SCIENCE
Banking
Abstract. Banking is the study of institutions that issue deposit claims, extend credit, intermediate financial risks, and provide continuing access to payment and liquidity services. The bank is both an economic organization and a balance-sheet institution whose assets and liabilities connect private borrowing to transferable purchasing power. This monograph develops a coherent account of banking across contractual theory, monetary accounting, risk management, institutional history, financial instability, competition, global development, technological change, and public safeguards. It relates the established theories of delegated monitoring, asymmetric information, liquidity insurance, demandable debt, and joint liquidity provision to observed banking operations and supervisory constraints. It distinguishes accounting identities from behavioral models, bank solvency from liquidity, firm-level efficiency from systemic stability, and standards from national law. Its documentary basis combines foundational journal research with official Basel, FSB, IMF, World Bank, Federal Reserve, Bank of England, and FDIC evidence verified for this edition. A synthetic conclusion identifies where the explanatory power of modern banking research is strong and where policy depends on jurisdiction-specific institutions, uncertain behavior, and endogenous responses to regulation. The analysis is interpretive and integrative; it proposes no new named theory and reports no independently estimated causal effects.
Keywords: Banking; commercial banks; deposit money; credit creation; financial intermediation; delegated monitoring; information asymmetry; bank runs; liquidity transformation; bank capital; financial regulation; resolution; deposit insurance; financial inclusion; digital banking; systemic risk
The Subject and Boundaries of Banking
Banking studies the organization, financing, governance, economic functions, and public consequences of banks. A bank is a legally constituted institution whose activities and authorizations depend on the applicable jurisdiction. The prototypical commercial bank accepts deposits repayable in money, extends credit, maintains accounts that allow payments, and manages the resulting contractual, liquidity, and solvency exposures. Many banking groups also undertake custody, securities, foreign-exchange, advisory, and wholesale-market activities. A universal definition must therefore identify the institutional logic without assuming that every licensed bank performs the same activity bundle.
A deposit is a claim of the depositor against its bank; a loan is a claim of the bank against a borrower. These two claims have different obligors, maturities, risk attributes, and payment roles. The public's demand for readily transferable deposit balances is supported by banks holding loans and securities that can be comparatively illiquid or long dated. Banking connects these different forms of claims in a contractual structure subject to market demand and public constraints. The Diamond–Dybvig analysis of deposits formalizes why withdrawal rights can provide valuable liquidity service while exposing banks to coordination failure.
The subject contains at least five analytically distinct levels: the customer contract; the legal entity and consolidated banking group; the network of banks and payment obligations; the national monetary and supervisory system; and global finance. Events at one level are transmitted to others. A single customer withdrawal is a routine payment; simultaneous withdrawals from many customers can change liquidity needs and expectations throughout an institution and its counterparties. A profitable bank can be vulnerable to a sudden funding shock, while a solvent bank cannot ensure every asset is immediately realizable at carrying value.
Banking intersects with monetary economics, financial intermediation, payment systems, accounting, corporate finance, risk theory, industrial organization, development economics, behavioral economics, information economics, law, and computer science. The present Science addresses the economic and institutional character of banking. It treats the operations of central banks, the macroeconomic conduct of monetary policy, and the general design of payment infrastructures as connected subjects rather than substituting for their independent disciplinary treatment.
Historical Development of Banking Institutions
Historical banking developed through arrangements for safekeeping, exchange, merchant finance, payments across places and currencies, deposit transfer, public-debt finance, and eventually large-scale industrial lending. Medieval and early-modern merchant bankers handled bills of exchange and credit among trading centers; public giro banks and private bankers addressed unreliable coin and settlement conditions. Across different centuries, deposit banking, exchange banking, central banking, and commercial lending emerged through overlapping institutional paths. Historical names such as bank, banker, exchange bank, and deposit bank should be interpreted in their period-specific legal and accounting contexts.
The rise of public monetary authorities and banknote systems altered the division between government-backed final settlement assets and private liabilities payable at par. Nineteenth-century branching networks, clearing associations, joint-stock banking, and lender-of-last-resort ideas supported larger financial systems. The twentieth century extended insured retail deposits, mass consumer credit, state and central-bank backstops, and financial supervision. International integration created banks operating through branches, subsidiaries, correspondent relations, securities markets, and foreign-currency funding.
The Great Depression, postwar regulation, deregulation and innovation, the global financial crisis of 2007–2009, and the bank disruptions of 2023 each changed the research agenda. These episodes demonstrate that the historical evolution of banking depends jointly on private contracts and the public legal framework for currency, chartering, insurance, and failure. The modern regulatory trajectory can be traced through the Basel Core Principles, which articulate minimum expectations for effective supervision across banking-system types.
Historical comparison must avoid a teleological account in which all banks supposedly converge toward a single universal corporate model. Cooperative banks, public development banks, Islamic banking arrangements, community lenders, investment-oriented banking groups, and digital entrants can embody different objectives, liabilities, ownership forms, and supervisory classifications. The institutional history supplies plausible mechanisms; identifying their causal consequences requires evidence specific to time and jurisdiction.
Theories of Why Banks Exist
Delegated monitoring and information costs
When ultimate investors cannot efficiently investigate each borrower, the financial intermediary may acquire information, administer contracts, monitor performance, collect payments, and enforce claims on their behalf. In Diamond's delegated-monitoring model, diversification allows the intermediary to economize on the cost of providing monitoring incentives. The logic explains how a banking organization can economize on information production and enforcement relative to atomized direct lending, within a specified setting of incentive problems and debt contracts.
