🇺🇸 Payments in the US

The real size of the US payments market

The US payments market differs from other large developed markets in one structural way: no single rail dominates. Cards lead by number of transactions, and ACH leads by value. Checks hang on, with average amounts Europe no longer sees, and instant payments are split between a private operator and the central bank. A payments leader arriving from Europe will usually look for an equivalent of SEPA: a single scheme backed by a single authority. The US has neither. Each rail has its own operator and its own rulebook.

The average ticket, meaning the value a rail carries divided by its number of transactions, is the key to reading this market. Cards carried 187.7 billion transactions worth $11.5 trillion, an average ticket of about $61. ACH carried 39.7 billion transfers worth $104.06 trillion, an average ticket above $2,600. Checks sit higher still, with 9.2 billion items worth $24.45 trillion. These gaps in ticket size reflect how usage splits across types of payer. Cards are the consumer’s rail; ACH and checks are the rails of business and government. Corporate collections setups are built around that divide.

One last structural marker is the number of banks in operation. In the first quarter of 2026, the FDIC counted 4,278 insured banks and savings institutions, and the NCUA counted 4,250 federally insured credit unions: more than 8,500 depository institutions in all. No other developed country is this fragmented. That fragmentation explains most of what follows: a private rules association (Nacha), the slow rollout of every new rail, and the outsized role of the large processors that run small banks’ systems.

ACH and Nacha: the rail that carries the value

The ACH Network, in service since 1972, carries payroll, government benefits, bill-payment debits, tax refunds, and most US B2B payments, in a setup that is unique in the world. A private association, Nacha, writes the rules, while clearing is handled by two competing operators: FedACH (run by the Federal Reserve Banks) and EPN, The Clearing House’s Electronic Payments Network. No other country has two operators competing under a single rulebook. That is why Nacha rules are written to be operator-agnostic. A company that sends ACH files does not choose the operator. Its bank does.

Same Day ACH, launched in 2016, is the rail’s incremental answer to instant payments, with three same-day settlement windows. The per-payment limit was raised to $1 million in March 2022. It is growing much faster than standard ACH, from 1.2 billion payments worth $3.2 trillion in 2024 to 1.4 billion payments worth $3.9 trillion in 2025 (Nacha, 2026). It is absorbing part of the demand that RTP and FedNow were targeting. For many corporate use cases it is a workable middle ground: faster than standard ACH, far cheaper to integrate than an instant rail, and already available at every bank.

Two instant rails: RTP vs. FedNow

An interbank instant payment system credits the payee’s account within seconds, at any hour, with finality. The US is the only major market where two interbank instant payment systems coexist without interoperating. The Clearing House’s RTP network opened in 2017 under the control of a consortium of large banks. It was the first entirely new interbank rail since ACH in 1974. The FedNow Service, run by the Federal Reserve Banks, opened in July 2023 with an explicitly political rationale: open access for any institution eligible for Fed services. Small banks and credit unions did not want to depend on infrastructure owned by their competitors. Both use ISO 20022, and both settle in central bank money. A payment cannot cross from one to the other.

These figures show adoption driven by treasury flows rather than by person-to-person payments. The average ticket on FedNow is $101,435, and the value carried by RTP grew fivefold in the same year its limit rose to $10 million. Large-value payments were the first to move to instant rails in the US: real estate closings, margin calls, merchant settlements, intercompany transfers, and brokerage account funding. Zelle had already held the instant person-to-person market since 2017. An instant payment use case built in the US is therefore a treasury play rather than a checkout play.

Two complementary features sit alongside these instant rails. Request for Payment (RfP), available on both RTP and FedNow, lets a creditor send a payment request into the debtor’s banking app, where the debtor approves it with a single tap. It is the only mechanism that gives US instant payments a billing use case, but adoption varies widely from bank to bank. A request for return of funds is the mechanism for recalling funds that have already settled. It is a recovery procedure with no guarantee of success, since the receiving bank is free to act on it or not. Banks are also free to set internal limits well below the network limit. The limit that actually applies to an account is therefore set by the bank that holds it, not by the operator.

Zelle, Venmo, Cash App: the P2P services that ate the check

Zelle, launched in 2017 by Early Warning Services (a joint venture of seven large US banks), has become the leading person-to-person payment service in the US. It is a service built into the banking apps of more than 2,200 institutions, not a separate app to install. Users install nothing, open no new account, and fund no balance. That is the opposite of the approach taken by PayPal and Block, whose services rely on a standalone app and a balance to top up. Zelle’s approach has produced the largest volumes in US P2P.

Beyond Zelle, nonbank players hold the younger half of the market. Venmo (PayPal, 2009) built its position on a social feature, a public activity feed, and has expanded into merchant acceptance. Cash App (Block, 2013) is the closest US example of a super app, combining P2P transfers, the Cash Card debit card, paycheck direct deposit, stock trading, bitcoin, and credit. It has a real foothold among underbanked consumers that neither Zelle nor the big banks reach. PayPal (1998) remains the leading alternative payment button in US e-commerce, with a hybrid model of e-money balances, ACH debit, and a card overlay.

