Buy now, pay later can generate more sales without improving a retailer’s economics—and may give merchants an incentive to raise sticker prices for customers who never use installment credit—according to a new model described by Washington University in St. Louis.
The findings challenge a common checkout-finance proposition: that adding another payment option is valuable because it converts abandoned carts and increases basket sizes. The researchers’ model instead focuses on what remains after the BNPL provider’s merchant fee and how that cost changes pricing and inventory decisions.
The study, titled Buy Now, Pay Later: The Hidden Effects of Consumer Liquidity on Retail Prices and Inventories, is forthcoming in Management Science. Washington University said its researchers ran millions of simulated scenarios. That distinction matters: the reported findings describe modelled incentives and outcomes, not a measured estimate of how much BNPL has already changed prices across the retail market.
More transactions do not necessarily mean more profit
BNPL providers generally pay the retailer upfront, collect installments from the customer and retain a portion of the transaction. Retailers may accept that cost because installment financing can help customers complete purchases they otherwise could not make immediately.
In the model, however, higher unit sales did not ensure higher profits. The university said a retailer’s effective price—the amount retained after financing costs—could decline even when the number of products sold increased. It also reported that the researchers found no simulated scenario in which BNPL turned an unprofitable product into a profitable one.
This is the key control issue for merchants and payment providers: authorization and conversion metrics can look stronger while contribution economics deteriorate. A merchant evaluating BNPL needs to separate incremental purchases from transactions that would have occurred through a lower-cost payment method, then account for provider fees, returns, fraud, servicing effects and any change in average order value. Gross sales alone cannot establish whether the payment option creates value.
The cost may reach customers who pay another way
The research also models a spillover beyond BNPL users. If a retailer uses one sticker price for customers paying through different methods, it may raise that price to offset the financing charge. Customers paying upfront can then bear part of the cost created by installment users.
Washington University summarized the study’s conclusion as wider BNPL adoption leading to higher prices for all consumers. That is a model result, not proof that every merchant responds this way. The university’s public report does not provide an observed average price increase, identify a uniform fee across providers, or show that price effects are the same across product categories.
PaymentsJournal, which reported on the research on August 26, quoted a Javelin Strategy & Research analyst saying BNPL transaction costs can be in the “5% plus” range and are typically higher than traditional credit-card costs. That comment illustrates why merchants may seek an offset, but it should not be read as a disclosed rate for every BNPL contract. Pricing can depend on the provider, merchant, product and commercial agreement.
Inventory is part of the payments decision
The model links checkout financing to stock levels as well as prices. Washington University said accepting BNPL could lead a retailer to hold less inventory because the cost of losing a potential sale changes once financing fees reduce the value retained from a completed transaction.
That finding broadens the operational question. BNPL is not only a tender button or credit product; under the study’s assumptions, its economics can affect merchandising decisions and product availability. Payments teams therefore need input from finance, pricing and inventory functions rather than evaluating the option only through checkout conversion.
What the evidence does—and does not—establish
The university’s report says the study is forthcoming and describes the modelling results, but it does not link to the full paper. The public material does not disclose enough methodology to independently assess all assumptions, estimate effect sizes or determine how closely the simulations match particular merchant sectors.
The report also cites broader consumer-risk indicators, including use of installment plans for necessities and missed payments. Those indicators add context about why the product’s cost distribution matters, but they do not demonstrate that the modelled retail-price effect caused household distress.
The defensible conclusion is narrower: BNPL fees can alter merchant incentives in ways that are hidden by sales growth, and the model predicts that some of the cost can be passed to non-BNPL shoppers through common retail prices. Establishing the real-world magnitude will require transaction-level or merchant-level evidence comparing prices, margins and inventory before and after adoption while accounting for differences among providers and product categories.
For payment and merchant-risk teams, the practical response is not to assume BNPL is either beneficial or harmful. It is to demand cohort-level economics, document who bears the financing cost, monitor substitution from cheaper payment methods and test whether conversion gains survive after all fees and operational effects are included.