Project Rialto and the drive toward cheaper global remittances

The world of cross-border retail payments—the lifeblood for millions of families sending money home—has long been plagued by a stubborn paradox. While domestic payments have become instantaneous and nearly free, sending money across national borders remains a “fragmented” journey of high costs, slow speeds, and opaque processes.

The recently released technical report on Project Rialto, a collaboration between the BIS Innovation Hub and the central banks of France, Italy, Malaysia, and Singapore, offers a compelling vision for how to fix this without tearing down the existing financial system.

Rather than proposing a radical replacement of existing financial rails, Project Rialto is a masterclass in pragmatic innovation. It operates on two distinct but tightly integrated layers:

• Layer 1 (The Customer Interface): Uses domestic Instant Payment Systems (IPS) to handle retail transactions. This means the user experience and regulatory oversight remain largely unchanged—familiarity is preserved.
• Layer 2 (The Settlement Engine): An experimental Cross-border DLT Network (XDN) where the actual “heavy lifting” of currency exchange and settlement occurs using tokenized central bank money (CeBM).

By isolating the complex FX and settlement activities from the retail processing, Rialto allows for rapid modernization without requiring every local bank to redesign its entire core infrastructure.

A cornerstone of the Rialto experiment is the use of Automated Market Makers (AMMs). Traditionally, foreign exchange in retail payments relies on manual pricing and prefunding, which are costly and slow, especially for less liquid currency corridors.

Rialto demonstrates that AMMs can provide continuous, algorithmic liquidity. In cases where a direct currency pair (like Euro to Singapore Dollar) is expensive or unavailable, the system automatically routes the trade through a vehicle currency (a third, highly liquid currency) within a single, atomic transaction. This ensures that even “exotic” currency routes can be settled efficiently on a Payment-versus-Payment (PvP) basis.

To bridge the gap between the algorithmic nature of AMMs (which can have price “slippage”) and the retail need for guaranteed quotes, Rialto introduces a specialized Foreign Exchange Provider (FXP). The FXP issues a binding quote to the customer and absorbs any slippage risk in exchange for a fee. This hybrid model brings together the efficiency of DeFi with the accountability of regulated finance.

The Rialto architecture is not just fast; it is robust. The report details how the system handles real-world failures:

• Compliance Rejections: If a destination bank flags a payment for AML/sanctions issues, the transaction is rejected before any money moves, and earmarked funds at the source are instantly released.
• Disruption Events: If a technical failure occurs mid-transaction, the system uses atomic transfer logic to ensure an “all-or-nothing” result, preventing a situation where money is debited from the sender but never reaches the receiver.

While technically feasible, the economic success of the Rialto model depends on a delicate balance of fees. Currently, retail FX costs average around 2%, while service fees range from 2.3% to 6%.

Rialto could theoretically lower these costs by reducing the need for prefunded nostro/vostro accounts, which currently tie up massive amounts of bank capital. However, AMMs require their own form of “locked” liquidity in pools, which presents a new trade-off for financial institutions to manage.

Project Rialto is more than a technical trial; it is a proof of concept for incremental progress. It suggests that the future of global finance is not necessarily a single, global digital currency, but rather a world of interoperable systems—where the speed of the internet meets the safety of central bank money.

For emerging economies in Asia and beyond, the lessons of Rialto offer a realistic roadmap toward faster, cheaper, and more transparent remittances.