Settlement is the moment a trade becomes final — when the seller delivers the security and the buyer delivers the cash, simultaneously and irrevocably. In traditional finance, this moment arrives two business days after execution (T+2, now T+1 in the United States since May 2024), facilitated by a chain of intermediaries: central counterparties, custodians, correspondent banks, and central securities depositories. Blockchain infrastructure promises to compress this interval toward zero, eliminate counterparty risk through atomicity, and reduce the $4 trillion in daily settlement fails that consume collateral across global markets. The reality, as of early 2026, is more nuanced — but the direction is unambiguous.
The BIS Taxonomy: Three Models of DvP
The Bank for International Settlements has established the definitive conceptual framework for delivery versus payment (DvP) in securities settlement. The 1992 BIS report introduced three models that remain the standard analytical lens:
Model 1 achieves gross, real-time DvP — every individual trade settles security and cash simultaneously, with no netting. This is the gold standard for finality and counterparty risk elimination, but it maximizes liquidity requirements because positions cannot offset each other before settlement.
Model 2 settles securities on a gross basis throughout the day but nets cash positions, settling the cash leg at end of day. This reduces cash demand relative to Model 1 but introduces intraday cash exposure.
Model 3 nets both securities and cash, settling the netted positions at end of day. DTCC’s current infrastructure operates predominantly in Model 3 — the netting engine dramatically reduces the volume of actual asset transfers, but participants are exposed to each other’s credit risk until settlement occurs.
Blockchain-based settlement primarily targets Model 1: atomic, real-time, gross DvP where the transfer of a security token and the corresponding cash payment occur in a single transaction that cannot be partially executed. This eliminates the principal risk — the risk that one leg settles while the other fails — that plagues traditional settlement infrastructure.
| DvP Model | Securities Settlement | Cash Settlement | Counterparty Risk | Liquidity Requirement |
|---|---|---|---|---|
| Model 1 (Gross/Gross) | Real-time, per-trade | Real-time, per-trade | Eliminated | Maximum |
| Model 2 (Gross/Net) | Real-time, per-trade | End-of-day net | Intraday exposure | Moderate |
| Model 3 (Net/Net) | End-of-day net | End-of-day net | Full day exposure | Minimum |
| Blockchain atomic | Instantaneous | Instantaneous | Eliminated | Maximum, but programmable |
Broadridge DLR: Scale Without Blockchain Purity
Broadridge Financial Solutions operates the Distributed Ledger Repo (DLR) platform, which by late 2023 had processed a single-day settlement volume of $384 billion in repurchase agreement transactions. The DLR platform uses a private, permissioned blockchain to enable intraday repos with same-day settlement — compressing what had been a T+1 or T+2 process for certain repo types into intraday finality.
The Broadridge DLR achievement is significant for what it demonstrates about institutional readiness: major banks including JPMorgan, Goldman Sachs, and Societe Generale are executing billions in daily settlement volume on distributed ledger infrastructure. But DLR operates in a closed ecosystem — the blockchain is visible only to Broadridge and its approved counterparties. Settlement finality is contractual rather than protocol-enforced. The platform is a DLT-enhanced version of traditional repo infrastructure rather than a fully blockchain-native system.
This distinction matters for the broader tokenization thesis. Tokenized treasuries and other real-world assets require settlement infrastructure that can interact with public or semi-public blockchain networks where token holders reside. The Broadridge model answers questions about scale and institutional willingness but leaves open the question of cross-network, protocol-enforced settlement finality.
DTCC Project Ion: Demonstrating T+0
The Depository Trust & Clearing Corporation, which processes over $2.5 quadrillion in annual settlement volume as the United States’ central securities depository, has been systematically testing blockchain-based settlement acceleration. Project Ion, announced in 2020 and expanded through multiple pilot phases, demonstrated T+0 settlement capability for US equities using distributed ledger technology.
The Ion demonstration involved real securities — US-listed equities already held in DTCC accounts — and achieved same-day settlement finality for transactions that would normally take two business days. The technical approach used a parallel DLT layer alongside DTCC’s existing infrastructure, allowing participating broker-dealers to opt into accelerated settlement for qualifying transactions.
The critical insight from Project Ion: T+0 settlement is technically achievable within the existing US securities regulatory framework, but it requires the simultaneous availability of both the security and the cash in settlement-ready form at the time of transaction. Under T+1 or T+2, brokers have time to locate securities and fund positions after execution. Under T+0, pre-funding or real-time funding mechanisms become mandatory — a structural change that affects how broker-dealers manage intraday liquidity.
T+1: The Bridge Step
The US equity markets’ transition to T+1 settlement, mandated by the SEC with an effective date of May 28, 2024, represents the first compression of the standard settlement cycle since the move from T+3 to T+2 in 2017. T+1 reduces the counterparty exposure window by 50%, decreases margin requirements at central counterparties, and reduces the pool of unsettled transactions outstanding at any given time.
