By Lee Reiners and Mark Hays
- Introduction
- A Politicized CEA and Predetermined Conclusions
- The Flawed Economics of the CEA’s Stablecoin Report
- Conclusion
- About
Introduction
Last month, the White House Council of Economic Advisers (CEA) released a report on payment stablecoins that asks a narrow question: what happens to bank lending if stablecoin issuers are prohibited from paying yield? The report concludes that such a prohibition would have a minimal effect on lending, implying that concerns about stablecoins disintermediating the banking system are overstated.[1]
That conclusion arrives at a particularly consequential moment. The Senate is actively negotiating the contours of a broader crypto market structure framework, including whether — and how — to address what has become known as the “payment of interest” loophole. Section 4(a)(11) of the GENIUS Act prohibits stablecoin issuers from paying interest or yield to token holders, reflecting a deliberate policy choice to prevent stablecoins from functioning as deposit substitutes. But the statute, as currently written, does not prohibit exchanges or other intermediaries from offering rewards on stablecoin balances.[2] Whether Congress moves to close this gap will shape the competitive dynamics of the stablecoin market and the extent to which stablecoins compete with traditional bank deposits.
The CEA’s analysis was quickly embraced by the crypto industry as evidence that such restrictions are unnecessary.[3] If limiting yield has little impact on bank lending, the argument goes, then policymakers should allow stablecoins — and the platforms that distribute them — to compete freely on returns. But that framing depends entirely on the assumptions embedded in the CEA’s model and the narrowness of the question it chose to answer.
On May 1, Senators Thom Tillis (R-N.C.) and Angela Alsobrooks (D-Md.) released compromise legislative text that, on its face, appears to reject the CEA’s core premise that stablecoin yield poses little risk to bank deposits and lending. The text states that it is the sense of Congress “that the payment of consideration by digital asset service providers to United States customers or users based on their payment stablecoin balances in a manner that is economically or functionally equivalent to the payment of interest or yield on an interest-bearing bank deposit may inhibit depository institutions’ key functions in the American economy.” Yet the same text contains a series of carveouts and ambiguities that would likely permit many existing stablecoin rewards programs to continue, while potentially further incentivizing stablecoin adoption at the expense of bank deposits.[4] It is therefore unsurprising that Coinbase — the primary beneficiary of such programs — quickly endorsed the compromise language.[5]
This paper argues that the CEA’s analysis is deeply flawed and substantially understates the potential impact of stablecoin growth on the banking system. More importantly, the report cannot be understood in isolation. It reflects a broader pattern of CEA analysis during the second Trump administration that relies on questionable assumptions and produces conclusions that align with the administration’s policy objectives, as well as an institutional context shaped by significant political and financial conflicts of interest. Taken together, these issues suggest that the CEA is not simply answering the wrong question but doing so in a way that risks distorting an important policy debate at a critical juncture.
A Politicized CEA and Predetermined Conclusions
Context on the CEA
The CEA reports directly to the President of the United States. Ask any former CEA staffer and they will tell you that the Council does not publish economic research that runs counter to the President’s policy agenda. That does not mean the analysis is fabricated, but it does mean that conclusions inconsistent with the White House’s priorities never see the light of day.
That institutional reality matters here because President Trump has not simply embraced crypto as a matter of policy — he has done so as a matter of personal financial interest. A recent report from Democracy Defenders Fund, co-authored by one of this report’s authors, examines how President Trump and his family’s extensive cryptocurrency business interests are shaping federal financial regulation and distorting competition in digital asset markets. Although the Trump family’s crypto-related activities span a wide range of ventures — including non-fungible tokens (NFTs), memecoins, and Bitcoin mining — the report focuses on a discrete and particularly consequential case: World Liberty Financial (WLFI) and its proprietary U.S. dollar-denominated stablecoin, USD1.[6]
The economics of stablecoins make the presidential connection particularly important. Stablecoin issuers generate revenue from the interest earned on their reserve assets. The faster a stablecoin grows, the larger its reserve base and the greater the income it generates. Critically, that growth is often driven by third-party rewards programs, offered by exchanges or affiliated platforms, that effectively pass through some portion of that reserve income to users in order to accelerate adoption. For example, in 2025 Circle paid Coinbase approximately $1.4 billion based on the amount of USDC held on Coinbase.[7] Therefore, the ability of third parties to offer rewards is not a side issue; it is central to the business model of both stablecoin issuers and their distribution partners.
Because the President is financially tied to a stablecoin that benefits from these dynamics, he has a direct economic incentive to ensure that such rewards remain permissible. In late December 2025, the world’s largest crypto exchange, Binance, launched the “USD1 Boost Program,” offering users an annualized percentage rate (APR) of up to 20 percent on USD1 balances of up to 50,000 tokens.[8] Although U.S. users are currently unable to access the primary Binance platform on which the Boost Program operates, Binance maintains a U.S. affiliate that offers a more limited suite of products[9] and that the firm has signaled it intends to expand.[10] Nothing in principle prevents Binance — or any other U.S.-based crypto exchange — from offering similar reward USD1 programs to U.S. customers. As a result, policies that restrict third-party rewards could directly constrain the growth, and therefore the profitability, of ventures in which the President has a financial stake.
Against that backdrop, it would be unrealistic to expect the CEA to produce analysis concluding that restrictions on third-party rewards are necessary or even economically meaningful. The more plausible interpretation is that the desired policy outcome — preserving the rewards channel —came first, and the economic analysis followed.
