RESEARCH DOSSIER

Offshore Bank Lending

Abstract. Offshore bank lending distributes origination, funding, credit exposure and borrower activity across jurisdictions and currency areas. This Research Dossier examines the economic organization of international bank credit with particular attention to offshore booking offices, foreign affiliates, balance-sheet consolidation, credit pricing, syndication, foreign-currency exposure, statistical measurement, crisis transmission and prudential governance. It integrates historical and institutional evidence with published empirical research and transparent hypothetical accounting examples. The analysis distinguishes creditor residence from bank nationality, immediate debtor from guarantor, gross claims from final-borrower financing, and funded balances from commitments. It establishes that the economic consequences of offshore bank lending depend on the structure of liabilities, credit allocation and the institutional capacity to absorb liquidity and solvency shocks. The resulting synthesis identifies the conditions under which international credit supports productive investment or transmits financial stress, and specifies the evidence required for causal and welfare analysis.

Keywords: Offshore bank lending; international banking; cross-border credit; eurocurrency; bank nationality; financial centers; loan syndication; foreign currency; country risk; banking statistics; credit supply; financial stability; prudential supervision

Scientific object and research method

Offshore bank lending comprises bank-credit transactions intermediated outside a specified home-market or currency perimeter, including loans booked through foreign affiliates and cross-border financial centers. In the strict eurocurrency sense, the contractual currency is foreign to both the lending office and the debtor. In the institutional sense used throughout this dossier, the offshore element consists in the location of intermediation outside the relevant home economy. A foreign-controlled bank's ordinary local-currency lending in its host economy and a domestic bank's foreign-currency lending to a domestic customer are adjacent categories; neither qualifies automatically as offshore lending. The central object is the credit asset and the economic consequences of its jurisdictionally distributed origination, funding and risk allocation. Offshore financial activity has several established meanings; this study identifies each transaction by its dimensions before applying a particular classification. A foreign-currency loan and a cross-border loan overlap when the relevant residence and currency tests are both satisfied.

The research question is how jurisdictionally distributed lending changes the formation, pricing, availability and transmission of bank credit. A commercial borrower negotiates a contract with an institution, receives purchasing power, undertakes a repayment obligation, and generates a financial asset for the creditor. The economics of that relationship extends beyond the place where the contract is booked. The office may procure funding from an affiliate; the group may allocate capital centrally; the debtor may use the proceeds in another economy; and a guarantor may accept part of the loss exposure. Each of these facts changes the interpretation of a particular statistic, the incidence of a particular shock, or both.

The method combines institutional history, comparative statistical analysis, balance-sheet reasoning and a critical synthesis of published empirical studies. The residence and consolidation perspectives of the BIS locational banking statistics and the consolidated banking statistics (BIS, n.d.) are the principal empirical infrastructure. Published findings are interpreted at the level of the study populations and designs. Numerical examples introduced by the author are hypothetical accounting or pricing exercises; they are not observations, forecasts or calibrated stress estimates. No independent econometric estimation, comprehensive live data extraction, or audit of individual institutions is claimed. The research cutoff is November 19, 2026. The most recent dated aggregate benchmark used is the BIS October 29, 2026 release for end-June 2026, with its mirror-access limitation documented in the evidence record. The research is an integrative, claim-scoped review rather than a systematic meta-analysis. The disciplinary focus is credit economics; tax legality, enforceability and insurance contracts enter as constraints on the economic relationship.

Historical formation of offshore bank credit

The expansion of eurocurrency banking made the geographic separation of money creation, funding, and lending an ordinary feature of international finance. A dollar deposit maintained at a banking office outside the United States represented a dollar-denominated liability of that office. A loan made from the resulting balance sheet could reach an enterprise in a third economy. The arrangement joined clients who wanted dollar assets with borrowers who required dollar finance. It also connected domestic financial institutions to wholesale markets whose effective capacity was much wider than their domestic deposit base.

The historical investigation by McCauley, McGuire and Wooldridge (2021), Seven decades of international banking, traces the transformation of eurocurrency banking through liberalization, financial innovation and crises. In the early market, regulatory differences and interbank dealing encouraged the concentration of foreign-currency transactions. Syndication subsequently enabled institutions to distribute individual loans among lenders and to finance borrowers that needed sums beyond the capacity or appetite of a single bank. Wider cross-border access and local foreign affiliates gradually changed the composition of international claims. These developments establish the historical contingency of any single definition of offshore credit: a narrow eurocurrency measure identifies a functional monetary segment, while a financial-center measure identifies the territorial geography of intermediation.

The fundamental economic development was a change in the feasible set of transactions. A bank could combine a currency funding source, specialist origination capability, legal entity and borrower that had previously been connected only through more restricted domestic arrangements. As competition intensified, the return on scarce banking capacity increasingly depended on global opportunities. The same mobility exposed borrowers to the withdrawal of credit from a lender whose own circumstances might have deteriorated in a distant economy.

Historical causation requires attention to simultaneous changes in exchange arrangements, capital-account policies, macroeconomic conditions and bank technology. Regulatory differences supplied incentives for offshore booking, but the persistence of established centers after liberalization also reflects financial infrastructure, expertise, repeated transactions and network scale. An institution that has assembled an experienced credit team and efficient settlement arrangements can retain a competitive advantage even when an original tax or reserve-rule difference narrows. This historical sequence therefore supports institutional explanations of location alongside price-based explanations.

Transaction geography and currency denomination

Four classifications are particularly useful. A residence-based cross-border loan has a bank office and debtor resident in different economies. Foreign-currency credit is denominated in a currency foreign to the debtor or, depending on the statistic, to the reporting office. A eurocurrency-market loan in the historical narrow sense involves a currency foreign to both the resident borrower and the lending office's location. A financial-center-based measure identifies lending from offices located in designated intermediating centers, using a stated and dated center definition.

Consider a Japanese-controlled bank with an office in London that lends dollars to a Brazilian manufacturing company. The lender's residence for the office-level statistic is the United Kingdom; the banking group is Japanese; the immediate debtor is Brazilian; and the debt is denominated in dollars. The currency area associated with the contractual unit of account is a fifth geographic dimension. The loan is a cross-border claim from the reporting office's perspective and is denominated in a currency foreign to both office and borrower. If an otherwise identical institution in Tokyo makes the loan, the residence dimension changes while the group nationality and debtor's payment obligation may remain the same.

A US office of the same bank making a dollar loan to a US resident provides an opposite example. The operation is local in residence and currency terms, although the bank is foreign-controlled. If a foreign-currency loan is granted by a domestic bank to a domestic customer, the transaction is local geographically but international in its currency characteristics. These distinctions determine eligibility for a statistical series. They are essential before testing any proposition about the effects of offshore lending.

