A country does not need to be sanctioned to become financially isolated.
There may be no government order, frozen central-bank reserve or dramatic announcement. The internet continues working. Local banks remain open. Businesses can log in to their accounts and see their balances.
But somewhere in New York, London, Frankfurt or Sydney, a global bank’s risk committee decides that maintaining a relationship with a smaller foreign bank is no longer worth the trouble.
The account is closed.
Then another international bank reaches the same conclusion. And perhaps another.
Soon, the local bank can still accept deposits and make domestic payments, but sending US dollars abroad becomes more expensive, slower or sometimes impossible. Importers struggle to pay suppliers. Exporters have difficulty receiving proceeds. Remittance companies lose their banking facilities. International charities cannot move funds. Even credit-card settlement and trade finance may become harder.
The country has not been formally excluded from global finance.
It has been de-risked.
The hidden infrastructure behind an international payment
Suppose a business in Samoa wants to pay a machinery supplier in the United States.
The Samoan business gives instructions to its local bank. But that bank does not necessarily have direct access to the US payment system. It may not maintain an account with the Federal Reserve, participate directly in Fedwire or possess an operational branch in New York.
Instead, it relies on a larger bank that does.
The Samoan bank holds a US-dollar account with an American, Australian or New Zealand bank. The larger institution processes dollar payments on its behalf.
This is correspondent banking.
The terminology initially sounds more complicated than the arrangement:
The bank providing international services is the correspondent bank.
The bank using those services is the respondent bank.
From the respondent bank’s perspective, the account it holds abroad is a nostro account meaning “our account with you.”
From the correspondent’s perspective, the same account is a vostro account “your account with us.”
The correspondent may provide more than a simple account. It can handle wire transfers, currency conversion, cash management, check clearing, trade finance, card settlement and access to international payment systems.
A bank may need separate relationships for different currencies. It might clear US dollars through one institution, euros through another and pounds through a third.
Sometimes a payment passes through several correspondents.
A bank in a small country sends an instruction to a regional bank. That regional bank uses a larger European institution, which then settles through its US branch. Each intermediary deducts fees, screens the payment and relies on information supplied by the institutions before it.
To the customer, it is one wire transfer.
To the financial system, it is a chain of balance-sheet entries and compliance decisions.
Why would a global bank close the relationship?
The simplest explanation is not that the smaller bank has necessarily done something wrong.
The relationship may simply offer an unattractive combination of low revenue, high compliance costs and potentially enormous downside.
Imagine a correspondent bank earns US$150,000 annually from a smaller foreign bank. To maintain the relationship, it must understand:
Who owns and controls the respondent bank;
Whether its regulator is credible;
How it verifies customers;
Whether it serves money-transfer companies;
How it monitors transactions;
Whether it permits nested relationships;
Which countries and industries appear in its payment flows;
How it handles sanctions alerts;
Whether politically exposed persons use its services; and
Whether the correspondent can obtain reliable information when something goes wrong.
The bank must also maintain analysts, transaction-monitoring systems, investigators, auditors, legal advisers and sanctions specialists.
If suspicious funds eventually pass through the account, the correspondent may face investigations, financial penalties, remediation costs and reputational damage far exceeding the revenue it earned.
The decision can therefore become brutally straightforward:
Why accept a theoretically enormous liability for a relatively small commercial return?
That calculation becomes even harsher when the respondent operates in a small economy with limited transaction volume.
A bank serving ten million profitable customers can distribute its compliance expenses across a large base. A bank serving a small island economy may require much of the same due diligence while generating a fraction of the revenue.
Correspondent banking is therefore vulnerable to an economic asymmetry: the smallest and least profitable relationships can require some of the most careful monitoring.
“De-risking” is not the same as managing risk
Banks are expected to apply a risk-based approach.
That means identifying the actual risks in a particular relationship and applying controls proportionate to them. A bank serving a well-regulated remittance company should not necessarily be treated in the same way as an institution unable to identify its customers.
De-risking is different.
The Financial Action Task Force uses the term for situations in which institutions terminate or restrict categories of relationships to avoid risk instead of assessing and managing it individually.
A correspondent may decide:
No money-services businesses;
No banks from a particular jurisdiction;
No transactions involving certain industries;
No customers connected with digital assets;
No respondent banks serving other financial institutions; or
No accounts generating insufficient revenue.
These rules are attractive because they are easy to administer. They can also exclude legitimate businesses along with dangerous ones.
The World Bank has warned that broad de-risking can undermine the objectives of financial-crime controls. When regulated providers lose banking access, transactions may move into informal, opaque or unlicensed channels where they become more difficult to monitor.
A bank reduces its own exposure. The financial system as a whole may become less transparent.
