The four numbers, two orders of magnitude apart.

Strip the discourse down to its load-bearing data and four figures describe the entire market. Outstanding supply, ~$321B. Gross on-chain transfer volume, ~$46T annualised. Adjusted economic volume, ~$4.2T. Real-economy payments, $350-550B. Each number is correct. They describe different things. Reporting "$46 trillion in stablecoin volume" without specifying which definition is in use is exactly how a regulator's confidence in market reporting erodes.

Underneath the headline, two contextual points carry the most analytical weight. First, BCG and Allium estimated that of approximately $62T of gross on-chain stablecoin transfers in 2025, only approximately $4.2T was economically meaningful after removing bots, intermediary routing and protocol mechanics⁹; of that, $350-550B was real-economy payments for goods and services, growing at ~60% YoY. Second, the IMF estimates the GENIUS Act reduced the market capitalization of listed U.S. incumbent payment firms by approximately 18%, or roughly $300B, when scaling the observed event-window response by a Polymarket-implied anticipation probability of 93%⁵.

Among the six majors, the positioning splits four ways. Tether is the offshore liquidity giant ($186B+ in USDT, $10B+ 2025 net profit, $141B+ U.S. Treasury exposure). Circle is the publicly listed U.S. challenger ($79B USDC; 80%+ of reserves in the BlackRock-managed Circle Reserve Fund at BNY Mellon). Paxos is the regulated white-label rail used by PayPal for PYUSD, supervised by NYDFS and OCC. Anchorage Digital Bank, N.A. is the only federally chartered crypto bank in the U.S. and the issuer of record for USDtb and USAT. Gemini issues GUSD under NYDFS oversight at small scale. Coinbase is not a fiat stablecoin issuer but is the largest USDC commercial counterparty: 50% of off-platform residual reserve revenue and 100% on Coinbase.

Behind the majors sits a mid-cap tier this report covers in unusual depth: USD1 ($4.61B), DAI ($4.60B), USDe ($4.35B), BUIDL ($3.17B), USDG ($2.79B), USDY ($2.12B), USDf ($1.62B), RLUSD ($1.60B), USDD ($1.53B), USDtb ($1.13B), United Stables ($1.06B), GHO ($583M), USD0 ($555M), YLDS ($503M). The mid-cap tier mixes four distinct archetypes (fiat-backed non-bank, crypto-collateralised, synthetic, tokenized-fund RWA), and the operator question is not "which is the next USDC" but "which archetype is the right tool for which use case."

Source Reliability Index.

The document draws on three reliability tiers. Readers should weight claims accordingly.

Top-10 Stablecoin Market Cap.

Outstanding supply is the most concentrated metric in the entire market. The top five stablecoins control roughly 88-90% of total supply. The top ten exceed 95%. Hundreds of stablecoins launch each year. Almost none survive past their first year of operation.

Speed to $1B in circulating supply

Gross On-Chain Volume vs Real Economic Activity.

Headline volume numbers can mislead. BCG and Allium analyzed approximately $62T of gross on-chain stablecoin transfers in 2025 and found only approximately $4.2T was economically meaningful after removing bots, internal contract routing, bridge mechanics, intermediary hops and pre-minted or synthetic tokens⁹.

The two waterfall charts below, reproduced from BCG and Allium's 2025 stablecoin payments report, decompose the same gap step by step. The first chart removes bots and protocol mechanics to get from gross to adjusted volume. The second chart removes CEX settlement, DEX, and treasury flows to get from adjusted to real-economy payments.

Real-economy bucket grew ~60% YoY against an ~$200T global payments market.

$4.2T adjusted breakdown

Chain Distribution and the Tron Question.

Chain choice is purchasing criterion tier-one, not an afterthought. USDT on Tron is the cheapest rail for retail cross-border. USDC on Ethereum or Base is the default for U.S. institutional flows. PYUSD on Solana benefits from sub-cent fees. The chain a stablecoin settles on reveals who the marginal user is.

Issuer-to-chain canonical deployment heatmap

A "USDC" on one chain may not have identical redemption or legal properties as a "USDC" on another chain. Dark = primary canonical deployment; lighter = secondary canonical; blank = bridged only or not deployed.

Geographic skew and the two markets.

The composition of the real-economy slice matters more than the size. B2B settlement accounts for roughly 40% of bilateral payment volume and is growing at about 65% a year. C2C, predominantly cross-border remittance, is 25%, expanding at roughly 75%. C2B, consumers paying digital and cross-border services, is another 25%. B2C payouts are the smallest segment, around 10%, constrained by the persistence of efficient domestic fiat rails in developed markets.

The geographic skew is sharper still. Surveys of active stablecoin holders place roughly 37% in Southeast Asia, 21% in South Asia, 11% in Africa, and 7% in MENA. North America and Western Europe combined sit below 10%. The institutional flows concentrate in the latter; the retail and remittance flows in the former. Two markets, one asset class.

The five archetypes, separated cleanly.

When people say "stablecoin" they usually mean the first archetype below, the fiat-backed non-bank token typified by USDT and USDC. That image is correct as a description of the market by volume, and badly misleading as a description of where the market is heading. Four other archetypes carry meaningful supply, address different counterparties, and increasingly sit on different regulatory tracks. A serious operator needs to hold all five distinctly in mind, because the operational implications of choosing one over another are not marginal, they are structural.

The archetypes are not points on a single quality scale, they are different products with different counterparties and different failure modes. A fiat-backed non-bank stablecoin can scale to hundreds of billions of supply because it monetises a float of short-dated US government paper; a bank deposit token, by contrast, is a representation of an underlying liability that already exists on a balance sheet and therefore cannot scale beyond that bank's deposit base. A tokenized-fund RWA is closer in form to a money-market product wrapped in a token than to either of the above. Treating these as interchangeable is the single most common diligence error we see in 2026 client engagements.

Four compounding moats. Why the top five extend their lead.

Concentration in stablecoins is not a temporary artefact of the market being young. It is the predictable output of four reinforcing moats, each of which strengthens with scale rather than weakening. Once an issuer crosses a threshold on any one of them, the cost for a challenger to overtake it grows non-linearly. Once an issuer crosses the threshold on all four, the lead is structural for at least a decade.

• Liquidity depth. Stablecoin pairs concentrate at the top because every market-maker, exchange, and OTC desk needs to quote the same instrument the rest of the market quotes. A new token that costs more to convert is a new token traders avoid. Liquidity begets liquidity, and the cost of crossing the spread on a non-leader token can be 5-15 basis points higher than on a leader, a death tax for any high-frequency use case.
• Redemption credibility. The single most-tested property of a stablecoin is whether you can get your dollar back at par, fast, when you want it. Tokens that have survived a 2023-style stress event with same-day redemption intact have an inherited credibility a newcomer cannot manufacture. Track record here is asymmetric: years of clean redemption add slowly to trust; a single failed redemption window deletes it instantly.
• Multi-chain presence. The leading tokens are present on twenty or more chains, including all of the high-throughput L2s, Solana, Tron, and increasingly Bitcoin via wrapped representations. A new entrant on a single chain is structurally cheaper to ignore. Building multi-chain presence requires bridge audits, native deployments, and integrations with every wallet, costs that fall heavily on year-one issuers and have already been amortised by the leaders.
• Integration depth. The largest tokens are wired into card networks, payroll providers, payment aggregators, remittance corridors, custodians, and treasury platforms. Each integration is a bilateral agreement with its own compliance review and legal opinion. The fixed cost of those integrations is paid once by the leaders and would need to be paid again, in full, by any challenger.

Six-issuer comparative card view.

The six majors carry the substantial majority of supply that institutional buyers actively diligence. Coinbase is included because it is the largest commercial counterparty inside the USDC franchise even though it is not a fiat stablecoin issuer.

NYSE: CRCL · Founded 2013

Circle, USDC, EURC

$79.0B USDC supply; €447M EURC (#1 EUR stablecoin)⁴

Reserves: 80%+ in BlackRock-managed Circle Reserve Fund (USDXX) at BNY Mellon; balance cash at BNY Mellon, Cross River, Customers Bank.

Reg: NYSE-listed parent; state MTLs; MiCA EMI; GENIUS Act permitted issuer.

Key risk: Coinbase concentration (50% off-platform residual / 100% on-platform); SVB-style banking partner risk (USDC depegged to $0.87 Mar 11 2023).

Private · El Salvador · Founded 2014

Tether, USDT, USAT, XAUT

$189.6B USDT in circulation (mid-May 2026); $193B reserves; $10B+ 2025 net profit¹

Reserves: ~82% T-bills/repos ($141B+ Treasury exposure); ~9% gold ($17B); ~4% Bitcoin ($8B); secured loans 5%; cash 5%.

Reg: Offshore; USAT track via Anchorage Digital Bank under OCC; not MiCA-compliant; delisted on major EU venues since mid-2025.

Key risk: Reserve auditability (Deloitte engagement for USAT first Big-4 audit²³); BTC and gold mark-to-market volatility (~8% of reserve).

Private · Series D $2.4B (Apr 2021) · NYDFS trust

Paxos, USDP, PYUSD, USDG, PAXG

PYUSD $3.50B; USDG $2.79B; USDP small; BUSD wound down Feb 2023 by NYDFS²⁸

Reserves: T-bills, overnight repos, demand deposits at insured depository institutions in segregated, bankruptcy-remote accounts; NYDFS-supervised monthly attestations.

