More money is coming into the South Pacific than ever. Fewer banks are willing to move it.

By Mark Townley, Managing Director, Ideas2Outcomes & Artipi (Banking & Payments Strategic Advisory) - July 2026

Three numbers give key insights into the South Pacific's decade ahead.

• (1) US$20 to 25 billion, the combined capital investment Papua New Guinea's two flagship resource projects could represent over the next ten years.

• (2) Forty-three per cent, (43%) remittances as a share of Tonga's GDP, the highest dependency in the world, with Samoa fourth at 28 per cent.

• (3) Sixty per cent, (60%) the fall in active correspondent banking relationships serving Pacific institutions since 2011, against a global average closer to 30 per cent; levels the Pacific Islands Forum describes as unsustainably low and a risk to the whole regional financial system.

  
  

Capital into one set of economies, income into another, and a small handful of correspondent banking relationships standing between all of it and the rest of the world. That chain is thinning for everyone: for Papua New Guinea's resource sector as much as for a Tongan household, and for Fiji, Samoa, Solomon Islands and Vanuatu equally.

More money is crossing that chain than at any point in the region's modern history, and there are fewer regulated ways to move it than at any point in a generation. That is not a Pacific inconvenience. It is a live question about whether the traditional global financial system can still serve small, remote, high-cost markets at all.

It is also, this year, a question with a room waiting for it. Sibos is the global financial industry's annual meeting, organised by Swift, bringing together the banks, market infrastructures, regulators, fintechs and technology providers that between them decide how money actually moves; the 2025 edition drew more than 12,500 participants.

The 2026 Sibos (Miami, USA) theme is digital finance for AI-driven economies, spanning cross-border payments, trade and settlement, digital assets and tokenisation, regulation and resilience. Every one of those items reads differently from the South Pacific.

Why the money is coming

Papua New Guinea remains the anchor. Its export flows run principally to Japan, Australia and China, and it draws close to half its imports from Australia alone. Resource development, both LNG and minerals, dominates the investment pipeline and brings the full apparatus of large project finance: multi-currency treasury, offshore procurement, contractor payment chains running to dozens of jurisdictions, and payroll for tens of thousands of workers.

Layered on top is infrastructure capital, much of it Chinese-funded and directed at energy, ports and telecommunications, alongside increased Australian and New Zealand development, security and climate finance, and tourism and fisheries revenue across the smaller island states.

Beneath the resource headline sits a second export economy that matters more than its dollar value suggests. Cocoa, coffee and palm oil are the principal cash crops; coffee alone is around one per cent of export value, but agriculture provides a subsistence livelihood for roughly 85 per cent of the population, and cash crops are how a large share of rural households touch the formal economy at all. Large scale property development across PNG becoming increasingly evident also.

The resource projects move very large sums through a handful of sophisticated counterparties; the cash crops move small sums through hundreds of thousands of smallholders. Both travel the same thinning chain, and a fixed compliance cost that is immaterial on an LNG payment is prohibitive on a container of coffee.

There is also a flow that barely appears in trade statistics. Mainline churches operate around 60 per cent of Papua New Guinea's schools and health services, funded through government partnership programmes, Australian faith-based partners and direct support from overseas congregations; comparable arrangements run through Solomon Islands and Vanuatu.

These are small, frequent, cross-border transfers to rural destinations, which makes them among the first to be repriced or refused when a chain thins.

Then there is the flow that matters most to households. Beyond the Tongan and Samoan dependencies already noted, Fiji's personal remittances passed one billion Fijian dollars in the year to September 2025. Papua New Guinea is the outlier, its cross-border money movement being corporate, government and resource-sector foreign exchange.

One region, two entirely different problems: institutional-grade trade and foreign exchange capability at one end, low-cost compliant retail corridors at the other.

Why the pipes are narrowing

The withdrawal of correspondent banking from the Pacific is usually explained as a risk story. It is more accurately an economics story with a risk trigger. A correspondent relationship is only worth holding if corridor revenue exceeds the cost of the compliance, screening, investigation and supervisory attention required to hold it, and individual Pacific economies do not generate the volume to clear that bar.

When financial crime scrutiny intensified globally and enforcement penalties began to bite, the rational response for a global bank was not to price the risk; it was to exit the corridor. Small jurisdictions went first.

The consequences are measurable. The average cost of sending a remittance into the Pacific sits around 10 per cent, roughly triple the United Nations target of 3 per cent. In the most affected markets, banks are down to a single remaining correspondent in a key currency, a single point of failure for an entire national economy.

Papua New Guinea's addition to the FATF grey list in February 2026 has prompted enhanced due diligence on PNG-linked transactions, adding cost precisely where cost was already the binding constraint.

Public capital has recognised the market failure. The World Bank's regional correspondent banking programme, approved in 2024 and in implementation through 2030, subsidises continued access for institutions at risk of losing their last link, while the Pacific De-Risking Group works on the regulatory environment. That is welcome and necessary. It is also, by design, a bridge rather than a destination; something has to change in the economics before the subsidy runs out.

The Sibos theme, read from the thin end of the network

Read that agenda in London, New York or Singapore and it is largely an efficiency agenda; margin, latency, straight-through rates, headcount. Read it from a Pacific bank and it is an access agenda, and considerably more urgent for it.

AI is not a productivity story; it is a corridor economics story. If financial crime compliance is the cost line that drove correspondent banks away, then anything that materially lowers the unit cost of screening, investigation and false-positive resolution changes the calculation on which those exits were based. That is the highest-leverage application of AI available to a small-market bank, and it is defensive before it is anything else.

