The Quiet Economics of Interchange

Interchange is the fee an acquiring bank pays an issuing bank each time a card transaction happens. It is a small number, usually well under two percent, and it is the most argued-about figure in the industry.

The arguments are usually about whether it is too high. That is the least interesting question about it.

What interchange is actually for

A card network is a two-sided market. It needs cardholders to be useful to merchants, and merchants to be useful to cardholders, and neither side joins first without a reason.

Interchange is the mechanism that moves value from the merchant side to the cardholder side. It funds the rewards, the fraud guarantees, the interest-free period, and everything else that makes carrying a card attractive. Without a transfer of this kind, the issuing side of the market has thin economics and the cards get worse.

That is the theory, and the theory is sound. Two-sided markets genuinely do need a mechanism to balance participation, and a price that runs in one direction is a reasonable way to build one.

Where it gets contentious

The difficulty is that interchange is set by the network, not discovered by negotiation between the parties who pay it and receive it. Merchants pay a price they did not agree to, determined by a body they are not part of, funding benefits that accrue to someone else’s customers.

Whether this is a problem depends on whether you think the network is setting the price to optimise the whole system or to favour issuers. Regulators in several jurisdictions have concluded the latter and capped it. The results have been mixed in an instructive way: merchant costs fell, and cardholder rewards fell with them, roughly as the theory predicted.

That outcome is worth sitting with. It suggests interchange was doing what it claimed to do, and that lowering it moved value between participants rather than creating or destroying it.

The second-order effects

The more interesting consequences of interchange are not about its level but about what it incentivises.

Because interchange varies by card type, issuers have a reason to push customers toward the products that earn more. Because it varies by transaction type, merchants have a reason to route transactions in particular ways. Because it varies by geography, a great deal of effort goes into where a transaction is deemed to occur.

None of this is fraud. It is rational response to a price structure. But it means a meaningful amount of the industry’s engineering effort goes into optimising against a fee schedule rather than into anything a customer would notice.

That is the part I find worth thinking about. A pricing mechanism designed to balance a two-sided market has, over decades, produced an entire secondary industry devoted to arbitraging its details. Whether that is an acceptable cost of an otherwise functional system is a genuinely open question, and it is a better question than whether the headline number is too high.