Hardy, McGuire and von Peter (2024b) distinguish banking groups that concentrate their international activity at home from groups that operate through affiliates in their geography of bank operations. For the economist, the comparative unit should be the lending office when investigating booking decisions, the group when investigating capital allocation, and the borrower when investigating investment and repayment. A jurisdiction can be important under the first question and substantially less important under the second. A map drawn for one analytical purpose should not be treated as a map of every economic relationship.

Booking, origination and the organization of banking groups

The credit decision can be decomposed into screening, negotiation, approval, booking, administration and loss absorption. One office can introduce a customer and analyze its financial statements; another can enter into the loan; a treasury function can finance the resulting asset; and a centralized committee can authorize its risk limits. The banking group's internal organization determines which functions are assigned to a particular jurisdiction. A legally accountable entity carries a contractual obligation and must have resources to meet its obligations even when other group offices contribute services.

A branch generally forms part of the bank legal entity, whereas a subsidiary is constituted as a separate legal entity. The exact responsibilities, insolvency arrangements, regulatory treatment and practical availability of transfers depend on governing law, corporate structure and supervision. The relevant economic distinction is the extent to which liquidity and capital can move within the group under normal and stressed conditions. A group-funded office may provide substantial lending with limited outside liabilities. A locally financed subsidiary can build stable customer relationships and obtain a funding base that is less synchronized with the group's wholesale liabilities. The institutional patterns discussed by Hardy et al. (2024b) imply that identical asset volumes can accompany very different liquidity dependencies.

A booking center may host teams providing genuine financial services, offices concentrating treasury and risk management, or relatively limited local operations that record contracts negotiated elsewhere. The term identifies a function; it does not establish that every institution in the jurisdiction lacks employees, substance or economic value. The appropriate empirical evidence includes staffing and decision authority, legal balance sheets, inter-office positions, transactions with unrelated customers, and local expenditure. These variables are rarely present in the same aggregate data set.

The authorial inference is that a change in booking share has an indeterminate welfare meaning until the underlying functions and constraints are identified. Moving an existing loan from one affiliate to another can alter local statistical positions with no contemporaneous change in the borrower's investment or cash flows. Moving credit origination to a more efficient specialist team can also reduce costs and increase the set of viable projects. The two cases require different evidence.

Balance sheets and the credit creation process

Bank credit is a balance-sheet relationship. In a simplified transaction, a bank grants a loan of 100 monetary units by crediting the borrower's deposit account with 100. Its assets and liabilities rise equally. When the borrower transfers the deposit to a seller at another bank, the lending bank must settle the payment, using its available settlement assets, borrowing or an internal funding transfer. Loan creation and the mobilization of settlement liquidity are related events with distinct constraints. There is no requirement that a specifically identified prior deposit be assigned to each loan, but the bank must be able to fund its resulting balance sheet and meet prudential requirements.

Suppose a parent office provides 100 units of inter-office funding to a branch, and the branch pays a borrower in another jurisdiction 100 units. After settlement, the branch has an external loan asset and an internal funding liability. The parent holds an internal claim and has transferred settlement assets or incurred a financing liability. In the consolidated group accounts the internal claim and liability cancel. The external loan remains. If the borrower then onlends all the funds to an unrelated enterprise, the resulting separate contractual claim belongs to that borrower's own balance sheet, not automatically to the banking group.

This example makes clear why the same 100 units of ultimate external bank credit can produce two cross-border claim entries when the internal funding crosses one border and the loan another. The 200-unit gross sum provides information about the network of settlement and funding obligations, but would overstate lending to the final customer if used as a unique volume measure. Consolidation removes the inter-office entry, but also removes information relevant to the distribution of liquidity across legal and geographic perimeters. The consolidation rules in CBS (BIS, n.d.) and the residence-based series should consequently be used together, without adding them.

For a lender the central feasibility conditions are creditworthiness, funding access, capital capacity, operational capability and compliance with the applicable laws. At the portfolio level, liquidity is affected by the timing, currency and location of cash flows. At the group level, losses consume common capital where the group bears the exposure. At the entity level, cash must be available to the institution that is legally required to make payment. The economic analysis of offshore lending must preserve all three levels, particularly when stress makes internal transfers more difficult.

Funding costs, pricing and credit allocation

For a new loan, a bank compares expected revenue with expected funding, operating, credit-risk, hedging and capital costs. The nominal interest rate is only one component of the transaction. Fees, maturity, collateral requirements, commitment usage, prepayment options, contractual covenants and the ability of the borrower to draw funds change the effective economics. Denomination in an international currency can widen funding opportunities, yet the relevant funding cost is the marginal cost of obtaining liquidity in that currency at the required tenor.

A simple illustrative annual spread calculation starts with funding at 4.0 percent, expected credit losses of 0.6 percent, operating and hedging costs of 0.5 percent, and an allocated capital cost of 0.9 percent. A loan yield of 6.0 percent covers these assumed components, before other taxes, fees, and target profit. An additional 0.5 percent target margin raises the illustrative required yield to 6.5 percent. If funding rises to 5.0 percent and the other components stay constant, the required yield rises to 7.5 percent. These arithmetic results describe the assumed accounting decomposition, not a universal pricing formula or estimated lending rate.

In practice, expected losses depend on borrower quality, recovery prospects, maturity and the probability of credit deterioration. Unexpected losses may create capital requirements, while the loan's uncertain drawdown profile may create liquidity requirements. Hedging costs and asset-liability transfer prices can vary across currencies. Treating risk charges and funding charges as independent constants is useful for a static illustration but not for a stress exercise in which they change together. Internal transfer prices designed for individual desks also need reconciliation to the external cost of resources for the group.

The choice of a lending location may reduce operating, settlement or search costs; it may permit access to investors or bank specialists and can reflect differing regulations and tax systems. The benefits must be compared with additional compliance, monitoring, hedging and recovery costs. Pogliani et al. (2022) describe the role of tax and regulatory considerations in the evolution of cross-border centers. Their evidence concerns incentives and center specialization. It does not identify the exact marginal cost saved on every offshore loan. The authorial economic conclusion is that cheaper booking can lower the borrower's financing cost when competition transmits the saving, whereas market power may allow the intermediary to retain some of it.

Loan products, syndicates and the distribution of risk

Term loans provide a specified principal amount with contractual maturity and repayment provisions. Revolving facilities give the borrower contingent access to credit over an agreed period, while trade-related lending finances a particular operating transaction or working-capital need. Long-maturity project finance may rely on expected project cash flows and security arrangements. Sovereign credit depends on public repayment capacity, currency arrangements and applicable contractual remedies. The economic significance of the offshore location varies across these products.

A syndicated facility pools commitments of multiple lenders and usually places coordination functions with arranging institutions and agents. If a facility has a committed amount of 500 million dollars shared equally by five banks, the initial equal commitment is 100 million each. A 60 percent aggregate draw, if distributed proportionally, produces 60 million in funded principal for each bank, with 40 million undrawn commitment outstanding. The full 500 million facility amount must not be counted as new funded credit five times. Fees and contractual risk distributions can be unequal even if commitments are equal.

