You can pay a street vendor in Bengaluru in two seconds. Yet sending the same amount to a family member in Singapore or a supplier in Dubai can still take days, pass through several banks, and cost a surprising slice of the money along the way.
That gap is one of the most stubborn problems in global finance. The good news is that 2025 and 2026 have brought real movement: new multilateral platforms, maturing CBDC experiments, and settlement techniques like atomic swaps that could finally make cross-border money behave like domestic money.
Why cross-border payments are still hard
A domestic payment moves inside one currency, one legal system and one settlement network. A cross-border payment must bridge all three.
Traditionally, that bridge is correspondent banking. Your bank may not have a relationship with the recipient's bank, so the payment travels through one or more intermediary banks that hold accounts with each other. Each step adds:
- Cost, through fees and foreign exchange spreads.
- Delay, as each bank processes, screens and settles.
- Opacity, because the sender often cannot see where the payment is or what it will cost in total.
- Risk, because one leg of the exchange can settle while the other has not.
Operating hours, time zones, compliance checks and inconsistent data formats multiply the friction.
What it costs today
The World Bank's Remittance Prices Worldwide database tracks the cost of sending $200 across hundreds of country corridors. Its data shows the global average cost was 6.49% in the first quarter of 2025, more than double the UN Sustainable Development Goal target of 3%.
The channel matters enormously:
- Banks were the most expensive channel, averaging around 9.50%.
- Digital providers averaged around 3.65%, much closer to the target.
- South Asia was the cheapest receiving region, at around 5.18%, helped by strong competition on routes into India, Pakistan and Bangladesh.
For a migrant worker sending money home every month, those percentages are not abstract. They are school fees, medicines and savings.
The G20 targets, and the reality check
In 2020, the G20 launched a roadmap to make cross-border payments faster, cheaper, more transparent and more accessible. The Financial Stability Board later set measurable targets. By end-2027, the ambitions include:
- Retail payments: global average cost no more than 1%, with no corridor above 3%.
- Speed: 75% of retail payments and remittances available to the recipient within one hour, and the rest within one business day.
- Wholesale payments: 75% credited within one hour, with full reconciliation by end of day.
- Access: every end user with at least one option for sending or receiving cross-border electronic payments.
- Remittances cost: in line with the UN goal of 3% on average by 2030, with no corridor above 5%.
In October 2025, the FSB's consolidated progress report was candid: policy milestones had been reached, but they had not yet translated into tangible improvements for end users at the global level, and satisfactory improvement in line with the 2027 timetable is unlikely. Fragmented regulation, inconsistent AML/CFT requirements, slow infrastructure upgrades and reliance on correspondent banking were cited as obstacles.
That is not a reason for pessimism. It is a clear signal that new infrastructure is needed.
Three big shifts in cross-border infrastructure
1. Linking fast payment systems: Project Nexus
Many countries now have instant domestic payment systems, like India's UPI. Project Nexus, launched by the BIS Innovation Hub Singapore Centre in 2022, aims to connect them. Instead of each country building a custom link to every other country, each payment system makes one connection to a shared platform and gains reach across the network.
The target experience is ambitious: cross-border payments from sender to recipient within 60 seconds in most cases.
In 2025, the central banks and payment system operators of India, Malaysia, the Philippines, Singapore and Thailand incorporated Nexus Global Payments (NGP), a not-for-profit company in Singapore, to take Nexus into live operation. The BIS has handed the project over to NGP and now serves in an advisory capacity. Public statements from partners have pointed to first live transactions around 2027.
2. Multi-CBDC platforms: mBridge
Project mBridge takes a different approach: a shared ledger where multiple central bank digital currencies can be exchanged directly. Initiated by the BIS Innovation Hub with the central banks of Hong Kong, Thailand, China and the UAE, it was joined by the Saudi Central Bank in 2024 and reached minimum viable product stage in mid-2024. In October 2024 the BIS announced it was handing the project to its partners.
mBridge demonstrates that real-time, peer-to-peer cross-border payments and FX transactions in central bank money are technically possible, while also highlighting how governance and geopolitics shape multilateral platforms.
3. Tokenised correspondent banking: Project Agorá
Project Agorá, involving seven central banks and more than 40 financial institutions, explored combining tokenised commercial bank deposits with tokenised central bank reserves on one programmable platform. Its May 2026 report showed the prototype could support atomic, multi-currency, always-on settlement, with smart contracts embedding compliance and conditional triggers.