Information is valuable because future borrower cash flows are uncertain and may be privately observed. Lending contracts must induce disclosure, effort, prudent risk selection, and repayment. A bank invests in screening and continuing monitoring using credit histories, project cash flows, collateral, covenants, covenanted reporting, and relationships. It trades off the marginal cost of information against expected losses and the opportunity cost of capital.
Liquidity insurance and demandability
The bank-run and liquidity-insurance model identifies a second rationale: deposit contracts permit households with uncertain consumption timing to obtain liquidity before long-term investments mature. Such contracts may support multiple equilibrium outcomes, including an inefficient run. The existence of demand deposits can thus be efficient even though the same contract can transmit panic.
Demandability also contributes to governance. Calomiris and Kahn's demandable-debt study models withdrawal rights as a disciplinary device where investors have limited information about the banker. A bank liability can simultaneously serve as payment money, a liquidity service, and a control mechanism. The cost of abrupt withdrawal is part of this contract design, not an accidental appendage.
Joint production of banking services
In Banks as Liquidity Providers, Kashyap, Rajan, and Stein explain the coexistence of deposit-taking and lending commitments through shared liquid-asset inventories. If deposit withdrawals and credit-line drawdowns are imperfectly correlated, one buffer may serve both activities. This result illuminates the bank as an organizer of contingent liquidity obligations rather than merely a conduit transferring a fixed stock of prior savings.
Agency, collateral, and economic activity
Holmström and Tirole's incentive model examines how the distribution of capital across firms, banks, and uninformed outside financiers changes investment and monitoring. Capital shortages can constrain projects even when technological returns are attractive, with more severe effects on poorly capitalized firms. Together, delegated monitoring, liquidity insurance, demandable-debt discipline, joint liquidity production, and capital-constrained agency form complementary explanations of banking. Their assumptions differ. No one model proves that the legal institution called a bank is uniquely optimal in every economic environment.
Bank Balance Sheets and Accounting Identities
The foundational accounting identity of a banking entity is assets = liabilities + equity. Assets normally include central-bank balances and currency, interbank claims, loans, securities, derivatives with positive value, and operational assets. Liabilities can include customer deposits, unsecured and secured wholesale borrowing, issued debt, derivatives with negative value, and other payables. Equity is the residual claim and loss-absorbing funding base under the applicable accounting and capital rules.
A simple illustrative balance sheet begins with assets of 100 units: reserves 15, loans 65, securities 20; liabilities of customer deposits 85 and equity 15. It balances because 15 + 65 + 20 = 85 + 15. These values are didactic assumptions, not a report about a real bank. A new loan for 10 units credited to a deposit account increases loans from 65 to 75 and deposits from 85 to 95, raising both sides to 110. In this instantaneous bookkeeping illustration the bank has created a private claim and a deposit liability simultaneously.
Now suppose the borrower transfers the entire 10 units to a customer at another bank. The first bank's new deposit liability may be extinguished through the outgoing payment, while its settlement reserves fall by 10 units if paid with its own existing reserves. Its loan remains. The receiving bank gains settlement balances and creates or credits a deposit for its customer. Funding composition and reserve distribution have changed even though the original lending contract remains on the first bank's asset side. Where the paying bank lacks adequate settlement balances, it must raise them through incoming flows, interbank borrowing, asset sales, secured funding, or central-bank facilities where eligible.
It follows that accounting creation of a deposit at origination is fully compatible with subsequent bank-level funding and liquidity constraints. Money Creation in the Modern Economy describes the contemporary credit-money mechanism while stressing the roles of bank profitability and monetary policy. A mechanical story in which banks simply pass through pre-existing customer deposits misses accounting creation; an unlimited-credit story misses capital, solvency, risk pricing, settlement, and demand constraints.
Bank accounting must distinguish book value, fair value, expected loss allowances, prudential adjustments, collateral value, and cash realizability. A debt security can have a fixed contractual redemption amount and a lower market price after interest rates rise. Whether the market change enters current accounting equity, regulatory capital, or disclosed unrealized gains depends on classification and applicable rules. Economic exposure can remain relevant even where accounting recognition differs.
Deposits, Loans, and the Creation of Money
The economics of a deposit depends on its denomination, legal promise, withdrawal rights, guarantee, price, transactional convenience, and operational accessibility. Sight deposits facilitate payments; term deposits contractually exchange some immediate availability for different remuneration; wholesale deposits and treasury balances can have highly mobile funding characteristics. Deposit-taking institutions must manage the likelihood and correlation of withdrawals and the composition of insured and uninsured balances.
A loan advances purchasing power to a borrower against a future repayment claim. Loan pricing must cover marginal funding cost, expected credit loss, operating cost, the opportunity cost of scarce capital and liquidity, uncertainty, and a suitable return to the bank's owners or public mandate. Credit contracts allocate state-contingent obligations through interest, maturity, amortization, collateral, security interests, guarantees, covenants, and remedies. Some lending is drawn immediately; commitments and revolving facilities make future funding demand contingent on customer decisions.