Conflicts of Interest and the Role of Stephen Miran
Concerns about the integrity of the CEA’s analysis are compounded by the unusual and highly controversial role played by Stephen Miran.
Between September 2025 and February 2026, Miran served as a Federal Reserve Board Governor while maintaining his role as Chair of the CEA through an unpaid leave — an arrangement that broke with longstanding norms designed to insulate monetary policy from White House influence. From the outset, the dual role drew sharp criticism. Democratic senators warned that retaining his status as a White House official while serving at the Fed would undermine the central bank’s independence and leave him beholden to the President.[11] Miran ultimately resigned as CEA Chair on February 3, 2026 to remain at the Fed, ending what the Wall Street Journal described as an “unusual dual role.”[12]
But the concerns did not end there. Since joining the Federal Open Market Committee (FOMC), Miran has repeatedly dissented in favor of lower interest rates, aligning with President Trump’s public calls for aggressive rate cuts. He dissented in September, October, and December 2025, each time favoring larger cuts than the Committee adopted, and again in January, March, and April 2026, when he pushed for cuts even as the Committee held rates steady.[13]
Dissent at the Fed is not uncommon, but it is relatively rare and typically reflects technical disagreements over inflation or labor market dynamics. Miran’s pattern — consistent dissents in the direction of easier policy — stands out, particularly given its alignment with the President’s stated preferences for lower interest rates. The most recent FOMC decision at the end of April is especially telling. While Miran was one of four voting members to dissent from the Committee’s policy action, he was the only one to do so in favor of lowering the Fed’s target rate. The other three dissented because they objected to language in the Fed’s policy statement suggesting that the next move in rates would be lower.[14]
Miran’s term as a Fed governor will conclude once Kevin Warsh is sworn in as the next Fed chair, but his short tenure on the Board reveals a troubling dynamic. Taken together, his dual role, the controversy surrounding his appointment, and his voting record raise serious questions about the independence of both the Federal Reserve and the Council of Economic Advisers during this period.
A Pattern of Questionable Analysis
The CEA’s stablecoin report reflects a broader pattern of analysis that relies on questionable assumptions and produces results that align neatly with the administration’s policy objectives.
Consider Chapter 10 of the 2026 Economic Report of the President, which attempts to quantify “The Economic Consequences of DEI.” The report bolsters the administration’s broad-based attacks on racial equity and civil rights by concluding that industries pursuing diversity initiatives were approximately 2.7 percent less productive by 2023.[15]
As Wall Street Journal columnist Greg Ip explains, the methodology underlying this conclusion is deeply flawed. Because firms do not disclose DEI policies in any systematic way, the authors construct a proxy by tracking changes in the racial composition of management. They then infer that increases in minority representation reflect the adoption of DEI policies and attribute subsequent changes in productivity to those policies.[16]
This approach assumes that changes in workforce composition are driven by DEI initiatives rather than broader demographic, educational, or industry-specific trends. It relies on a before-and-after comparison without a credible control group, making it impossible to establish causation. And it effectively assumes that nothing else in the economy changed between 2015 and 2023, a period marked by a pandemic, massive fiscal stimulus, supply chain disruptions, and structural shifts in the labor market.
It also highlights a more fundamental methodological issue: even if it were the case that DEI policies reduced productivity at individual firms, competitive forces would mitigate the aggregate impact. Firms that performed worse would lose market share to more productive competitors, limiting the effect on overall output. The CEA’s analysis ignores this basic equilibrium dynamic, leading to conclusions that are not just questionable, but economically incoherent.
In February 2026, the CEA published a 16-page report claiming the Consumer Financial Protection Bureau (CFPB) has cost consumers between $237 billion and $369 billion since 2011, including $222 billion to $350 billion in higher borrowing costs.[17]
However, as Georgetown Law professor Adam Levitin put it in a critique of the report, “everything in it is wrong, starting with the first sentence.”[18] The paper begins with basic factual errors, overstating the number of CFPB rules and formal advisory opinions by more than 200 and mischaracterizing the Bureau’s statutory authority. It then attributes a 16-basis-point mortgage rate premium for loans with debt-to-income (DTI) ratios above 43 percent — which fall outside the Ability-to-Repay safe harbor — to CFPB regulation, while ignoring that the underlying ability-to-repay requirement was established by Congress, that the CFPB created the safe harbor for qualified mortgages, and that higher-DTI borrowers are inherently riskier. The analysis further compounds these errors by extrapolating its mortgage findings to auto loans and credit cards using consumer complaints as a proxy for regulatory intensity, effectively treating evidence of consumer harm as evidence of excessive regulation. Like the CEA’s DEI analysis, the result is not a serious attempt at causal inference or welfare analysis, but a selective and profoundly methodologically flawed exercise designed to reach a predetermined conclusion.[19]
But this analysis came as the Trump administration was taking aggressive, and unprecedented, actions to dismantle the CFPB.[20] Since taking office, the administration has issued stop-work orders halting supervision, enforcement, and rulemaking; shuttered the agency’s headquarters; fired large portions of its staff; and sought to slash the remaining workforce by more than half while significantly curtailing its core functions.[21] The result has been an 80 percent decline in the agency’s processing of consumer complaints, leaving thousands of complaints unaddressed and effectively sidelining the CFPB’s statutory mission.[22] Against that backdrop, the CEA report reads less like neutral economic analysis and more like an attempt to provide intellectual cover for a policy agenda aimed at eliminating the nation’s primary consumer financial regulator.
