You buy a $60 pair of shoes. The shop does not receive $60. Most people know that much, and most people assume the missing money went to Visa or Mastercard.
It did not. The largest piece of it went to your own bank — the one that issued the card you just tapped. That transfer is called interchange, and once you can see which direction it flows, a surprising amount of everyday card behaviour stops being mysterious: why the corner shop has a card minimum, why some merchants nudge you toward debit, and where the cashback on a premium card actually comes from.
The one-sentence answer
Interchange is a payment from the acquirer — the merchant’s bank — to the issuer — the bank that gave you your card. Visa describes its own version as exactly that: “Visa uses interchange reimbursement fees as transfer fees between acquiring banks and issuing banks for each Visa card transaction.”
Two consequences follow, and both are counter-intuitive.
The card network is not the recipient. Visa and Mastercard set and publish the schedules, and they charge both banks separately for running the rails — those are scheme fees, a different and much smaller line. Interchange itself passes between the two banks. Blaming “the network fee” for a merchant’s card costs points at the wrong number.
The merchant does not literally pay interchange. Visa is blunt about this too: “Merchants do not pay interchange reimbursement fees — merchants negotiate and pay a ‘merchant discount’ to their financial institution that is typically calculated as a percentage per transaction.” A merchant has one commercial relationship, with its acquirer. Interchange is the biggest input into the price that acquirer quotes, so the cost lands on the merchant regardless — but it lands as one component of a bill, not as a fee the merchant hands to a bank it has never met.
Notice where interchange physically happens: on the interbank leg. When the two banks square up, the issuer sends the acquirer the sale amount less interchange. Nobody writes a cheque labelled “interchange.” It is withheld. That leg runs on the network’s settlement cycle, which is a separate timeline from when the shop’s account is credited — the gap is mapped out in clearing vs settlement.
For how the full merchant discount splits between issuer, network, and acquirer, see the four-party model.
Why the money runs that way
Card payments are a two-sided market. The system needs shoppers and merchants at the same time, and the costs of getting them are lopsided.
The issuer carries almost everything expensive: it fronts the money on a credit card, absorbs fraud losses when a stolen card gets used, staffs a call centre for disputes, runs the risk engine that answers in under a second, and funds whatever perks made you pick that card. The merchant, meanwhile, captures a sale that would otherwise have walked out of the shop.
Interchange moves value from the side that captures the sale to the side that carries the cost and the risk. Every design decision downstream — rewards budgets, which cards a merchant likes, what regulators eventually did about it — is a consequence of that one flow.
What moves the rate
There is no single interchange rate. A network’s published schedule runs to hundreds of rows, and which row a given tap lands on depends on four things.
Card type. Debit costs the merchant side less than credit; the issuer is not lending money or carrying a revolving balance. Premium consumer and commercial cards sit at the top. The pattern is visible even in law: the EU’s interchange regulation sets a lower ceiling for consumer debit than for consumer credit.
Merchant category. Every merchant carries a four-digit MCC (Merchant Category Code). Supermarkets, fuel, utilities, transit, and charities commonly sit in cheaper categories — high volume, thin margins, or public interest. Restaurants, travel, and general retail generally do not.
How the card was presented. A chip or contactless tap with the card physically there is the cheapest evidence there is. A number typed into a website is the most expensive, unless the merchant supplies something that closes the gap — 3-D Secure authentication or a network token instead of a raw card number. Those checkout steps that feel like friction are partly a merchant buying a better rate.
Geography. Domestic — issuer and acquirer in the same country — is cheaper than cross-border, which adds currency conversion, longer dispute chains, and more fraud exposure. This is one reason a card that behaves cheaply at home can be expensive for a merchant overseas; the full route is in cross-border card payments.
None of these are negotiable by the merchant. A shop negotiates its acquirer’s markup. The interchange underneath is set by the schedule and determined by facts about the transaction.
Downgrades: paying more for a paperwork error
Each best-rate row comes with conditions — submit the transaction for clearing within a time window, include the authentication result, populate the enhanced data fields, match the authorization to the presentment. Miss one and the transaction is downgraded to a costlier row.
There is no error message and no alert. The merchant simply finds a worse number on the month-end statement. A misconfigured terminal, a batch closed late on a bank holiday, or a checkout that stopped sending 3DS results can quietly move a slice of a shop’s volume to a higher tier for weeks.
For a cardholder, this explains small mysteries: the extra authentication step, the receipt that arrives a day late, the shop that insists on the chip rather than the tap. Someone upstream is protecting a rate. If you want to see how the layers stack on a specific card, our fee calculator models them.
