In the spring of 1915, the brand new Federal Reserve had a logistics problem.

Twelve regional Reserve Banks were now clearing payments across district lines, and every imbalance between them had to be settled in gold. Gold, inconveniently, is heavy.

So the Federal Reserve Board created the Gold Settlement Fund. Each Reserve Bank kept at least $1 million in gold with the Treasury in Washington. The banks telegraphed what they owed one another, and the Board settled up by changing entries in a ledger. The gold mostly stayed put. Only the bookkeeping moved.

I love this story because it captures something the early Fed had to learn the hard way: Money can move in a decentralized way, but settling it requires a center.

More than a century later, stablecoins are relearning that lesson. They move value across borders in seconds, without the traditional chain of correspondent banks. But a dollar stablecoin sent to Mexico is not pesos in the recipient's bank account.That's the last mile problem for stablecoins. The transfer settles on-chain almost instantly, but conversion and payout still run through infrastructure that lives off the blockchain. Until recipients get the currency they actually spend, speed alone won't drive adoption.

What the Fed's Economists FoundCopy anchor linkCopied

In March, three Federal Reserve economists published a FEDS Note that walks through this problem with a simple example: A customer of a small U.S. bank sends $1 to a customer of a small Mexican bank.

Today, that dollar passes through four banks. The small U.S. bank sends it to a large U.S. correspondent, which sends it to a large Mexican bank with a U.S. branch, which sends pesos to the small Mexican bank. Each large bank can charge a fee (on top of wire fees), and somewhere along the way the USD becomes pesos with little transparency on the FX rate.

With a stablecoin, the two small banks transact directly. The Mexican bank credits its customer in pesos. Then it sells the stablecoin to a large Mexican bank, because it doesn't want to carry the currency risk.

The stablecoin removed the correspondent. It didn't remove the large bank. The pesos still had to come from somewhere. That somewhere is why Western Union and MoneyGram spent decades building pickup networks.

The authors explain why the large bank sticks around. Cross-border payments carry high fixed costs: A foreign presence to exchange currency, plus the staff and systems to run anti-money-laundering checks. Large banks spread those costs across millions of transactions. Small banks can't, so they rent a large bank's presence. That's why more than 60% of wholesale payments route through at least one intermediary, even as the number of active correspondent banks has fallen about 30% in a decade.

Their conclusion: Stablecoins eliminate some costs, not all. On-chain transfers are cheap, but getting on and off the chain can add cost, and banks will keep leaning on large intermediaries for currency inventories and multi-jurisdiction compliance.

The last mile is the price of a local presence, and volume decides who can offer it cheaply.

The Three Costs of the Last MileCopy anchor linkCopied

Scale matters in three places:

1. FX. Whoever turns a dollar stablecoin into pesos takes on currency risk and charges for it. FX is a volume business. Large players aggregate flows, hold currency inventories, and manage exchange-rate risk more efficiently, so they get better pricing.
2. Compliance. Moving money on a blockchain doesn't exempt anyone from verifying customers and screening payments. Every hop in a payment chain can mean another redundant check, and the cost per check falls as volume rises.
3. Funding. To pay recipients abroad, businesses often prefund accounts with local partners before a payment is even requested. Across a dozen currencies, that ties up serious working capital. The alternative is postfunding, where the payout partner pays first and collects later. That requires a partner with the balance sheet to carry the risk in between.

Who Runs the Last Mile?Copy anchor linkCopied

That's the backdrop for one of the more consequential stablecoin deals this year.

In August, Mastercard completed its acquisition of BVNK, a stablecoin infrastructure company. Last month, the two launched Mastercard Global Money Platform, which pairs BVNK's stablecoin stack with Mastercard’s global payments infrastructure, including a component — Mastercard Move — that many folks (including me, until recently) are unfamiliar with.

Mastercard Move is a payout network that runs alongside the card rails and reaches more than 17 billion bank account, card, wallet, and cash pickup endpoints. Combined with BVNK, it now enables banks and fintech companies to embed an account that holds fiat and stablecoin balances side by side, converts between them, and pays out from stablecoin balances in local currency, directly into a recipient's bank account.

And they’ve designed a second way to bridge the last mile: a card on Mastercard’s rails that lets people spend stablecoins anywhere Mastercard is accepted. That gives the last mile two routes; Move delivers fiat, and the card lets customers hold the USD equivalent while keeping full local currency utility (wherever cards are accepted).

Here's how I think this new platform will map to the three costs we covered above:

1. FX. Mastercard already converts currencies at enormous volume every day for card transactions that cross borders. Platform customers should be able to convert at rates shaped by that volume, not their own.
2. Compliance. A payment that moves from a stablecoin balance to a local bank account through one network passes through fewer hands, which should mean fewer redundant checks. Move also bundles fraud prevention, validation, and risk management, so fintech companies won’t need to rebuild that stack corridor by corridor.
3. Funding. One account holding fiat and stablecoins is one pool of liquidity, not a pile of prefunded balances scattered across markets. And Mastercard should be able to bring the balance sheet and bank relationships to support postfunding, and enough existing flow to net payments against each other before money moves.

That's the logic of the acquisition and this new product announcement. BVNK built the part of the trip stablecoins are good at. Mastercard owns the part they aren't.

A Relay RaceCopy anchor linkCopied

Cross-border payments are starting to look like a relay race. Stablecoins carry value across the border. Established networks run the last and most expensive leg, into the recipient's account, in the currency they need. Nobody watching a relay asks which runner mattered most. The handoff is the race.

The Fed figured this out in 1915. Twelve banks could move money independently, but settling it took a shared ledger in Washington. Stablecoins are on the same path. And this time, the center already exists.

Editor’s Note — This article was made in collaboration with BVNK. As with all content in Fintech Takes, this article was written, edited, and published by me, Alex Johnson. I hope you enjoyed it!