SWIFT built a ledger for them. JPMorgan settles billions with them. The FDIC is writing rules about them. Tokenized deposits are the banking system’s answer to stablecoins, and understanding the difference decides how you read the next five years of digital money.
Summary
- A tokenized deposit is a commercial bank deposit represented as a token on a blockchain, issued by the bank itself and maintaining a one-to-one relationship with money on the bank’s balance sheet.
- Unlike a stablecoin, the money never leaves the bank. It stays available to fund lending, remains covered by deposit insurance up to statutory limits, and stays inside the supervisory perimeter of banking regulation.
- The category has moved from pilot to production: SWIFT launched a shared ledger with 17 global banks in July, JPMorgan’s Kinexys settles institutional payments today, and a consortium including Bank of America and BNY targets a 2027 network.
- A technical distinction matters more than it sounds: a non-transferable tokenized deposit that settles between banks behaves like account money, while a freely transferable deposit token behaves more like a bank-issued stablecoin, and regulation treats the two differently.
- The stakes are structural. Stablecoins pull deposits out of banks into reserve assets; tokenized deposits keep them in. Which model wins the institutional corridor shapes bank funding, credit creation, and what a dollar on a blockchain actually is.
The most consequential money on blockchains this year is not a cryptocurrency and not a stablecoin. It is ordinary bank deposits, the money in checking accounts, wearing a token as a coat. In July, SWIFT switched on the 17-bank ledger built for this instrument with 17 of the world’s largest banks built specifically to move this instrument. JPMorgan already settles institutional payments with its own version. A consortium of American banking giants is building a shared network for 2027, and the FDIC’s stablecoin rulemaking carves out space to address how deposit insurance applies to it. The instrument is the tokenized deposit, and the reason it deserves twenty minutes of any crypto reader’s attention is that it is the banking system’s structural answer to the $300 billion stablecoin sector: a digital dollar that does everything a stablecoin does while never leaving the bank. Whether that is the point or the problem depends on where you sit, which is exactly what this guide unpacks.
The definition, precisely
A tokenized deposit is a digital representation of a claim on a commercial bank, recorded on a blockchain or distributed ledger, issued by the bank that holds the deposit, and redeemable one-to-one against it.
Every clause is doing work. It is a claim on a bank, the same legal object as the balance in a checking account, which means it is commercial bank money, the kind that makes up the overwhelming majority of what people and firms actually use as dollars. It is issued by the bank itself, not by a third party holding the bank at arm’s length. It maintains one-to-one correspondence with a deposit that remains on the bank’s balance sheet, so tokenizing a million dollars does not move a million dollars anywhere; it changes the record-keeping technology for money that stays put. And it lives on a ledger, which is what gives it the properties deposits never had: settlement in seconds, availability at 3 a.m. on a Sunday, and the ability to be composed into programmable payment logic.
The cleanest way to hold the concept: a stablecoin is a new kind of money issued by a new kind of company, while a tokenized deposit is the oldest kind of money with a new kind of plumbing. For readers who want the other side of the comparison, crypto.news has also explained the competing model.
How it differs from a stablecoin, mechanically
The two instruments look identical at the point of use, a dollar-denominated token that moves on a ledger and settles fast, and are opposites underneath. Three differences carry all the weight.
Where the money sits. When a customer buys a stablecoin, dollars leave their bank account and land in the issuer’s reserve portfolio, Treasury bills, repo, money funds, custodial accounts, outside the banking system. The bank loses a deposit; the reserve assets sit sterile with respect to lending. With a tokenized deposit, nothing leaves. The deposit stays on the bank’s balance sheet, funding loans exactly as before, while the token circulates as its mobile representation. Multiply across a sector and this is the difference between digital dollars that drain bank funding and digital dollars that preserve it, which is why the Federal Reserve’s research treats stablecoins as a disintermediation risk and tokenized deposits as the banks’ countermove.
