Buy Now Pay Later Economics: How BNPL Actually Works

Buy now, pay later has grown from a niche checkout option into a $560 billion global market. In 2025, over 91 million Americans used BNPL products, and providers like Klarna, Affirm, and Afterpay processed tens of billions in gross merchandise volume per quarter. But the business model behind BNPL is widely misunderstood: consumers assume it is free, merchants assume it is expensive, and regulators are still figuring out what category it belongs in. This article breaks down the actual economics.

What BNPL Is and How It Differs From Credit Cards

BNPL is a point-of-sale lending product. The consumer splits a purchase into installments (typically four payments over six weeks), and the BNPL provider pays the merchant upfront in full, minus a fee. The consumer repays the provider over time.

The core distinction from traditional credit cards: in a card transaction, the issuing bank extends revolving credit, and revenue flows primarily through interchange fees (1.5-2.5%) and consumer interest on carried balances. In BNPL, the provider is simultaneously the lender and the payment method. There is no revolving credit line, no card number, and (in the pay-in-4 model) often no interest charged to the consumer.

This creates a fundamentally different incentive structure. Card networks profit from transaction volume and outstanding balances. BNPL providers profit primarily from merchant discount rates that are significantly higher than card interchange. The question is why merchants agree to pay more.

The Major Players and Their Scale

Four providers dominate BNPL globally, each with a distinct revenue model and market positioning.

Klarna went public in September 2025, listing on the NYSE at a $15.1 billion valuation after raising $1.37 billion. Affirm reached GAAP operating profitability in its June 2025 quarter, posting $69.2 million in net income. These milestones suggest the BNPL model is maturing past the growth-at-all-costs phase: for deeper analysis of how payment providers generate revenue, see card network economics.

Revenue Streams: Where the Money Comes From

BNPL revenue is not monolithic. Providers earn from four to five distinct streams, weighted differently depending on their product mix.

Merchant Discount Rate

The merchant discount rate (MDR) is the fee merchants pay on each BNPL transaction. It typically ranges from 2% to 8% of the transaction value, with an average of 4% to 6%. This is roughly double the total cost of a standard credit card transaction (2.5-3.5% including interchange, network fees, and processor markup).

For Klarna, merchant fees represent approximately 57% of total revenue (around $1.6 billion in 2024). For Affirm, merchant network revenue accounted for 27.4% of FY2025 revenue ($882.7 million), with the lower percentage reflecting Affirm's heavier reliance on consumer interest from longer-term loans.

Consumer Interest on Longer-Term Loans

Short-term pay-in-4 products are typically interest-free, but longer-term financing (6, 12, or 24 months) carries APRs ranging from 0% to 36%. Affirm's interest income reached $1.608 billion in FY2025: 49.9% of total revenue. Klarna earns approximately 24% of revenue from interest income. This stream is where BNPL most resembles traditional consumer lending.

Late Fees

Klarna and Afterpay charge late fees on missed payments, generating roughly 9% of Klarna's revenue (~$254 million in 2024). Affirm charges zero late fees as a stated company policy: their entire model depends on merchant fees and interest income. The late fee question is central to the regulatory debate, because it determines whether BNPL functions more like a payment product or a consumer credit product.

Additional Revenue

Klarna has built an advertising business worth roughly 6% of revenue, monetizing its shopping app's traffic. Both Klarna and Affirm earn card network revenue from their branded card products (Klarna Card, Affirm Card), generating interchange on transactions that flow through Visa or Mastercard rails.

Revenue composition matters: Affirm and Klarna look similar from the consumer side, but their economics are structurally different. Affirm earns half its revenue from interest, meaning its unit economics depend on loan duration and credit performance. Klarna earns over half from merchant fees, meaning its economics depend on transaction volume and merchant retention.

Why Merchants Pay More Than Card Interchange

The most common question about BNPL economics: why would a merchant pay 4-6% when they could process a credit card for 2.5-3.5%? The answer comes down to measurable conversion benefits and risk transfer. For a broader look at what merchants pay across all payment methods, see merchant payment acceptance costs.

Conversion and Basket Size

• Checkout conversion rates increase 20-30% when BNPL is offered as a payment option
• Average order values rise 15-40% compared to non-BNPL transactions
• Up to 40% of BNPL-driven sales come from new customers who would not have purchased otherwise

For a retailer with 30% gross margins, a 25% increase in average order value on a 5% MDR is a net positive. The marginal contribution from the larger basket exceeds the incremental payment cost, especially for higher-margin categories like fashion, electronics, and home goods.

Risk Transfer

When a consumer uses BNPL, the merchant receives the full purchase amount upfront (minus the MDR). The BNPL provider assumes 100% of the credit risk. If the consumer defaults on their installments, the merchant keeps the payment. This is fundamentally different from credit card transactions, where merchants face chargeback risk, friendly fraud, and dispute costs that typically add 0.5-1.5% to the effective cost of card acceptance.

When accounting for chargebacks, fraud costs, and the value of guaranteed payment, the effective premium for BNPL over cards narrows considerably.