Loan participations, assignments, guarantees, credit protection and collateral may distribute pieces of the expected loss exposure. A sale in the secondary market changes who holds the asset without necessarily supplying additional money to the borrower. An undrawn commitment can create a future liquidity obligation during stress, precisely when borrowers are most likely to need cash. An investor buying a portion of a loan may bear credit risk without providing the original origination service.

The concentration of origination capability and diversification of ultimate credit exposures can coexist. For a borrower, that structure may facilitate access to a larger pool of capital. For a group of lenders, a shared contract can reduce the idiosyncratic exposure of each institution but also create a common position to the same obligor. The economics of diversification should therefore examine the correlations among borrower defaults and lender funding shocks rather than merely count the parties to the contract. Syndicated loan databases add important contract-level evidence but represent a selected market segment, often weighted toward larger borrowers.

Foreign currency, settlement and borrower vulnerability

A loan denominated in a leading international currency is attractive when the borrower earns revenues in that currency, purchases inputs in that currency, or faces a more liquid lending market than in its home currency. It can also present a substantial mismatch when the borrower receives local-currency revenues and promises a fixed foreign-currency payment. In the simplest illustrative case, a borrower owes 1 million dollars and the exchange rate moves from 10 to 12 units of domestic currency per dollar. The local-currency principal equivalent rises from 10 million to 12 million, a 20 percent increase, before interest, hedges and changes in receipts.

An exporter with dollar revenues may have a natural offset. A domestically oriented company may use a forward or swap to hedge contractual payments. The hedge introduces counterparties, settlement terms, collateral arrangements and possible renewal requirements. A borrower can be economically exposed even when accounting measures report limited short-run mismatch; its revenue contract might expire before the debt matures. Equally, a large foreign-currency liability held against durable earnings in the same currency is different from an unhedged liability of the same amount.

The lender may obtain currency liquidity by borrowing directly in that currency or using market instruments that transform funding from another currency. A bank's net open exchange position can thus be small while its need to renew funding in a particular currency remains material. A deterioration in FX swap liquidity can affect its marginal loan price and capacity even in the absence of a large direct exchange-rate loss.

Becker, Schmeling and Schrimpf (2024), in Global bank lending and exchange rates, report evidence from syndicated relationships over 2000–2021 using granular instrumental variables. Their summary associates a one-standard-deviation increase in net US dollar lending flows with a 36 basis point appreciation of the dollar and describes stronger effects when intermediary funding is constrained. The sample and identification assumptions govern this result; it is not an exchange-rate elasticity transferable to individual loans. The mechanism nevertheless illustrates a crucial economic feedback: credit demand in a foreign currency stimulates bank funding demand, while shifts in funding prices and exchange rates influence loan supply and debt-service capacity.

Currency selection also has a geographic and sectoral structure. Emter et al. (2024) use bilateral banking positions in their study of the determinants of loan currency. Their results associate loan denomination with trade invoicing and document a more regional euro market and a more global dollar market. They identify the Cayman Islands' nonbank financial sector as important in dollar lending and the United Kingdom as a key location in euro lending. These associations concern the structure of observed bank-credit relationships; they do not establish that every individual transaction in a financial center is tax-driven.

The currency of income, the currency of reported assets, and the currency of repayment can differ. Consider a hypothetical exporter with one-year contracted dollar revenue of USD 4 million, dollar input costs of USD 1 million, and dollar principal plus interest due of USD 2.5 million. Its contracted net dollar operating receipts cover the payment 1.20 times. If dollar revenue declines by 20 percent while input costs remain fixed, net receipts become USD 2.2 million and coverage falls to 0.88. A stable currency denomination cannot protect against quantity, counterparty and timing shocks. A borrowing assessment should therefore attend to the joint distribution of operating receipts and debt obligations, even where the accounting currency exposure is small.

For a lender choosing among otherwise comparable dollar and euro credit assets, the relevant comparison is the marginal risk-adjusted return after currency-specific funding, hedging and capital costs. A lower euro coupon can be consistent with a higher expected return to a euro-funded institution. Conversely, a dollar-funded borrower with dollar receipts may accept a higher nominal coupon to reduce mismatch. These are analytical implications of the transaction's cash-flow structure, not predictions about market rates.

Country risk and the allocation of loss

Credit risk concerns the possibility of failure to meet an obligation and the associated recovery shortfall. Country risk encompasses events and policies in the jurisdiction affecting a portfolio of borrowers or connected assets. Transfer risk concerns impediments to remitting funds needed for repayment, potentially even when the borrower has adequate resources in local currency. These categories overlap in crisis circumstances, but a banking group can be exposed to country risk through local activity as well as through cross-border loans.

An offshore borrower could be a financial affiliate of an enterprise headquartered elsewhere. If the liability is guaranteed by the parent, a risk assessment must consider the terms, enforceability and creditworthiness of the guarantor. The immediate debtor's residence remains relevant to its own payment capacity and applicable legal processes. The country attributed to the exposure after recognized risk transfers answers a different question. The BIS study of country risk by Hardy et al. (2024a) explains how consolidated positions and additional exposure categories help identify such vulnerabilities.

In a hypothetical 100-unit claim on a corporate affiliate, a recognized guarantee from a parent covering 70 units may justify attributing 70 units to the guarantor's country and 30 units to the direct debtor's country, subject to the relevant reporting and guarantee rules. The nominal loan principal remains 100 units. Reallocation changes the attributed country exposure without creating additional credit. If the guarantor fails at the same time as the borrower, the value of the protection may be lower than expected. This example is a conceptual allocation, not a reporting instruction for partial guarantees under a specific national regime.

Market prices can respond to perceived risks well before contractual default. A fall in collateral value changes potential recovery, while exchange controls can interrupt payment despite unchanged production capacity. Risk transfers can be valuable because they diversify counterparty exposure or strengthen recovery, yet their effectiveness depends on the state of the world in which they are called upon. An empirical comparison of immediate and guarantor-basis statistics must not treat the difference as a direct observation of losses avoided. It is a measure of attributed exposure under stated reporting conventions.

Measuring stocks, flows and loan commitments

The BIS locational banking statistics organize international balance sheets by reporting location, sector, currency and other available dimensions. They are suitable for studying networks of inter-office and interbank claims, geographic intermediation and denominational structure. The official LBS reporting metadata identifies coverage, quarterly frequency and reporting breaks. Researchers should select an explicit instrument measure, distinguishing total claims, credit and loans/deposits where available; a series covering bank claims can include instruments beyond the loan product of interest.

