| Investor | Approach | Verdict | Conviction | In one line |
|---|---|---|---|---|
| A | Quality and long-term value | Watch | 6 | Superb economics built on a mandate the regulator is dismantling |
| B | Deep value and margin of safety | Avoid | 8 | Five-year average earnings, not the trailing print, set the value he will pay |
| C | Global macro and reflexivity | Watch | 6 | The short base is crowded and the adoption data does not exist |
| D | Growth and fundamental stock picking | Watch | 7 | The multiple is attractive; the post-policy growth is not proven |
No Buys of four. Fair Isaac needs three to enter a basket, so it does not.
Bottom line
We met on Fair Isaac the week the FHFA put a regulator-sponsored competitor on equal footing with its score, and the committee's answer was three Watches and one Avoid. The number that matters is $887M, the average of the four seats' probability-weighted year-three free cash flow, against the roughly $1,060M path the price already embeds by then. The seats expect less cash than the market pays for, and still nobody says Sell, because the one figure that would settle it, the mortgage share of Scores revenue, is not filed by anyone. The 10-K is due by November 30, 2026, and last year's release came November 5, 2025, so the ambiguity has a deadline.
Key takeaways
- The profit pool is one segment. Scores supplied 88% of segment operating income in the latest quarter at a 90.8% margin (10-Q, filed July 29, 2026). Software, the supposed diversifier, saw its operating income fall 19.0%.
- The balance sheet was built for the old world. Negative equity, buybacks at 1.83 times fiscal 2025 free cash flow, and no debt maturity schedule on file, our arithmetic.
- Nobody can size the exposure. The mortgage share of Scores revenue is not disclosed, so the threatened slice could be modest or most of the segment.
- What splits the committee. A waits for a filing that proves durability, B finds a 152% premium over his normalized value, C waits for the crowded short to be wrong, D likes the multiple but not the unproven growth.
The business in 60 seconds
Fair Isaac reports two segments. Scores sells the FICO credit score, the system the Financial Times says "currently reigns supreme in the mortgage market", alongside card, auto and other lending (October 3, 2026). Software sells decisioning and analytics platforms to banks, largely by subscription. In the quarter to June 30, 2026, Scores produced $416.89M of operating income on $458.90M of revenue, and Software $55.04M on $215.29M (10-Q, filed July 29, 2026). Scores is the business; Software is a shrinking second act.
Scores revenue is pulls times price, and the filings give neither. The segment grew 41.5% while originations sat at half their 2021 level, which points to price as the engine. The score is delivered through the three credit bureaus, and those bureaus own VantageScore. The distributor and the challenger are the same firms.
The capital structure is a choice, not a need. Capex was $8.92M in fiscal 2025 against $769.88M of free cash flow, but buybacks were $1.41B and equity is negative $4.10B. Pay is keyed to shareholder return relative to the 55th percentile of the S&P 500 (DEF 14A, filed January 27, 2026).
What changed
- The regulator ended the mortgage exclusivity. FHFA director Bill Pulte instructed Fannie Mae and Freddie Mac to approve all lenders to use VantageScore, removing the surcharge that penalized the rival score, and said the bureaus "have been overcharging Americans for far too long. This will end soon." (Bloomberg, September 4, 2026; Financial Times, October 3, 2026)
- A second cut is under study. The agencies are studying bi-merge credit reports, two bureaus instead of three, which would cut score pulls per loan before any share is lost. (Bloomberg, September 4, 2026)
- The last filed quarter predates the policy. Revenue grew 25.7% and operating income 38.1% in the quarter to June 30, 2026 (10-Q, filed July 29, 2026), so the annual report will show a nearly untouched quarter.
- The guidance on file is stale by one policy regime. Fiscal 2026 GAAP net income guidance of $825M to $850M predates it (results release, Exhibit 99 to the 8-K).
- Funding costs moved against the debt. The 10-year yield rose 45 basis points in 30 days to 5.24%, and high-yield spreads widened 59 (FRED, October 1, 2026), against $5.58B of total debt with no maturity schedule on file.
The bull case on Fair Isaac
The honest bull case is that permission is not adoption, and the price already pays for stagnation. Investor A, who still will not buy, concedes the engine: "The scores franchise has some of the best unit economics I have seen." Investor D goes further on price: "the current valuation implies required annual free cash flow growth of only 2.1% over the next 10 years," a hurdle D finds modest against the 16.6% a year the business delivered over four fiscal years. Investor C supplies the reflexive leg: "the 8.98% short base is trapped against a 6.97% FCF yield and a shrinking share count," and the count did shrink 16.1% over four years. If switching costs bind, the earnings base holds and a stagnation multiple re-rates without the old growth rate returning.
The bear case on Fair Isaac
The bear case attacks the starting cash flow, not the growth rate. Investor B prices the whole company off five-year average net income of $471.94M at an assumed 12x, which supports $5.66B of equity value against a $14.28B market cap, and concludes: "The current valuation does not adequately compensate me for that uncertainty." Investor A refuses for a different reason: "Still, I do not buy a business when I cannot judge how durable its main earnings stream is, and certainly not one carrying this much debt." The debt is the amplifier. Negative equity leaves no book cushion, refinancing meets a rising 10-year, and Software cannot help while its operating income falls.