Credit Risk and Default Rates

BNPL providers are lenders, and lending means credit losses. The industry has generally maintained lower charge-off rates than traditional credit cards, but the comparison is not straightforward.

Klarna reported an unpaid loan rate of 0.54% in 2025 and credit losses of $137 million in Q1 2025 (up 17% year over year). The Richmond Federal Reserve found that the top five BNPL lenders had charge-off rates of 1.83% in 2020, rising to 2.39% in 2021. These figures look favorable compared to credit cards, but the short duration of BNPL loans means that annualized loss rates can be significantly higher than they appear.

A more concerning data point from CFPB research: 61% of BNPL loans originated in 2021-2022 went to subprime or deep subprime borrowers, and 63% of users maintain multiple simultaneous BNPL loans across different providers. Because BNPL loans are not consistently reported to credit bureaus, a consumer can accumulate obligations across Klarna, Affirm, Afterpay, and PayPal without any single provider seeing the full picture.

The stacking problem: Without centralized reporting, BNPL providers underwrite each loan in isolation. A consumer with four simultaneous pay-in-4 loans across different providers has effectively created an undisclosed credit facility. This is the systemic risk that regulators are most focused on addressing.

How BNPL Integration Works

BNPL providers have built three distinct integration models, each with different technical requirements and merchant tradeoffs. These models interact with existing payment rails and gateway infrastructure.

Online Checkout

The original BNPL model: a button or widget appears at checkout alongside credit card and PayPal options. Implementation ranges from pre-built plugins for Shopify, WooCommerce, and Magento (requiring zero code) to full API and SDK integrations for merchants with custom checkout flows. The consumer selects BNPL, authenticates with the provider, gets an instant credit decision, and completes the purchase. The merchant receives funds within one to two business days.

Virtual Cards for Universal Acceptance

Virtual cards solve the coverage gap. A consumer opens their BNPL app, requests a one-time virtual card number, and adds it to Apple Wallet or Google Wallet. At checkout (online or in-store via NFC tap), the transaction flows through standard card network rails. The merchant sees a normal Visa or Mastercard transaction: no BNPL integration required.

This model is strategically important because it decouples BNPL from merchant agreements. The Affirm Card (2.3 million active cardholders as of mid-2025) and Klarna Card let consumers use installment payments at any merchant that accepts cards, without the merchant opting in or paying BNPL MDR. In these cases, the provider earns standard card interchange instead of the higher MDR.

In-Store and Point of Sale

Afterpay's integration with Block's Square POS terminals enables native BNPL at physical retail. Cash App launched Afterpay integration in March 2025, allowing in-store BNPL via QR code. Klarna offers in-store via its app, generating a barcode that POS terminals scan. The operational challenge is training retail staff and handling returns on partially paid installment plans.

Settlement and Cash Flow

BNPL settlement mechanics differ from traditional card processing in a way that matters for merchant cash flow.

In a standard card transaction, the acquirer settles funds to the merchant in T+1 or T+2 (one to two business days after the transaction). The clearing process involves the card network, issuing bank, and acquiring bank.

In BNPL, the provider pays the merchant the full purchase amount (minus MDR) upfront, regardless of whether the consumer has made any installment payments yet. Settlement timing varies: Afterpay typically settles next business day, Klarna operates weekly or bi-weekly settlement cycles, and Affirm settles within a few business days but may hold rolling reserves of 5-10% for three to six months as a risk buffer.

The key difference is that the merchant gets paid in full before the consumer finishes paying. The BNPL provider finances the gap, which is why these companies require significant capital or credit facilities to operate. Klarna's 2024 balance sheet reflected over $90 billion in gross merchandise volume flowing through this financing cycle.

Regulatory Evolution

BNPL has existed in a regulatory gray zone. Pay-in-4 products were structured to avoid triggering consumer credit regulations, because they charge no interest and impose minimal fees. That exemption is closing across multiple jurisdictions.

United States: Federal Retreat, State Action

The CFPB under the Biden administration issued an interpretive rule in 2024 classifying BNPL as open-end credit subject to Truth in Lending Act protections. The current administration withdrew that rule on May 12, 2025, calling it “procedurally defective” and noting that BNPL products are “generally structured as closed-end loans,” making open-end credit regulations a poor fit. The CFPB confirmed in June 2025 that it will not issue a revised BNPL rule.

States are filling the gap. New York enacted BNPL regulation in its FY2026 budget, requiring providers to obtain a license from the New York Department of Financial Services and capping fees in line with existing state lending limits. Meanwhile, Nevada moved to ease regulatory burdens on BNPL companies. The result is an emerging patchwork of state-level rules.

United Kingdom: FCA Regulation

The UK legislated on July 14, 2025 to bring BNPL under FCA regulatory scope. New rules come into force on July 15, 2026, requiring creditworthiness assessments, precontract disclosure, complaints handling, and support for customers in financial difficulty. A registration window opens May 15 to July 1, 2026, with firms given six months to apply for full authorization.

