perspective

Cross-Border 2030

By 2030, cross-border payments won't be dominated by a single rail. They'll stratify across four parallel layers: wholesale tokenized, regulated stablecoin, fintech local accounts, and legacy correspondent.

By Anton Titov, Founder · Plexo Institute

Published 2026-05-17

A four layer map of the 2030 cross border market: wholesale tokenized settlement, regulated stablecoin clearing, fintech local accounts, and legacy correspondent banking.

Reader Brief

By 2030, cross-border payments won't be dominated by a single rail. They'll stratify across four parallel layers: wholesale tokenized, regulated stablecoin, fintech local accounts, and legacy correspondent - each serving different counterparties.

Reading Guide

Four ideas to anchor the read.

The perspective starts from a simple premise: the 2030 market is not one modernized rail. It is a stratified market where each architecture wins the flows that fit its counterparty profile, transaction size, and regulatory burden.

Four-layer end state by 2030: wholesale tokenized, regulated stablecoin, fintech local accounts, and legacy correspondent.

The 2030 cross-border market is not a single modernized rail. Each layer captures a defined counterparty segment. Layer 1, with BIS Project Agora as a reference and JPM Kinexys, Partior, and Societe Generale EUR CoinVertible as adjacent examples, handles wholesale interbank value. Layer 2, regulated stablecoin operators under MiCA, GENIUS, UAE PTSR, MAS PSA, and ADGM FRT, handles B2B and SME flows where local banking is hard. Layer 3, Wise, Airwallex, Deel, and successors, handles consumer and SME flows in mature corridors with viable local banking. Layer 4, correspondent banking, retreats to high-value corridors with complex documentary requirements where legal certainty justifies the capital cost.

$1 quadrillion IMF projection by 2030. No single architecture absorbs that scale.

IMF Working Paper WP/25/127 projects global cross-border payment volume reaches $1 quadrillion annually by 2030. The implication: the question is not which architecture wins, but which architecture captures which segment. Layer 1 will likely handle 10-30% of wholesale interbank value, with small transaction count but large absolute value. Layer 2 will likely handle the largest share of EM-to-EM and a significant share of G7-to-EM flows. Layer 3 will dominate consumer remittance in mature corridors. Layer 4 contracts but does not disappear; it concentrates on G7-to-G7 and major-EM-to-G7 high-value flows.

Three forces shaping the 2030 mix: regulatory convergence, Basel IV implementation, and non-USD stablecoins.

Force 1, regulatory convergence, asymmetrically benefits Layer 2. Regulated stablecoin operators were the segment most limited by regulatory ambiguity, and they gain the most from clarity. Force 2, Basel IV, accelerates the shift from Layer 4 to Layers 2 and 3 in EM corridors; McKinsey estimates 10-20% of correspondent banking volume could shift to alternative layers by 2028. Force 3, non-USD stablecoins, reduces USD intermediation through EURC under MiCA, UAE AED stablecoins, Singapore SGD, and UK GBP. Layer 2 becomes efficient for corridors that previously required a USD hop. Layer 4 loses relative advantage in non-USD corridors where it had provided liquidity depth.

Strategic implication: cross-layer ambitions fail, while layer-specific optimization wins.

Layer 1 operators, such as bank consortium networks, cannot compete with Layer 3 fintechs on consumer UX. Layer 3 fintechs cannot compete with Layer 1 banks on wholesale legal certainty. Layer 2 operators that chase Layer 1 institutional flows often lose their SME base in the process. The successful operators by 2030 will be the ones who chose their layer early and optimized for its specific variables: settlement finality and regulatory alignment for Layer 1; speed, cost, and corridor reach for Layer 2; consumer UX and mature corridor density for Layer 3; legal enforceability and high-value documentation for Layer 4.

The Stratified End State

The 2030 cross-border payment market will look like four parallel layers operating at different scales.

The 2030 cross-border payment market will not look like a single modernized version of 2025. It will look like four parallel layers operating at different scales, for different counterparties, through different rails. IMF estimates global cross-border flows will reach $1 quadrillion by 2030 [1]. No single architecture absorbs that. Each layer captures a defined segment.

Stratification map showing wholesale token settlement, regulated coins, local accounts, and legacy banks sorted by transaction value, counterparty certainty, and corridor depth. — The 2030 market is a stratified routing surface: each layer wins the flow whose certainty, size, and corridor constraints match its strengths.

$1Q Projected annual global cross-border payment volume by 2030. IMF WP/25/127 projection [1].
4 Distinct settlement layers operating in parallel. Each layer serves a different counterparty segment.

