Payment market trends going into 2026 — what 2025 actually told us
Every January the payments industry produces a wave of trend pieces, most of which read like the same trend piece with the year on the cover updated. 2026 is genuinely different, and not because anything dramatic happened in a single quarter. It is different because several long-running shifts that spent most of 2025 in pilots and regulatory consultations have, between October and Christmas, crossed the line from "interesting" to "default." Looking at the payments market as of mid-January 2026, the honest summary is that the rails, the products on top of them, and the behaviour of the people using them have all moved at the same time — and the businesses still optimising around the 2023 stack are now visibly out of step.
The single biggest infrastructure story is that real-time payment rails are no longer a regional curiosity. The EU's Instant Payments Regulation made SEPA Instant the default euro credit transfer in 2025, with full reachability across the euro area and the price cap that finally killed the "instant premium." The UK's New Payments Architecture continued its phased migration, with Faster Payments volumes hitting new highs and commercial Variable Recurring Payments going live across a growing list of banks. In the United States, FedNow participation crossed well past thirteen hundred institutions during 2025 and is now a credible national rail alongside RTP. Brazil's Pix is processing more transactions per month than every card scheme in the country combined, and India's UPI continues to set the pace internationally. The practical consequence is that "instant, 24/7, structured-data, low-cost" is now the baseline expectation for a domestic payment in most of the markets that matter, rather than a premium product.
On top of those rails, the most visible product trend is the maturing of account-to-account commerce. Pay-by-bank, which spent years as the option no one clicked, has hit selection rates of 20 to 40 percent in the categories where it has been presented well — wallet top-ups, high-ticket retail, tax payments, charity giving, government services, and an expanding list of subscription and utility businesses that have switched their primary rail away from cards. The shift is not ideological; it is the boring result of the flow being faster, using the banking app the customer already trusts, and on mobile being fewer taps than typing a card number. Specialist providers have stopped competing on raw bank coverage — that is commoditised — and started competing on conversion, risk scoring, refund orchestration and the reconciliation experience their merchants actually live with.
The stablecoin story moved decisively from "crypto thing" to "treasury thing" during 2025. MiCA has been fully in force across the EU since the end of 2024, and by January 2026 the first generation of regulated euro-denominated e-money tokens — Circle's EURC, Société Générale–Forge's EURCV, and a growing cohort of bank-issued tokens — are in production use for cross-border B2B settlement and acquirer-to-merchant payouts, not just pilots. In the United States, the federal payment stablecoin framework passed in 2025 gave banks and licensed non-banks an explicit lane to issue and settle in dollar-backed tokens under prudential supervision. The use case that actually drives adoption inside tier-one acquirers is mundane: T+0 settlement to merchants on weekends, structured remittance data that an ERP can reconcile without human help, and the elimination of correspondent-bank credit risk on the last leg of a cross-border payout. None of this is consumer-facing, and that is precisely why it is finally working.
Cards are not going away — they remain the universal fallback, they own travel and rewards and a long tail of small-ticket commerce — but the card stack itself is being rebuilt around tokenisation. Network tokens have moved from "nice to have" to "table stakes," with Visa and Mastercard both pushing aggressive adoption targets and the major PSPs quietly switching merchants over by default. The conversion uplift is real (a few percentage points on recurring billing, more on cross-border), the fraud profile is materially better, and the operational benefit of automatic credential lifecycle updates — no more failed renewals when a card expires — has become impossible to ignore. Click to Pay, after years of false starts, is finally seeing meaningful merchant adoption in the EU and the UK on the back of SCA exemptions and a genuinely improved UX. The card-not-present experience in 2026 looks less like "type your 16 digits" and more like "approve in your wallet."
