Boston Fed Working Paper: How Stablecoins and Tokenized Funds Compare to Money Market Fund Vulnerabilities

Boston Fed Supervisory Research on Stablecoin Vulnerabilities

The Federal Reserve Bank of Boston’s Supervisory Research and Analysis Unit has published a pivotal working paper (SRA 26-01) that provides the most rigorous comparison to date between stablecoin vulnerabilities and the well-documented fragilities of traditional money market funds. Authored by Kenechukwu Anadu from the Boston Fed alongside Patrick McCabe, JP Perez-Sangimino, and Nathan Swem from the Federal Reserve Board, the paper was first presented at the Second Conference on Stablecoins and Tokenization co-hosted by the Boston and New York Federal Reserve Banks in May 2025.

The research arrives at a critical juncture for the financial industry. With U.S. money market funds managing $7.7 trillion and the stablecoin market reaching approximately $300 billion in capitalization—a sixty-fold increase from $5 billion in 2019—the intersection of traditional and digital financial products creates risk pathways that supervisors are only beginning to understand. The paper establishes that vulnerability analysis must consider how features combine rather than examining them in isolation, a principle that has immediate implications for banks, asset managers, and compliance teams navigating this rapidly evolving landscape.

Five Features That Make Money-Like Products Fragile

The Boston Fed framework identifies five distinct features that create vulnerability to investor runs when they appear in combination. Liquidity transformation occurs when issuers convert illiquid assets into liquid liabilities, giving early redeemers a first-mover advantage. Threshold effects create cliff-edge dynamics—discontinuous changes in expected payoffs such as a fund “breaking the buck” when its net asset value drops below $0.995. Moneyness measures the degree to which a product is treated as a cash equivalent with the “no-questions-asked” property described by economist Bengt Holmstrom. Contagion effects propagate stress from one product or issuer to related products. Reactive investor bases determine how rapidly confidence shocks translate into mass redemptions.

The central analytical insight is that these features are not independently dangerous. The paper points to equity and bond mutual funds as evidence: despite engaging in significant liquidity transformation, these products have never experienced industry-wide runs because they lack the additional combination of threshold effects, high moneyness, and reactive institutional investors that made money market funds vulnerable in September 2008 when the Reserve Primary Fund broke the buck and again during the March 2020 market stress. This combinatorial approach to risk assessment represents a methodological advance over simpler checklists that evaluate features one at a time.

Structural Versus Non-Structural Stablecoin Vulnerability Drivers

One of the most practically useful distinctions in the Boston Fed framework is the classification of vulnerability features as structural, non-structural, or hybrid. Structural features—liquidity transformation and threshold effects—arise from business models and legal frameworks. They are relatively predictable and unlikely to change without regulatory intervention. Non-structural features—particularly the composition of the investor base—reflect how products are perceived and used in practice. They are more malleable and substantially harder to forecast.

Moneyness and contagion effects are classified as hybrid features containing elements of both categories. Moneyness requires structural prerequisites such as principal stability and reliable liquidity, but it also depends on social convention and network externalities—investors are more willing to regard a product as cash-like when others do the same. This creates a self-reinforcing cycle that can shift rapidly in either direction. The Boston Fed warns that because the products examined are novel and still unfamiliar to many investors, non-structural features are “especially likely to shift” as adoption patterns mature, making current assessments inherently tentative.

For risk managers and compliance professionals, the practical implication is clear: monitoring structural features alone provides an incomplete picture. The most dangerous scenarios emerge when non-structural features evolve—for example, when a stablecoin’s investor base shifts from sophisticated crypto traders to risk-averse institutional cash managers who expect stable-value performance.

Stablecoin Reserve Composition and Liquidity Risk

The Boston Fed rates stablecoin liquidity transformation as “uncertain” compared to traditional money market funds—a notably cautious assessment that reflects genuine complexity on both the asset and liability sides of stablecoin balance sheets. On the asset side, reserve compositions vary dramatically across issuers. As of June 2025, Tether holds 20% of its reserves in non-traditional assets: secured loans (6.2%), bitcoins (5.5%), precious metals (5.4%), “Other Investments” (3.0%), and a minimal allocation to corporate bonds (0.01%). These holdings would be far less liquid during a stress event than the Treasury securities and repo agreements that dominate MMF portfolios.

The two largest stablecoins collectively hold approximately $137 billion in U.S. Treasury securities and $57 billion in Treasury repo—holdings that dwarf many individual asset managers but remain modest relative to the $2.7 trillion in Treasury securities and $1.9 trillion in Treasury repo held across the broader MMF industry. This scale difference matters because forced liquidation of stablecoin Treasury positions during a run would create less market price impact than comparable MMF liquidations, but the growth trajectory suggests this buffer may narrow rapidly.