Who stands behind it. A tokenized deposit carries the full apparatus of banking: deposit insurance up to statutory limits, the bank’s capital and supervision, and, behind the bank, access to the Federal Reserve’s discount window. A stablecoin carries the issuer’s reserves and, under the GENIUS Act, a legal priority for holders in insolvency plus full-reserve requirements, real protections, but no insurance and no central bank. The FDIC has confirmed the insurance line between the two: stablecoin wallets get no pass-through deposit insurance, while the FDIC’s own stablecoin rulemaking addresses insurance treatment of tokenized deposits precisely because they are deposits.
What it may pay. The GENIUS Act prohibits payment stablecoin issuers from paying interest on the coin itself, a line Congress drew to stop stablecoins from becoming uninsured savings accounts. A tokenized deposit is a deposit; a bank can pay interest on it the way it pays on any account. In a world of meaningful rates, that asymmetry is not a footnote, it is a business model, and it is one reason banks believe the institutional corridor is winnable.
The distinction inside the category
Here the vocabulary gets sloppy across the industry, and one analytical cut, articulated most clearly by researcher Noelle Acheson, brings it into focus: a tokenized deposit is not the same thing as a deposit token, and the difference is transferability.
In the strict model, a tokenized deposit moves only between customers of banks in the network, and when it moves between banks, the banks settle behind the scenes, the token a customer of Bank A holds is always a claim on Bank A, and transferring value to a customer of Bank B means Bank A’s token is burned, interbank settlement occurs, and Bank B mints its own. This is account money with better rails: the customer relationship, the compliance perimeter, and the claim structure all stay intact, which is why regulators are comfortable with it and why SWIFT’s ledger, which coordinates exactly this burn-settle-mint choreography across institutions, is built this way.
In the looser model, a deposit token is a bearer-style instrument: freely transferable to anyone with a wallet, circulating like a stablecoin while claiming deposit status. This version makes bank money composable with open networks, and it makes regulators nervous, because a freely circulating claim on a bank held by strangers to the bank starts to blur into a bank-issued stablecoin, raising exactly the insurance, run-risk, and know-your-customer questions the strict model avoids. Where each jurisdiction draws this line will quietly determine whether tokenized deposits remain an interbank instrument or grow into a public one, and it is the single most important open design question in the category.
Who is building what
The category crossed from white papers to production over roughly eighteen months, and three architectures now compete.
The single-bank model is live. JPMorgan’s Kinexys settles institutional payments with tokenized deposits today, has extended onto public infrastructure including Base and the Canton network, and proves the concept at the only scale that matters, real money, real clients. Its structural limit is reach: one bank’s token moves one bank’s money, and every large bank running its own rail recreates the fragmentation problem that correspondent banking exists to solve.
The shared-network model is the answer to that limit, and it launched in earnest on July 9, when SWIFT’s blockchain-based ledger went live for initial use with 17 banks across six continents, Citi, HSBC, UBS, BNP Paribas among them, built on Hyperledger Besu in nine months. The ledger validates and coordinates tokenized-deposit movements between member banks around the clock, with final settlement through existing rails, and its pitch is distribution: SWIFT connects more than 11,000 institutions, a footprint no single bank or startup can match. A parallel American effort through The Clearing House, backed by JPMorgan, Bank of America, Barclays, and BNY, targets a 2027 launch, meaning even the shared-network lane already has competing networks.
The public-facing frontier is where the deposit-token question lives: experiments in making bank-issued tokens usable in open on-chain environments as settlement assets and collateral. This is the smallest lane today and the one with the largest implications, because it is where bank money and DeFi composability would actually meet.
The honest limitations
The category’s advocates describe it as stablecoins without the risk. The description omits four things.
Tokenized deposits are permissioned by construction. Every holder is a bank customer inside a compliance perimeter; there is no permissionless access, which means the instrument does nothing for the populations and corridors where stablecoins found their strongest product-market fit, users the banking system serves badly or not at all. A fintech in Lagos paying a supplier in Shenzhen holds USDT because it cannot hold a JPMorgan deposit; that fact does not change when the deposit grows a token.
They are also only as good as the network effects they achieve. Money is useful in proportion to who accepts it, and a tokenized deposit accepted inside one consortium is a better wire transfer, not a new form of money. The proliferation of competing networks, SWIFT’s, The Clearing House’s, each mega-bank’s own, raises a real fragmentation scenario in which the category succeeds technically and still fails to produce a unified instrument.
