Two different approaches to airline settlement are taking shape, and they place financial responsibility in different parts of the distribution chain. Airlines spent about $5 per passenger on payments in 2025, against IATA’s forecast net profit of $4.50 per passenger for 2026, so the distinction matters.
When Mastercard’s Chiara Quaia spoke to TDN in May, her argument about NDC settlement was short. A single OTA booking can now contain separate NDC orders from several airlines, each with its own payment, servicing and settlement requirements, and traditional settlement arrangements fragment under that load. Her proposed fix is for the OTA to act as merchant of record: it collects one payment from the traveller, then pays each airline in the background, with virtual cards as the enabling tool.
The diagnosis is credible. Airline settlement was built around tickets, and a basket of orders from several carriers does not sit neatly inside a ticket-shaped process. But a diagnosis is not yet a remedy. The approaches now on the table are not strictly rival settlement systems, because they work on different parts of the payment chain. What separates them is where each places financial control, cost and risk.
Different parts of the payment chain
IATA is building an order-based clearing framework. It describes Settlement with Orders as a framework for settling orders between partners, with automated exchange of settlement data and fully automated reconciliation. Its Clearance Manager is designed to receive and validate claims, combine them on a set schedule, net them out and tell each party its net position, after which the payer remits funds to IATA and IATA passes them to the payee. IATA says the service can settle orders whether or not they sit within BSP governance. The underlying Clearance and Settlement Standard has been approved, but IATA describes Clearance Manager as a platform still in development and is recruiting launch partners. The next test is whether the framework can move from an approved standard to an operational service with sufficient industry participation.
Mastercard’s proposal sits at a different point. Making the OTA merchant of record settles who collects and processes the traveller’s payment, and virtual cards are the tool for paying each airline. Amadeus offers a useful comparison on the same question. It has said an airline can choose to be merchant of record, alongside the arrangement common today in which the agency takes that responsibility, and it offers travel agencies a way to pay airlines with airline co-branded Mastercard single-use virtual cards, for airlines that prefer not to carry the role. These are choices about payment architecture and commercial responsibility, not a separate settlement framework.
Each aims to tie money to an order, and an OTA could in principle act as merchant of record while taking part in an order-based clearing arrangement, depending on the commercial and technical terms agreed between the parties. What differs is who handles the funds, who earns on them and who answers when something goes wrong.
The number that frames the choice
On 27 October IATA published the figure against which any settlement fix should be tested. Airlines incurred $24.5 billion in payment costs to process $1.05 trillion in sales in 2025. Total payment costs rose 10.4% on 2023, while sales grew 7.7% over the same period. The average cost came to $5 per passenger, against IATA’s forecast net profit of $4.50 per passenger for airlines in 2026. The two measures cover different years and are not identical accounting measures, and the $24.5 billion covers airline payment costs generally, not NDC settlement or virtual cards specifically.
Most of that is fees. IATA put payment fees at $20.7 billion, or 84.5% of the total, with a further $2.9 billion in operating costs such as systems, delayed cash flow, staff and compliance, and $0.9 billion lost to fraud or uncollected sales.
By that measure the question for any new settlement model is plain: does it reduce the $5, or move it? Cards carried the highest average fees in IATA’s analysis, at 2.53% against an estimated 1.97% across all payment methods. A virtual card is still a card. Interchange applies, and one virtual card provider notes that issuers commonly rebate part of it to the OTA, while BSP cash carries no card processing fee. In some merchant of record arrangements, the party receiving the rebate and the party ultimately bearing the acceptance cost are not the same.
That does not make the model wrong, and its backers have a case. Virtual cards allow a unique number per transaction, which simplifies reconciliation, and analysis written on behalf of Amadeus argues they carry lower fraud exposure than a lodged or consumer card, although they do not remove fraud, chargebacks or payment costs altogether. The same company claims its virtual card product can reduce overall payment costs by up to 70%. That is a vendor figure, and the result depends heavily on what it is measured against, which is why airlines will want it tested against their own BSP and card mix before relying on it.
Questions the interview left open
The first is working capital. Adyen, which processes payments for OTAs, describes the merchant of record model as creating a structural timing gap between when customer funds arrive and when suppliers must be paid. A large OTA can carry that. In markets where many intermediaries are smaller and currency exposure is a daily problem, whoever funds the gap takes on a financing business they may not have set out to run.
The second is liability. Hahn Air, which acts as merchant of record for agents, has itself described exposure to chargebacks and disputes. Take a hypothetical itinerary across three airlines that a traveller later disputes. The OTA answers to the card scheme, yet the airline that did not operate its segment has already been paid. This is an illustration, not a reported case, but the Quaia interview does not say how liability would be allocated, and in a multi-airline basket the answer decides how attractive the role is.
The third is fit with local payment habits. IATA’s 2025 Global Passenger Survey found that 29% of passengers prefer a digital wallet and 8% a local instant payment solution. Those are global figures, not African or MENA ones, and they should be read that way. Mastercard also points to wallets, tokenised credentials and account-to-account payments in its own inclusion work. IATA Pay, which supports local instant payments, has launched in 34 countries with 38 payment schemes, according to IATA’s 27 October announcement. The test for any card-centred settlement design is whether it can accommodate markets where account-to-account payments, wallets and local payment systems matter alongside cards.
Where this leaves airlines
Settlement for orders is being shaped by IATA, the card networks, intermediaries and airlines, and the arrangements they settle on will decide who handles the money and who answers for it. The technology matters less than the financial architecture around it, which sets who earns on the flow and who carries the risk. An airline that treats the decision as plumbing will learn the price only after the plumbing is built. Carriers in emerging markets, where currency conversion is already a cost, have little room to learn it late.
