Tokenized deposits are the base layer real-world asset infrastructure depends on. Without them, tokenized assets settle against legacy payment rails or stablecoin trust models, which means the asset is digital and the money is not. That gap is why most RWA pilots stall before production. Close it, and the rest of the stack becomes buildable.
Most institutional writing on RWA infrastructure treats the asset side as the work and the cash side as an implementation detail. That framing is backwards. The asset side of tokenization is largely solved. Issuers can mint claims, register ownership, and track transfer on a ledger. The unsolved piece is how the money moves against the asset in the same instant, on the same rail, under the same compliance regime. Until that cash leg exists as programmable money on the same infrastructure, what looks like RWA infrastructure is a set of digital certificates wired to a settlement system designed for paper.
The asset side is not the hard part
RWA infrastructure has a cash-leg problem, not an asset-leg problem. Tokenizing a Treasury, a private credit instrument, or a fund share is now routine. What remains operationally broken is the settlement against bank money in real time, because when the asset is on-chain and the cash is not, settlement risk has been relocated, not removed.
The DTCC and SEC proved the point in December 2025. The SEC no-action letter to DTCC cleared participants to record security entitlements for the Russell 1000, Treasuries, and major index ETFs on distributed ledger, with a pilot launch scheduled for H2 2026. The DTCC announcement frames the initiative as a bridge to tokenized custodied assets. What neither document covers is the cash leg. Entitlements get a new representation. The payment against those entitlements still clears through existing rails.
That is the gap. The asset-side unlock happened. The cash-side unlock has to happen separately, and it is happening through a different category of infrastructure.
Tokenized deposits are a regulated form of the same instrument banks already issue
A tokenized deposit is a commercial bank deposit represented on a programmable ledger. It is the bank’s liability, not a claim on a private issuer. Deposit insurance applies where the jurisdiction provides it, and the instrument remains subject to the prudential regime that governs the bank’s balance sheet. The token is a form factor; the legal substance is unchanged.
The European Banking Authority established this cleanly in its December 2024 Report on Tokenised Deposits: tokenisation does not alter the fundamental nature of the deposit claim, and tokenised deposits are distinct from e-money tokens under MiCAR. The FDIC position is the same. In its 2025 statement on key policy issues, the agency committed to providing certainty that “deposits are deposits, regardless of the technology or recordkeeping deployed.” US legislation reinforces the point. The GENIUS Act, signed into law in July 2025, creates a federal regime for payment stablecoins while explicitly preserving tokenized deposits as a separate category that can pay yield, carry deposit insurance, and remain regulated under existing banking law.
This is why tokenized deposits slot into institutional balance sheets without the bespoke risk analysis stablecoins require. Treasury teams do not need a new policy for a deposit they already hold. They need a new settlement mechanism for the deposit they already understand.
The institutional cash leg is already moving from pilot to production
Major institutions have stopped debating tokenized deposits and started running them in real-value environments. The pattern across jurisdictions is consistent: the money side is being built as programmable bank money, not as a crypto-native settlement token.
JPMorgan’s Kinexys platform is the largest live example. In November 2025 the firm deployed JPMD on the Base public blockchain, with B2C2, Coinbase, and Mastercard as initial participants. Naveen Mallela frames the proposition directly on that JPMD launch announcement: “security of bank-backed deposits and settlement, combined with the speed and flexibility of 24/7, near real-time blockchain transactions.” JPMorgan’s platform disclosures show cumulative notional above $1.5 trillion, daily volume around $2 billion, and year-over-year growth of roughly ten times. In Hong Kong, the HKMA launched EnsembleTX in November 2025, graduating Project Ensemble from proof of concept to real value. SFC CEO Julia Leung described the initiative as a step toward interbank settlement of tokenised deposits in real time, 24/7. Standard Chartered followed in December with a tokenized deposit solution built with Ant International, covering HKD, CNH, USD, and SGD across Hong Kong and Singapore.
These are not pilots in the conventional sense. They are live settlement rails moving institutional money, and they share one architectural assumption: bank money, represented as tokens, is the settlement asset for tokenized finance.
The BIS and the Fed have converged on the same conclusion
The official sector has not left the question open. The central bank view is that tokenised deposits preserve the singleness of money because they settle in central bank money between banks. Stablecoins, as bearer instruments, can deviate from par. This was the argument in BIS Bulletin 73 by Garratt and Shin in April 2023, and it is the argument the BIS has sharpened since.
The BIS Annual Economic Report 2025 formalises the position: the next-generation monetary and financial system will take shape on a tokenised unified ledger, with tokenised commercial bank deposits and tokenised wholesale central bank money as the two settlement-side components of the BIS trilogy, alongside tokenised government bonds. Project Agorá is the operational expression: seven central banks and more than 40 private institutions testing exactly this combination, with a report expected in the first half of 2026. The FSB’s October 2024 report on tokenisation adds the risk-side case, flagging operational fragility and liquidity mismatch as the specific vulnerabilities the cash leg needs to address. The Fed’s December 2025 FEDS Note on banks in the age of stablecoins frames tokenised deposits as a defensive response available to banks seeking to retain their deposit base as programmable cash moves on-chain.
Four different institutional bodies, one conclusion. Bank money on a programmable ledger is the settlement asset the tokenised system is being built around.
