Regulatory frameworks for stablecoins approved between 2024 and 2026 share a starting point: one-to-one backing. Consensus breaks when defining reserve asset composition. The GENIUS Act permits currency and Treasury bills with residual maturity of 93 days or less, plus overnight repos collateralized by sovereign debt.
MiCA requires a minimum of 30% in bank deposits and raises the threshold to 60% for issuers designated as significant. The Bank of England adopted a 70/30 model with 30% held in non-remunerated central bank deposits. Three jurisdictions, the same backing ratio, three risk profiles.
Public discussion focuses on the coverage percentage. The parameter defining token behavior during stress is different: collateral composition, legal treatment against creditors, and speed of liquidation without loss. The ratio is a solved problem. The reserve is not.
The reserve as monetary policy instrument
When a rule requires 60% of the reserve in bank deposits, the stablecoin float channels into the banking system. When a rule requires Treasury bills, the float goes to the short-term sovereign debt market. The choice is not neutral. Each regime decides which balance sheet absorbs resources and which captures associated yield.
The UK model takes the reasoning to an explicit form. Non-remunerated deposits at the central bank withdraw liquidity and transfer float yield to the monetary authority. The US model operates in inverse form: the issuer retains Treasury yield and books it as operating income. The difference affects issuer cost structure, redemption pricing, and capacity to sustain low-margin operations.
Bank deposits: the requirement introducing credit risk
Requiring a high percentage of bank deposits introduces direct exposure to credit institutions within the reserve asset. An instrument designed to reduce credit risk on the liability side incorporates credit risk on the asset side. In a banking stress scenario, a deposit at an institution under pressure is the least liquid reserve component, not the most liquid.
Prudential logic justifying the requirement—proximity to central bank money—does not hold when the depository institution faces a run.
Tether declined to seek authorization under MiCA and cited the deposit requirement as a central obstacle. Circle chose a structure of separate entities by jurisdiction. Both responses are rational. Neither solves the underlying problem: a global issuer cannot maintain a single reserve satisfying two regulators with incompatible criteria.
Redemption timelines and liquidity mismatch
Par redemption timelines differ more than ratio convergence suggests. The US regime sets two business days, with extension to seven calendar days when requests exceed 10% of circulating supply in 24 hours. MiCA establishes an immediate right without fees.
The Bank of England requires redemption within 24 hours for systemic issuers, without suspension power. Singapore grants five business days.
A promise of 24-hour redemption against a Treasury bill reserve settling T+1, without access to a central bank liquidity facility, creates an operational mismatch. The mismatch worsens on weekends, holidays, and market closures.
The UK proposal for a liquidity facility addresses the gap. Without a central bank backstop, the issuer covers the mismatch with cash or bank lines and passes cost to the fee structure or reserve yield.
Yield prohibition and issuer economics
Major jurisdictions prohibit paying interest to the holder. The United States, European Union, Singapore, Hong Kong, United Arab Emirates, and Canada agree.
Coordination eliminates regulatory arbitrage on the liability side but concentrates float yield in the issuer. The result is a business model whose margin depends on the spread between reserve yield and operational and compliance cost.
The practical consequence: a more restrictive reserve composition means lower available yield and a narrower margin. A regime requiring 60% in bank deposits remunerated below money market rates compresses issuer income. Yield prohibition and composition restriction operate as combined controls on sector profitability.
Capital, segregation, and the parameter holders ignore
Own capital requirements do not converge. MiCA sets 2% to 3% own funds. The Bank of England requires the greater of six months of operating expenses or the cost of a recovery and orderly wind-down plan. Hong Kong sets HK$25 million paid-up capital.
The UAE requires Dh15 million initial plus capital linked to circulating supply. Loss absorption capacity varies by orders of magnitude among regimes declaring the same backing standard.
The decisive parameter is patrimonial segregation. The UK regime requires a statutory trust. Hong Kong and Australia require trust structures.
The relevant question is not how many assets back the token, but what happens to the assets when the issuer becomes insolvent and whether the assets remain outside the custodian’s bankruptcy estate.
A full reserve held in an omnibus custodian account, without accounting separation or individual identification, does not protect the holder against a custodian creditor.
Frameworks requiring identification and prohibiting rehypothecation solve part of the problem. Frameworks requiring only custody at a regulated institution do not.
Foreign issuer treatment: fragmentation by design
Japan requires credit risk category 1-2 and a minimum of ¥100 trillion in outstanding bonds from the foreign issuer, a threshold excluding most candidates. The UAE prohibits use of foreign tokens for payment of goods and services.
The Swiss draft treats non-Swiss tokens as crypto assets without backing, regardless of home regulation. The European Union does not recognize equivalence automatically. Singapore incorporates a recognition regime. The United States has a federal framework.
The predictable result is an industry organized as a set of local entities. Each entity maintains its own reserve, meets its own capital requirements, and reports to a different authority.
Token liquidity fragments, interoperability between jurisdictional versions becomes an infrastructure problem, and cross-border redemption becomes more expensive. Cost transfers to the end user through spreads.
What should be harmonized
The alternative is mutual recognition based on equivalence of outcomes: reserve quality, effective segregation, enforceable redemption right, periodic disclosure, and independent audit. Composition can differ if the prudential outcome is equivalent.
Frequency, auditor standard, and report scope vary across regimes. An attestation confirming reserve amount without verifying segregation and absence of liens covers a fraction of the risk. Australian quarterly reports and monthly external audits required in the UAE differ in scope and frequency.
Treatment of tokenized money market funds as reserve assets also merits review. Several frameworks do not admit the funds, despite replicating the liquidity profile of instruments authorized. Exclusion reduces options without reducing risk.
Reserve requirements ceased to be a technical detail and became the principal instrument of prudential and monetary policy over the sector.
The backing ratio is the visible part and the least informative. Parameters determining holder risk are collateral composition, legal structure of segregation, and liquidity availability at redemption.
Issuers segmenting operations by jurisdiction and assuming multiple compliance costs will capture market access. Issuers seeking a single global structure will face rising costs and access restrictions.
For the holder, the relevant metric is not the declared backing percentage, but the legal enforceability of a claim over a defined asset set.