The CEA’s stablecoin report follows this same pattern. It frames the issue narrowly, relies on restrictive and often unrealistic assumptions, and ultimately produces a result that aligns neatly with the administration’s policy preferences regarding regulation of the crypto industry, or deregulation as the case may be.[23]
The Flawed Economics of the CEA’s Stablecoin Report
The CEA Asks a Very Narrow Question
The CEA stablecoin study asks a very narrow question: what happens to bank lending if stablecoins are prohibited from paying yield today? As the American Bankers Association’s Sayee Srinivasan and Yikai Wang note in their response, this is simply the wrong question to ask.[24]
There is little evidence that current stablecoin rewards programs are meaningfully disintermediating bank deposits. Coinbase — the largest U.S. crypto exchange — has offered “rewards” on USDC balances (currently at 3.5 percent APY for select customers[25]) for years without triggering alarm from the banking sector. Even if one assumes that the roughly $300-$320 billion in dollar-denominated stablecoins would otherwise sit in bank deposits, that is trivial relative to the roughly $19 trillion in commercial bank deposits.[26]
Moreover, current stablecoin rewards are broadly in line with yields available on government money market funds. This is not a coincidence. Under the GENIUS Act, permissible reserve assets for stablecoins — including case cash, short-term Treasuries, repos, and short-government bond repurchase agreements — closely resemble the eligible assets for government money market funds. These funds have been widely accessible to retail investors for decades, yet banks are not arguing that they should be banned or that they pose an existential threat to deposit funding. But as Srinivasan and Wang correctly argue, the relevant question is not about today’s equilibrium, it is about what happens to bank deposits and lending if stablecoins scale.[27]
Even the more conservative public forecasts contemplate substantial growth. Standard Chartered has projected the stablecoin market could reach roughly $2 trillion by 2028, Bernstein has forecast roughly $4 trillion over the next decade, and even J.P.Morgan — which dismissed trillion-dollar projections as “far too optimistic” — still estimated the market could reach $500 billion by 2028. Therefore, the relevant policy horizon is not the market as it exists in April 2026, but the one the industry is actively building — with the support of policymakers — over the next several years.[28]
A Preview from the Last Crypto Cycle and a Look to the Future
The crypto industry’s recent history offers a clear preview of how this could evolve. During the 2020—2022 cycle — prior to the FTX collapse — crypto platforms aggressively marketed “yield,” “earn,” and “rewards” products to retail users, often promising double-digit returns on assets positioned as low-risk alternatives to savings accounts.
In reality, those yields were driven by highly leveraged lending, opaque proprietary trading, rehypothecation of customer assets, and, in some cases, outright fraud. When crypto asset prices declined and counterparties failed, these programs quickly unraveled. Voyager, BlockFi, Celsius, Genesis, and Terraform Labs all offered above-market yields to attract deposits — and all ultimately failed or entered bankruptcy during the crypto winter. In hindsight, these “rewards” were not a sustainable financial innovation, but a customer acquisition strategy.
While today’s stablecoin rewards on U.S. exchanges are more modest, there is nothing in law or regulation that prevents a return to these dynamics. Although some of these earlier products were alleged to be unregistered securities by state regulators and the Securities and Exchange Commission (SEC), the agency is unlikely to pursue similar cases under the Trump administration given its more accommodating stance toward the sector. Therefore, rewards could once again be subsidized by stablecoin issuers seeking to grow market share, by exchanges pursuing aggressive user acquisition strategies, or some combination of the two.
That possibility becomes more plausible, not less, as crypto platforms move toward what SEC Chairman Paul Atkins has called “super-apps” — platforms that can offer a broad range of financial products “under one roof with a single license.” Firms like Coinbase and Robinhood increasingly offer integrated platforms spanning crypto trading, equities, derivatives, and even event contracts.[29] Their business models rely heavily on cross-selling and customer retention.
Offering above-market rewards on stablecoin balances could become a rational loss-leading strategy: attract customers with yield, then monetize them across trading, lending, and wealth management products. Traditional banking provides a clear precedent — Wells Fargo’s infamous “Eight is Great” cross-selling strategy comes to mind.[30]
Crucially, these rewards do not even need to be funded internally. For example, Coinbase’s integration with decentralized finance protocols like Morpho allows users to earn yield through on-chain lending markets while remaining within a centralized interface.[31] This “DeFi mullet” —centralized front end, decentralized back end — obscures the underlying risk while preserving the user experience. For customers, the source of the yield is largely irrelevant.[32]
For now, platforms may be cautious. Offering aggressively high yields ahead of a finalized market structure bill could invite regulatory scrutiny or political backlash. But if regulatory clarity is achieved, the incentives to compete on yield — and scale rapidly — will be significant.
Another potential driver of explosive growth is “agentic commerce” the use of autonomous AI agents to conduct transactions on behalf of users. In this paradigm, software agents could continuously manage liquidity, execute trades, pay for services, and optimize financial decisions in real time. —
Stablecoins are a natural settlement asset for this model. On Circle’s February 2026 earnings call, Circle CEO Jeremy Allaire described AI agents as “a major new demand driver” for Circle’s stablecoin network and said that in a world of “tens or even hundreds of billions of AI agents,” the velocity of money could become “multiple orders of magnitude higher” than it is today.[33] In such a world, stablecoin demand would not be limited to human portfolio allocation decisions; it would be embedded in machine-driven financial systems operating at scale. This is a fundamentally different demand profile than anything captured in the CEA’s static model.
