RESEARCH DOSSIER
International Correspondent Banking and Offshore Finance
Abstract. International correspondent banking provides the account and payment services through which banks maintain international settlement capacity. This monograph studies its economic role within the global offshore financial system, examining account relationships, double-entry settlement, transaction costs, foreign-currency liquidity, financial geography, network structure, the decline in active correspondents, and distinctions among message data, financial stocks, flows and effective access. It evaluates de-risking and financial integrity, trade and household outcomes, small-economy vulnerability, sanctions and network fragmentation, settlement and operational exposures, and international payment reform. The analysis synthesizes dated primary institutional materials and original scholarly studies, using explicitly illustrative accounting and economic examples rather than asserting new empirical estimates. It concludes that offshore financial capacity depends on both the location and ownership of claims and the practical availability of safe, lawful and resilient cross-border settlement. The research contributes an integrated economic interpretation while preserving source-specific measurement limits.
Keywords: international correspondent banking; offshore finance; nostro and vostro accounts; cross-border payments; foreign-currency liquidity; financial networks; de-risking; financial inclusion; settlement risk; offshore financial centers; banking statistics; international trade
Introduction
International finance requires the capacity to complete obligations denominated in currencies and payable in jurisdictions different from those in which the underlying economic activity occurs. A bank in a small financial center may accept deposits, administer wealth, distribute investment income and finance transactions for nonresident clients while depending on institutions elsewhere for the practical settlement of those activities. The architecture that connects its liabilities and payment instructions to a wider set of settlement assets is the subject of this study.
International correspondent banking is the provision of banking services by one bank to another across jurisdictional boundaries, especially account holding, payment execution, clearing access and related services. Offshore finance is the organization of financial claims, intermediation and associated functions across jurisdictional or currency-area boundaries. Their intersection concerns how jurisdictionally distributed balance sheets acquire usable international payment capacity. It concerns both specialist financial centers and ordinary economies that participate in global trade and finance.
The central economic question is how the topology, pricing, contractual availability and liquidity requirements of correspondent networks influence the location, scale, resilience and social incidence of offshore financial activity. A jurisdiction can offer competitive legal and administrative services yet have limited access to particular foreign-currency settlements. A large financial center can offer excellent access and charge financial institutions elsewhere for the services required to obtain it. The allocation of payment capacity is therefore a material economic dimension of the offshore system.
The study distinguishes payment processing from foreign-currency funding, banking from messaging, network access from transaction volume, and statistical location from ultimate economic exposure. It examines historical development, balance-sheet mechanics, institutional incentives, empirical network contraction, trade and remittances, small-economy vulnerability, sanctions, and the prospects for technological and policy reform. Legal obligations are considered insofar as they influence economic behavior; jurisdiction-specific legal interpretation belongs to the corresponding legal research.
This is a qualitative and accounting-based research synthesis with dated official evidence and published empirical findings available through November 18, 2026. Numerical examples are explicitly illustrative. No new sample, regression, estimated global network, causal coefficient or new named theory is claimed. Full analytical articles, official definitions, official report summaries and author abstracts are differentiated by access depth. Conclusions based on a summary retain that summary's scope.
Concepts and the Units of Analysis
A correspondent relationship is an arrangement through which a service-providing bank supplies services to a respondent bank. The respondent may be domestically important, internationally active or specialized in a financial-center business. The correspondent can also be a respondent of another institution for a different currency or market. The same bank may therefore supply and purchase correspondent services simultaneously.
The CPMI (2016) correspondent banking report treats the arrangement as a set of banking services and identifies improvements in data, identification and communications. The relevant unit is typically the bank-to-bank relationship, qualified by currency, service, contractual limits and operating location. An account relationship can exist without being active during a particular observation window. A payment message can indicate activity without revealing the complete chain or contractual conditions.
A nostro account is a bank's deposit account at another bank, viewed from the depositing bank's perspective. The account is a claim on the account provider. A vostro account is the same type of balance viewed as a liability by the account-maintaining bank to the other bank. These terms identify accounting perspectives rather than different instruments. Currency denomination, governing contract, location of the booking office and ownership of the banking group are separate properties.
Clearing involves the exchange and reconciliation of instructions and, where relevant, calculation of obligations before settlement. Settlement discharges obligations through transfers of an agreed settlement asset in accordance with applicable rules. Finality concerns the legal and operational point at which settlement becomes final. A financial message records or instructs an operation; its transmission alone establishes neither the movement of settlement assets nor the availability of money to the beneficiary.
Offshore banking, offshore financial centers and international correspondent banking overlap without being coextensive. A correspondent supporting a manufacturing exporter in a conventional domestic economy belongs to the payment network even if neither customer is located in a specialist financial center. Conversely, an offshore investment vehicle may own marketable securities without acting as a correspondent bank. The connection is functional: internationally distributed financial activity demands usable means to meet monetary obligations.
An offshore relationship can be classified according to both currency area and institutional residence. Two non-US institutions may transfer dollar claims to one another. The resulting exposure belongs to an international dollar system, even where the underlying obligations and economic participants lie outside the United States. The settlement path and the final recipient of the underlying economic benefit remain additional facts to establish.
Historical Formation of International Banking Networks
Correspondent banking emerged as a solution to geographic dispersion and the difficulty of maintaining a banking office in every commercial market. Bilateral agency and account relationships permitted local institutions to extend the effective geographic reach of their services while retaining independent balance sheets. Commodity trade, bills of exchange, merchant banking and access to major centers provided economic reasons to maintain foreign relationships. The historical rise of the network follows the evolution of trade, credit instruments and communications; particular dates of origin vary by instrument and jurisdiction.
International financial centers became valuable aggregation points for claims on firms, sovereigns and banks. The growth of offshore currency markets in the second half of the twentieth century added a distinction between the monetary area issuing a currency and the location of banks intermediating claims in that currency. Aldasoro and colleagues (2026) connect the early BIS locational banking statistics to the development of eurocurrency markets in their study of offshore financial activity. These markets created a statistical problem because lending and deposit taking by foreign offices could be economically significant even where the associated claims were booked beyond the currency issuer's territory.
A separate historical problem arose from the sequencing of settlement across time zones and currencies. Glowka and Nilsson (2022), in their study of FX settlement risk, identify the 1974 failure of Bankhaus Herstatt as a consequential example of principal risk. An exchange transaction can leave one party exposed when it transfers the currency it owes before receipt of the other currency. This vulnerability connects the history of cross-border banking to the subsequent development of settlement arrangements that reduce asynchronous delivery risk.
Electronic messaging accelerated the transmission of instructions and enabled standardized communication among institutions. The growth of messaging systems did not eliminate the balance-sheet relationships that made the instructions economically effective. A bank could transmit an impeccably formatted message and still lack a funded account or authorized transaction route. The historical distinction between transmission and settlement therefore survived the shift from paper to digital infrastructure.
The global financial crisis of 2007–2009 intensified the importance of capital, liquidity, funding and risk management for internationally active banks. Rice, von Peter and Boar (2020) describe the subsequent retreat of correspondent banks as resulting from a variety of commercial and institutional changes. The trend must be placed in a period of bank consolidation, business restructuring, stronger compliance expectations and altered risk-return assessments.