Consolidated statistics remove intragroup positions and organize the global external positions of a reporting banking group by the country of its controlling parent. The CBS methodological overview describes immediate counterparty and guarantor bases. Consolidated foreign claims can incorporate locally booked positions as well as direct cross-border claims; a foreign bank's locally financed local-currency loan therefore differs from an international loan originated abroad. A cross-border-only question must use the corresponding component, while a global credit-exposure question may require the broader aggregate.

The change in a dollar-valued stock between two quarters is not necessarily the net amount newly lent. In a simple case, a fixed balance of 100 euro units translates into 110 dollar units at one reporting date and 120 at the next. The dollar stock rises by 10 even though the euro principal has not changed. Recognition of impairments, loan sales, writeoffs, reclassifications and changes in reporting population can further alter measured positions. The appropriate adjusted change series or currency-specific reconciliation distinguishes valuation from transaction effects as far as the source permits.

Origination records have different units. A signed commitment is neither an amount necessarily drawn nor an outstanding balance at a later reporting date. A refinancing can replace an old loan with a new one without increasing total funded credit. When estimating net flows, an analyst must account for repayment, sale, writeoff and any methodological adjustments. When estimating access to finance, new commitments, maturity, collateral and rejected applications may be closer to the scientific object. An aggregate series cannot identify application-level exclusion or counterfactual supply.

A reproducible empirical workflow records the data release vintage, reporting locations, creditor sector, counterparty sector and residence, currency, instrument, position type, consolidation basis and treatment of missing or suppressed values. It states whether the sample includes intragroup claims and how revisions are handled. The selected denominator must match the numerator's instrument and reporting population. Combining claims from one source with denominators from another without aligning these definitions introduces a potentially large error.

A distinction is needed between the statistical term credit and the individual loan contract. In BIS locational banking statistics, bank credit comprises loan-and-deposit claims plus holdings of debt securities; total claims additionally include derivatives with positive market value and other residual instruments. Interbank deposits are claims of one bank on another. These asset-side observations differ from deposits held by a bank as liabilities to depositors. The instrument filter required for an exclusively loan-focused study is narrower than the credit aggregate used in the statistical commentary. Even the loans-and-deposits category cannot always isolate conventional term loans or identify recipients of onward finance.

The consolidated data answer a different question. International claims comprise cross-border claims and local claims in a foreign currency; foreign claims additionally include local claims in the currency of the host economy. Both depend on the reporting group and counterparty country. Consequently, a study of local credit supplied by foreign banks can contain ordinary domestic-currency loans that the institutional definition of offshore lending does not automatically include. The group-nationality and office-residence views should be retained as distinct data products rather than combined into a single volume measure.

Contemporary statistical benchmarks

Quarterly statistical releases permit a dated comparison of the scale of cross-border credit with broader international bank claims. The BIS (2026a) end-March 2026 statistical release reported outstanding cross-border claims of USD 47.6 trillion and credit of USD 39.5 trillion. The latter grew by 11 percent year on year, with an exchange-rate- and break-adjusted increase of USD 1.7 trillion over the quarter. In the same release, cross-border credit to nonbank financial institutions rose by USD 651 billion. These figures concern the full reporting universe of cross-border banks and are not an estimate of lending through offshore affiliates alone.

The BIS (2026b) end-June 2026 statistical release, dated October 29 and available in a copy hosted by the Austrian banking association, reported cross-border credit of USD 40.7 trillion and annual growth of 11 percent. The reported adjusted quarterly credit increase was USD 877 billion, consisting of USD 560 billion in loans and deposits and USD 317 billion in securities holdings. The original BIS listing corroborates the date and headline stock and growth; the PDF copy supplies the component amounts. Published releases may differ in revision vintages.

The last two values yield a simple authorial calculation: loans and deposits represented approximately 63.9 percent of that quarter's adjusted bank-credit increase (560 divided by 877), while debt securities represented about 36.1 percent. These shares describe the instrument composition of the change, not the proportion of offshore banking assets held in loans, the stock composition, or the proportion lent to productive enterprise. The difference between reported credit stocks of USD 39.5 trillion and USD 40.7 trillion must also not be read as the adjusted transaction flow. The releases' rounded current-dollar outstanding amounts, currency conversion, statistical adjustments and possible revisions enter that comparison.

The cross-border distribution of lending calls for a further decomposition into booking office residence, bank nationality and counterparty location. An analyst who applies a financial-center list to office-residence data estimates a type of geographic intermediation. An analyst who identifies branches and subsidiaries of groups outside their home economies estimates a distinct institutional population. Both selections can be useful; the denominator, quarter, instrument and sector must be identical before shares are compared. In particular, the paper's affiliate share discussed below is an estimate for its own sample and definition, and cannot be multiplied mechanically by the latest bank-credit stock to derive a current worldwide offshore loan estimate.

A credible empirical extension would join the locational table's currency, instrument and sector variables with available nationality information, preserve stocks and adjusted changes separately and report the overlap between the two offshore definitions. Suppressed values and categories not jointly available must remain explicit. The present dossier presents checked published observations and transparent arithmetic; it does not claim to have downloaded or reproduced the underlying confidential reporting microdata.

Financial centers and the economic geography of lending

Financial centers concentrate particular functions because repeated transactions reduce search costs, skilled intermediaries become available, market infrastructure deepens, and banks can serve clients across economic regions. Specialized jurisdictions may facilitate particular products or forms of corporate organization. The same physical concentration can generate geographic diversification of final borrowers. Neither the size of the local economy nor the location of an office alone identifies the scale of final economic activity financed by that office.

Pogliani, von Peter and Wooldridge (2022) examine cross-border financial centers as intermediaries with unusually large external financial activity relative to domestic economic scale. A center's liabilities and claims have their own geographic distributions. Asymmetric distributions provide evidence of intermediation across economic regions rather than a simple picture of local saving funding local investment. The measured population, sectors and thresholds are properties of the study's classification.

The March 2026 study by Aldasoro et al., offshore activity in BIS statistics, provides a particularly relevant example. Its discussion of Graph 7 reports that less than 60 percent of bank cross-border claims are extended from home offices, with foreign affiliates accounting for the remainder. It also identifies substantial credit to China and Brazil from offshore offices of their own national banking groups. These are aggregate findings in that study's statistical universe; the affiliate share is not a global total of tax-haven loans, and own-nationality lending through foreign affiliates does not establish the source of the group's underlying funding.

These observations motivate a systematic distinction between inward foreign credit and inward credit from abroad. A borrower can receive funding from a bank office abroad whose controlling group is domestic. Another borrower can be financed locally by a bank whose parent is foreign. The domestic funding base, the ownership of the banking institution and the legal location of the facility then yield different policy implications. A creditor-location concentration measure would assign importance to financial centers; a lender-nationality concentration measure can reveal dependence on a much smaller number of banking systems.