The numbers reveal a shrinking network
Correspondent banking has not vanished. The majority of countries still maintain significant international connections, and payment volumes have continued to grow.
But those payments are increasingly concentrated among fewer relationships.
BIS data showed that the number of active correspondent banking relationships declined by approximately 25% between 2011 and 2020. The volume and value of transactions increased even as the number of relationships fell.
Data through 2022 showed that the downward trend continued, with some of the sharpest losses occurring around 2020.
This does not mean international payments declined by 25%. It means more activity became concentrated through fewer banks and fewer routes.
Concentration can create efficiency. It can also create fragility.
If a local bank has five correspondents and loses one, it still has alternatives. If it has only one functioning US-dollar relationship, the correspondent possesses enormous practical influence over the local bank’s business.
The local institution may technically remain licensed and solvent. Yet a decision by one foreign compliance committee can determine whether it can continue offering international wires, serving remittance companies or financing imports.
That is not merely vendor concentration. It is a form of financial dependence.
Why small states are especially vulnerable
The problem is particularly serious for island economies in the Caribbean and Pacific.
These countries often have several characteristics that concern correspondent institutions:
Small transaction volumes;
High dependence on remittances;
Large numbers of cross-border transactions relative to population;
Tourism and cash-intensive industries;
Limited supervisory resources;
Offshore companies or international financial services;
Citizenship-by-investment programs;
Geographic exposure to larger high-risk regions; and
Local banks that cannot afford sophisticated compliance systems.
None of these characteristics automatically means that a jurisdiction facilitates financial crime.
But they can produce an unfavorable risk-and-revenue profile.
The Pacific provides a striking example. Several island countries depend heavily on Australia and New Zealand for employment, trade, remittances and banking connections. Workers abroad send money home to support family consumption, education and housing.
If remittance operators lose their bank accounts, the problem does not remain inside the financial sector. Families receive less money or pay more to obtain it.
The IMF reported in 2025 that Samoa had experienced a decade of de-risking, resulting in withdrawn correspondent relationships, greater fragility and increased concentration—particularly for money-transfer operators.
The threat became serious enough for the World Bank and Pacific governments to establish a regional initiative specifically intended to preserve access to correspondent banking. By 2025, the World Bank described it as a US$77 million project covering eight Pacific countries, with additional participation under development.
That is remarkable.
Development finance is ordinarily associated with roads, schools, electricity or public health. Here, public resources are being used to preserve something less visible but equally foundational: the ability of a country’s banks to reach the rest of the world.
What happens when a relationship is lost?
The immediate effect depends on which services the correspondent provided.
If a bank loses one of several dollar correspondents, it may reroute payments through another. Customers may notice only increased fees or longer processing times.
If the lost relationship was its only route, the consequences become more severe.
Payments take longer
A direct payment may be replaced with a longer chain involving two or three intermediaries. Each institution performs sanctions and AML screening. Each may request additional information.
Costs rise
Every intermediary charges a fee. The respondent bank also incurs expenses finding a replacement, upgrading systems and meeting more demanding compliance requirements.
Trade finance becomes harder
A local bank’s letter of credit is valuable only if foreign suppliers and confirming banks trust the institution behind it. Weak correspondent access can make the instrument less acceptable.
Remittance companies lose accounts
Banks may stop serving money-transfer operators to protect their own correspondent relationships. The local bank fears that the foreign correspondent will view MTO activity as insufficiently transparent.
International businesses leave
An exporter that cannot receive foreign payments reliably—or an importer that cannot pay suppliers—may relocate banking and corporate activity elsewhere.
Financial inclusion deteriorates
Formal channels become slower or more expensive. Customers may turn to cash couriers, informal value-transfer systems, personal accounts or unregulated cryptocurrency intermediaries.
The system becomes less safe precisely because regulated access has narrowed.
The problem of nesting
One of the most sensitive issues in correspondent banking is nesting.
Suppose a major US bank provides a dollar account to Regional Bank A. Regional Bank A then uses that account to process transactions for Banks B, C and D.
The US bank may have approved Bank A, but it may have little visibility into the customers and controls of B, C and D.
The relationship has acquired additional layers.
Nesting is not automatically improper. Regional correspondent arrangements are often necessary because not every small bank can maintain direct relationships with major global institutions.
The risk arises when nesting is undisclosed, poorly controlled or so extensive that the ultimate correspondent cannot understand who is using its infrastructure.
The same issue appears when a payment company places multiple financial institutions, agents or downstream fintechs under one master account.
A proposal may describe this as an elegant aggregation model. The bank may see undisclosed financial institutions using its payment rails.
For a consultant, this is why drawing the funds flow is not enough. One must also draw the customer and institutional hierarchy:
Who contracts with the bank?
Who originates the underlying transactions?
Which party performs KYC?
Which institutions have customers of their own?
