Stablecoins are still being sold as the obvious answer for digital payments. The Bank for International Settlements is not buying it, at least not yet, and says they still fail the basic test of money at scale.
- BIS verdict: stablecoins still fall short as payment money
- Preferred path: tokenized bank deposits inside regulated rails
- Main tension: open crypto settlement vs controlled bank settlement
- Regulatory reality: major jurisdictions are not aligned
At the Federal Reserve’s Jackson Hole symposium on Aug. 28, Pablo Hernández de Cos, the BIS General Manager, argued that stablecoins do not yet credibly work as a payment method at scale.
That is a familiar warning from the BIS, the global central bank club that tends to look at monetary plumbing the way surgeons look at a scalpel: useful, but not something to swing around because it looks new. Still, the warning matters. Stablecoins are no longer a crypto side quest. They are built into trading, transfers, treasury flows, and the bigger “digital dollar” pitch that says private tokens will somehow become everyday money.
De Cos pushed back on that idea. His preferred alternative is tokenized bank deposits, which stay inside the regulated banking system and settle in central bank money.
“Tokenised deposits offer a more direct path to harness tokenisation while preserving the monetary system’s foundations, ” de Cos said.
That is the core of the BIS argument. It is not anti-digital. It is anti-chaos. The institution wants programmable payments without building a parallel monetary layer that tries to act like cash while depending on private reserves, private redemption promises, and blockchain bridges that can be as elegant as a shopping cart with one broken wheel.
Why the BIS says stablecoins fail the money test
De Cos said stablecoins struggle with three features that matter for money: singleness, interoperability, and financial integrity.
Singleness means a unit of money in the same currency should be worth the same as every other unit. If something is meant to be one dollar, it should reliably redeem for one dollar. Stablecoins can trade above or below that level during stress, which means the “one token equals one dollar” promise can wobble when confidence does.
Interoperability means money should move smoothly across systems. Stablecoins can work well inside a single venue or blockchain, but moving them between chains often requires bridges, centralized intermediaries, or wrapped assets. Those solutions work, but they also add risk, complexity, and more ways for things to break.
Financial integrity is the AML and counterterrorist financing piece. Public blockchains are open by design, which is part of the appeal. It is also why regulators worry about enforcement. Open rails are great for access and censorship resistance; they are a headache if your main obsession is tracking bad actors and blocking illicit flows.
De Cos did not call for a blanket ban. He suggested stablecoins and tokenized deposits could coexist if regulators clearly define their roles and add safeguards. Under that model, tokenized deposits would handle most daily and wholesale payments, while stablecoins would be left to narrower uses, including decentralized lending.
That is a lot less glamorous than the “stablecoins will replace banks” crowd likes to pretend, but it is closer to how institutions actually think. Most serious money systems are not built on slogans. Annoying, yes. True, also yes.
What stablecoins get right, and what still goes wrong
Stablecoins usually aim to stay pegged to a fiat currency by holding reserves such as cash, bank deposits, repurchase agreements, and short-term government debt. The U.S. Treasury said the GENIUS Act requires permitted payment stablecoins to maintain one-for-one reserves using cash, deposits, repos, or Treasury securities with remaining maturities of 93 days or less.
That reserve model is supposed to support redemption at par. In calm markets, it often works well enough. In stressed markets, the cracks show.
De Cos warned that stablecoins can trade above or below one dollar under pressure, which means the secondary market price can drift away from the claimed unit of account. If redemptions spike, issuers may need to sell Treasury bills or pull bank deposits to meet demand. Large-scale selling of short-term government paper could strain funding markets. That does not mean apocalypse. It does mean the plumbing matters when the market gets cranky.
There is also the banking-system angle. If households or firms move deposits from banks into stablecoins, banks may have to pay more to keep funding. That could raise funding costs and, in turn, affect credit availability. The size of that effect depends on scale, user behavior, and whether stablecoins actually substitute for deposits or mostly replace cash-like balances. So yes, the risk is real, but it is not a law of physics.
De Cos also pointed to a possible upside. If stablecoin demand comes from outside the United States, it could increase demand for U.S. government debt and lower borrowing costs. That is the bullish Treasury argument in plain English: more buyers for short-term U.S. paper is usually good for the issuer, even if those buyers come via crypto infrastructure.
He also said BIS modeling suggests the overall economic effect would likely be modest, depending on reserve composition, government debt levels, and whether demand is domestic or foreign. In other words, stablecoins are not some magical macro lever. They are a payments tool with second-order effects, not a sovereign debt cheat code.
Tokenized deposits: less hype, more control
Tokenized deposits are digital representations of commercial bank deposits built on programmable infrastructure. The key difference from stablecoins is simple: the claim remains on the bank.
That matters because bank deposits already sit inside a regulated framework with capital, liquidity, resolution, supervision, and customer-protection rules. From the BIS perspective, that makes tokenized deposits the cleaner foundation for programmable payments. The money stays inside the system instead of being recreated outside it and then patched together with reserve assets and redemption promises.
But tokenized deposits are not some flawless utopia in a suit.
