Introduction

When the Office of the Comptroller of the Currency (OCC) proposed its stablecoin liquidity framework to implement the GENIUS Act, it drew on a regulatory tradition closer to that of government money market mutual funds (MMFs) than the liquidity rules governing banks. The OCC’s Notice of Proposed Rulemaking (NPR) imposes a uniform standard that, despite lower percentage floors, is stricter than the MMF standard in terms of what assets qualify. It also creates a structural dependency on large demand deposit positions that could increase credit and competitive risks. Figure 1 shows that stablecoin reserve allocations to deposits vary widely across issuers today, even among U.S.-domiciled issuers subject to the GENIUS Act regime, which means the adjustment burden the rule imposes will vary considerably. Moreover, because the liquidity requirement is expressed as a ratio, its “always-on” floor design may force issuers to hold more than the stated minimum in practice.  Deposit placement networks, like IntraFi’s Insured Cash Sweep, offer a practical path toward addressing the most immediate of these risks, but they do not resolve the underlying question of whether uniform thresholds are the right calibration instrument for a heterogeneous sector.

Why Liquidity Regulation Exists

The theoretical case for liquidity regulation rests on a fundamental vulnerability inherent in financial intermediaries: maturity transformation. Institutions that borrow short and lend long are structurally exposed to runs because they fund illiquid, long-dated assets with short-term, demandable liabilities. Diamond and Dybvig (1983) showed that such institutions face multiple equilibria. In a good equilibrium, patient depositors wait and the institution remains solvent. In a bad equilibrium, the fear of others withdrawing triggers a self-fulfilling run even against a solvent institution. Rochet and Vives (2004) extend this framework to show that coordination failures can trigger runs even against solvent institutions when creditors act strategically and interbank markets are imperfect, reinforcing the case for liquidity regulation as a preventive backstop. The social cost of runs extends beyond the failing institution itself. Forced asset liquidations depress prices below fundamental values, what Shleifer and Vishny (1992) termed “fire sales”, imposing losses on otherwise healthy institutions holding similar assets (Brunnermeier and Pedersen, 2009, Calomiris et  al. 2026).

These externalities provide the core rationale for regulatory intervention to require intermediaries to maintain adequate liquidity. In other words, left to their own devices, intermediaries will tend to hold less liquidity than is socially optimal, because they do not fully internalize the systemic costs of their illiquidity (Allen and Gale, 2000). That said, some illiquidity can be desirable when structuring intermediation. Calomiris and Kahn (1991) and Calomiris et al. (2026) show that demandable debt can serve a disciplining function: the credible threat of withdrawal by informed depositors constrains managerial risk-taking in ways that more stable funding would not. Whether that logic extends to stablecoin issuers depends on how transparent their reserve portfolios actually are in practice. Reserve assets are categorically more observable than bank loans. Treasuries and repo instruments carry market prices, and issuers publish periodic attestations. However, the Silicon Valley Bank (SVB) episode illustrated that material information asymmetries can persist even within an apparently transparent framework. Circle disclosed aggregate reserve amounts but not the specific institution holding its deposits, leaving holders unable to observe the concentrated exposure to SVB. To the extent reserve composition is disclosed with a lag or specific counterparties are unnamed, some disciplinary value from demandable claims may remain, but the agency problem is considerably smaller than in bank lending.

Because stablecoin holders can demand redemption at par on short notice, it is also necessary to maintain sufficient assets that are not only of short duration but also can be liquidated immediately. If reserve assets cannot be liquidated at full value in the same timeframe, then the issuer faces a gap that can trigger run dynamics structurally similar to a bank run (Gorton and Zhang, 2023). Ahmed et al. (2024) formalize this in a stablecoin-specific model, showing that the quality and transparency of reserve assets have distinct effects on run risk, and that large negative public shocks to reserve values can trigger runs even when initial reserve quality is high.^[1] This vulnerability was on display in March 2023, when Circle’s USDC temporarily lost its peg after $3.3 billion in reserves were frozen at SVB.

In setting liquidity requirements for stablecoin issuers, the policy question is less about whether they need liquidity requirements (they plainly do) than about how those requirements should be designed. The dominant approaches in existing financial regulation offer two broad templates: uniform portfolio composition standards, as in SEC Rule 2a-7 governing money market funds, and stress-based buffer requirements calibrated to each institution’s risk profile, as in the Basel III Liquidity Coverage Ratio (BCBS, 2013). Each reflects a different set of assumptions about the homogeneity of the regulated population and the reliability of stress models. The LCR-style approach works in part because banks can classify the risk characteristics of their liabilities (e.g., whether they are retail or wholesale, historically stable or unstable). While assets held by stablecoin issuers are observable through public disclosures, issuers typically cannot identify their liability holders with the same granularity that banks can, which limits the practical applicability of a liability-weighted framework to this market. The OCC’s answer, described in the next section, draws primarily from the MMF playbook.

