State of Stablecoins

The Wall Street Journal reported on August 26 that JPMorgan — the most prominent public sceptic on stablecoins, which had consistently argued a bank with the Kinexys platform needs no token — has held internal discussions about issuing its own stablecoin. The bank's statement was carefully hedged: no plans to issue, but all options would be evaluated depending on customer demand and the regulatory landscape. The second part of the report carries more weight: more than a dozen financial institutions, including Bank of America, Wells Fargo and Santander, are working on a separate joint stablecoin for corporate clients — dollar first, then the euro and the other G7 currencies.

This is a reversal of the position around which the banking storyline of recent months was built. Until August the banks' answer was framed as tokenized deposits instead of a token: JPM Coin and Kinexys, The Clearing House network, the SWIFT ledger. The structural difference banks now concede publicly: a tokenized deposit is tied to one bank's balance sheet and lives inside its perimeter, while a stablecoin travels across public networks between wallets and applications. If a corporate client wants the latter, a deposit token will not hold it. A material caveat: this is reporting on preliminary discussions, not a product announcement — no structure, timing or issuer has been disclosed.

On August 25, 39 state bankers associations announced BankChain Alliance — a nationwide blockchain network “owned by the industry, designed by the industry, governed by the industry”. The participating associations represent 3,283 banks with $21.8 trillion in combined assets (FDIC call report data, March 31, 2026). Launch is targeted for 2027. Interim chair is Kathy Kraninger, head of the Florida Bankers Association and former director of the CFPB. The network is designed for tokenized deposits and regulated stablecoins at the same time.

This is the missing bottom tier of the structure the US banking system has been assembling all year. Until now the banks' answer was a story of the largest players: The Clearing House network with JPMorgan, Bank of America, Citi and Wells Fargo, the SWIFT ledger, Cari Network for regional banks. BankChain gives blockchain rails to a community bank that cannot run its own ledger. What is genuinely new: the bank lobby, for the first time, is not opposing deposit tokens to stablecoins but building one network for both — even as the same associations pressed the Senate in July to tighten the CLARITY Act's yield provisions. Two qualifications: the announced participants are associations, not banks, and no bank has publicly taken an ownership stake; no technology partner, consensus model or interoperability plan with FedNow, RTP or CHIPS has been disclosed. At least four competing bank blockchains are now under construction in the US.

On August 19, HSBC and Standard Chartered executed the first live cross-border interbank transfer in tokenized deposits on SWIFT's blockchain-based ledger. Obligations were recorded on HSBC's Tokenised Deposit Service and Standard Chartered's tokenized-deposit infrastructure. The bank's own wording describes the architecture precisely: the SWIFT ledger acted as a secure orchestration layer that matched and netted obligations between the two banks before final settlement through existing systems. The transaction builds on the ledger's July launch, whose pilot involved 17 banks.

The architectural detail matters more than the milestone: the SWIFT ledger operates not as a settlement network but as an orchestrator. Banks are building a 24/7 layer on top of correspondent accounts without abandoning them — the exact opposite of the thesis that stablecoins replace the correspondent chain. The pilot has produced a working transaction, giving the deposit camp its first proof of operation rather than another announcement. The choice of participants is telling: in the same week Standard Chartered took the other side of the argument — on August 24 it became the first bank distributor of HKDAP, Hong Kong's first regulated HKD stablecoin, issued by Anchorpoint Financial, in which the bank is majority shareholder. Declared use cases: money market fund subscriptions, settlement with asset managers, intragroup and cross-border payments, a tokenized MMF service in Q4. One bank is building both rails at once.

On August 17 the US Treasury proposed the Section 3 rule under the GENIUS Act — the first major implementing regulation. It defines when a payment stablecoin is issued in the United States and when a platform is selling to a US person. Foreign issuers avoid US-issuer status under four conditions: the issuer is not located in the US; there is reasonable belief each recipient is outside the US; appropriate policies and controls are implemented; and there is no US-directed advertising or solicitation. The ability to comply with US lawful orders and reciprocal arrangements is a separate requirement. From July 18, 2028, digital asset service providers may not offer Americans stablecoins issued by anyone other than a permitted or qualified foreign issuer; peer-to-peer transfers and self-custodial wallets are carved out. Federal Register publication followed on August 18; comments close October 19. Two days later, at the Wyoming Blockchain Symposium in Jackson Hole, Comptroller Jonathan Gould committed the OCC to a final version of its 376-page February NPRM by November, with issuer applications processed from the start of 2027.

The legislative track has stalled: the CLARITY Act vote did not happen before the August recess, the first procedural vote — cloture on the motion to proceed — is set for September 15, and the stablecoin-yield dispute remains unresolved. In these conditions the framework is being written by agencies, and it now has a calendar. GENIUS takes effect on the earlier of two dates — January 18, 2027, or 120 days after final rules — so a November OCC publication effectively synchronizes both. For banks the practical meaning is in Gould's figure: since January 2025 the OCC has received 40 applications for new bank charters, 23 of them tied to digital assets. The material contradiction for risk assessment: the industry is commenting on rules for a statute parts of which the CLARITY Act may rewrite.