When consolidation is impossible across nonbank intermediaries, a complete mapping requires separate evidence. An analyst should resist replacing unobserved ownership or use of proceeds with a geographic guess. The correct conclusion from the presence of offshore centers in a credit network is that intermediary location matters for interpreting statistical data and exposure; the identity of the ultimate source of saving or recipient of expenditure remains a separate empirical question.

Direct credit, indirect credit and nonbank linkages

A borrower can access international bank credit directly by obtaining a loan from an office abroad. Alternatively, a resident bank may borrow externally and then finance a local enterprise. In the second case the enterprise has a local liability but its creditor depends in part on external finance. Ehlers, Hardy and McGuire (2025), in their international dimensions of credit study, distinguish the origin of funding, denomination and creditor structure. This approach is vital for detecting channels that a simple enterprise external-debt ratio can miss.

Consider a foreign bank lends 100 to a resident bank, which lends 90 to enterprises and retains 10 as liquid assets. Direct bank-to-enterprise cross-border credit is zero in this stylized transaction. The resident banking sector has a 100-unit external loan liability, while enterprises receive 90 in domestic bank credit. Counting both claims as 190 units of final borrower financing confuses sequential intermediation with an expansion of final resources. However, disregarding the first claim would obscure the foreign lender's ability to influence domestic credit through its renewal decision.

Bank lending to nonbank financial institutions introduces comparable layers. A finance company can borrow from a bank and lend to households; a fund can borrow to finance securities positions; a broker may use credit for settlement. Each creates an asset and a liability with different liquidity and valuation characteristics. Aldasoro et al. (2026) emphasize the limitations of aggregate bank data in identifying the nationality or underlying operations of nonbank counterparties. The statistical nonbank category should never be equated exclusively with final nonfinancial borrowers or entirely with shadow credit.

The policy relevance depends on the destination and terms of funds. An expansion of bank-to-nonbank credit may support risk distribution and market liquidity. It may also increase leverage in portfolios whose rapid collateral demands or asset sales transmit shocks back to banks. Supervisors need measures of credit lines, collateral concentration, maturities and connected exposures. Economists studying production need a link between finance and real investment rather than assume that all lending to an intermediary finances fixed capital formation.

International credit cycles and the transmission of shocks

International loan supply can respond to borrower-country conditions, deterioration in the lending group, changes in the booking jurisdiction, or shifts in the global currency market. A lender may reduce new transactions because its expected recoveries have worsened; it may also withdraw from healthy borrowers when its own capital or liquidity becomes scarce. Observed lending contraction therefore combines changes in credit demand, bank supply, contractual maturity and reallocation among borrowers and offices.

Cetorelli and Goldberg (2010), in Global banks and international shock transmission, study emerging-market credit during the 2007–09 crisis. Their staff-report abstract identifies direct international lending, foreign-affiliate domestic lending and external funding of locally owned banks as distinct transmission channels. Their findings help explain why a shock to the lender can affect enterprises several transactions removed from that lender. The study's cross-market crisis evidence supports the existence of the channels, with magnitudes contingent on the studied systems and episode.

The syndicated market additionally permits examination of changes in the portfolio location of newly originated loans. Giannetti and Laeven (2012), in their flight home effect study, report an approximate 20 percent increase in the home bias of originations when the lender experiences a banking crisis in its country of origin. This is a relative home-bias result in their sample, not a 20 percentage point change in all foreign credit or a prediction for a particular bank. Its significance is that the geography of credit can change even when foreign borrowers have comparable quality.

The historical sequence of bank credit cycles includes the Latin American debt crisis, the Asian financial crisis and the global financial crisis. Different combinations of sovereign exposure, short-term external borrowing, exchange arrangements and wholesale funding shaped these episodes. Contraction across multiple jurisdictions can produce synchronized reductions in production and investment. The comparative lesson is to investigate which contractual exposures were short-dated, which lenders faced common funding shocks, and which borrowers could substitute alternative funding.

A locally funded affiliate may continue to lend when its parent reduces direct cross-border lending, provided it retains independent liquidity, adequate capital and authority. A parent facing binding group capital constraints may still reduce the affiliate's risk budget. Conversely, a parent with ample liquidity may support an affiliate in a localized crisis. The direction of internal resource transfer is conditional on the distribution of shocks and restrictions. Offshore credit should therefore be studied through the topology of the banking group rather than presumed either universally unstable or universally stabilizing.

The strength of a lending relationship moderates withdrawal. De Haas and van Horen (2011) examine the behavior of major international banks in Running for the Exit. Their working-paper evidence associates continued crisis lending with borrower-country familiarity, repeat borrowers and local co-lending networks. This provides a useful qualification to aggregate homeward reallocation: two banks exposed to the same global shock may differ in how much information and organizational capacity they have accumulated in a destination market.

A hypothetical credit line illustrates the role of substitution. A manufacturer has a USD 12 million offshore line, of which USD 8 million is drawn; its usable domestic credit lines total USD 3 million. If the offshore commitment is reduced to USD 8 million at renewal, funded debt does not immediately change, but the manufacturer loses USD 4 million of contingent financing capacity. Domestic credit can restore at most USD 3 million of usable capacity if still fully available; the residual shortfall is USD 1 million. A stock-based indicator could show stability even while the enterprise's capacity to expand or absorb an interruption deteriorates. This is a conditional scenario, not an observed loss or a forecast.

A further distinction concerns maturity. A short-term refinancing freeze may affect investment quickly if the borrower depends on continuously renewed finance, while a longer-duration loan with no immediate principal repayment can temporarily absorb the same funding shock. An undrawn commitment made by a stressed bank might not provide the same reliable liquidity as cash already held. Maturity, contractual cancellation provisions, collateral and alternate credit capacity therefore belong in the analysis of crisis exposure alongside nominal outstanding amounts.

Empirical identification and research design

A causal analysis of offshore credit must establish what would have happened without a particular funding shock, policy change or booking arrangement. Country-level correlations are informative about aggregate comovement, but they rarely isolate supply. A reduction in foreign bank lending during recession may reflect fewer profitable projects, a conscious bank adjustment, the mechanical expiration of commitments, or all three.

A loan-level study could compare borrowing from differently exposed lenders by enterprises borrowing from more than one institution. Borrower-by-period fixed effects absorb shocks common to those relationships, conditional on measurement quality and sample selection. Lender exposure to an externally timed shock can then be related to changes in the credit offered to the same borrower. Such a design requires credible differential exposure, adequate pre-shock observations, and attention to endogenous bank matching, changes in relationship composition and sample survivorship. Multi-bank firms may differ from firms reliant on one bank, which limits external validity.

A second design compares the same lender's offices across jurisdictions around a home or host policy change. A group-wide exposure measure and a location-specific regulation measure can be distinguished; borrower demand must be addressed separately. Policy changes can respond to the very credit growth being studied, so simple before-and-after estimates can be biased. Event studies should assess pretrends and anticipation, and consider heterogeneous treatment timing. An instrument must have a defensible exclusion restriction, not merely predict the loan variable.