ISO 20022 is not a messaging upgrade; it is the raw material. Papua New Guinea's central bank completed its migration in October 2025, and comparable programmes are underway across the region. Structured data is what makes AI-based screening viable in the first place; treat the migration as a compliance exercise and the opportunity is filed away with it.

Tokenisation is not a crypto conversation; it is a routing conversation. For a region whose problem is a shrinking intermediary chain, the interesting property of tokenised settlement is that value and its compliance data move together and settle with fewer hops. The strategic risk is not volatility; it is exclusion. If a tokenised settlement layer forms among the major economies and the Pacific is not interoperable with it, the region will have been de-risked twice.

The region is already testing that proposition, and the results are instructive. Vanuatu passed a Virtual Asset Service Provider Act in March 2025 and has since tabled a stablecoin bill, building a licensing regime aligned to FATF standards. Fiji went the other way and prohibited virtual asset service providers outright, explicitly to buy time while supervisory capacity is built.

Papua New Guinea's central bank has run a central bank digital currency proof of concept while warning citizens away from crypto entirely, and the Solomon Islands central bank piloted a digital representation of its own currency from 2023. Tonga came closest to the radical option, with a parliamentary push to make bitcoin legal tender expressly to cut remittance costs; it never reached legislation.

Read that as four governments disagreeing and you miss the point. Every one has concluded the existing rails are inadequate, and each has hit the same wall: the technology is possible the easier part. What determines whether an alternative rail works is legal definition, licensing capacity, supervisory skill, custody, connectivity and public trust. Which is why the quieter work matters more than the headlines.

What has actually advanced in Tonga is not a bitcoin bill but a Payment Systems and Services Bill, drafted with the National Reserve Bank of Tonga under the PACER Plus trade agreement with World Bank technical input, to license payment providers, regulate electronic money, and bring informal remittance channels inside the supervisory perimeter for the first time.

In Solomon Islands, that same central bank has entered the second phase of a unified national QR standard with Australian government support: technical standards and business rules co-developed with banks and mobile money operators, then pilot testing, staged rollout and public education. No new asset class, no new currency. Standards, governance, integration and adoption, in that order.

The four-party problem

The region is not without scale. Bank of South Pacific (BSP Financial Group) operates across seven Pacific countries and accounts for roughly half of South Pacific banking deposits and lending; Westpac and ANZ retain positions in Papua New Guinea and Fiji, BRED operates in Fiji and Solomon Islands, and a tier of national and development banks sits beneath them across the region.

Concentration of that kind is part of the answer to the sub-scale problem, since a correspondent that cannot justify a dozen small relationships may well justify one large and well-supervised one. It is not the whole answer. No bank, whatever its share, legislates a payments act, builds a national switch, sets a virtual asset perimeter or trains a country's financial crime analysts. None of this is delivered by one actor. It requires all four to move forward together, and each has a different clock speed.

Regulators set the perimeter. Pacific central banks move at very different speeds on instant payments, virtual assets, custody, digital identity and data, and a corridor operates at the pace of its slowest supervisor. The regional prize is shared foundations, particularly on electronic know-your-customer and the virtual asset perimeter, negotiated once rather than a dozen times.

Infrastructure determines what is possible. National switches, real-time payments rails, interoperable QR, resilient connectivity from ocean data cables to low-earth-orbit satellite, and identity systems: decade-length builds with political dependencies, and no bank controls them.

Bank systems carry the accumulated debt. Across the region's major and minor banks alike, the honest position is batch cores, fragmented channel stacks, partial API coverage and reconciliation performed by people. You cannot run a 24/7 real-time proposition on a platform that closes for end-of-day, and no amount of AI at the front compensates for that.

People and process is the constraint nobody funds. The binding scarcity is not licences or hardware; it is payments operations specialists, financial crime analysts and data engineers resident in the region. Capability that flies in and flies out does not compound.

The uncomfortable implication is that any single party moving alone achieves very little. A bank that modernises into a regulatory vacuum stalls. A regulator that mandates ahead of the systems creates unfunded compliance. Infrastructure built without the skills to operate it becomes an expensive monument. The institutions making real progress treat this as one programme with four owners, not four programmes with one.

For Sibos 2026; an engaging, relevant and topical discussion

None of what is needed by 2030 is exotic; most of it is standard practice somewhere else already.

Instant payment systems in each market that connect to each other, so Pacific currencies can be exchanged directly rather than through Australian or United States dollars that clear offshore. ISO 20022 end to end, with the data actually used. Shared eKYC and financial crime utilities at regional scale, because the compliance economics only work when the cost is spread. Regional gateway banks, where material volume sits behind a well-managed and well-supervised correspondent relationship rather than being split twelve ways, unable to meet risk or profitability thresholds.

Cross border financial flows across South Pacific are real, growing and increasingly strategic. So the question for anyone walking the floor in Sibos is not whether the South Pacific is bankable. It is whether the economics of serving it can be re-engineered, using precisely the tools this year's theme is about, so that willingness returns and opportunity to prosperity can thrive.

Sources: Pacific Islands Forum, 2025 Annual Report on the Pacific Strengthening Correspondent Banking Relationships Project; World Bank; Swift and the Sibos 2026 conference programme; Bank of Papua New Guinea; Reserve Bank of Fiji; National Reserve Bank of Tonga; Central Bank of Solomon Islands; Vanuatu Financial Services Commission; PACER Plus Implementation Unit; BSP Financial Group; Australian Department of Foreign Affairs and Trade; United States Department of State; UNDP Pacific; ANZ Research; FATF.