Every payment processor shows you their rates. None of them explain how they actually make money.
The published rate (2.6% + $0.10, 2.9% + $0.30) is just the visible layer. Underneath it sit revenue streams that never appear on any merchant statement: interchange variance, float income on reserves, float income on settlement timing and network rebates at scale. These hidden economics generate billions annually and explain why certain business models dominate while others struggle to compete.
Interchange variance: the margin inside flat pricing
PayFacs like Square and Stripe price differently than traditional processors. The difference matters more than most merchants realize.
Traditional acquiring uses interchange-plus pricing. The processor passes through interchange (which varies by card type) and adds a fixed markup. A small merchant might pay interchange plus 50 basis points; a large merchant might negotiate down to interchange plus 8. The processor’s margin is visible and compresses as merchants gain power.
PayFacs capture the entire merchant discount rate because they are the merchant of record to the card networks. When Square prices at 2.6% + $0.10, they’re not marking up interchange. They’re setting a flat rate that captures interchange variance.
The variance is where the real margin lives. Consider a $100 transaction:
A regulated debit card (Durbin-covered) costs roughly $0.26 in interchange. Square collects $2.70. That’s 244 basis points of margin.
A corporate rewards card costs roughly $2.50 in interchange. Square still collects $2.70. That’s 20 basis points.
Same pricing, wildly different economics. PayFac profitability depends almost entirely on portfolio mix. Consumer-focused merchants with high debit usage (coffee shops, quick-service restaurants, food trucks) generate dramatically better economics than B2B merchants taking corporate cards.
This explains merchant category restrictions that seem arbitrary. The prohibition on certain business types isn’t just risk management. It’s margin protection. PayFacs quietly discourage categories with unfavorable interchange profiles even when the risk is perfectly acceptable.
Reserve float: the profit center on holdbacks
The standard explanation for reserves (10% holdback for 180 days to cover chargeback windows) describes the structure but misses the economics.
Reserves aren’t just risk mitigation. They’re float income. An acquirer holding $100 million in merchant reserves at 5% rates earns $5 million annually. Pure margin that appears nowhere on merchant statements.
The 180-day hold ties to chargeback windows. Card network rules allow disputes up to 120 days from transaction date for most cases, extended to 540 days for certain fraud claims. The 180-day window covers common cases while releasing funds before the tail risk period. But that timeline is a starting point, not a rule.
Reserve structures vary more than processors admit:
Upfront reserves require capital deposits before processing begins. Common for travel, events and subscription businesses with long fulfillment windows. Typically 50-100% of estimated monthly volume.
Rolling reserves withhold a percentage of each settlement, releasing on a lagged basis. A 10% rolling reserve with 180-day release means your day-one holdback releases on day 181. You reach steady state around day 180, then the reserve stays constant.
Velocity-based reserves trigger when processing patterns deviate from baseline. A business averaging $50,000 monthly that suddenly processes $200,000 might see 50% or more of the spike held pending review.
Negotiating room exists because reserves are profit centers:
Processing history transfers. Bring statements from previous processors showing chargeback rates below 0.5%. Acquirers will reduce reserves based on documented history, but you have to ask.
Collateral substitution. Some acquirers accept letters of credit or CDs in lieu of rolling reserves. Your capital stays productive while they maintain protection.
Reserve caps. Rather than straight percentages, negotiate caps: 10% of processing up to $100,000 maximum.
Volume commitments. Processors accept higher per-merchant risk in exchange for guaranteed revenue. A three-year processing commitment can meaningfully reduce reserve requirements.
Every dollar released early costs the acquirer yield. Knowing this changes the negotiation.
Settlement float: the overnight money machine
Payment processors control timing between when they receive funds and when merchants receive settlement. That gap generates float income that subsidizes published processing rates.
Card network settlement happens on a 2-3 day cycle. The processor receives funds, pools them across all merchants and distributes through ACH. During the holding period, those funds sit in interest-bearing accounts.
Block’s financials make this explicit. Their 2023 10-K shows Cash App and Square holding billions in customer funds. Q3 2024 reported $303 million in interest income from corporate cash and customer balances. That figure grew substantially as rates rose from near-zero.
PayPal’s 2023 annual report shows $36.2 billion in customer accounts and funds receivable. At current rates, that position generates hundreds of millions annually in interest income.
A processor with $10 billion in daily transaction volume might average $15-20 billion in float when accounting for weekends, holidays and settlement timing. At 5% rates, that’s $750 million to $1 billion annually. Revenue that doesn’t appear on any merchant pricing page but absolutely affects what rates they can offer.
Instant payouts reveal the economics explicitly:
Square charges 1.75% for instant deposits. Stripe charges 1% for Instant Payouts. PayPal charges 1.75% for instant transfers.
These fees seem steep for moving money that already belongs to the merchant. But the processor isn’t taking duration risk. They’re advancing funds against tomorrow’s certain settlement. They already hold your money; they’re releasing it early. At 1.75% for a two-day advance, the annualized rate exceeds 300%.
A business processing $100,000 monthly using instant deposits on every transaction pays $1,750 annually in fees. That’s roughly the interest income the processor would otherwise earn on that merchant’s float.
Network rebates: scale economics that smaller players can’t access
Above $1 billion in annual processing, Visa and Mastercard offer volume incentives that meaningfully improve unit economics. A processor doing $5 billion might earn 5-8 basis points back on every transaction.
These rebates flow straight to margin. They’re not passed through to merchants. And they’re structurally unavailable to smaller competitors.
This creates a compounding advantage. Higher volume generates rebates. Rebates fund more competitive pricing. Competitive pricing attracts more volume. The flywheel is difficult to break into.
Network rebates also explain why processors fight hard to retain large merchants even at thin visible margins. The rebate economics might make a seemingly unprofitable relationship quite profitable when the full picture is visible.
How these economics compound
These revenue streams don’t exist in isolation. They compound.
A PayFac earns interchange variance on every transaction, float income during settlement windows, interest on reserves they hold and network rebates at scale. Layer those together and you understand why Square can price at 2.6% flat while traditional processors struggle at interchange-plus models with better headline rates.
The merchant sees: 2.6% + $0.10 The processor captures: interchange variance + settlement float + reserve float + network rebates
This is why the industry resists transparency. Interchange-plus pricing looks merchant-friendly until you realize processors make their real margin on timing, not spread. Flat-rate pricing looks expensive until you understand the full stack of economics it bundles.
Merchants who negotiate effectively understand all four layers. Optimizing visible rates while ignoring float and reserves leaves value uncaptured. Negotiating power exists precisely because these are profit centers, not cost centers.
The shift toward real-time payments and regulatory scrutiny might change some of these dynamics. But as long as money sits somewhere between transaction and settlement, someone will find a way to earn yield on it.