Trade Finance and Assurance: Why Faster Rails Have Not Closed the Gap

The building society holding my mortgage in Saint Vincent lost its correspondent-banking relationship. With it went the route for sending money to my account from abroad, even though nothing about the account itself had changed. An article I co-authored with Brian Canup for Trade Finance Global (James and Canup, 2023) argued that paper-heavy, intermediary-heavy trade finance left smaller firms behind. It cited an estimated $5 trillion shortfall in financing for SMEs and suggested that decentralised finance could open access to them. Three years of evidence now allow that argument to be tested, and the question is whether rebuilt payment infrastructure can close the trade-finance gap for smaller firms.

The trade-finance gap is the difference between the trade finance firms seek and the finance banks provide, currently estimated at $2.5 trillion (Asian Development Bank, 2026). De-risking occurs when banks restrict or end whole categories of relationship instead of managing each counterparty individually (Financial Conduct Authority, 2023). The problem falls between those two definitions: whether faster, digital systems address the reasons finance is withheld, or only change how it moves once granted.

Right about the friction

Gaps in trade-finance provision fall most heavily on small and medium-sized enterprises (World Trade Organization, 2016). Legal reform has since addressed part of the documentary burden. The UK's Electronic Trade Documents Act 2023 gave qualifying electronic trade documents the same legal effect as paper, building on the Model Law on Electronic Transferable Records (UNCITRAL, 2017). This resolved the long-standing difficulty that an electronic record could not be "possessed" in the way a paper bill of lading could. Yet the global gap has stayed at $2.5 trillion since 2023 (Asian Development Bank, 2026). The loss of my own mortgage provider's correspondent relationship fits that pattern. The likely cause was the correspondent's confidence in the respondent bank, and faster documents would not necessarily have restored it. Reducing documentary friction is valuable, but it may not on its own widen access for the smaller firms the gap affects most.

The problem beneath the payment

Swift's blockchain-based ledger went live for initial use in July 2026, with seventeen banks across six continents preparing to pilot tokenised cross-border payments on it (Swift, 2026). J.P. Morgan's Kinexys business has also taken deposit tokens into live institutional use through JPM Coin. Both are the providers' own announcements, however, and they describe new capability, not evidence of wider access. These developments change how value moves between institutions that already deal with one another. The literature on correspondent banking points somewhere else. The withdrawal of correspondent relationships has been presented as a case for policy action (International Monetary Fund, 2016), tracked through correspondent banking data (Financial Stability Board, 2017) and made the subject of a dedicated de-risking strategy (US Treasury, 2023). None of these sources locates the problem in settlement speed. For a building society in Saint Vincent, a faster rail would still have needed a correspondent willing to accept the relationship. Settlement innovation may therefore improve execution while leaving the decision to extend access where it was.

De-risking and de-banking may amount to an over-compliance dilemma (de Koker and Casanovas, 2024). A correspondent bank weighs financial-crime exposure, sanctions risk, the quality of the respondent's controls, the reliability of its supervision and the commercial return on offer. The cost of reaching a defensible view on those questions is likely to be largely fixed per relationship, whereas revenue tends to scale with the relationship's size. In my view, regulation sets the price of confidence, and each bank's own assessment then decides whether a relationship still covers it. The Caribbean recurs as a case throughout this literature, so often that the Center for Strategic and International Studies has asked whether de-risking there has become a "new normal". Saint Vincent fits that pattern, and small respondent banks may lose access where the cost of confidence exceeds the value of the relationship.

The work that did not travel

An incoming custodian in naval logistics accepts responsibility for a stores account only after it has been audited, because signing for another person's record needs evidence behind it. Correspondent banking rests on a similar logic at a larger scale. Good practice exists on anti-money-laundering and counter-terrorist-financing risk management for emerging-market banks (International Finance Corporation, 2019), and their correspondents largely rely on those controls without observing them directly. Tokenisation and shared ledgers, by contrast, are often judged by what they remove, whether a reconciliation step, a paper original or an intermediary. Intermediaries, however, also perform assurance work. They identify parties, examine documents, monitor transactions and keep the evidence behind each decision. Removing an intermediary relocates the assurance work it performed, and the new owner of that risk is rarely named when the move happens.