The counterargument: wholesale CBDCs and regulated stablecoins
The strongest case against tokenized deposits as the base layer has two parts. The first is that wholesale central bank digital currency is a cleaner settlement asset because it eliminates commercial bank credit risk. The IMF Fintech Note 2025/011 on central bank exploration of tokenized reserves makes this argument carefully: central bank money is the apex settlement instrument, and tokenised reserves extend that quality to on-chain environments. The second is that regulated payment stablecoins, now codified under the GENIUS Act, are a faster path to programmable cash at retail scale than bank deposit infrastructure, which is slower to mobilise across heterogeneous bank balance sheets.
Both arguments are correct on their own terms. Neither displaces tokenized deposits as the base layer for the next five to ten years.
Wholesale CBDC is the long-duration destination for interbank settlement, not the medium-term rail for institutional RWA flows. Project Agorá is the evidence: central banks are testing wholesale CBDC with tokenised commercial bank deposits, not as a replacement for them. Stablecoins occupy a different role. Under the GENIUS Act, they are preserved as a regulated cash rail but explicitly distinguished from deposits. They cannot pay yield, do not carry FDIC insurance, and, as the NY Fed Staff Report SR 1179 sets out, sit outside the bank balance sheet in a way that alters the credit-intermediation picture when scaled. For an asset manager settling a tokenised Treasury trade or a PE firm executing a capital call, the settlement asset needs to be yield-bearing, insured where relevant, and already in the counterparty’s treasury framework. That is a deposit.
The default base layer for institutional RWA settlement for the remainder of this decade is tokenized bank deposits. CBDC and stablecoins extend the money spectrum. They do not replace it.
The cash leg is where RWA market forecasts actually converge
The market-size case for tokenized finance collapses without a programmable cash leg. McKinsey’s June 2024 analysis estimates a $2 trillion tokenized market by 2030 in its base case, with cash and deposits as the largest single category at roughly $1.1 trillion, ahead of loans and bonds at around $0.3 trillion each. BCG’s October 2024 analysis of tokenized funds projects AUM above $600 billion by 2030 and makes the dependency explicit: rising demand is conditional on the parallel materialisation of regulated on-chain money, including tokenized deposits and CBDCs. Citi’s 2023 research reaches a similar structure, with $4 trillion in tokenized private markets and up to $5 trillion in CBDCs contributing to the adoption curve.
The earlier, often-quoted BCG and ADDX estimate of $16 trillion by 2030 predates the institutional tokenized deposit infrastructure that is now coming online. Delivery has lagged the forecast for a predictable reason. The asset-side infrastructure was ready before the cash side was. The cash side is the rate-limiter.
What “base layer” actually means architecturally
The base layer is the primitive every higher layer depends on and cannot substitute away. For tokenized finance, that primitive is bank-issued programmable money with a verifiable reserve relationship and protocol-level compliance. Capital routing, yield distribution, secondary trading, and asset-level settlement all execute against that primitive, which is why the settlement asset’s legal quality decides whether the stack holds.
Fnality’s Simone Cortese frames this well in “Rails Aren’t Rules”: institutional scale depends on “the quality, legal certainty and settlement finality characteristics of the settlement asset itself.” Settlement, he notes, “defines when obligations are extinguished, when exposure truly closes and when capital can be redeployed safely.” This is the reason, according to Firebrand Research, settlement failures have cost the financial system over $914 billion in the past decade, a figure cited in JPMorgan’s Kinexys, Chainlink, and Ondo cross-chain DvP test in May 2025. The test settled tokenised Treasuries against deposit tokens across chains. It worked because a programmable cash leg existed.
FractiFi builds this base layer for institutions outside the scope of JPMorgan or Standard Chartered: tokenized bank deposits with a protocol-enforced reserve invariant, compliance embedded at the protocol level, and a settlement layer that allows mid-tier banks, asset managers, PE firms, and family offices to operate against the same primitives the tier-one banks built internally. Oliver Wyman and JPMorgan Onyx, in their 2023 joint paper, argued that deposit tokens extend the prudential and monetary-policy benefits of the banking system into programmable form, making them a structurally different instrument from stablecoins rather than a competing one. The architectural implication is the same. Deposits do not merely settle the system. They are the system’s foundation.
The operational test for RWA infrastructure
A simple test separates RWA infrastructure from asset-side experiments: ask where the cash comes from when the asset settles. If the answer is a wire, an ACH, a correspondent line, or a stablecoin that exits the ledger within 24 hours, the infrastructure is incomplete, because the asset is on-chain while the money is not and settlement finality still depends on rails the tokenised layer does not control.
If the answer is a tokenised bank deposit that settles in the same transaction, carries the compliance identity through to the counterparty, and is backed by a verifiable reserve on the issuing bank’s balance sheet, the infrastructure is whole. Everything above it, including capital routing, waterfall execution, secondary markets, and cross-chain DvP, becomes buildable at that point, and tokenization graduates from wrapper to settlement system.
The institutions that understand this distinction are already building to it. Those that do not will spend the next three years running pilots that look successful in isolation and fail to scale, because the cash leg they assumed would appear has not appeared, and no one else will build it for them.
The operational gap between tokenized assets and programmable bank money is the single largest unsolved piece of RWA infrastructure for institutions outside the top-tier banks. FractiFi builds the tokenized deposit rails, settlement layer, and protocol-level compliance engine that mid-tier banks, asset managers, private equity firms, and family offices need to settle tokenized assets against bank-backed money without assembling a vendor stack from scratch. The infrastructure is designed for institutions moving from pilot to production, with three bank integration models depending on where the counterparty sits in its digital infrastructure journey. More detail on the team and company is on the about page. If you are evaluating how to close the cash-leg gap in a live deployment, we would like to talk.