The main open question in US P2P is legal: who bears the loss when the customer has been manipulated? Federal law, through Regulation E (12 CFR 1005), protects consumers against unauthorized transfers. A payment the customer authorized, even as the result of a scam, falls outside that scope. In December 2024 the CFPB sued Early Warning and three of its owner banks to obtain refunds for this “induced” fraud, then dropped the case on March 4, 2025, with prejudice. The New York Attorney General picked up the fight on August 13, 2025, suing Early Warning in New York state court. In the US, when the federal regulator steps back, the states step in. The protection a consumer gets therefore depends on the state they live in, not on federal law alone.

Paze is an online checkout solution launched in 2023 by Early Warning to win back control of the checkout from the digital wallets. It works by automatically pre-provisioning partner banks’ cards, with no enrollment required from the customer. The operator reports more than 200 million eligible cards in 2026 and nine issuers (Bank of America, Capital One, Chase, Citi, Elan, PNC, Truist, U.S. Bank, Wells Fargo), while acknowledging that merchant adoption has been slow. The metric that matters for its rollout is merchant adoption, not the number of eligible cards.

PIN debit, Durbin, and the dual-routing requirement

In the US, a debit card carries at least two separate networks that the transaction can be routed over. Alongside the “signature” brand (Visa Debit, Debit Mastercard), there is at least one unaffiliated PIN network. This requirement comes from the Durbin Amendment to the 2010 Dodd-Frank Act. It is implemented by the Federal Reserve’s Regulation II (12 CFR part 235), in force since 2011. It has two parts, an interchange cap and a routing choice requirement, and for merchants the second one matters more.

The cap, by contrast, applies only to debit and only to large issuers, those with at least $10 billion in assets. It is set at 21 cents per transaction, plus 5 basis points of the transaction value, plus a 1-cent fraud-prevention adjustment. Other issuers are exempt, and the measured gap is wide. Credit card interchange is not capped at all. That is what sets the US market apart from Europe. Most of the difference in acceptance costs between the two markets comes from this difference in how credit is treated.

Credit cards: the economics of rewards and what they cost

Rewards credit cards, very common in the US, give cardholders something back for using them: airline miles, hotel points, cash back, travel insurance, or lounge access. These programs are funded by interchange, which has no legal cap, and by revolving interest. In 2024, credit cards carried 67.1 billion transactions worth $6.51 trillion, growing 9.6% a year since 2021. For the first time in a decade, credit is growing faster than debit (Federal Reserve Payments Study, 2024 data). The rising acceptance costs borne by US merchants therefore come from a product mix drifting upmarket, not from higher rate schedules: each premium card costs more than the last.

American Express is a three-party network, acting as both issuer and acquirer, so it sits structurally outside the Durbin cap and long stood outside the routing rules. Its OptBlue program hands acquiring for small merchants to third-party PSPs, a US-specific mechanism that explains why Amex is now accepted in places where it once was not. Discover Network, the only other network that is both issuer and operator, was acquired by Capital One for $35.3 billion in a deal that closed on May 18, 2025. About 25 million Capital One debit cards have already moved to the Discover network, and migration of the credit portfolio began in late July 2026. No vertical integration on this scale had happened in decades. The deal automatically takes volume away from Visa and Mastercard.

A surcharge is an extra fee charged to customers who pay by credit card, and a cash discount is a price reduction for customers who pay cash. Network rules allow surcharging on credit cards up to the merchant’s actual cost of acceptance. Visa has capped surcharges at 3% since April 15, 2023 (down from 4%), and Mastercard caps them at 4%. Surcharging is prohibited on debit and prepaid cards, even when they are run as “credit.” Merchants must notify their acquirer 30 days in advance and disclose the policy both in store and online. State law adds another layer: a few states and territories ban surcharging outright (Connecticut, Massachusetts, Puerto Rico), while others cap it, such as Colorado at 2%. The list keeps changing. Check the rules state by state before any nationwide rollout, since consumer litigation over surcharging is common.

What digital didn’t kill: checks, cash, and billers

Checks are still an everyday noncash payment instrument in the US. They carried 9.2 billion items worth $24.45 trillion in 2024, an average of about $2,650, more than 40 times the average card payment. Check use is falling fast (down 5.9% a year by number between 2021 and 2024, down 2.5% by value), but from a high base. It is concentrated in B2B, real estate, regulated professions, local governments, and insurers (Federal Reserve Payments Study, 2024 data).

The Check 21 Act, in force since October 28, 2004, digitized check presentment without doing away with checks. A bank can present an image of a check and, if needed, produce a substitute check with the same legal standing as the original. That made remote deposit capture possible: depositing a check by taking a photo with a phone. It also explains why checks have lasted so long. Businesses rely on two controls: positive pay, which matches each presented check against the file of checks actually issued, and its counterpart, ACH positive pay. Check fraud remains an active internal control issue in the US.

Bill payment in the US follows a model with no European equivalent. In Europe, billers collect by direct debit under a mandate. In the US, bills are mostly paid through an intermediary. The payer goes to a biller portal, to ACI Speedpay, or to a convenience store counter with a barcode, and the biller is credited electronically. The result is a whole layer of intermediaries, the biller service providers, that every billing software vendor has to integrate with. That is why recurring direct debit (PPD) remains a minority option in the US next to payer-initiated payments.