The supply of new bank credit can initially create deposits, but the subsequent fate of those deposits depends on payments and portfolio choices across the system. Repayment of bank principal extinguishes the corresponding bank loan asset and, when funded by a deposit in the banking system, reduces deposit liabilities. The precise distribution across banks depends on settlement and funding operations. Asset purchases, government transactions, cash withdrawals, and interbank transfers introduce additional channels.
At the macroeconomic level, aggregate bank lending and deposit creation are influenced by borrowing demand, creditworthiness, interest-rate policy, macroeconomic expectations, competition, regulation, and banks' own capital and liquidity capacity. Funding-shock research in 2026 documents why a bank replacing departing deposits with higher-cost wholesale finance faces a different profit outlook even when it continues to own its loans. The banking system's accounting capacity to create inside money therefore coexists with economic and institutional constraints on its scale and stability.
Contracts, Screening, and Credit Allocation
Credit allocation is a problem of choice under uncertainty and private information. A bank may know less than the borrower about underlying project quality, effort, intentions, and available alternatives. Charging a higher interest rate can change the pool of applicants or alter borrowers' incentives; the marginal increase in contractual yield need not produce a corresponding increase in expected actual repayment. The classic Credit Rationing in Markets with Imperfect Information studies circumstances in which loan quantities may be constrained even when some rejected borrowers would accept the stated interest rate.
Screening uses income and cash-flow evidence, leverage, payment history, sector outlook, project evaluation, collateral and relevant qualitative factors. Monitoring follows through periodic financial statements, account activity, covenant tests, site visits, collateral valuations, and restructuring negotiations. Loss experience is not a sufficient indicator of underwriting quality if macroeconomic conditions or risk appetite change sharply. Lending decisions must account for correlated exposures and the possibility that familiar customer segments share the same adverse shock.
Relationship lending builds information across repeated interactions. Berger and Udell's small-firm evidence finds that longer relationships were associated with lower rates and a reduced likelihood of collateral requirements in its study population. Interpreting these results requires attention to selection: stronger firms may sustain longer relationships, and both relationship quality and credit terms depend on unobserved borrower traits. The research establishes a meaningful empirical association within its setting, not an unconditional rule that loyalty reduces credit costs for everyone.
Banks also allocate credit through standardized scoring, syndication, securitization, collateral lending and guarantees. The marginal value of relationship information varies by firm transparency, loan size, availability of transaction data and competition. Lending design is therefore an institutional choice between multiple ways of acquiring and using information.
Liquidity Transformation and Bank Runs
A bank promises access to its liabilities at contractual dates while investing in assets whose sale or collection may require time. Liquidity transformation is the transformation of the timing and marketability of claims. Maturity transformation is related but distinct: a long-dated asset funded by a short-term deposit has both repricing and refinancing exposure, but even some short-duration loans can be difficult to sell without loss. Liquidity risk is the inability to meet obligations when due at tolerable cost; solvency risk concerns whether the economic value of resources can cover obligations.
A run occurs when withdrawals or withdrawals threatened by depositor expectations generate a self-reinforcing demand for liquidity. The Diamond–Dybvig multiple-equilibrium result demonstrates a run mechanism in which expectations matter. A fundamentally weak bank may also experience rational withdrawals in response to losses or poor governance; an observed run alone does not identify the contribution of fundamentals versus coordination.
Modern payment technology can shorten the time between concern and outflow. The FDIC's 2026 depositor-flight study analyzed transaction-level data from Silicon Valley Bank, Signature Bank, and First Republic Bank, finding particularly rapid flight among large uninsured depositors. The result points to the distribution and concentration of individual balances, not simply the aggregate amount of deposits, as important funding-risk dimensions. It does not license a claim that all uninsured depositors respond in the same way.
A bank's liquidity defense includes cash and central-bank reserves, unencumbered marketable securities, cash inflows, stable liabilities, collateral mobilization capacity, diversified funding sources, and rehearsed contingency plans. The Basel Liquidity Coverage Ratio requires a stress-based comparison of eligible liquid assets with projected net outflows over 30 days for in-scope institutions. Even a strong measured ratio cannot exhaust all adverse scenarios or guarantee uninterrupted market access. Liquidity management is a dynamic process involving behavior, funding networks, operational readiness, and confidence.
Bank Funding, Capital, and Risk
Bank capital is a cushion for losses and an incentive-bearing financing claim. The accounting equity ratio compares equity with assets. Regulatory capital ratios compare defined eligible capital with risk-weighted exposures or leverage-denominator measures. A bank can satisfy one ratio and face difficulty under another because the denominators, permitted capital instruments, and risks captured differ. A bank's cash liquidity cannot be inferred from its regulatory capital position.
The Basel risk-based capital framework specifies minimum internationally agreed capital requirements and related buffers. In the consolidated framework's current applicable minimum structure, Common Equity Tier 1 capital must be at least 4.5 percent of risk-weighted assets, Tier 1 at least 6 percent, and total capital at least 8 percent, before relevant buffers and additional constraints. These are Basel standards; country law and transitional implementation determine each banking organization's binding duties. Risk weights are regulatory approximations rather than direct probabilities of default or reliable upper bounds on loss.
The Net Stable Funding Ratio promotes more stable financing relative to the liquidity and maturity characteristics of assets and off-balance-sheet commitments. It complements short-horizon stress liquidity measurement. Both liquidity and funding ratios depend on prescribed assumptions about the behavior of deposits, wholesale investors, loan drawdowns and financial markets.
