The Multi-Rail Mandate: Building for Optionality

Payments are converging toward a common customer expectation: instant, always-on, cross-border-capable and increasingly programmable. Yet the infrastructures and settlement models used to deliver that experience continue to multiply, spanning domestic instant-payment systems, interconnected cross-border networks, modernized correspondent banking, stablecoins and tokenized deposits.

For leadership teams, the strategic challenge has shifted from choosing the right payment rail to building an institution that can adapt to many rails. The strategic question is no longer which rail will win. Multiple rails will coexist, and the optimal rail will vary by customer, use case, geography, economics and regulatory environment. What matters is whether banks and payment service providers can deliver that outcome, regardless of the rail, corridor, scheme or settlement model behind it.

Competitive advantage increasingly lies in institutional optionality: the technology, data, fraud, risk, liquidity, governance and operating capabilities that work in real time and can be reused across payment models. The objective is not to become rail-blind. It is to ensure that the next rail, cross-border corridor or settlement model does not trigger another multiyear transformation.

The winners will not necessarily be the institutions that predicted the right rail. They will be the ones that made deliberate, no-regret investments in reusable capabilities and can integrate new payment types quickly as customer demand and market structures evolve.

Payment Rails Are Proliferating and Becoming Increasingly Connected

Instant-payment infrastructure is now widespread, and cross-border interoperability is advancing. Fast payment systems are now live in over 100 jurisdictions worldwide. The next wave of investment is increasingly focused not simply on building another domestic rail, but on connecting rails, abstracting their complexity and extending real-time capabilities across borders.

Bilateral links between domestic systems already work in production (Unified Payments Interface and PayNow-UPI) and are giving way to multilateral designs to let domestic systems connect once and reach multiple markets (Nexus Global Payments). At the same time, established networks such as SWIFT are modernizing, and tokenized deposits and programmable settlement models are moving from experimentation toward production use cases. Tokenization can also change the settlement model itself, bringing money, assets and transaction logic onto programmable platforms to enable conditional or atomic settlement and reduce reconciliation and manual handoffs (Project Agorá).

The expectation is that banks will increasingly operate a portfolio of payment rails and settlement mechanisms, instead of one payment model replacing every other model.

The Hard Problem Is Now Inside the Institution

The infrastructure can move money in seconds, yet many financial institutions still process payments over days and within specific windows.

Swift research found that approximately 80% of the elapsed cross-border payment time occurs after the payment reaches the beneficiary bank, often because of regulatory controls, manual processes and the absence of 24/7 infrastructure and operations.

A rail can settle in seconds, but the customer will not experience speed if:

Fraud operations close overnight or lack real-time decisioning
Treasury manages liquidity around business hours
Exceptions wait for a next-day queue
Reconciliation only runs overnight
Core systems cannot post and confirm continuously
Data is lost or degraded as a payment moves across systems

This is a point of customer friction when customers pay more for instant payments that behave like traditional payments.

Turning optionality into an operating model requires five key capabilities that are strategic design choices:

Move to 24/7 Operations
Always-on payments eliminate the banking calendar for a transaction but not necessarily for the bank behind it. Banks need continuous monitoring, clear overnight and weekend decision rights, intelligent and automated exception management and service levels designed around minutes rather than the next business day.

A 24/7 rail sitting on top of a five-day operating model is not a 24/7 payments capability.

Assess Risk and Fraud From the Start of Payment
Instant payments compress the control window. Final and irrevocable payments leave significantly less time to identify suspicious activity, investigate alerts or reverse errors. The U.S. Federal Reserve has identified the irrevocable, real-time nature of instant payments as a fraud-management challenge.

The operating model must shift from detect and repair to predict and prevent. That requires:

Real-time fraud scoring and behavioral signals
Pre-transaction controls proportionate to risk
Continuously refreshed customer, counterparty and device intelligence
Automated alert triage with governed escalation for decisions related to sanctions and financial crime
Clear accountability for model performance, exceptions and customer remediation

Manage Liquidity Continuously, Not at End of Day
A payment system that settles continuously creates a treasury problem if liquidity is still managed around traditional market hours. The challenge cuts across payment models: Instant payment services require liquidity to be positioned and managed continuously, while stablecoins and tokenized deposits introduce additional funding, redemption and settlement liquidity requirements. At scale, stablecoins could also shift deposits and alter bank funding dynamics.

The European Central Bank has identified 24/7 liquidity sourcing and stress management as a central challenge created by instant payments.

Banks need real-time balance visibility, intraday forecasting, pre-funding rules, automated funding thresholds, stress scenarios and clear governance for weekend and overnight funding events.

