Fifty-plus banks across 16 countries, representing more than $10 trillion in combined assets under management, launched Project Pangea on June 23, 2026, at Point Zero Forum in Zurich. The initiative targets T+0 atomic settlement for the EUR-KRW foreign exchange corridor — a direct challenge to th...
"Project Pangea upgrades the fragmented foreign exchange model of today with direct, atomic currency swaps using stablecoins." — Fernando Vazquez, Chainlink
Fifty-plus banks across 16 countries, representing more than $10 trillion in combined assets under management, launched Project Pangea on June 23, 2026, at Point Zero Forum in Zurich. The initiative targets T+0 atomic settlement for the EUR-KRW foreign exchange corridor — a direct challenge to the T+2 standard that currently governs the $14 trillion-per-day global FX settlement market.
The project uses Chainlink's Cross-Chain Interoperability Protocol (CCIP), Chainlink Runtime Environment (CRE), and a purpose-built Pangea L1 blockchain to execute payment-versus-payment stablecoin swaps routed through existing Swift ISO 20022 messaging infrastructure. The consortium expects live transactions within 12 months. If it works, the template could extend to other currency corridors where $1.4 trillion still settles daily with full counterparty risk exposure, according to the Bank for International Settlements' 2025 Triennial Survey.
This report examines the technical architecture, participating institutions, economic rationale, and systemic risks of an effort that attempts to restructure a piece of the oldest and largest financial market in the world.
The BIS Triennial Survey, published in June 2026 with April 2025 data, measured over $14 trillion in gross FX obligations settled daily. The breakdown:
| Settlement Method | Daily Volume | Share | Risk Level | |---|---|---|---| | PvP (CLS and others) | $5.2T | 36% | Eliminated | | Intragroup netting | $5.0T | 35% | Mitigated | | Pre-settlement netting | $2.15T | 15% | Mitigated | | Timing controls | $0.4T | 3% | Mitigated | | Gross bilateral (no mitigation) | $1.4T | 10% | Full exposure |
The $1.4 trillion settling with full counterparty exposure represents a persistent systemic concern. Of that amount, 25% ($347 billion) involves trades eligible for PvP but settling bilaterally anyway. The remaining 75% ($1.05 trillion) cannot access PvP services because counterparties lack access (57%), currency pairs are ineligible (36%), or the trade type is excluded (37%).
CLS Bank, launched in 2002, settles 18 major currencies and processes roughly $8 trillion daily. But it covers only 40% of CLS-eligible trades. Non-CLS currencies — including the Korean won — account for $2.8 trillion (19%) of daily settlement, of which only 12% uses PvP.
The structural cost is substantial. The BIS estimates approximately $27 trillion sits in nostro and vostro accounts globally to support cross-border settlement. For firms operating across 20-plus jurisdictions, intraday borrowing against stranded liquidity costs 40 to 50 basis points, according to industry data.
Progress has been real but slow. Gross bilateral settlement dropped from 32% of daily volume in 2006 to 10% in 2025. But the absolute dollar exposure remains large because the FX market itself grew. Settlement fails remain low at 0.01% of obligations ($1.5 billion), but the systemic risk of a single large counterparty failure during the settlement window has not been eliminated.
Project Pangea is organized around four core entities:
Chainlink provides the middleware layer — CCIP for cross-chain token transfers, Data Streams for real-time FX pricing, and CRE for orchestrating Swift-to-blockchain translation.
FairSquareLab operates the onchain FX settlement engine, a multi-currency stablecoin liquidity pool anchored to oracle-verified prices rather than automated market maker bonding curves. It also built the Pangea L1 blockchain.
UniKA (Unified Korea Alliance) represents the Korean banking contingent. Its steering committee includes Shinhan Bank, JB Bank, Kbank, FairSquareLab, and OBDIA. More than 10 Korean commercial banks participate.
Qivalis is an Amsterdam-based consortium of 37 European banks developing a MiCA-compliant euro stablecoin. Members include ING, BNP Paribas, UniCredit, Deutsche Bank, CaixaBank, BBVA, ABN AMRO, Rabobank, Nordea, and Intesa Sanpaolo.