Valuation
What the price requires. The sheet's reverse arithmetic on trailing free cash flow of $996.02M: 2.1% growth a year for ten years, 3.0% after that, discounted at 9.74%, the 10-year yield of 5.24% plus 4.5 points. On our arithmetic that path puts year-three free cash flow near $1,060M. Delivered growth over the last four fiscal years was 16.6% a year. The market is paying for stagnation.
What the committee expects. Each seat's scenarios, weighted by its own probabilities. Year-three free cash flow, in millions:
| Investor | Bear | Base | Bull | Probability-weighted, year 3 |
|---|---|---|---|---|
| A | 35%, about $650M | 45%, about $1.05B | 20%, about $1.3B | $960M |
| B | 35%, $341M | 45%, $691M | 20%, $1,062M | $643M |
| C | 35%, about $650M | 45%, about $1.05B | 20%, about $1.25B | $950M |
| D | 30%, $620M | 50%, $1.05B | 20%, $1.42B | $995M |
| Average | $887M |
The average sits 11% below today's trailing free cash flow and 16% below the stagnation path the price embeds. Three of four seats call the price fair or attractive while their own arithmetic expects less cash in three years than the market pays for today. The reconciliation is a reset weighed against a recovery, not a meltdown, and B, the one seat treating trailing earnings as unsustainable, lands 34% below the average of the other three. One caveat: B projects from fiscal 2025 revenue, the others from roughly today.
Equity against enterprise. Free cash flow yields 6.97% on the market cap and 5.09% on enterprise value of $19.56B, the cap plus $5.28B of net debt. Net debt is 5.3 times trailing free cash flow. The equity looks cheaper than the business.
Our arithmetic, from the filings. Trailing free cash flow of $996.02M less the $185M to $188M of fiscal 2026 stock compensation guidance gives owner earnings near $810M, a 4.14% yield on enterprise value. Stock compensation was 7.9% of fiscal 2025 revenue and the guidance embeds it rising. Cash statements that treat stock pay as free overstate what owners keep.
Comparables
| Company | Currency | Market cap | EV | Revenue growth, last FY | Operating margin | P/E | Price / FCF | FCF yield | EV / after-tax operating income | Net debt / operating income | ROIC |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Fair Isaac Corp (FICO) | USD | $14.3B | $19.6B | 15.9% | 46.5% | 17.5x | 14.3x | 7.0% | 20.7x | 4.3x | 64.9% |
| S&P Global Inc. (SPGI) | USD | $114.3B | $123.2B | 7.9% | 42.2% | 23.9x | 20.6x | 4.9% | 23.0x | 1.3x | n/a |
| Verisk Analytics, Inc. (VRSK) | USD | $21.5B | $25.2B | 6.6% | 43.7% | 23.6x | 19.0x | 5.2% | 24.0x | 2.7x | n/a |
| Moodys Corp /De/ (MCO) | USD | $76.4B | $81.8B | 8.9% | 43.4% | 27.4x | 25.8x | 3.9% | 28.8x | 1.5x | n/a |
| Equifax Inc (EFX) | USD | $16.4B | $20.3B | 6.9% | 18.0% | 23.8x | 14.9x | 6.7% | 23.9x | 3.4x | n/a |
| Transunion (TRU) | USD | $12.2B | $16.9B | 9.4% | 18.7% | 17.4x | 17.6x | 5.7% | 24.4x | 5.5x | n/a |
Source: each company's own filings (SEC EDGAR for US filers; company reports under IFRS for Canadian filers) and the Yahoo Finance price, computed the same way for every row. EV is market cap plus net debt. n/a means an input could not be sourced; see the notes. Every n/a is in the ROIC column, where invested-capital inputs were not sourced for the five peers.
FICO has the highest growth and the highest operating margin in this set, and the lowest price to free cash flow: 14.3x against 19x to 26x for the three data-and-ratings franchises. The bureaus, its distributors and the owners of its challenger, trade at 23.8x and 17.4x earnings. If the equal-footing policy transfers profit rather than share, the bureaus sit on the receiving end, and the market has priced a generic de-rating at FICO rather than a transfer to them. November can close that gap.