The result is a more selective network in which older banking relationships coexist with newer digital channels. The institutional evolution is cumulative: communications efficiency, settlement safety, liquidity availability and legal confidence each address different constraints. Technologies often improve one element while leaving others to the institutions that already provide the underlying monetary and banking services.
Balance-Sheet Mechanics and Payment Settlement
Consider a stylized example in which Bank R operates in financial center A, Bank S operates in jurisdiction B, and both maintain dollar accounts at Bank C. The transaction is expressly lawful, fully authorized, denominated in dollars and free of fees and exchange-rate conversion for the purpose of illustrating accounting. A client of R pays 100 units to a client of S. R has sufficient transferable dollar balances with C.
At initiation, R debits its customer's deposit liability by 100 and credits, or reduces, its asset claim on C by 100 when the interbank transfer is settled. In C's books, the deposit liability to R decreases by 100 and the deposit liability to S increases by 100. S increases its claim on C by 100 and increases its liability to the recipient by 100. Across the three banks, the account changes are consistent with the movement of purchasing power from the first customer to the second.
| Institution | Asset change | Liability change |
|---|---|---|
| Bank R | Claim on C: -100 | Payer deposit: -100 |
| Bank C | No change in this same-book example | Deposit due to R: -100; deposit due to S: +100 |
| Bank S | Claim on C: +100 | Beneficiary deposit: +100 |
The example demonstrates an internal book transfer at one correspondent, not a transfer of physical currency or an assertion about central-bank reserve movements. If the banks held accounts with different correspondents, the two account providers might use an appropriate payment system or another settlement relationship to transfer value. The accounting then includes additional claims, liabilities and possibly settlement balances. A complete real-world ledger depends on the institutions, clearing arrangements, currencies and terms involved.
The ability to execute the transfer depends on the state of the claims. R needs usable funds or permissible credit, accurate beneficiary information, availability of the account service, financial-crime and sanctions controls, and settlement processing capacity. A positive book balance is insufficient if the balance is restricted, subject to a payment outage, held in another currency or unavailable within the required operating window.
In a transaction involving two currencies, agreement on a price for exchange must be distinguished from the discharge of the two currency obligations. The Basel Committee's foreign exchange risk guidelines describe payment-versus-payment as a conditional connection between the final currency transfers. If one leg fails, the arrangement protects against certain principal losses by preventing the corresponding final delivery; it does not necessarily cause the intended trade to settle or eliminate replacement-cost and liquidity consequences.
The offshore location of the account-holder does not by itself determine the final settlement institution. A dollar transfer among customers outside the United States may be completed by internal book transfer, by an interbank arrangement involving US settlement infrastructure, or through other authorized settlement structures. The analyst must identify the mechanism rather than assume every transaction uses identical institutions or crosses the same number of borders.
Correspondent Services and the Economics of Intermediation
The correspondent supplies a service bundle. Account maintenance permits the respondent to hold balances in a particular currency. Payment processing handles instructions, customer or bank transfers, confirmations, reconciliation and errors. Related arrangements can support foreign exchange, bank-to-bank credit or trade finance. The bundle varies by institutional agreement and by the provider's assessment of risk and profitability. Access to one service need not imply access to all others.
The costs of providing these services include technical infrastructure, staff, operational controls, legal and supervisory compliance, liquidity, capital where applicable, and losses associated with fraud, operational failure or counterparty exposures. The revenue side can include fees, interest margins, bundled services and wider client relationships. Some expenses respond directly to transactions; others are largely incurred when maintaining an active institutional relationship regardless of realized payment volume.
An elementary break-even calculation illustrates the importance of scale. Let F be the annual relationship-specific fixed expense, c the incremental expense per transaction, p the average revenue per transaction, and q annual transactions. Holding all other revenues and costs constant, operating surplus equals q(p-c)-F. When p exceeds c, the arithmetic threshold is q=F/(p-c). This is a conventional accounting illustration, not an empirically estimated banking model or a complete decision rule.
With a hypothetical F of USD 120,000 and an incremental margin of USD 6, the threshold is 20,000 transactions per year. Doubling the fixed expense doubles the threshold under the stated assumptions. A corridor serving a small economy can thus be unprofitable for a provider even where the average transaction is legally routine and the respondent itself remains solvent. The result says nothing about actual fees or current compliance costs in any institution.
The social value of a correspondent relationship includes the transactions it makes possible for third parties. Exporters, importers, individuals, governments and other intermediaries may benefit from continued access while the bank internalizes a narrower revenue stream. Where the difference is substantial, a provider's commercially reasonable exit can have economic costs beyond the value of the direct banking contract.
Bundling complicates the analysis. A correspondent can sustain a low-revenue payment operation because foreign exchange, trade finance, custody or a wider institutional relationship contributes additional value. An apparent decline in the profitability of payment processing does not necessarily predict a termination. Conversely, adequate aggregate revenue can fail to compensate for tail risks or operational exposures that management considers outside its appetite. Relationship-level profitability is therefore an important but incomplete explanatory variable.
Foreign-Currency Liquidity and Offshore Dollar Banking
Currency denomination establishes the unit in which monetary obligations must be met. A bank may fund itself in one currency and invest or lend in another, or raise and deploy funds in the same foreign currency. Correspondent account balances provide one means of meeting payment obligations. Offshore dollar financing extends far beyond correspondent balances, encompassing loans, securities, derivative-based funding and intragroup transfers.
The CGFS (2020) report US dollar funding describes the dollar's importance to international borrowing and the attractiveness of deep dollar markets. Its approximate half-share of the cross-border loan and international debt securities measures belongs to the report's stated period and instruments. The estimate must not be treated as the dollar share of global payments, worldwide money or current offshore wealth.
Funding liquidity is the ability to obtain funds when liabilities become due. Settlement liquidity is the ability to deliver the required settlement asset at the time and place of settlement. They interact but can diverge. A bank could have adequate medium-term dollar funding while experiencing an intraday shortage at a particular account provider. Another could maintain substantial immediately usable balances while its wholesale funding structure is fragile.
Internationally operating banking groups can allocate funds among branches and subsidiaries, but the elasticity of those internal transfers depends on contractual and regulatory restrictions, collateral, operating time, market conditions and the needs of the parent. The CGFS (2026) foreign-currency liquidity report emphasizes the role of intragroup funding, the risk of funding-market disruption and rollover exposure from derivative hedges. These observations establish why local banking exposure and group-wide liquidity must be examined together.
Prefunding provides a buffer at the cost of tying up assets that might otherwise earn income or serve other obligations. Suppose a bank holds USD 8 million in a transferable correspondent balance and the net usable yield forgone relative to an alternative admissible liquid asset is 2 percentage points per year. The illustrative annual carry cost is USD 160,000. This excludes differences in liquidity value, credit risk, regulatory treatment and any service charges. The cost can rationally be incurred to reduce the probability or consequences of late payment.