The most genuinely new trend, and the one most likely to be over-claimed in marketing decks this quarter, is agentic commerce. AI assistants buying on behalf of users — researching, comparing, and ultimately initiating payment — moved from demos to early production during 2025, and the payments industry has been scrambling to catch up. The schemes have published agentic-commerce frameworks defining how an AI agent is identified, how its mandate is scoped, and how liability is allocated when an autonomous purchase goes wrong. Tokenised credentials with explicit per-merchant or per-category limits, delegated authentication flows, and richer dispute primitives are the technical answer. The honest assessment in January 2026 is that the rails are mostly ready, the user behaviour is not yet at scale, and the businesses that will benefit first are those with high-frequency, low-judgment repurchase categories — groceries, household consumables, business supplies — rather than discretionary retail.
The interesting question in 2026 is no longer which rail is fastest. It is which combination of rails, tokens and identity primitives a merchant should default to for each flow.
Consumer behaviour has shifted in ways that are easy to miss because they are not driven by any single product. Wallet penetration on mobile has crossed the point where Apple Pay and Google Pay are the default tap on iOS and Android in most developed markets, with conversion lifts on iOS routinely 10 to 20 percent above raw card forms. BNPL has matured from a growth story into a regulated product: the UK's BNPL regime came into force in 2025, the EU's revised Consumer Credit Directive applies the same logic across the bloc, and the providers that survived the consolidation are now profitable, supervised, and integrated into mainstream checkouts rather than bolted on. Subscription fatigue is real and measurable, and the businesses that have switched recurring billing to pay-by-bank or VRP are seeing meaningfully lower involuntary churn. And privacy expectations have hardened: customers visibly prefer flows where they do not have to hand a 16-digit number to a merchant they will never buy from again, which is part of why wallets and pay-by-bank are winning.
Business needs have moved in parallel. Treasury teams that spent 2023 worrying about fraud now spend 2026 worrying about settlement speed, working capital efficiency and reconciliation quality — because instant rails and ISO 20022 data have made all three genuinely improvable for the first time. Finance teams want structured remittance data end-to-end, not free-text memo fields. Risk teams want device, identity and behavioural signals integrated into a single decision, not bolted onto a card authorisation as an afterthought. Product teams want one acceptance layer that abstracts cards, wallets, A2A and stablecoins behind a single API, with routing decisions made on real-time cost, conversion and settlement-time data rather than a quarterly procurement spreadsheet. The orchestration layer — once a niche category sold to enterprise merchants — is now where most of the interesting commercial action is happening.
Projecting from 2025 into the rest of 2026, four things look close to certain. First, the share of euro retail payments running on SEPA Instant — directly or via pay-by-bank — will continue to climb at the expense of cards in the categories where the UX already favours it. Second, regulated stablecoins will move from acquirer pilots to a default settlement option for cross-border B2B and marketplace payouts, with euro-denominated EMTs catching up to dollar-denominated ones inside the EU. Third, network tokens and Click to Pay will keep eroding the "type your card number" experience until it survives mostly as a fallback. Fourth, agentic checkout will spend most of 2026 in the trough of disillusionment, with the durable winners being the unglamorous infrastructure pieces — agent identity, scoped mandates, delegated authentication — rather than the consumer-facing assistants themselves.
The risks are equally clear. Fragmentation between national instant rails outside the EU continues to make cross-border real-time payments harder than they should be, and the various interlinking initiatives have so far promised more than they have delivered. Stablecoin liquidity remains concentrated in dollar-denominated tokens, which is uncomfortable for European policymakers and inconvenient for European businesses that price in euros. Agentic commerce introduces a genuinely new liability surface that the schemes' frameworks have addressed in principle but not yet stress-tested in volume. And the cost of operating in payments — compliance, fraud, scheme fees, certification — keeps rising in ways that quietly squeeze smaller PSPs and consolidate the market further.
The simple test for any payments roadmap entering 2026 is whether it treats instant rails, tokenised credentials and structured ISO 20022 data as the baseline and builds upward from there, or whether it still treats them as projects to scope. The businesses in the first group are spending their time on conversion, routing and treasury optimisation. The businesses in the second group are spending their time explaining to their board why their cost of payments is going up while everyone else's is going down. That, more than any single trend, is the line that defines the year ahead.