pay.2nth.ai Tree rails card-schemes
rails · Card Schemes · Leaf

The rules everyone pays to follow.

A card scheme is not a bank and rarely touches your money. It is a rulebook, a brand, and a clearing network — and it sets the economics for every party that plugs in. Understand the four-party model and the rest of cards falls into place.

Four-party model Interchange Scheme fees Mandates Domestic schemes

A rulebook with a logo on the front

Schemes monetise reach and rules, not balances.

A card scheme (Visa, Mastercard, Amex, mada, Jaywan) is the entity that defines how a card payment moves between the cardholder’s bank and the merchant’s bank. It owns the brand, the technical standards, the dispute rules, and the fee structure — but in the dominant model it does not issue cards, does not sign up merchants, and does not lend money.

The mental trap is treating the scheme as “the payment company.” It is closer to a standards body that also runs a switch and charges rent. The cards in wallets, the terminals in shops, and the credit risk all live with banks. The scheme sets the terms on which those banks cooperate — and competes hard for which logo ends up on the plastic.

The four-party model

Four roles. The scheme sits in the middle and never holds the cardholder relationship.

The default structure for Visa and Mastercard is the four-party model — four roles, not four companies (the scheme itself sits in the middle as a fifth actor that does not take a financial position in the transaction).

RoleWhoWhat they do
CardholderThe consumer or businessHolds the card, makes the purchase
IssuerCardholder’s bankIssues the card, carries credit risk, approves or declines, receives interchange
AcquirerMerchant’s bank / PSPSigns up the merchant, routes transactions, pays interchange, sets the merchant discount rate
MerchantThe shopAccepts the card, pays the merchant service charge
SchemeVisa / MastercardOperates the network, sets rules and fees, settles between issuer and acquirer

Money flows acquirer → scheme → issuer at settlement; interchange flows the other way (acquirer pays issuer). The scheme nets the positions and instructs settlement — it is a switch and a clearing house wearing a brand.

When the scheme is also the bank

Closed-loop means higher capture, thinner acceptance — and increasingly, hybrid models.

In a three-party (closed-loop) scheme — classic American Express, Diners Club — the scheme is the issuer and the acquirer. There is no separate interchange-paying acquirer; the scheme deals directly with both cardholder and merchant. That is why Amex historically charged higher merchant fees and why acceptance was thinner: one party captured the whole spread.

The line has blurred

Amex and Diners now license banks to issue and acquire on their networks in many markets, making them look four-party in practice. Treat “three-party” as a model, not a permanent label for a brand.

Regulators noticed

The EU and others applied interchange-style caps to three-party schemes once they operate through licensed third parties — the closed-loop exemption is narrower than it was.

mada, Jaywan, and the South African gap

Local rails for local spend; co-badging for the rest. South Africa is a notable holdout.

A domestic card scheme is a nationally-owned network that keeps card processing — and the fees — inside the country. The economic argument is sovereignty and cost: routing local transactions through a local switch avoids paying international scheme fees on purely domestic spend.

mada (Saudi Arabia)

The national debit scheme, mandatory acceptance for merchants, with roughly 93% share of card payments. Cards are typically co-badged with Visa or Mastercard for use abroad — domestic rails for local spend, international rails for travel.

Jaywan (UAE)

Launched in 2024 by Al Etihad Payments, a Central Bank of the UAE subsidiary. The UAE’s first domestic scheme, co-badgeable with Visa, Mastercard and UnionPay for global reach.

South Africa — no domestic scheme

As of 2026 South Africa has no domestic card scheme; the market is split roughly Visa ~51% / Mastercard ~48%. The SARB consulted on feasibility (2021) but has not launched one. Domestic-rails ambition currently runs through account-to-account (PayShap), not cards.

The pattern

India’s RuPay, Turkey’s Troy, Brazil’s Elo — the same playbook. A central bank or bank consortium builds local rails to cut cost and dependence on two US networks.

Who pays whom, and why it is contested

Interchange funds the issuer. Scheme fees fund the scheme. Merchants feel both.

Two separate fees get confused constantly. Interchange is paid by the acquirer to the issuer — it is the largest component of what a merchant pays, and the scheme sets the rate but does not keep it. Scheme fees (assessment, switching, licensing, data fees) are what the scheme actually earns — charged to both issuers and acquirers for using the network.

FeePaid byPaid toSet by
InterchangeAcquirerIssuerScheme (capped by regulators in EU/UK/AU)
Scheme / assessment feesIssuer & acquirerSchemeScheme
Merchant service chargeMerchantAcquirerAcquirer (bundles interchange + scheme + margin)

Interchange is the most regulated number in payments: the EU caps it at 0.2% (debit) / 0.3% (credit) on consumer cards; South Africa runs a regulator-set interchange determination through the SARB. Scheme fees are far less transparent and have drifted upward as interchange got capped — the schemes recovered margin where regulators were not looking.

The lever that moves the whole market

Connect to a scheme and you inherit its calendar.

The scheme rulebook is binding contract law for everyone who connects. It governs chargebacks, fraud liability, brand usage, security mandates (EMV, PCI, 3-D Secure), and acceptance rules. Schemes enforce change through mandates — dated requirements that issuers and acquirers must implement or face fines and liability shifts.

Mandates are non-negotiable deadlines

A scheme mandate (e.g. a new authentication requirement or message-format change) lands with a compliance date. Miss it and you eat liability shifts or get fined. Roadmaps in the card world are largely set by the schemes’ release calendars, not yours.

The liability shift is the real weapon

Schemes rarely force behaviour directly; they move chargeback liability to whichever party failed to adopt (EMV, 3DS). The economics do the enforcing.

Reading the card economics

You inherit the rules. Compete on execution, not on the rulebook.

If you are a merchant or PSP, the number that matters is the all-in MSC, not the headline rate. Acquirers quote blended or interchange-plus pricing; interchange-plus is almost always more honest because it separates the regulated pass-through (interchange) from the scheme fees and the acquirer’s own margin. A blended rate hides where your money goes.

When a domestic scheme is worth it

Domestic rails pay off only at national scale with regulator backing — you are building a public utility, not a product. For a single PSP, the answer is integrate the existing schemes well, not build your own.

Cost of getting fees wrong

Mispricing interchange categories (wrong MCC, wrong card-present/not-present flag) leaks margin on every transaction and is invisible until someone audits the scheme-fee line. In high-volume acquiring this is where money quietly disappears.

In South Africa specifically

There is no domestic card alternative to fall back on, and interchange is regulator-determined, so your lever is operational: correct transaction qualification, surcharging rules, and routing. Card strategy here increasingly means a card-vs-PayShap conversation, not a Visa-vs-Mastercard one.

The honest summary: you cannot out-negotiate a scheme on rules or mandates, and you rarely beat interchange caps. The wins are in qualification, transparency of pricing, and knowing which transactions belong on cards versus account-to-account rails at all.

Where this sits in the tree

Primary sources