Avdjiev et al. (2020), in Home sweet host, decompose lending growth into home, host and common components to investigate prudential spillovers. Their reported results vary with the type and origin of policy and with US monetary conditions. The important methodological consequence is that merging bank nationality with reporting residence can obscure the independent variation necessary to assess policy channels.

An additional possible design uses a sharp change in booking location within a stable group and compares loans to similar borrowers before and after. It could reveal a change in funding, spreads or maturities associated with the new organizational arrangement. But booking choice is endogenous: a group may choose relocation because its preexisting portfolio, funding costs or regulatory position has changed. A credible comparison requires either an externally determined reorganization or strong supplementary identification evidence. The present monograph proposes these as research strategies and derives interpretation conditions; it reports no original regression estimates.

Empirical publication should provide a data dictionary, preprocessing rules, code, sample exclusions, confidence intervals, and sensitivity to sector and jurisdiction classifications. Statistical significance by itself does not quantify welfare or systemic importance. The policy question is whether the change affects credit available to borrowers, the allocation of risk, or the expected cost of financial distress. These are separate outcomes.

Liquidity management, capital and banking supervision

Banking groups coordinate credit decisions under capital and liquidity constraints that can apply at multiple levels. The group may have sufficient aggregate capital and cash while a particular legal entity faces a shortfall. Assets denominated in one currency may be difficult to turn into cash in another during stress. The appropriate risk analysis therefore maps available liquid resources by location, currency, legal entity and settlement date, as well as modeling consolidated losses.

The Basel Committee's revised Core Principles for effective banking supervision (BCBS, 2024) establish a global supervisory baseline. Their application depends on national legal and supervisory frameworks. The principles support the economic objective of observing and managing the risks arising from geographically distributed operations. The relevant measures must prevent blind spots between a home supervisor concentrating on consolidated exposures and a host supervisor charged with the stability of institutions in its jurisdiction.

The Liquidity Coverage Ratio standard (BCBS, 2013) specifies a short-term stress horizon of 30 calendar days, supported by high-quality liquid assets. Its normal-times minimum is distinguished from the intended use of liquid buffers in stress. A hypothetical bank with 120 units of eligible liquid assets and 100 units of prescribed net outflows has a ratio of 120 percent under those assumed values. If usable resources at a particular office fall to 70 while the local liabilities become due, the group ratio alone does not resolve the local settlement problem. The example does not reproduce the regulatory calculation, which requires detailed definitions, caps, haircuts and applicable scope.

Capital requirements address loss absorption and risk-taking incentives. Liquidity regulation addresses the need to meet obligations when due. A foreign-currency asset can meet a bank's investment objective while increasing currency-specific refinancing needs. A loan with sound expected profitability can become unviable to originate when incremental capital usage or currency funding costs exceed the return. A bank able to increase loan margins may pass some of the increase to the borrower, while clients with few substitutes may bear a larger adjustment.

Home-country and host-country policies can interact through internal booking and lending decisions. A group-wide constraint can leave relatively little room for reallocation among affiliates; a location-specific constraint may change where the group books business. Whether this redistribution reduces consolidated risk depends on the post-change risk exposures and funding arrangements. Evaluation requires evidence about group behavior, rather than the local volume reduction alone.

Central bank liquidity and public stabilization

Foreign-currency liquidity problems expose a structural difference between the currency of a contractual promise and the jurisdiction in which the promising intermediary operates. A local central bank can ordinarily provide its own currency subject to its legal framework and operations. It does not independently issue another central bank's money. Currency liquidity can instead come from reserve assets, market transactions, and authorized international arrangements.

The October 31, 2013 announcement of standing bilateral swap arrangements by the Federal Reserve and five other central banks established an enduring network intended to supply foreign-currency liquidity under agreed conditions. The Federal Reserve's central bank liquidity swaps page (2023) also identifies the role of dollar facilities in limiting strains in credit provision. The arrangements are between central banks; downstream access depends on the facilities operated by the relevant authority. An unsecured borrower does not automatically obtain the benefit as a contractual entitlement.

Liquidity support can interrupt the feedback in which funding scarcity induces forced credit contraction or asset sales, leading to additional funding losses. The economic justification is the avoidance of distortions caused by a general shortage of settlement liquidity rather than the preservation of every risky investment. The distinction between liquidity and solvency is crucial: a facility that enables timely settlement does not restore the value of a loan whose borrower is unable to repay.

Public backstops also shape institutional incentives. An intermediary that can expect emergency liquidity may undertake longer or less liquid commitments, which can support useful maturity transformation but can create an incentive to economize excessively on private buffers. Institutional design therefore considers eligibility, collateral, price, supervision, and the distribution of losses. The global arrangement remains differentiated across countries and currencies. Countries or institutions lacking direct access to a relevant backstop can face tighter conditions in a global shortage. This asymmetry may affect the pricing and geographic composition of lending even in normal periods.

Development, competition and distributive consequences

International bank lending can expand investment opportunities where domestic financial markets have limited capacity. A firm may use an offshore loan to purchase equipment, finance exports or build a project that requires longer maturities or larger amounts than local lenders offer. The introduction of foreign bank competition may improve services, lower financing costs and disseminate expertise. These mechanisms depend on borrower eligibility, productive opportunities, financial infrastructure and contract terms.

The benefits can be unevenly distributed. Large companies with external earnings, transparent financial reporting, recognized collateral and specialist advice may obtain credit on terms unavailable to smaller domestic firms. Local intermediaries may be better at collecting information about small clients, even when they have a less favorable funding cost. A cross-border expansion can consequently finance internationally connected enterprises while leaving credit-constrained domestic enterprises relatively unchanged. A welfare evaluation requires information about the marginal borrower admitted to credit and the projects actually undertaken.

The cost of financing should include fees, exchange-risk management, contingent collateral demands, and the cost of distress. A lower headline coupon does not establish cheaper economic financing if its denomination and maturity introduce additional risk. The correct comparison is with feasible alternatives for the same borrower and project. Differences in borrowing cost can reflect productivity-enhancing efficiency, credit selection or implicit support. Observational comparisons of interest-rate averages can therefore exaggerate or understate the benefits of access.

The host jurisdiction can earn fees, wages and tax receipts from supporting the lending process; the borrower jurisdiction can gain productive capacity; and the creditor group's home economy can receive profits and bear part of the losses. The relative distribution is a question for incidence analysis. A rapid expansion in external credit can also increase asset-price sensitivity, foreign-currency indebtedness and crisis-related output costs. The economic gain from more resources today must be assessed alongside the obligations created for tomorrow.