Key risks
Ranked by lasting damage and how soon the evidence arrives. Raised by says where each risk entered the record.
| # | Risk | Why it matters | Severity | Raised by | First signal |
|---|---|---|---|---|---|
| 1 | Equal footing becomes adoption | The surcharge that protected mortgage pricing is gone; switching is unmeasured | High | Bloomberg, September 4, 2026; Financial Times, October 3, 2026; all four seats | VantageScore share of agency-delivered loans |
| 2 | A bi-merge mandate | Two bureau reports instead of three cuts pulls per loan | High | Bloomberg, September 4, 2026; Investors C and D | The FHFA study outcome, no filed date |
| 3 | Leverage into a pricing reset | $5.28B net debt, negative equity, no maturity schedule, refinancing into a rising 10-year | High | Analyst; Investors A and B; desk arithmetic | Coupons on new debt, or net debt above 6x free cash flow |
| 4 | Trailing earnings overstate the base | Five-year average net income is $471.94M against $815.02M trailing | High | Investor B; desk arithmetic | Fiscal 2027 guidance at the results release |
| 5 | The distributor owns the challenger | The bureaus deliver FICO's scores and own VantageScore | High | Financial Times, October 3, 2026; Analyst | Bureau bundle pricing, and FICO's bureau contract renewals |
| 6 | Capital allocation while besieged | Buybacks ran ahead of free cash flow; pay is keyed to relative shareholder return | Medium | Analyst; Investors A and B; DEF 14A, filed January 27, 2026 | Buyback pace and debt issuance in the 10-K |
| 7 | Software keeps shrinking | The diversifier is 12% of segment operating income and falling | Medium | 10-Q, filed July 29, 2026; Investors A and C | Software growth and margin in the next 10-Qs |
| 8 | Litigation over past disclosures | Investigation announcements establish no finding | Medium | forum (unverified); Investor C; Analyst | A filed complaint, or a regulator opening an inquiry |
| 9 | Workforce concentration in India | About one-third of staff; the 10-K names India-Pakistan tensions | Low | 10-K Item 1A | Business continuity disclosures |
What to watch
| Date | Event | What it decides |
|---|---|---|
| Early November 2026, by the November 5, 2025 pattern, not a scheduled date | Fiscal 2026 results release | The first guidance written after the policy change |
| By November 30, 2026, SEC deadline | 10-K for the year to September 30, 2026 | Annual Scores revenue and margin, debt, and whether buybacks stayed ahead of free cash flow |
| Late January 2027, by the January 28, 2026 pattern, not a scheduled date | First quarter of fiscal 2027, the first full post-policy quarter | Whether Scores revenue growth goes negative, the line three seats named as their breaker |
| Pending, no filed date | FHFA bi-merge study outcome | Whether score pulls per loan fall from three bureaus to two |
Owning it onchain
xStocks issues FICOx, a token tracking Fair Isaac. We have not judged whether it is investable: our only quote came before regular US trading opened on October 5, 2026, when market makers quote thinly. Backpack's FICO token is not verified on Jupiter, so it is not investable.
xStocks says its tokens are not offered in the United States or to US persons. Sunrise blocks users in the United States and Canada from Backpack's tokens. This is not a recommendation.
The committee record
| Investor | Verdict | What would change its mind |
|---|---|---|
| A | Watch 6 | Double-digit Scores growth after the policy moves it toward Buy. A bi-merge adoption or a mortgage price concession moves it to Avoid |
| B | Avoid 8 | GAAP net income of at least $1.19B with net debt no greater than $5.28B in the fiscal 2028 10-K breaks its thesis. Net debt above $6.0B in the fiscal 2026 10-K breaks its resilience case |
| C | Watch 6 | A bi-merge mandate flips it bearish regardless of the yield. Fiscal 2027 non-GAAP guidance below about $866M breaks the stagnation read |
| D | Watch 7 | A pause or rescission of the directive, or private investors rejecting non-FICO collateral, moves it to Buy. A bi-merge mandate moves it to Avoid |
Data
| Item | Figure | Source |
|---|---|---|
| Market cap | $14.28B | $661.25 times 21,597,635 SEC shares, October 2, 2026 |
| Revenue, trailing twelve months | $2.39B | SEC EDGAR, periods to June 30, 2026 |
| Net income, trailing twelve months | $815.02M | SEC EDGAR, periods to June 30, 2026 |
| Free cash flow, trailing twelve months | $996.02M | SEC EDGAR, computed |
| Scores segment margin | 90.8%, 88% of segment operating income | 10-Q, filed July 29, 2026 |
| Total debt | $5.58B, liquid assets $304.54M | Balance sheet, June 30, 2026 |
| Shareholders' equity | Negative $4.10B | Balance sheet, June 30, 2026 |
| Fiscal 2026 GAAP net income guidance | $825M to $850M | Results 8-K, Exhibit 99, filed July 29, 2026 |
| Buybacks, fiscal 2025 | $1.41B, 1.83x free cash flow | Filings, our arithmetic |
| Short interest | 8.98% of shares, 4.03 days to cover | FINRA, settlement September 15, 2026 |
Method. Four investors with different philosophies, each on a different AI model, read one identical fact sheet and voted independently; a name needs three Buys to enter a basket. A fifth, non-voting analyst wrote the business and risk review. The fact sheet was built from SEC filings and the FHFA policy as reported by Bloomberg and the Financial Times, with forum discussion as labelled unverified context. Every verdict is later scored under a rule we published before the first results.
Correction. This article first called FICOx investable on a 1.24% round-trip cost. That quote was taken before regular US trading opened, when market makers quote thinly. We now judge that cost only during regular US market hours, and we have removed the figure, the verdict and the mint.
As of October 5, 2026, BagsCapital manages no outside capital and holds no position in Fair Isaac. Nothing here is investment advice.