The harder issue is inside the bank
Most informed readers no longer need convincing that several forms of digital money may coexist.
That part is increasingly accepted.
The underplayed issue is what that coexistence does to the middle of the bank.
When stablecoins, tokenised deposits, CBDC-linked flows, and tokenised assets start showing up against different use cases, banks often respond one use case at a time. That is where the real cost begins.
Controls get duplicated. Treasury visibility fragments. Exception handling stops looking consistent. The next use case becomes more expensive than it should be because the bank is quietly building another middle.
The market question is no longer whether several forms can coexist. The bank question is whether the core can absorb them.
They do not land in the bank the same way
This is not just a broader menu of digital instruments.
These forms change the bank in different ways.
The money question
Stablecoins, tokenised deposits, and CBDC-linked flows are not interchangeable wrappers.
- Stablecoins are claims on a private issuer and its reserves.
- Tokenised deposits are commercial-bank money in a new form.
- CBDC-linked flows sit closer to central-bank money.
That changes where the liability sits, who issues the money, how redemption works, what happens to deposits and reserves, and what Treasury and Operations have to manage in practice.
The asset question
Tokenised assets create a different layer again.
A tokenised bond, fund unit, deposit receipt, or other digital asset does not answer the money question on its own. It adds asset servicing, custody, lifecycle events, books and records, and settlement choreography to the picture.
That is why the money question and the asset question should not be collapsed into one.
One is about the form of money moving through the flow. The other is about the asset being issued, transferred, settled, serviced, or redeemed.
The practical base case is coexistence
Banks should assume mixed combinations of money forms, networks, and tokenised assets are likely to persist for some time.
Not because the market is undecided in some abstract sense, but because institutions are solving different problems and moving at different speeds.
That means the practical task is not to wait for one winner. It is to decide what the bank builds once and what it is willing to vary at the edge.
The architecture consequence follows quite quickly
Once that is accepted, the practical consequence is hard to avoid.
If the bank may need to support several money forms and several digital assets, it cannot keep rebuilding the middle every time a new combination turns up.
That is how the same bank capability ends up funded three times.
One team builds around stablecoin connectivity. Another builds around tokenised-deposit issuance. A third adds custody and servicing for tokenised assets. Each move can be defended locally. The trouble starts later, when management discovers it has multiple control patterns, multiple exception processes, and no clean way to extend into the next use case without reopening the whole architecture question.
This is why the centre of gravity should stay on one reusable core, many instruments.
What should be common, and what should vary
The core should carry the repeatable bank capabilities.
That usually means things like:
- controls, approvals, and monitoring
- Treasury and liquidity treatment
- positions, balances, and status visibility
- orchestration and exception handling
- reconciliation, reporting, and audit-ready traceability
- the repeatable parts of issuance, redemption, custody, and servicing workflows
Variation should sit where the instrument genuinely changes the shape of the problem.
That includes the legal model, network behaviour, settlement path, asset lifecycle, and customer journey.
The mistake is letting those edge differences drive a fresh middle every time.
The practical test
The question is no longer whether the market can support more than one form.
The question is whether the bank can.
If a second or third form turns up next year, does the bank extend the core, or quietly fund another middle?