Cash is still a rail in its own right in the US, even though ATM withdrawals are falling in both number and value. Allpoint (NCR Atleos, 2002) is the country’s largest surcharge-free ATM network, built on machines in major retail chains. It is essential for neobanks and prepaid programs that have no ATM fleet of their own. Velera (formerly CO-OP Solutions, renamed in 2024) plays the same shared role for credit unions. For a card program aimed at mainstream US consumers, access to cash is part of the product’s basic terms of use.

A regulatory patchwork: federal, state, and private rules

US payments regulation is split across three overlapping layers, with no single payments regulator and no equivalent of PSD2. A US compliance plan has to cover all three. The federal layer involves several agencies, each with its own scope. The state layer handles licensing, fifty times over. The private layer covers Nacha rules and card network rules, contractual texts whose breach cuts off access to the rail far faster than any regulatory action.

A money transmitter license is the license a state grants to move funds on behalf of others, and it is the first practical hurdle a new entrant faces. It must be obtained in every state where customers live. Each state sets its own net worth, surety bond, and permissible investment requirements, and applications go through the NMLS. The Money Transmission Modernization Act, a model law backed by the CSBS, is gradually harmonizing these regimes. As of February 26, 2026, 31 states had adopted it in whole or in part, with Louisiana joining the list during the year and Virginia’s law taking effect on July 1, 2026 (CSBS, 2026). Harmonization is progressing, but it is neither complete nor uniform. The process takes several years, and a product launch timeline has to account for that.

• Regulation E (12 CFR 1005) governs consumer electronic fund transfers: capped customer liability for unauthorized transactions, and an error-resolution procedure with deadlines in business days. What tips the balance is whether the payment was authorized, not whether the customer suffered a loss.
• Regulation Z (12 CFR 1026) covers credit: billing dispute rights, protections specific to credit cards, and disclosure requirements. It is the legal basis for consumer chargebacks, separate from the network rules.
• Regulation CC governs the availability of deposited funds and check collection, including Check 21.
• Regulation II (12 CFR 235) sets the debit interchange cap and the requirement for two unaffiliated networks.
• UCC Article 4A governs large-value funds transfers (Fedwire, CHIPS). It is not consumer law. Risk is allocated in a fundamentally different way, and Article 4A decides who bears the loss when payment details are wrong.
• The Bank Secrecy Act and FinCEN’s implementing regulations require MSB registration, a compliance program, and regulatory reporting.

A dollar-backed stablecoin ecosystem has grown up alongside these rails. USDC (Circle, 2018) is the institutional benchmark. PYUSD (issued by Paxos for PayPal since 2023) is the first dollar-backed stablecoin created for a major payments company. RLUSD (Ripple, 2024) serves as the settlement asset in Ripple Payments. USA₮ (Tether, issued by Anchorage Digital Bank since January 27, 2026) is Tether’s way to operate under US law without bringing USDT’s global supply under it. This market can still reverse course, as two precedents show. The NYDFS ordered a halt to BUSD issuance in February 2023, and Diem was wound down in early 2022 without ever going live.

Accepting payments in the US: what breaks, what costs money, who to know

This final section sums up six issues that come up in every US market entry project, in the order in which problems tend to surface. The first three concern licensing, debit routing, and the consumer ACH return window. They drive a project’s timeline far more than technical choices do.

ROUTING
  - Do you route PIN debit as PINless on e-commerce transactions?
  - Which unaffiliated networks: STAR, Accel, NYCE, PULSE, SHAZAM?
  - Does the network mix appear in monthly reporting, line by line?

PRICING
  - Interchange++ or blended? If blended, what is the repricing clause?
  - Is the cost broken down by card type (regulated debit, exempt debit,
    standard credit, premium credit, commercial)?
  - Do you pass Level II / Level III data on commercial cards?

ACH
  - Who is the ODFI? What reserves, what limits, what holding period?
  - Account validation on the first WEB debit: included, or billed?
  - What unauthorized return rate triggers an alert on your side?

INSTANT
  - Are you connected to RTP, FedNow, or both?
  - What per-payment limit does YOUR bank apply (not the network)?
  - Do you support Request for Payment, and with which issuers?

COMPLIANCE
  - What license do you operate under: MTL in each state, partner bank, agent?
  - Who is the depository institution, and how is the insurance structured?
  - OFAC screening: which parties, how often, with what audit log?

Installment payments in the US cover three distinct business models. Interest-free pay in 4 is funded by the merchant fee (Afterpay, Klarna, Zip, Sezzle). Longer-term credit discloses its interest and targets larger purchases (Affirm, the largest homegrown player, since 2012). The third model is installment payments outside e-commerce: Sunbit finances auto repair, dental, optical, and veterinary care at the counter, at more than 30,000 service locations in 47 states. Then there is Bread Financial, a large private-label store card issuer that also offers installment plans. Its model relies heavily on late fees, which sets it apart from the BNPL specialists. PayPal Pay in 4 is enabled within the existing PayPal button, with no additional merchant integration.