T+1 is not blockchain settlement. It uses existing DTCC infrastructure, NSCC clearing, and DTC delivery mechanisms. But it has a meaningful relationship to tokenization: it establishes the operational expectation that settlement can be faster, creates systems and workflows optimized for accelerated settlement, and forces broker-dealer back offices to modernize reconciliation and affirmation processes that had accumulated technical debt for decades. The firms that successfully adapted to T+1 have infrastructure that is better positioned to handle the further compression toward T+0 or atomic settlement.
The international dimension adds complexity: European markets, which predominantly settled at T+2, now face a mismatch with US T+1 that creates settlement timing challenges for cross-border transactions. This mismatch may actually accelerate European interest in tokenized settlement infrastructure that can operate independently of calendar-based cycles.
JPMorgan Kinexys: Atomic Settlement at Scale
JPMorgan Chase’s Kinexys platform (formerly Onyx/JPM Coin) provides blockchain-based intraday settlement for institutional clients, primarily using a dollar-denominated digital currency — JPM Coin — to transfer value between JPMorgan accounts on a permissioned blockchain. The platform has processed hundreds of billions in transactions since its launch and includes participation from institutional clients including Goldman Sachs and Siemens.
The Kinexys architecture achieves genuine atomicity within its permissioned network: the debit and credit occur in the same transaction, with no settlement lag. For intra-JPMorgan transactions — payments between accounts both held at the bank — this represents true DvP. The limitation is the network boundary: settlement finality applies only within the Kinexys network, and cross-network transfers require bridge mechanisms that introduce latency and counterparty risk.
JPMorgan has positioned Kinexys as the institutional cash leg for blockchain-native securities settlement. In theory, a security token issued on a compatible blockchain could be exchanged atomically for Kinexys-settled digital dollars. The practical realization of this architecture requires the cross-chain interoperability infrastructure to connect security token networks with the Kinexys settlement layer — a challenge that Chainlink CCIP and similar protocols are attempting to solve.
Chainlink CCIP: The Cross-Chain Settlement Protocol
Chainlink’s Cross-Chain Interoperability Protocol (CCIP) has emerged as a candidate standard for cross-chain tokenized asset settlement. CCIP provides a messaging protocol that allows smart contracts on different blockchain networks to communicate and execute coordinated transactions — a prerequisite for cross-chain DvP where the security token and the payment token may reside on different networks.
The practical significance: BlackRock’s BUIDL fund operates on Ethereum, while Franklin Templeton’s FOBXX operates across Stellar and Polygon, and various other tokenized assets exist on Avalanche, Solana, and private chains. For an investor to exchange BUIDL shares for Polygon-native stablecoins, or for a market maker to arbitrage tokenized Treasury positions across chains, cross-chain settlement infrastructure is required. CCIP’s architecture uses a commit chain with risk management network to validate cross-chain messages before execution, providing a security layer that simple bridge protocols lack.
Chainlink has announced partnerships with SWIFT — the interbank messaging network — and with several major banks to test cross-chain settlement for tokenized assets. The SWIFT partnership is particularly notable: it suggests that traditional financial messaging infrastructure may integrate with blockchain settlement rather than compete with it.
Settlement Finality: Why Blockchain Choice Matters
Not all blockchains provide equivalent settlement finality, and the distinction is consequential for institutional use. Settlement finality — the point at which a transaction is irreversibly complete — varies by blockchain consensus mechanism.
Ethereum’s proof-of-stake consensus provides probabilistic finality that approaches certainty after approximately 12 minutes (2 epochs), though the protocol introduced “single slot finality” as a research priority to reduce this window. Solana’s tower BFT consensus provides near-instant finality within seconds. Hyperledger Fabric and other permissioned blockchain frameworks can provide immediate, deterministic finality within a closed validator set.
For settlement systems, the finality window defines the period of residual counterparty risk after transaction execution. A 12-minute finality window on Ethereum is dramatically better than T+1, but it is not zero. Permissioned blockchains with deterministic finality can match or exceed the settlement certainty of real-time gross settlement (RTGS) systems used by central banks, but at the cost of decentralization and the compliance burden of managing a permissioned validator set.
The institutional custody infrastructure layer interacts directly with settlement finality: custodians must decide at what point in the finality window to credit client accounts, and their decisions have implications for client liquidity and risk exposure. These are not merely technical questions; they are credit and operations policy decisions that custodian risk committees must resolve.
The Endgame: What True Atomic Settlement Requires
Achieving genuine atomic, real-time, gross DvP for tokenized securities at institutional scale requires the convergence of several infrastructure components that remain in development: a widely adopted digital cash instrument (central bank digital currency, tokenized bank deposits, or high-quality stablecoin) that can serve as the cash leg; cross-chain interoperability standards mature enough to carry institutional settlement risk; legal frameworks that recognize on-chain settlement finality as legally binding discharge of payment obligation; and regulatory clarity that allows custodians to hold the relevant digital cash instruments as qualified custodians.
Each of these requirements is being addressed simultaneously across different regulatory and technical workstreams. The trajectory suggests that institutional-grade atomic settlement is achievable within a multi-year horizon — but it is a systems integration challenge of the first order, not merely a software engineering problem.