The European Central Bank will activate Project Pontes on 21 September 2026, connecting distributed ledger technology platforms to the Eurosystem's TARGET real-time gross settlement infrastructure. The system enables atomic delivery-versus-payment settlement of tokenized securities in central ban...
"Central banks should go on-chain too. That means bringing central bank money into the tokenised environment." — Isabel Schnabel, Executive Board Member, European Central Bank
The European Central Bank will activate Project Pontes on 21 September 2026, connecting distributed ledger technology platforms to the Eurosystem's TARGET real-time gross settlement infrastructure. The system enables atomic delivery-versus-payment settlement of tokenized securities in central bank money — a capability no private stablecoin or commercial bank token can replicate.
The launch follows a 2024 pilot that settled €1.6 billion across 64 institutions in nine jurisdictions and 58 use cases. Clearstream, the Deutsche Börse-owned post-trade services provider, completed end-to-end testing in August 2026. Registration for the initial go-live cohort closed on 7 August 2026.
Pontes operates against a backdrop in which global tokenized assets reached approximately €38 billion in February 2026, up from €7.4 billion at the start of 2024. The eurozone's traditional financial markets stand at €241 trillion. The ECB's bet: if central bank money stays off-chain while capital markets tokenize, dollar-denominated stablecoins — currently commanding over $304 billion in global supply versus less than $1 billion in euro-pegged tokens — become the default settlement layer for European markets.
Project Pontes is a Eurosystem-operated DLT platform that bridges market-facing blockchain infrastructure to TARGET Services, the eurozone's existing payment backbone. The Governing Council approved Pontes on 1 July 2025 as the near-term component of a two-track strategy. The second track, Appia, targets a full architecture blueprint by 2028.
At its core, Pontes enables one function: settling tokenized wholesale transactions — repo, collateral management, bond issuance — in central bank money rather than commercial bank deposits or stablecoins. Settlement finality sits on the Eurosystem DLT layer, not solely in the T2 real-time gross settlement system as before.
The initial launch is described by the ECB as a "focused version with limited operating parameters." Complete technical standards, eligibility criteria for participants and assets, and interoperability specifications for different DLT platforms remain partially unspecified. Institutions that missed the 7 August registration deadline can apply for later participation windows.
Between May and November 2024, the ECB conducted settlement experiments with 64 market participants. The pilot covered 58 payment and securities settlement scenarios across nine jurisdictions. Total settled value reached €1.6 billion in central bank money.
The experiments validated two critical capabilities: interoperability between market infrastructures and delivery-versus-payment (DvP) settlement, where the cash and securities legs of a transaction settle simultaneously or not at all. This atomic settlement eliminates the counterparty risk embedded in traditional T+1 or T+2 settlement windows.
Clearstream announced on 17 August 2026 that it would engage in a series of end-to-end tests ahead of the 21 September launch. The joint testing program focused on demonstrating interoperability between Clearstream's market infrastructure and the Eurosystem DLT platform. Institutions must demonstrate appropriate testing completion to their national central bank before accessing the production environment.
Pontes offers two parallel settlement mechanisms:
Cash Tokens Model: Tokenized representations of central bank money that exist natively on the Eurosystem's DLT platform. These are direct claims on the central bank — functionally identical in credit quality to reserves held in a central bank account, but programmable and composable within DLT environments.
Trigger Model: The DLT initiates a conventional payment routed through the T2 RTGS system. The settlement occurs in existing central bank infrastructure, but the instruction originates on-chain. This model provides an on-ramp for institutions not yet ready for full DLT-native settlement.
The dual model reflects a pragmatic assessment. Not every institution will migrate to DLT-native operations simultaneously. The trigger model preserves integration with existing systems while the cash token model demonstrates the full value proposition — programmability, atomicity, and composability.
As Schnabel noted in her 28 August 2026 Jackson Hole speech: "Atomicity implies that the legs of a transaction settle together or not at all, thereby eliminating settlement risk." She added: "Programmability means that settlement can be made conditional on a set of rules that are executed automatically."
The ECB's urgency around Pontes stems from a quantifiable market asymmetry. According to DeFiLlama data from August 2026, global stablecoin supply stood at $304.6 billion. Euro-pegged tokens accounted for less than $1 billion. Dollar-denominated stablecoins dominate the on-chain settlement layer by a ratio exceeding 300:1.
Under MiCA regulation, stablecoin issuers operating in Europe face reserve requirements: a minimum 30% held with credit institutions for standard issuers, rising to 60% for issuers deemed "significant." Full authorization deadlines passed on 1 July 2026.