A Uniform Liquidity Framework

The GENIUS Act establishes a reserve asset menu for stablecoin issuers: cash, Federal Reserve balances, demand deposits at insured depository institutions (not the same thing as insured deposits), short-term Treasuries, Treasury-backed repurchase agreements, and shares of government MMFs investing in these assets. Congress left it to the primary federal stablecoin regulators to establish liquidity requirements, leaving the form of those requirements to regulatory discretion.

The OCC’s NPR answers that mandate with a quantitative framework with two options. The proposal would impose fixed percentage minimums of at least 10 percent of reserve assets as daily liquid assets and at least 30 percent as weekly liquid assets, each measured against total outstanding issuance.^[2] Under Option A, these serve as a safe harbor within a broader principles-based standard. Under Option B, they become binding minimums for all issuers.

Either way, the structure differs from conventional bank liquidity regulation. Rules like the LCR model each institution’s stressed outflows and calibrate a liquid asset buffer according to each bank’s liability profile. The OCC’s NPR instead imposes uniform thresholds regardless of issuer size or portfolio composition. The OCC itself acknowledged it referenced SEC Rule 2a-7, the similarly uniform rules governing government MMFs, when drafting these requirements. Under Rule 2a-7, government MMFs must hold at least 25 percent of assets as daily liquid assets and 50 percent as weekly liquid assets.

MMF-style rules create predictable, uniform minimum portfolio compositions rather than risk-sensitive, institution-specific buffers. That approach has logic in a mature, homogeneous industry. But stablecoins are an emerging and evolving market, and issuers may turn out to vary considerably in their business models, risk profiles, and funding structures. A one-size-fits-all standard may be harder to justify as that diversity becomes clearer (as Figure 1 illustrates).

Along with liquidity rules, the NPR establishes a capital regime. Notably, however, the OCC took the opposite approach for capital, citing issuer diversity as the reason for a case-by-case framework, though it did not explicitly reconcile that choice with its uniform liquidity thresholds.^[3] 

The Liquidity Rule has a Deposit Problem

The OCC’s 10 percent daily floor is lower than the 25 percent required of government MMFs, but that comparison understates other aspects of the proposed stablecoin framework. The qualifying asset definitions, WAM cap, and maximum maturity limits are all considerably more restrictive than their MMF equivalents, although redemption timing is somewhat more lenient, allowing up to T+2 versus T+1 for government funds (see table below).

Of those differences, the qualifying asset definitions may have the most consequential practical effect. Under SEC Rule 2a-7, government MMFs can count cash, bank deposits, U.S. Treasury securities of any maturity, and securities maturing within one business day toward their daily liquidity requirement. By contrast, stablecoin issuers can count only demand deposits at banks or credit unions and Federal Reserve balances. Because the rule applies identically regardless of how an issuer’s reserves are structured, an issuer with a large capital buffer, or one holding a portfolio that is heavy in overnight reverse repos or maturing T-bills may be well-positioned to meet any reasonable liquidity standard, yet faces the same mandatory deposit floor as one with a less liquid portfolio. Because interest-bearing Fed accounts are not yet broadly available to non-bank stablecoin issuers, commercial bank demand deposits will be the means for meeting the daily requirement.^[4] A stablecoin issuer with $20 billion in reserves could end up maintaining at least $2 billion in commercial bank demand deposits. The vast majority of that amount likely will be uninsured (in accounts that exceed the $250,000 FDIC insurance limit).

Beyond the qualifying asset definition, a further design tension arises from the always-on nature of the floor. Because the daily liquidity requirement is expressed as a ratio against outstanding issuance, an issuer sitting at exactly the 10 percent threshold cannot use any of its daily liquid assets to fund redemptions without immediately falling into breach. In a stress scenario, the issuer is therefore effectively forced to liquidate other reserve assets such as Treasuries, or possibly borrow in repo markets, to replenish deposits while simultaneously meeting redemption demands.^[5]

The always-on design also compounds the deposit problem in normal times because a prudent issuer will hold deposits above the 10 percent floor to avoid approaching the breach threshold. As a result, the effective demand deposit requirement in practice would exceed the stated minimum and increase uninsured deposit exposure beyond what the rule’s face terms already imply. The tension is considerably more acute under Option B, where the thresholds are hard binding minimums with no discretionary relief. Option A’s principles-based safe harbor preserves regulatory flexibility to assess whether an issuer drawing down its buffer during genuine stress was acting prudently, allowing the buffer to function as intended rather than as a floor that cannot be touched without triggering a violation.

Three Unintended Consequences

Pushing stablecoin issuers toward demand deposits has advantages. Demand deposits are the most liquid instrument on the reserve menu: available same-day, with no contractual duration and no mark-to-market price risk from interest rate changes.^[6] But there are also costs of reliance on them as reserves. These costs fall on all issuers equally, regardless of their existing portfolio composition or risk profile, which is a direct consequence of the rule’s uniform design.