The combined AUM of participating institutions exceeds $10 trillion. The consortium committed to achieving live T+0 settlement on the Europe-Korea corridor within 12 months of the June 23 announcement.
Project Pangea operates on a three-layer architecture designed to avoid requiring banks to replace existing internal systems:
Layer 1 — Banking (Swift/ISO 20022): A bank initiates an FX trade through a standard ISO 20022 message sent via Swift. No custom API integration is required. This is the same messaging format banks already use for cross-border payments, meaning the 11,000+ institutions on the Swift network can theoretically participate without infrastructure upgrades.
Layer 2 — Connectivity (Chainlink CRE + CCIP + Data Streams): CRE translates the ISO 20022 message into an onchain settlement instruction. It identifies the counterparty, retrieves the current FX rate from Chainlink Data Streams, and routes the transaction to the appropriate settlement chain via CCIP. Once executed, CRE translates the resulting onchain event back into a Swift message for the originating bank's systems.
This bi-directional translation — Swift to blockchain, blockchain to Swift — is the core technical contribution. UBS has already adopted this pattern for tokenized fund subscriptions and redemptions through the Chainlink Digital Transfer Agent (DTA) standard.
Layer 3 — Settlement (Smart Contracts on Pangea L1, Ethereum, Polygon): The atomic PvP swap executes on FairSquareLab's settlement engine. Both sides of the trade — EUR stablecoin for KRW stablecoin — either settle simultaneously or not at all. There is no window where one bank has paid and the other has not, eliminating the principal risk that defines the T+2 settlement gap.
A notable design decision: on the Pangea L1, oracle data updates are guaranteed to execute ahead of all other transactions in each block. This ensures every FX swap settles against the current market price rather than a stale rate — a protection against front-running and MEV extraction that public chains do not natively provide.
The settlement layer depends on two regulated stablecoins that do not yet exist in production:
Qivalis EUR Stablecoin: On May 20, 2026, Qivalis expanded from 12 to 37 member banks across 15 European countries. The stablecoin will be 100% backed by euros and high-quality liquid assets held by regulated custodians, under planned supervision by De Nederlandsche Bank (DNB). In March 2026, Qivalis selected Fireblocks for tokenization, wallet infrastructure, custody, and compliance tooling. Launch is targeted for H2 2026. The stablecoin is designed to be MiCA-compliant from inception.
KRWQ (Korean Won Stablecoin): Developed by IQ and Frax Finance, KRWQ became the first KRW stablecoin with automated, real-time reserve verification through integration with Chainlink Proof of Reserve and Data Streams. This provides continuous onchain proof of off-chain reserves — a requirement that most existing stablecoins still handle through periodic third-party attestations.
Neither stablecoin has launched for production use. Project Pangea's 12-month timeline is therefore contingent on regulatory approvals in both Europe and South Korea, a non-trivial dependency.
The choice of EUR-KRW is strategic. EU-South Korea bilateral trade in goods reached approximately €124.25 billion in 2025, with average annual growth of 5.3% since 2011 according to the European External Action Service. The Republic of Korea is the EU's eighth-largest trade partner in goods. The project targets more than $150 billion in annual trade-related FX flows along this corridor, according to Yahoo Finance.
KRW is not among the 18 currencies settled through CLS Bank. This means EUR-KRW trades fall into the non-CLS category where only 12% use PvP settlement. The corridor represents exactly the kind of underserved market where settlement risk is highest and incumbent infrastructure is weakest.
South Korea has also made structural FX market reforms that enable this approach. In July 2024, trading hours were extended until 2:00 a.m. KST, overlapping with London and New York sessions. By mid-2025, 16-21% of daily KRW trading occurred during extended hours. The Registered Foreign Institution (RFI) system, introduced in 2023-2024, allows accredited foreign banks and asset managers to trade KRW directly onshore, with approximately 40 RFIs participating by late 2024.
