A research paper published this week quantifies what happens when bank deposits move at blockchain speed. The number is large enough to reshape how banks fund the economy, and the crypto industry is building the pipes without acknowledging the consequences.
Summary
- A research paper published on August 25 found that tokenized deposits could reduce U.S. bank lending capacity by $580 billion if the technology reaches widespread adoption, roughly 5% of total bank lending.
- The mechanism is straightforward: banks lend against stable deposits, and if deposits can move on chain in minutes instead of days, the deposit base becomes less stable, forcing banks to hold more liquid reserves and lend less.
- LayerZero and Keeta launched tokenized bank deposits across Ethereum, Solana, Base, and Keeta in July 2026, covering nine fiat currencies and making the theoretical risk operationally real.
- The Bank of England endorsed tokenized deposits as belonging in UK payments infrastructure, and South Korea began trialing them for government spending, indicating that adoption pressure is coming from regulators, not just startups.
- The $580 billion figure assumes a moderate adoption scenario. The paper’s high adoption model projects a reduction of $1.2 trillion in lending capacity, a number that would force structural changes to how U.S. banks fund mortgages, small business loans, and commercial real estate.
The crypto industry has spent two years building infrastructure to put bank deposits on chain. The banking industry has spent two years worrying about what happens when it works. A new research paper puts a number on the worry, and the number is large enough that both sides should be paying closer attention.
How bank lending actually works
This section requires explaining something that most crypto coverage skips entirely: the mechanics of fractional reserve banking and why deposit stability is the load bearing wall of the entire system.
When a customer deposits $1,000 at a bank, the bank does not keep $1,000 in a vault. It keeps a fraction, typically 3% to 10% depending on the bank’s risk profile and regulatory requirements, and lends the rest. That $900 or $970 goes to a mortgage borrower, a small business, or a commercial real estate developer. The borrower spends it, and the recipient deposits it at another bank, which lends most of that out again. This is the money multiplier, and it is the engine that converts $22 trillion in U.S. bank deposits into $12 trillion in bank lending.
The system works because deposits are sticky. A customer who deposits money on Monday does not withdraw it on Tuesday. The bank can rely on a statistical floor, the amount that will remain regardless of individual withdrawals, and lend against that floor with reasonable confidence.
Regulatory frameworks formalize this assumption. Basel III assigns stability scores to different deposit types. Retail deposits from individuals receive the highest stability weighting because individuals rarely move their entire balance in a single day. Corporate deposits receive lower scores because businesses manage cash more actively. Interbank deposits receive the lowest scores because banks move money constantly.
The Liquidity Coverage Ratio, a core Basel III metric, requires banks to hold enough high quality liquid assets to cover 30 days of net cash outflows under stress. The calculation assumes that retail deposits experience outflows of 3% to 10% over 30 days. Corporate deposits face outflow assumptions of 20% to 40%. These percentages determine how much of each deposit type a bank can lend out.
Tokenized deposits threaten to reclassify every deposit into the highest outflow category, because the technology makes any deposit as mobile as an interbank transfer.
What the paper found
The research paper, published on August 25, modeled three scenarios for tokenized deposit adoption in the U.S. banking system.
In the low adoption scenario, covering 5% to 10% of total deposits, the impact on lending capacity is modest: roughly $120 billion, absorbed through minor adjustments to reserve ratios and overnight funding markets. Banks would barely notice, and the adjustment would be indistinguishable from normal quarter to quarter fluctuations in deposit levels.
In the moderate scenario, covering 15% to 25% of deposits, lending capacity falls by $580 billion. This is the headline number, and it represents a meaningful contraction. To put it in context, $580 billion is roughly the total outstanding balance of U.S. auto loans, or about one third of all outstanding commercial and industrial loans. A contraction of that magnitude would not cause a crisis, but it would tighten credit availability for borrowers at the margin, precisely the small businesses and first time homebuyers who are most rate sensitive.
In the high adoption scenario, covering 35% to 50% of deposits, the reduction reaches $1.2 trillion. At that level, banks would need to fundamentally restructure their funding models, shifting from deposit funded lending to wholesale funding markets, securitization, or Federal Home Loan Bank advances. Each of these alternatives is more expensive than deposits, which means the cost of borrowing rises for everyone. The paper estimates that average mortgage rates could increase by 15 to 30 basis points under the high adoption scenario, and small business loan rates could rise by 25 to 50 basis points.
