Goldman Sachs Was a Holdout in August. It Just Turned Up as a Founding Member of a 21-Bank Stablecoin Company.
Subscribe Free. The weekly briefing is free, always.
Twenty-one financial institutions are forming a company to issue stablecoins pegged to G7 currencies. A dollar token first, a euro token after, targeted at the first half of 2027. Bank of America, Citigroup, Goldman Sachs and Mitsubishi UFJ are named. The other seventeen are not, at least not yet.
Five weeks ago we told you Goldman was a holdout. It is now a founder.
That is the part worth your time. Not the announcement. The reversal.
What We Told You in June, and Why It Was Only Half Right
On June 7, writing about JPMorgan, Bank of America, Citi and Wells Fargo building a shared network through The Clearing House, we made a structural argument. Banks were choosing tokenized deposits instead of stablecoins, and the reason was capital efficiency. A tokenized deposit stays on the issuing bank’s balance sheet and can be lent against under fractional reserve rules. A stablecoin’s reserves sit in cash and short-term Treasuries and cannot be lent at all. Banks keep the spread. Regulators keep money creation inside the perimeter.
On August 2, those same four consolidated their separate blockchain projects onto shared rails. We told you then to watch two names: Goldman Sachs and State Street, the large institutions that had not signed on.
Then on August 15 we moved. Writing about Wells Fargo, we argued the banks meant to own both forms of programmable money and would use whichever one fit the flow. That was the right call. This week retires the June framing for good.
Deposits versus stablecoins was never the question. It was always deposits and stablecoins, divided by how far the dollar has to travel.
Why a Bank Needs Two Kinds of Digital Dollar
Here is the thing a tokenized deposit cannot do easily. It cannot simply leave the banking perimeter.
A Citi deposit token is a claim on Citi. It settles beautifully against a Bank of America deposit token, so long as both sit inside a shared utility with a common rulebook. A Federal Reserve account sits at the end of that chain. That is exactly what The Clearing House is building, and it solves the domestic interbank problem. Interoperability beyond it is a design choice rather than an impossibility, but every hop needs a legal and settlement arrangement negotiated in advance. Move outside that perimeter, to a counterparty in Singapore with no correspondent relationship, and the claim has nowhere to land.
A stablecoin behaves more like a bearer instrument. It travels to whoever holds it, with no relationship negotiated first. That is its advantage, and it is the reason Tether found its market in places where dollar bank access is thin rather than in Manhattan.
So the banks are running two tracks. Tokenized deposits for domestic institutional settlement, where the counterparties are each other. A stablecoin for cross-border and cross-currency flow, where they are not. The G7 scope in this announcement is the tell. A euro token is not a domestic American product. Neither is a joint venture with Japan’s largest bank in it.
Which brings us back to Goldman. Goldman does not have a $2 trillion retail deposit base to protect. For Bank of America and Citi, a bank-issued stablecoin cannibalizes their own funding, which is why they spent a year building the version that does not. Goldman has almost nothing to cannibalize and a great deal of cross-border institutional flow to win. A holdout on the deposit project becomes a founder on the stablecoin one. I read the holdout as reluctance. It was specialization.
So who loses? Not the technology, and not stablecoins as a form. The loser is the independent issuer’s privileged position between tokenized markets and the banking system, where it currently charges for the passage. Run the chain the banks are assembling: deposit franchise, tokenized deposit, bank-sponsored stablecoin, tokenized securities settlement, cross-border flow. They are not trying to kill the instrument, they are trying to own both ends of it.
What This Means for Your Dollars
Three things change, and one does not.
Issuer economics get harder. Circle earns the overwhelming majority of its revenue from interest on reserves, the same spread banks have collected for decades, and it currently keeps nearly all of it. Twenty-one banks arriving with their own funding costs and their own distribution put that arrangement under pressure. It could give way through lower fees, reserve-sharing, institutional pricing, or simple loss of share. Where regulators allow it, some may reach holders. What looks hard to sustain is an issuer keeping essentially all of the reserve income while bank-sponsored alternatives compete for the same balances.
Tokenized securities get a native settlement asset. Today they mostly settle against USDC, which means a bank-issued bond token settles against a non-bank liability. A bank consortium stablecoin removes that mismatch, which could help make tokenized collateral usable in size. The settlement asset is one input among several, not the whole problem.
Access tightens rather than widens. This will be permissioned. KYC’d corporates, qualified purchasers, institutional counterparties. If you are waiting for a permissionless alternative to Tether with a Wall Street balance sheet behind it, this is not that, and it was never going to be.
What does not change is your position this year. Nothing here ships before 2027. Announcements are not products, and the banks have announced a great deal since 2019 that never reached a customer.
What to Watch Next
Two signals, and they are more specific than usual.
First, the legal structure. Shared equity stakes and co-equal board seats mean a company. Nonbinding memoranda and a working group mean another Project Guardian, another Canton, another decade of white papers. The difference will be visible in the incorporation documents, not the press release.
Second, watch whether this venture and The Clearing House project ever mention each other. If the overlapping banks describe them publicly as one strategy with two instruments, the two-track reading is right and both survive. If each pretends the other does not exist, it is a turf fight between the transaction bank and the markets side, and one of them quietly dies in 2027. I have watched that specific fight happen inside three institutions. It usually ends with the side that controls the client relationship winning.
And a smaller thing: the full list of twenty-one has not been published. Watch whether JPMorgan and Wells Fargo, the two loudest voices on tokenized deposits, are on it. If they are, the strategy is coordinated. If they are not, the industry has just split into two camps, and the camp with Goldman in it is the one selling to everybody else.
Presented by The Bridge. Weekly institutional research on blockchain, agentics, and tokenization, written for hedge funds, asset managers, and corporates. Because you read OMNM, the retail edition is yours for $349 (normally $399): thebridgenewsletter.com/signup?ref=omnm. OMNM co-host Douglas Borthwick co-founded The Bridge with Steve Kraus; we may earn a commission.
From the Author
Old Men, New Money co-host Douglas Borthwick has written on this in depth.
Go Deeper — Related Education Modules
Related Guides
Never Miss an Issue
The weekly briefing is free, always.
Join 38,000+ professionals getting weekly analysis on the convergence of traditional finance and digital assets — delivered straight to your inbox.