Payments in Stablecoins: Companies Must Also Measure Failures

For a company, reducing the cost of an international payment is only part of the problem. On October 16, 2026, Cointelegraph relayed data attributed to Borderless.xyz on settlements funded in stablecoins. They describe short delays on several local networks, but also strong differences in success rates depending on the destination. The interest in the topic goes beyond the headline price: an operation must be executed, received, and correctly reconciled with the corresponding invoice.

Speed does not describe every transaction

The report indicates median delays of two minutes in Brazil and three minutes in Mexico. It also reports a completion rate of 98.6 % in Mexico, versus 73 % in Costa Rica. These indicators concern the circuits observed by the provider. They are neither a universal ranking of countries nor a measurement of all banks. A fast median does not mean that every transfer completes within that time.

This distinction changes the reading of the service. Speed informs about the observed delays; completion indicates the proportion of operations that succeeded within the studied scope. For an urgent invoice, the existence of a digital balance sent does not necessarily resolve the question of its availability in the local account. The payment may travel across a blockchain before depending on a service provider and a destination banking system.

The local network remains part of the product

In its public report on the second quarter of 2026, published on July 13, Borderless explicitly distinguishes circuits that are accessible from those that actually carried payments. The document also describes corridors with multiple providers and others that rely on a single provider. These observations do not prove the results of the following quarter, but they illuminate a structural limitation: the announced coverage of a network and its operational capacity are not identical concepts.

Nexorena’s analysis is that a second provider can offer a continuity option, without eliminating all risks. Procedures still need to be able to identify a blocked payment, avoid a double settlement, and retrieve the reference useful for accounting. The time mobilized by teams to resolve an incident also forms part of the payment experience. A commission saving does not alone describe this operational burden.

Measuring what was actually delivered

On September 18, 2026, Borderless presented an update to its comparison methodology. The company explains that it groups payments by corridor, direction, and size, while separating fixed fees from proportional components. It says it prioritizes completed transactions rather than only the announced prices, with a flag when observations are insufficient. This approach acknowledges that a quote and a payment actually executed are not necessarily equivalent.

To monitor a payment solution, the indicators must therefore remain complementary: final cost, delay, completion rate, and quality of exception handling. The figures relayed in October reinforce the interest of stablecoins as a possible component of international settlements; they do not demonstrate uniform reliability. The real change for companies will be less visible in a spectacular single transaction than in repeated, documented, and reconciled operations without permanent intervention from finance teams.

A regular measurement should distinguish finalized operations, those pending, and those refunded, together with a resolution time. This reading helps avoid confusing an available network with a fully executed service.