Reg: NYDFS trust + OCC oversight; MiCA via Paxos Issuance Europe; MAS via Paxos Digital Singapore. Broadest multi-jurisdictional footprint of any pure-play issuer.

Key risk: Partner brand dependency (PayPal for PYUSD); BUSD-style regulatory event risk; Anchorage stepping back from USDG²⁴.

Private · Federal trust bank · Charter Jan 2021

Anchorage Digital, USDtb, USAT

Only federally chartered crypto bank in U.S.; Tether's $100M investment Feb 5 2026²².

USDtb reserves: Predominantly BlackRock BUIDL tokenized U.S. Treasury fund; sole bank issuer; USAT reserves custodied at Cantor Fitzgerald.

Reg: OCC federal trust charter²⁵; permitted issuer pathway under GENIUS Act. Bo Hines (ex-White House Crypto Council) CEO of Tether USAT.

Key risk: Federal charter durability (2022 OCC consent order precedent); BUIDL single-vehicle concentration; partner risk on USAT.

Nasdaq: GEMI (Sep 2025) · NYDFS trust

Gemini, GUSD

$40-48M GUSD in circulation, down from peak ~$470M in 2022. Bluechip A rating (SMIDGE)⁵⁴.

Reserves: Cash, U.S. T-bills, government MMF shares; monthly BPM LLP attestation; NYDFS supervision.

Reg: NYDFS trust charter; state MTLs; SEC Earn case settled Feb 2024 ($1.1B returned via Genesis); broader 2024 case dropped under current admin.

Key risk: Scale (~0.05% of USDT); distribution contained to Gemini; product not the center of company strategy.

Nasdaq: COIN · Direct listing Apr 2021

Coinbase, USDC distribution + Base

Not a fiat stablecoin issuer. 50% off-platform residual on USDC reserves; 100% on USDC held in Coinbase products²¹.

Q1 2026: Stablecoin revenue $305M (largest S&S component); USDC balance at Coinbase ~$19B; S&S 44% of total revenue; ~25% of all USDC in custody⁸.

Strategic: Operates Base (Ethereum L2). May 13 2026: official USDC treasury deployer on Hyperliquid as Aligned Quote Asset (AQA), displacing USDH³⁷.

Key risk: Circle dependency on the residual share; USDC custody concentration; competitive pressure from JPMD and other bank stablecoins.

The Mid-Cap Tier. $0.5B-$5B issuers, chain footprint and archetype.

Behind the top-five concentration curve sits a $20B+ aggregate mid-cap layer that neither the daily press coverage nor most issuer reports map cleanly. Sixteen assets, sized below. The RWA flag marks tokenized-fund and yield-bearing instruments where price is not pegged to par.

World Liberty Financial USD

USD1

$4.61B

Fiat-backed non-bank · WLFI-aligned, multi-chain launch.

EthereumBSCSolanaAptosTronMonad

Dai

DAI

$4.60B

Crypto-collateralised · Legacy MakerDAO, continues alongside USDS.

EthereumPolygonFantomBSCArbitrumOP Mainnet

Ethena USDe

USDe

$4.35B

Synthetic / delta-neutral · Funding-rate yield. USDtb is the regulated counterpart.

EthereumSolanaMegaETHBSCTONPlasma

PayPal USD

PYUSD

$3.50B

Fiat-backed non-bank · Paxos for PayPal. Full deep-dive in Part V.

EthereumSolanaArbitrumSeiPolygonFlow

BlackRock USD RWA

BUIDL

$3.17B

Tokenized U.S. Treasury fund · Principal reserve backing of USDtb.

EthereumSolanaAptosBSCAvalancheOP Mainnet

Circle USYC RWA

USYC

$2.97B

Tokenized yield fund (price ~$1.13) · Hashnote, acquired by Circle.

BSCEthereum

Global Dollar

USDG

$2.79B

Fiat-backed non-bank · Paxos Digital Singapore under MAS. GDN consortium.

X LayerSolanaEthereumInkHyperliquid L1

Ondo US Dollar Yield RWA

USDY

$2.12B

Tokenized short-duration U.S. Treasury (price ~$1.12) · Yield-bearing.

EthereumPlumeSeiSolanaStellarMantle

Falcon USD

USDf

$1.62B

Yield-bearing synthetic · Falcon Finance.

EthereumBSC

Ripple USD

RLUSD

$1.60B

Fiat-backed non-bank · Ripple. NYDFS-supervised.

EthereumXRPL

USDD

USDD

$1.53B

Crypto-collateralised + yield · Tron ecosystem.

TronEthereumBSCBittorrent

Ethena USDtb

USDTB

$1.13B

Fiat-backed non-bank · Anchorage-issued, BUIDL-backed. GENIUS-compliant counterpart to USDe.

EthereumSolana

United Stables

U

$1.06B

Fiat-backed non-bank · Multi-chain new entrant.

BSCTronEthereum

GHO

GHO

$583M

Crypto-collateralised · Aave-issued.

Ethereum

Usual USD

USD0

$555M

Tokenized-RWA-backed (BUIDL, USYC) · DeFi-native distribution.

EthereumArbitrum

YLDS RWA

YLDS

$503M

Tokenized yield instrument · Figure-issued.

SolanaProvenanceEthereumStellar

What lives below the duopoly. The 17% in three structural slices.

USDT and USDC together hold ~83% of total stablecoin supply (USDT alone at ~59%). Everything else fights over the remaining ~17%, but that remainder is not homogenous. It splits cleanly into three structural categories with different issuer types, different regulatory frames, and very different growth trajectories. The breakdown below uses the mid-cap data plus USDS, the largest token outside the top two.

The asymmetry is the real signal.

The asymmetry is the load-bearing fact of the segment in 2026. A regulated EUR rail layer exists. The instruments are MiCA-compliant. The volumes are not there. Two reasons compound: domestic euro payment rails (SEPA, instant SEPA, TIPS) are already cheap and fast, so the substitution case for a EUR stablecoin is weaker than for a USD stablecoin in a low-trust currency corridor; and the MiCA reserve rules, particularly the cap on bank counterparty exposure and the 60% central-bank cash floor proposed for the largest EMTs, compress issuer economics relative to GENIUS-Act-licensed competitors. The next 12-24 months will tell whether EUR supply can cross the $5B aggregate threshold; until it does, the global stablecoin economy is a USD economy with a thin EUR appendix.

The EUR stablecoin layer. Five issuers, ~$650M combined.

The MiCA-authorised EUR stablecoin segment is small but legible. Five named tokens carry essentially all of the supply, each with a different issuer model and a different distribution thesis. Circle's EURC is the clear leader by an order of magnitude; the rest is fragmented across bank-issued (EURCV from SocGen FORGE), a stablecoin-tech pure-play (EURI from Eurite), an e-money institution (EURE from Monerium), and an exchange-aligned MiCA EMT (EURQ from Quantoz). All five trade slightly above par at the moment of this snapshot because each interacts with EUR rails at a slight premium to the spot reference rate, and the unit price is reported in dollar equivalent.

Circle EURC

EURC

€447M

MiCA EMT · Circle Internet Group · the dominant EUR EMT, multi-chain native deployment.

EthereumSolanaBaseAvalancheStellarSonic

EUR CoinVertible

EURCV

€91M

MiCA EMT · SocGen FORGE · bank-issued, integrated with Canton Network for institutional settlement.

EthereumSolana

Eurite

EURI

€59M

MiCA EMT · Eurite SA · stablecoin-tech pure-play, paired with Binance for distribution.

EthereumBSC

Monerium EUR emoney

EURE

€31M

E-money institution · Monerium · oldest EUR token in production, SEPA on/off-ramp native.

GnosisEthereumArbitrumPolygonLineaNoble

Quantoz EURQ

EURQ

€20M

MiCA EMT · Quantoz · exchange-aligned (Kraken, Bitstamp), multi-chain including Algorand and XRPL.

EthereumAlgorandPolygonXRPL

The strategic question for the EUR segment is whether MiCA's reserve rules will hold or will be relaxed in 2026-2027 to allow more competitive economics. The proposed 60% central-bank cash floor for the largest EMTs caps the yield issuers can earn on reserves at the deposit-facility rate, which compresses the issuer margin versus a US payment-stablecoin issuer holding T-bills. Until that gap closes, EUR supply will remain a thin appendix to a USD-dominated tokenized-cash market.

The bank entrants turning the corner.

Through 2023 and 2024, bank-issued tokenized cash sat in the "interesting pilot" category, sandboxes, regulatory testbeds, intra-group settlement only. In 2025 that changed materially. A cluster of named programmes crossed from pilot into production, and 2026 is the first year in which institutional treasury teams are realistically able to choose a bank deposit token for cross-border, B2B, or intra-day funding flows.

Three observations matter. First, the bank entrants are not building a single shared rail; they are building separately, on different chains, with different governance and different counterparty access. The pre-2025 expectation that banks would converge on one consortium token has not happened. The 2026 reality is parallel ecosystems with selective interoperability. Second, the public-chain choice (Base for JPMD, Ethereum for EURCV) is no longer controversial. The compliance and operational risk of using a public L2 is now seen as manageable. Third, none of these programmes compete head-on with USDT or USDC for trader use cases, they compete with correspondent banking, repo, and intra-group cash management.

