Affirm Is Outgrowing the Checkout Button
Executive summary
Affirm's fiscal 2026 results are easy to read as a straightforward buy-now, pay-later growth story. Gross merchandise volume rose 37% to $50.2 billion, active consumers reached 27.8 million and transactions per active consumer increased to 7.0. The company reported another quarter of profitable growth and pointed toward $100 billion of annual volume.
That interpretation misses the more important change in the operating model. Affirm is becoming less dependent on appearing as a financing button at a directly integrated merchant. Its card and virtual-card products carry Affirm into stores and websites where the merchant may not have chosen Affirm at checkout. Meanwhile, a large funding and loan-sale system turns short-duration consumer receivables back into capacity for the next transaction.
The contrarian thesis is that Affirm's next stage will be decided less by consumer awareness of BNPL than by its ability to run two external networks at once: card rails for acceptance and capital markets for balance-sheet velocity. That broadens distribution and raises purchase frequency. It also makes interchange economics, merchant incentives, funding availability, credit performance and loan-buyer demand central to growth.
This is not an argument that merchant checkout integrations no longer matter. Direct integrations still contribute substantial volume and richer merchant economics. It is an argument that the headline category label no longer describes the full system operators are buying into or competing against.
Market context
The conventional BNPL model starts at the merchant. A retailer integrates a provider, presents financing at checkout and often pays a fee that reflects the provider's claim that payment flexibility can increase conversion or basket size. Distribution is negotiated merchant by merchant or acquired through a commerce platform.
Affirm now spans three routes: direct merchant point-of-sale integrations, direct-to-consumer products led by Affirm Card, and wallet or platform partnerships. The card route changes the boundary. A consumer can use Affirm through established card acceptance, then pay in full or apply to pay over time. The merchant does not need to display an Affirm button for the transaction to enter Affirm's system.
The latest quarter makes the shift visible. In the quarter ended June 30, 2026, direct-to-consumer GMV rose 49% to $4.7 billion, and Affirm said the growth was driven entirely by Affirm Card, whose GMV rose 124% to $2.8 billion. For the full fiscal year, $17.5 billion of GMV ran through card-issuing partners, up 47%.
Volume expansion also requires funding. Most transactions generate a loan. Affirm purchases many loans from originating bank partners, originates some through subsidiaries, holds some receivables, sells others to third-party buyers and securitization vehicles, and continues to service sold loans. A consumer-facing transaction therefore depends on an institutional capital chain behind it.
The mainstream interpretation is that scale and better underwriting are turning BNPL into a profitable mainstream payment method. The alternative view is that scale comes from extending beyond the merchant integration and recycling credit assets efficiently. Profitability can improve, but the system's fragility migrates toward funding spreads, loan-sale execution and the economics of third-party rails.
Findings
Finding 1
Affirm is gaining frequency faster than it is adding consumers.
The user base is growing, but engagement is doing more work. Active consumers rose 21% in fiscal 2026, from 23.0 million to 27.8 million. Transactions per active consumer rose 20%, from 5.8 to 7.0. Total GMV rose 37%. Those rates are consistent with a service moving from episodic large-ticket financing toward more regular use.
The product mix supports that reading. Interest-bearing monthly installment loans still represented 70% of fiscal 2026 GMV, but that share declined from 74% in fiscal 2024. Pay-in-X rose to 16% of GMV and 0% APR monthly installments rose to 14%. Shorter or merchant-supported products are not replacing interest-bearing loans; they are widening the set of occasions.
| Metric | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| GMV, $ billions | 26.6 | 36.7 | 50.2 |
| Active consumers, millions | 18.713 | 23.003 | 27.782 |
| Transactions per active consumer | 4.9 | 5.8 | 7.0 |
| Interest-bearing installment share of GMV | 74% | 72% | 70% |
| Pay-in-X share of GMV | 15% | 14% | 16% |
| 0% APR monthly installment share of GMV | 11% | 13% | 14% |
Source: Affirm fiscal 2026 Form 10-K, fiscal years ended June 30, 2024-2026. Units are shown in the table. Values are company-reported; active consumers are those with at least one transaction in the preceding 12 months. Shares use GMV net of refunds. Limitations: the table does not identify unique purchase occasions, credit quality, customer income or incremental merchant sales.
