CPMI and Cross-Border Payments: How the BIS Committee Is Using Tokenization to Fix International Transfers

Cross-border payments are expensive, slow, and opaque — and they have been for decades. Sending money internationally through the correspondent banking system takes an average of one to five business days, costs an average of 6.5% for the remittance corridor that matters most to migrant workers (sending money home to developing countries), and provides limited transparency about fees, exchange rates, and arrival times. The World Bank tracks remittance costs through its Remittance Prices Worldwide database, and the consistent finding is that the cost of moving money internationally is far higher than the cost of the economic service being provided.

The G20 identified cross-border payment reform as a priority during Saudi Arabia’s 2020 presidency and established quantitative targets: average cost below 3% by 2027, end-to-end processing within one hour for the majority of cross-border transactions, and universal access to payment services for populations currently excluded. These targets are ambitious. Meeting them requires addressing structural features of the correspondent banking system that have resisted reform for decades.

The Committee on Payments and Market Infrastructures — the BIS committee responsible for payment standards — is the primary international technical body tasked with implementing the G20 cross-border payments roadmap. Its role in identifying tokenization and CBDC interoperability as key tools for achieving the G20 targets has been the most significant institutional endorsement of tokenization as payment infrastructure anywhere in the international system.

Why Correspondent Banking Is Expensive

Understanding why CPMI is looking at tokenization requires understanding why the current system is expensive. International transfers typically move through a chain of correspondent banks — institutions that maintain accounts with each other in different currencies and process transfers on each other’s behalf. A dollar payment from the Philippines to Mexico might pass through a Philippine bank, its US dollar correspondent bank in New York, the receiving bank’s US dollar correspondent in New York, and finally the receiving bank in Mexico. Each step involves fees, each correspondent bank uses liquidity to fund the transfer before being reimbursed, and the end-to-end process requires multiple bilateral reconciliation processes that take days.

The cost of correspondent banking is not primarily technology — it is the cost of liquidity management (banks must hold balances in each other’s accounts to facilitate transfers), compliance (each bank applies its own AML screening to each transfer), and the overhead of bilateral reconciliation (each correspondent relationship requires settlement processes that consume operational resources).

Tokenization addresses this cost structure directly. If a dollar payment can be represented as a tokenized dollar — a digital token that carries the value of a dollar and settles finally when transferred from one address to another — then the transfer from the Philippines to Mexico can potentially be executed as a single atomic transaction, without the chain of correspondent banks and without the multilateral reconciliation problem. The liquidity management cost is reduced by same-day or real-time settlement. The compliance burden is potentially reduced by shared infrastructure that applies AML screening once rather than at every node in the chain.

Project Nexus: Linking National Payment Systems

CPMI’s most concrete contribution to the cross-border payments agenda is Project Nexus — a platform designed to link domestic instant payment systems across multiple countries into an interoperable cross-border network. Nexus is not a new payment rail. It is a connector layer that translates between different national systems’ messaging formats, compliance requirements, and settlement mechanisms, allowing a payment sent from a domestic account in one country to arrive at a domestic account in another country in near-real time.

The Nexus approach has the advantage of building on existing domestic infrastructure rather than requiring countries to build new cross-border payment systems from scratch. Countries that have invested in domestic real-time payment systems — the UPI in India, the PromptPay in Thailand, the PayNow in Singapore — can connect to Nexus and immediately offer cross-border capability to their existing customer bases. The technical work is in the translation layer, not in the payment system itself.

CPMI’s Nexus work has engaged with tokenization as a settlement mechanism. If participating countries’ central banks issue wholesale CBDCs, those CBDCs can serve as the cross-border settlement asset on the Nexus platform — enabling payment finality in central bank money without the correspondent banking chain. This is the link between the cross-border payments agenda and CBDC development: CBDC becomes valuable not only for domestic payment modernisation but as the settlement instrument for international payment infrastructure.

The Policy Legitimisation of Tokenization

The significance of CPMI’s cross-border payments work for the broader tokenization policy debate is the institutional framing it provides. When the BIS committee responsible for payment standards tells the G20 that tokenization and CBDC interoperability are tools for achieving the G20’s cross-border payment targets, it moves tokenization from the speculative-innovation category to the infrastructure-policy category.

This framing matters for national regulators deciding how to treat tokenized payment instruments. A tokenized dollar used in a Nexus cross-border payment context looks different from a speculative crypto token under this framing — it is infrastructure solving a defined policy problem, validated by the international standards body responsible for payment system standards.

The policy legitimisation effect is cumulative. IMF research on cross-border payment efficiency, World Bank data on remittance costs, CPMI standards work on Nexus, and FSB analysis of tokenized asset settlement all reinforce the same policy conclusion: tokenization of payment instruments and financial assets is a legitimate tool for addressing real inefficiencies in the international financial system. The combined weight of these institutional endorsements makes it harder for national regulators to maintain purely restrictive postures toward tokenized payment instruments.

Remaining Challenges

The CPMI’s work has been honest about the challenges that technical solutions cannot resolve. Domestic legal frameworks for instant payment systems vary in ways that create friction at international boundaries. AML compliance requirements differ across jurisdictions, and harmonisation is politically difficult. Foreign exchange conversion — the economic operation that underlies most cross-border payments — requires liquid FX markets that tokenization does not create.

The most optimistic reading of the CPMI agenda is that it solves the plumbing problems in cross-border payments — making the technical infrastructure faster and cheaper — while the underlying economic and regulatory challenges (exchange rate risk, capital controls in some markets, varying KYC/AML standards) require parallel policy work that CPMI cannot drive alone. The pessimistic reading is that technical solutions to payment infrastructure problems are necessary but insufficient, and that the G20 targets require political commitments on regulatory harmonisation that the international community has consistently failed to make.

The realistic assessment is somewhere between: Nexus and CBDC interoperability will meaningfully reduce the cost and time of cross-border payments in corridors where participating countries are willing to align their regulatory frameworks. Universal transformation of the correspondent banking system is a longer-term ambition.