The ECB asks Brussels to cut stablecoin reserves loose from bank deposits: the same filing tightens the yield ban so deposits don't leave anyway.

Unhook the float from bank deposits, then wall off every route back to it.

Strip the filing down to what was actually requested, versus what is merely fixed by the calendar.

One document, two rule changes and a deadline. Nothing here is law, and nothing here is scheduled to become law this year.

Here is what the ESCB is asking Brussels to change, holder side and issuer side, in the same filing.

• The two asks solve two different risks. The reserve fix protects banks from a stablecoin run. The yield fix protects bank deposits from stablecoin competition.
• Both branches are requests, not rules. Nothing in the diagram is in force. The Commission still owns the legislative pen.

The reserve fix and the yield fix defend the same institutions from opposite directions. Unhooking issuer reserves from bank deposits closes the channel that turned Circle's USDC into a March 2023 de-peg event, when Circle's disclosed $3.3 billion of reserves at Silicon Valley Bank tied a stablecoin's credibility to one failing lender's balance sheet. Remove the mandatory deposit tie and a future stablecoin run stops being a guaranteed bank run too. But reserves that no longer have to sit in bank deposits are also reserves no longer obligated to fund bank balance sheets at all, so the ESCB pairs that loosening with a tighter yield wall: block every indirect route, lending, staking, borrowing, loyalty rewards, back to a return on holding a stablecoin, so euro holders have no incentive to treat one as a savings substitute and pull deposits out through crypto rails instead.

The reserve fix borrows a number that already exists. The European Banking Authority drafted the 40%/60% liquidity-bucket thresholds as technical standards in 2024; the ESCB is not asking Brussels to invent new plumbing, it is asking Brussels to promote draft numbers already sitting in a technical standard into MiCA's own text. That is a faster ask than it looks, if the Commission agrees to make it.

The issuers this would rewrite are not abstract. Qivalis, the 37-bank euro stablecoin consortium tracee covered in June, and Circle's EURC both size reserve treasuries around today's guaranteed deposit slice. A liquidity-bucket test instead of a fixed deposit share is a different treasury design, not a rounding adjustment.

• This is a response, not a proposal. The Commission holds the pen. Nothing in MiCA's own text changes by 30 September, or this year.
• The timeline stretches past 2026 either way. The Commission's review is expected to shape legislative changes in 2027 at the earliest, and an actual MiCA amendment still has to clear the EU's ordinary legislative procedure, Parliament and Council both, a process that typically runs 12 to 24 months from proposal to force.
• The yield-ban expansion reaches outside MiCA's own perimeter. Crypto lending, borrowing and staking are not currently MiCA-regulated activities at all. Extending a MiCA prohibition to reach them may need a different legal basis than a straightforward amendment, a gap the ESCB's response does not resolve.
• This reverses the ECB's own position from four months ago. At the Nicosia ECOFIN on 22 May 2026, the ECB rejected the same reserve-floor easing when Bruegel proposed it, calling the reserve drag deliberate policy. The ESCB's filing does not explain what changed.
• The filing flags a gap it does not fix. MiCA has no legal basis today for issuing the same stablecoin both inside and outside the EU, a separate multi-issuance problem the ESCB raises but leaves for a later round.

This is the third euro-adjacent settlement or reserve move tracee has tracked inside seven days. Pontes went live on 21 September, wiring wholesale tokenised-asset settlement to central bank money. The FCA opened its UK stablecoin authorisation gateway the same week. Different regulators, different currencies, the same underlying question: what backs a token that claims to be money, and who is on the hook when holders ask for it back at once.

The Bank of England's own draft Code of Practice for systemic sterling stablecoins runs the opposite direction from the ECB's ask: a 70% gilts, 30% central bank cash split, with the cash tranche earning no interest, effectively mandating a central-bank tie rather than removing a bank one. An issuer running both a euro and a sterling stablecoin book off one treasury desk is not choosing between two drafts of the same rule. It is building toward two different reserve architectures at the same time.

The ECB is not asking Brussels for a looser stablecoin regime. It is asking for reserves that no longer read as a hidden subsidy to euro-area banks, paired with a yield wall wide enough that a stablecoin can never read as a savings account either. Both asks serve the same institution: the banking system MiCA was written to protect from stablecoins, not because reserves would shrink, but because deposits stay wherever holding a stablecoin looks the least attractive.

Watch three things before the Commission's next legislative step:

• Whether the 2027 legislative proposal keeps the liquidity-bucket swap or the 60% floor. That is what tells an issuer which treasury architecture to build toward.
• Whether the yield-ban expansion survives its own legal-basis problem. Lending and staking sit outside MiCA today; a prohibition that reaches them may need a different instrument entirely.
• Whether the EU and UK reserve architectures keep diverging. A euro book on liquidity buckets and a sterling book on a 70/30 gilts-and-cash split is two treasury designs for one desk to run.