Why hundreds launch and almost none survive.

Across the last seven years roughly five hundred stablecoin or stablecoin-like instruments have been launched, announced, or piloted. About one in twelve is still active at meaningful supply. The mortality rate is not a function of bad technology, most of these instruments work mechanically exactly as designed. The mortality rate is a function of four failure modes that launch decks systematically underweight.

Distribution starvation is the dominant killer by a wide margin and the failure mode most absent from the typical stablecoin pitch deck. A new token can have the cleanest reserves in the market, audited monthly by a Big Four firm, MiCA-licensed, and still die quietly over twelve months as no exchange volunteers a listing and no wallet integrates a third dollar-pegged token when the first two work fine. Distribution in stablecoins is not a problem an issuer can hire its way out of, because the parties whose distribution matters have no incentive to onboard a token that does not already have volume. The asymmetry is brutal.

Tabbed issuer detail.

PYUSD Growth Trajectory.

PYUSD launched on Ethereum on August 7 2023. Solana came in May 2024, the YouTube U.S. creator payouts in December 2025, and Visa Direct/BVNK in January 2026. The result: +680% to +726% YoY growth by early 2026, with figures depending on snapshot date.

PYUSD Distribution Channel Architecture.

Stablecoin Insider classifies PYUSD's growth as distribution-led⁷. Three categories of distribution drove PYUSD's growth in 2025-2026: PayPal/Venmo native consumer access; programmatic rails and enterprise payments (Visa Direct via BVNK, Worldpay, Fiserv, Stripe); and creator-economy/platform-payout integrations (YouTube).

PYUSD vs USDC vs USDT , side by side.

Quality of growth, three growth-model groups

Source: Stablecoin Insider, May 2026⁷. PYUSD grew fastest in percentage terms; USDT and USDC grew fastest in absolute terms.

The yield-risk question

Two distinct regulatory threads bear on PYUSD's yield-rewards program: OCC proposed rulemaking implementing the GENIUS Act (raised questions in Mar 2026 about whether affiliate yield arrangements are permissible; comment period closed May 1 2026); and the proposed CLARITY Act with language that would limit yield on payment stablecoins. Coindesk Mar 1 2026 reporting¹⁷ indicates the yield-ban language is contested and may not survive markup, but the situation is fluid.

What "1:1 backed" actually means in 2026.

The phrase "1:1 backed" is now a regulatory term of art rather than a marketing line. Under the frameworks that hardened across 2024 and 2025, a payment stablecoin must be backed by reserves with a market value at least equal to the par value of the tokens outstanding, measured continuously, with a defined set of eligible assets, and disclosed at a frequency the supervisor sets.

"Market value" is no longer marked at issuance prices. The leaders now report holdings at end-of-day mark for short-dated Treasuries and a conservative mark for repo positions, with haircuts that anticipate the loss the reserve would take in a forced sale within twenty-four hours. This is closer to a money-market fund convention than to the historical stablecoin convention, and the move tightened the reported reserve cushion across the segment by 30-80 basis points compared to the pre-2024 reporting style.

"Eligible assets" is a narrower category than it was. The composition of the typical top-five reserve in 2026 is dominated by US Treasury bills with maturities under three months, overnight repurchase agreements collateralised by Treasuries, and balances at correspondent banks. Commercial paper, money-market fund holdings of riskier instruments, and any meaningful credit exposure have been written out of the eligible list, partly by regulation and partly by issuer choice in anticipation of regulation. The reserve looks more like a sovereign cash position than a treasury yield product.

"Continuously" is not literal but it is close. The supervisory expectation is that the issuer's internal systems mark reserves at least daily and can produce a real-time view on supervisor request. Public disclosure is monthly for the largest issuers and quarterly for smaller ones; the gap between supervisory frequency and public frequency has compressed steadily since 2023 and is expected to disappear for the top tier by 2027.

Reserve composition, line by line.

The typical fiat-backed stablecoin reserve in 2026 looks like the chart below. The shares are an average across the top-five issuers; individual issuers vary by 5-10 percentage points around these midpoints.

Three observations from this composition. First, the reserve is effectively a sovereign-cash portfolio. Roughly 90% of the typical reserve has direct or indirect Treasury exposure, and the credit risk on the remaining 10% is concentrated in a handful of G-SIB bank counterparties. The risk profile of a top-five stablecoin reserve in 2026 is closer to that of a Treasury money-market fund than to that of any pre-2023 stablecoin reserve. Second, the yield on this composition, at current Fed Funds, is approximately 4.3% gross, of which the issuer typically retains the full amount as float income (most tokens still do not pay yield to holders). The economics of running a top-five stablecoin are, at scale, extraordinarily favourable: a $50B float at 4% gross yields $2B in annual revenue against operating costs of low hundreds of millions. Third, the concentration of custody is high. Three custodians hold the bulk of Treasury reserves for the top five tokens combined, and three correspondent banks hold the bulk of the operating cash.

Redemption windows, and what breaks them.

The most important property of a stablecoin is the redemption guarantee. The promise that one token, returned to the issuer, will be exchanged for one dollar, within a defined window, is what gives the token its peg. Every break in that promise, every paused redemption, every par-discount, every "temporary" gate, is an event the market remembers and the issuer cannot un-remember.

The redemption windows currently published by the top issuers fall into three patterns. The first pattern, used by the largest fiat-backed non-bank issuers, is a same-day redemption window for whitelisted institutional counterparties above a minimum size (typically $100k-$1M), with longer windows for retail-routed redemptions. The second pattern, used by some MiCA-licensed issuers, is a defined T+1 or T+2 redemption window for all holders, with a published holiday calendar. The third pattern, used by some smaller issuers, is a discretionary window, redemption "as soon as practicable", which the market increasingly discounts.

Three failure modes break redemption windows. The first is reserve liquidity failure: the reserve exists, has the right composition, but cannot be liquidated fast enough or at par to meet the redemption pressure within the published window. T-bills are not infinitely liquid; in stressed conditions the issuer can take 5-30 basis points of mark-to-market loss to liquidate at the speed redemption requires. The second is operational failure: the redemption infrastructure (application UI, KYC review, wire instructions, banking partner processing time) gets overwhelmed. The reserve is fine; the plumbing is not. The third is regulatory or counterparty failure: a banking partner closes, freezes, or restricts the issuer's operating accounts, and the issuer cannot move the cash to meet redemption even if it is legally theirs.

The 2023 USDC depegging episode was a textbook case of the third failure mode. The reserve composition was fine. The operational redemption pipeline was fine. A banking partner became unavailable over a weekend, and the market priced in the possibility that the issuer could not access its own cash. The token recovered fully within seventy-two hours, but the episode shaped reserve design across the segment for the following twenty-four months. Every top-five issuer now operates with at least three independent banking partners.

Capital buffers and prudential cushions.

A stablecoin reserve at exactly 1:1 with circulating supply has no cushion against the small losses that arise even in normal markets, fractional spreads on T-bill sales, operational debits, fee leakage, accounting reconciliation gaps. Every serious issuer therefore holds a cushion above 100% of circulating supply, funded either by issuer equity or by retained earnings on the reserve float.

Across the top five, the disclosed cushion in 2026 ranges from approximately 0.5% to approximately 4% of circulating supply. The lower end is sufficient to absorb ordinary mark-to-market noise on a daily basis. The upper end is sufficient to absorb a stress event with several percentage points of forced-sale loss on the reserve without touching the par value of outstanding tokens. The MiCA technical standards require an "own funds" requirement of at least 2% of the value of reserve assets for asset-referenced tokens, and similar minimums have been proposed in the US framework.

A second buffer that has emerged in the past two years is the redemption reserve: a separate operating cash position, typically 1-3% of circulating supply, held at the most liquid correspondent banks, available for redemption without any need to liquidate Treasuries or unwind repo. The redemption reserve is what allowed the largest issuers to meet stressed redemption flow in 2024 episodes without any visible mark-to-market drag on the published reserve composition. It is now a competitive minimum.

Five stress scenarios we model for clients.

In our advisory work, the five stress scenarios below are the load-test cases an issuer should be able to survive without breaking redemption at par. If the answer to any of these is "we have not modeled this," the answer to the scenario is "this issuer is one stress event away from a problem."

• Scenario A, banking-partner outage. A correspondent bank holding 5-15% of operating cash becomes unavailable for 72 hours. Can redemption continue at the published window using only the remaining partners and the redemption operating reserve?
• Scenario B, 10% redemption wave in 48 hours. A market event prompts redemption of 10% of circulating supply over two business days. Can the reserve be liquidated at par, in the required window, without forcing a mark-to-market hit that consumes the equity cushion?
• Scenario C, major CEX halts withdrawals. A large exchange suspends USD-onramps to USDC/USDT. APs cannot arbitrage secondary-market price deviations back to NAV. Can the issuer's direct mint/redeem channel honor at par fast enough?
• Scenario D, treasury rate shock or repo freeze. A 100-basis-point parallel shift in the short-dated Treasury curve over a single day, or temporary disruption to the Treasury repo market. What is the mark-to-market hit on the reserve, and does it consume the equity cushion?
• Scenario E, repo counterparty insolvency. A repo counterparty representing 20% of the reverse-repo book becomes insolvent. The collateral is delivered; what is the operational delay in liquidating it, and does that delay break the redemption window?

