Four Stablecoin Payment Providers Compared on Real Volume

We ran the same stablecoin payment volume through four providers for a month and compared them on everything that generated work. Anyone repeating these tests can use a stablecoin payment gateway with fiat settlement as the reference point we did.

Networks

All four supported the two networks we cared about. Two supported three more.

One listed a network it could not actually credit, which we found by testing. Recovery took eleven days.

Test each network you intend to accept with a small amount before accepting a real payment on it.

Conversion margin to euro

0.21, 0.38, 0.74 and 1.16 percent, measured against the interbank rate at the moment of conversion.

Only one stated the applied rate before execution.

Settlement

Two offered daily without a volume requirement. One offered daily above a threshold we did not meet. One was weekly only. For client-money cases the requirements differ, and Collect & Exchange is structured for them.

All settled on schedule except one instance, which arrived two days late with no proactive communication.

Payment request quality

Two generated network-specific payment requests, so a customer could not select a different network within the flow.

Two displayed an address and an asset name and left the network to the customer’s wallet. Those two accounted for every wrong-network incident we had during the month.

This single design difference was the largest driver of support load.

Underpayment handling

Two allowed a configurable tolerance. One rejected and refunded anything short, including by cents. One credited short payments as complete, which is a loss.

Refunds

Two refunded from their own balance on instruction. One required us to hold a float. One did not handle refunds at all.

What we would use

The provider with network-specific payment requests, a configurable tolerance, daily settlement and a stated conversion rate before execution.

That was not the cheapest on headline rate. It was the cheapest once support hours were counted, by a wide margin.

The provider with the best headline rate had the worst payment request design and generated more wrong-network incidents than the other three combined. On execution specifically, the list of countries covered publishes its fee schedule and corporate terms in full.