European Union: Consumer Credit Directive 2

The EU's revised Consumer Credit Directive (CCD2) removes the previous exemption for interest-free credit agreements. Member states must transpose CCD2 by November 2025, with implementing measures applying from November 2026. The directive expands its scope to loans up to €100,000 (previously €200-75,000), meaning even small BNPL transactions fall under regulatory requirements including creditworthiness assessments, APR caps, and enhanced transparency. The Netherlands is going further, explicitly banning BNPL for minors.

The Unit Economics Under Pressure

BNPL unit economics face structural challenges on multiple fronts.

The MDR is under downward pressure as merchants gain leverage. Early BNPL adoption commanded MDRs of 6-8%, but competition among providers has pushed average rates toward 3-5% for large merchants. Meanwhile, credit losses scale with loan volume, and providers must hold capital against their lending book. Regulatory compliance costs are rising across all jurisdictions.

The virtual card and card-linked models earn lower revenue per transaction (standard interchange of 1.5-2.5% instead of BNPL MDR of 4-6%), but they dramatically expand the addressable market. Affirm's revenue as a percentage of GMV was 8.5% in early 2025: a blended rate reflecting high-MDR direct integrations, lower-MDR card transactions, and consumer interest income.

The path to profitability for most BNPL providers runs through diversification: becoming a broader financial platform rather than a single-product lender. Klarna's advertising and shopping app revenue, Affirm's card product, and Afterpay's Cash App integration all reflect this strategic shift.

What BNPL Reveals About Payment Infrastructure

BNPL's rise exposes inefficiencies in existing payment rails. Merchants pay 4-6% for BNPL partly because the underlying cost stack of card processing, settlement cycles, and intermediary fees is already 2.5-3.5%. BNPL providers layer lending economics on top of these existing costs.

Consider what happens when a consumer makes a BNPL purchase through a virtual card: the transaction touches the BNPL provider (lender), the card network (Visa/Mastercard), the issuing bank, the acquiring bank, and the payment processor. Each intermediary takes a cut, and settlement still takes T+1 or T+2. The merchant waits days to access funds, and the BNPL provider finances installment gaps using warehouse credit lines at institutional rates.

This layered cost structure is exactly what stablecoin payment rails and programmable money are designed to compress. If the payment leg of a BNPL transaction could settle instantly and directly, without card network intermediaries, the economics change significantly. For a detailed comparison of how these rails differ, see stablecoin rails vs. traditional payment infrastructure.

BNPL on Stablecoin Rails

The theoretical case for running BNPL on stablecoin infrastructure is straightforward: remove intermediaries from the payment leg so the lending margin can be lower, or the merchant fee can shrink while the provider keeps the same margin.

A stablecoin-based BNPL flow could work as follows: the consumer commits to an installment schedule, the stablecoin payment is sent directly to the merchant's wallet with instant finality (no T+2 settlement delay), and repayment installments are handled programmatically on-chain or on a Layer 2. No card network fees, no acquiring bank markup, no clearing delay.

Spark, a Bitcoin Layer 2 that supports dollar-denominated stablecoin transfers, is one infrastructure layer where this model could operate. Transfers on Spark settle instantly, cost fractions of a cent, and support programmable conditions through its leaf architecture. A BNPL provider building on Spark could settle to merchants in real time, enforce repayment schedules programmatically, and eliminate the card network toll that currently adds 1.5-2.5% to every virtual card BNPL transaction.

The practical challenges are real: merchant adoption of stablecoin payments is still early, consumer UX needs to match the simplicity of tapping a card, and regulatory clarity around stablecoin lending is evolving. But the cost arbitrage is significant enough that the direction of travel is clear. For context on how embedded finance is already blurring the line between fintech and banking, and how programmable money enables automated payment logic, these dynamics are directly applicable to the next generation of installment products.

The cost compression opportunity: Current BNPL transactions on card rails carry 1.5-2.5% in network and processing fees before the lending margin is applied. On stablecoin rails like Spark, the payment processing cost drops to near zero, meaning either merchants pay less or providers keep more: or both.

Looking Ahead

BNPL is not going away. The Richmond Federal Reserve assessed in 2026 that U.S. BNPL transaction volume reached approximately $70 billion, representing about 1.1% of total credit card spending. That share is growing, and regulatory frameworks across the U.S., UK, and EU are formalizing BNPL as a recognized lending product rather than an unregulated payment method.

The competitive dynamics will increasingly favor providers with lower cost structures. As MDRs compress and compliance costs rise, the providers that can reduce their payment processing overhead will maintain margins. This is where infrastructure choices matter: real-time payment rails, stablecoin settlement, and payment orchestration platforms will determine which BNPL models are economically sustainable at scale.

If you are building payment products and want to explore how stablecoin rails can reduce settlement costs, the Spark developer documentation covers the SDK, API integration, and the programmable transfer model. For consumers interested in dollar-denominated Bitcoin wallets, General Bread is a Spark-powered wallet that demonstrates how stablecoin payments work in practice. You can also explore the interactive tools on this site for calculators and comparisons relevant to payment economics.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.