A sound international credit assessment follows the use of funds, the contractual allocation of risks and the capacity of ultimate borrowers to repay. Transactions that relocate existing liabilities without supporting additional economic activity require a different interpretation from genuinely incremental financing. Evidence of a large gross position in a small jurisdiction establishes the scale of recorded transactions, not a measure of local welfare produced by them.

Published evidence supports the relevance of the macroeconomic channel. Aldasoro et al. (2020), in international lending and emerging markets, investigate 22 emerging-market economies from the first quarter of 1990 through the fourth quarter of 2018 using granular instruments. Their reported results connect positive cross-border bank-credit shocks with easier financial conditions, domestic credit growth, economic activity and higher housing prices, and show weaker estimated responses in economies with greater restrictions on capital inflows. These findings apply to cross-border bank credit in the specified sample. They neither isolate every offshore booking arrangement nor establish a single welfare effect of inflow restrictions.

The dual implication is important. If additional credit supports productive investment, its direct economic benefit can be substantial. If the same expansion is accompanied by leverage and rising collateral values, increasing sensitivity to future reversal can alter its intertemporal consequences. Observed output growth and observed housing-price appreciation should be treated as distinct outcomes, not interchangeable measures of social benefit. An adequate welfare study combines production, financial access, distributional outcomes and expected crisis costs rather than assuming that a larger financing stock implies a superior allocation.

Compliance, institutions and the boundaries of economic inference

Offshore bank lending may occur under diverse regulatory, tax and reporting regimes. The price and availability of credit respond to differences in legal certainty, administrative costs, transparency, collateral enforcement and the cost of compliance. These institutional variables affect the value of transactions and the risk of losses, and can influence which locations attract particular functions.

Economics can compare the incentives and outcomes of institutional arrangements, while detailed determination of legality requires applicable legislation, supervisory rules, case law and specific facts. A cross-border loan is a standard instrument of international finance. Its offshore classification alone does not determine whether the creditor or borrower is complying with tax, reporting, anti-money-laundering or sanctions obligations. The assessment of individual conduct requires independent evidence and due process. This monograph offers no advice on bypassing obligations, hiding control or frustrating oversight.

From the perspective of a legitimate lender, opacity about the counterparty's identity, sources of repayment and contractual support makes credit assessment more expensive and uncertain. Better verification of borrower finances can reduce information asymmetry, although it also entails administrative costs. A credible information framework can benefit the market by enabling comparable risk assessment and reducing doubts about the counterparties with which a bank deals. The distribution of the costs and benefits depends on the scope and design of the obligations.

Statistical confidentiality and public disclosure are related but distinct. Granular supervisory data can be necessary for risk management without being appropriate for unrestricted publication. Aggregate statistical disclosure may protect individual borrowers but create difficulty in identifying highly concentrated exposures. Researchers should document suppressed data and avoid imputing secret activities simply because a statistical cell is unavailable. The significance of incomplete information is the resultant uncertainty, rather than proof of misconduct.

Comparative loan economics and stress analysis

The allocation of a loan through a financial center can change costs without changing the borrower's underlying project. The relevant counterfactual is the loan feasible under another institutional arrangement, with comparable maturity, contingent provisions, contractual risk and currency exposure. If the alternatives have different cash-flow structures, an apparent pricing advantage cannot be interpreted independently of that structure.

Consider a hypothetical firm with a one-year domestic-currency loan offer at an annual interest rate of 9 percent and a one-year dollar offer at 5 percent. Assume both have the same principal equivalent at the initial exchange rate, no fees and bullet repayment. If the domestic currency depreciates by 7 percent against the dollar, the unhedged dollar loan's domestic-currency repayment is 1.05 multiplied by 1.07, or 1.1235 times the original domestic-currency principal. Its realized domestic-currency funding cost is 12.35 percent, exceeding the domestic alternative by 3.35 percentage points. Its break-even depreciation against the domestic loan is approximately 3.81 percent (1.09 divided by 1.05, less one). This comparison omits taxes, risk adjustments, forward premiums, transaction costs and differences in default remedies. It illustrates why the lower stated interest rate is insufficient evidence of lower economic financing cost.

A separate hypothetical lender has a portfolio of USD 100 million, expected annual interest income of USD 6 million and marginal funding expense of USD 3.5 million. Operating and hedging costs total USD 1 million. The annual pre-credit-loss contribution is USD 1.5 million. Suppose expected loss under a one-period approximation is exposure multiplied by probability of default and loss given default. With a probability of 2 percent and loss given default of 40 percent, expected loss is USD 0.8 million; the illustrative expected contribution is USD 0.7 million. A stress scenario with a 5 percent probability and 50 percent severity has an expected loss of USD 2.5 million; holding other terms fixed, expected contribution becomes negative USD 1 million. The calculation does not model unexpected loss, regulatory capital, correlation, discounting or a full lifetime provision. It shows how changes in the economics of credit risk can dominate the apparent advantage of a booking location.

Liquidity stress can arise even while this portfolio remains solvent on an expected-value basis. Assume a bank office owes USD 30 million over the next thirty days and can reliably mobilize USD 22 million in the relevant currency during that interval. It faces an illustrative USD 8 million cash gap, subject to the timing and certainty of receipts and obligations. An affiliate's surplus in another currency or country does not remove the gap unless the bank can lawfully and operationally transform and transfer those funds in time. The example is a currency-and-location-specific cash-flow analysis, not a regulatory liquidity coverage ratio calculation.

The cases establish an analytical ordering for project evaluation. Examine the feasibility of borrower payments in the relevant currency, the lender's cash-flow capacity and the allocation of expected and unexpected loss. Next compare the economic contribution of the additional loans with the contribution of alternative arrangements. Finally assess system spillovers arising from correlated borrower shocks, concentrated funding exposures and simultaneous loan withdrawals. The ordering applies established banking-economic distinctions to offshore credit; it is not presented as a newly discovered theoretical system.

In a comparative empirical design, costs and benefits should be attributed to the actors who receive or incur them. Lower spreads accrue initially to borrowers, service income to intermediaries, and tax revenues according to applicable jurisdictional rules. Expected distress losses may fall on equity holders, contract counterparties, insured or uninsured creditors, borrowers and, in particular institutional settings, public balance sheets. The identity and magnitude of these effects are conditional on actual contracts and institutions. They cannot be inferred from the lender's booking address alone.

An additional question concerns whether the apparent benefit reflects genuine new financing. Rebooking a loan between two offices of the same consolidated banking group, with no change to the borrower's principal, pricing, duration or enforceable rights, does not automatically expand the borrower's investment possibilities. A refinancing that lowers the expected financing cost may create an economic benefit without increasing principal, and a commitment extension can improve financing resilience without immediately altering funded balances. These cases explain why the appropriate dependent variable in offshore credit research is determined by the question being asked rather than by the availability of a single aggregate series.