Project Pangea creates a new value chain that redistributes settlement economics:
Banks stand to gain from reduced nostro account prefunding requirements. If atomic settlement eliminates the T+2 window, capital currently trapped in correspondent banking accounts can be redeployed. For the 50-plus institutions involved, even marginal improvements in capital efficiency across a $150 billion annual corridor would be material.
Chainlink captures value through enterprise fees converted to LINK tokens held in Chainlink Reserve, according to project documentation. LINK currently trades at approximately $7.98 with a market capitalization of $5.8 billion. The gap between Chainlink's infrastructure ambitions — processing for institutions managing $10 trillion in assets — and its current token valuation represents either an opportunity or a mispricing, depending on whether institutional adoption translates to meaningful fee revenue.
FairSquareLab monetizes the settlement engine and Pangea L1 infrastructure. The oracle-priority block design creates a controlled environment that may attract institutional users uncomfortable with public chain MEV dynamics, but concentrates infrastructure dependency in a single entity.
Stablecoin Issuers (Qivalis and KRWQ operators) generate yield on reserves — the standard stablecoin business model. The 37-bank Qivalis consortium distributes this revenue across member institutions, creating a cooperative structure unusual in stablecoin markets where Circle and Tether dominate.
Regulatory Contingency: Both stablecoins require regulatory approval before live settlement can begin. The Qivalis EUR stablecoin needs DNB authorization under MiCA. KRWQ faces South Korea's evolving crypto regulatory framework. Either delay breaks the 12-month timeline.
Counterparty Concentration: The Pangea L1 is operated by FairSquareLab, a single entity. Oracle priority ordering eliminates MEV but introduces a trust assumption. If the oracle feed fails or is manipulated, atomic settlement becomes atomic failure. The system trades one form of counterparty risk (bilateral settlement) for another (infrastructure dependency).
Liquidity Bootstrapping: Atomic PvP settlement requires sufficient stablecoin liquidity on both sides of the trade at all times. For a $150 billion annual corridor, the liquidity pool must handle peak-day volumes without slippage. How this liquidity is sourced, incentivized, and maintained is not fully specified in public documentation.
Scalability Beyond EUR-KRW: The architecture is designed for one corridor. Expanding to other non-CLS currency pairs (there are dozens) requires new stablecoin issuers, new regulatory approvals, and new banking consortia for each corridor. The template may be replicable, but replication is neither automatic nor cheap.
Incumbent Response: CLS Bank has not been static. CLSNet, launched in 2018, provides bilateral netting for 120-plus currencies. CLS could extend PvP services to KRW and other non-CLS currencies, potentially undermining Pangea's value proposition before it reaches production. The BIS has explicitly recommended expanding PvP membership access and eligible currency coverage.
Oracle and Price Feed Risk: Using oracle-verified prices rather than AMM bonding curves improves price accuracy under normal conditions but creates single-point-of-failure risk. During periods of extreme FX volatility, oracle latency or data-source outages could halt settlement entirely.
Project Pangea is the most specific institutional attempt to date to apply blockchain-based settlement to a defined FX corridor. It is not a whitepaper or a testnet experiment — it involves named banks, identified stablecoins, a concrete currency pair, and a public timeline.
The economic logic is sound. The EUR-KRW corridor is underserved by existing PvP infrastructure. Korean FX market reforms have created conditions for direct onshore access. The Swift-to-blockchain translation layer addresses the integration barrier that has stalled most institutional blockchain adoption.
But the project faces compounding dependencies: two unissued stablecoins, two regulatory regimes, a single-entity settlement chain, and a liquidity bootstrapping challenge for a $150 billion annual corridor. Each is solvable in isolation. Together, they represent a coordination problem that has historically been the graveyard of multi-bank blockchain initiatives.
The BIS data makes one thing clear: FX settlement risk is a real, measurable, trillion-dollar problem. Whether Project Pangea or CLS extension or some other approach solves it for non-CLS currencies, the market will not wait indefinitely. The $1.4 trillion settling without risk mitigation is $1.4 trillion of daily systemic exposure that regulators have identified, measured, and flagged.
The next 12 months will determine whether 50 banks can coordinate faster than incumbents can adapt.