Five instruments are commonly grouped as digital money. Only two can carry institutional settlement at scale. The line between them is not technological. It is the legal status of the claim, bank deposit liability versus everything else, and it determines what can serve as the cash leg of a tokenized market.
The five instruments, defined
Five categories of digital money differ on issuer, claim, and survival of the claim if the issuer fails. CBDCs are sovereign liabilities issued directly by central banks. Tokenized deposits are bank deposit liabilities recorded on a programmable ledger inside the issuing bank. Deposit tokens are bank claims that circulate across institutions on shared rails. Stablecoins are private issuer obligations backed by reserves the holder does not own. Wrapped cash is a tokenized claim against cash held in escrow.
The five differ on who issues them, what the holder actually owns, and whether the claim survives an issuer failure intact. The Bank for International Settlements distinguishes between tokenized money that settles in central bank money and private bearer instruments that do not, arguing in BIS Bulletin 73 that the bearer structure introduces departures from par value.
Three categories deserve precise definitions because the market uses them interchangeably. A tokenized deposit is a bank liability in token form, recorded within the issuing bank's ledger and excluded from MiCAR scope because, as the European Banking Authority found in December 2024, tokenisation does not alter the regulatory qualification of the underlying deposit. A deposit token is similar in legal substance but travels on shared public or consortium rails, carrying the issuing bank’s credit risk into a multi-bank environment. A stablecoin is a claim on a private issuer backed by reserves the holder does not own.
CBDCs are the simplest category to define and the most uneven in practice. The Atlantic Council CBDC Tracker records 146 countries exploring CBDC designs, with only three retail launches live and all three reporting slow adoption. Wholesale CBDC pilots, including mBridge with $55.49B in volume, are running at scale where retail variants have stalled.
The comparison table: issuer, claim, backing, legal regime, programmability, settlement role
The six dimensions that matter are the issuer, the legal nature of the claim, what backs it, which regulatory regime applies, how programmable it is, and whether it can serve as a settlement asset. Read across each row and the legal discontinuity becomes visible.
| Instrument | Issuer | Claim | Backing | Legal regime | Programmable | Settlement role |
|---|---|---|---|---|---|---|
| CBDC (wholesale) | Central bank | Sovereign liability | Central bank balance sheet | Central bank law | Yes | Cash leg — ideal |
| CBDC (retail) | Central bank | Sovereign liability | Central bank balance sheet | Central bank law | Limited today | Retail only, no wholesale settlement |
| Tokenized deposit | Commercial bank | Deposit liability | Bank assets (FDIC/DGS insured) | Banking law + deposit insurance | Yes | Cash leg — institutional grade |
| Deposit token | Commercial bank | Deposit liability (issuing bank) | Bank assets (issuing bank) | Banking law | Yes | Cash leg — multi-bank rails |
| Stablecoin | Private issuer | Contract claim on reserve pool | Reserves (holder has no direct title) | E-money / bespoke statute | Yes | Trading, payments — not settlement |
| Wrapped cash | Custodian / escrow | Claim on escrowed cash | Segregated cash | Contract / escrow law | Limited | Crypto bridge only |
The columns that matter most are the claim and the legal regime. A tokenized deposit holder owns the same legal claim as a traditional depositor at the same bank. A stablecoin holder owns a contract that points at a pool of reserves they do not have direct title to. That is the discontinuity.
The legal discontinuity between deposit liabilities and everything else
The legal regime governing the claim, not the technology recording it, determines what an instrument can do at institutional scale. Deposit liabilities sit inside banking law with established resolution regimes, insurance frameworks, and balance sheet treatment. Stablecoins and wrapped cash sit outside it, governed by e-money rules or bespoke statutes that emerged in the last 24 months. Every other property follows from this split.
FDIC Acting Chairman Travis Hill stated the principle plainly in an April 2025 policy update: “deposits are deposits, regardless of the technology or recordkeeping deployed.” That sentence does the work of an entire regulatory paper. A deposit recorded as a token is still a deposit. A token that is not a deposit, regardless of how it is marketed, does not become one by adopting deposit-like features.
The GENIUS Act, summarized in the NY Fed Liberty Street Economics historical perspective on stablecoins, codified this separation in US law. Stablecoins must be fully backed 1:1 by safe liquid assets, may not pay interest or yield on stablecoin balances, and holders receive priority claims in bankruptcy. None of these features make a stablecoin into a bank liability. They make it a regulated non-bank instrument with a defined ceiling on what it can do.
Banks holding stablecoin-issuer reserve deposits face the cost of this separation directly. A December 2025 Federal Reserve FEDS Note classifies these balances as “wholesale and typically uninsured” with higher regulatory runoff assumptions than retail deposits. The same dollar sitting in a tokenized retail deposit account is treated as a stable retail liability. The same dollar parked at a bank as stablecoin reserve backing is treated as flight-prone wholesale funding. The technology is identical. What changes is the legal status of the claim.
Singleness of money: why bearer instruments break at scale
Money functions as money only when one dollar trades at par with every other dollar, and bearer instruments lose that property when stress hits. The BIS calls this the singleness of money, and the 2025 Annual Economic Report concluded that stablecoins “do not deliver singleness of money, elasticity, and integrity.” Tokenized deposits, which settle into central bank money, preserve it.
The historical parallel is the pre-Fed era of national bank notes, when notes from different banks traded at discounts based on issuer creditworthiness and distance to redemption. A note from a New York bank circulated at par in Manhattan and at a discount in St. Louis. That system was abandoned in favor of a singular currency for a reason: every payment becomes a credit assessment, and money stops working as money. The NY Fed essay draws the parallel directly, noting that stablecoins from different issuers already trade at small but persistent spreads in stressed conditions.