The Tillis and Alsobrooks “Compromise” Could Hasten Deposit Outflows
The Tillis and Alsobrooks “compromise” attempts to resolve the very issue the CEA largely dismisses: whether allowing stablecoins to deliver yield — directly or indirectly — will accelerate the migration of deposits out of the banking system.[34] The legislation splits the difference by prohibiting “any form of interest or yield… solely in connection with the holding of… payment stablecoins; or on a payment stablecoin balance in a manner that is economically or functionally equivalent to the payment of interest or yield on an interest-bearing bank deposit,” while permitting “rewards or incentives based on bona fide activities or bona fide transactions that are not economically or functionally equivalent to the payment of interest or yield on an interest-bearing bank deposit.”[35]
This framework closely mirrors existing credit card rewards programs, which is ironic given that stablecoin advocates frequently position these instruments as a superior alternative to cards and their merchant-funded interchange fees. It is also ironic in a more fundamental sense: the compromise implicitly incentivizes greater centralization of a technology originally designed to enable peer-to-peer payments without intermediaries. In practice, under the compromise text, users can only earn transaction-based rewards if they custody their assets with centralized platforms that monitor and monetize their activity.
The compromise text also leaves significant ambiguity to be resolved through rulemaking. Regulators will need to determine whether rewards are paid “solely in connection” with holding a balance or are sufficiently tied to activity to fall outside the prohibition, a distinction that may ultimately hinge on how programs are structured and marketed. More importantly, regulators must assess whether a given reward is “economically or functionally equivalent” to interest on a bank deposit. That is an inherently subjective determination, and one that invites aggressive structuring by industry participants. Given the current regulatory environment, it is reasonable to expect that these ambiguities will be resolved in ways that favor continued stablecoin growth.[36]
Compounding this problem, the compromise includes a “non-exhaustive list” of permissible activity-based rewards that effectively codifies many existing stablecoin incentive programs. For example, rewards are explicitly permitted for “providing liquidity for market-making activity, posting of collateral in connection with trading, or otherwise putting assets at credit or investment risk,” which would encompass arrangements like Coinbase’s integration with lending protocols such as Morpho. The text also permits rewards tied to staking and other common exchange-based services.[37]
Perhaps most importantly, the compromise ignores the realities of how crypto platforms are structured and how users actually behave. As recent history has shown, users do not meaningfully distinguish between different sources of yield — they care about returns and ease of use. Exchanges understand this and will structure their offerings on the back end to comply with the letter of the law, while preserving a front-end experience that looks largely unchanged: deposit stablecoins and watch your balance grow over time.
This dynamic is reinforced by recent guidance from the SEC’s Division of Trading and Markets, which states that certain “covered user interfaces” that enable users to initiate crypto asset securities transactions through self-custodial wallets generally do not need to register as broker-dealers, provided they meet specific conditions, such as not taking custody of user funds, not providing investment advice, and not exercising discretion over routing or executing orders.[38] In practice, this gives large crypto platforms a clearer path to offer web and app-based interfaces that allow users to access DeFi protocols and yield-generating strategies through self-custodial wallets without triggering traditional broker-dealer obligations. Because major U.S. exchanges already offer both custodial and self-custodial wallet products, they can seamlessly direct users to these interfaces, enabling them to convert stablecoins into tokenized securities or other crypto assets and earn yield through decentralized markets, all while maintaining a unified and simplified user experience. The result is a regulatory architecture that facilitates the expansion of “DeFi mullet” models — centralized front ends paired with decentralized back ends — further blurring the line between prohibited interest and permissible rewards.
The inclusion of transaction-based rewards further tilts the playing field in a way that meaningfully reshapes platform behavior. By explicitly permitting incentives tied to “transaction, payment, transfer, conversion, remittance, or settlement activity,” the compromise encourages users to hold stablecoins as their default transactional medium within crypto platforms. In practice, this means that stablecoins become the preferred funding layer for all activity — crypto trading, tokenized equities, derivatives, and even event contracts. The economics are straightforward: why hold idle dollars on a platform earning nothing when stablecoins generate rewards every time they are used? Over time, this creates a powerful incentive for users to keep their entire trading balances in stablecoins, not just for payments but as the base asset for all financial activity on the platform.
In that environment, the distinction between prohibited “interest” and permissible “rewards” becomes largely academic. The economic reality is that users are being compensated for holding and using stablecoins in ways that closely resemble deposit interest. Rather than limiting disintermediation, the compromise may accelerate it, facilitating a steady migration of funds out of the banking system and into stablecoin ecosystems that replicate many of the functions of deposits without being subject to the same regulatory constraints.
This dynamic is critical because it directly contradicts the central premise of the CEA’s analysis. The report assumes that stablecoin growth — whether driven by yield or other incentives — has only a negligible effect on bank deposits and lending. But if the compromise itself encourages users to hold stablecoins as the default asset for all financial activity on these platforms, then the relevant question is no longer whether yield matters at the margin, but how large and persistent deposit substitution becomes as these ecosystems scale.
Where the CEA Model Breaks Down
The dynamics described above, particularly the potential for stablecoins to become the default funding layer for a wide range of financial activity, are absent from the CEA’s framework. Instead, the report concludes that stablecoin expansion poses little risk to bank lending, a result that rests heavily on a key simplifying assumption: that stablecoin reserve flows simply recycle deposits within the banking system.
In their baseline scenario, a depositor withdraws $1 from Bank A to purchase a stablecoin. The issuer uses that $1 to buy a Treasury bill from a dealer, and the dealer deposits the proceeds at Bank B. Aggregate deposits, in this framework, remain unchanged.[39] But this assumption obscures two critical issues.