On the liability side, redemption mechanisms introduce their own dynamics. Tether requires a minimum $100,000 redemption with a 0.10% fee. USDC uses a tiered structure charging 0.03% for $2–5 million redemptions, 0.06% for $5–15 million, and 0.10% above $15 million. These fees create a partial deterrent to panic redemptions—an important structural difference from money market funds where redemptions are typically free—but they also mean that most stablecoin holders rely on secondary market sales rather than direct issuer redemptions, shifting the vulnerability from the issuer’s balance sheet to secondary market liquidity.

Tokenized Money Market Funds and the Ownership Transfer Debate

Perhaps the most consequential finding in the Boston Fed working paper concerns tokenized money market funds and the pivotal question of whether blockchain tokens can legally effect transfers of underlying fund share ownership. Currently holding almost $1.5 billion in net assets—with the broader category including vehicles like BlackRock’s BUIDL fund exceeding $6 billion—tokenized MMFs sit at a regulatory crossroads that will fundamentally determine their vulnerability profile.

If tokenization merely provides a more convenient digital wrapper for existing fund shares—with all transfers still processed through traditional transfer agents—then tokenized MMFs add little new risk beyond what the underlying fund already presents. The token becomes a cosmetic innovation. However, if tokens can broadly transfer MMF share ownership on public blockchains, enabling 24/7 trading, instant settlement, and collateral posting, then tokenization simultaneously increases the product’s utility and amplifies every vulnerability of the underlying fund. The paper notes that the SEC Crypto Task Force, following Commissioner Hester Pierce’s February 2025 statement, has signaled that this transferability question is a priority.

The Boston Fed identifies a particularly intriguing possibility: if tokenized MMF shares could be posted as margin collateral, investors facing margin calls during market stress could pledge additional tokens rather than redeeming MMF shares for cash. This would directly reduce the redemption pressure that drives traditional MMF runs. It represents a rare case where innovation could actually mitigate rather than amplify existing vulnerabilities—but only if the broader transferability framework is established in a way that enables collateral use without creating excessive secondary market liquidity risks.

Cross-Product Contagion: A Novel Financial Stability Risk

The Boston Fed working paper’s identification of cross-product contagion channels represents its most significant contribution to the financial stability literature. Traditional contagion analysis focuses on within-product contagion—stress at one MMF spreading to other MMFs, or problems at one stablecoin prompting runs on competing stablecoins. The paper demonstrates that MMETFs and tokenized MMFs introduce an entirely different dynamic: cross-product contagion where events in one category of money-like product trigger destabilizing responses in a fundamentally different category.

The mechanism is straightforward but powerful. Money market ETFs, first introduced in 2024 with approximately $4 billion in net assets, trade at market-determined prices that fluctuate throughout the trading day. Traditional government and retail prime MMFs maintain stable $1.00 NAVs through price rounding. If market stress causes MMETF prices to decline—even modestly—MMF investors may interpret the decline as evidence that their own stable-NAV shares are overvalued at $1.00. This informational signal could trigger MMF redemptions even when no new information about MMF asset quality has emerged. The paper characterizes this as a mechanism that adds to aggregate financial system vulnerability, particularly because the $7.7 trillion MMF sector dwarfs the nascent MMETF market.

An analogous cross-product contagion channel exists for tokenized MMFs: sophisticated investors observing a discount between token prices and underlying fund NAV could execute an arbitrage strategy—buying discounted tokens, converting to fund shares, and redeeming at full NAV—that accelerates precisely the kind of information-driven run that proved devastating during the 2008 financial crisis.

GENIUS Act Regulatory Impact on Stablecoin Risk Profiles

The Guiding and Establishing National Innovation for U.S. Stablecoins (GENIUS) Act, signed into law in July 2025, introduces the first comprehensive federal regulatory framework for “payment stablecoins” in the United States. The Boston Fed working paper provides a nuanced assessment of the Act’s impact on vulnerability features, finding that it simultaneously reduces some risks while creating or amplifying others—a classic regulatory trade-off that supervisors must navigate carefully.

On the risk-reduction side, the GENIUS Act imposes reserve requirements that partially align with money market fund rules under SEC Rule 2a-7. Payment stablecoin issuers must hold reserves in high-quality liquid assets with Treasury maturity capped at 93 days—actually stricter than MMF weighted average maturity requirements. The Act also mandates capital requirements, regular auditing, and transparency standards that should reduce the opacity currently surrounding certain issuers’ reserve compositions. Additionally, the Act’s prohibition on redemption suspensions for payment stablecoins eliminates one type of threshold effect that can precipitate panic.

However, the Boston Fed warns that greater regulation paradoxically increases certain vulnerability features. By legitimizing payment stablecoins as regulated financial products, the GENIUS Act is likely to increase their moneyness substantially—attracting risk-averse institutional investors who expect cash-equivalent stability. Greater standardization across regulated stablecoins also increases contagion potential: if one regulated issuer experiences stress, investors may question all regulated issuers more quickly because the products appear more homogeneous. The paper further notes that the Act’s prohibition on redemption suspensions, while reducing threshold effects, also removes a crisis management tool that gives issuers breathing room during acute stress events.