Bank risk comprises credit losses; market-price and foreign-exchange changes; interest-rate risk in the banking book; funding and liquidity shocks; counterparty and settlement failures; operational, cyber and fraud losses; conduct and legal exposures; concentrated customers, sectors and geographies; and systemic correlations. Risks may reinforce each other. A rise in interest rates can lower bond prices, alter deposit competitiveness, weaken borrowers, raise bank funding costs, and shift depositors' expectations. Capital and liquidity buffers address different parts of this joint transmission process.
Pricing, Profitability, and Competition
Bank profitability depends on the margins between asset yields and liability costs, fee income, loan losses, funding structure, operating efficiency, taxes, and the capital base supporting operations. Net interest income is interest earned on interest-bearing assets less interest paid on interest-bearing liabilities. Net interest margin divides that income by a chosen average earning-asset denominator; comparability requires consistent accounting definitions. Return on assets and return on equity convey different sensitivities to leverage. A high return on equity may reflect durable competitive advantage, unusually high risk, or a thin equity base.
Deposit rates depend on customer alternatives, convenience, branch or digital service quality, deposit guarantees, switching costs, product bundling, and banks' funding needs. Loan rates depend on expected losses, market rivalry, borrower information, capital constraints, and collateral. Rate-setting is rarely a single national interest margin; it involves multiple customer segments and instruments with different maturities, option values and costs.
Banking industrial organization studies concentration, entry barriers, geographic markets, network effects, information accumulation, and government guarantees. Concentration may allow scale economies, but can also generate market power or large interconnected institutions. Low concentration does not automatically make the system safe: widespread competitive pressure can change lending standards, risk-taking, or the resilience of institutions. Conversely, large diversified banks can spread some idiosyncratic losses while creating substantial systemwide consequences if they fail.
For credit markets, information and contracting conditions are integral to competition. The credit-rationing literature demonstrates why supply allocation may adjust on nonprice dimensions. The relationship-lending findings show that accumulated soft information may be valuable for certain borrowers. Digital comparison tools and open interfaces may reduce some customer search and switching costs while creating new platform dependencies and data advantages. Competition analysis must examine both prices and the stability of promised financial services.
Risk Management and Bank Governance
The bank's board and senior management select strategy, risk appetite, internal control responsibilities, and the tolerance for concentration or maturity mismatch. An effective risk system combines independent challenge, reliable data, internal audit, scenario analysis, model validation, legal compliance, escalation pathways, and the ability to respond to adverse information. Risk committees and formal limits are useful only when decision-makers understand, enforce, and revise them.
Credit-risk measurement separates probability of default, exposure at default, loss given default, expected loss and unexpected loss concepts. Loan grading and provisioning respond to information about borrower condition and the applicable accounting standard. Portfolio analysis studies correlations among exposures, including the same real estate market, supply chain, sector, currency, or funding channel. Stress tests ask how bank capital, earnings, and liquidity evolve under defined adverse paths; results are conditional on scenarios, data and management assumptions.
Market risk and interest-rate risk require mapping cash flows, repricing dates, durations, embedded optionality and hedges. A bank may appear well hedged under an assumed deposit-duration model yet suffer when depositor behavior changes abruptly. Operational resilience covers payments, authentication, outsourced technology, data integrity, cyber incidents, and continuity of access to critical records. Incident risks can produce balance-sheet consequences through withdrawals, business interruption, compensation, regulatory action, or reputational loss.
The Federal Reserve's SVB supervisory review describes management failures, supervisory deficiencies, and a failure of supervisory escalation as intertwined explanations of the 2023 collapse. Its lesson is institutional: technical indicators matter only when organizations connect them to governance, credible funding plans and decisive action. The 2024 supervisory principles provide the corresponding public benchmark for responsibilities, powers, independence, risk management and timely intervention.
Prudential Regulation and Supervision
Banking regulation is justified by the economic consequences of failure, depositors' limited information, payment interdependence, asymmetric incentives, public guarantees and the capacity of financial shocks to disrupt real activity. Its instruments include licensing and fit-and-proper assessments; capital and liquidity requirements; limits on exposures; governance rules; inspections and data reporting; recovery planning; depositor protection; conduct safeguards; and resolution powers. Regulatory architecture differs between countries, and the word supervision can refer to multiple authorities with overlapping responsibilities.
The Basel Core Principles form a common international reference for effective supervision: the 2024 framework contains 29 principles, the first thirteen focusing on supervisors' powers and responsibilities and the remaining sixteen on prudential requirements for banks. The principles are not a mechanical international banking code automatically enforceable in each state. Assessment of real compliance requires examining national powers, incentives, practice, outcomes, and enforcement.
Capital regulation reduces some risks but can generate risk-weight arbitrage if exposures are inaccurately measured or shifted outside the regulated perimeter. Liquidity regulation supports preparedness, but measured flows may understate exceptional deposit behavior or failure of presumed collateral markets. The LCR stress standard and NSFR funding requirements therefore complement, rather than replace, supervisory judgment, contingency liquidity access, and direct inspection of asset quality and funding concentrations.