First, it ignores distributional effects. As others have pointed out, the deposits leaving Bank A are far more likely to originate from community banks, while the deposits arriving at Bank B are overwhelmingly concentrated at large, systemically important institutions. In practice, stablecoin reserves are custodied at institutions like BNY Mellon and managed by firms like BlackRock. Circle’s own disclosures illustrate this concentration: of the $11.6 billion in cash reserves backing USDC, $11 billion is held in deposit at systemically important financial institutions.[40]
Second, the model assumes that all proceeds from Treasury sales ultimately return to the banking system as deposits. That is far from guaranteed. As Philip Basil, Christopher Appel, and Amanda Fisher at Better Markets point out, “it is unreasonable to assume that the 88 percent of stablecoin reserve holdings in the form of Treasuries and repos will entirely recirculate through the banking system.” As they note, for this to be true, “then bank deposits would roughly equal the value of all financial assets, which is not even close to being the case.” Therefore, the proceeds from selling treasuries to stablecoin issuers are more likely to “end up reinvested in other financial assets, not in bank deposits.” [41]If deposits do not fully recycle, the implications for bank funding — and lending — are materially different.
These distributional dynamics have real economic consequences. As Jill Castilla, CEO of Citizens Bank of Edmond, warned:
A yield-bearing stablecoin offered through a national platform like Walmart would pull liquidity out of rural and community banks and into centralized balance sheets managed far from the customers and businesses those deposits once supported.[42]
Community banks play a disproportionate role in lending to small businesses and local economies. As Better Markets notes, banks under $1 billion in assets direct roughly 78 percent of their commercial and industrial lending to small businesses. When deposits migrate away from these institutions, the impact falls most heavily on Main Street.[43]
As the ABA’s Srinivasan and Wang note, a community bank that loses deposits has to replace that funding, and quickly, often through higher-cost wholesale funding such as Federal Home Loan Bank advances or capital-markets borrowing, or by raising the rate it pays depositors to keep them from leaving. Either way, its cost of funds rises.[44] That means higher borrowing costs for households and small businesses, and less lending to the local economy.
Furthermore, the CEA’s analysis is conditioned on a post-Global Financial Crisis “ample reserves” framework, in which “banks hold excess reserves above regulatory minimums, so deposit reshuffling across banks does not force balance sheet contraction.”[45] But that assumption is doing enormous work. The President’s own nominee for Fed Chair, Kevin Warsh, has been explicit in his desire to shrink the Fed’s balance sheet and move toward a scarcer reserves regime[46] — a position he reiterated during his Senate Banking Committee confirmation hearing.[47] While many economists are skeptical of the Fed’s ability to quickly reduce their balance sheet[48], if that shift were to occur, the CEA’s core mechanism breaks down. In a world of scarce reserves, banks receiving deposits from stablecoin-related flows may be forced to hold a larger share of those funds as reserves to meet liquidity and regulatory constraints, rather than deploying them into loans. At the same time, banks losing deposits may be forced to contract lending outright. In other words, the same deposit reshuffling the CEA treats as neutral could, under a different and entirely plausible monetary policy regime, result in a net reduction in credit. The CEA briefly acknowledges this possibility in an appendix but declines to quantify it[49], despite the fact that President Trump had nominated Warsh several months prior to the report’s publication and his views on the Fed’s balance sheet were well known.[50]
The Repo and ON RRP Problem
The CEA further assumes that Treasuries and overnight reverse repurchase agreements (repos) have equivalent effects on bank deposits. This is a significant oversimplification.
Of the $78.8 billion in Circle’s USDC reserves, $46.3 billion consists of overnight reverse treasury repos.[51] And of the roughly $193 billion in Tether’s USDT reserves, 13.1 percent are held in overnight treasury repos (Circle and Tether account for more than 80 percent of dollar denominated stablecoin market capitalization).[52] In an overnight reverse repo, the stablecoin issuer will lend cash to a counterparty that provides Treasury securities as collateral. The trade unwinds the next day, and the stablecoin issuer will get their cash back plus a small amount of interest (the repo rate).
Roughly 90 percent of Circle’s reserves are held in Circle Reserve Fund, a government money market fund managed by BlackRock.[53] According to the fund’s disclosures, its repo counterparties are primarily large financial institutions.[54] If those counterparties redeposit the cash they receive at commercial banks, then the CEA’s assumption — that funds ultimately remain within the banking system — may hold. But that is only one possible outcome.
If, instead, this chain ultimately connects to the Federal Reserve’s overnight reverse repurchase agreement (ON RRP) facility, the result is very different: the cash is effectively parked at the Fed and no longer exists as a bank deposit. In that case, stablecoin reserve flows exit the banking system entirely.
ON RRP was created in 2013 as a tool to ensure that the Fed could lift the effective federal funds rate (the interest rate the Fed targets to conduct monetary policy) above zero when it decided it was appropriate to do so. Because ON RRP allows certain nonbanks (like money market funds) to place cash at the Fed overnight in exchange for Treasuries, the ON RRP rate acts as a floor, because these institutions won’t lend cash in private markets for less than what they can earn from the Fed. The Fed then uses the ON RRP (the floor) in tandem with the interest on reserve balances (IORB) rate (the ceiling) as its target policy rate.[55]
During the Fed’s tightening cycle in 2022, ON RRP usage surged dramatically, peaking at approximately $2.5 trillion.[56] This increase coincided with declining bank reserves and deposit outflows, as money market funds began offering higher yields than banks, which were slow to raise deposit rates as the Fed increased its policy rate. The result was a shift of funds out of the banking system and into money market funds — and ultimately onto the Fed’s balance sheet. As researchers at the Kansas City Fed note, rising ON RRP usage was driven not only by deposit migration, but also by limited investment alternatives and changes in money market structure.[57] Similarly, analysis from the Atlantic Council highlights how money market funds increasingly preferred placing cash at the Fed — where they face virtually no counterparty risk — rather than lending to private-sector borrowers.[58]
Although ON RRP balances have since declined as conditions normalized, the episode demonstrates an important point: under certain market and policy conditions, large quantities of cash can migrate out of the banking system and onto the Fed’s balance sheet. In a future tightening cycle — particularly one involving a much larger stablecoin market — the same dynamics could reemerge.