The top-three issuers can survive all five scenarios with public disclosure that allows external analysts to verify the survival. The next tier of issuers can survive A, B, and D but have less robust answers to C and E. The long tail beyond the top ten cannot credibly answer any of the five. This is not a moral judgment about the long tail, it is a structural observation that the discipline required to survive these scenarios is expensive, and the float economics of a sub-billion token do not pay for the discipline.

Regulatory Environment & Timeline.

Regulation is the single most decisive factor in which stablecoin issuers scale. Three jurisdictions matter most: United States, European Union, and Singapore, with Hong Kong, UK and UAE as fast followers.

Regulatory standing by issuer

United States. GENIUS Act, NYDFS, OCC.

The Guiding and Establishing National Innovation for U.S. Stablecoins Act (S.1582) is the first federal regulatory regime for payment stablecoins. Senate passed 68-30 on June 17 2025; House passed July 17 2025; Presidential signature July 18 2025. Effective date: the earlier of 18 months after enactment or 120 days after primary federal regulators issue final implementing rules¹²¹³⁵².

Permitted issuer categories

• Federally chartered banks (national commercial banks; national trust banks like Anchorage).
• OCC-supervised non-bank issuers above the size threshold (~$10B AUM line).
• Certain state-chartered entities under defined parameters.

Reserve composition rules

Cash, demand deposits at insured depository institutions, U.S. Treasury bills (~93 days or less), overnight reverse repos collateralized by Treasuries, and government money market fund shares. Commercial paper, corporate bonds, secured loans, gold and crypto are prohibited. This is a meaningful constraint on Tether specifically, whose current $193B reserve composition includes $17B gold, $8B Bitcoin, and ~5% secured loans, all of which would be impermissible under GENIUS Act for any U.S. payment-stablecoin issuer. USAT's path through Anchorage Digital Bank addresses this by using a Cantor-Fitzgerald-custodied, T-bill-only reserve segregated from USDT's general reserve.

The CLARITY Act and yield question (open as of May 2026)

The proposed CLARITY Act, separate from the GENIUS Act, contains language that would limit yield on payment stablecoins. Coindesk Mar 1 2026 reporting¹⁷ indicates the OCC's proposed rulemaking under the GENIUS Act does not in fact ban yield rewards, with the OCC comment period closing May 1 2026. Industry reporting suggests the yield-ban language is contested and may not survive markup. Readers should treat the yield risk as material but not as a baseline outcome.

European Union. MiCA in production.

Regulation (EU) 2023/1114¹⁴ entered force in 2023 with stablecoin provisions phased in from June 2024 and full EU-wide enforcement from July 1 2026. MiCA classifies stablecoins as either e-money tokens (single fiat reference) or asset-referenced tokens (basket reference). E-money tokens require an authorized EMI or credit institution, full reserve backing, custody at regulated EEA institutions, monthly attestations and redemption rights at par.

The structural problem for non-EU issuers: MiCA contains no third-country equivalence regime. Non-EU issuers must establish full legal presence inside the EU and obtain direct authorization. USDC, EURC and USDG are MiCA-compliant. USDT has been delisted on multiple EU exchanges since mid-2025; Banxa reports USDC volumes on its platform surged 400% in 2025 while USDT volumes dropped 60% in Europe, almost entirely driven by MiCA-related delistings⁷. From March 2026, e-money token custody in the EU may require both MiCA and PSD2 licenses.

The most visible MiCA outcome in 2026 is the emergence of EUR-denominated EMTs as a credible asset class. Pre-MiCA, EUR stablecoins existed but had thin liquidity and limited institutional adoption. Post-MiCA, several MiCA-authorised EUR tokens have meaningful circulating supply, with the largest in the low-single-digit billions of euros equivalent, small next to USDT and USDC but growing fastest in the segment. EURCV (SocGen FORGE), EURC (Circle's EUR token), EURI, and EURS each occupy distinct niches in the European institutional and retail market.

Asia. Three regimes, three logics.

Asia is the region where the regulatory perimeter is least uniform, and where the differences between regimes are most consequential for operators. The three lines that matter most are Hong Kong, Singapore, and Japan; each has chosen a different starting point and built outwards from that choice.

Hong Kong. HK Stablecoin Ordinance enacted August 2025¹⁵; licensing rules went live in early 2026; first licenses expected mid-2026. The HKMA stablecoin licensing regime is the most ambitious in Asia in scope, it covers issuance, custody, and intermediation, and the most conservative in approval pace. The result is a clean regulatory perimeter with very few authorised issuers, a pattern that the HKMA appears to consider a feature rather than a bug.

Singapore. MAS launched the Single-Currency Stablecoin framework in 2023¹⁶. The broader Digital Token Service Provider regime became operational June 2025. USDG is issued by Paxos Digital Singapore Pte. Ltd. under the MAS framework as the institutional vehicle for the Global Dollar Network. The Singapore framework provides analogous bankruptcy-remoteness protections to NYDFS trust supervision, but the case law is thinner.

Japan. Japan was an early mover, defining electronic payment instruments under the revised Payment Services Act with effect from 2023. The Japanese framework constrains stablecoin issuance to licensed banks, trust companies, and money transfer agents, and requires JPY-pegged stablecoins to be fully backed by trust-segregated cash. The Japanese approach has produced a small but operationally clean JPY stablecoin market, with bank-issued tokens dominant.

United Kingdom. The FCA framework and the contested 40% rule.

The UK chose a deliberate, late-but-coherent path. The FCA and PRA framework for fiat-backed stablecoins came into force in stages across 2024-2025, with a clear distinction between systemic and non-systemic issuers and a defined supervisory perimeter that aligns broadly with EU and US standards. The UK is intentionally designing for compatibility with the largest existing programmes rather than for the creation of a UK-native stablecoin segment.

The contested 40% central-bank-deposit rule. The UK proposal that 40% of backing assets be held as unremunerated central bank deposits is contested and would create asymmetry with the GENIUS Act's more flexible reserve rules⁵¹. If implemented, the 40% rule would materially compress issuer economics for UK-licensed stablecoins versus GENIUS Act-licensed competitors. The UK final framework is expected late 2026 or 2027.

The Bank of England's Digital Pound work, a CBDC programme rather than a stablecoin programme, has progressed in parallel but at a pace that means a retail CBDC is not expected before 2027 at the earliest. The 2026 UK position is one in which private stablecoins serve the institutional and B2B segment under FCA authorisation, and the central bank's CBDC work is treated as a separate stream that does not, in the medium term, displace the private offering.

UAE and Brazil

UAE Payment Token Regulation in force from 2024. Brazil's BCB Instruction No. 701/2026 treats stablecoin transactions as foreign exchange operations and requires VASP authorisation.

Cross-border traps and the half-finished perimeter.

The regulatory perimeter has hardened inside each major jurisdiction; the perimeter between jurisdictions remains soft and is where the most consequential operational mistakes happen. We see four cross-border traps repeatedly in our advisory work.

• The "we are EU-authorised so we can serve EU users via a non-EU wallet" assumption. Not true. The wallet routing matters. A MiCA-authorised token routed through a non-EU venue to an EU retail user can still create regulatory exposure for both the issuer and the venue.
• The "we are US-onshore so we can serve EU institutions" assumption. Selectively true, depending on the institution's domicile and the route by which the EU institution holds the token. The reverse-solicitation logic that applied in pre-MiCA Europe no longer applies cleanly.
• The "we use an offshore issuer so we are outside the perimeter" assumption. Not true. The major jurisdictions have all adopted some form of extraterritorial reach for offshore issuers serving their domestic users. The supervisor's reach is determined by the user base, not by the issuer's flag.
• The "tokenized deposit token is just a stablecoin with a bank label" assumption. Not true, and dangerously misleading. Bank deposit tokens are regulated under banking law, with different reserve requirements, different insolvency treatment, and different supervisory regime. This is the subject of Part VIII.

The legal form, side by side.

The headline distinction is a legal one. A stablecoin, in every major jurisdiction's regulatory framework, represents a claim against a reserve held by a non-bank or bank issuer for the purpose of issuance. A tokenized deposit, by contrast, represents a claim against a commercial bank, specifically, it is a token-form representation of a bank deposit liability that already exists on the bank's balance sheet. The token is a new container for an old liability. The stablecoin is a new liability backed by a new pool of assets. A tokenized-fund RWA is something different again: a tokenized share of a regulated investment fund, where the holder is a fund shareholder.

The three instruments look identical at the user-interface layer, a token in a wallet with a value pegged or tracked. They are very different instruments underneath. The single most consequential difference is the insolvency treatment: a stablecoin holder is paid from a segregated reserve pool that ideally is sufficient to make all holders whole at par; a tokenized deposit holder is a deposit creditor of a bank, with deposit insurance up to the scheme limit and the risk of bank insolvency above it; a tokenized-fund RWA holder is a fund shareholder, with the protection of fund-segregation rules. The differences cascade from those roots.

Why corporate treasuries pick deposit tokens.

The fastest-growing use case for tokenized deposits in 2026 is corporate treasury. The reasons are operational, not ideological.

