Interchange moves over $100 billion a year from merchants to card-issuing banks. The networks set it, the issuers receive it, the merchants pay it — and three continents regulate it three different ways.
You tap your card for a $100 dinner. The restaurant receives about $96.80. Three parties split the missing $3.20, and the biggest slice — around $2.30 — goes to your own bank, the one that issued your card.
That slice is called interchange. You never see it, the restaurant can't refuse it, and it quietly pays for your cashback and airport lounges. This chapter explains who sets it, why it's different on every card and in every country, and what happens when governments try to cap it.
The same $100 credit-card purchase generates wildly different interchange depending on where it happens. This single chart explains half of global payments politics:
Eight terms do the heavy lifting in every fee conversation. Plain definition, real example, why you should care.
On every card purchase, the merchant's side pays a fee to the bank that issued the card. The network sets it, the issuer receives it, the merchant pays it.
Why it matters: it moves $100B+ a year, sets the price of card acceptance worldwide, and is the most fought-over number in payments.
The bank that gave you your card. It approves or declines every tap, lends you the money on credit cards, eats the loss if you never pay — and receives interchange.
Why it matters: rewards cards exist because issuers compete for you with money they collect from merchants.
The bank (or payment company standing in front of one) that lets a shop accept cards, collects the money from issuers, and deposits it minus fees.
Why it matters: everything a merchant can negotiate — rates, payout speed, disputes — happens with the acquirer, never with Visa or your bank.
The total percentage a merchant loses on a card sale: interchange + network fees + the acquirer's own markup, blended into one number.
Why it matters: when merchants complain about "card fees," they mean MDR. Interchange is its biggest ingredient, but not the whole recipe.
1 basis point = 0.01%. The industry negotiates fees in basis points because the volumes are so large that hundredths matter.
Why it matters: read any payments contract or news story and the units are bps. Now you can convert instantly: 100 bps = 1%.
A 4-digit code assigned when a merchant signs up, declaring its line of business. Interchange tables price by MCC — groceries and fuel get discounts, small retail pays standard.
Why it matters: a merchant's 4-digit label is destiny — it moves rates, rewards eligibility, and whether transactions get declined. The MCC lookup tool decodes them all.
Card-present: the physical card met the terminal (chip, tap). Card-not-present (CNP): the number was entered remotely — online, in-app, by phone. CNP fraud runs ~3× higher, so CNP interchange prices higher.
Why it matters: this is why online businesses pay more to accept cards, and why they spend so much on fraud screening — every fraud-risk improvement is priced in bps.
The fee the card network (Visa, Mastercard) charges per transaction — around 0.13–0.14% per side in the US, plus small per-item fees. Separate from interchange, which the network sets but never touches.
Why it matters: it explains the networks' business model (tiny toll, planetary volume) and where extra fees like international assessments get attached.
US interchange tables run to hundreds of rate categories. They all reduce to four dials:
A basic card might carry ~1.5% interchange; a premium travel card ~2.4% or more, plus 10¢. The richer the rewards, the higher the interchange that funds them. Example: a $4 coffee on a metal premium card can cost the café more than double what the same coffee costs on a basic debit card.
Card-present (chip/tap) earns the lowest rates. Online (card-not-present) pays ~0.3–0.5% more because fraud runs roughly 3× higher. Keyed-in phone orders pay the most. The interchange table is a fraud-risk map.
Supermarkets, fuel stations, and utilities get discounted categories (thin margins, huge volume, political leverage); small retail pays standard rates. Example: US supermarket interchange categories can run ~1.2–1.5% while small-ticket retail pays 2%+. Your 4-digit MCC code is destiny.
Most rates are % + fixed fee (e.g., 1.65% + 10¢). On a $200 purchase the 10¢ is noise; on a $2 purchase it's 5% by itself. This is why some kiosks and newsstands set card minimums — and why micro-payments never worked on card rails.
One $100 restaurant bill on a premium US rewards card. Here is everyone's position when the books close tonight:
A $100 café sale on a premium credit card, and what the shop keeps — the regulation gap made visible:
Whenever interchange grows big enough, governments step in — every continent has its own battle:
Capped big-bank debit at 21¢ + 0.05% (+1¢ fraud adjustment) — but left credit untouched, which is why US credit interchange stayed the rich world's highest and rewards cards flourished. Still contested: a 2025 federal court ruling vacated the Fed's fee standard, reopening the fight. Meanwhile a 2024 antitrust settlement trims credit rates ~10bps over five years.
Consumer cards capped at 0.3% credit / 0.2% debit, full stop. Result: rewards cards mostly vanished from Europe, annual fees rose — and merchants saved billions. Pick your trade-off: US-style rewards or EU-style merchant costs. You can't have both.
The Reserve Bank moved first, capping credit interchange via weighted averages (~0.5%) and — uniquely — allowing surcharging so card costs became visible at checkout. Australia is the long-run natural experiment everyone else cites.
India didn't cap the fee — it abolished it for UPI and RuPay debit merchant payments. Adoption exploded to 23B+ transactions a month; the unresolved question — who funds the rails? — is now a permanent policy debate. The most radical answer any country has given.
Five questions people actually ask about card fees, answered without hedging.
The rate card behind the rate card: downgrades, the fee split, commercial cards, and crossing borders.