They can create closed networks. Liquidity can get trapped. Smaller banks may face higher implementation costs. And 24/7 transferability can bring new operational risk, because payments that never sleep also never stop demanding resilience. Innovation is fun right up until the system gets used the way users actually use systems.
Still, the BIS sees tokenized deposits as a more coherent path for programmable money because they preserve settlement in central bank money. That is the anchor. If two dollars are going to be treated as two dollars, the settlement layer has to deliver the same final value without drama.
The regulatory map is fragmented
A Financial Stability Institute study published one day before de Cos’s speech compared stablecoin regulation in the United States, the European Union, the United Kingdom, Hong Kong, and Singapore. The conclusion was blunt: the rules are all over the place.
That matters because stablecoin issuers do not live in one country in practice. They structure entities, affiliates, custody arrangements, and reserve operations across jurisdictions. The more fragmented the rulebook, the easier it is for firms to pick the most convenient regime and optimize around the rest.
Under the U.S. framework discussed in the study, lending, staking, proprietary trading, and custody are generally outside a payment stablecoin issuer’s core permissions under the GENIUS Act. The European Union, the United Kingdom, and Hong Kong allow certain extra activities with separate authorization or consent.
That difference is not just legal trivia. It changes what kind of business a stablecoin issuer can be, what risks it can take, and how easily it can expand into adjacent services. The more permissioned the structure, the less room there is for cowboy behavior. That is both the point and the problem.
The study also flagged a group-level gap. In some cases, restrictions apply to the issuing legal entity, not every company in the same corporate group. That creates an obvious loophole: the issuer can be ring-fenced while riskier business sits nearby in the corporate family tree, ready to cause reputational or contagion spillover if something blows up.
One caveat: the FSI study says its conclusions represent the authors’ views and do not necessarily reflect the position of the BIS or its member central banks.
The policy fight is bigger than stablecoins
Stablecoin backers often argue that these tokens will strengthen dollar demand and deepen Treasury markets. U.S. Treasury Secretary Scott Bessent said in July 2025 that stablecoins are “a revolution in digital finance.”
The bullish case is easy to understand. Stablecoins can move dollars across borders quickly, extend dollar access to users outside the banking system, and create more demand for short-term U.S. government debt. For the dollar, that is a nice geopolitical and financial story.
But the tradeoff is just as real. If demand comes mainly from domestic users shifting deposits out of banks, banks lose a source of funding. If redemptions spike, issuers may need to dump Treasury bills. If compliance is too weak, illicit finance gets easier. If the system fragments across chains and issuers, interoperability becomes a mess.
So the real argument is not “stablecoins good” versus “stablecoins bad.” It is whether programmable money should be built on open crypto rails or inside regulated bank infrastructure. One path is open, messy, and permissionless. The other is cleaner, more controlled, and much more to the liking of central bankers. Pick your poison.
Project Agorá shows the other direction
The BIS is not just criticizing stablecoins from the sidelines. It is also helping test what a more controlled tokenized settlement system could look like.
Project Agorá involves seven central banks and more than 40 private financial institutions. It has tested cross-border settlement using tokenized commercial bank money and central bank reserves. In plain terms, it is an attempt to make cross-border payments programmable without handing the keys to a loose pile of private crypto infrastructure.
The project has moved from prototype work toward real-value testing. That matters because demos are cheap and real money is not. Once actual value is on the line, the nice PowerPoint claims either survive or get exposed as theater.
Agorá is important because it shows where the institutional camp is heading: not away from tokenization, but toward tokenization with guardrails, oversight, and settlement anchored in central bank money.
Key takeaways
-
Can stablecoins become mainstream money?
Not under the BIS’s current framework. De Cos says they still do not credibly function as payment money at scale, especially under stress. -
Why does the BIS prefer tokenized deposits?
Because they remain claims on regulated banks and settle in central bank money, which preserves the existing monetary framework. -
Are stablecoins useless?
No. The BIS still sees narrower roles for them, including decentralized lending and some crypto-native uses. -
Do stablecoins help the U.S. Treasury market?
They can, especially if demand comes from outside the U.S. But if they pull deposits out of banks, they can also raise funding pressure inside the financial system. -
Are global stablecoin rules aligned?
Not even close. The U.S., EU, U.K., Hong Kong, and Singapore are taking meaningfully different approaches to issuer permissions and extra activities. -
What is the real battle here?
Whether programmable money will be built on open stablecoin rails or on regulated bank infrastructure with tighter oversight and central bank settlement.
What this means for crypto
The BIS is not saying digital money is fake. It is saying the rails matter, the reserve model matters, and regulatory oversight matters even more.
Stablecoins are already important inside crypto markets for trading, transfers, collateral, and liquidity. That is a real use case. No need to pretend otherwise. But that is not the same thing as being ready to serve as universal money for the broader economy.
Tokenized deposits may be less sexy, less open, and less aligned with crypto’s permissionless ideal. They are also more consistent with the way central banks and regulators want the system to function. That tension is not going away anytime soon.
For now, the message from Jackson Hole is pretty clear: stablecoins have carved out a meaningful role, but the BIS is not ready to call them serious money at scale.