Macroprudential policy considers system-wide credit booms, correlated balance sheets, cross-bank contagion, and common exposures. Countercyclical buffers, borrower-based limits, systemic capital surcharges, and liquidity measures address different channels. A coherent public framework seeks to preserve payment and credit functions while allowing shareholders and creditors, under applicable laws, to bear losses. Evaluating regulation demands counterfactual analysis, implementation dates, incentives, and recognition that periods of calm are an imperfect test of resilience.
Deposit Protection, Recovery, and Resolution
Deposit insurance protects eligible depositors within legally established limits and conditions and can strengthen confidence in bank liabilities. Coverage varies by jurisdiction, institution, depositor, account category, currency, and ownership structure. An insured depositor's experience depends on the speed of reimbursement or continuity of account access as well as the nominal protection limit. Insurance can change depositor monitoring and bank risk incentives, making pricing, supervision, and credible loss allocation relevant.
In the bank-run insurance analysis, a sufficiently credible guarantee can remove the self-fulfilling run equilibrium under the model's conditions. More recent quantitative deposit-insurance research explicitly studies the balance between fund accumulation, premium design, productive deployment of deposits and moral hazard. Its numerical counterfactual findings are conditional on calibrated assumptions, not ready-made coverage rules for every economy.
Recovery and resolution are separate stages. Recovery actions attempt to restore the viability of a still-operating bank by raising capital, selling assets, reducing risk or replacing funding. Resolution applies legal tools to manage a failing or likely-to-fail institution while preserving essential services and allocating losses according to national law. The FSB Key Attributes articulate international goals and powers for orderly resolution, cross-border cooperation, information sharing and planning. The April 2024 version principally updates guidance for financial market infrastructures while retaining the original twelve Key Attributes.
A credible resolution regime must connect legal authority to operational execution: deposit records, payment continuity, valuations, claims hierarchy, funding, transfers, communications, and cross-border coordination. It cannot eliminate the economic cost of a failed investment portfolio. It can change the timing, distribution and external spillovers of that loss.
Banking Crises and Contagion
Banking crises combine asset losses, funding shortages, collapsing confidence, institutional interdependence and policy responses in varying proportions. A solvency crisis can generate a liquidity crisis; a liquidity crisis can force asset sales and generate solvency losses. Banking distress can transmit through interbank exposures, collateral calls, common asset holdings, deposit migration, credit contraction, and expectations about the availability of government support.
The Systemic Banking Crises Database: 1970–2025 provides a structured cross-country record of crises, policy interventions and associated costs. Its operational definition requires both systemic financial distress and substantial policy intervention. Counts therefore depend on coding decisions and source completeness. Comparing fiscal costs or output losses across crises requires consistency in accounting conventions and counterfactual assumptions.
In March 2023, Silicon Valley Bank failed after rapid balance-sheet growth, significant uninsured funding concentration, and exposure to interest-rate losses. The official supervisory findings identify failures of risk management and regulatory oversight. The 2023 Federal Reserve funding-risk assessment situates those vulnerabilities within broader deposit and interest-rate shifts, while noting that the failed banks were unusual in the extent of their uninsured funding concentrations. The 2026 failed-bank depositor study adds transaction-level evidence about the size and speed of depositor withdrawals.
A generalization from these events is that risk aggregation must include asset duration, deposit concentration, funding mobility, collateral usability, and operational capacity to execute liquidity plans. A bank can carry large liquid-looking securities but face material losses if they must be sold rapidly after interest rates rise. A decline in confidence can affect both asset prices and the cost of maintaining liabilities. The October 2026 financial-stability assessment reports broadly resilient global banks alongside vulnerable pockets and growing cross-sector risks. Aggregate resilience and institution-specific vulnerability are compatible facts.
Banking and the Macroeconomy
Banks transmit macroeconomic conditions to households and firms through loan supply, deposit pricing, payment reliability, risk preferences, and the valuation of collateral. Interest-rate changes alter the prices of new credit and deposits, but also the present value of existing assets, the burden of variable-rate debt, and incentives to shift among bank and market instruments. Bank lending responds to borrowers' prospects and demand while simultaneously influencing investment and spending.
The financial-intermediation and real-sector model describes how shortages of capital among firms and intermediaries can restrict investment disproportionately among constrained firms. The liquidity and leverage study demonstrates a related amplification mechanism in mark-to-market intermediary balance sheets. Their domains are not identical: dealer leverage and deposit-bank credit do not share all funding structures or accounting treatments. Their common implication is that financial intermediaries adjust balance-sheet capacity endogenously as market conditions change.
Bank-based finance can smooth consumption, fund housing, provide working capital and spread information costs. It can also intensify boom-bust cycles when collateral values, borrower leverage, and bank balance sheets reinforce each other. Empirical interpretation must distinguish a contraction caused by declining credit demand from one caused by a bank funding or capital shock. Loan quantities alone do not identify the causal channel.
Government-bond holdings create a further interaction between banks and sovereigns. Government debt may be an important liquidity asset or collateral, while abrupt repricing or sovereign distress can weaken bank balance sheets. Conversely, costly bank support can affect public finances. The 2026 global risk review emphasizes sovereign yield shocks and bank–nonbank linkages as relevant stability channels. These interactions warrant coordinated macroeconomic, prudential and fiscal analysis without merging the distinct responsibilities of banks and central banks.