A bearer instrument that settles peer-to-peer with no central reconciliation introduces this same problem in token form. When USDC briefly traded below par during the March 2023 Silicon Valley Bank weekend, the discount was a market price for issuer credit risk, not a technical anomaly. The instrument behaved as designed. Hyun Song Shin made the contrast explicit, arguing in the BIS 2025 report that “tokenisation of deposits and central bank money means that both the primary means of payment as well as the settlement function of central bank money can be integrated seamlessly on the same programmable platform.” Deposits settle through central bank money. A stablecoin settles into another claim on the same private issuer.
What can serve as the cash leg of a tokenized market
A tokenized security needs a cash leg that is legally final, balance-sheet eligible, and immune to issuer-credit reassessment during settlement, and only tokenized central bank money and tokenized commercial bank deposits clear that bar. BIS Project Agorá, the seven-central-bank initiative bringing together more than 40 private institutions, was designed around exactly this premise. The project description states that “settlement using central bank money (the safest and most liquid settlement asset) is crucial for a stable financial system as it eliminates credit risk.”
Live institutional volume confirms the architecture. JPMorgan’s Kinexys platform has processed more than $3 trillion in total volume since inception and runs above $7 billion in average daily volume, with the firm explicit that its Blockchain Deposit Accounts “function similarly to traditional Demand Deposit Accounts” and are not stablecoins. HKMA launched the EnsembleTX pilot environment in November 2025 to enable real-value tokenized deposit transactions, including tokenized money market fund settlements with real-time liquidity, with the pilot running through 2026. Fnality became the first DLT-based payment system to receive UK Settlement Finality designation in December 2024, with participant funds held at the Bank of England.
Stablecoins now circulate at scale and keep growing. RWA.xyz tracks roughly $305 billion in total stablecoin market cap with monthly transfer volume near $7.7 trillion. None of that volume is institutional securities settlement. It is crypto trading, cross-border payments, and on-chain treasury operations where the legal status of the claim matters less than the speed and reach of the rail. FractiFi designs settlement infrastructure on the assumption that the cash leg must be a deposit liability, because every institutional counterparty we have engaged operates under treasury mandates that cannot accept a non-bank claim as settlement.
Where the spectrum framing breaks down, and where it still helps
The strongest version of the spectrum argument is that all five instruments are programmable forms of value transfer, share common token infrastructure, and increasingly interoperate through bridges, atomic swaps, and shared messaging standards. A treasurer evaluating digital cash options sees five technology choices that move dollars on a blockchain. From a developer’s perspective, the API surface looks similar. From a user’s perspective, the UX converges. There is real signal in that view.
The framing breaks down at the legal layer, which is the layer that determines what risk managers, auditors, and resolution authorities can accept. The Financial Stability Board’s October 2024 report on tokenisation identifies five risk categories that emerge as tokenized financial systems scale: liquidity and maturity mismatch, leverage, asset price and quality risks, interconnectedness, and operational fragilities. The FSB’s scope explicitly excludes CBDCs, signaling that sovereign instruments occupy a separate risk plane from private ones. Within the private instruments, the regulatory treatment of a deposit liability and a stablecoin diverge on every one of those five categories.
The spectrum framing is useful for understanding what an institution can hold in a single wallet and route through a single integration layer. It is misleading as soon as the question becomes which instrument can settle a primary issuance, post as collateral against a securities trade, or close out an interbank exposure. Fnality’s framing is the clearest: in The Next Era of Digital Money, the firm separates deposit tokens, tokenized deposits, and stablecoins as distinct instruments while noting that “you still need a risk-free settlement asset to close out interbank exposures generated by all these digital forms of money.” The spectrum view does not see that requirement. The legal view treats it as the central question.
What this means for institutions deciding today
Institutions evaluating digital money infrastructure should choose instruments based on the legal status of the claim, not the marketing of the rail. For trading, payments, and crypto-adjacent flows, stablecoins are the established instrument and will remain so. For institutional settlement, primary issuance, and balance-sheet-eligible cash legs, the choice is between tokenized deposits, deposit tokens, and wholesale CBDCs where available.
Stablecoins are not a bridge to tokenized deposits. They are a different instrument addressing different use cases under a different regulatory regime, and treating them as a transitional technology delays the work of integrating the actual settlement layer. The GENIUS Act’s prohibition on paying interest on stablecoin balances is not a constraint to be engineered around. It is a structural feature that defines what a stablecoin is in US law, and it forecloses the institutional treasury use case that requires yield-bearing on-chain cash.
JPMorgan, HSBC, Citi, Fnality consortium banks, and HKMA participants are building the infrastructure at the deposit layer. The BIS Agorá conclusions, scheduled for H1 2026, will set the reference architecture for tokenized central bank and commercial bank money integration. Institutions that wait for stablecoin regulation to converge with deposit law are waiting for something that is not coming. The two will remain distinct categories, governed by distinct rules, capable of distinct functions.
The institutions that move first on tokenized deposit infrastructure will integrate into the BIS Agorá reference design as it lands. The institutions that pilot on stablecoin rails will rebuild on deposit rails when their counterparties refuse to settle in non-bank claims. The legal discontinuity is permanent, and architecture choices made now should reflect it.
FractiFi builds tokenized deposit infrastructure for mid-tier banks, asset managers, private equity firms, and family offices that need an institutional-grade cash leg for tokenized markets but cannot build the rails independently. The decisions you make this year on which instrument to integrate determine which counterparties you can settle with in 2027 and beyond. If you are mapping the digital money stack against your settlement requirements, we would like to talk. The starting point is usually our tokenized deposit infrastructure overview and the Day 1 pillar on why tokenized deposits are the missing foundation of RWA infrastructure.