On August 18 the FASB issued the exposure draft “Cash Equivalents — Disclosure Enhancement and Evaluation of Certain Digital Assets”. The board proposes not to change the definition of cash equivalents but to add illustrative examples showing when a dollar token meets it. Three criteria: a contractual right to redeem on demand; a direct redemption right with the issuer for a known amount of cash; and segregated reserves of at least 1:1 against circulating supply, held in short-term highly liquid assets with annual disclosure of composition. The project began after a 6:1 board vote in October 2025. Public comments are open until November 19.

For corporate treasuries this is arguably the most consequential item of the month. While a stablecoin is booked as an “other digital asset”, a treasurer cannot hold it in place of cash: liquidity metrics, covenants and the cash flow statement all suffer. Tokens that pass the tests would sit in the accounts alongside Treasuries, commercial paper and money market funds. The accounting standard is the missing link between “the stablecoin as a settlement rail” and “the stablecoin as an asset a treasurer will keep on the balance sheet overnight”. The counter-argument to keep in view: Jack Castonguay (Hofstra University) argues the FASB is changing the spirit rather than the letter of the definition, and that after the depegs and runs of recent years admitting stablecoins to the cash line is premature.

DG FISMA's targeted consultation on reviewing MiCA, opened on May 20, gained momentum in August and was extended: feedback is now accepted until September 30, 2026. The focus is on access rules for third-country issuers and a possible equivalence regime, multi-issuance schemes (the same token issued by several entities in different jurisdictions), and extending the framework to tokenized payment instruments and deposits, DeFi, staking and lending. The Commission's formal report to Parliament and the Council under Articles 140 and 142 is due by June 30, 2027, and may be accompanied by a legislative proposal.

MiCA de facto pushed Tether out of the European market: delistings ran through Coinbase Europe, Crypto.com, Binance and Kraken, and on August 31 Revolut ended USDT support for European clients. Reviewing third-country access means that outcome is no longer treated as final. The direct trigger is the GENIUS Act and US stablecoin policy: a framework adopted in May 2023 will be recut for the post-GENIUS world. The second review track matters just as much for banks: extending MiCA to tokenized payment instruments and deposits covers precisely what European and US banks are building now. The other side of the story: on August 26 Revolut launched its own euro stablecoin EURR, issued by Bridge's Luxembourg entity — Stripe's payments arm — under a CSSF licence; five days before the USDT cut-off, the bank took the vacated niche with its own product. Timing caveat: the review remains a consultation procedure, not an adopted decision.

The Kansas City Fed's Jackson Hole symposium (August 27–29) ran under the theme “Financial Innovation: Implications for Payments and Policy”; the first keynote of the new Fed Chair Kevin Warsh was devoted to financial innovation. On August 28, BIS General Manager Pablo Hernández de Cos said stablecoins are not a credible means of payment at scale and called for anchoring tokenization in central bank money. His three charges: singleness of money — stablecoins have no mechanism forcing redemption at par, and token-for-token exchange runs through a secondary market that may clear away from parity; interoperability — tokens on different networks do not meet without risky or costly workarounds; financial integrity — balances sit mostly in self-custodial wallets, complicating the AML/CFT perimeter. His alternative: tokenized deposits. A day earlier, on August 27, HM Treasury announced amendments to the Financial Services and Markets Bill giving the Bank of England a secondary statutory objective — to support innovation in payment systems and digital money, including systemic payment systems using digital settlement assets.

De Cos's thesis lands squarely on the August banking line: the deposit camp now has both an institutional theory from the BIS and working infrastructure — from The Clearing House and the SWIFT ledger to BankChain Alliance. The British construction matters legally: the objective is secondary, subordinate to the primary task of financial stability, and does not oblige the Bank to support innovation where it threatens stability; the Bank will report annually to Parliament on progress. The practical meaning is a changed negotiating position: the Bank of England has been the most cautious major central bank on stablecoins all year, and the industry now has a formal lever to demand justifications. The British track is also advancing technically: on August 12 a consortium of Polygon Labs, NOBO Finance and Dun & Bradstreet entered phase 2 of the Digital Pound Lab with a trade finance scenario — an exporter settled in a stablecoin, a UK importer paying in a simulated digital pound, both legs closing in a single flow. Caveats: the UK amendments are announced, not enacted — the bill is before the House of Lords on September 7 and 9; the Digital Pound Lab is a sandbox without real money, and no launch decision has been taken.

