America’s smaller lenders want programmable money. They may not enjoy what it does to their deposits.
For three years the argument about digital money has been staged as a duel. On one side sit the banks, defending the deposit. On the other sit Circle, Tether and their imitators, building a money layer outside the regulated perimeter. On August 25th a third answer appeared. Thirty-nine American state banking associations announced the BankChain Alliance, a shared blockchain network intended for launch in 2027 and ultimately to be owned by the banks that use it. It is designed to carry tokenised deposits, stablecoins, programmable payments and automated settlement. A technology partner has not yet been chosen.
The logic is mutualisation. A community lender in New Hampshire cannot build what JPMorgan Chase has built, and cannot afford to concede programmable money to those who can. Sharing the plumbing lets thousands of small institutions keep the customer, the deposit and the regulated relationship, while renting the technology from themselves. It is the correspondent-banking bargain rewritten for a ledger.
The rest of the week supplied evidence that the plumbing is the hard part. Swift completed its first live tokenised-deposit transaction, between HSBC and Standard Chartered, connecting two bank-issued deposit tokens without either leaving its home platform. Seventeen banks across six continents are to join the pilot, an impressive number until measured against Swift’s roughly 11,500 members. In Switzerland, six banks including UBS, PostFinance and Zürcher Kantonalbank began testing a franc stablecoin in a live sandbox, where the interesting question is not issuance but whether one bank’s token redeems at par against another’s, and how any of it reconciles with SIC, the country’s real-time settlement system. Tokenised money is proliferating faster than the means of connecting it.
Then the awkward part, published the same day as BankChain’s announcement. Economists at the Federal Reserve Bank of Dallas asked what happens to a bank when its deposits become programmable. Their answer is that such deposits become mobile. Money that can move instantly, and increasingly can be moved by software hunting yield without a human deciding to, is money that stops behaving like a relationship and starts behaving like wholesale funding. The authors estimate, illustratively, that a 10% reduction in the average life of a deposit would remove roughly $580bn of ten-year-equivalent duration capacity from the American banking system. They are careful to call this exploratory rather than predictive. It is nonetheless the first serious official attempt to price the second-order cost of the thing every bank now says it wants.
That is the trade BankChain’s members are making, and it is worth naming plainly. Deposits are the cheapest funding in finance because they are lazy. Programmability is an argument for making them less so. A technology adopted to defend the deposit franchise may, at scale, dissolve the property that makes the franchise valuable.
The commercial models on offer do not yet resolve this. Revolut’s new euro stablecoin, EURR, issued by Bridge and rolling out first in Denmark, Poland and Portugal, follows the company’s familiar method: price the product at nothing, acquire the customer, monetise elsewhere. Standard Chartered’s distribution of the Hong Kong dollar token HKDAP is the opposite, a business-to-business efficiency play sold into existing institutional workflows. Neither is chasing the token itself, which is sensible, since stablecoin payments generated only about 2% of crypto-sector revenue last year.
Banks have concluded, correctly, that they cannot win this from outside. The harder question for the thirty-nine associations is not whether they can build the rails. It is whether, once built, the money stays where they put it.