Credit risk. A stablecoin issuer holding billions in demand deposits at a single bank carries the vast majority of that balance in uninsured territory, creating direct credit exposure to that bank. The failures of SVB and Signature Bank in 2023 were a sharp reminder of what that exposure can mean. The OCC recognized this risk, explicitly noting in the NPR that a $1 billion issuer keeping 10 percent of reserves in deposits would need to spread them across 400 accounts to maintain full FDIC coverage, and that in practice issuers would likely concentrate at fewer institutions. The OCC’s response was to address the risk downstream, through concentration limits and a minimum insured deposit requirement for large issuers, and to consider but ultimately not propose a 0.40 percent capital charge on uninsured deposits. What the NPR does not acknowledge is that the uninsured deposit exposure is itself a product of the liquidity rule’s qualifying asset definition. The rule creates the dependency it then attempts to manage.

Deposit concentration. Although the OCC’s concentration limit requires issuers to keep less than 50 percent of their daily reserve at any one bank (and 40 percent of all reserves), they likely will gravitate toward the largest banks, which have the most robust operational infrastructure, the deepest relationships, and the implicit market expectation of too-big-to fail protection that makes uninsured deposits as good as insured. Stablecoin reserves likely will concentrate at the top of the banking system.

Credit disintermediation of community banks. Community banks rely on deposit funding to originate small-business loans, mortgages, and other local credit. A regulatory structure that channels stablecoin reserves toward large institutions could disadvantage community lenders and their customers. As stablecoin issuance grows and depositors substitute toward stablecoins, reserves concentrate at large banks. Community banks that would otherwise hold those deposits lose a corresponding funding base, which could constrain their lending capacity.

A Partial Solution: Deposit Placement Networks

Deposit placement networks address all three problems at once. The most prominent is IntraFi’s Insured Cash Sweep (ICS) service, with similar offerings from R&T Deposit Solutions, StoneCastle Cash Management, and others.^[7] A depositor places funds at a participating institution, which sweeps the balance in sub-$250,000 increments across a network of other FDIC-insured banks. Each piece is fully insured. The depositor keeps a single banking relationship while gaining coverage for the full balance.

On credit risk. Spreading deposits across dozens or hundreds of institutions eliminates the concentrated uninsured exposure. More reserve deposits fall within FDIC coverage limits, reducing credit-risk exposure from bank failure.

On deposit concentration and credit disintermediation. ICS partially reverses the distributional problem. Approximately 95 percent of IntraFi’s roughly 3,000 member banks are community banks (or institutions with less than $10 billion in assets). A stablecoin issuer routing its daily liquidity through ICS would redistribute that funding back to the banks most exposed to stablecoin-driven deposit outflows, partially offsetting the disintermediation that stablecoin growth might otherwise bring.

The OCC has already signaled a similar instinct. Under proposed §15.11(d), issuers with outstanding issuance of $25 billion or more must hold at least 0.5 percent of reserve assets as insured deposits each business day, capped at $500 million. Deposit placement networks are a practical mechanism for meeting that requirement at scale and going well beyond it.

Risks and limitations. Deposit placement networks are not without complications. Coordinating sweep activity across a large bank network introduces operational complexity that a direct deposit at a single institution does not. There are also intraday liquidity questions: ICS deposits typically settle next-day, which may not satisfy same-day needs in a stress scenario. And issuers using ICS would need to monitor the NPR’s 40 percent concentration limit, which applies across all reserve exposures (deposits, custodied securities, repo counterparties) at any single eligible institution. If a stablecoin issuer’s ICS sweeps are routed disproportionately through one receiving bank, whether by network design or that bank’s capacity, inadvertent violations of that threshold are possible. Issuers would need to structure their reserves and clearing arrangements with these constraints in mind.

Deposit placement networks address the distributional and credit-risk consequences of the qualifying asset definition and the always-on nature of the policy. They do not resolve the underlying question of whether uniform thresholds are the right calibration instrument for a sector this heterogeneous; a question the OCC’s own capital framework, with its issuer-by-issuer approach, implicitly acknowledges.

Bottom Line

The OCC’s stablecoin liquidity framework reflects a reasonable first instinct: stablecoins are demandable at par, and reserve assets must be genuinely liquid. But the instrument chosen is blunt. By restricting daily liquidity to bank deposits and Fed balances, the framework risks pushing issuers toward large, uninsured deposit concentrations with predictable consequences for credit risk and the competitive position of community banks. The always-on ratio design compounds this problem; because issuers at the threshold cannot draw on their daily liquid assets to fund redemptions without immediately falling into breach, prudent issuers will hold deposits above the stated minimum as a precautionary buffer, further increasing uninsured deposit exposure beyond what the rule’s face terms imply.

Both problems are better addressed under Option A’s principles-based safe harbor, which preserves the regulatory discretion to treat buffer drawdowns in genuine stress as prudent rather than non-compliant, than under Option B’s binding minimums. Deposit placement networks offer a practical complement, maximizing FDIC coverage and redistributing reserves toward the smaller banks most exposed to stablecoin-driven outflows. As issuance grows, the case for recalibrating these thresholds to reflect each issuer’s actual risk profile will only strengthen.

---