• Existing banking relationship. The corporate already has a deposit account at the issuing bank. The tokenized deposit is a token-form representation of that same relationship, with no new counterparty risk introduced.
• Accounting and tax simplicity. A tokenized deposit is, for accounting purposes, a deposit. Existing treasury policies, internal control frameworks, and audit treatments apply unchanged. A stablecoin requires accounting policy work that depends on jurisdiction.
• Yield. Most stablecoins do not pay holders yield. A bank deposit, including a tokenized deposit, pays the prevailing deposit rate. Across a multi-billion-dollar corporate treasury position, the foregone yield on a stablecoin equates to tens of millions of dollars per year.
• Deposit insurance and rank. A tokenized deposit is insured up to the scheme limit and ranks as a deposit creditor above the bank's general unsecured creditors. A stablecoin is, in most jurisdictions, not deposit-insured.
• Regulatory clarity for the corporate. A corporate treasurer holding tokenized deposits is doing what their treasury policy already permits. Holding stablecoins may require policy approval, board notification, and ongoing reporting. The friction is asymmetric.

Why retail and crypto-native users stay with stablecoins.

The corporate-treasury logic reverses when the user is a retail wallet holder or a crypto-native counterparty. The decision matrix changes on most of the same dimensions, but the signs flip.

• Existing banking relationship. A retail crypto user typically does not have a deposit account at the bank issuing the deposit token. The stablecoin is available in their existing wallet with zero onboarding.
• Permissioning. Tokenized deposits are typically transferable only between KYC'd parties at the issuing bank. The stablecoin is bearer-like and transfers without the receiver onboarding.
• Cross-chain availability. Stablecoins are present on twenty or more chains. Tokenized deposits are typically present on one or two.
• Use cases. A user wanting to participate in DeFi, hold a stable asset across exchanges, or settle a peer-to-peer payment outside the banking system has a use case the deposit token cannot serve.

The cumulative effect is the mirror image of the corporate case: for retail and crypto-native users, the stablecoin is the lower-friction choice in almost every cell, and the deposit token wins only where the user is willing to onboard to a specific bank in exchange for the deposit's protections and yield.

The RWA-yield bridge. BUIDL, USYC, USDY, YLDS.

A third tier sits adjacent to both stablecoins and tokenized deposits: the tokenized money-market fund / tokenized U.S. Treasury fund. BUIDL ($3.17B, BlackRock), USYC ($2.97B, Circle/Hashnote), USDY ($2.12B, Ondo), YLDS ($503M, Figure) are the canonical examples. The instruments are not 1:1 USD-pegged. Their price tracks NAV, which means USYC trades around $1.13 and USDY around $1.12 as the underlying Treasury yield accrues. They behave like tokenized shares of a money-market fund, because that is what they are.

The strategic significance of this tier sits in two places. First, BUIDL is the principal reserve backing for Anchorage's USDtb, the regulated counterpart to Ethena's offshore USDe. The tokenized fund is now used as collateral inside a payment-stablecoin reserve, which is a meaningful architectural innovation: the BUIDL holder is a BlackRock fund shareholder, the USDtb holder is a stablecoin holder whose backing pool consists of those fund shares plus operating cash. The chain runs investment-fund regulation → bank-issuer charter → payment-stablecoin authorization. Second, the RWA-yield tier competes with stablecoins for "yield-bearing dollar exposure on chain". The institutional users who would otherwise hold an idle USDC balance now have the option of holding a yield-bearing tokenized fund instead.

Where the supply lives, chain by chain.

The mental model of stablecoin volume as primarily an Ethereum phenomenon has not been correct since 2022, and remains decisively wrong in 2026. The largest single chain by stablecoin volume is Tron, by a margin that surprises operators who have not measured it recently.

Card networks integrating stablecoin settlement.

The single most-watched payments development of 2024-2025 was the card networks moving from pilot to production on stablecoin settlement. Visa, Mastercard, and the major card-issuing fintechs now support stablecoin settlement to issuing banks and to selected acquirer flows. The mechanics are not exotic, the card transaction settles in stablecoin between issuer and acquirer, with the consumer-facing leg unchanged, but the implications are large.

For the issuer, the value is settlement speed and FX. Daily settlement, which used to be a T+2 affair across borders, becomes near-real-time. Cross-border FX, which used to require correspondent banking with embedded markups, can settle in a stablecoin pair with the FX leg priced explicitly. For the acquirer, the value is similar, faster settlement, fewer intermediaries, narrower FX spread. For the cardholder, nothing changes.

The volume on these rails is still small relative to the total card-network throughput, a low single-digit-percent share of cross-border card settlement, growing fast. The 2026 trajectory suggests this share will be 10-15% by end-2027 if the current adoption rate continues, and the inflection driver is not technology (which is solved) but bank-side adoption: the card-issuing banks signing on to use stablecoin settlement as the default rather than the optional path.

Remittance corridors. The quiet displacement.

Remittance is the use case where stablecoins have already won the most clearly, and where the win is the most invisible to mainstream coverage. Traditional remittance corridors, particularly the US-Latin America, Gulf-South Asia, and intra-Africa corridors, now move a meaningful share of their volume on stablecoin rails, typically with a fiat on-ramp and off-ramp at each end and stablecoin (predominantly USDT, often on Tron) for the cross-border leg in the middle.

The economics make the substitution unavoidable. A traditional remittance corridor charges 5-8% in fees and FX spread combined. A stablecoin-routed corridor with regulated on/off-ramps charges 1-2.5%. A direct peer-to-peer stablecoin transfer with informal on/off-ramps can come in under 1% combined. The cost differential is not marginal; it is a 4-7x spread over the incumbent remittance providers.

The corridors where the stablecoin share is highest are the ones with the worst incumbent service and the most established crypto on-ramp infrastructure. Turkey, Argentina, Nigeria, and Vietnam are the four corridors we observe most consistently in our advisory work, with stablecoin-routed flow representing 20-40% of estimated cross-border consumer flow in each.

The Western Union response: USDPT. Western Union announced USDPT (Stable by Western Union) for launch on Solana via Anchorage Digital Bank in 2026, planned for 40+ countries⁵⁸. The incumbent that was most exposed to stablecoin displacement is the incumbent that is now building the most institutionally regulated stablecoin-based remittance product. The pattern says: distribution beats arbitrage. Western Union's customer relationships and compliance posture are worth more than the rail-cost arbitrage that a pure-play crypto competitor would offer.

Securities settlement and the cash leg.

The use case for tokenised cash that has crossed most decisively into production in 2025-2026 is securities settlement. Tokenised bonds, tokenised money-market funds, tokenised repo, and tokenised collateral mobility have all moved from pilot to production at scale at major banks, and the cash leg of these transactions is increasingly tokenised, either as a bank deposit token or as a wholesale settlement token like Fnality or Partior.

The economic case is delivery-versus-payment without the multi-day delay. A bond traded on a tokenised platform with a tokenised cash leg can settle atomically, both sides exchanging simultaneously in a single transaction, rather than across two-to-three-business-day settlement cycles with embedded counterparty risk. Industry estimates of the savings at scale are in the tens of billions of dollars per year across the major securities markets.

The 2026 reality is that the largest issuance platforms, SocGen FORGE, HSBC Orion, Goldman DAP, JPM Onyx Digital Assets, BNY's tokenisation platform, the Singapore Project Guardian network, are all live with cash-leg tokenisation, and the volume of tokenised securities settled with tokenised cash is now in the low tens of billions of dollars per year and growing roughly 60-80% year-on-year.

Notably, the cash leg in these transactions is usually a bank deposit token or a wholesale settlement token, not a public stablecoin. The corporate and institutional counterparties prefer the bank-issued instrument for the same accounting, regulatory, and credit-line reasons that apply elsewhere (Part VIII). The public stablecoins have a different role in securities settlement, primarily as the cash leg for retail-accessible tokenised products and for cross-border settlement where the institutional rails do not yet reach.

The interoperability problem, and its current state.

The unresolved infrastructure problem of 2026 is interoperability, between chains, between rails, and between the public stablecoin world and the bank-issued tokenised world. The progress has been real and the gaps remain real.

Cross-chain interoperability for the same stablecoin has largely been solved. CCTP (Circle's Cross-Chain Transfer Protocol) for USDC, and similar mechanisms for USDT, allow native burn-and-mint transfers across chains without the bridging risk that defined the 2022-2023 era. A user can move USDC from Ethereum to Base to Arbitrum to Solana without a third-party bridge holding inventory. The bridge exploits that dominated 2022 are a smaller share of the risk surface in 2026, though they have not disappeared for the tokens that have not deployed native cross-chain mechanisms.

Cross-rail interoperability, moving a public stablecoin into a permissioned bank-deposit-token network and back out, is partially solved. Some banks operate atomic-swap mechanisms at their network perimeter, allowing a corporate counterparty to exchange a tokenised deposit for a public stablecoin at the bank's quoted rate. The state of the art is moving towards bidirectional, atomic interoperability, but no consensus standard has emerged.

Cross-stablecoin interoperability, atomic swaps between USDC and EURC, or between USDT and a regional fiat token, is solved at the DEX layer for public chains and in development for permissioned environments. The institutional users of these flows route them through regulated FX intermediaries, and the unregulated DEX route remains for the retail and crypto-native segments.

