BlockRiver Blog | Stablecoin Rails Challenge the Nostro Trap: What Treasury Teams Need to Know About Multi-Currency Settlement

Corporate treasury has always operated under a structural constraint that rarely gets named directly: the nostro trap. To move money across borders, banks must pre-fund accounts in foreign currencies at correspondent institutions. This creates a cascade of idle capital, balances sitting in yen accounts overnight while Tokyo sleeps, euros repositioned before London opens, dollars parked while waiting for New York to come online. The precision required resembles air traffic control, except the planes are billions of dollars moving through correspondent channels that never close, and the cost of underestimating is failed payments while the cost of overestimating is capital earning nothing.

For multinationals with subsidiaries across jurisdictions, the same dynamic plays out at the corporate level. Foreign exchange restrictions, capital controls and limited cross-border pooling often result in trapped cash. Companies in North America can improve their cash flows through cash concentration, minimising idle balances and reducing borrowing costs. In much of Latin America, however, pooling is constrained by capital controls, taxes or both, necessitating different ways to manage their money. The result is liquidity that cannot be reallocated in real time, a drag on return on deployed capital that treasury teams absorb as a cost of doing business.

This is the context into which new payment rails are emerging. The correspondent banking model is not disappearing, but it is contracting. A 2019 BIS study found a continued decline in the number of correspondent banking relationships; between 2012-2019, active relationships in the global network declined by about 20%, though reductions varied across regions. The sharp decline in correspondent banking over the past decade has raised concerns that the associated disruptions in cross-border payments could hamper international trade and economic growth. The Pacific region has been hit particularly hard: since 2011, the Pacific has seen a 60 per cent drop in correspondent banking relationships, double the world average.

De-risking, driven by compliance costs and low transaction volumes in high-risk corridors, has concentrated correspondent relationships among fewer institutions. Concerns on the part of large international banks about regulatory compliance with AML and customer due diligence requirements have led some banks to shed their correspondent banking relationships with smaller banks, often in emerging markets. According to the Bank for International Settlements, rising costs and uncertainty about how far due diligence should go to avoid regulatory sanction are cited by banks as among the main reasons for this de-risking. For corporate treasury, this means fewer options, longer chains, and more capital required to pre-fund the accounts that remain.

The legacy system is not standing still. The global financial community has reached a major milestone: ISO 20022 is now the standard language for cross-border payments worldwide. ISO 20022 elevates customer experience in today's fiat currency system in support of the G20 goals for international payments, enabling faster, more efficient and data-driven payments. On 22 November 2025, the coexistence period between MT and ISO 20022 messages for cross-border payments ended, marking the start of a new era of richer, more structured payments data. SWIFT GPI has delivered real improvements in speed and transparency: around 60% of GPI payments reach the beneficiary bank within 30 minutes, and nearly all are credited within 24 hours.

But speed is not the same as capital efficiency. The fundamental architecture still requires pre-positioning funds across currencies and time zones. Estimate too conservatively, and customer payments fail. Estimate too aggressively, and millions in capital sit idle earning minimal returns while shareholders demand better performance. This is where the alternative rails present a different proposition entirely.

The BIS Project Agorá prototype demonstrates that tokenised commercial bank deposits can be successfully combined with the trust and safety of tokenised central bank reserves on a shared platform. The project has shown the possibility of completing atomic settlement of wholesale cross-border transactions. The prototype enables atomic, multi-currency settlement of wholesale cross-border payments, which could occur on an around-the-clock basis if implemented. This promises to reduce reconciliation burdens, manual intervention and other operational frictions, key sources of delay, cost and payment failure in today's cross-border system.

The private sector is not waiting for central bank infrastructure to mature. JPMorgan Chase, Citigroup, Bank of America, Wells Fargo, and other major US commercial banks plan to launch a tokenized deposit network in the first half of 2027. The initiative represents Wall Street's most coordinated competitive response to stablecoins to date. JPMorgan's JPM Coin launched on Coinbase's Base network in late 2025 for institutional clients and has since expanded toward the Canton Network. JPMorgan positions the product as a direct bank deposit claim with on-chain programmability. Citi announced integration of its Token Services platform with its 24/7 USD Clearing solution, which supports 250+ banks across 40+ markets and enables multibank cross-border instant payments.

On July 18, 2025, President Trump signed the GENIUS Act into law. This landmark bill is the first major crypto legislation in the United States. The GENIUS Act creates licensing and regulatory requirements for domestic payment stablecoin issuers and standards for participation in the US payment stablecoin market by foreign stablecoin issuers. The regulatory uncertainty that kept institutional treasury teams on the sidelines is resolving, not perfectly, but meaningfully.

Banks are using stablecoins for always-on treasury operations. Instead of pre-funding accounts globally, institutions can move USDC instantly as needed, optimizing liquidity and reducing trapped capital. For large-value transactions along well-liquid corridors, the economics are compelling. The American Banker analysis is direct: stablecoins are carving out a niche in large-scale institutional trading and wholesale liquidity, while for small-value transactions, the costs of on- and off-ramps can diminish the benefit.

This is not a wholesale replacement of correspondent banking. The G20 has made faster, cheaper, more transparent and more inclusive cross-border payments a priority and has set targets to achieve this by end-2027. The Roadmap, endorsed by the G20 Leaders, lays out a comprehensive set of actions across 19 building blocks. But jurisdictions' implementation of the Roadmap's policy recommendations has been uneven, and improvements in end user outcomes remain limited. It is unlikely that the G20's quantitative targets will be met by end-2027.

The implication for corporate treasury is operational, not speculative. The question is whether your current liquidity allocation model, the buffers held across currencies, the pre-funding required for each correspondent relationship, the capital trapped in accounts you cannot efficiently rebalance, is the only architecture available, or whether it is simply the one you inherited. The rails are multiplying. The capital efficiency gains are real for specific use cases. Whether treasury owns this evaluation or waits for policy to approve it determines whether the decision is made proactively or reactively.

The nostro trap is not a law of physics. It is an artifact of infrastructure constraints that are now being challenged from multiple directions, by central bank experiments, by major bank consortiums, and by regulated stablecoin issuers. Treasury teams that understand this are not making a crypto bet. They are stress-testing a capital allocation model against infrastructure that may no longer require it.