Frequency matters because a payment network becomes more valuable when the customer can reuse it across contexts. The Affirm Card can reach in-store purchases and merchants without a direct Affirm integration. In fiscal 2026, volume processed through card-issuing partners rose to $17.5 billion, approximately 35% of total GMV. The prior-year amount was about $11.9 billion, derived from Affirm's disclosed 47% growth rate.
This is useful distribution, but it is not free distribution. Card network revenue rose 27%, slower than the 47% increase in partner-processed GMV. Merchant incentives recorded against card network revenue rose 144%. Mix, interchange and incentives determine yield. A card transaction can expand acceptance while producing different economics from a direct merchant financing offer.
Finding 2
The funding system is part of the product, not a back-office utility.
Affirm purchased $40.2 billion of loans from originating bank partners and directly originated $9.5 billion during fiscal 2026. It sold loans with $21.9 billion of unpaid principal to third parties, up 39% from $15.8 billion a year earlier. The company retained servicing relationships on many sold receivables.
This recycling lets transaction volume exceed the equity capital that would be required if every loan remained on Affirm's balance sheet. Funding capacity reached $30.0 billion at fiscal year-end, up from $26.1 billion. Affirm estimated that the capacity could support more than $70 billion of annual GMV, based on a weighted-average loan duration of about five months.
| Funding and distribution measure | FY2025 | FY2026 | Change / transformation |
|---|---|---|---|
| Total GMV, $bn | 36.7 | 50.2 | +37% |
| GMV processed through card-issuing partners, $bn | 11.9* | 17.5 | +47% |
| Loans sold, unpaid principal, $bn | 15.8 | 21.9 | +39% |
| Year-end funding capacity, $bn | 26.1 | 30.0 | +15% |
| Merchant network revenue as % of GMV | 2.4% | 2.3% | -0.1 percentage point |
| Average warehouse/securitization funding debt, $bn | 5.9 | 7.5 | +27% |
| Funding-cost growth | n/a | +7% | Favorable pricing partly offset higher balances |
Sources: Affirm fiscal 2026 Form 10-K and fiscal Q4 2026 shareholder letter. Period: fiscal years ended June 30, 2025 and 2026, except funding capacity at each year-end. Units are dollars in billions, percentages and growth rates. *FY2025 card-partner GMV is derived as $17.5bn divided by 1.47 and rounded to one decimal. Limitations: capacity is not utilization; loan sales depend on market conditions; funding channels carry different duration, recourse and spread economics.
The favorable fiscal 2026 pattern is real: average warehouse and securitization debt increased 27%, while funding costs increased only 7%, partly because of better pricing. Gain on loan sales rose 56%, aided by both higher volume and market conditions. The danger is treating that performance as a permanent property of software. Credit markets can reprice faster than consumer demand, and gain-on-sale economics can change even if application conversion is stable.
For merchants, this means vendor resilience cannot be assessed only through approval rates and checkout uptime. Funding diversification, committed capacity, asset-sale demand and loss performance affect whether the provider can keep extending offers through a cycle.
Finding 3
General acceptance broadens the network while weakening a simple merchant-fee story.
Affirm's merchant network revenue grew 30% in fiscal 2026, slower than GMV. Merchant network revenue as a percentage of GMV slipped to 2.3% from 2.4%. The company explained that direct-to-consumer products generally earn lower merchant revenue, while the mix of 0% APR products, commercial agreements and incentives also changes the rate.