Payments, Clearing, and Settlement
Banks support payment activity by maintaining customer accounts, authenticating instructions, verifying balances, transmitting messages, participating in clearing arrangements and settling obligations. A transfer between customers at the same bank can be implemented by changes in internal deposit-ledger entries. A transfer between banks creates or changes an interbank settlement obligation; systems differ in the use of central-bank reserves, net settlement, correspondent balances and designated settlement assets.
Settlement finality is a legal and operational property of payment arrangements. The customer's experience of an immediate app confirmation may precede or differ from irrevocable interbank settlement. Operational resilience therefore involves identity and fraud controls, message integrity, availability of connectivity, capacity for intraday liquidity, and recoverability after technology failure.
Commercial-bank deposits are widely used purchasing-power claims because individuals and firms expect them to exchange at par through payment systems, subject to national protections and institutional arrangements. Payment convenience increases the value of deposits and affects deposit pricing. Conversely, an outage, legal restriction, or loss of confidence in settlement can damage banks even when their lending portfolio has not changed. The joint-liquidity-provision analysis helps explain why contingent payments and lending commitments make liquid-asset management a core banking function.
Digital payment innovators can provide interfaces and transaction services without becoming deposit-taking banks. Analytically, the institution holding the liability, the party initiating the instruction, the entity operating the rail, and the ultimate settlement authority may be four different organizations. Mapping these layers is essential for tracing risk, customer claims and regulatory responsibility.
International Banking and Emerging Markets
Cross-border banking supplies trade credit, foreign-currency finance, correspondent services, international payments, and risk transfer. A multinational bank operates under distinct national laws through branches, subsidiaries, and affiliated legal entities. Consolidated financial statements can conceal the location of trapped liquidity, ring-fenced capital, deposit guarantees and local loss-bearing resources. Host and home supervisors may have different information, legal powers and incentives during stress.
Currency mismatch is especially important where banks borrow or accept deposits in a reserve currency while assets and borrower cash flows are denominated in domestic currency. Exchange-rate depreciation can increase effective borrower burdens, reduce collateral value in foreign-currency terms, and create funding stress. Cross-border wholesale funding can improve financial depth during normal times while exposing local credit to global liquidity and risk-appetite cycles.
The international resolution framework treats cross-border cooperation and credible information exchange as necessary elements of failure management. Yet the real capacity to recognize and transfer losses depends on legally enforceable arrangements and local operational preparedness. Bank groups are not equivalent to a single globally consolidated pool of immediately transferable resources.
Emerging and developing economies display widely different banking structures: state-owned banks, foreign-bank subsidiaries, microfinance-linked intermediaries, credit cooperatives, mobile-money providers and informal credit markets. The IMF 2026 financial-access brief reports that nonbanks now hold more loan accounts than banks in low-income countries, while banks retain most loan accounts across the economies covered overall. The relationship between banking development and productive growth depends on institutional quality, borrower opportunity, supervision, financial literacy, inequality, and shocks. More bank credit by itself is not a sufficient metric of development.
Banking, Inclusion, and Economic Development
Financial inclusion asks who can access and meaningfully use safe and affordable accounts, credit, savings instruments, and payment services. The appropriate unit of analysis can be the adult, household, enterprise, or transaction. The Global Findex 2025 survey draws on nationally representative fieldwork conducted in 2024 and reports indicators across 141 economies; its definition of account ownership includes regulated nonbank and mobile-money providers. That definition matters when translating account-ownership percentages into claims about the size of the banking customer base.
Access and use differ. Opening an account does not imply regular savings, affordable credit, sufficient consumer protection, or resilience to financial shocks. Bank branches may remain necessary for cash and some advisory needs, while digital accounts can expand geographic access where internet, identification and device availability permit. The price, identity rules, language accessibility, trust, and security of services determine whether formal availability becomes effective financial participation.
Credit inclusion also raises underwriting tradeoffs. Broadening credit to opaque borrowers may require more expensive information collection and careful risk assessment; excluding them can leave profitable investment unrealized. Credit scoring can reduce some administrative costs but may also import bias through incomplete or unrepresentative data. The task of the researcher is to evaluate both borrower outcomes and bank portfolio outcomes, with explicit time horizons and counterfactuals.
Supply-side Financial Access Survey data and demand-side survey evidence answer distinct questions. Institution-reported account counts, transaction counts and branch figures cannot automatically be equated with the number of unique adults served. A person may hold multiple accounts, and a financial-service user may use a provider that is not classified as a bank. Careful measurement is part of economic interpretation, not a technical appendix.
Digital Banking and the Changing Financial Perimeter
Digital banking involves customer interfaces, remote onboarding, automated compliance, payment application interfaces, account data exchange, cloud infrastructure, analytics, and digitally mediated lending. The technology modifies the cost structure, geographic reach, customer experience, information flows, and speed of transfers. It also creates concentration and operational dependencies in platforms, identity systems, telecommunications, data suppliers, and service providers.
Open-banking arrangements can permit authorized data access by third parties under country-specific laws and standards. Their potential effects include lower customer switching costs, more personalized services, and expanded comparison of products. Their risks include consent failures, cybersecurity exposures, data misuse, and fragmentation of responsibility. The mere presence of an API does not establish lawful authorization or beneficial consumer outcomes.
Artificial intelligence can assist transaction monitoring, fraud detection, underwriting, customer support, internal controls and research. Its impact depends on governance, data quality, validation, drift detection, cyber resilience, outsourcing, and human accountability. The FSB's October 2026 AI sound practices propose twelve considerations spanning organization-wide governance and the lifecycle of AI systems, including explainability, monitoring, oversight and third-party risk. Those recommendations are guidance and should be distinguished from binding bank-regulation provisions in each jurisdiction.