Layer 1: Wholesale Tokenized Settlement

Large institutional flows between major financial centers settle through tokenized deposit networks, wholesale CBDCs, or shared ledgers.

Large institutional flows between major financial centers settle through tokenized deposit networks, wholesale CBDCs, or shared ledger systems operated by central bank consortia. BIS Project Agora, involving 7 central banks and 41 private institutions in this analysis, is the reference architecture. Bank-issued alternatives, including JP Morgan Kinexys, Partior, and Societe Generale EUR CoinVertible, occupy adjacent positioning.

Who uses this layer: G7 and major-economy banks for wholesale interbank flows, typically above $10M per transaction.

Wholesale tokenized settlement addresses the specific needs of large balance sheet institutions: central bank finality, atomic PvP settlement, regulatory alignment across multiple jurisdictions, and integration with existing RTGS systems. BIS Papers No. 167 provides the source technical framing [2]. This layer does not reach retail or SME flows. The governance, onboarding, and compliance overhead is designed for institutions, not for payment operators. By 2030, it will likely handle a meaningful share of wholesale interbank value, with scenario estimates varying from 10% to 30% of wholesale flows, but small absolute counts of transactions.

Layer 2: Regulated Stablecoin Clearing

Licensed non-bank operators settle B2B and consumer flows through regulated stablecoin networks.

Licensed non-bank operators settle cross-border B2B and consumer flows through regulated stablecoin networks with multilateral clearing infrastructure. This is the scaled version of the fiat-sandwich architecture currently emerging. By 2030, this layer will handle the largest share of EM-to-EM flows and a significant share of G7-to-EM flows.

Stablecoin flow estimates: $10T+ adjusted trailing-12-month volume, growing to tens of trillions by 2030 as the layer matures.

Current stablecoin payment volume estimates vary widely because raw transfer volume and adjusted payment-like volume use different filters. Visa/Allium reports more than $10T in adjusted global transaction volume over the latest trailing 12 months [3]. McKinsey projects stablecoin supply will reach $1.9-4 trillion by 2030 [4], implying annual payment volume in the tens of trillions at current velocity ratios. The addressable market is large enough that this layer can accommodate substantial growth without disrupting other layers. Regulatory frameworks, including MiCA, GENIUS Act, UAE PTSR, MAS PSA, and ADGM FRT, establish the operational perimeter. Six Pathways describes the current state of this layer.

Layer 3: Fintech Local Account Networks

Local-account networks keep scaling where banking access is viable and corridor volume is dense.

Wise, Airwallex, Deel, and successors continue to scale the local-account model for consumer and SME corridors. By 2030, this layer covers most consumer remittance flows in mature corridors, including US-Mexico, UK-India, and EU-Philippines, and increasingly absorbs SME treasury flows below institutional thresholds.

Where this layer wins: balanced high-volume corridors where local banking access is viable.

The fintech local-account model depends on maintaining bank accounts in every operating jurisdiction. This works well in corridors where local banks accept fintech clients, regulatory frameworks tolerate the business model, and volume is high enough to justify the operational overhead. It does not scale to corridors where banks refuse to serve fintechs, including much of Francophone Africa and some Central Asian markets, or where regulatory frameworks are ambiguous. In those corridors, Layer 2, regulated stablecoin clearing, takes share.

Layer 4: Legacy Correspondent Banking

Correspondent banking contracts into the use cases where no alternative has equivalent legal enforceability.

Correspondent banking does not disappear by 2030. It retreats to its highest-value corridors and to use cases where no alternative has equivalent legal enforceability: large M&A payments, sovereign debt flows, and trade finance with complex documentary requirements.

Continued contraction: global CBR count likely falls 30-40% below the 2011 peak by 2030.

The BIS CPMI data shows correspondent banking relationships declined roughly 22% globally between 2011 and 2019 [5]. Basel IV implementation from 2025 to 2028 is expected to accelerate the decline as trade finance capital charges rise. By 2030, cumulative decline from the 2011 peak will likely reach 30-40%. This does not mean correspondent banking dies. It means the footprint concentrates in G7-to-G7 and major-EM-to-G7 corridors where the economics still work. EM-to-EM corridors continue migrating to Layers 2 and 3. The CBR Exodus documents the structural decline.

What Drives The Stratification

Each layer exists because it optimizes for different variables in the cross-border payments problem.

The four-layer end state is not accidental. Each layer exists because it optimizes for different variables in the cross-border payments problem. The table shows which variables each layer prioritizes.

Evidence And Sources

This raw HTML export preserves source visibility for crawler and contractor review. Indexing decision: index, follow.