The CPMI (2022) central-bank liquidity bridges report examines how collateral held in one jurisdiction could support intraday liquidity for a related institution elsewhere under a carefully governed arrangement. The proposal responds to fragmentation in the location of collateral and usable money. A private respondent has no automatic right to such support, and an intraday bridge serves a different purpose from a central-bank foreign-currency funding backstop.
An offshore balance-sheet stress analysis should therefore identify the currency of liabilities, maturity structure, immediate liquidity by location, contingent contractual draws, admissible collateral, dependence on private correspondents and potential changes in intragroup flows. A domestic central bank may issue its own currency elastically under applicable rules but cannot issue another sovereign currency. Access to official foreign-currency liquidity arrangements depends on eligibility and institutional decisions, rather than the mere presence of dollar liabilities.
Financial Geography and Currency Hierarchies
Correspondent relationships connect places with different economic functions. A jurisdiction may specialize in fund administration, entity registration or private wealth services; settlement and foreign exchange may be supplied through financial institutions in larger markets. A large global financial center can simultaneously provide account relationships, financial markets, clearing infrastructure, legal services and treasury management. A small, externally oriented financial center may have high financial assets relative to its domestic output and a comparatively limited independent settlement infrastructure.
Aldasoro and colleagues (2026) show why offshore activity requires a view beyond residence-based financial positions. Investors, issuers and banks can act through foreign affiliates, separating the place of recording from the ownership or control of a claim. The same principle applies to analysis of international payment networks: the location of the account-holding office may differ from the location of the group making strategic decisions, the beneficiary's residence and the jurisdiction of final settlement.
Currency hierarchies derive in part from network benefits. A widely used currency offers a dense set of potential buyers, sellers, lenders, borrowers and account providers. Its use in trade and finance encourages further investment in settlement infrastructure and services. These effects are supported by market depth, institutional confidence and availability of liquid instruments; currency internationalization depends on more than technical design.
The same density creates dependencies. Banks can find many commercial uses for widely accepted monetary claims while a small number of infrastructure providers or correspondent groups become particularly consequential. Matvos and Neiman's (2026) global payments network study identifies broad connectivity in dollar and euro networks and important institutional hubs. The authors' account is based on observed banking relationships rather than an exhaustive map of every settlement transaction.
A jurisdiction-level measure of prominence should consequently distinguish external balance-sheet intensity, gross payment activity, transaction value, the supply of services to nonresident banks, and the concentration of providers offering access. None of these measures alone establishes the scale of tax advantages, legal secrecy, financial wrongdoing or domestic financial development.
The geographical position of an offshore center can change when an international banking group consolidates its offices, even if the center's legislation is unchanged. It can also change through technological interoperability, diversification of payment arrangements, the emergence of new client demand or greater restrictions on cross-border transactions. Geography here is an economic outcome of continuously reorganized institutions and claims, not simply a map of physical bank headquarters.
Network Structure, Concentration, and Resilience
The network can be represented as directed relationships among banks, carrying attributes for currency, service, jurisdiction, operating hours, limits and transaction availability. A country-to-country graph aggregates away these attributes. It is useful for some comparisons but does not provide a complete account of vulnerability. A bank can maintain a dollar relationship and lack a usable euro relationship or retain incoming payment access while facing different restrictions on outgoing services.
The CPMI's quantitative review methodology identifies active correspondents from observed payment messages and active corridors from directed jurisdiction pairs. The note expressly states that complete payment chains cannot be identified from the published observations. Accordingly, a corridor-level link is an observed activity indicator, not proof that an ultimate payer and beneficiary directly settled an underlying transaction without intermediaries.
Standard network quantities have clear economic interpretations when used with appropriate data. Degree counts connections; weighted degree considers the volume associated with connections; concentration measures the distribution of a specified activity across counterparties; and reachability assesses the availability of a path under a defined network representation. Betweenness or related intermediary measures require reliable topology and assumptions about how routes are selected. An aggregate corridor table does not supply the institution-level information necessary to estimate all these quantities.
For a stylized local banking market, consider two correspondent providers receiving 80% and 20% of payment instructions. The Herfindahl-Hirschman index of these shares is 0.8 squared plus 0.2 squared, or 0.68 on a zero-to-one scale. Four providers with equal shares have an index of 0.25. The former portfolio is more concentrated according to that measure. Its actual resilience also depends on whether the providers share an upstream correspondent or other critical service, their backup capacity and whether the smaller provider can absorb reallocated traffic.
Suppose a respondent is linked to two providers that rely on the same downstream settlement institution. If the institution becomes unavailable, both relationships may lose a particular functional capability simultaneously. Two contracts therefore need not supply two economically independent channels. Where permissible, assessments of diversification should account for common group ownership, liquidity sources, settlement infrastructure, jurisdictional restrictions and operational dependencies.
Centralization can improve normal-time efficiency. Concentrated volume can support fixed investments in processing, analytics, liquidity management and regulatory controls. It can also make particular interruptions more costly. The appropriate welfare comparison accounts for both efficiency gains and potential interruption costs rather than assuming that a greater number of links is unconditionally desirable.
Access maintained by a rarely used backup may be valuable even where little transaction activity is observed. Its value can be lost if contractual eligibility, funding arrangements or testing have lapsed. A meaningful measure of resilience consequently asks whether substitute services are usable at the required time, capacity, currency and level of risk, not merely whether an institution appears in a directory.
Empirical Evidence on Network Contraction
The Financial Stability Board's 2017 banking data report combines a survey of more than 300 banks in nearly 50 jurisdictions with SWIFT-derived evidence. It reports a 6% decline in active correspondent relationships across currencies between 2011 and 2016, and a heterogeneous regional pattern. The statistics describe observed activity within the specified reporting construction, not an estimate of all bilateral account contracts outstanding.
Rice, von Peter and Boar (2020) analyze subsequent developments in their global retreat study. They associate differences in the contraction with governance, deficiencies in controls against illicit financing, trade, growth and changes in bank strategies. The identification of corridors from locational banking data uses the economic connection between interbank balances and correspondent arrangements. Such a proxy also includes intragroup and other interbank positions and cannot independently establish that a specific account contract exists.
The European Central Bank's June 2025 international role of the euro report summarizes the later CPMI series: approximately 90,000 active correspondent observations and 9,000 payment corridors in 2022, broadly 20–30% fewer than about a decade earlier. Its definitions and Chart 12 use observed message activity and directed corridors. The underlying series ends in 2022 for that analysis, rather than the date of the ECB publication.
The figures from different documents should not be spliced without obtaining the series and checking its treatment of reporting populations, currency aggregation, directed pairs and observations. A decline in measured active relationships is consistent with consolidation of activity onto fewer providers. An economy may retain substantial total payment flows even as the number of channels falls. Increases in traffic on remaining links and exits of little-used links can occur together.
Two further interpretations require care. First, a decrease in active correspondents can reflect altered messaging practices, inactivity or replacement rather than the legal termination of every previously observed account. Second, an increase in observed corridors need not establish better access for low-income households or small firms. Payments may serve a narrow set of institutions, be expensive or require services beyond the reach of particular potential customers.