And India's own path
The RBI has stated in its Annual Report 2025-26 that it will explore bilateral and multilateral cross-border CBDC pilots in 2026-27 with select use cases. It signed an MoU on digital asset collaboration with the Monetary Authority of Singapore and has held discussions with MAS and the Central Bank of the UAE on operationalising a cross-border CBDC pilot. India is also participating in BIS projects including Mandala (Phase 2) and Rialto.
The richer data layer: ISO 20022
Infrastructure is only half the story. Payments also need a common language.
On 22 November 2025, the coexistence period between legacy MT messages and ISO 20022 for cross-border payments on Swift ended. Payment instructions between financial institutions on Swift must now use ISO 20022, which carries structured data about parties, purposes and remittance information.
Richer data means better screening, fewer false positives, faster straight-through processing and far easier reconciliation, provided the systems on either end can use it.
What is an atomic swap?
Now for the concept that ties this together.
An atomic swap is an exchange in which both sides settle together, or neither settles at all. The word "atomic" comes from computing, where an atomic operation is indivisible: it either completes fully or does not happen.
In finance, this principle has long existed under other names:
- Payment versus Payment (PvP): one currency is delivered only if the other currency is delivered.
- Delivery versus Payment (DvP): a security is delivered only if payment is made.
The traditional problem is settlement risk, sometimes called Herstatt risk: one party pays out its currency, but the counterparty fails before delivering the other currency.
How atomic swaps work on programmable ledgers
On a programmable platform, an atomic swap can be enforced by code rather than by trust or intermediaries. A simplified flow looks like this:
- Agree terms. An Indian business wants to pay a supplier in Singapore. The rupee amount, the Singapore dollar amount and the exchange rate are fixed.
- Lock both legs. The rupee-side value and the Singapore dollar-side value are locked in smart contracts, often using cryptographic conditions such as hash time-locks.
- Verify conditions. The contract checks that both legs are properly funded and that compliance checks have passed.
- Settle together. Both transfers execute in one indivisible step.
- Or unwind. If any condition fails or a time limit expires, both locks release and funds return to their owners. No one is left exposed.
The result is near-instant finality with no counterparty settlement risk on the exchange itself.
Why atomic settlement matters for real people and businesses
- Families receive money faster, with less value lost in the middle.
- Exporters and importers stop tying up working capital while waiting for settlement.
- Treasurers can see exactly when money moved, without chasing correspondent confirmations.
- Banks reduce the liquidity they must pre-fund in foreign accounts.
- Regulators get cleaner, more transparent transaction records.
What still has to be solved
Atomic swaps are powerful, but a credible picture includes the hard parts:
- Legal finality: laws must recognise ledger settlement as final.
- Compliance across jurisdictions: sanctions screening, AML and capital flow rules differ by country. Projects like BIS Mandala are exploring ways to embed these checks into transaction protocols.
- Liquidity: someone must provide currency on both sides, especially for less-traded corridors.
- Interoperability: different CBDC systems, fast payment networks and bank ledgers must connect securely.
- Governance: multilateral platforms need clear rules on access, oversight and dispute resolution.
How E1 envisions cross-border money
E1's vision is a world where money moves across borders without crossing borders: instantly, transparently and with the trust of central banks behind it.
E1 is building cross-border payment infrastructure designed around the RBI's CBDC framework and atomic swaps, so that both sides of a currency exchange settle together. Our approach combines:
- CBDC-native settlement designed for faster, lower-cost transfers with blockchain-based security and sidechain flexibility for scale.
- Datachain interoperability, E1's layer designed to enable trustless communication and asset transfer across heterogeneous ledgers, so one network can settle with another.
- Chainpay orchestration, designed to accept familiar payment methods and translate successful transactions into cross-chain settlement events.
- A cross-chain enterprise settlement layer, designed so that once a fiat payment clears, cross-border invoices settle and trade documents synchronise across institutional networks.
- Automated cross-border payroll and contractor payouts, designed to route payments through the most efficient rail to a recipient's local account or wallet.
The global targets are clear, and the infrastructure to reach them is finally taking shape. E1's goal is to help connect those pieces into one seamless experience, where sending money to Singapore feels as simple as paying a vendor down the street.
Sources