Insurance is bounded. Deposit insurance covers up to the statutory limit per depositor per bank, which protects retail balances fully and institutional balances barely; a corporate treasurer holding nine figures in tokenized deposits is an uninsured creditor of the bank above the cap, exactly as with ordinary deposits. The instrument inherits banking’s protections and also banking’s fine print.
And the model is untested in a run. Tokenized deposits settle at all hours, which cuts both ways: the same rails that move corporate treasury on Sunday morning can move a panic on Sunday morning, faster than any deposit flight in history. Bank supervisors have noticed; it is one reason the strict, non-transferable design keeps winning approvals.
What could still go wrong
A category moving this fast earns a section on its failure modes, and tokenized deposits have four worth taking seriously, none of them exotic.
The first is the interoperability trap. Every architecture described above, single-bank rails, SWIFT’s shared ledger, The Clearing House network, mints tokens that work within its own perimeter. History’s parallel is instructive: early wire transfer and card networks fragmented for decades before consolidating, and the consolidation was driven by merchants and users refusing to hold seventeen incompatible instruments. A corporate treasurer offered JPMorgan tokens, SWIFT-coordinated tokens, and consortium tokens, each with different settlement finality and legal terms, may reasonably decide the pilot era is someone else’s problem and keep wiring. The category’s success requires the networks to interconnect, and the incentives to interconnect are weakest for exactly the largest banks whose participation matters most, because a proprietary rail that works is a moat.
The second is the run-dynamics question, which deserves more respect than the marketing gives it. A tokenized deposit inherits the bank’s credit risk, and always-on settlement means the deposit can leave at any hour a holder gets nervous. The 2023 regional banking crisis showed what smartphone-speed withdrawals do to a bank funded by concentrated, sophisticated depositors; token rails compress the same dynamic further. Supervisors have levers, the non-transferable design, settlement windows, position limits, but every lever traded against the always-on convenience that is the product’s selling point. The instrument’s safety case and its value proposition are, at the margin, the same dial turned in opposite directions.
The third is regulatory divergence on the deposit-token boundary. If one major jurisdiction blesses freely transferable deposit tokens while another confines banks to the strict interbank model, bank money itself forks: a transferable claim on a Singapore or London bank circulating on open networks while American bank tokens stay walled. That is not hypothetical, jurisdictions are already writing different answers, and the arbitrage it invites, banking migrating to wherever bank money is allowed to be most bearer-like, is the kind regulators historically respond to late and harshly.
The fourth is the quiet dependency on stablecoin rules. The competitive case for tokenized deposits leans on asymmetries the law that bans stablecoin interest created, stablecoins cannot pay interest, stablecoins carry no insurance, and asymmetries written by one Congress can be rewritten by another. A future amendment permitting yield-bearing regulated stablecoins, an idea already circulating in the CLARITY Act fights over activity-based rewards, would collapse the banks’ cleanest advantage overnight. The banks are building on ground the law currently tilts toward them, and the tilt is a policy choice, not a property of the technology.
None of these kill the category; each shapes what version of it survives. The strongest honest forecast is conditional: tokenized deposits win the regulated institutional corridor if the networks interconnect, if supervisors hold the transferability line without strangling the product, and if the legislative tilt endures. Three ifs is not a guarantee. It is, however, a much shorter list than the one stablecoins faced a decade ago, which is the fairest way to size the two contenders.
Why the fight matters
Strip the technology away and the tokenized-deposit-versus-stablecoin contest is a fight over the two-tier monetary system, the arrangement where central banks serve banks and banks serve everyone else, and every reader in crypto has a stake in the outcome.
If tokenized deposits win the institutional corridor, on-chain finance gets absorbed into banking: programmable settlement arrives, but issuance, access, and control remain with chartered institutions, and the deposit-funding model that finances lending survives digitization intact. If stablecoins win it, a parallel monetary layer keeps growing outside bank balance sheets, with narrower backing, broader access, and the disintermediation consequences the Fed’s researchers keep modeling. The likeliest outcome is partition, banks holding the regulated core, stablecoins holding the open edge, with the boundary contested for years at exactly the seams this guide has mapped: transferability rules, insurance treatment, and the interest-rate asymmetry the GENIUS Act wrote into law.
