Research/Stablecoins

Too Few Banks: The Concentration Risk in Stablecoin Reserve Custody

Most major stablecoins rely on a handful of banking partners for reserve custody. What happens if one of those banks fails or debanks an issuer.

bcMaoNov 22, 2026

The largest stablecoins in circulation collectively hold over $200 billion in reserves. Those reserves sit in bank accounts and custody arrangements at a surprisingly small number of financial institutions. When one of those banks fails, as three did in March 2023, the consequences ripple through the entire stablecoin ecosystem: pegs break, contagion spreads to unrelated tokens, and billions of dollars become temporarily inaccessible.

This article maps the banking relationships behind major stablecoins, examines the structural forces that concentrate reserves into so few institutions, and evaluates whether proposed solutions actually reduce the risk.

Who Banks the Stablecoins

Every fiat-backed stablecoin needs at least one bank to hold its cash reserves and at least one custodian for its Treasury securities. In practice, the web of relationships is narrower than most users realize. The table below maps the known banking and custody partners of the four largest dollar-denominated stablecoins as of late 2026.

StablecoinIssuerPrimary Banking / Custody PartnersReserve Composition
USDCCircleBNY Mellon (custody), BlackRock (Circle Reserve Fund manager), Cross River Bank (transaction banking)~84% Circle Reserve Fund (short-term Treasuries), ~16% cash at banking partners
USDTTetherCantor Fitzgerald (Treasury custody), Deltec Bank, Britannia Bank & Trust, Capital Union Bank, AnsbacherPrimarily U.S. Treasury bills, with cash at multiple Bahamas-based banks
PYUSDPaxos Trust CompanyPaxos (qualified custodian under NYDFS), undisclosed banking partners for cash depositsMostly overnight Treasury reverse repurchase agreements, remainder in cash deposits
USDBBraleRegulated banking partners (names not publicly disclosed), segregated custodyU.S. Treasury bills and cash equivalents in bankruptcy-remote accounts

Several patterns stand out. USDC concentrates the vast majority of its reserves with a single asset manager (BlackRock) and a single custodian (BNY Mellon). Tether relies on a cluster of small Bahamas-based banks plus Cantor Fitzgerald for Treasury custody. PYUSD funnels its reserves through Paxos, itself a trust company rather than a traditional bank. Each structure creates a different kind of counterparty risk, but all share the same underlying vulnerability: too few institutions bearing too much weight.

Case Study: The March 2023 Banking Crisis

The week of March 8 to 12, 2023, demonstrated what happens when stablecoin banking concentration meets a bank run. Three crypto-friendly banks collapsed in rapid succession: Silvergate Bank announced voluntary liquidation on March 8, Silicon Valley Bank (SVB) was seized by regulators on March 10, and Signature Bank was shut down on March 12.

The USDC Depeg

On March 10, Circle disclosed that approximately $3.3 billion of USDC reserves were held as uninsured deposits at SVB. That represented roughly 8% of the total $40 billion USDC reserve at the time. Within hours, USDC broke its dollar peg on secondary markets, falling as low as $0.88 according to reporting from The Block.

The depeg was not caused by a shortfall in Circle's total assets. It was caused by the uncertainty of whether Circle could access its funds at a failed bank, and by the resulting wave of redemption requests that Circle could not process over a weekend when banks were closed.

Contagion without exposure: DAI, MakerDAO's stablecoin, also lost its peg during the SVB weekend despite having no direct SVB deposits. Because DAI's Peg Stability Modules held large amounts of USDC as collateral, USDC's depeg mechanically dragged DAI down with it. A Federal Reserve research note later described stablecoins as "run-able liabilities" susceptible to self-reinforcing confidence crises.

Payment Rails Disrupted

The banking failures did not just threaten reserves. Silvergate's SEN (Silvergate Exchange Network) and Signature's Signet platform were the two primary real-time payment networks that let crypto firms move dollars 24/7. When both disappeared in the same week, stablecoin issuers and exchanges lost their ability to process fiat settlements outside of business hours. The dollar on-ramps and off-ramps that the entire stablecoin payment rail depended on simply stopped working.

Resolution

USDC's peg was restored after the FDIC invoked the systemic risk exception and guaranteed all deposits at SVB and Signature, including uninsured balances above the standard $250,000 limit. Circle moved its banking relationship to Cross River Bank and expanded its use of BNY Mellon for reserve custody. But the episode exposed a fragility that regulation had not addressed: stablecoin reserves are, by definition, too large for deposit insurance to cover, and the pool of banks willing to hold them is small.

Why So Few Banks

The concentration of stablecoin reserves in a handful of institutions is not an accident. It reflects structural constraints on both sides of the relationship.

Regulatory Pressure

Between 2022 and 2024, U.S. banking regulators conducted what critics have called "Operation Chokepoint 2.0": a campaign of informal pressure discouraging banks from serving crypto companies. A November 2025 House Financial Services Committee report documented at least 30 entities that lost banking access during this period. A June 2024 lawsuit led by Coinbase resulted in the release of letters showing regulators had asked financial institutions to "pause" crypto banking activities.

