Margin capture across the payment stack

You swipe your card for a $100 dinner. The restaurant never sees $100. By the time the money settles into its bank account, roughly $97 lands, and the missing $3 has already been split between six or seven companies you've never heard of, each defending a sliver of margin that gets thinner every year. Understanding exactly who takes what, and why, is the fastest way to understand power in fintech.

The anatomy of a $100 transaction

Follow the money from swipe to settlement:

• Interchange fee (~1.5-2.0% in the US, capped at 0.3% for credit and 0.2% for debit in the EU under the Interchange Fee Regulation): paid by the merchant's bank to the cardholder's bank. This is the single largest slice, and it's set by card networks but pocketed by issuing banks (Chase, Capital One, BBVA).
• Network/scheme fee (~0.13-0.15%, estimate): kept by Visa or Mastercard for running the rails, authorization, clearing, and fraud tools.
• Processor/acquirer markup: taken by the company that connects the merchant to the networks (Stripe, Adyen, Fiserv). Highly variable, often 0.1-0.5% plus a flat fee.
• Fintech/ISV take rate: if a software platform (Toast, Shopify, Square) is embeddingembeddingAn embedding is a numerical vector that represents data (text, images, or items) in a way that captures meaning, so similar items sit close together in space.View full definition → payments, it adds its own margin on top, often 0.3-1% or a per-transaction fee.

So on $100: roughly $1.80 goes to the issuing bank, $0.15 to the network, $0.20-0.50 to the acquirer/processor, and the embedded fintech might keep $0.30-1.00. The merchant nets somewhere around $96-97. These are illustrative, order-of-magnitude splits, actual contracts vary by industry, card type, and negotiating power.

This is why "payments" isn't one business. It's a stack, and each layer has a different owner, a different regulator, and a different growth ceiling.

Who sits where: the real players

Issuers hold the deepest pool historically: interchange plus interest income on revolving credit. But they carry credit risk and regulatory capital burden, and in the US they're constrained by the Durbin Amendment (part of Dodd-Frank), which caps debit interchange for large banks.

Card networks (Visa, Mastercard) run a two-sided platform with extraordinary operating leverageoperating leverageThe degree to which a company's cost base is fixed rather than variable, which magnifies how profit reacts to changes in revenue.View full definition →: they don't lend, don't hold deposits, and don't take credit risk. Their toll-road position is why they trade at software-like margins despite being 60-year-old infrastructure. Together they process the vast majority of global card volume, a duopoly that regulators watch closely on both sides of the Atlantic.

Acquirers/processors (Fiserv, Global Payments, Adyen, Stripe) are the plumbers connecting merchants to networks and issuers. This layer has seen the most disruption: legacy acquirers built on batch processing lost share to APIAPIApplication Programming Interface: a standardised interface that lets applications communicate and exchange data without knowing each other's internal workings.View full definition →-first players like Stripe and Adyen, who won by selling developer experience, not basis points.

Embedded fintechs and platforms (Toast for restaurants, Shopify for e-commerce, Square/Block for SMBs) sit closest to the end customer. They monetize payments as a feature of a broader software product, which is why they can often out-earn pure payment processors: they're not just taking a payment margin, they're taking a share of the merchant's total software spend.

Regulators shape all of this without taking a cut. The US lacks a single interchange regulator for credit (debit is capped via Durbin), while the EU's Interchange Fee Regulation directly caps both credit and debit interchange, which is a major reason European card economics look structurally different from American ones. The Consumer Financial Protection Bureau and, in the EU, the European Banking Authority, oversee consumer protection and open banking mandates (like PSD2, the EU's second Payment Services Directive) that reshape who can access payment data and rails at all.

Why margins compress as you move down the stack

New entrants rarely attack interchange or network fees directly, those are protected by regulation, scale, and bank relationships that took decades to build. Instead, they compete at the acquirer and software layer, where differentiation is technical (APIs, uptime, fraud tools) rather than structural.

The problem: that layer is also the easiest to commoditize. Once Stripe proved API-first acquiring was viable, dozens of copycats emerged, and price competition intensified. A processor charging 30 basis points in 2015 might be defending 10-15 basis points in 2026, an estimate, but directionally accurate for a market where merchants increasingly multi-home across processors and negotiate hard.

The way out is vertical integration: capturing more of the stack instead of a thinner slice of one layer. Block's acquisition of Afterpay, or Shopify building Shopify Payments instead of just plugging into Stripe, are both attempts to move from "fee on a transaction" to "share of merchant lifetime valuelifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition →."

A worked example: same transaction, two structures

Pure processor model: A standalone acquirer processes a $100 transaction, keeps 20 basis points ($0.20), pays out everything else. Volume is the only lever. To make $1 million in margin, they need to process $500 million in volume.

Embedded software model: A vertical SaaS platform (like Toast) processes the same $100 transaction but also charges the restaurant a monthly software fee and takes a slightly higher blended payments margin, say 60 basis points, because payments are bundled with point-of-sale software, inventory, and payroll tools the merchant won't easily switch away from. To make $1 million in margin from payments alone, they need only about $167 million in volume, and they have a second, stickier revenue line on top.

This is the strategic logic behind almost every "fintech becomes a bank" or "software company adds payments" move you'll see in the market. Owning the customer relationship lets you capture margin that a pure infrastructure player cannot.

Power dynamics: who can squeeze whom

The balance of power is asymmetric and shifting:

• Networks over everyone: Visa and Mastercard set interchange schedules and network rules that issuers, acquirers, and merchants must simply accept. Their bargaining power comes from network effects: merchants must accept the cards consumers carry.
• Large merchants over acquirers: Amazon or Walmart can negotiate custom processing rates far below list price, and can credibly threaten to build in-house payment infrastructure.
• Regulators over networks and issuers: The EU's interchange caps directly transferred billions in margin from banks to merchants. The US has seen periodic legislative pushes (the Credit Card Competition Act, still debated as of 2026) to extend Durbin-style caps to credit interchange, which would reshape issuer economics again.
• Big Tech as a wildcard: Apple Pay and Google Pay don't charge consumers, but Apple has historically taken a small cut from issuers for transactions on its wallet, another thin layer inserted into the stack, illustrating how new intermediaries can still find room to extract margin even in a "mature" system.

🎬 [VIDEO: "How Visa and Mastercard Make Money" - youtube.com/@CNBC - a concise breakdown of the network business model and why it scales so profitably without taking credit risk]

For a deeper technical reference on how funds actually move between banks, the Federal Reserve's payment systems overview is a solid, free primer on clearing and settlement infrastructure in the US.

Key Takeaways

• A card transaction splits margin across issuers, networks, acquirers, and embedded fintechs, each layer owned by different players with different regulatory exposure and negotiating leverage.
• Interchange (paid to issuing banks) is usually the largest fee component; it's capped in the EU by regulation but largely uncapped for credit in the US, creating structurally different economics across regions.
• New entrants tend to compete at the acquirer/software layer because interchange and network fees are protected by regulation and scale, but that layer commoditizes fast, compressing margins for pure processors.
• Vertical integration (bundling payments with software, lending, or a broader platform) is the primary strategy fintechs use to escape margin compression and capture more of the total transaction value.
• Regulatory intervention (Durbin Amendment, EU Interchange Fee Regulation, PSD2) is not a side issue in payments, it is one of the primary forces redistributing margin between banks, networks, and merchants.