In such an environment, stablecoin reserve flows could amplify this shift. Because stablecoin reserves are heavily intermediated through money market funds, they are already structurally linked to ON RRP-eligible counterparties. If yields on ON RRP are attractive relative to private markets, or if counterparty risk rises during periods of stress, those funds may preferentially flow to the Fed rather than back into the banking system.
The implication is straightforward but important: the CEA’s assumption that stablecoin reserve flows leave aggregate bank deposits unchanged does not hold in a world where ON RRP plays a central role. Under plausible market conditions, those flows can reduce deposits in the banking system, with direct consequences for bank funding and lending.
Policy Achilles Heel: Master Accounts and IORB
The CEA analysis also fails to account for potential policy changes that could significantly amplify these effects.
One such change is granting stablecoin issuers access to Federal Reserve master accounts. A master account would allow issuers to hold reserves directly at the Fed, bypassing commercial banks entirely. In that scenario, stablecoin growth could directly displace bank deposits if issuers choose to hold a meaningful share of reserves at the Fed, although the attractiveness of this option would be limited if those balances earn no interest.
The CEA’s paper contemplates a related scenario in which stablecoin issuers place reserves at commercial banks that are required to back those deposits one-for-one with central bank reserves or near-reserve assets. This strict treatment — while not required under the GENIUS Act — effectively sequesters those deposits from the credit intermediation process. The CEA notes that only 12 percent of stablecoin reserves were held in cash as of 2025, but that figure is not fixed.[59] As Better Markets observes, issuers could shift significantly toward cash if regulators impose stricter custody requirements, if redemption pressures increase, or if conditions in Treasury markets change.[60]
In the CEA’s own worst-case scenario — where all stablecoin reserves are held in segregated deposits and the Fed operates in a scarce-reserves regime — a ban on stablecoin yield increases bank lending by $531 billion.[61] But that scenario becomes far more plausible if stablecoin issuers are given direct access to the Fed. Already, at least one crypto firm (Kraken) has obtained a master account[62], and multiple stablecoin issuers have applied for — or received — OCC national trust charters, making them eligible to do the same.[63]
That access could expand further. Federal Reserve Board Governor Waller has floated the idea of “skinny” master accounts, which would allow nonbanks to hold balances at the Fed, albeit with potential caps.[64] Even limited access would meaningfully change the transmission mechanism the CEA assumes.
More consequential still is the possibility that stablecoin issuers with master accounts could earn interest on reserve balances. The Financial Services Regulatory Relief Act of 2006 authorizes the Fed to pay interest on reserves, but only to depository institutions. National trust banks — which are not permitted to take deposits — are therefore not currently eligible to receive IORB.[65] However, this constraint is not immutable. On March 2, 2026, the Office of the Comptroller of the Currency (OCC) finalized amendments to its existing rule on national bank chartering to affirm the authority of national trust banks to engage in non-fiduciary activities, including certain custody and safekeeping activities.[66] Given the crypto industry’s growing influence over policymakers and Congress, it is not difficult to envision regulatory or legislative changes that would allow stablecoin issuers to earn IORB — whether through reinterpretation, rulemaking, or statutory amendment.
If stablecoin issuers can hold reserves directly at the Fed — and especially if they can earn interest on those balances — the CEA’s core assumption that deposits simply “recycle” within the banking system breaks down entirely. Under those conditions, stablecoin growth would displace bank deposit funding and reduce lending.
Conclusion
The CEA report falls flat for another reason as well: it treats stablecoins as safer than the real-world record suggests. The report describes them as “safe assets” that “may help guard against runs and improve the stability of the payment system.” The truth is much closer to the opposite. In 2023, BIS economists found that not a single fiat-backed stablecoin had maintained parity with its peg at all times, regardless of size or backing, and that there was “currently no guarantee” issuers could redeem in full and on demand.[67] USDC’s depeg during the 2023 Silicon Valley Bank failure is the clearest illustration. Circle disclosed that $3.3 billion of USDC reserves were stuck at SVB, and USDC fell as low as $0.88 before recovering after U.S. authorities intervened to guarantee all uninsured deposits at SVB.[68] Absent that intervention, a run on USDC was likely, and its survival would have been far from assured.
The BIS has since sharpened the critique. In its 2025 Annual Economic Report, BIS concluded that stablecoins “fall short” of the requirements for serving as the backbone of the monetary system when judged against the three core tests of singleness, elasticity, and integrity. They fail singleness because they are tagged to specific issuers and can trade at varying exchange rates, undermining the “no questions asked” principle of money. They fail elasticity because their cash-in-advance structure requires full upfront funding rather than allowing the kind of balance-sheet expansion and liquidity provision that banks and central banks provide. And they fail integrity because their bearer-like features, self-hosted wallets and pseudonymous public blockchains complicate anti-money-laundering and other safeguards.[69] Stablecoins may well persist as a niche or subsidiary instrument. But they are not sound money, and they are certainly not a reason to be cavalier about impairing bank credit intermediation.