10.1 Reserve Duration and Liquidity Profile.

Stablecoins are maturity-transformation businesses. The issuer takes demand deposits (the stablecoin is a redeemable-on-demand obligation) and invests the proceeds in assets that, while short-duration, are not all overnight. The spread between the redemption-on-demand promise and the realized maturity of reserves is the principal mechanism by which a peg breaks under stress.

10.2 Redemption Stress Modeling, Five Scenarios.

Under a defined adverse scenario, can the issuer honor stablecoin redemption at par within an operational time window, and at what haircut on the marginal sale of reserves? Building on the prudential stress framework in Part VI, this section maps each scenario to specific issuer mitigants.

10.3 Treasury Market Dependency Risk.

Most major fiat-backed stablecoins now depend structurally on short-duration U.S. Treasury bills, the Treasury repo market and government MMFs for the bulk of their reserves. Aggregate stablecoin-driven Treasury bill demand is in the $200-220B range as of Q1 2026.

Three concentration vectors

• Sovereign concentration. All major fiat-backed stablecoins denominated in USD and backed with U.S. Treasury claims. Principal-amount risk correlated with U.S. sovereign credit and dollar funding markets.
• Duration risk. WAM under 60-90 days limits mark-to-market sensitivity but does not eliminate it under severe rate moves.
• Monetary-policy dependency. Federal Reserve standing repo facility, discount window access for member-bank reserve custodians, and Treasury auction functioning are operational requirements.

10.4 Banking Concentration and Counterparty Mapping.

Bank deposit concentration was the proximate cause of USDC's March 2023 depeg and is the single most institutional-relevant operational risk for U.S.-regulated stablecoins. Each issuer has diversified its own banking partners since 2023; at the system level the diversification is limited because the universe of banks willing and able to bank a stablecoin issuer at scale is small.

10.5 OFAC, Sanctions and Compliance Controls.

Stablecoin issuers operate one of the most consequential compliance functions in private money: issuer-controlled freezing of token balances at specific blockchain addresses in response to law-enforcement requests or sanctions designations.

10.6 Smart-Contract and Upgradeability Risk.

Stablecoin tokens are smart-contract artifacts. The contract's architecture, admin keys, upgrade pathway and pause functions determine what the issuer can and cannot do unilaterally.

The tension: The contract features that enable issuer-controlled freezing for sanctions/law-enforcement compliance (10.5) are the same features that enable issuer-controlled freezing for any reason. USDC, PYUSD, USDP, GUSD, USDG and USDtb all operate under U.S. or U.S.-equivalent regulatory supervision that constrains discretionary use of admin powers. USDT's discretionary perimeter is less constrained by an equivalent supervisor.

10.7 Cross-Chain and Bridge Risk.

A "USDC" on one chain may not have identical redemption or legal properties as a "USDC" on another chain. This is one of the most consequential nuances institutional readers miss.

Chain finality differences

• Ethereum mainnet: ~12.8 minutes finalization (two epochs)
• Solana: Sub-second finality with weaker probabilistic guarantees
• Base and other Optimistic L2s: 7-day fraud-proof window for full settlement to Ethereum

Stablecoin redemption pipelines must be designed against the worst-case finality of the chain the redemption is initiated on.

10.8 Bankruptcy and Legal Priority Waterfall.

In an issuer insolvency, who has priority claim to the reserve assets, and through what legal process? One of the most important institutional diligence questions and one of the least well-explored in crypto research.

10.9 Authorized Participant and Redemption Concentration.

Stablecoin redemption mechanics are analogous to ETF authorized-participant structures: a small number of vetted counterparties have the right to mint and redeem at scale, and the rest of the market depends on those counterparties for arbitrage between secondary-market price and net asset value.

Redemption access by issuer

• USDC. Circle Mint for institutional clients under KYC. Coinbase is by far the largest single counterparty holding ~25% of all USDC. AP set diversified across major market makers.
• USDT. Direct mint/redeem through Tether for vetted KYC counterparties. AP set smaller and more opaque than USDC. Opacity is itself an institutional risk axis.
• PYUSD. Direct mint/redeem through Paxos for institutional clients; consumer redemption through PayPal and Venmo. AP set small relative to USDC and USDT.
• USDG. Paxos Digital Singapore + Global Dollar Network consortium members as APs. Consortium structure provides AP redundancy.
• USDtb and USAT. Anchorage Digital Bank for institutional clients. AP set being built.
• GUSD. Mint and redeem largely contained to the Gemini venue; thin AP set outside Gemini.

10.10 A Practical Risk Map for 2026 Operators.

Pulling Part X together, the risk map for an operator integrating stablecoins or tokenised deposits in 2026 is the following. The risks are listed in order of how often we see them mismanaged in our advisory work.

Six things we are confident about.

These six expectations carry the highest probability in our internal forecasting work. Each is grounded in a structural driver from the preceding parts and would require a substantial surprise to be falsified.

Four things we are watching closely.

These are the scenarios with high uncertainty and high consequence. They are the ones that, if they happen, will require us to rewrite material parts of this report.

• USDT's onshore-vs-offshore split. If Tether obtains a route to onshore US distribution under the federal framework, the implication is a meaningful expansion of its addressable market. If Tether does not, the geographic specialisation we forecast accelerates. The most-watched single regulatory event of the next 18 months.
• The fate of yield-bearing stablecoins. Several issuers have explored yield-bearing structures. Regulatory treatment of these instruments (as securities, as e-money plus yield distribution, as money-market fund tokens) is unresolved in most jurisdictions. A clear ruling, particularly in the US, would either open a large new category or close it definitively.
• The path of CBDC retail rollouts. The Digital Euro, Digital Pound, and digital RMB programmes are at various stages. None is expected to displace private stablecoins in the medium term, but a fast-track retail rollout in one major economy could compress the addressable market for private alternatives. Low probability, high consequence.
• The tokenisation of money-market funds. The tokenised MMF segment (BUIDL, USYC, USDY, YLDS) is structurally adjacent to the stablecoin segment and competes for similar use cases. The boundary will be redrawn over the next 18 months, and the question of how much "stablecoin volume" effectively becomes "tokenised MMF volume" is open.

The single statistic to watch.

If we could publish only one number quarterly going forward, it would be the share of stablecoin transaction volume that touches a regulated institutional rail at some point in the journey. Today that share is in the 10-20% range. By end-2027, on the trajectory in front of us, it should be in the 40-50% range. The slope of that line is the single most informative leading indicator for the segment.

It matters because of what it measures. The share captures everything: the regulatory perimeter hardening (more flows route through authorised entities), the bank-deposit-token growth (institutional users prefer regulated rails), the card-network adoption (regulated institutions in the routing path), the corporate-treasury adoption (institutional counterparties on both ends), and the securities-settlement growth (regulated dealers and custodians). A single number that aggregates all of this is the cleanest summary of whether the segment is integrating into the regulated financial system or remaining adjacent to it.

If the share rises faster than 5-8 percentage points per year, the integration is accelerating. If it stalls, it suggests one or more of the drivers in this report is not playing out as expected.

Spotlight, The Coinbase / Hyperliquid USDC Deal.

On May 13 2026 Coinbase announced it had become the official treasury deployer of USDC as an Aligned Quote Asset (AQA) on Hyperliquid³⁷. Hyperliquid had been working toward its own USDH stablecoin; the Coinbase deal displaces USDH in favor of USDC. This is the most consequential single distribution win for USDC in 2026 to date and demonstrates the commercial muscle Coinbase brings to the USDC franchise.

Mechanics of the deal

• Coinbase as treasury deployer. Coinbase deploys Hyperliquid USDC reserves and shares a majority of USDC reserve yield with Hyperliquid³⁷.
• USDH brand transition. Coinbase acquired rights to USDH brand assets. USDH remains redeemable for USDC or fiat during a migration period.
• AQA designation. USDC becomes the canonical settlement asset across Hyperliquid's HIP-1 through HIP-4 markets, displacing the prior USDH plans.
• Scale. Trade press³⁸³⁹ reports USDC on Hyperliquid had reached approximately $5B in total prior to the deal [REPORTED].

What the deal signals strategically

1. Coinbase is willing to share reserve yield down the stack to win distribution at scale. Prior to the Hyperliquid deal, the precedent for Coinbase sharing reserve yield with a downstream partner was unusual. The willingness to share a majority of USDC reserve yield with Hyperliquid reflects a recognition that distribution lock-in at the venue level is worth giving up margin.

2. USDC's role as quote currency is consolidating against USDT in U.S.-regulated venues. Hyperliquid is the largest decentralized perpetuals venue by volume. Designating USDC as AQA over USDT or USDH is a structural endorsement of USDC for institutional-grade DeFi.

3. The Circle-Coinbase joint franchise becomes even more inseparable. The 50% off-platform residual + 100% on-platform structure already meant Coinbase was the dominant commercial actor in USDC. Adding the Hyperliquid mechanism deepens the entanglement.

Spotlight, Tether Reserve Composition & 2025 Profit.

Tether's Q4 2025 attestation by BDO¹ is the single most-referenced primary document in this report. Total USDT circulation surpassed $186B at Dec 31 2025; total reserve assets climbed to nearly $193B; total direct and indirect U.S. Treasury exposure surpassed $141B; gold ~$17B; Bitcoin ~$8B; excess reserves $6.3B; net 2025 profit exceeded $10B.