The speed problem
The critical variable is not how much deposits move, but how fast they move. Traditional bank transfers through ACH take one to three business days. Wire transfers settle within hours but cost $25 to $50 and are typically reserved for large transactions. Neither mechanism threatens deposit stability because the friction creates natural resistance to movement.
Even FedNow, the Federal Reserve’s instant payment system launched in 2023, processes transfers in seconds but imposes transaction limits and operates within the existing banking framework. A FedNow transfer moves money from one bank account to another, but both accounts remain within the banking system. The deposit leaves one bank and arrives at another, preserving the aggregate deposit base.
Tokenized deposits are different. A transfer on Ethereum’s base layer settles in roughly 12 seconds. On Solana, it takes under a second. On a Layer 2 like Base, settlement is near instantaneous for the user, with finality following within minutes. More importantly, the deposit can leave the banking system entirely, moving into DeFi protocols, smart contract escrow, or cross chain bridges where no bank holds the underlying balance.
The paper models the impact of settlement speed directly. At one day settlement, the effect on deposit stability is negligible. At one hour settlement, it becomes measurable. At near instant settlement, which is what blockchain infrastructure provides, the deposit stability models that underpin Basel III capital requirements break down entirely, because the statistical assumptions about how long deposits remain were calibrated for a world where moving money takes days, not seconds.
This is not a theoretical concern. The tokenized deposit infrastructure is already live. LayerZero and Keeta deployed tokenized bank deposits across four chains in July 2026. USBC, Uphold, and Vast Bank launched the first retail tokenized U.S. dollar deposits in late 2025. The pipes exist. The question is how much volume they carry and how quickly that volume grows.
Who is building this and why
The builders fall into three categories, each with different motivations and different risk profiles.
Fintech infrastructure companies like LayerZero and Keeta are building the plumbing. Their business model is transaction fees and protocol revenue. More deposit movement means more revenue. They have no incentive to consider the systemic effects on bank lending because those effects are externalities, costs borne by borrowers and the broader economy while the revenue flows to the infrastructure provider.
LayerZero’s deployment covers nine fiat currencies across four blockchains. Keeta’s architecture allows any bank to issue tokenized deposits on its platform, abstracting the blockchain layer so that depositors interact with a familiar banking interface while their funds exist as on chain tokens. The cross chain interoperability means a deposit tokenized on Ethereum can move to Solana in minutes, a level of fungibility that traditional banking infrastructure cannot match.
Banks themselves are experimenting cautiously. JPMorgan’s Kinexys platform processes tokenized deposit transfers between institutional counterparties. MUFG, SMBC, and Mizuho in Japan are piloting tokenized government bonds settled through tokenized central bank reserves. The Bank of Japan’s sandbox uses tokenized central bank reserves as the settlement asset, which is as close to a central bank digital currency as Japan has come without officially launching one. These pilots are controlled environments with known counterparties and limited scale, but the technology they validate is the same technology that, at scale, could destabilize their own deposit bases.
Regulators are the wild card. The Bank of England explicitly endorsed tokenized deposits as part of UK payments infrastructure. Sarah Breeden, the Bank’s deputy governor for financial stability, said tokenized deposits belong in the UK’s future payments architecture alongside stablecoins and a potential digital pound. South Korea is trialing tokenized deposits for government operational spending. The GENIUS Act’s stablecoin framework implicitly endorses the underlying technology by creating a regulated category for digital dollars that compete with bank deposits for the same customer balances.
Regulators are simultaneously promoting the technology and responsible for managing the systemic risk it creates. The contradiction is not lost on central bankers, but the competitive pressure from China’s digital yuan pilots and the private sector’s first mover advantage leaves regulators feeling that the alternative to managed adoption is unmanaged adoption, which is worse.
The stablecoin connection
Tokenized deposits and stablecoins are often discussed as competitors, but the systemic risk analysis reveals them as complements that amplify the same underlying pressure on bank balance sheets.
Stablecoins like USDC and USDT are backed by Treasury bills, commercial paper, and bank deposits. When a user buys $1,000 of USDC, Circle deposits that $1,000 at a partner bank. The bank lends against it. The deposit is still in the banking system; it has just been intermediated through a stablecoin issuer. Circle’s reserve management acts as a buffer, because Circle does not withdraw its deposits based on individual user redemptions. It manages aggregate flows, smoothing the volatility.