The only way to ensure these limitations are less likely to create systemic problems is to close the third-party rewards loophole for stablecoin yield. A prohibition that binds only the issuer while allowing exchanges, affiliates, and other intermediaries to pass reserve income through to holders is not much of a prohibition at all. If stablecoins are going to be treated as “payment stablecoins,” then Congress should not allow them to become yield-bearing deposit substitutes through semantic gamesmanship and distribution agreements. Preserving valuable bank credit intermediation, especially at community banks, and protecting consumers from instruments that have repeatedly failed to behave like genuinely stable money requires banning third parties from paying rewards on stablecoins.
About
Lee Reiners is a Lecturing Fellow at the Duke Financial Economics Center.* At Duke, Reiners has taught classes on FinTech Law and Policy, Cryptocurrency Law and Policy, Financial Regulatory Policy, Climate Change and Financial Markets, and Cybersecurity Law and Policy. Reiners is widely recognized for his commentary and analysis on cryptocurrency regulation, and he has testified three times on the subject before the U.S. Congress.
Mark Hays is the Associate Director for Cryptocurrency & Financial Technology at Americans for Financial Reform Education Fund (AFREF) and Demand Progress. AFREF is an independent, nonprofit organization founded by a coalition of more than 200 civil rights, community-based, consumer, labor, small business, investor, faith-based, civic groups, and individual experts. AFREF fights for a fair and just financial system that contributes to shared prosperity for all families and communities.
*Title and affiliation are provided for identification purposes only. Any views expressed are the author’s own and do not reflect the views of Duke University or its affiliate
ENDNOTES
[1] White House Council of Economic Advisers. “Effect of Stablecoin Yield on Banking Lending.” April 2026.
[2] 12 USC § 5903 (a)(11).
[3] Grewal, Paul. Coinbase. “The CEA Report is finally out.” X.com. April 8, 2026.
[4] “Vault: Tillis-Alsobrooks cinch deal on stablecoin yield.” Punchbowl News. May 1, 2026.
[5] Shirzad, Faryar. Coinbase. “The final rewards text in the CLARITY Act is now public.” X.com. May 1, 2026.
[6] Reiners, Lee A. et al. “Political Capital: How the Trump Family’s Crypto Ventures Are Distorting the Market.” Democracy Defenders Fund. March 11, 2026.
[7] Circle Internet Group, Inc. Securities and Exchange Commission (SEC). Form 10-K. Fiscal year-ending (FYE) December 31, 2025 at 68.
[8] Binance. [Announcement]. “Binance USD1 Booster Program: Enjoy Up to 20% APR with USD1 Flexible Products – Limited at 50,000 USD1.” Binance Square, December 23, 2025.
[9] Kharif, Olga and Annie Massa. “Binance Weighs US Strategy as Billionaire Zhao Regains Clout.” Bloomberg. December17, 2025.
[10] Braun, Helene. “Binance.US plots comeback with derivatives, prediction markets push.” CoinDesk. April 6, 2026.
[11] Minority Press Release. U.S. Senate Committee on Banking, Housing and Urban Affairs. ”Warren on Fed Nominee Stephen Miran: If We Allow the Fed to Become a Political Football, Americans Will Pay the Price.“ September 24, 2025.
[12] Grossman, Matt. ”Stephen Miran, Temporary Fed Governor, Resigns From White House Role.” Wall Street Journal. February 3, 2026.
[13] Press Releases. Federal Reserve. ”Monthly Monetary Policy Meeting Summaries.” September 17, 2025, October 29, 2025, December 19, 2025, January 28, 2026, March 18, 2026, April 29, 2026.
[14] Cox, Jeff. “Fed holds rates steady but with highest level of dissent since 1992.” CNBC.Com. April 29, 2026.
[15] White House. “2026 Economic Report of the President.” April 2026 at Chapter 10: The Economic Consequences of DEI.
[16] Ip, Greg. “White House Study Says DEI Hurts Productivity.” Wall Street Journal. April 13, 2026.
[17] White House Council of Economic Advisors. “Estimating the Cost of the Consumer Financial Protection Bureau to Consumers.” February 17, 2026.
[18] Levitin, Adam. “The Council of Economic Advisers Discredits Itself.” Credit Slips. February 17, 2026.
[19] Levitin, Ibid.
[20] Cowley, Stacey. ”Trump Administration Scales Back Plan to Dismantle Consumer Protection Bureau.” New York Times. April 1, 2026.
[21] Cowley, Ibid.
[22] Minority Staff Report. U.S. Senate Committee on Banking, Housing and Urban Affairs. “Left in Limbo: How President Trump’s Efforts to Dismantle the Consumer Financial Protection Bureau Force Consumers to Face Corporate Greed Alone.“ February 2025.
[23] Stone, Peter. “Critics slam deregulation of crypto as Trump family expands its footprint in industry.” The Guardian. April 14, 2025.
[24] Srinivasan, Sayee and Yikai Wang. “The CEA studied the wrong question on stablecoin ‘yield’ and community banks.” ABA Banking Journal. April 13, 2026.
[25] Coinbase. “Earn unlimited 3.50% rewards on USDC exclusively with Coinbase One.” Available at https://www.coinbase.com/usdc. Accessed May 4, 2026.
[26] Board of Governors of the Federal Reserve System. ”Deposits, All Commercial Banks [DPSACBW027SBOG].” Federal Reserve Bank of St. Louis as of May 4, 2026.