That is not necessarily deterioration. Lower direct merchant revenue can be rational if card distribution adds occasions, consumer retention, interest income or loan-sale value. It does mean that operators should not evaluate expansion by GMV alone. The relevant unit is contribution after merchant incentives, network costs, funding costs, credit losses, servicing and the capital required for the receivable.
Affirm's fiscal 2026 net income also needs a stable denominator. The company released a significant portion of its domestic deferred-tax valuation allowance, producing a roughly $1.5 billion non-cash tax benefit. That accounting event reflected improved evidence of profitability, but it is not recurring transaction economics. Revenue less transaction costs, operating income, credit provision and cash generation provide cleaner views of repeatable performance.
The deeper strategic change is two-sided. The card expands consumer-led distribution beyond direct merchant integrations. The funding network expands institutional distribution of the resulting assets. Affirm sits between merchants, consumers, card partners, originating banks and capital providers. The system can compound if each side increases the value of the others. It can also transmit pressure from one side to the rest.
Implications for operators
- Merchants should compare routes, not logos. A direct integration, wallet placement and card transaction can put the same provider in the purchase flow with different fees, data access, dispute processes and conversion effects. Measure contribution margin and incrementality by route.
- Payments teams should monitor acceptance yield. Track revenue and gross profit per dollar of volume across direct merchant, card and wallet channels. Rising GMV with falling yield may be an efficient expansion or an expensive subsidy; the cohort evidence decides.
- Finance leaders should include provider funding in continuity reviews. Request information on committed capacity, maturity concentration, loan-sale channels, triggers and stress performance. A checkout provider that originates credit has a different dependency map from a processor that only moves funds.
- Consumer-product teams should separate frequency from indebtedness. More transactions per active consumer can show useful everyday adoption, but it can also increase the need for clear obligation views, repayment sequencing and hardship handling across simultaneous plans.
- Competitors should not copy the card without the asset engine. Universal acceptance can acquire transactions, but sustainable economics require underwriting, servicing, funding and capital allocation that work at the same frequency.
Risks & open questions
The thesis would weaken if direct merchant integrations remain the dominant source of incremental GMV and gross profit, while card-led volume plateaus or requires persistent incentives. It would also weaken if funding and loan-sale economics prove largely insensitive to credit-market conditions.
Public reporting does not provide a full route-level contribution-margin bridge. Card-partner GMV combines Affirm Card, virtual cards and some platform or merchant arrangements. The categories therefore do not map perfectly to consumer-led distribution. Company disclosures also do not show the credit performance, repeat rate or merchant incrementality of each route in a common cohort table.
Credit quality remains the main external test. Faster frequency and broader acceptance are valuable only if underwriting retains enough signal outside a directly integrated merchant context. Funding capacity is another test: unused capacity is not the same as durable, attractively priced capital. Finally, card networks and issuing partners can change economics or rules that Affirm does not control.
The mainstream BNPL frame will remain useful for consumer regulation and credit-risk analysis. It is becoming less useful for strategy. The operator now has to understand a payment network, a consumer-credit platform and an asset-distribution system at the same time.
Appendix / methodology notes
This report uses Affirm's fiscal 2026 Form 10-K, shareholder letter and earnings materials published August 27, 2026. Figures are transcribed from company filings unless explicitly marked as derived. The only derived value is fiscal 2025 GMV processed through card-issuing partners: $17.5 billion divided by 1.47, rounded to $11.9 billion.
GMV is Affirm's reported transaction value net of refunds and is not revenue. Active consumers have completed at least one transaction in the prior 12 months. Product shares are percentages of GMV, not loan balances. Funding capacity is a company estimate of available capacity across channels and should not be read as cash, committed drawdown or deployed capital.
No valuation conclusion is offered. The analytical test is whether route-level revenue less transaction costs, credit outcomes and capital requirements improve as card-led frequency and external asset distribution grow. A stronger future chart would compare direct merchant, card and wallet cohorts on approval rate, repeat frequency, gross yield, funding cost, loss rate and contribution margin; those data are not publicly disclosed in a comparable form.