Define the Strategic Architecture: Standardize the Core and Modularize the Edge
Most banks did not build one payments architecture; they accumulated many. Each rail can bring its own message translation, routing, fraud logic, reconciliation and operating process, so adding another rail often means building another vertical stack.

One fix is a common core supported by modular adapters for what is genuinely rail-specific, though not everything abstracts equally well. Message translation and reconciliation can run through one shared core, with a thin adapter per rail handling formatting and rail-specific rules. Liquidity and fraud abstract less cleanly: Correspondent banking still runs on pre-funded nostro balances, a fundamentally different liquidity model than real time settlement against central bank reserves, and instant, irrevocable payments compress the fraud-detection window in ways card payments do not.

However, a fully unified core is not the only strategic starting point. A lightweight orchestration layer or payments hub with shared routing logic calling separate rail-specific services can capture much of the benefit faster, trading some duplication for speed. The appropriate design should follow the institution’s target capabilities, existing estate, risk appetite and sequencing constraints. For more, see the companion article, “The Payment Hub as a Bank Accelerator: Turning Optionality Into Execution.”

Govern Capabilities, Not Individual Rails
Payments investment is commonly funded rail by rail: one program for domestic instant payments, another for cross-border modernization, another for digital assets, and so on. That structure can duplicate spending and preserve fragmentation even as the bank modernizes individual components.

Leadership should establish enterprise ownership, investment governance, and measurable reuse targets for the capabilities that sit beneath every rail. The goal is not to centralize every decision but to ensure that shared capabilities are deliberately owned, funded and reused to optimize efficiency by reducing technical and operational overhead.

What Leadership Should Decide Now

The decisions below establish the “why” and the enterprise direction.

Banks do not need to predict the winning infrastructure before acting. They need to separate rail decisions, which can be staged by market and use case, from readiness decisions, which increasingly cannot.

Start With Customer Outcomes, Not Rails
Map what priority customer segments require and where today’s internal model falls short. The optimal rail should follow the required outcome and not the other way around.
Prove Reuse Before Scaling
Choose a commercially important payment flow and deliver it using shared capabilities such as common data, real-time risk controls, liquidity visibility, standardized reconciliation and continuous operations. Then, reuse those capabilities across a second rail, corridor or use case to ensure capabilities are truly shared.
Fund and Govern Shared Capabilities as Infrastructure
Treat 24/7 operations, real-time risk, continuous liquidity, structured data and orchestration as reusable enterprise capabilities rather than repeatedly funding them within individual rail programs.

The payoff is commercial as well as technical: faster product and corridor launches, lower incremental operating and control costs, stronger fraud prevention, more effective use of payment data, greater resilience and less risk of losing valuable flows to fast-moving competitors.

That is the no-regret strategic roadmap: make fewer bets on which rail will win and invest more consistently in the institutional capabilities needed to operate across plausible futures. Once leadership has set that direction, the next question is implementation. A companion article, “The Payment Hub as a Bank Accelerator,” examines when and how a payment hub can translate this strategy into a common processing and operating model, and what must change around the technology for the promised speed to reach clients.

Footnotes:

1: Frost, Jon, et al., “Fast payments: design and adoption,” BIS Quarterly Review (Mar. 4, 2024).

2: Monetary Authority of Singapore, Press Release, “Launch of Real-time Payments between Singapore and India” (Feb. 21, 2023).

3: Bank for International Settlements, “Project Nexus: enabling instant cross-border payments” (Aug. 27, 2025).

4: Swift, Press Release, “Swift’s blockchain ledger ready for use as 17 banks set to pioneer tokenised cross-border payments on trusted global infrastructure” (July 9, 2026).

5: Bank for International Settlements, “Project Agora: Exploring tokenisation of wholesale cross-border payments” (May 27, 2026).

6: Swift, “Unlocking last mile speed in cross-border payments” (July 20, 2026).

7: Vall-Llobera, Joseph and Goldberg, Susan, “Key Risk Considerations for Implementing Instant or Real-Time Payments,” Federal Reserve Bank of Atlanta (Feb. 16, 2023).

8: Jørgensen, Thomas Broeng, et al., “Sound practices for intraday liquidity risk management,” European Central Bank, Banking Supervision (Nov. 13, 2024).

9: Bank for International Settlements, Committee on Payments and Market Infrastructures, “Correspondent Banking – final report” (July 13, 2016).

The views expressed herein are those of the author(s) and not necessarily the views of FTI Consulting, Inc., its management, its subsidiaries, its affiliates, or its other professionals.