The Bank of England's stablecoin rulebook starts too late by design

A rulebook needs somebody to obey it

The Code of Practice the Bank of England published for stablecoin issuers has, on the day of publication, no addressee. Not one firm is bound by it. That is not an oversight in the drafting. It is the architecture.

A stablecoin is a digital token engineered to hold a fixed value against ordinary money — one coin, one pound — so that it can be used to make payments and settle obligations. The Bank of England, the United Kingdom’s central bank, has now issued a statement of policy and a draft set of rules for the issuers of such tokens. But the rules are written for systemic issuers. The adjective is load-bearing. It means the Code reaches an issuer once the Bank has formed the view that the issuer’s failure or disruption would matter to the financial system as a whole.

Regulation that arrives when the risk already has

Compare this with how a bank is supervised. A bank is authorised before it takes its first deposit. The obligations come first, the business second, and the sequence is the point: the regulator sets the terms of entry rather than reacting to what walks through the door.

A designation-gated regime inverts that. The obligations attach when the firm has already become important. So the phase this Code does not govern is precisely the phase that generates the exposure — the years in which a token accumulates holders, gets wired into other firms’ plumbing, and builds up a redemption promise it has never had to honour under stress. By the time the Bank concludes the issuer is systemic, the thing worth worrying about has already been built.

This is why publishing a rulebook is the cheaper half of the job. The expensive half is publishing the trigger.

The cliff edge nobody has priced

Designation, as the item frames it, is something the Bank does to a firm rather than something the firm elects into. An issuer therefore cannot know the date on which its cost base changes shape. It can guess, and it will hire people to guess, but it cannot plan against a threshold it has not been shown.

The consequence is a step change in obligations landing at the moment of maximum growth and minimum slack. Governance, resilience testing, reporting, whatever the final Code demands — all of it arrives at once, on a timetable set elsewhere. A firm that has to retrofit systemic-grade controls under a deadline it did not choose is a firm more likely to do it badly. Graduated obligations, disclosed in advance and phased in, would produce better compliance than a single unheralded step.

What “draft” and “policy statement” together tell you

The Bank published both on the same day. A statement of policy is normally the settled part; the draft is the wording. Read together, they suggest the framework has been decided and what remains open is expression. Firms will read it that way, and they will be broadly right.

That is defensible — consultations cannot be infinite — but it places a duty of candour on the Bank. If the shape of the regime is fixed and only the drafting is live, say so, and say which questions genuinely remain open. A consultation that invites comment on settled matters wastes everyone’s submissions and corrodes the credibility of the next one.

The strongest case for waiting

There are two good arguments against me. The first is proportionality: it would be absurd to impose central-bank-grade supervision on a token with a hundred users, and the Bank’s remit is systemic risk, not every payment experiment. The second is gaming. Publish a numerical threshold and firms will park just underneath it, splitting entities and shaving volumes to stay outside the perimeter.

The proportionality point is right about the content of rules and wrong about the timing of their arrival. You can scale obligations to size without withholding the map. A published ladder is proportionate; an undisclosed trapdoor is not.

The gaming point is stronger, and I concede part of it. But an issuer that deliberately restrains its own growth to stay below a disclosed line has been curbed by disclosure — which is a policy outcome, not a failure of one. And unpublished lines are gamed too. They are simply gamed in private, through lobbying and supervisory relationships, where nobody outside can see who got what.

The Bank is also the incumbent

One uncomfortable adjacency. The same institution has recently announced a delay to the November 2026 standards release for its real-time gross settlement system — the plumbing through which sterling payments finally settle. Australia’s payments regulator, meanwhile, has been consulting on what its own settlement system should do in a tokenised environment.

No bad faith is implied, and none should be. But a central bank that both operates the incumbent rails and writes the rulebook for private substitutes should be conspicuously clear about the two hats. Part of the demand for privately issued payment tokens is a verdict on the speed of the public alternative. That is a reason to publish the perimeter early, not late.