Integrated analytical synthesis

Offshore bank lending simultaneously rearranges economic capacity and the visibility of financial relationships. The preceding analysis supports four connected conclusions. First, the place where a contract is recorded is an economically meaningful piece of information, especially for understanding settlement, administrative activity and local supervision. It acquires explanatory power when situated alongside group control, currency and borrower identity.

Second, a group may reallocate lending geographically without increasing global external credit. An office-level expansion can reflect a booking change or an inter-office funding leg. Conversely, consolidated credit expansion may occur with relatively little change in a particular center's position. Statistics of the appropriate scope reveal both dynamics when they are read together.

Third, exposure and liquidity have different geographies. A loan loss can reach the owners of a controlling institution, while a settlement shortage arises at the office or legal entity that owes a payment. A parent's guarantee may alter attributed country risk but leave the economic location of production unchanged. The conditions for the repayment of credit, for the absorption of losses and for the timely settlement of payments require separate analysis.

Fourth, the contribution of offshore lending to economic welfare depends on additional financing, resource allocation and stability. A lower cost of financing can support productive investment, but a rapid expansion of vulnerable credit may impose costs on borrowers, lenders and the wider economy. The distribution of those consequences follows balance sheets, institutions and contractual relationships rather than a single territorial classification.

Taken together, these conclusions supply an authorial synthesis of the established economic and statistical literature. They are operational distinctions for disciplined research; the monograph does not assert a new named theory. The appropriate next empirical step is to link disaggregated banking exposures with borrower-level outcomes while preserving legitimate confidentiality and accounting for endogenous portfolio allocation. International statistics provide the starting infrastructure, whereas causal interpretation and evaluation require appropriately designed studies.

Conclusion

Offshore bank lending is a major form of international financial intermediation whose defining economic relationships extend across the geographical boundaries of individual bank offices. The study has traced its historical formation, institutional organization, contractual structure, funding requirements, quantitative measurement and transmission of risk. Its central finding is that the location of intermediation and the allocation of the resulting credit exposure must be observed jointly. A coherent assessment also identifies the currency of cash flows, the location of repayment capacity and the extent to which intermediaries depend on external or intragroup liquidity.

The historical evidence demonstrates both the expansion of access to credit and the transformation of the ways international banking responds to financial shocks. The statistical evidence establishes that gross transactions cannot be interpreted as unique final-borrower financing and that nationality-based and residence-based series answer different questions. Published empirical work gives support to lender-side shock transmission and the interaction between bank credit, currency funding and regulation. The scope of each result remains that of its original research design.

The economic assessment of international credit should proceed from the additional resources supplied to borrowers, the costs and maturity of those resources, the capacity to service debt, and the distribution of expected and unexpected losses. Supervisory and stabilization measures matter because the economic consequences of loan origination are not confined to the institutions and countries represented on one contract. The analytical contribution of this dossier is the integration of these perspectives into a consistent account of the economics of offshore bank credit. It establishes a basis for further comparative and causal research while distinguishing institutional description, documented findings, illustrative calculations and the author's inferences.

References

  • Aldasoro, I., Beltrán, P., Grinberg, F., and Mancini-Griffoli, T. (2020). The macro-financial effects of international bank lending on emerging markets. BIS Working Papers No. 899, November 6. Research record
  • Aldasoro, I., Hardy, B., von Peter, G., and Wooldridge, P. (2026). International finance through the lens of BIS statistics: offshore activity. BIS Quarterly Review, March 16, pp. 81–96. Research record
  • Avdjiev, S., Hardy, B., McGuire, P., and von Peter, G. (2020). Home sweet host: Prudential and monetary policy spillovers through global banks. BIS Working Papers No. 853, April 6. Research record
  • Bank for International Settlements (BIS). (2026a). Statistical release: BIS international banking statistics and global liquidity indicators at end-March 2026. July 31, pp. 1–3. Research record
  • Bank for International Settlements (BIS). (2026b). Statistical release: BIS international banking statistics and global liquidity indicators at end-June 2026. October 29, p. 1. Copy hosted by Bankenverband Österreich. Research record
  • Bank for International Settlements (BIS). (n.d.). Consolidated banking statistics: Overview. BIS Data Portal. Accessed November 18, 2026. Research record
  • Bank for International Settlements (BIS). (n.d.). Locational banking statistics: Overview. BIS Data Portal. Accessed November 18, 2026. Research record
  • Basel Committee on Banking Supervision (BCBS). (2013). Basel III: The Liquidity Coverage Ratio and liquidity risk monitoring tools. Bank for International Settlements, January 7. Research record
  • Basel Committee on Banking Supervision (BCBS). (2024). Core principles for effective banking supervision. Bank for International Settlements, April 25. Research record
  • Becker, J., Schmeling, M., and Schrimpf, A. (2024). Global bank lending and exchange rates. BIS Working Papers No. 1161, January 15. Research record
  • Board of Governors of the Federal Reserve System. (2013). Federal Reserve and other central banks convert temporary bilateral liquidity swap arrangements to standing arrangements. Press release, October 31. Research record
  • Board of Governors of the Federal Reserve System. (2023). Central bank liquidity swaps. Policy tools page, updated March 19. Research record
  • Cetorelli, N., and Goldberg, L. S. (2010). Global banks and international shock transmission: Evidence from the crisis. Federal Reserve Bank of New York Staff Reports No. 446, May. Research record
  • De Haas, R., and van Horen, N. (2011). Running for the exit: International banks and crisis transmission. De Nederlandsche Bank Working Paper No. 279, February 3. Research record
  • Ehlers, T., Hardy, B., and McGuire, P. (2025). International finance through the lens of BIS statistics: the international dimensions of credit. BIS Quarterly Review, March 11, pp. 89–106. Research record
  • Emter, L., McQuade, P., Pradhan, S.-K., and Schmitz, M. (2024). Determinants of currency choice in cross-border bank loans. BIS Working Papers No. 1184, May 3. Research record
  • Giannetti, M., and Laeven, L. (2012). The flight home effect: Evidence from the syndicated loan market during financial crises. Journal of Financial Economics, 104(1), 23–43. Research record
  • Hardy, B., McGuire, P., and von Peter, G. (2024a). International finance through the lens of BIS statistics: bank exposures and country risk. BIS Quarterly Review, September 16, pp. 69–86. Research record
  • Hardy, B., McGuire, P., and von Peter, G. (2024b). International finance through the lens of BIS statistics: the geography of banks' operations. BIS Quarterly Review, December 10, pp. 69–86. Research record
  • McCauley, R. N., McGuire, P., and Wooldridge, P. (2021). Seven decades of international banking. BIS Quarterly Review, September, 61–75. Research record
  • Pogliani, P., von Peter, G., and Wooldridge, P. (2022). The outsize role of cross-border financial centres. BIS Quarterly Review, June, 1–15. Research record