What's not in the Q4 attestation

• Counterparty-level bank disclosure. Offshore banking partners that hold the cash portion are not named at counterparty level.
• Weighted Average Maturity (WAM). Precise WAM for the T-bill + repo portion is not in the public attestation at the detail an institutional buyer would request.
• Operational redemption capacity. Bilateral SLAs and intraday operational redemption capacity under stress are not publicly disclosed.
• Secured loan counterparty detail. The ~5% secured loan component is described in composition but not at counterparty level.

The Deloitte audit engagement (March 2026)

In March 2026 Coindesk reported that Tether had engaged Deloitte as its first Big Four auditor for USAT reserves and signaled progress toward a broader audit of USDT²³. A full GAAS audit of USDT reserves has been the most-asked-about disclosure item for a decade. The Deloitte engagement is the first concrete step toward that audit.

The 2021 CFTC finding in context

The CFTC Order of October 15, 2021 found that Tether held sufficient fiat reserves to back USDT only approximately 27.6% of the days in a 26-month sample period 2016-2018³⁰. The Q4 2025 attestation shows full backing plus $6.3B excess reserves. The trajectory between these two data points is the central narrative for Tether's transparency progress: the CFTC finding is the floor; the current attestation is the present; the Deloitte audit is the next confirming step.

Spotlight, The Anchorage B2B Issuance Platform.

Anchorage Digital Bank's commercial model has shifted toward third-party stablecoin issuance under its OCC charter. The platform structure: partners bring brand, customer base and distribution; Anchorage provides federally regulated issuance, custody and reserve management. The B2B model captures issuance fees, reserve management fees and custody fees, not direct share of reserve interest. This differentiates from Circle's economic model where 50%+ of reserve residual flows out to Coinbase.

Five products on the Anchorage platform (May 2026)

The Tether investment, Feb 2026. Tether invested approximately $100M in Anchorage in February 2026²², adding to a cap table that already included Andreessen Horowitz, KKR, Goldman Sachs, Visa and BlackRock (total funding ~$487M across rounds). The Tether investment reinforces the strategic alignment of USAT through Anchorage.

The USDG stepping-back report, May 11 2026. In May 2026 Coindesk reported that Anchorage was stepping back from the USDG Global Dollar Network consortium²⁴. The reported reason is strategic prioritization of products on which Anchorage is the sole issuer (USDtb, USAT, USDPT, USDGO) rather than consortium-shared issuance like USDG. This reinforces the thesis that Anchorage's commercial model maximizes the value of the OCC charter when Anchorage is the bank of record.

Spotlight, The New Mid-Cap Entrants Reshaping the $1B-$5B Band.

The mid-cap matrix in Part III mapped the entire $0.5B-$5B layer. This spotlight zooms in on the three entrants we think will most affect competitive dynamics over the next 12-18 months: USD1, USDe, and BUIDL. Each represents a different archetype, each addresses a different audience, and each is competing for share in a part of the market the top-five did not previously contest.

USD1 (World Liberty Financial USD), $4.61B

USD1 reached $4.6B by May 2026 from a 2025 launch, the fastest non-PYUSD ascent in the regulated-issuer space and a top-five entrant within months. Six chains at launch (Ethereum, BSC, Solana, Aptos, Tron, Monad) signals a multi-chain-first product strategy rather than the EVM-first pattern of earlier entrants.

The political-brand context is the load-bearing variable. USD1 is issued by World Liberty Financial, the Trump-family-affiliated venture, and its scale-up overlapped with the first six months of the second Trump administration. The supply ramp is partially attributable to politically aligned distribution channels and partially to a credible technical push through BSC and Tron, two chains where USDT incumbency is heaviest. Untangling the two contributions is hard from public data alone, and any institutional buyer evaluating USD1 should treat the brand-political risk as a distinct diligence axis from the standard issuer-risk frame.

The strategic question is which lane USD1 occupies. If it positions as a USDT-on-BSC alternative for retail and emerging-market flows, it competes with USDT directly on chain economics. If it positions as a politically aligned alternative to USDC in the US-onshore institutional market, it competes for share Circle has not lost in 2026. The reserve composition disclosures and regulatory authorisation path will decide which of the two it becomes. Until those disclosures land at the depth GENIUS Act issuers will be held to, the supply ascent is the leading data point but not yet a verdict on the model.

USDe (Ethena), $4.35B

USDe is the largest synthetic / delta-neutral stablecoin in the market. The protocol takes long crypto positions hedged by perpetual shorts and captures the funding-rate spread as yield to holders. Its $4.35B supply makes it larger than PYUSD; its archetype puts it in a different regulatory category. USDe is technically a synthetic asset, not a payment stablecoin, and that distinction is the basis for the Anchorage-issued USDtb (the regulated counterpart). The strategic significance of USDe is not its size but the precedent it sets: that a synthetic dollar instrument can scale into the multi-billions on DeFi-native demand alone. If the regulatory perimeter eventually constrains synthetic stablecoins, USDe is the data point that the constraint is being drawn against.

BUIDL (BlackRock), $3.17B

BUIDL is the canonical tokenized U.S. Treasury fund and the principal reserve backing for USDtb. Its growth (reportedly +30%/month at certain points in 2025-2026⁷) reflects institutional demand for tokenized-Treasury exposure on chain. The strategic significance is the role-reversal it implies: BUIDL is both an end-instrument (used by institutional holders directly) and an upstream input to a stablecoin (used by Anchorage as USDtb's reserve). The chain of investment-fund regulation → bank-issuer charter → payment-stablecoin authorization is a new architectural pattern. Other tokenized funds (Franklin OnChain US Government Money Fund, USDY, USYC, YLDS) are following similar paths.

Case Studies, Three Events Every Buyer Should Know.

Three industry events from 2022-2026 illustrate the failure modes a stablecoin risk committee should pressure-test against. Reading them in sequence makes the diligence axes from Part X concrete.

1. The USDC SVB depeg, March 10-13 2023

Sequence of events: Silicon Valley Bank failed on Friday March 10, 2023, with Circle holding approximately $3.3B of USDC reserves in SVB deposits as part of its cash-portion banking partner network⁴⁵. Over the weekend, secondary-market USDC traded down to approximately $0.87, a roughly 13% deviation from par. On Monday March 13, the FDIC's systemic risk exception confirmed full access to SVB deposits, Circle's redemption pipeline through Cross River Bank and other partners resumed, and the USDC peg recovered.

2. NYDFS BUSD wind-down, February 13 2023

Sequence: NYDFS directed Paxos to stop new issuance of BUSD (Binance-branded USD stablecoin issued by Paxos) on February 13 2023²⁸. BUSD circulation at the time was approximately $16-20B at peak. Paxos honored redemptions and the BUSD circulation declined from ~$20B to under $50M over the following months.

What it teaches: A U.S. state supervisor (NYDFS) can act against a Paxos-issued stablecoin tied to a counterparty (Binance) without insolvency or fraud at Paxos itself. Regulatory event risk can compress a $20B product to functionally zero within months. This precedent informs how a buyer should weight the "regulatory event risk" axis for any white-label rail.

3. Tornado Cash OFAC freeze, August 8 2022

Sequence: OFAC designated the Tornado Cash mixer and added associated Ethereum addresses to the SDN list on August 8 2022²⁹. Circle (USDC) froze tokens at the designated addresses the same day. The freeze was operational within hours of the OFAC action.

What it teaches: Issuer-controlled freezes are operationally available within hours of an OFAC action. Tether has frozen over $2.9B of USDT to date, including a $344M Iran-linked freeze in April 2026²⁶²⁷.

Issuer Economics, Who Captures the Reserve Yield.

Stablecoin economics are fundamentally net interest income on reserves. At prevailing T-bill yields of 4-5%, the gross interest pool for each issuer scales linearly with reserve size. What matters strategically is how that pool is split between the issuer, partners (Coinbase, PayPal, Global Dollar Network consortium), and end-user rewards.

What to bet on. What to ignore.

For an operator allocating capital, attention, or product roadmap to this segment in 2026, the action implications of the report are concrete. Below is the short version of what we tell clients in initial engagements.

• Bet on the rail layer over the token layer. The token layer is settled at the top. The rail layer, routing logic, multi-instrument wallet support, deposit-token integrations, card-network connections, is unsettled and where the competitive advantage will accrue over the next twenty-four months.
• Bet on regional specialisation over global ambition. A token serving a specific corridor or use case the leaders do not is in a defensible niche. A token competing head-on with USDC for the US-onshore institutional market is fighting four compounding moats from day one.
• Bet on regulated rails over offshore arbitrage. The era of regulatory arbitrage is closing. Every major jurisdiction has drawn its perimeter, and the cost-benefit of building inside the perimeter has flipped relative to operating outside it.
• Bet on the RWA-yield bridge. BUIDL, USYC, USDY, YLDS are growing fastest, the regulatory perimeter is more permissive (investment-fund regulation rather than payment-stablecoin), and the institutional buyers want yield-bearing dollar exposure on chain. The boundary between this segment and stablecoins is where the next 18 months of supply migration will happen.
• Ignore the "next generation algorithmic stablecoin" pitches. The category is closed by the regulatory work of 2024-2025. Variants will continue to appear and continue to fail outside the perimeter; none will reach institutional scale.
• Ignore the "stablecoins will replace banks" pitches. They will not, in any horizon we can foresee. The two adjacent markets, public stablecoins and tokenised bank deposits, will coexist, with banks capturing the institutional and corporate-treasury share.
• Ignore the headline-grabbing CBDC announcements unless they include retail rollout dates inside the next 24 months. Wholesale CBDC pilots are interesting but slow; retail CBDC is the disruptive event, and the timelines for the major economies do not put it inside the 2026-2027 window.