Unchanged since 2023

Survey responses from more than 110 trade-finance providers between 2023 and 2025 put the gap at $2.5 trillion, unchanged in absolute terms since 2023 and down only from 10.6% to roughly 10% of global trade (Asian Development Bank, 2026). The survey's January 2026 release reported rejection rates of 41% for SMEs and 40% for large and mid-cap corporates, the first time the two have sat so close together. Capital constraints, pricing, country risk and transaction size are also likely to influence those figures, so better verification alone could not plausibly close a gap of that scale. Large counterparties, with established banking histories and professional compliance functions, tend to absorb the continuing cost of being understood more easily than smaller ones, whose ownership and trading behaviour may need reassessing long after onboarding. The evidence supports a narrower claim: some viable trade goes unfinanced because establishing confidence in the counterparty costs more than the relationship is worth.

The whole argument in miniature

On a naval stores account, what a supplier says about itself arrives on paper, and what it delivered is tested at the goods-in check, where the two do not always agree. A risk programme built only on the first carries a gap it cannot see, and trade finance faces the same distinction between assertion and evidence. The model law (UNCITRAL, 2017) and the Electronic Trade Documents Act 2023 show where technology earns its place. Electronic records can strengthen the evidence once their integrity and transfer can be demonstrated, and shared ledgers can reduce discrepancies where participants rely on a common record. Regulatory concern about blanket de-risking (Financial Conduct Authority, 2023) points to the harder requirement, which is assessment of each counterparty on its merits. The most useful tools lower the cost of producing, validating and reusing evidence about a counterparty, including public-source verification. Tools that only speed up payments between parties who already trust each other contribute least to closing the gap.

Worth less than it looks

Entries on a naval stores account trace back to a named signatory, and a record is only as defensible as the person who signed for it. The US de-risking strategy and UK regulatory research both focus on how institutions decide on individual relationships (US Treasury, 2023; Financial Conduct Authority, 2023). Automation can gather evidence, flag inconsistencies and calculate a risk indicator. The institution would still need to know who may accept a relationship, on what conditions and when that acceptance should be revisited. Where one institution cannot rely on another's verification, digitalisation may simply repeat the same checks faster and leave the duplication in place. An automated assessment without clear authority behind it may become more consistent without becoming more defensible, and a dashboard that delivers consistency alone is worth less than it looks.

The cost of knowing

The two problems behave differently. Settlement delay tends to yield to better infrastructure. Knowing a counterparty yields only to cheaper, more reusable evidence about it, and much of that evidence is already held inside the institution, in self-attestations accepted at face value and in assessments that one team completes and another repeats from scratch. Independent, public-source verification could add a further layer at modest cost, provided someone with authority is prepared to accept it. Because I took over stores accounts only once they had been audited, my own view is that the first place to look is the institution's own files. The record of why a relationship was declined, or never offered, is likely to say more about the gap than any new settlement architecture could.

Confidence before settlement

The answer, then, is a partial one. Documentary reform and tokenised settlement have begun to reduce paper, reconciliation and delay, and the 2023 article's diagnosis of paper-based processes and costly cross-border payments remains sound. Even so, rebuilt payment infrastructure is unlikely on its own to close the trade-finance gap for smaller firms. The harder problem looks to be the cost of establishing and maintaining confidence in a counterparty, together with the question of who remains accountable once another party or system has performed part of that assessment. The revision to the 2023 argument is therefore one of sequence: before proposing a new rail, the first question is why the institution on the existing rail withdrew. The next line of research this raises is how public, authoritative and independently verifiable information might lower the cost of that confidence without creating another opaque score. Legitimate trade may only become easier to finance once the evidence supporting trust can travel as readily as the payment.

The views expressed here are my own and are based on general observations about trade finance, third-party risk and financial inclusion. They draw only on published sources and do not represent the position of the Royal Navy, the Ministry of Defence or any organisation with which I work. Nothing published here reproduces any organisation's internal policy, procedure or assurance output.