SWIFT gpi vs. Local Rail Interconnects: Choosing the Right Cross-Border Stack

The choice of cross-border payment infrastructure is no longer a binary decision between “use SWIFT” and “find an alternative.” A layered landscape has emerged: SWIFT gpi has materially improved the base correspondent banking experience; bilateral interoperability between national real-time payment rails is creating direct corridor connections; and card network B2B products from Visa and Mastercard are competing for corporate payment volume in the mid-market. Understanding what each approach actually delivers — and where each falls short — is the operational knowledge that treasury and finance teams need to make intelligent routing decisions.

What SWIFT gpi Delivers

SWIFT’s Global Payments Innovation initiative, launched in 2017, tackled the most acute pain points in correspondent banking without requiring banks to rebuild their underlying infrastructure. Three improvements were substantive.

Speed commitments: gpi introduced a “same day use of funds” standard, requiring that payment instructions be processed within the business day. By 2023, SWIFT reported that over 50% of gpi payments were credited within 30 minutes, and approximately 40% within five minutes. These figures represent a genuine step change from the two-to-three day norms of pre-gpi correspondent banking on major corridors.

End-to-end tracking: Every gpi payment carries a unique end-to-end transaction reference (UETR) that allows the sending bank — and, in implementations that expose it, the corporate sender — to track the payment’s status through each correspondent hop. The gpi Observer service monitors adherence to service levels and provides exception tracking. For corporate treasuries managing hundreds of supplier payments across multiple currencies, real-time visibility into payment status reduces the manual investigation burden substantially.

Fee transparency: gpi introduced requirements for correspondent banks to pass payment amounts without deduction unless explicitly authorized, and to disclose any deductions that do occur. This addresses the “short payment” problem that made reconciliation difficult under traditional MT messaging — recipients now receive the expected amount, and any FX conversion and fee information is structured and machine-readable.

Where gpi still falls short is the last mile and the FX layer. SWIFT gpi optimizes the inter-bank messaging and correspondent chain coordination; it does not change the economics of FX conversion, which remains determined by individual bank treasury pricing policies. Spreads on commercial FX through gpi-enabled bank accounts remain meaningfully wider than the interbank rates available through specialist payment platforms. Additionally, gpi’s tracking and speed improvements are contingent on all banks in the payment chain being gpi-enrolled. Payments that exit the gpi network into non-enrolled correspondent banks — which still occurs, particularly on less-traveled corridors — lose their tracking capability and speed guarantees.

How Bilateral Local Rail Links Work

A structurally different approach to cross-border payments is emerging from the interconnection of national real-time payment rails. Rather than routing international payments through correspondent banks, bilateral rail links allow a domestic payment instruction in one country to trigger a domestic payment in another country, with a currency conversion happening at the junction.

The Singapore-Thailand QR code interoperability link, operational since 2021, is the clearest example of this model in practice. A Thai consumer with a Thai bank account can scan a PayNow QR code at a Singaporean merchant using their Thai banking app. The Thai bank sends a PromptPay instruction, a bilateral gateway handles the THB/SGD conversion at a pre-agreed rate structure, and the merchant receives a domestic PayNow credit in SGD. The payment clears in seconds and costs close to the domestic transaction fee for both parties.

By 2025, Singapore had established similar bilateral real-time QR interoperability links with Malaysia (DuitNow), Indonesia (QRIS), India (UPI), and the Philippines (InstaPay). These are bilateral agreements — each corridor requires a separate negotiation between the two central banks, technical integration between the two national payment operators, and a liquidity and FX arrangement between designated settlement banks. This makes scaling them labor-intensive: 10 countries with bilateral links to all others requires 45 separate corridor agreements.

Project Nexus, the BIS Innovation Hub initiative, is designed to solve the multilateral scaling problem. Rather than requiring bilateral agreements between every pair of participating countries, Nexus defines a common API and data standard that national instant payment systems plug into once. Any two Nexus-connected systems can then exchange payments without a separate bilateral negotiation. India, Malaysia, Philippines, Singapore, and Thailand are the founding members, with the Nexus network targeted to be operational for broader use by 2026-2027. If it delivers at scale, a corridor like Malaysia-Philippines — currently expensive and slow due to limited banking connectivity — becomes as fast and cheap as a domestic transfer.

Visa B2B Connect and Mastercard Move

The card networks have entered the corporate cross-border payments space with dedicated infrastructure that sits outside both the correspondent banking network and the real-time rail ecosystem.

Visa B2B Connect, launched commercially in 2019, is a non-card, account-to-account network specifically designed for high-value corporate payments. It uses distributed ledger technology to enable direct, bank-to-bank settlement between enrolled financial institutions, bypassing the traditional correspondent chain. Visa positions it for payments in the $1,000-$1,000,000 range — too large for card transactions, too important for correspondent banking delays. Coverage spans over 100 countries. Settlement can occur within one to two business days, with better FX transparency than traditional wires.

Mastercard Move is a broader product suite covering both consumer and business disbursements, leveraging Mastercard’s network reach to push funds to bank accounts, debit cards, and wallets globally. For corporate use cases — paying international contractors, distributing insurance claims across borders, settling marketplace payouts — Mastercard Move offers speed and reach advantages over traditional wires on covered corridors.

Both card network products occupy a specific niche: they work best for organizations that are already embedded in the Visa or Mastercard ecosystem, have predictable high-volume corridors, and value the simplicity of a single API integration over optimizing for minimum FX spread. They are not cost-minimization tools — the network economics include card-network margin — but they compete on speed, reliability, and operational simplicity.

Choosing the Right Approach

The decision framework for cross-border payment routing in 2026 looks roughly as follows. For high-value, complex, multi-bank treasury operations — intercompany transfers, large supplier payments, regulated financial flows — SWIFT gpi remains the standard, particularly on major USD, EUR, GBP corridors where gpi enrollment among correspondent banks is near-universal and the tracking and compliance infrastructure is battle-tested.

For retail and SME payments on corridors with live real-time rail interconnects — SGD-THB, SGD-MYR, SGD-INR, and increasingly the broader Nexus network — local rail links offer materially better speed and cost. A payments product designed for Southeast Asian workers sending remittances home should not be routing through SWIFT when a bilateral real-time rail link exists.

For corporate disbursements at scale — marketplace payouts, gig economy contractor payments, insurance claims — the card network B2B products and specialist platforms like Airwallex and Wise Business offer the best combination of coverage, speed, and operational simplicity.

The longer-term trajectory points toward a more fragmented but more capable routing landscape. SWIFT gpi will continue improving, ISO 20022 data richness will increase straight-through processing rates, and the Nexus multilateral network will (if it delivers) dramatically improve the economics of intra-ASEAN and ASEAN-India payments. The treasury teams and payment product managers who understand the strengths and boundaries of each layer will route intelligently across them; those who default to a single provider will leave speed and cost optimization on the table.