The published data support the conclusion that the correspondent network experienced a broad and uneven contraction during the observed period. They do not support a single universal measure of the size of the entire offshore system. An empirical assessment of access should combine dated relationship measures with actual availability of services, fees, rejection and delay data, and measures of customer-level outcomes.
Statistics of Positions, Flows, and Payment Activity
The empirical analysis of correspondent banking draws on three fundamentally different observable families. Payment-message data describe the sending and receiving of financial instructions under a specified network's coverage. Banking statistics measure outstanding claims and liabilities, or changes estimated from positions. User-level payment data measure prices, speed, accessibility and other realized service outcomes. Each family answers a different question.
The BIS locational banking statistics organize positions from the residence of the reporting banking office and offer currency and counterparty dimensions, including intragroup positions in the relevant aggregates. Quarterly changes adjusted for exchange-rate movements and documented breaks approximate underlying changes in claims and liabilities. They are different from gross payment turnover, since offsetting transactions may leave outstanding positions unchanged.
The BIS consolidated banking statistics aggregate the claims of reporting banking groups and eliminate intragroup claims. Immediate-counterparty and guarantor measures address distinct risk concepts. Consolidation is valuable when identifying the banking system bearing credit exposure, while the locational perspective is essential when identifying the location of particular funding and booking functions. Neither is sufficient alone to reconstruct all payment messages or ultimate beneficiaries.
Assume a bank begins the quarter with a nostro balance of 100, receives 900 of settled inflows and pays 850 of settled outflows, with no other adjustments. The closing balance is 150 and the net change is 50, although gross activity across the account is 1,750. Counting all displayed receipts and disbursements as separate independent final economic transactions would also be inappropriate when the same underlying transaction passes through more than one intermediary. The example makes the identification problem explicit.
Currency changes introduce a further measurement issue. A position of one million euros translated into dollars at two different exchange rates changes in dollar terms even if the euro balance and economic claims have not changed. The LBS adjustment methodology addresses exchange-rate and reporting breaks at the appropriate aggregate level. The adjusted change still cannot identify the gross sum of payment instructions that occurred between observations.
The analyst should record the unit of observation, direction, report date and underlying reference period, coverage of institutions, cross-border definition, currency denomination, treatment of intragroup positions, treatment of exchange-rate effects and whether statistics represent observed or inferred relations. Research comparing countries must also account for statistical confidentiality and missing bilateral positions. Missing values are not zero claims or absent relationships.
| Observable | Typical unit | Main use | Principal limitation |
|---|---|---|---|
| Active correspondent | Bank activity in defined data | Network participation | Contract status and full chains unknown |
| Active corridor | Directed jurisdiction pair | Geographic connectivity | Frequency, capacity and economic purpose may differ |
| Locational positions | Banking office and counterparty | Place and currency of financial claims | Stocks are not message or transaction turnover |
| Consolidated positions | Banking group and exposure basis | Group-wide credit risk | Intragroup legs removed |
| Remittance price | Quoted service and corridor | Consumer service costs | Quotes do not measure every actual transaction |
Empirical research should document each source vintage rather than construct a synthetic total by adding the units in this table. The statistical architecture is an essential part of the scientific interpretation, especially in an offshore setting where the same underlying financing can appear through multiple institutional locations.
De-Risking, Compliance, and Relationship Exit
Financial-crime controls are material conditions of access to correspondent banking. Institutions face legal obligations, supervisory expectations, operational costs and potential losses from misuse of their services. Respondent banks can differ in ownership transparency, customer profiles, transaction quality, supervisory environment and capacity to respond to information requests. A correspondent assesses these factors alongside commercial considerations.
De-risking in the policy debate commonly concerns avoiding categories of risk through relationship withdrawal or restrictions instead of managing risk on an appropriately differentiated basis. The World Bank's de-risking surveys commissioned in the context of international fact-finding identify the concern across correspondent banking and remittance services. A survey of institutions captures reported experiences and business judgments; it is not an exhaustive census of unlawful activity or proof of the reasons for each termination.
The FATF's (2016) correspondent banking guidance clarifies how risk-sensitive measures apply to these relationships. Its official summary explains that the FATF recommendations do not impose customer due diligence on each individual customer of the respondent bank. Enhanced diligence should correspond to the risk involved. This does not remove applicable sanctions requirements, information requests or a bank's obligation to assess its relationship. National law and supervision can introduce additional requirements.
A respondent's ability to carry out effective due diligence and produce reliable information has economic value. The provider may economize on repeated information collection when standards, controls and reliable evidence are in place. An improvement in institutional integrity can improve the potential attractiveness of a relationship, yet profitability, business priorities, currency liquidity and other constraints may still prevent its continuation.
Relationship termination can arise from many causes, including mergers, risk limits, withdrawal from particular business lines, compliance concerns, deteriorating financial condition or broader changes in market strategy. Calling all exits de-risking erases material distinctions. Conversely, the description of an exit as a business decision does not establish that the decision was unrelated to perceived regulatory exposure.
Demand-side and supply-side feedbacks can reinforce exit. If payment services become costlier, clients may use competing banks or other authorized providers, reducing the respondent's volume and revenue. A lower volume may make the remaining relationships harder to justify commercially. An isolated loss can thereby become a more persistent accessibility problem.
Blanket exclusion can also shift transactions toward less observable channels and reduce access to legitimate payment services. The FATF policy concern links inclusion and integrity: proportionate, effective risk controls can support the use of transparent and accountable systems. The appropriate economic policy must preserve the ability to reject prohibited transactions and relationships whose risks cannot be managed while reducing avoidable losses of lawful access.
International Trade, Firm Behavior, and Trade Finance
A cross-border payment is often the final or intermediate step in a real commercial relationship. A firm must pay for imports, receive export revenues, meet contractual deadlines and obtain working capital. Payment reliability influences whether trade can be initiated, continued or expanded. A reduction in international banking access can therefore affect real activity even where domestic credit and production technology initially remain unchanged.
A correspondent relationship may contribute directly to payment execution and indirectly to other banking products. Trade finance, documentary arrangements, foreign exchange services and payment collection can be affected when banks withdraw particular services. Payment delay can increase working-capital needs, especially for businesses with limited cash reserves or small margins. Firms may also bear an administrative burden when moving accounts or establishing different authorized payment arrangements.
Borchert, De Haas, Kirschenmann and Schultz (2023) use data on severed relationships and firm exports in their EBRD working paper Broken Relationships. The published author summary reports declines in exports on both extensive and intensive margins for corporate clients of affected banks, partial substitution into domestic sales, and adverse changes in total revenues and employment. These findings concern an emerging European sample and the terminations studied by the authors; the present review does not independently re-estimate their econometric specifications.
The extensive margin concerns whether firms participate in exporting or particular export activities; the intensive margin concerns the amount traded by those active in the relevant activity. A fixed switching cost can influence participation while recurrent payment costs can reduce transaction sizes or profitability. Distinguishing the margins therefore assists interpretation of economic damage rather than merely recording a decline in total exports.
