Summary

  • Fiscal-Q4 2026 Affirm Card GMV rose 124% to US$2.842 billion and active Card consumers rose 125% to 5.2 million, giving the product a disclosed 19% attach rate.
  • The adoption record is not a Card profit disclosure. Direct-to-consumer volume carries different merchant economics, while pay-over-time transactions still depend on issuers, originators, underwriting, funding and loan buyers.

Analysis

The most consequential part of a card transaction may occur before the card touches a terminal. An Affirm user can decide in the app whether to pay in full or request instalment credit, then spend online or in a shop that never built Affirm into its checkout. The card therefore changes who controls distribution. Affirm no longer has to wait for every merchant to place its financing button beside the price.

The adoption numbers are substantial. In fiscal Q4 2026, Affirm reported US$2.842 billion of Affirm Card GMV, 124% more than a year earlier. Active Card consumers reached 5.2 million, up 125%, and the attach rate was 19%. Across the platform, GMV rose 36% to US$14.1 billion, active consumers increased 21% to 27.8 million and transactions per active consumer reached 7.0, up 20%.

Those figures establish use. They do not establish one fixed conversion from Card GMV to earnings. GMV is purchase value net of refunds, not revenue, cash received or credit retained. A pay-in-full purchase and an interest-bearing instalment loan can cross the same piece of plastic while creating different fees, balance-sheet positions and risks.

The fiscal-2026 Form 10-K makes that distinction visible. Total revenue rose 32% to US$4.261 billion. Merchant network revenue increased 30% to US$1.150 billion; card network revenue increased 27% to US$294 million; interest income rose 27% to US$2.047 billion; gains on loan sales rose 56% to US$596.6 million; and servicing income increased 44% to US$173.1 million. These are related outputs of the network, not interchangeable descriptions of Card economics.

Merchant network revenue fell from 2.4% to 2.3% of GMV. Affirm says its direct-to-consumer products, including Affirm Card, generally earn lower merchant revenue and are predominantly interest-bearing. That is not necessarily a deterioration. The card can reach purchases outside an integrated merchant relationship, deepen consumer frequency and generate interchange or interest instead. But it means a faster Card mix can depress one take-rate measure while improving the total network.

The card-processing record shows a second boundary. GMV processed through card-issuing partners rose about 47% to US$17.5 billion, while card network revenue rose 27% for the year. Merchant incentives recorded as a reduction of card network revenue increased US$18.4 million, or 144%. The two growth rates must not be divided into a makeshift Card margin: issuer-processed GMV includes Affirm Card, virtual debit cards and some merchant-integration flows, and the revenue line reflects interchange sharing, merchant mix and incentives. The periods also differ from the quarterly 124% Card figure.

Credit begins when a user chooses to pay over time. Affirm purchased US$40.2 billion of loans from originating banks during the year, up 34%, including US$6.3 billion of 0% APR instalment loans. Losses on loan-purchase commitments rose 29% to US$311.9 million. Credit-loss provision rose 29% to US$796.7 million as the average held-for-investment loan balance increased 27%.

The period-end ageing table is a reason for vigilance, not a credit-collapse headline. Loans more than three days past due totalled about US$488.4 million on a US$9.466 billion amortised-cost base, compared with US$339.5 million on US$6.958 billion a year earlier. That is approximately 5.16% versus 4.88% by calculation. The 30-to-119-day balance was approximately 2.51% versus 2.46%. Portfolio age, duration, product and origination mix affect both comparisons. Modifications for borrowers in financial difficulty rose from 0.17% to 0.25% of receivables, still a small share but an indicator worth carrying forward.

Affirm does not keep every loan. It sold US$21.9 billion of unpaid principal during fiscal 2026, 39% more than a year earlier. Gains on those sales increased 56%, which the company attributed to greater volume and favourable transaction economics driven mainly by market conditions. Average principal serviced for third-party investors and off-balance-sheet securitisations rose 42% to US$9.1 billion. A loan sale can release balance-sheet capacity and preserve servicing income, but its price is set partly by buyers' appetite, funding markets and estimates of recourse, servicing and risk sharing.

There is no evidence of an immediate funding squeeze in the filing. Affirm had US$2.603 billion of cash and available-for-sale securities, US$5.3 billion of available funding-debt capacity and US$675 million of revolving capacity. Yet the operating chain remains concentrated at important junctions: Celtic Bank and Lead Bank originated substantially all platform loans, while Evolve Bank & Trust and Stride Bank were the two Affirm Card issuers at fiscal year-end.

The annual profit headline also needs its own ledger. Affirm produced US$417.0 million of operating income after an US$87.3 million operating loss the prior year. Net income was much larger at US$1.930 billion because it included an US$1.437 billion income-tax benefit, mainly a release of the valuation allowance against domestic deferred tax assets. That accounting benefit is valid, but it neither measures Card profitability nor equals recurring operating cash.

Affirm Card has passed an important distribution test. It is used by millions of consumers and is becoming a meaningful share of platform transactions. The next test is harder: show that repeated use continues to produce durable revenue after merchant and card incentives, credit provision, processing, funding and loan-disposition costs. The card can widen the network even when a familiar take rate falls. The evidence must follow the whole route rather than stop at the terminal.

Sources