Definition
Delivery versus Payment (DvP) is the settlement principle under which the transfer of securities from a seller to a buyer is conditioned on the simultaneous or near-simultaneous transfer of funds from the buyer to the seller — ensuring that no party is exposed to principal risk during the settlement process. DvP was formally articulated as a global standard by the Bank for International Settlements’ Committee on Payment and Market Infrastructures (BIS CPMI) in its 1992 report “Delivery Versus Payment in Securities Settlement Systems,” which identified the elimination of principal risk as the central objective of securities settlement system design. Principal risk — the risk that a seller delivers securities but does not receive payment, or that a buyer pays but does not receive securities — is distinct from credit risk (risk of counterparty insolvency) and market risk (risk of price movements), but can cause equally catastrophic losses when settlement fails.
The BIS defines three DvP models with different operational characteristics. DvP Model 1 achieves simultaneous gross settlement of both the securities leg and the cash leg — every individual trade settles immediately and independently, providing the purest form of DvP protection. DvP Model 2 settles the securities leg on a gross basis (individually) but nets the cash leg across a batch of trades, reducing cash payment flows while maintaining security-level specificity. DvP Model 3 nets both legs — a batch of securities and cash flows is accumulated and settled on a net basis, meaning participants need only deliver or receive the net position across all trades in the batch — providing maximum liquidity efficiency but accepting some degree of mutual interdependence among the netted transactions. Most major global settlement systems use DvP Model 3 (including DTCC’s DTC for US equities), accepting netting efficiency in exchange for the pure gross settlement of Model 1.
Key Facts
- The DTCC’s DTC (Depository Trust Company) settles over $2 quadrillion in securities transactions annually using a DvP Model 3 approach, netting down the gross obligations of its participants to a small fraction of gross volume.
- US equity markets moved to T+1 settlement in May 2024, reducing the window of principal risk exposure from 48 hours to 24 hours — a meaningful but incremental improvement from the blockchain tokenization perspective of true T+0 atomic DvP.
- Broadridge’s Distributed Ledger Repo (DLR) platform achieves near-Model 1 DvP for bilateral repo agreements, allowing institutions to execute intraday repo with immediate securities delivery and simultaneous payment — saving estimated $1 billion+ in annual collateral costs across its user base.
- The Continuous Linked Settlement (CLS) system achieves DvP-equivalent settlement for foreign exchange transactions by simultaneously delivering both currency legs of FX trades, eliminating Herstatt risk (the FX equivalent of principal risk) for the major currency pairs.
- DTCC’s Project Whitney (2021) studied the application of DvP to tokenized securities and concluded that true atomic DvP was achievable on blockchain platforms but required a coordinated cash leg — either a wholesale CBDC or regulated stablecoin — to function with institutional-grade reliability.
- The DTCC’s 2024 T+1 implementation eliminated 11 hours of settlement processing time daily, freeing approximately $500 million in clearing fund requirements — illustrating the capital efficiency gains available from settlement compression, which T+0 atomic DvP would amplify further.
- The Reserve Bank of Australia completed a proof-of-concept for DvP settlement of tokenized equity and tokenized cash (as a synthetic CBDC) in 2023, demonstrating that regulatory-grade atomic DvP settlement is technically feasible in a controlled environment.
Relevance to Tokenization
DvP is the conceptual heart of the tokenization settlement value proposition. When tokenization advocates promise “faster, cheaper, safer settlement,” the specific settlement improvement they describe is a move from DvP Model 3 (netting, T+1 batch settlement) to atomic DvP Model 1 (instant, gross, simultaneous delivery and payment). The efficiency gains from this transition are large and measurable: clearing fund requirements drop, collateral is freed for other uses, counterparty risk exposure windows shrink from hours to seconds, and failed settlements — which occur at a rate of approximately 2-5% in traditional markets and trigger complex reconciliation processes — become technically impossible.
The critical enabler of blockchain DvP is the on-chain cash leg. For securities tokenized on Ethereum, the cash payment must also be an Ethereum token that can be included in the same atomic transaction as the security token transfer. This requires either: (1) a regulated stablecoin like USDC or PYUSD that is widely accepted as equivalent to cash by institutional parties, (2) a tokenized bank deposit (JPMorgan’s Kinexys Digital Payments, Societe Generale’s forge EUR-coin) that carries the credit quality of a commercial bank, or (3) a wholesale central bank digital currency (wCBDC) issued by the Federal Reserve that carries the credit quality of sovereign money. The GENIUS Act’s stablecoin framework and ongoing Federal Reserve research into wCBDC are therefore not peripheral to tokenization — they are prerequisites for the most transformative aspect of the tokenization value proposition.
For tokenized securities platforms seeking to implement genuine atomic DvP, the practical design question is which cash instrument to use. USDC is the most widely deployed option, with strong reserve backing (100% US Treasuries) and institutional acceptance growing rapidly. Tokenized bank deposits offer regulatory familiarity (bank deposits are federally insured and well-understood credit instruments) but are currently fragmented across individual bank implementations without a common standard. A wholesale CBDC would provide the ideal combination of universal acceptance, sovereign credit quality, and technical suitability for atomic settlement — but Federal Reserve implementation is not expected before 2027-2029 at the earliest, based on current research timelines.
Related entries: Atomic Settlement, GENIUS Act, Smart Contract