Firm-level causal research must account for the conditions that cause a correspondent relationship to be severed. For example, a banking institution may lose foreign relationships during a period of deteriorating local economic conditions. A naive comparison of subsequent exports before and after the termination would combine the effect of termination with changes in underlying demand, firm solvency or bank health. Credible designs need a comparison group, defensible timing, relevant controls and sensitivity to spillovers and concurrent events.
A generalization from one study to specialist offshore jurisdictions is a conditional inference. A financial-center entity that carries out internationally distributed investment administration may face different transaction types, liquidity requirements and alternatives than a manufacturing exporter. The study supplies empirical evidence that the loss of payment and banking connectivity can affect real economic activity, while the magnitude and transmission channels must be studied in each market.
Remittances, Households, and Inclusive Access
Cross-border remittances connect workers and households located in different economies. Payment service providers can use banks, agent networks, account intermediaries and domestic payment infrastructures. Consumers may see a single displayed transfer fee even though the full economic cost includes conversion, intermediary charges and recipient-side deductions. The institutions maintaining correspondent access influence the service economics behind the retail interface.
The World Bank's Remittance Prices Worldwide portal supplies comparable quoted price information across a defined set of corridors. The snapshot retrieved for this study, displaying an August 18, 2025 update, reported an average of 6.36% and coverage of 367 corridors. The number is a dated portal indicator, not the November 2026 realized global average, and it should not be interpreted as applying to each corridor or transaction.
A simple illustrative transaction shows why transparency matters. A household sending USD 200 may face a USD 4 explicit charge and a currency-conversion margin equivalent to USD 6. Holding all other costs aside, the total economic transfer expense is USD 10 or 5% of the principal. If the beneficiary also faces a USD 2 receiving charge, the corresponding total becomes USD 12 or 6% of the principal. Actual quote methodologies and exchange-rate benchmarks must be specified before comparing providers.
A fixed fee has a higher proportional incidence on a small transfer than on a large one. If relationship maintenance costs are spread across a low-volume corridor, institutions may raise flat or minimum charges; this can affect low-income customers disproportionately. Faster payment can reduce household uncertainty and meet urgent needs even if the monetary fee does not change. Hence the economic value of progress includes timeliness, predictable total charges and reliable access.
A reduction in correspondent relationships is neither a sufficient nor necessary explanation for every increase in remittance prices. Exchange-rate market conditions, retail competition, agent distribution costs, available payment modes, consumer protection and local infrastructure can independently alter observed prices. Causal inference requires data connecting upstream bank service changes to specific retail providers, corridors and customers.
Access also depends on whether customers can actually meet eligibility and verification requirements and use a suitable service. A policy that merely adds a payment platform need not reach rural or financially excluded recipients. The welfare question is whether legitimate households can receive and send value at a reasonable cost, within predictable times, through accountable service arrangements.
Offshore Financial Centers and Small-Economy Vulnerabilities
An externally oriented financial center can collect fees and supply services to nonresident individuals, enterprises and financial institutions. A substantial share of its economic activity may depend on the ability to receive foreign-currency payments, make distributions, settle investments, collect subscriptions or pay international service providers. In that sense, correspondent capacity can be an enabling input into the local export of financial services.
A small domestic banking market presents a distinct commercial problem. Gross cross-border service demand may be low, concentrated in certain institutions, variable over time or dominated by relatively few customers. The fixed cost of maintaining a relationship cannot always be distributed across a sufficiently broad volume of activity. In specialist offshore markets, aggregate financial claims may be large even where the number and composition of transactions are modest compared with those processed in larger centers.
The 2017 IMF working paper by Alleyne and colleagues, Loss of Correspondent Banking Relationships in the Caribbean, describes the experience of Caribbean banks in a period of withdrawals. The authors report that replacement or multiple relationships moderated observed aggregate economic effects while services became more costly and some activities lost access. The indexed article identifies The Bahamas, Belize and St. Lucia as case studies. Detailed country-level causal magnitudes are not inferred from the accessible abstract and indexed excerpts.
The evidence implies that a regional aggregate can conceal sector-specific disruption. A local bank that successfully replaces a correspondent may still face higher minimum balances, restricted services, longer processing or a different risk review. Firms and households dealing with that bank may experience increased expense despite unchanged formal indicators of national banking access.
Offshore financial-center activity must also be separated from the experience of small economies generally. A tourism-dependent island, remittance-receiving country or agricultural exporter can experience correspondent withdrawal without having a major nonresident financial sector. Conversely, a large economy with substantial international financial business can possess diversified networks and extensive market access. The Pacific example in the next section is used to study small-market vulnerability, not to classify the participating countries as offshore centers.
Where financial intermediation contributes to the domestic economy, service disruption may affect employment and tax receipts. Large recorded external assets alone do not determine the magnitude of that effect. It depends on how local service production, wages, tax liabilities and domestic spending relate to the affected activities. A scientific assessment therefore requires both sector-level domestic data and evidence of actual international banking access.
Comparative Evidence from the Pacific and Caribbean
The World Bank (2025) announced the participation of Solomon Islands in a regional Pacific correspondent banking project. According to that announcement, available correspondent banking services in Solomon Islands declined by 57% between 2011 and 2022. The institution characterizes the resulting risk as potential financial disconnection and describes technical assistance, regional reforms and a standby arrangement for emergency services.
The reported 57% is tied to the World Bank's description of available correspondent banking services and the specified country-period. It is not a finding that 57% of transactions, customers or total financial assets disappeared. The announcement establishes the program design and stated motivation. It does not by itself evaluate subsequent program effectiveness or prove a current level of access.
Comparing the Pacific program with the Caribbean evidence illustrates differences in economic dependence. Both settings involve small markets and the difficulty of sustaining international services. The region-specific drivers, institutional structures, payment needs, financial-sector specialization and regulatory settings vary. Effective comparisons should therefore isolate common commercial mechanisms while retaining locally specific evidence about transaction volumes, currency demand, market structure and real economic activity.
Regional cooperation may address scale constraints. Coordinated data requirements, compatible supervisory practices, shared infrastructure and pooled technical expertise can reduce duplication and help institutions meet internationally expected controls. Regional arrangements can also organize contingency provision for temporarily underserved markets. Such arrangements require clear responsibility, financial viability, eligibility, governance and effective compliance.
A pooled relationship or shared technical system can introduce its own concentration risk. For example, several economies obtaining a common service through a single provider may be more vulnerable to interruption affecting that provider. The policy design must compare the reduction in fixed expenses and the availability benefits against the costs of shared dependencies. A regional mechanism does not automatically supply independent backup arrangements for every country.
The relevant performance indicators include actual sustained service coverage, availability by currency and transaction type, total end-user costs, time to settle, provider switching ability, risk-control quality and the fiscal or institutional cost of supporting the arrangement. The value of the intervention cannot be inferred from announced financing alone. It should be assessed with observed outcomes and attention to what would have happened in the absence of the program.