[27] Srinivasan, Ibid.
[28] Singh, Rashika. “J.P. Morgan wary of stablecoin’s trillion dollar growth bets, cuts them by half.” Reuters. July 3, 2025.
[29] Atkins, Paul. “Speech: American Leadership in the Digital Revolution.” Securities and Exchange Commission. July 31, 2025.
[30] Skroupa, Christopher P. “The Wells Fargo Fake Account Scandal: A Timeline.” Forbes. April 28, 2017.
[31] Allison, Ian and AI Boost. “Coinbase Adds USDC Lending With Morpho and Steakhouse Financial.” CoinDesk. September 18, 2025.
[32] Zhou, JJ. “Pioneering the DeFi Mullet: Coinbase in the Front, DeFi, in the Back.” Morpho Blog. May 6, 2025.
[33] Circle Internet Group, Inc. [Transcript]. Fourth Quarter 2025 Earnings Call. February 25, 2026.
[34] Thorn, Alex. “CLARITY ACT – text of tillis (R) / alsobrooks (D) compromise on stablecoin yield is out now.” X.com. May 1, 2026 (draft text posted as PDF image).
[35] Thorn, Ibid.
[36] Thorn, Ibid.
[37] Thorn, Ibid.
[38] Securities and Exchange Commission (SEC). Division of Trading and Markets. “Staff Statement Regarding Broker-Dealer Registration of Certain User Interfaces Utilized to Prepare Transactions in Crypto Asset Securities.” April 13, 2026.
[39] WH CEA ”Stablecoin Yield”, Ibid.
[40] Circle. “Transparency and stability.” Available at https://www.circle.com/transparency. Accessed on May 4, 2026.
[41] Basil, Phillip and Amanda Fischer. “Policymakers Shouldn’t Yield on Stablecoin Interest.“ Better Markets. April 2026.
[42] Castilla, Jill. LinkedIn Post. April 13, 2026.
[43] Basil and Fischer, Ibid.
[44] Srinivasan, Ibid.
[45] WH CEA ”Stablecoin Yield”, Ibid.
[46] Kelly, David. “The Investment Implications of Shrinking the Fed’s Balance Sheet.“ J.P. Morgan Asset Management, Notes on the Week Ahead. March 24, 2026.
[47] Warsh, Kevin. Member and Chairman Designate, Board of Governors of the Federal Reserve System. Nomination Hearing. U.S. Senate. Testimony before the Committee on Banking, Housing, and Urban Affairs. April 21, 2026.
[48] Smith, Colby. “How Kevin Warsh Could Shrink the Fed’s Footprint in Financial Markets.” New York Times. April 24, 2026.
[49] WH CEA ”Stablecoin Yield”, Ibid.
[50] Guida, Victoria and Sam Sutton. “Trump picks Kevin Warsh to lead Federal Reserve, tapping central bank critic.” Politico. January 30, 2026.
[51] Circle, ”Transparency and stability.“, Ibid.
[52] Tether. “Transparency.” Available at https://tether.to/en/transparency/?tab=reports. Accessed on May 4, 2026.
[53] Circle Internet Group, SEC10-K, Ibid. at 15.
[54] BlackRock. [Statement] “BlackRock Funds – Circle Reserve Fund.” 2025 Annual Financial Statements and Additional Information, April 30, 2025, at 3.
[55] Afonso, Gara et al. “How the Fed’s Overnight Reverse Repo Facility Works.” Liberty Street Economics, Federal Reserve Bank of New York. January 11, 2022.
[56] Federal Reserve Bank of New York. ”Overnight Reverse Repurchase Agreements: Treasury Securities Sold by the Federal Reserve in the Temporary Open Market Operations [RRPONTSYD].“ retrieved from FRED, Federal Reserve Bank of St. Louis. May 5, 2026.
[57] Marsh, W. Blake and Rajdeep Sengupta. “Is Bank Capital Regulation Driving Continued Use of the Reverse Repurchase (ON RPP) Facility?” Economic Bulletin, Federal Reserve Bank of Kansas City. December 16, 2022.
[58] Tran, Hung. “Fed reverse repos hit a new record: An unhealthy development.” Atlantic Council. January 9, 2023.
[59] WH CEA ”Stablecoin Yield”, Ibid.
[60] Basil and Fischer, Ibid.
[61] WH CEA ”Stablecoin Yield”, Ibid.
[62] Kraken. [Release]. “Kraken becomes first digital asset bank to receive a Federal Reserve master account.” March 4, 2026.
[63] Gannon, Pet. “OCC conditionally approves 5 crypto firms for national trust bank charters.” Axios. December 12, 2025.
[64] Lutz, Sander. “Fed Pushing Ahead With ‘Skinny’ Master Account Plan for Crypto Banks.” Decrypt/Yahoo Finance. December 19, 2025.
[65] Financial Services Regulatory Relief Act of 2006. Public Law 109-351. October 13, 2006 at § 201.
[66] 91 Fed. Reg. 40. March 2, 2026 at 9977 to 9982.
[67] Kosse, Anneke et al. “Will the real stablecoin please stand up?” Bank of International Settlements, BIS Papers No 141. November 2023.
[68] Howcroft, Elizabeth and Rishabh Jaiswal. ”Circle assures market after stablecoin USDC breaks dollar peg.” Reuters. March 11, 2023.
[69] Bank of International Settlements. “Annual Economic Report 2025.” June 29, 2025 at Part III. The next-generation monetary and financial system.