On August 25, Rosie Levy and Srini Ramaswamy published on the Dallas Fed's site a calculation of what deposit tokenization does to a bank balance sheet. The logic: instant settlement shortens a deposit's weighted average life and raises its beta — the sensitivity of the deposit rate to the central bank rate. Per H.8 data, about 80% of US banks' interest rate risk ($5.8 trillion of $7 trillion in ten-year equivalents) rests on the duration properties of deposits; a 10% cut in weighted average life removes roughly $580 billion of maturity transformation capacity, a 10% rise in beta roughly $700 billion. The empirical anchor is Brazil's Pix (about 200 million active users, about $650 billion in monthly transactions). On August 27 the New York Fed released Staff Report No. 1202, “Stablecoins Meet the Mundell–Fleming Trilemma” (Pablo Azar, Maryam Farboodi, Nish Sinha): nine crisis episodes in eight countries over 2021–2025, wallets tagged by country attributes in ENS registrations, about 4.47 million wallet-event-week observations across the 19 largest dollar stablecoins. The result: in the week a crisis begins, tagged wallets are 1.8 percentage points more likely to receive dollar stablecoins against a 9.8% baseline; outbound activity adds 1.3 percentage points two weeks later; there is no pre-trend in the two weeks before the shock.

The Dallas Fed paper moves the stablecoins-versus-tokenized-deposits dispute from product announcements to balance sheet numbers — and carries an uncomfortable conclusion for banks. A tokenized deposit looks like the safe alternative: it sits inside existing regulation and pays interest to the holder. But instant settlement destroys the friction on which bank economics rests. Banks either shift to term debt, becoming economically similar to non-bank lenders, or grow their buffer of high-quality liquid assets. The flight into tokenized deposits imports the very problem banks adopt them to escape. The New York Fed paper answers the second, macroeconomic question: under Mundell–Fleming a country cannot hold a fixed exchange rate, free capital movement and an independent rate at once, and historically the third corner was pinched through banks. Stablecoins give a household a route to the dollar around the domestic bank. Both caveats are material: the Dallas piece is a research publication, not a Fed position, and its authors decline to estimate adoption probability; the New York study does not show stablecoins toppling any currency, and the sample is wallets already tied to stablecoin activity, not a country's population.

On August 13, Tether announced that KPMG US had issued an unqualified opinion on the full financial statements of Tether International, S.A. de C.V. for the year ended December 31, 2025 — the first complete audit in the company's history. Reserves exceeded liabilities by $6.814 billion at the reporting date; the auditors physically counted every gold bar. The current picture is different: the Q2 2026 attestation (BDO, published July 31) showed a buffer of $4.11 billion against $8.23 billion at the end of Q1, on assets of $187.75 billion versus $183.64 billion of liabilities. On August 5, Circle reported Q2: revenue and reserve income of $701 million (+7% year over year), net income $48 million, adjusted EBITDA $143 million (+8%), diluted EPS $0.18, USDC in circulation $73.3 billion (+19% year over year), on-chain volume $14.8 trillion (+151% year over year). The reserve return rate fell 66 basis points.

The audit removes the oldest argument against the largest issuer: the promise dates back to 2017, and the NYAG ($18.5 million) and CFTC ($41 million) penalties were imposed precisely for reserve claims without audit confirmation. For Tether it is the key asset ahead of entering the US market under the GENIUS Act. Comparing the two dates gives a more restrained picture: the audited reserve surplus of $6.81 billion at end-2025 had compressed to $4.11 billion by mid-2026 — a cushion of about 2.2% over $183.6 billion of liabilities. Circle's results show the other half of the issuer economy: on-chain volume grows at triple-digit rates while income declines with rates — a model where revenue is set by the reserve return rate, not by turnover, is sensitive to monetary easing regardless of usage growth.

Total stablecoin market capitalization fell from a May peak of about $322.1 billion to $307.6 billion by August 2 and $304.4 billion by September 1. June's $11.4 billion decline was the sharpest monthly contraction since May 2022, and the episode overall ranks as the third-largest drawdown in the industry's history. The peg held throughout: per Cumberland, USDT traded in a $0.9988–0.9992 corridor and USDC above $0.9997. Settlement infrastructure was redistributed in parallel. On August 3, Mastercard closed its $1.8 billion acquisition of BVNK ahead of schedule, taking a London stablecoin-payments infrastructure with about $30 billion in annual volume, support for more than 150 currencies and presence in more than 200 markets — the first major payment network to buy stablecoin infrastructure outright. On August 5, Visa moved stablecoins into production on Visa Direct: USDC prefunding and payouts via Zerohash across more than 18 billion endpoints in 195 countries. On August 18 the consequence surfaced: Visa issued an RFP for a new stablecoin settlement partner, BVNK having gone to a competitor.

Contraction in capitalization alongside record transaction volume marks a change of function: the stablecoin is becoming a settlement instrument rather than a savings one. The main hypothesis is structural: the GENIUS Act's prohibition on paying interest pushes income-seeking capital into tokenized Treasuries and other RWA. Cumberland's reading refines the picture — an orderly exit from crypto assets, not a crisis of confidence in stablecoins, as the peg's behavior confirms. Hence the direct link to the FASB proposal: if the accounting standard admits stablecoins as cash equivalents, a corporate holder gains a motive to hold the token that is not tied to yield. The second line: competition between payment networks has moved from issuing tokens to owning distribution and settlement infrastructure. Mastercard bought the contractor outright; Visa works through partnerships and is now seeking a replacement — while both networks participate in the Open USD consortium and joined the founding validator cohort of Circle's Arc blockchain, announced August 5 alongside BlackRock, ICE, DTCC and Standard Chartered.