As yield-bearing stablecoins grow, do bank deposits shrink, or just move to different places? What then happens to bank loans? The answer to those questions is key as Congress and regulators craft a regulatory framework for stablecoins in pending legislation and rulemakings.
Some public commentary has suggested that the growth in yield-bearing stablecoins backed by Treasury bills will not reduce the level of bank deposits because the deposits would just move around, and therefore that the total amount of deposits would not go down. For example, a recent Reuters commentary stated: “The first problem with the banks’ argument is that it’s not possible to pull deposits from the system on aggregate.”[1] That is incorrect. If the stablecoin offers transaction services similar to bank deposits as well as yield, the coins will compete with deposits in the provision of those services, but the coins will not be a source of funding for loans to businesses and households. Once the dust is settled and interest rates and balance sheets adjust, deposits will decline and so will bank lending.
Transaction-Based Thinking
The incorrect story that deposits will not be reduced typically starts with an example tracing out a chain of transactions. That’s the problem. An evaluation of how the financial system will adjust to the introduction of yield-bearing stablecoins requires examining the big picture: How will interest rates adjust? How will everyone rebalance their assets and liabilities? Tracing transactions can lead you to whatever outcome you wish.
For example, a recent paper by the Council of Economic Advisors started out with this story:
Suppose a household withdraws $1 from Bank A and purchases a stablecoin. The issuer receives the dollar and buys a Treasury bill from a dealer. The dealer, having sold the bill, deposits the $1 proceeds at Bank B. The result: Bank A loses a deposit, Bank B gains one. Aggregate deposits in the banking system are effectively unchanged—they have simply moved from one institution to another.[2]
But consider what might happen next. Suppose, as seems reasonable, everybody started off with portfolios of assets and liabilities they were happy with. They will then seek to return to similar portfolios.
The dealer now needs to replenish its stock of Treasury bills, and it has more deposits than it wants, so it buys Treasury bills from the household reducing the deposits at Bank B and raising them at Bank A. The household now has more deposits than it wants (because it also has stablecoins), so it uses some of the deposits to repay an outstanding loan from Bank A. Deposits and lending fall.
The second set of transactions does not prove that stablecoins reduce deposits, but nor does the first prove that deposits are unchanged.
System-Based Thinking
Alternatively, consider the problem using the analysis in “Stablecoins and Banking: Deposit Dynamics, Financial Stability, and Regulatory Design,” (2025) by Lin William Cong with financial support from Coinbase, Paradigm, PayPal and Stripe. As the author correctly notes, the consequences extend beyond the initial transaction: “When a household acquires a stablecoin, the corresponding deposit liability typically shifts from the buyer’s bank to the issuer’s bank, and only the issuer’s subsequent reserve-management choices — such as investing in T-bills or reallocating funds across custodians — determine whether aggregate deposits fall.”[3] Cong calibrated his model to match the behavior of the U.S. financial system.
In the paper, Cong argues that the threat of stablecoins that pay interest will lead banks to compete more aggressively for deposits, boosting deposits by 1.5 percent, that’s about $300 billion for current levels of deposits ($19 trillion). That $300 billion boost occurs when there are zero stablecoins, just the threat of stablecoins.[4] But because in the model yield-bearing stablecoins and deposits are perfect substitutes, once the stablecoins actually begin to circulate, they reduce deposits. When the amount of stablecoins goes up, deposits don’t just move around, the amount of deposits goes down, dollar for dollar.[5]
Cong states: “Stablecoin adoption is believed to expand substantially, reaching roughly four trillion dollars by 2030.”[6] In that case, using this analysis sponsored by the stablecoin industry, deposits would therefore decline by $4 trillion. Deposits would have first risen by $300 billion then fallen by $4 trillion, so that’s $3.7 trillion in destroyed deposits, far from deposits just moving around.
Moreover, according to the paper’s model, calibrated to match the behavior of the U.S. financial system, loans decline by the same percentage as deposits. In the projected future with $4 trillion of stablecoins, deposits would decline by 19 percent on net, so loans would also decline by 19 percent. That’s a $2.7 trillion reduction in lending.[7]
These outcomes are implausibly large, but analysis published at about the same time by an economist at the Federal Reserve Board – “Banks in the Age of Stablecoins: Some Possible Implications for Deposits, Credit, and Financial Intermediation,” by Jessie Jiaxu Wang – estimated that the introduction of stablecoins would destroy deposits and reduce lending by somewhere between $65 billion and $1.26 trillion, with the high end occurring if stablecoin issuers get access to master accounts at the Fed.[8]
Why Such Different Outcomes?
As noted, the incorrect intuition that deposits are just moved around arises from thinking about the problem by tracing through the transactions and stopping the analysis there: I send you a payment, my bank deposit goes down and yours goes up. That obvious result has led some to conclude that deposits cannot be created or destroyed, they can only be moved around. However, there is no “law of the conservation of deposits.”