Can the correspondent identify the ultimate originator and beneficiary?
Does it know that downstream institutions exist?
An undisclosed nested relationship can threaten the entire banking arrangement.
Can technology solve de-risking?
Technology can reduce some of the cost.
Standardized KYC utilities can help correspondents obtain reliable information about respondent banks. Digital identity can improve customer verification. Better payment messages can carry more complete originator and beneficiary information. APIs can support ongoing monitoring instead of periodic document requests.
Artificial intelligence may help identify unusual transaction patterns or summarize complex investigations.
But technology cannot remove the commercial judgment at the center of the relationship.
A correspondent still decides whether the expected revenue justifies the compliance, reputational and regulatory exposure. Better data makes that decision more informed; it does not guarantee a favorable answer.
Nor can a smaller bank solve the problem merely by purchasing expensive transaction-monitoring software.
The correspondent will examine whether the system is properly configured, whether alerts are investigated, whether management acts on findings and whether the respondent’s regulator tests the controls.
A sophisticated dashboard attached to a weak compliance culture remains a weak control.
Can stablecoins bypass correspondent banking?
Stablecoins can provide an alternative channel for transferring dollar-linked value.
A business may purchase a stablecoin locally, transfer it to a supplier or intermediary and redeem it elsewhere. This can operate outside the traditional chain of correspondent accounts and banking hours.
That may reduce friction, especially in underserved corridors.
But it does not eliminate the need for financial access.
Someone must convert local currency into the stablecoin. Someone must provide liquidity. The recipient may need to redeem the token into a bank account. The issuer’s reserves are normally held within the conventional financial system. Exchanges and market makers still need banking relationships.
Stablecoins can change the location of the banking dependency without removing it.
They may also introduce new questions:
Is the issuer regulated?
Who can redeem directly?
Is the token liquid in the destination market?
Which sanctions controls apply?
Can wallets be frozen?
Is the conversion spread stable during market stress?
Can a local institution legally offer the service?
For a country facing correspondent withdrawal, stablecoins may create a useful supplementary rail. Treating them as a complete substitute for bank connectivity would be dangerous.
Trade finance, tax payments, government transactions, card settlement and large corporate flows still depend heavily on conventional institutions.
What actually improves the situation?
There is no single repair.
The most effective responses combine several elements.
Better local supervision
A correspondent is more comfortable when it trusts the respondent bank’s regulator. Strong examinations, enforcement and sector-specific guidance can reduce the need for the foreign bank to independently reconstruct the entire risk environment.
Credible bank-level controls
Respondent banks must demonstrate, not merely claim that they understand their customers, agents, MTOs and downstream institutions.
Better economics
A relationship that generates adequate revenue is easier to preserve. Smaller banks may aggregate services regionally, share utilities or negotiate collective solutions that create sufficient scale.
Safe payment corridors
A corridor can be assessed as a defined system: who sends the money, why it is sent, how workers are identified, which providers participate and what the actual financial-crime risks are.
The IMF’s work on Samoa explores this approach. Instead of treating every remittance transaction as an abstract high-risk transfer, a safe-corridor assessment examines the specific labor, remittance and regulatory structure connecting Samoa with Australia and New Zealand.
Multiple routes
No bank, MSB or national financial system should depend on one correspondent if credible alternatives can be maintained.
Redundancy costs money. Losing the only international settlement route costs much more.
The larger lesson
The modern financial system is not a single network to which every licensed institution has equal access.
It is a hierarchy.
At the center are central banks, reserve currencies, major clearing systems and large international banks. Around them sit regional correspondents, domestic banks, payment institutions, money-transfer companies, agents, merchants and customers.
Every participant depends on someone nearer the center.
A local license gives an institution permission to operate. It does not compel a foreign bank to open an account, provide dollars or process its customers’ payments.
That distinction is one of the most important in financial services:
Regulatory access is not the same as operational access.
A company can obtain an MSB registration, payment-institution license or crypto authorization and still be commercially unable to operate because no credible institution will provide settlement, safeguarding or correspondent services.
De-risking makes that distinction visible at the level of an entire economy.
The country remains sovereign. Its banks remain licensed. Its money still circulates domestically.
Yet its connection to international money can depend on a small number of private institutions located thousands of miles away.
No formal sanction is required. Sometimes the quietest form of financial isolation begins with an account-closure letter.
Practical takeaway: When assessing a cross-border payment or licensing opportunity, do not stop at the local license or domestic bank account. Map the complete chain to final settlement: the correspondent banks, currencies, nostro accounts, downstream financial institutions, prohibited customer categories, alternative routes and termination rights. Ask what happens if the principal correspondent exits with 30 or 60 days’ notice. If the business cannot answer that question, its international payment capability is rented, not secured.