For now, the practical takeaways are three. A tokenized deposit is bank money with new rails, insured and supervised, and structurally unavailable to anyone outside a bank relationship. A stablecoin is new money with open rails, reserve-backed and uninsured, and structurally available to anyone with a wallet. And the institutions that spent a decade dismissing blockchains have now committed, with 17 banks, a 53-year-old cooperative, and the world’s largest asset managers in the room, to putting the oldest money in the world on them. Whatever else that signals, it settles one argument: the rails were never the controversial part. The money was. For the adjacent cash-market structure, crypto.news has also explained the other regulated cash instrument on-chain.
Frequently asked questions
What is a tokenized deposit in one sentence?
It is a commercial bank deposit represented as a token on a blockchain, issued by the bank holding the deposit, redeemable one-to-one, and left on the bank’s balance sheet, so it settles like a crypto asset while remaining ordinary, insured bank money underneath.
How is that different from a stablecoin?
Three ways. The money stays inside the bank and keeps funding loans, whereas stablecoin purchases move money out of banks into issuer reserves. It carries deposit insurance up to statutory limits and bank supervision, whereas stablecoin holders rely on reserves and legal priority with no insurance. And banks may pay interest on it like any deposit, while the GENIUS Act bars stablecoin issuers from paying interest on their coins.
Are tokenized deposits FDIC-insured?
As deposits, yes, up to the statutory limit per depositor per bank, and the FDIC’s current stablecoin-era rulemaking addresses their insurance treatment explicitly. The practical caveat is the cap: retail balances are fully covered, while institutional holders above the limit are uninsured bank creditors for the excess, exactly as with conventional accounts. Stablecoin wallets, by contrast, carry no pass-through insurance at all.
What is the difference between a tokenized deposit and a deposit token?
Transferability. A tokenized deposit in the strict sense moves only among customers of participating banks, with interbank settlement behind each transfer, preserving the account relationship. A deposit token is freely transferable to any wallet, circulating like a bank-issued stablecoin. Regulators are far more comfortable with the first model, and where jurisdictions draw this line will shape whether the instrument stays interbank or becomes public.
Who actually uses tokenized deposits today?
Institutions, not retail. JPMorgan’s Kinexys settles real institutional payments and has extended to public infrastructure including Base and Canton. SWIFT’s shared ledger launched initial use in July 2026 with 17 global banks coordinating tokenized-deposit movements around the clock. A Clearing House consortium including Bank of America and BNY targets 2027. Retail-facing versions remain experimental almost everywhere.
Do tokenized deposits threaten stablecoins?
In the institutional corridor, directly: for regulated entities moving money between themselves, an insured, interest-capable, supervised instrument is a strong competitor. In open corridors, not really: tokenized deposits require a bank relationship, so exchange settlement, DeFi collateral, and unbanked-adjacent remittances remain stablecoin territory. The likely outcome is partition rather than a winner-take-all, with the boundary set by regulation as much as preference.
What are the main risks or limits?
Permissioned access excludes everyone outside member banks. Competing networks risk fragmenting the category into non-interoperable islands. Insurance is capped, leaving large institutional balances exposed above the limit. And always-on settlement is untested under stress, since the same 24/7 rails could accelerate a deposit run faster than any in history, which is partly why supervisors favor non-transferable designs.
Why does this matter for someone holding crypto?
Because it defines the competition. The growth path stablecoins were assumed to own, institutional settlement, corporate treasury, tokenized-asset plumbing, is exactly where banks are now deploying an instrument with insurance and interest attached. How that contest resolves shapes stablecoin demand, the reserves feeding Treasury markets, and which digital dollar becomes default in each corridor. This is educational context, not investment advice.
Disclaimer: This article is for information and educational purposes only and does not constitute financial, investment, or legal advice. Product structures, insurance treatment, and regulatory rules described here vary by jurisdiction and are subject to change. Always do your own research. Information is accurate as of July 20, 2026.
