2026 cut to $0.99 per pull, which removed the price increase that prompted the referral
- Price parity holds the line on volume: lenders pull both scores but FICO stays the primary on the large majority of conforming files, and the $65 funded-loan fee scales with a gradually normalising origination market
- Non-platform ARR falls below a third of software ARR during FY2027, so Platform's 62% growth and 148% net retention finally show up in reported Software revenue — segment growth returns to double digits
- FY2027 non-GAAP EPS of ~$48–50 at 30–32×, in line with Moody's and below S&P Global: ~$1,450–1,600/share, bracketing the post-print sell-side cluster (RBC $1,525, Clear Street $1,600, Wells Fargo $1,450, BofA $1,400)

### Bull — $2,050

The $0.99 price converts the market to FICO Score 10T, mortgage volumes recover, the auto/card/personal-loan pricing ladder delivers, and Platform ARR compounds past $700M — restoring a premium multiple on a broader base.

- The $0.99 upfront price does what it was designed to do: 10T displaces Classic FICO as the mortgage standard on the back of the July 2026 Fannie/Freddie performance datasets, and the $65 funded-loan fee proves worth more per closed loan than the old per-pull royalty
- FICO executes the auto, card and personal loan pricing ladder that management flagged on the Q3 call, restoring a Scores pricing lever independent of mortgage and independent of the FHFA-supervised channels
- FICO Platform becomes the decision-management backbone for 12+ major banks during the AI-driven refresh cycle, ARR passes $700M by FY2028 and the non-platform drag is fully lapped — shifting the revenue mix toward higher-multiple recurring software
- Originations normalise alongside; FY2029 non-GAAP EPS reaches ~$62 and the market pays 33× for a franchise with restored pricing power and a genuine software engine: ~$2,050/share

---

InvestMoat is an open-source research and education framework. Nothing here is financial advice. Past performance does not guarantee future results.