Many transactions create or destroy deposits. When a bank extends a loan, it does so by providing the borrower a deposit. Those are newly created deposits. The bank’s assets and liabilities go up by the same amount. If a customer repays a loan, he or she reduces his or her deposits. Deposits are destroyed. If a bank issues commercial paper, the deposits of the investor are destroyed. The bank’s assets are not changed, but its liabilities have shifted from deposits to commercial paper.
In fact, the idea that deposits can be created or destroyed is embedded in economists’ understanding of how the financial system and macroeconomy works. In the money multiplier that many of us learned in undergraduate economics, bank loans create deposits. Moreover, economists model the demand for money by households and businesses as depending on economic activity and the opportunity costs of holding deposits. The whole premise is that businesses and households choose their desired level of deposits based on transaction need and opportunity cost (the interest rate on the deposits less the interest rate on alternatives like Treasury bills or money market funds).
That is, one needs to view the problem from the 30,000-foot level and take into account how everything will adjust. The foundational change is that yield-bearing stablecoins that provide transaction services similar to deposits are introduced. That change will spark a nearly infinite sequence of transactions that play out over time. Once everything settles, businesses and households will meet some of their transaction needs with stablecoins and fewer with deposits. Stablecoin issuers will grow and banks will shrink. While stablecoins only fund a portfolio of Treasury bills (in this case), deposits at banks fund both Treasury bills and loans. The new financial system supports less lending.
Simultaneously, interest rates all adjust so that everyone is content with their portfolios of assets and liabilities in the new equilibrium. The interest rate on Treasury bills will fall, while deposit rates and loan rates will rise.
And of course, stories involving one household or one bank making decisions are just metaphors for how the millions of households and businesses and the thousands of banks and other financial institutions will behave. It’s not the same individual household that buys a stablecoin that sells a Treasury bill. In the stories we economists tell, the household is standing in for all depositors, all potential purchasers of stablecoins, and all nonbank investors in Treasury bills.
While this note is focused on the result that deposits will be destroyed, a separate problem is that as interest-bearing stablecoins grow, even if issuers invest in bank deposits, deposits will move around and take different forms. As described in the ECB paper “Toss a stablecoin to your banker,” if stablecoin issuers invest in bank deposits (as allowed under the Genius Act), liquidity regulations will treat those deposits (which likely will massively exceed the FDIC deposit insurance limit of $250,000) as unreliable under stress and require banks to back them with an equal amount of liquid assets (e.g., Treasury bills). That is, the deposits at banks from stablecoin issuers cannot fund loans. In the end, the outcome would be similar: lending would decline.
[1] “Why US banks’ stablecoin pleading is flawed,” Stephen Gandel, February 12, 2026. https://www.reuters.com/commentary/breakingviews/why-us-banks-stablecoin-pleading-is-flawed-2026-02-13/
[2] “Effects of Stablecoin Yield Prohibition on Bank Lending,” p. 2. https://www.whitehouse.gov/research/2026/04/effects-of-stablecoin-yield-prohibition-on-bank-lending/
[3] Page 12.
[4] As we observe in “Even Crypto-Funded Research Affirms That Yield-Bearing Stablecoins Reduce Bank Deposits and Lending,” this initial rise where the latent threat of an interest-bearing stablecoin increases deposits and loans is premised on the view that each bank has some monopoly power in the supply of deposit accounts. That premise is implausible when there are over 9,000 depository institutions in the United States and a new deposit account is merely a mouse-click or phone-tap away for most consumers. https://bpi.com/wp-content/uploads/2026/01/Even-Crypto-Funded-Research-Affirms-That-Yield-Bearing-Stablecoins-Reduce-Bank-Deposits-and-Lending.pdf
[5] Cong does not provide the details of his model in his paper, but he states that he uses the model in “Bank Market Power and Central Bank Digital Currency: Theory and Quantitative Evidence,” by Jonathan Chiu, Seyed Mohammadreza Davoodalhosseini, Janet Jiang, and Yu Zhu, February 20, 2022. The analysis by Chiu and colleagues concerns the impact of an interest-bearing central bank digital currency (CBDC). Cong uses the model to analyze the impact of interest-bearing stablecoins. In the Chiu model, households consider deposits and CBDC to be perfect substitutes (p. 6), so we assume households in Cong’s model treat deposits and interest-bearing stablecoins as perfect substitutes. Households choose the amount of electronic payment instruments (deposits or stablecoins) to hold, and then divide that amount between the two instruments depending on their yields.
[6] Page 10.
[7] In the model used by Chiu and his colleagues (and therefore, we think, Cong, although he doesn’t share his precise model) loans end up being a simple multiple of deposits (p. 16 of Chiu and colleagues). They also find that if stablecoins grow too large, banks will all go out of business and there will be no lending. It may be that the “all banks fail” outcome occurs when there are $4 trillion in stablecoins circulating (the amount Cong cites as expected). Neither paper provides sufficient information to know.
[8] Wang, Jessie Jiaxu, “Banks in the Age of Stablecoins: Some Possible Implications for Deposits, Credit, and Financial Intermediation,” Federal Reserve, Dec. 17, 2025.