A critical regulatory gap persists for non-payment stablecoins—including crypto-backed and algorithmic varieties—which the GENIUS Act does not cover. These products “could continue to operate without GENIUS Act restrictions,” creating a dual system where regulated payment stablecoins coexist alongside unregulated alternatives with potentially higher risk profiles.

Stablecoin Moneyness and the No-Questions-Asked Property

The Boston Fed classifies stablecoin moneyness as “likely similar” to traditional money market funds—a striking assessment given that stablecoins were created barely a decade ago and initially served a niche audience of cryptocurrency traders. The concept of moneyness draws on economist Bengt Holmstrom’s 2015 characterization of the “no-questions-asked” (NQA) property: an asset is money-like when it is perceived as so safe and liquid that its price relative to the unit of account is effectively fixed and investors need not investigate whether they can transact at par.

Within the digital asset ecosystem, stablecoins already function with near-NQA status. They serve as the primary medium of exchange for DeFi protocols, trading platforms, and cross-border transactions. In 2024, roughly two-thirds of all stablecoin transactions were for DeFi and trading-related activities—contexts where participants treat USDT and USDC as functional equivalents of dollars without questioning their redeemability. The GENIUS Act is expected to expand this NQA perception into mainstream payment contexts, potentially making stablecoins the first digital-native money-like products to achieve true mass-market moneyness.

The fragility of moneyness is illustrated by the USDC de-pegging event during Silicon Valley Bank’s failure in March 2023. When Circle disclosed $3.3 billion in SVB exposure—representing 8.3% of USDC reserves—the secondary market price collapsed to an intraday low of $0.88. This 12% deviation from par demonstrated how quickly the NQA property can evaporate: once investors begin asking questions about reserve safety, the self-reinforcing cycle that sustains moneyness reverses into a self-reinforcing cycle of redemptions and price declines.

Dynamic Liquidity Fees and Missing Stablecoin Safeguards

The Boston Fed working paper highlights an important regulatory asymmetry: since October 2024, institutional prime money market funds have been required to impose dynamic liquidity fees when daily net redemptions exceed 5% of fund assets. This mechanism, adopted by the SEC in 2023, forces redeeming investors to bear the liquidity costs their withdrawals impose on remaining shareholders. It directly addresses the first-mover advantage that drives run dynamics by making early redemption during stress periods more expensive.

No equivalent mechanism exists for stablecoins. While stablecoin issuers like Tether and USDC impose redemption fees and minimum thresholds, these are fixed-rate charges that do not scale with redemption pressure. During a genuine crisis—when redemption volumes spike and reserve liquidity becomes constrained—a fixed 0.10% fee provides minimal deterrence compared to a dynamic fee that increases as stress intensifies. The absence of a scalable fee mechanism means that stablecoins lack one of the most effective tools that regulators have developed for managing run risk in traditional money market funds.

The paper also notes that current stablecoins retain the option to suspend redemptions entirely during extreme stress—a blunt instrument that creates its own threshold effects but does provide a last-resort circuit breaker. The GENIUS Act removes this option for payment stablecoins, meaning that the regulated stablecoin sector will operate without either a dynamic fee mechanism or a suspension option. This represents a structural vulnerability gap that the supervisory community may need to address as payment stablecoins grow toward the $1.5 trillion scale that industry analysts project.

Supervisory Outlook for Digital Money-Like Products

The Boston Fed working paper concludes by identifying the key developments that will shape the vulnerability landscape for each product category. For money market ETFs, the critical question is whether regulators will require in-kind redemptions—a structural change that would substantially reduce liquidity transformation vulnerabilities but faces practical challenges given that repo ownership is difficult to transfer in-kind. For tokenized MMFs, the SEC Crypto Task Force’s determination on token-based ownership transfers will be watershed: it will either keep tokenized MMFs as cosmetic innovations with limited additional risk or transform them into powerful—and potentially destabilizing—financial instruments.

For stablecoins, the implementation of the GENIUS Act creates what amounts to a natural experiment in regulatory design. The paper identifies several variables that will determine outcomes: whether payment stablecoins capture the bulk of market growth or non-payment stablecoins continue to dominate; how quickly institutional investors adopt regulated stablecoins as cash management tools; whether reserve compositions converge toward higher-quality liquid assets under regulatory pressure; and whether secondary market infrastructure develops sufficient depth to handle stress-period trading volumes without severe price dislocations.

The broader message for supervisors and financial institutions is that the landscape of money-like products has become fundamentally more complex. Products that would have been classified as entirely separate asset categories a decade ago now create interconnected vulnerability pathways through shared investors, cross-product price signals, and common reserve assets. The Boston Fed’s five-feature framework provides a structured methodology for tracking these evolving risks—but the authors emphasize that current assessments are tentative precisely because the non-structural features most likely to drive future crises are also the ones most likely to change unpredictably as adoption patterns shift and regulatory frameworks mature.