For two years, large banks have been debating which stablecoin to standardise on. The unspoken assumption was that the digital-cash market would consolidate the way payment cards did — a small number of dominant tokens, and the bank's job was to back the right one early.
This week, Sygnum, a Swiss digital-asset bank, reported the opposite is happening. Institutional clients are no longer asking which stablecoin will win. They are asking for infrastructure that supports several at once. That shift is small in news terms but large in architectural terms, and it is worth understanding.
What Sygnum actually observed
Sygnum's clients are institutional treasurers, asset managers, and corporate finance teams using digital cash for settlement and payments. Their behaviour has changed in two specific ways.
First, they have stopped asking Sygnum to recommend "the right stablecoin." They now assume their workflows will involve USDC, USDT, JPYC, eventual euro-denominated coins, and bank-issued tokens like the one Japan's three megabanks plan to launch in 2027. The question is no longer which one — it is how to route between them.
Second, they have abandoned the earlier preference for private, permissioned blockchains. The new requirement is for tokenised-cash instruments that interoperate on public chains, with the issuer treated as a parameter and the rails treated as substrate. The reasons are practical: liquidity, neutrality, lower operating cost, and regulator visibility.
Both shifts point the same direction. The market is moving from a "pick the winner" mindset to a "route between many" mindset.
Why the monoculture assumption broke
Three forces broke it at roughly the same time.
Counterparty concentration is a regulator-flagged risk. A bank with most of its tokenised-cash exposure in a single issuer is one bad audit away from a balance-sheet problem. Regulators have started asking large institutions to demonstrate diversification across stablecoin issuers, the same way they require diversification across counterparties for any other instrument.
Jurisdictions are fragmenting, not consolidating. The US GENIUS Act, the EU's MiCA framework, Japan's stablecoin regulations, Singapore's, and the UAE's are not converging. They are setting incompatible rules about reserves, redemption, foreign-issuer access, and licensing. No single issuer can serve every market under every rulebook. A bank operating in five jurisdictions needs five-jurisdiction coverage, which means multiple issuers.
Public-chain economics finally beat private-chain economics. The original case for permissioned consortium chains — privacy, predictable fees, governance control — was undermined by improving public-chain throughput, lower gas costs, and the realisation that institutional privacy can be handled at the application layer rather than the chain layer. Permissioned chains became more expensive than the alternative they were supposed to be cheaper than.
Banks didn't choose pluralism because it's better. They chose it because monoculture turned out to be a bigger risk than they could underwrite.
What multi-token rails actually look like
Think of it like a foreign-exchange desk, not like a single-currency bank. An FX desk does not pick "the right currency." It quotes prices in many currencies, routes flows between them, and treats each currency as a parameter of the trade, not the identity of the desk.
Tokenised-cash rails are starting to behave the same way. The bank operates the routing infrastructure. The token issuer is selected per transaction based on counterparty preference, jurisdiction, liquidity, and cost. The chain is selected based on settlement guarantees and fees. The interesting product is no longer the token — it is the layer that picks between tokens.
| Old model | New model |
|---|---|
| Pick the winning stablecoin | Route across many issuers |
| Private/consortium blockchain | Public chain as neutral substrate |
| Bank locks in to one technology stack | Bank operates the routing layer; issuers are pluggable |
| Token is the product | Routing and compliance are the product |
This is what mature payments infrastructure usually looks like. Visa and Mastercard do not issue currencies; they route between them. Correspondent banks do not pick a single ledger; they bridge ledgers. The same shape is appearing in tokenised cash, just compressed into a much shorter timeline.
What this means for builders
If you are building anything that touches stablecoins, three things change.
Stop hard-coding the issuer. Treat the stablecoin as a runtime parameter. Your code should not care whether the transfer is USDC, USDT, or a future bank-issued euro token. If it does, you will be refactoring within eighteen months.
The interoperability layer is the product. The defensible position is not "we support USDC" or "we built our own stablecoin." It is "we route between any tokenised cash instrument with full compliance and provenance." That is what the next generation of payments infrastructure looks like.
Compliance is now multi-jurisdictional by default. If your product handles tokenised cash, you need to think about reserve rules, redemption windows, and foreign-issuer restrictions across at least the US, EU, and major Asian markets. The era of "ship in one jurisdiction and worry about the rest later" is closing for any institutional-grade product.
Conclusion
The institutional shift Sygnum is describing is not a crypto-market story. It is an infrastructure-maturity story. Markets that look like they are about to consolidate around one winner often turn out to look like FX desks instead — many issuers, neutral rails, and routing layers as the durable product.
Tokenised cash is following that path. The teams that understand it early will build the connective tissue, not the tokens. That is where the durable value sits.