Tokenized deposits cut out the intermediary. When a user holds a tokenized deposit, they hold a direct claim on the bank. There is no stablecoin issuer sitting between the depositor and the bank. That directness is marketed as an advantage, eliminating counterparty risk from the stablecoin issuer, but it also means the depositor can withdraw at blockchain speed without Circle or Tether serving as a shock absorber.
The Revolut stablecoin launch in Europe illustrates the competitive dynamics. Revolut has 50 million users who can now hold euros in a stablecoin form. If those users shift from bank deposits to Revolut’s stablecoin or to tokenized deposits, the net effect on European bank lending capacity follows the same pattern the research paper describes for the U.S.
The combined effect of stablecoins and tokenized deposits is larger than either alone. Stablecoins pull deposits out of the banking system and into reserve managed pools. Tokenized deposits keep deposits in the banking system but make them volatile. Both reduce the stable deposit base that banks use to justify long term lending.
The section a competitor could not write
Every existing analysis of tokenized deposits focuses on either the technology (how they work) or the opportunity (how much faster payments become). This piece examines the second order effect that neither the crypto industry nor the banking industry wants to discuss openly.
The crypto industry does not want to discuss it because acknowledging that tokenized deposits reduce lending capacity undermines the narrative that blockchain technology is purely additive. If putting deposits on chain means fewer mortgages, fewer small business loans, and higher borrowing costs, the political and regulatory response will be hostile. The industry has spent years arguing that crypto creates new financial access. The research paper suggests it could restrict existing access by destabilizing the lending infrastructure that funds the real economy.
The banking industry does not want to discuss it because acknowledging the risk validates the technology’s power. If tokenized deposits are not a threat to deposit stability, there is no reason to oppose them. If they are a threat, it means the technology works exactly as described, moving money faster and more efficiently than legacy rails. That admission attracts more investment, more builders, and faster adoption, accelerating the very dynamic banks fear.
The research paper breaks this silence by quantifying the cost. $580 billion in reduced lending capacity is not an existential threat to the U.S. banking system, but it is large enough to change behavior. Banks would need to raise deposit rates to retain customers, increase wholesale funding at higher cost, or reduce lending to lower risk categories. All three responses have consequences for borrowers who depend on affordable credit.
What the Fed would do
The Federal Reserve has not publicly addressed the research paper’s findings, but the institutional response is predictable based on how the Fed handled previous deposit stability threats, including the money market fund reforms of 2010 and 2014 and the SVB deposit flight crisis of 2023.
If tokenized deposit adoption reaches the moderate scenario, the Fed would likely adjust Liquidity Coverage Ratio requirements to classify tokenized deposits as less stable than traditional deposits, assigning them outflow rates of 40% to 60% instead of the 3% to 10% applied to standard retail deposits. This would increase the amount of high quality liquid assets banks must hold against tokenized deposit balances, effectively pricing in the faster withdrawal risk and reducing the lending capacity impact by forcing banks to hold more reserves from day one.
The Fed could also impose holding period requirements or withdrawal speed limits on tokenized deposits, similar to the gates and fees that money market funds implemented after the 2008 financial crisis and strengthened after the March 2020 liquidity stress. These measures would reduce the systemic risk but would also eliminate the speed advantage that makes tokenized deposits attractive in the first place, potentially killing adoption.
A more creative response would involve the Fed launching its own tokenized settlement system through FedNow or a future central bank digital currency, allowing deposits to move quickly within a system the Fed controls and monitors in real time. This would preserve the speed benefit while keeping the systemic risk management within the central bank’s perimeter.
What would prove this analysis wrong
Three developments would invalidate the $580 billion projection.
First, if tokenized deposits adopt voluntary speed limits, settling in hours instead of seconds, the deposit stability impact drops sharply. Some implementations already include programmable settlement delays that can be configured by the issuing bank. If these become standard, the paper’s extreme speed scenarios do not materialize, and the impact reverts to the low adoption model even at higher volume.
Second, if banks create new lending products specifically designed for volatile deposit bases, the lending capacity reduction could be offset. Variable rate loans that reprice in real time, for example, would match asset duration to the shorter deposit duration, preserving lending volume at the cost of transferring interest rate risk to borrowers. Overnight repo style lending, already common in institutional markets, could expand to consumer credit.