The ECB has rejected proposals to ease euro stablecoin reserve rules and denied backstop access for stablecoin issuers to central bank facilities. Schnabel's Jackson Hole argument was direct: "A stablecoin issuer has no independent capacity to expand liquidity elastically" during periods of financial stress. She cited the 1907 U.S. banking panic as precedent for what happens when private money issuers cannot create liquidity in crisis conditions.
The policy position is explicit: stablecoins function as "complements, not substitutes" for central bank money. In institutional securities settlement, where counterparty risk management is paramount, the ECB intends tokenized central bank money to be the default cash leg. The implication for stablecoins is containment — useful for payments, trading, and decentralized market access, but structurally excluded from the settlement layer that matters most to institutional participants.
An ECB macroprudential bulletin published in April 2026 analyzed 183 tokenized bonds issued between August 2018 and November 2025. The data reveals an accelerating market: 88% of those issuances occurred in the most recent three years, with two-thirds of issuers domiciled in Germany.
Matching 41 tokenized bonds against 546 conventional bonds, the ECB found statistically significant efficiency gains:
These findings carry a caveat: the sample is small, the market is nascent, and 91% of issuers were non-financial and financial corporations. Public entity issuances — such as those by the European Investment Bank and the World Bank — represent a growing but still minority share.
The settlement infrastructure gap is the binding constraint. Without central bank money available on-chain, tokenized bonds settle in commercial bank deposits or require off-chain settlement legs, negating the efficiency gains of atomic DvP. Pontes addresses this gap directly.
DLT operates continuously. Central banks do not. This mismatch is the primary operational constraint on Pontes at launch.
The ECB conducted a consultation that received 125 responses on the question of extended TARGET operating hours. The initial Pontes launch in September 2026 will operate within limited hours. A version with 24/7 functionality is targeted for 2028.
The problem is structural, not merely operational. As the ECB consultation documents note, if central bank funds remain locked in conventional RTGS accounts during closed hours, out-of-hours DLT transactions cannot proceed. The alternative — pre-positioning capital on DLT platforms — creates liquidity fragmentation and potentially reduces interest income for participating institutions.
The Bank for International Settlements acknowledged "the need for central banks to move towards 24/7 in a DLT world" following its Project Agorá prototype, which involved seven central banks across five time zones. The ECB's phased approach — limited hours at launch, full continuous settlement by 2028 — mirrors the cautious timeline adopted by most central bank digital infrastructure projects.
Pontes is explicitly designed as a bridge. Appia, the second track approved alongside Pontes, aims to deliver a full architecture blueprint by 2028. The project is evaluating three infrastructure models:
Schnabel signaled a preference for the third option: "A unified ledger processing millions of transactions may become a bottleneck, whereas several specialised ledgers can distribute the workload."
The Appia decision will determine whether the eurozone builds a centralized settlement monopoly or a federated network. The choice has direct implications for the competitive position of private market infrastructure operators like Clearstream, Euroclear, and emerging DLT-native platforms.
Pontes does not operate in isolation. Circle's Arc mainnet launches five days earlier, on 16 September 2026, as a Layer-1 blockchain for institutional stablecoin settlement with Wall Street validators. In the United States, the Chainlink-Bottomline partnership announced in September 2026 targets cross-chain payment infrastructure for 600+ banks processing $16 trillion annually through Swift channels.
The divergence is notable. The U.S. approach relies on private infrastructure operators and stablecoin issuers to build settlement rails. The ECB approach asserts central bank primacy: the ultimate settlement asset must be a direct claim on the central bank, not a token backed by a reserve portfolio.
Both approaches face the same constraint: institutional adoption requires settlement finality that legal and compliance teams can underwrite. The ECB's advantage is that central bank money carries zero credit risk by definition. The private sector's advantage is speed — Circle and Chainlink are already operational, while Pontes launches with "limited operating parameters" and reaches full capability only in 2028.
Project Pontes represents the first production-grade deployment of tokenized central bank money by a major central bank. The economic logic is straightforward: if €241 trillion in European financial markets tokenizes while central bank money remains analog, the settlement layer defaults to dollar-denominated private tokens. The ECB has decided that outcome is unacceptable.
The September 21 launch is operationally limited. Full 24/7 capability arrives only in 2028 with Appia. The practical impact in the intervening period depends on adoption velocity — how quickly institutions integrate Pontes settlement into existing workflows and whether the efficiency gains documented in tokenized bond markets (40% lower yield spreads, 27% tighter bid-ask spreads) prove reproducible at scale.
The broader significance is structural. Pontes establishes a template for how central banks intend to operate in tokenized markets: not as regulators standing outside the system, but as infrastructure operators providing the settlement asset itself. Whether that model scales or calcifies into a bottleneck depends on decisions — particularly the Appia architecture choice — that remain unresolved.