Tokenized deposits, stablecoins, central-bank digital currencies and account-based commercial-bank money are legally and economically distinct claims. A new ledger representation does not by itself change the obligor, the source of loss absorption, or the rights of the holder. Researchers should distinguish settlement design, enforceable redemption rights, balance-sheet treatment, governance, and operational failure modes. The October 2026 stability report notes both potential tokenization functionality and prospective new channels for liquidity mismatches, leverage and technological dependencies.
Technological progress increases the relevance of a basic question: which institution has promised payment, which entity holds liquid resources, and who bears losses when a promise cannot be honored? Banking retains a contract-and-balance-sheet center even when the visible user interface changes dramatically.
Research Methods and Evidentiary Limits
Banking research uses formal theory, institutional histories, legal comparison, bank call reports, supervisory returns, transaction records, market prices, payment statistics, surveys, firm-level credit files, and natural or designed policy experiments. The reliability of a conclusion depends on the identification strategy and the proximity between measured variables and the economic mechanism of interest.
Theory clarifies possible mechanisms under explicit assumptions. The canonical bank-run model is a rigorous analysis of deposit contracts and expectations; it is not a dataset establishing how every real bank fails. The delegated-monitoring framework identifies an incentive solution; it does not establish the effect of all lending intermediaries in every regulatory environment. The liquidity-synergy study advances both theory and empirical examination, but its empirical predictions must be interpreted in the selected institutional setting.
Empirical designs need appropriate comparison groups and treatment definitions. A decline in lending after a capital shock may reflect borrower demand or bank supply. Banks experiencing a run are not randomly selected from the population of banks. The failed-bank transaction study gives unusually detailed depositor behavior for particular institutions, but generalizing from failures demands caution about selection and differences in deposit insurance or customer concentration.
The updated crisis database is useful for systematic international comparison, but crisis coding, intervention thresholds, fiscal-cost accounting and observation windows need to be held consistent. The household financial-access survey observes self-reported individuals and their behaviors, while the administrative financial-access dataset summarizes institution-reported products and transactions. Their denominators, coverage and reporting units differ.
Policy studies should state the jurisdiction, effective date, legal status, regulated population and degree of implementation. The Basel capital rules and Core Principles describe international standards and supervisory benchmarks; actual application requires national legal and supervisory evidence. This monograph uses numerical examples only where expressly identified as illustrative accounting arithmetic. It does not report independent estimation, meta-analysis, replication, primary historical archival discovery, interviews, or external peer review.
Integrative Analysis and Research Agenda
Across the evidence, the bank is intelligible as an organized set of promises rather than as a single product. Customers value the ability to transfer deposits, receive assured or contingent credit, and preserve access to payment services. To deliver those promises, the bank acquires risky assets, evaluates borrowers, maintains capital, and assembles funding that is behaviorally reliable under uncertain conditions. These activities are simultaneous and causally entangled.
The first integrative finding is that bank-credit creation and bank funding constraints belong to the same explanation. Loan origination can create a deposit immediately, as the Bank of England accounting explanation shows; the continued viability of that asset requires capital, asset quality, funding and settlement access. Treating the accounting entry as the complete theory of lending mistakes an identity for a behavioral equilibrium.
The second finding is that deposit liquidity is both a banking service and a source of endogenous vulnerability. The liquidity-insurance result, demandable-debt discipline argument, and modern depositor-flight evidence describe complementary mechanisms: valuable withdrawal options, incentives for banker discipline, and rapid collective outflow. Their joint interpretation helps explain why depositor protections, supervision, liquid resources, and resolution arrangements are economic complements rather than isolated safeguards.
The third finding concerns the boundaries of a bank. Delegated monitoring and relationship information can improve credit allocation; technology can replace or commoditize portions of that work; nonbanks can perform similar financing functions through different balance sheets. What remains distinctive is the precise combination of regulated promises, liquidity-bearing liabilities, settlement connections, contractual rights and public safeguards. A statement about the economic function of a bank must therefore be connected to an actual legal entity and jurisdiction.
The fourth finding is that the measurement of resilience requires several horizons. A loan portfolio can perform acceptably while a sudden run overwhelms funding; a liquid bank can carry insufficient loss-absorbing capital; a capitalized bank can become operationally unavailable after a cyber incident. The short-horizon liquidity standard, stable-funding standard, and capital framework address different dimensions. They must be read together with governance, market access, customer concentration and credible resolution.
The fifth finding is that banking modernization changes the speed and distribution of uncertainty. Digital services can expand access, but also accelerate withdrawals and increase reliance on common technology providers. This motivates a research agenda in which financial inclusion, deposit concentration, digital identity, third-party resilience, systemwide liquidity, and customer trust are studied as interacting variables rather than independent success measures.
Open scientific problems include the behavior of large uninsured depositors under instantaneous payment capabilities; causal effects of different resolution and deposit-guarantee designs; effective stress measures for contingent liquidity and nonbank-bank linkages; identification of credit supply in the presence of digital platform lending; governance and validation of adaptive underwriting models; cross-border liquidity portability under stress; and the welfare effects of regulation when innovation alters the institutional perimeter. These are research questions grounded in the reviewed literature, not newly asserted theories or claims of priority.