Sanctions, Financial Integrity, and Network Fragmentation
International payment infrastructure can be affected by public restrictions on permitted transactions and counterparties. Financial sanctions can restrict access to assets, services or institutions within the applicable legal regimes. Their economic effect depends in part on the centrality of the affected services and the ability of institutions to maintain lawful alternative business relationships. The specific legality of any transaction remains a distinct jurisdiction-sensitive question.
Matvos and Neiman's (2026) study of financial sanctions reports evidence from the 2021–2025 period that targeted institutions lost access to major-currency correspondent networks. The authors also report spillovers to institutions not individually targeted in countries where sanctions intensified, and limited average shifts away from the dollar outside heavily sanctioned economies. This is working-paper evidence on network connectivity, not a comprehensive assessment of every sanction's welfare or enforcement effects.
The economic distinction between targeted restrictions and wider effects is important. Banks may reassess the risks of relationship maintenance across a whole market because transaction screening, ownership assessment and legal uncertainty become costlier. Lawful customers of non-targeted institutions may then experience service restrictions or increased prices. Evidence of a spillover does not demonstrate either illegality by those customers or a failure of every individual compliance decision.
The ECB (2025) analysis of payment fragmentation describes correspondent retrenchment and the growing relevance of geopolitical incentives for developing payment links. Currency and infrastructure choices may reflect political as well as commercial considerations. These motives can coexist with efficiency objectives such as lower costs, shorter processing times or increased resilience.
The distribution of network power depends on the availability of economically meaningful substitutes. A transfer network may offer technical communication in many places but still require liquidity, interoperable settlement, regulatory authorization and acceptance of the currency delivered. Those additional conditions constrain how easily a market can substitute away from an established arrangement. The study does not provide methods for avoiding sanctions or financial-crime controls.
The welfare consequences of fragmentation can include higher duplication costs, a reduction in economies of scale, narrower currency liquidity, less predictable service arrangements and changes in the distribution of compliance risks. In some settings, carefully designed diversification can improve resilience. Assessing the outcome requires separating lawful redundancy and potential efficiency gains from the effects of market segmentation and restrictions on available financial services.
Settlement Risk, Operational Risk, and Financial Stability
Correspondent banking creates exposures through deposits, credit arrangements, foreign-exchange obligations, processing dependencies and settlement timing. These exposures transmit stress across balance sheets and institutions. Their materiality differs between a fully funded same-book transfer and a complex operation requiring several settlement arrangements or contingent liquidity.
Credit exposure arises when the respondent relies on the correspondent to honor its liabilities or when one institution extends credit to another. Liquidity exposure arises when the institution cannot obtain the requisite currency or settlement asset on time. Operational risk concerns interruptions, errors, cyber incidents, reconciliation failures and deficiencies in internal or outsourced systems. A service can fail for operational reasons even when the parties have adequate capital and liquidity.
Foreign-exchange principal risk deserves particular treatment. Glowka and Nilsson (2022) report in FX settlement risk that approximately USD 2.2 trillion of daily deliverable FX turnover in April 2022 remained exposed to settlement risk, compared with an estimated USD 1.9 trillion in April 2019. The 2022 figure was around one-third of deliverable turnover. These are dated estimates of transaction value at risk, not realized losses, not offshore-only FX flows and not amounts outstanding on balance sheets.
Payment-versus-payment reduces the risk of final delivery of one leg without the other. The Basel FX risk guidelines explain the continuing need to manage currency-specific liquidity, correspondent exposures and potential FMI disruptions. Netting can reduce the amount that must actually be delivered, but legal enforceability and the consequences of a settlement failure determine the economic protection. If obligations revert to gross amounts following disruption, previously sufficient liquidity may become inadequate.
A correspondent's default or abrupt withdrawal can affect respondents even where balance-sheet credit losses are limited. Banks may need to acquire new service arrangements, reestablish account balances and inform customers while service contracts and due diligence are being completed. Revenue falls if payments cannot be executed, and treasury costs rise if liquidity must be redistributed.
Stress testing should consequently combine several distinct scenarios: failure or exit of an important account provider; loss of shared infrastructure; disruption to intragroup funding; currency funding stress; operational delays; and a surge in urgent payment demand. These scenarios should be assessed without treating overlapping exposures as independent. A resilient institution must maintain workable financial and operational capacity under plausible stress, rather than relying on the sum of nominal available account balances.
Information, Interoperability, and Technical Reform
Cross-border payments combine financial transfers with data about the parties, amounts, currencies, references and the applicable service arrangements. Information quality affects automated processing, screening, reconciliation and the customer's ability to understand the service. Different message formats and inconsistent interpretation of fields can increase manual reviews, errors and delays.
The CPMI's (2016) technical recommendations for correspondent banking address identity, data exchange and messaging concerns. The use of common identification and data conventions can reduce costly duplication and facilitate the transmission of accurate information. The benefits depend on consistent application, accountable maintenance, reliable source information and practical acceptance across institutions.
The 2026 CPMI ISO 20022 requirements provide updated harmonization guidance for structured cross-border payment data. A common message format can support interoperability, but identical syntax does not ensure agreement on regulatory permissions, business meaning, validation quality or the usability of the intended settlement arrangement. Implementation and the quality of information are empirical questions.
An illustrative institutional comparison clarifies the difference. Bank A introduces a structured messaging format and experiences fewer manual repairs, while Bank B introduces the same format but still receives incomplete or inaccurate underlying customer data. Bank B can continue to experience delays and requests for further information. The standard provides an opportunity to improve the information process rather than a guarantee of uniform results.
Interoperable payment systems require agreement about access, data, recovery time objectives and service continuity, dispute resolution, settlement assets, foreign exchange arrangements, oversight and responsibilities. Integration can reduce duplicated technical connections, but operational and legal compatibility must be maintained across the participating jurisdictions.
Progress in messaging should therefore be assessed alongside end-user outcomes, safety and financial resilience. A higher rate of automated processing can be beneficial while fees remain unchanged if competition is weak or funding requirements continue to be expensive. The efficiency of the entire payment service cannot be identified from a message-processing indicator alone.
Alternative Payment Arrangements and the Future of Correspondent Banking
Bech, Faruqui and Shirakami (2020) review cross-border and multicurrency designs in Payments without Borders. They distinguish improvements to existing chains, remote access to payment infrastructures, the linking of domestic payment systems and dedicated cross-border arrangements. These approaches address different elements of the same economic demand for safe and efficient settlement.
Connecting existing domestic infrastructures may yield scale benefits by using established local participation networks and standards. It can reduce the need to establish separate bilateral technical connections for every pair of systems. Nevertheless, the system still requires arrangements for currency conversion, liquidity, financial integrity, governance and accountable settlement. These functions can be supplied by banks, nonbank service providers or properly governed financial market infrastructures according to applicable rules.
The BIS Innovation Hub's (2024) Nexus blueprint sets out a design for linking domestic instant payment systems, informed by collaboration with central banks and operators in Indonesia, Malaysia, Singapore, the Philippines and Thailand. The published report establishes the project design and participating design-stage institutions. It does not establish a universal network or the subsequent deployment status of every participating jurisdiction.
