Buy Now, Pay Later: The Economics Behind the Free Installments
Four payments, no interest, approved in seconds. BNPL feels like free money because someone else is paying, and the someone is the merchant, betting you will buy more than you meant to.
The Button That Ate Checkout
Buy now, pay later splits a purchase into installments, classically four payments over six weeks, zero interest, approved in a soft credit check that takes seconds at checkout. Klarna, Afterpay, Affirm, and PayPal built the category from a Scandinavian curiosity into a standard checkout option on hundreds of billions of dollars of annual commerce, with usage heaviest among the young shoppers this site\'s readers know personally. The consumer math looks impossible, credit with no interest, so the analyst\'s first question applies, who pays. The answer is the merchant, on purpose, and the reasons illuminate both retail psychology and the credit cycle risk hiding in the category.
The Merchant Pays for Your Restraint to Fail
BNPL providers charge merchants roughly 2 to 8 percent of each transaction, several times the card interchange this site covers separately, and merchants pay it happily for measurable returns, conversion and cart size. Installment framing dissolves sticker shock, a 200 dollar purchase becomes four 50s, and providers advertise meaningful lifts in checkout completion and average order value. Understand what is actually being sold, the merchant is buying reduced purchase friction, which is a polite name for weakened restraint, and the provider is monetizing its position as the friction remover. Around that core sit the secondary engines, late fees when installments miss, longer interest bearing plans for bigger tickets, where Affirm concentrates, and app driven shopping referrals, the providers evolving into marketplaces that sell merchants placement, not just payments.
BNPL\'s pitch to merchants is the mirror image of its pitch to users. To you, easier payments. To the store, customers who complete more purchases and choose bigger ones. Both pitches are true, and their combination is the business model.
The Risk the Model Carries
The provider fronts the merchant full payment and collects from the consumer over weeks, so it wears the credit risk, underwritten in milliseconds from thin data. Loss rates in the low single digits of volume are the norm in good times, manageable against the merchant fee, and the six week loan term lets providers re underwrite the entire book many times a year, a genuine structural advantage over card lenders. The systemic worry is what regulators and rating agencies call phantom debt, BNPL balances have historically been invisible to credit bureaus, so a consumer can stack installments across four providers, each seeing only its own slice, and the true leverage of young borrowers appears nowhere until it fails all at once. Survey data consistently shows meaningful minorities of users paying late, and usage skews toward exactly the credit thin borrowers least able to absorb a shock. The category has never been through a proper recession at scale, which is the sentence that should appear in every BNPL equity model.
The Regulators Arrive, Unevenly
The rulebook is being written in real time and diverges by geography. In the US, the CFPB\'s move to treat BNPL like credit cards was withdrawn in 2025 with the change of administration, leaving federal oversight light, so the action shifted to states, New York has proposed licensing and rules for providers, and to Congress, where lawmakers are pressing the credit bureaus on why BNPL data still barely reaches credit files. The bureaus and larger providers are slowly wiring reporting anyway, since visibility serves the incumbents against reckless entrants. Britain ends the debate in July 2026, bringing BNPL fully under its financial regulator with affordability checks and complaint rights. The direction everywhere is the same, the product is credit, and it is being regulated into admitting it, which raises costs and, again, favors scale.
The Bottom Line
BNPL is merchant funded credit, stores pay several percent for higher conversion, providers wear a fast turning book of small loans, and the zero interest headline is subsidized by the psychology it enables. The model\'s genuine innovations, instant underwriting, six week duration, checkout distribution, are real, and so are its untested edges, invisible stacked leverage, recession naive loss models, and a regulatory perimeter closing jurisdiction by jurisdiction. For a student, the analytical takeaways are the durable ones, always locate the payer behind free, and never trust a lending model that has only seen good weather.