References.

All inline numbered superscripts in this document resolve to the entries below. Sources are listed in order of first appearance. Tier 1 (primary), Tier 2 (aggregator), Tier 3 (trade press) tags retained where relevant.

1. Tether International, S.A. de C.V. Q4 2025 Attestation Report by BDO Italy (Dec 31 2025). Total USDT circulation, reserve composition, U.S. Treasury exposure, gold, Bitcoin, secured loans, excess reserves, 2025 net profit. [Tier 1]
2. CoinMarketCap. Stablecoin supply data and historical price-supply series, May 2026. [Tier 2]
3. CoinGecko. 2025 Annual Report. Supply growth and market structure data. [Tier 2]
4. CoinGecko Q1 2026 stablecoin rankings. Market cap, share, holder distribution. [Tier 2]
5. Copestake, Englander, Martinez Peria, Villegas-Bauer. Stablecoins and the Future of Payments: Evidence from Financial Markets. IMF Working Paper WP/26/52, March 2026. 58 pages. [Tier 1]
6. Tipranks. PYUSD distribution and Visa Direct / BVNK coverage, February 2026. [Tier 3]
7. Stablecoin Insider. Mapping the Top 100 Stablecoins and Their Future. The 2026 Stablecoin Issuance Landscape. May 2026. 61 pages. [Tier 2]
8. Coinbase Global, Inc. Q1 2026 earnings deck and 10-Q filings. Stablecoin revenue, USDC custody balance, S&S revenue mix. [Tier 1]
9. Boston Consulting Group and Allium Labs (Batra, Zevin, Mathur, Quitete, Bravo, Li). Stablecoin Payments: The Truth Behind the Numbers. White paper, January 2026. [Tier 1]
10. DefiLlama. On-chain stablecoin supply by chain, mid-April 2026 snapshot. [Tier 2]
11. Paxos / PayPal. PYUSD launch press release, August 7, 2023; PayPal/Paxos Solana expansion press, May 2024. [Tier 1]
12. U.S. Congress. Guiding and Establishing National Innovation for U.S. Stablecoins (GENIUS) Act, S.1582. Senate 68-30 (Jun 17 2025); House (Jul 17 2025); Presidential signature (Jul 18 2025). [Tier 1]
13. White House Fact Sheet on GENIUS Act, July 2025. [Tier 1]
14. European Union. Regulation (EU) 2023/1114 (MiCA). Full enforcement from July 1, 2026. [Tier 1]
15. Hong Kong Stablecoin Ordinance, enacted August 2025. [Tier 1]
16. Monetary Authority of Singapore. Single-Currency Stablecoin framework (2023); Digital Token Service Provider regime (operational June 2025). [Tier 1]
17. Coindesk. State of Crypto, March 1, 2026. OCC proposed rulemaking under the GENIUS Act and CLARITY Act yield question. [Tier 3]
18. Yahoo Finance. CRCL share-price page, mid-May 2026 (Circle Internet Group market cap reference). [Tier 3]
19. Eco.com. USDC Reserve Composition Explainer, 2026. Bank counterparty and reserve-fund detail. [Tier 3]
20. Crane Data. Government MMF rankings and USDXX coverage. [Tier 2]
21. Circle Internet Group. S-1 IPO Filing, April 2025. Coinbase residual revenue split (50% off-platform / 100% on-platform); 2024 financials. [Tier 1]
22. Coindesk. "Tether invests ~$100M in Anchorage Digital." February 5, 2026. [Tier 3]
23. Coindesk. "Tether taps Deloitte for first USAT reserve report." March 3, 2026. [Tier 3]
24. Coindesk. "Anchorage stepping back from USDG Global Dollar Network." May 11, 2026. [Tier 3]
25. Office of the Comptroller of the Currency (OCC). National trust-bank charter approval for Anchorage Digital Bank, N.A., January 2021. [Tier 1]
26. Tether.io transparency posts and law-enforcement cooperation pages (cumulative freeze totals, 275+ agencies, 59 jurisdictions). [Tier 1]
27. Coindesk. "Tether supports freeze of more than $344 million in USDT in coordination with OFAC and U.S. law enforcement." April 24, 2026. [Tier 3]
28. New York Department of Financial Services (NYDFS). Press release directing Paxos to halt BUSD issuance, February 13, 2023. [Tier 1]
29. U.S. Department of the Treasury, Office of Foreign Assets Control (OFAC). Tornado Cash designation and SDN list addition, August 8, 2022. [Tier 1]
30. U.S. Commodity Futures Trading Commission. Order against Tether Holdings et al., October 15, 2021 ($41M civil penalty; 27.6% backing-day finding for 26-month sample 2016-2018). [Tier 1]
31. Fortune. YouTube creator-payout coverage on PYUSD, December 11, 2025. [Tier 3]
32. Coindesk. YouTube U.S. creator payouts with PYUSD, December 12, 2025. [Tier 3]
33. Coindesk. "PayPal and USD.AI launch $1B PYUSD incentive program at 4.5%." December 18, 2025. [Tier 3]
34. PayPal. "Buy. Hold. Earn Rewards." Press release introducing 4% variable PYUSD rewards, April 23, 2025. [Tier 1]
35. Blockhead. "PYUSD now available in 70 markets." March 18, 2026. [Tier 3]
36. WalletConnect. State of Stablecoin and Crypto Payments 2026. 62 pages. Section 5: holder penetration; Cahill Gordon & Reindel UK 40% rule analysis. [Tier 1]
37. Coinbase. "Coinbase and Hyperliquid: Aligning Markets on Hyperliquid to USDC." Coinbase blog, May 13, 2026. [Tier 1]
38. The Block. Hyperliquid-Coinbase AQA coverage, May 14, 2026. [Tier 3]
39. FXStreet. Hyperliquid AQA reporting, May 14, 2026. [Tier 3]
40. Decrypt. Coverage of SEC probe into Coinbase-Circle USDC revenue split, August 2025. [Tier 3]
41. Cerutti, Firat, Hengge, Sagawa. Stablecoin Shocks. IMF Working Papers, 2026. [Tier 1]
42. Ahmed and Aldasoro. Stablecoins and safe asset prices. Bank for International Settlements, 2025. [Tier 1]
43. Ferrari Minesso and Siena. Private money and public debt: U.S. stablecoins and the global safe asset channel. European Central Bank, 2026. [Tier 1]
44. Snowberg, Wolfers, Zitzewitz. How Prediction Markets Can Save Event Studies. NBER Working Paper 16949, 2011. [Tier 1]
45. Circle public statement, March 11, 2023; Bloomberg, Financial Times and CNBC contemporaneous coverage of SVB failure and USDC depeg. [Tier 1]
46. Cointelegraph. Circle-Cross River banking partnership coverage. [Tier 3]
47. Paxos newsroom. Global Dollar Network / USDG consortium announcement, November 2024. [Tier 1]
48. Coindesk. USAT launch with Tether and Anchorage Digital Bank, January 27, 2026. [Tier 3]
49. Coindesk. Tether 2025 net-profit and reserve coverage, January 30, 2026. [Tier 3]
50. Coindesk. DeFi yield-compression coverage, April 7, 2026. [Tier 3]
51. Cahill Gordon & Reindel. UK proposed 40% central bank deposit reserve analysis, cited in WalletConnect 2026. [Tier 1]
52. Latham & Watkins, Paul Hastings, Sidley Austin, Stinson LLP. Published legal analyses of the GENIUS Act, 2025-2026. [Tier 1]
53. Adrian et al. Understanding Stablecoins. IMF Departmental Papers, 2025. [Tier 1]
54. Bluechip. SMIDGE-framework stablecoin ratings (Gemini GUSD A; comparative issuer-design scoring). [Tier 2]
55. Chainalysis. 2024 Global Crypto Adoption Index. [Tier 2]
56. Visa Onchain Analytics dashboard. [Tier 2]
57. Backed Finance. Press release, November 2024 (Morpho RWA / PYUSD integration). [Tier 1]
58. Western Union. Investor Relations press releases (USDPT announcements October 2025; 2026 Solana launch detail). [Tier 1]
59. JPMorgan Chase. Kinexys press releases (June 2025 first USD-denominated deposit tokens; November 2025 institutional availability on Base; January 2026 Canton Network announcement). [Tier 1]
60. Anchorage Digital Bank, N.A. USDtb Reserve Attestations; October 2025 / January 2026 USDtb press releases. [Tier 1]

Reliability flags used in document

• [REPORTED], Trade-press claim with corroborating publication but not primary-confirmed.
• [REPORTED, unverified], Trade-press claim without primary corroboration. Treat as directional.
• [EST], Author calculation or analyst estimate.
• [PROJ], Forward-looking projection.
• [VIEW], Opinion, thesis or interpretation by the document author.

Glossary of Key Terms.