Retail instant payments can offer faster availability for some transactions while wholesale funding, securities settlement, high-value commercial transfers and trade finance continue to demand additional capabilities. A specialist offshore center whose clients primarily deal with large financial positions may benefit from very different infrastructures than a household sending modest remittances. Payment functionality should be matched to the economic activity.
The distinction between tokenized money and a tokenized message or claim is also analytically relevant. A technology can represent an obligation on a new ledger without establishing the issuer's ability to meet redemption claims, the governing rules or the availability of conversion into other settlement assets. The economic questions concerning creditworthiness, liquidity, access and finality persist across technical representations.
Competition between payment arrangements can alter fee structures and institutional roles. Services with fewer intermediary stages may lower certain variable costs, while their operators incur new fixed costs and governance obligations. Existing banking institutions may participate in new arrangements by providing liquidity, custody, currency conversion or compliance services. A change in the technical path need not mean the disappearance of bank intermediation.
Evidence of a pilot, blueprint or announced implementation should not be conflated with sustained production availability, observed market share or lower realized costs. Valid comparison requires the same transaction definition, reference period and total-cost measurement, and must account for differences in settlement quality, accessibility and consumer protection.
Policy Evaluation and the International Reform Agenda
The G20 Roadmap addresses the price, speed, availability and transparency of cross-border payments. The FSB's cross-border payment targets differentiate wholesale, retail and remittance markets and identify quantitative goals, generally with an end-2027 horizon. End-user transparency includes the information needed to understand the overall expense, expected time, payment status and conditions of the service.
The FSB's (2025) annual progress report describes policy milestones and improvements in aspects of payment speed while finding relatively small aggregate improvements for service users and persistent costs. The FSB's (2026) consolidated progress report of November 16 indicates that international policy work under the Roadmap was largely complete, but implementation and end-user improvements remained uneven. The latter finding is drawn from the retrieved executive-summary text; no new chart values are reproduced.
The evolution of these assessments illustrates the difference between regulatory or technical outputs and economic outcomes. Issuing recommendations, building a new technical interface or adopting a message standard are implementation milestones. Access to a service, the cost paid, time to receive usable funds and system reliability are outcomes. The latter require observed measurement across specific types of transaction and affected populations.
Potential policy responses can be grouped by the economic problem addressed. Technical and information reforms target avoidable processing costs. Proportionate and effective supervision targets integrity while reducing unnecessary uncertainty. Competition and interoperability target structural access barriers. Carefully governed liquidity arrangements target fragmented resources. Contingency measures target essential service continuity. Each policy has costs, dependencies and constraints that must be measured as well as its intended benefits.
A small-economy access intervention, for example, should establish whether the main constraint is insufficient transaction volume, weaknesses in financial-crime controls, lack of appropriate counterparties, unusable foreign-currency liquidity or the cost of compliance with divergent requirements. An intervention directed at documentation alone will have limited effect when the dominant barrier is insufficient scale. Similarly, an access subsidy that fails to address real integrity weaknesses can impose risks on providers and public authorities.
The long-term objectives are efficient allocation of payment capacity, adequate risk controls, meaningful access for legitimate activity and sustainable provision. Governments and international institutions should evaluate policies based on observed reductions in barriers, costs and vulnerability, not by the number of announcements or connections nominally established. Attribution of improvements to a specific intervention requires credible evaluation rather than simple comparison of aggregate indicators over time.
Research Methods and Reproducibility
Research on this subject requires evidence matched to the claim. Regulatory reports and central-bank documents establish definitions, official programs and supervisory concerns. The original data documentation establishes the population and unit being measured. Empirical studies establish findings under their research designs. Historical descriptions need archival or historical scholarly evidence. Illustrative accounting examples clarify conditional mechanisms rather than measure their prevalence.
An appropriate descriptive dataset for studying a respondent's correspondent access would require a stable institution identifier, group ownership, location of the relevant office, currency, direction, services offered, activity period and available contractual capacity. Payment data would ideally identify transaction type, intermediary dependencies and completion status under applicable confidentiality rules. In public research these details are often unavailable, so estimates must disclose coverage limitations.
A time-series analysis should distinguish mergers or identification changes from actual service exits. A geographical comparison should account for the size and composition of the domestic economy, extent of nonresident banking, currency use, market concentration and reporting limitations. Where confidential data cannot be obtained, research can describe the resulting identification gap and use publicly accessible proxies only for questions they can legitimately answer.
For causal inference one may study event timing, compare affected and suitable unaffected units, examine pre-event trends and test sensitivity to changes in group composition. A relationship exit can affect firms banking with other institutions through competition or shared network dependencies. An ordinary difference-in-differences design that assumes no spillovers therefore requires a substantive defense of that assumption or a method that accounts for the exposure of comparison units. This work proposes no new estimation framework and performs no econometric replication.
For an assessment of the economic cost of relationship loss, useful outcomes include actual transaction prices, payment delays, rejected or returned transactions, service substitution, trade participation, working-capital requirements and customer switching. Bank-level indicators should be linked to firm or household outcomes where possible. Aggregate exports alone cannot isolate the effects of payment interruptions, and remittance quotes alone cannot measure the welfare of all users.
Researchers should identify source access levels and cite actual documents at the point of each material factual claim. Estimates must retain the original period and population. Privately held transaction details and personal data should be protected; reported results should avoid inferring crimes or individual motives from offshore geography. Documentary allegations, administrative findings and adjudicated facts require different evidentiary treatment.
Replication of the mechanisms illustrated in this monograph involves checking the double-entry examples, arithmetic and definitions. Replication of cited empirical research would require original datasets, code and the relevant permissions. The present work accomplishes the former and reviews the published findings at the accessible source depth. It does not claim the latter.
Integrated Economic Synthesis
The economic connection between international correspondent banking and offshore finance is the provision of usable payment and settlement capacity for financial claims whose ownership, currency, recording location and intended economic use can be distributed across jurisdictions. The connection has consequences at the transaction level, at the balance-sheet level and at the scale of the international financial network.
At the transaction level, an account relationship translates an underlying monetary obligation into a feasible settlement process when the needed currency, information, authorization and operating arrangements are available. The relationship is economically valuable because it allows a respondent to provide internationally usable financial services without supplying every infrastructure and service itself. Its value depends on the conditions of access as well as the nominal existence of the contract.
At the balance-sheet level, international intermediation creates a demand for liquid money at particular institutions, in particular currencies and at particular times. Offshore currency financing can support large positions and substantial economic activity, while payment execution places immediate constraints on local currency-specific liquidity and account availability. Settlement records and financial exposures therefore provide complementary perspectives.
At the network level, the concentration of service providers can be both a source of efficiencies and a potential vulnerability. Large providers can sustain costly infrastructure and controls by processing activity for many institutions. Jurisdictions depending on a small set of providers bear a greater risk that particular commercial, liquidity or operational decisions reduce their service availability. The magnitude of that risk depends on actual substitutability and shared dependencies.
The distributional consequences reach beyond banks. Firm export participation, household remittance availability, international investment administration and essential imports depend on practical payment access. Evidence from the Caribbean and Pacific documents small-market service pressures, while firm-level findings from emerging Europe support a real-economy transmission mechanism. These different sources do not supply one transferable global estimate; they do establish why the economic effects merit examination at multiple scales.