Why regulators capped it, and what happened next
Interchange has an unusual property: the party paying it does not choose it, and the party choosing it does not pay it. You pick the card, the merchant absorbs the cost. Competition between networks for issuers therefore pushes rates up, not down, because a higher rate is what attracts banks to issue your card. That inversion is why regulators intervened rather than leaving it to the market.
In the EU, Regulation (EU) 2015/751 caps interchange on consumer debit and consumer credit cards. Its definition of the fee matches the direction described above: a fee “paid for each transaction directly or indirectly … between the issuer and the acquirer involved in a card-based payment transaction.” The carve-outs are as important as the cap: the limits do not apply to commercial cards issued to businesses, to cash withdrawals, or to three-party schemes where one company both issues cards and signs merchants.
In the US, the Durbin Amendment and the Federal Reserve’s Regulation II take a narrower and more surgical approach. It covers debit only, and only for large issuers — “certain small debit card issuers, government-administered payment programs and reloadable general-use prepaid cards are exempt from the interchange fee limitations.” It also does something the EU rule does not: it forces routing choice. Issuers and networks may not restrict debit transactions “to less than two unaffiliated networks,” and may not inhibit “a merchant’s ability to direct the routing of a debit transaction over any network that the issuer has enabled to process it.”
The follow-on effects are the part worth internalising.
- The money relocates to the carve-outs. When consumer cards are capped and commercial cards are not, issuing commercial cards becomes relatively more attractive. When domestic is capped and cross-border is not, cross-border volume matters more. Regulation reshapes the product mix, it does not delete the incentive.
- Reward budgets tighten where caps bite. Issuer rewards are funded substantially by interchange. Cap the input and the output shrinks — expect thinner cashback, annual fees, or perks moved behind a subscription in heavily capped markets, and richer offers in uncapped ones. If a card’s rewards look far better in one region than another, the regulatory line is usually the reason.
- Routing rules turn debit into a bidding market. Where a merchant can choose among networks for the same debit card, networks compete on the merchant’s side rather than only on the issuer’s — a different competitive shape entirely from credit.
What interchange explains at the till
Merchants steering you to debit. Debit is materially cheaper for them to accept. A prompt asking “debit or credit?”, a discount for one and not the other, or a surcharge on credit are all the same instinct.
Card minimums at small shops. Interchange schedules typically pair a percentage with a fixed amount per transaction. The percentage scales with the sale; the fixed piece does not. On a $3 coffee that fixed component is a much larger share of the ticket than on a $300 jacket, which is why small merchants set minimums and why some of them still prefer cash.
Where premium-card perks come from. Lounge access and generous cashback are largely paid for by the acceptance side. A premium card is, mechanically, an arrangement where merchants across the economy subsidise the benefits of the people carrying the most valuable cards.
Why small merchants use flat-rate aggregators. A single blended price per transaction is easy to understand and easy to budget, and the aggregator absorbs the risk that a customer turns up with an expensive card. The trade is visibility: a blended merchant never learns that its costs are being driven by one card type or one broken checkout setting. Larger merchants move to itemised pricing — interchange, scheme fee, and acquirer markup listed separately — precisely so they can find and fix the downgrades.
Why prepaid and crypto cards have strange economics. Most crypto cards are legally prepaid or e-money products riding debit-style rails, and merchants generally accept them without complaint. But the exemptions cut in the issuer’s favour: in the US, reloadable general-use prepaid cards are explicitly exempt from Regulation II’s caps, so a prepaid program is not held to the ceiling that applies to a large bank’s debit card. That is one of the quieter reasons prepaid programs are attractive to launch. It is also why crypto-card cashback rarely rests on interchange alone — subscriptions, token lockups, and the conversion spread do a lot of the work. Our card comparison tool lays the offers side by side.
Quick answers
- Who pays interchange? The acquirer pays it to the issuer. The merchant bears it indirectly, inside the merchant discount its own bank charges.
- Does Visa or Mastercard keep it? No. They set and publish the schedules and charge both banks separate scheme fees; interchange transfers between the banks.
- Can a merchant negotiate its interchange rate? No. It negotiates the acquirer’s markup and can influence which rate row it qualifies for by how it submits transactions.
- Why is my card cheaper for a shop than my friend’s? Different card type, different rate row. Debit and basic credit sit below premium and commercial.
- Does interchange come out of my purchase? Not from your side of it. You are billed the full amount; the deduction happens on the merchant’s side of the ledger.
- When is the fee actually applied? At clearing, when the network prices the transaction — see clearing vs settlement and the fee stack in the payment authorization guide.
Interchange will not change what you personally pay for the shoes. It will explain, reliably, why the shop wanted your debit card, why the checkout asked for a code, and why the card in your other pocket earns three times the cashback.