Third, if adoption stalls below 10% of total deposits, the low scenario applies and the impact is within the range that existing capital buffers can absorb without behavioral changes. Adoption is not guaranteed to reach the moderate scenario, and the friction of opening tokenized deposit accounts may limit uptake to technologically sophisticated users who represent a small fraction of total deposits.
What to watch
LayerZero and Keeta transaction volume. These platforms provide the clearest real time signal of how fast tokenized deposit adoption is growing. Monthly volume crossing $10 billion would put the system in the low adoption scenario. $100 billion would approach moderate.
Fed commentary on deposit stability. Any mention of tokenized deposits in Federal Reserve speeches, meeting minutes, or Financial Stability Reports would signal that the $580 billion scenario has entered the regulatory conversation. Watch the November 2026 Financial Stability Report specifically.
Basel Committee updates. The Basel Committee on Banking Supervision reviews capital requirements for digital assets periodically. A reclassification of tokenized deposits in the liquidity coverage ratio framework would be the first regulatory acknowledgment of the speed risk at the global standard setting level.
Bank deposit rate movements. If major U.S. banks begin raising deposit rates in markets where tokenized deposit alternatives are available, it would confirm that deposit competition is already affecting bank behavior, even before adoption reaches the paper’s moderate scenario.
Central bank digital currency timelines. A Fed CBDC or expanded FedNow tokenized settlement system would provide a government controlled alternative to private tokenized deposits, potentially capping adoption of private solutions at a level below the paper’s risk thresholds.
What are tokenized deposits?
Tokenized deposits are bank deposits represented as digital tokens on a blockchain. They give the depositor a direct claim on the issuing bank, the same as a traditional deposit, but allow transfers at blockchain speed instead of through traditional banking rails like ACH or wire transfers.
How could tokenized deposits reduce bank lending?
Banks lend against stable deposits, relying on statistical models that assume most depositors will not withdraw their money on any given day. If deposits can move in seconds instead of days, the deposit base becomes less predictable. Banks must hold more liquid reserves to cover faster potential withdrawals, leaving less capital available for loans.
How much lending could be affected?
A research paper published August 25, 2026, projects that moderate adoption of tokenized deposits could reduce U.S. bank lending capacity by $580 billion, roughly the total outstanding balance of U.S. auto loans. High adoption could reduce it by $1.2 trillion.
Are tokenized deposits the same as stablecoins?
No. Stablecoins are issued by non bank entities like Circle or Tether and backed by reserves including Treasury bills and bank deposits. Tokenized deposits are issued by banks and represent a direct deposit claim. Stablecoins add an intermediary between the depositor and the bank. Tokenized deposits remove it, giving the depositor direct access to withdraw at blockchain speed.
Which banks are experimenting with tokenized deposits?
JPMorgan runs Kinexys for institutional tokenized transfers. In Japan, MUFG, SMBC, and Mizuho are piloting tokenized government bonds settled through tokenized central bank reserves. Vast Bank in the U.S. launched the first retail tokenized dollar deposits in late 2025. LayerZero and Keeta deployed multi chain infrastructure covering nine fiat currencies in July 2026.
Would the Federal Reserve intervene?
The Fed has not publicly addressed the research. Based on precedent from money market fund reforms and the SVB crisis response, the Fed would likely adjust liquidity requirements, impose settlement speed limits, or launch its own tokenized settlement system if adoption reaches levels that threaten deposit stability.
How fast can tokenized deposits move?
On Ethereum, settlement takes roughly 12 seconds. On Solana, under one second. On Layer 2 networks like Base, near instantly from the user’s perspective. This speed, compared to one to three business days for ACH transfers, is what makes tokenized deposits both attractive as a product and risky as a systemic factor.
Should I be concerned about tokenized deposits?
This is educational analysis, not investment advice. Tokenized deposits offer faster payments and broader access to banking services. The systemic risk to bank lending is real but depends on adoption rates that remain uncertain. The technology is in early deployment, and regulatory responses will shape outcomes significantly over the next two to three years.
Disclosure: This content is provided by a third party. Neither crypto.news nor the author of this article endorses any product mentioned on this page. Users should conduct their own research before taking any action related to the company.