The resulting policy objective is to preserve safe and accessible international payment capacity while managing the financial and operational risks produced by cross-border intermediation. Reforms must address the specific constraint responsible for high costs or inadequate services. More standardized documentation may reduce administrative burdens, new technical links may improve processing, and carefully managed liquidity arrangements may reduce fragmentation. None independently resolves every source of economic vulnerability.
The central authorial inference is that an offshore jurisdiction's financial capacity should be assessed through the combination of its recorded claims and effective access to the institutions and assets needed to discharge its monetary obligations. A large stock of foreign claims or a large count of registered financial entities is incomplete evidence of international service capability. Conversely, low absolute positions may coexist with economically indispensable payment relationships. This is an analytical conclusion derived from the accounting, statistical, institutional and empirical evidence reviewed, not a new named theory or quantitative measure.
Conclusion
International correspondent banking supplies the account, liquidity and institutional services that enable internationally distributed economic participants to conduct monetary transactions. It is a central component of offshore finance because offshore activity separates financial claims and obligations across institutional residences, ownership structures and currency areas. The ability to transfer value under those conditions is a practical economic resource.
The global network is structured by market scale, currency use, compliance capacity, commercial strategy, liquidity arrangements and settlement infrastructure. The observed contraction in correspondent activity from the early 2010s to 2022 was broad but uneven. Evidence concerning transaction services, firm exports and small economies indicates that the loss of a relationship can matter even where aggregate payment activity persists or alternative arrangements are eventually established.
Financial geography requires more than identifying asset locations. Residence-based statistics record the location of banking offices, consolidated measures trace banking-group exposures, and payment-message statistics describe active connections within their reporting coverage. The categories cannot be added together as interchangeable estimates. Each has a defined scientific role.
The economics of correspondent services exhibits fixed costs, liquidity commitments and additional exposures that can be difficult to recover in low-volume markets. Scale and centralization can facilitate efficient service delivery, while dependence on common providers can reduce resilience. Institutional integrity and proportional risk management are important conditions for sustainable access; lawful service maintenance and effective controls should be considered jointly.
The international reform agenda has produced significant standards, technical designs and policy coordination. As the FSB's November 2026 assessment indicates, improvements in the experience of payment users remain uneven. This distinction reinforces the need to measure sustained access, total realized cost, time to usable funds, transaction safety and the viability of alternative arrangements.
The principal conclusion is that correspondent connectivity is an economically consequential input into the operation of offshore financial centers and the wider international economy. Its contribution must be evaluated alongside financial positions, real economic activities, currency liquidity and the institutional capacity to manage risk. This integrated perspective permits a more accurate understanding of international capital intermediation, the distribution of economic benefits and vulnerabilities, and the conditions for reliable lawful participation in global finance.
References
Aldasoro, I., Hardy, B., von Peter, G., & Wooldridge, P. (2026, March 16). International finance through the lens of BIS statistics: Offshore activity. BIS Quarterly Review.
Alleyne, T. S. C., Bouhga-Hagbe, J., Dowling, T., Kovtun, D., Myrvoda, A., Okwuokei, J. C., & Turunen, J. (2017, September 29). Loss of correspondent banking relationships in the Caribbean: Trends, impact, and policy options. IMF Working Paper No. 2017/209. https://doi.org/10.5089/9781484320686.001.
Bank for International Settlements Innovation Hub. (2024, July 1). Nexus: Enabling instant cross-border payments. Bank for International Settlements.
Bank for International Settlements. (n.d.-a). Locational banking statistics: Overview. BIS Data Portal. Accessed November 18, 2026.
Bank for International Settlements. (n.d.-b). Consolidated banking statistics: Overview. BIS Data Portal. Accessed November 18, 2026.
Basel Committee on Banking Supervision. (n.d.). Foreign exchange risks. Basel Framework consolidated guidelines, RMA 20. Accessed November 18, 2026.
Bech, M. L., Faruqui, U., & Shirakami, T. (2020, March 1). Payments without borders. BIS Quarterly Review.
Borchert, L., De Haas, R., Kirschenmann, K., & Schultz, A. (2023, December). Broken relationships: De-risking by correspondent banks and international trade. EBRD Working Paper No. 285. European Bank for Reconstruction and Development.
Committee on Payments and Market Infrastructures. (2016, July 13). Correspondent banking. CPMI Papers No. 147. Bank for International Settlements.
Committee on Payments and Market Infrastructures. (2022, September 7). Liquidity bridges across central banks for cross-border payments. CPMI Papers No. 209. Bank for International Settlements.
Committee on Payments and Market Infrastructures. (2023, final data vintage 2022). CPMI quantitative review of correspondent banking data. Bank for International Settlements, online methodological note and data collection.
Committee on Payments and Market Infrastructures. (2026, February 26). Harmonised ISO 20022 data requirements for enhancing cross-border payments: Updated report. CPMI Papers No. 230. Bank for International Settlements.
Committee on the Global Financial System. (2020, June 18). US dollar funding: An international perspective. CGFS Papers No. 65. Bank for International Settlements.
Committee on the Global Financial System. (2026, March 12). Foreign currency funding risk and cross-border liquidity. CGFS Papers No. 71. Bank for International Settlements.
European Central Bank. (2025, June). The international role of the euro. European Central Bank, section on cross-border payments, Chart 12.
Financial Action Task Force. (2016, October). Guidance on correspondent banking services. Official guidance summary hosted by the Financial Stability Board.
Financial Stability Board. (2017, July 4). FSB correspondent banking data report. Financial Stability Board.
Financial Stability Board. (2021, targets endorsed by G20; online resource subsequently updated). G20 targets for enhancing cross-border payments. Financial Stability Board, target and monitoring resource.
Financial Stability Board. (2025, October 9). G20 roadmap for cross-border payments: Consolidated progress report for 2025. Financial Stability Board.
Financial Stability Board. (2026, November 16). G20 roadmap for enhancing cross-border payments: Consolidated progress report for 2026. Financial Stability Board, executive summary, pp. 1–2.
Glowka, M., & Nilsson, T. (2022, December 5). FX settlement risk: An unsettled issue. BIS Quarterly Review.
Matvos, G., & Neiman, B. (2026). Financial sanctions and the global payments network. Becker Friedman Institute, University of Chicago, working paper; NBER Working Paper No. 35453, DOI 10.3386/w35453.
Rice, T., von Peter, G., & Boar, C. (2020, March). On the global retreat of correspondent banks. BIS Quarterly Review, 37–52.
World Bank Group. (2015). World Bank Group surveys probe de-risking practices. Financial Sector, official survey summary.
World Bank. (2025, June 5). Solomon Islands joins regional initiative to safeguard financial lifelines as Pacific Banking Project goes live. Press Release No. 2025/079/EAP.
World Bank. (n.d.). Remittance Prices Worldwide. Data portal, last update August 18, 2025 in retrieved snapshot. Accessed November 18, 2026.