Even after President Trump signed an executive order in August 2025 formally ending debanking, many banks remained cautious. As Custodia Bank CEO Caitlin Long noted in March 2025: "It's premature to say that debanking is over."

Compliance Costs

Serving a stablecoin issuer requires a bank to handle unusually large, fast-moving deposit flows, maintain enhanced AML/KYC monitoring, and accept the reputational risk of association with crypto. For most community and regional banks, the compliance burden outweighs the revenue from holding deposits. The result is a small set of institutions willing to take on the business, each of which becomes a single point of failure for one or more major stablecoins.

Correspondent Banking Dependencies

Tether's banking structure illustrates a subtler form of concentration. Its primary banking partners (Deltec, Britannia, Capital Union, Ansbacher) are all small, Bahamas-based institutions. None of them hold Federal Reserve master accounts. To move U.S. dollars, each must rely on a correspondent bank in the United States. If a major U.S. bank decides to terminate its correspondent relationship with any of these institutions, Tether's ability to process dollar redemptions could be disrupted even without a direct bank failure.

The Circular Dependency

The fundamental problem is structural: stablecoins are backed by bank deposits, but banks can unilaterally terminate the relationship. This creates a circular dependency where the stablecoin's reliability depends on the continued willingness of its banking partners to serve it.

Unilateral Termination Risk

Banks can close accounts with 30 to 60 days' notice, and sometimes less in cases of regulatory concern. When a bank exits a stablecoin relationship, the issuer must find a replacement and transfer billions of dollars, all while maintaining the peg. This is what happened to Tether multiple times: it moved from Wells Fargo to Noble Bank (Puerto Rico) in 2017, then to Deltec (Bahamas) in 2018, and later added Britannia in 2023 after U.S. banks further restricted crypto relationships.

The Insurance Gap

FDIC deposit insurance covers $250,000 per depositor per institution. A stablecoin with $40 billion in circulation would need accounts at 160,000 banks to be fully insured. Only about 4,500 FDIC-insured institutions exist in the United States. Circle's own SEC filings acknowledge that its reserve balances "far exceeded" the FDIC limit and describe the uninsured deposit risk as one the company has judged "worth accepting" given the alternative of a Treasury-backed fund structure.

Risk FactorUSDCUSDTPYUSD
Number of known banking partners3 (BNY Mellon, Cross River, Citizens Trust)5 (Cantor, Deltec, Britannia, Capital Union, Ansbacher)Paxos + undisclosed banks
Largest single-entity concentration~84% in one fund (BlackRock-managed)Cantor holds majority of Treasury reservesPaxos manages all reserves under NYDFS
Jurisdiction of banking partnersUnited StatesBahamas, United States (Cantor)United States (NYDFS-supervised)
Fed master account accessIndirect (through banking partners)None (correspondent banking only)Indirect (through banking partners)
Prior depeg from banking failureYes (SVB, March 2023)No direct depeg from banking failureNo (launched August 2023, post-crisis)
Monthly third-party attestationYes (Deloitte)Yes (BDO Italia)Yes (Withum)

What Regulators Are Doing

The GENIUS Act, signed into law in 2025, establishes the first comprehensive federal framework for stablecoin reserve management. Its rules, finalized by the OCC in late 2026 with a January 2027 effective date, directly address several dimensions of banking partner concentration.

Reserve Requirements

The law mandates 1:1 backing in specified high-quality liquid assets: U.S. dollars, insured deposits, short-term Treasury securities, and certain repurchase agreements. Reserves must be segregated from operating funds and held with qualified custodians. Rehypothecation is prohibited except for narrowly defined purposes. Each month, the issuer's CEO and CFO must personally certify the accuracy of reserve composition reports.

Concentration Limits

The New York State Department of Financial Services has proposed supplemental rules that would impose explicit concentration limits on reserve custody. Under the proposal, issuers with $25 billion or more in outstanding stablecoins would be required to hold at least 0.5% of reserves (capped at $500 million) in FDIC-insured deposits. The FDIC itself has separately proposed allowing issuers to keep up to 40% of reserves at a single insured bank, though banking watchdog groups have objected that such concentration could threaten the Deposit Insurance Fund in a stress scenario.

Regulatory gap: The GENIUS Act requires qualified custodians and segregated reserves, but it does not mandate a minimum number of banking partners. An issuer could technically comply with the law while holding 100% of its cash reserves at a single bank, as long as that bank meets custodial standards. Diversification requirements, where they exist, come from state-level rules rather than the federal framework.

Potential Solutions

Several approaches could reduce the concentration of stablecoin reserves in a small number of banking institutions. Each comes with its own tradeoffs.

Tokenized Treasury Reserves

The most direct way to reduce banking partner concentration is to hold fewer reserves in bank deposits. Tokenized Treasury products like BlackRock's BUIDL fund (which reached approximately $2.85 billion by late 2025) allow stablecoin issuers to hold short-term government securities on-chain, with settlement and custody handled by regulated fund administrators rather than commercial banks. BlackRock has launched specific tokenized Treasury products designed for stablecoin issuers under the GENIUS Act framework.

This approach reduces bank dependency for the securities portion of reserves but does not eliminate it entirely. Issuers still need banking partners for the cash portion (typically 15 to 20% of total reserves) used to process minting and redemption flows.

Federal Trust Charters

Circle received OCC approval to establish a national trust bank in July 2026, making it the first major stablecoin issuer to hold its own federal banking charter. This allows Circle to serve as its own qualified custodian for reserve assets, potentially reducing dependence on third-party banks. However, a trust charter does not grant access to the Federal Reserve's payment systems or deposit insurance, so it addresses custody concentration without fully solving the deposit and settlement problem.

Multiple Banking Partners

Spreading cash reserves across more institutions is the simplest diversification strategy, but it runs into practical limits. Each new banking relationship requires onboarding, compliance integration, and ongoing monitoring. Many banks impose their own limits on crypto-related deposits. Circle's SEC filings indicate the company caps deposits at 10% of any single bank's deposit base, which limits how many dollars it can place at smaller institutions.

Direct Fed Access

Several proposals have called for giving stablecoin issuers direct access to Federal Reserve master accounts, which would allow them to hold reserves as central bank deposits rather than commercial bank deposits. This would eliminate both counterparty risk and deposit insurance limitations. Custodia Bank applied for a Fed master account and was denied in 2023. The political landscape has shifted since then, but no stablecoin issuer currently holds direct Fed access, and the Fed has shown reluctance to extend master accounts to non-traditional financial institutions.

SolutionReduces Banking DependencyLimitations
Tokenized TreasuriesYes, for securities portion (80%+ of reserves)Cash portion still needs bank partners; fund administrator becomes new dependency
Federal trust charterPartially (issuer becomes own custodian)No Fed payment access; no deposit insurance; regulatory burden of running a bank
Multiple banking partnersYes, through diversificationFew banks willing; compliance cost per relationship; impractical to fully insure reserves
Direct Fed accessYes, would eliminate commercial bank dependencyNo issuer has been granted access; politically uncertain; precedent-setting
Expanded FDIC coverageReduces loss-given-failure, not concentrationUnder preliminary consideration only; scope and eligibility unclear

How This Affects the Broader Ecosystem

Reserve banking concentration is not an abstract risk. It affects every application, protocol, and user that depends on stablecoins for settlement, collateral, or payments. When USDC depegged in March 2023, the impact cascaded through DeFi protocols across Ethereum, Arbitrum, and other chains. Lending markets saw liquidations. AMMs experienced abnormal slippage. Stablecoin depeg contagion is a systemic risk that originates not from smart contract vulnerabilities but from the traditional banking system.

For payment networks built on stablecoins, the implications are direct. If a stablecoin's banking partner fails or terminates the relationship, every wallet, merchant, and payment flow denominated in that stablecoin is affected. This is why the choice of reserve structure and banking partners is not just a compliance question: it is a core infrastructure dependency.

Implications for Bitcoin-Based Stablecoins

Reserve banking concentration affects stablecoins across every chain and layer. On Spark, USDB is issued by Brale, which manages reserve custody under U.S. money transmitter licenses with reserves held in Treasury bills and cash equivalents in segregated, bankruptcy-remote accounts. The same banking partner dependencies that affect USDC and USDT apply to any fiat-backed stablecoin, including those in the Spark ecosystem. Understanding who holds the reserves, and how concentrated that custody is, matters for anyone building or using stablecoin-based payment infrastructure.

For developers integrating stablecoins into wallets and payment flows, the Spark SDK documentation covers how USDB operates within the Spark protocol. Users looking for a practical starting point can explore General Bread, a Spark-powered wallet that supports both BTC and USDB.

What to Watch

Several developments in 2027 and beyond will determine whether stablecoin reserve concentration improves or worsens.

  • The OCC's finalized GENIUS Act rules take effect January 18, 2027, introducing qualified custodian requirements and mandatory reserve segregation across the industry.
  • The FDIC's proposed stablecoin deposit insurance framework, if finalized, could change the economics of reserve diversification by reducing the cost of a single-bank failure.
  • Circle's federal trust bank, once fully operational, will test whether issuer-as-custodian models reduce concentration or simply shift it.
  • Tokenized Treasury adoption by stablecoin issuers will determine how much of the reserve stack moves from bank deposits to on-chain fund shares.
  • The growth of new stablecoin issuers and their banking choices will either diversify or further concentrate the system-wide banking partner set.

The run risk analysis for stablecoins is incomplete without understanding the banking layer beneath the reserves. Reserve transparency reports tell you what assets back a stablecoin. They rarely tell you how many institutions stand between those assets and an actual dollar redemption, or what happens when one of those institutions disappears.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.