Conclusion
Banking is the institutional organization of credit, deposit liabilities, liquidity access, payment promises, and financial risk. Its economic function emerges through information production, contractual enforcement, liquidity insurance, intermediation, and monetary accounting. Its weaknesses arise from the same architecture: asset and liability maturities differ, information is distributed unevenly, funding depends on confidence, asset values respond to changing conditions, and individual choices have collective consequences.
The principal schools of banking research provide a cumulative, conditional explanation. Delegated monitoring accounts for economies in information acquisition; asymmetric-information models explain credit constraints; liquidity-insurance models explain the value and vulnerability of callable deposits; incentive models reveal governance effects; joint-liquidity analyses explain the conjunction of deposit-taking and contingent lending. Balance-sheet accounting supplies the formal consistency that must connect these mechanisms to actual bank operations.
The scientific evaluation of banking requires precise levels of analysis. A contract is not a consolidated banking group; a solvency measure is not cash availability; a global standard is not automatically national law; account access is not bank-only membership; an observed crisis is not a randomized experiment. The institutional record and verified literature demonstrate that durable banking systems depend on profitable intermediation, credible claims, adequate loss absorption, reliable liquidity, competent supervision, and executable failure arrangements.
The continuing research task is to measure how these elements evolve with interconnected markets, rapid payments, technological concentration and changing borrower behavior. Banking remains an indispensable object of economic investigation because the everyday promise of transferable money rests on long-lived, risky, and institutionally governed claims.
References
- Achugamonu, Faith, Tim Schmidt-Eisenlohr, and Matthew P. Seay. 2026. “Assessing Bank Resilience to a Funding Shock.” FEDS Notes, February 17. https://www.federalreserve.gov/econres/notes/feds-notes/assessing-bank-resilience-to-a-funding-shock-20260217.html.
- Adrian, Tobias, and Hyun Song Shin. 2010. “Liquidity and Leverage.” Journal of Financial Intermediation 19 (3): 418–437. https://doi.org/10.1016/j.jfi.2008.12.002.
- Basel Committee on Banking Supervision. 2024. “Core Principles for Effective Banking Supervision.” Basel Framework, effective April 25, 2024. https://www.bis.org/committees/bcbs/basel-framework/standard/bcp.
- Basel Committee on Banking Supervision. 2026-accessed consolidated edition. “Liquidity Coverage Ratio.” Basel Framework. https://www.bis.org/committees/bcbs/basel-framework/standard/lcr.
- Basel Committee on Banking Supervision. 2026-accessed consolidated edition. “Net Stable Funding Ratio.” Basel Framework. https://www.bis.org/committees/bcbs/basel-framework/standard/nsf.
- Basel Committee on Banking Supervision. 2026-accessed consolidated edition. “Risk-Based Capital Requirements.” Basel Framework. https://www.bis.org/committees/bcbs/basel-framework/standard/rbc.
- Berger, Allen N., and Gregory F. Udell. 1995. “Relationship Lending and Lines of Credit in Small Firm Finance.” Journal of Business 68 (3): 351–381. https://scholarcommons.sc.edu/fin_facpub/10/.
- Board of Governors of the Federal Reserve System. 2023. “Funding Risks.” Financial Stability Report, May. https://www.federalreserve.gov/publications/2023-may-financial-stability-report-funding-risks.htm.
- Board of Governors of the Federal Reserve System. 2023. “Review of the Federal Reserve’s Supervision and Regulation of Silicon Valley Bank.” Press release and review summary, April 28. https://www.federalreserve.gov/newsevents/pressreleases/bcreg20230428a.htm.
- Calomiris, Charles W., and Charles M. Kahn. 1991. “The Role of Demandable Debt in Structuring Optimal Banking Arrangements.” American Economic Review 81 (3): 497–513. https://business.columbia.edu/faculty/research/role-demandable-debt-structuring-optimal-banking-arrangements.
- Diamond, Douglas W. 1984. “Financial Intermediation and Delegated Monitoring.” Review of Economic Studies 51 (3): 393–414. https://doi.org/10.2307/2297430.
- Diamond, Douglas W., and Philip H. Dybvig. 1983. “Bank Runs, Deposit Insurance, and Liquidity.” Journal of Political Economy 91 (3): 401–419. https://doi.org/10.1086/261155.
- Federal Deposit Insurance Corporation. 2026. “FDIC Releases Staff Study of Deposit Flows at Three Failed Banks in Spring 2023.” May 14. https://www.fdic.gov/news/press-releases/2026/fdic-releases-staff-study-deposit-flows-three-failed-banks-spring-2023.
- Financial Stability Board. 2024. “Key Attributes of Effective Resolution Regimes for Financial Institutions,” revised version, April 25. https://www.fsb.org/2024/04/key-attributes-of-effective-resolution-regimes-for-financial-institutions-revised-version-2024/.
- Financial Stability Board. 2026. “Sound Practices for Responsible Adoption of Artificial Intelligence (AI): Final Report.” October 15. https://www.fsb.org/2026/10/sound-practices-for-responsible-adoption-of-artificial-intelligence-ai-final-report/.
- Holmström, Bengt, and Jean Tirole. 1997. “Financial Intermediation, Loanable Funds, and the Real Sector.” Quarterly Journal of Economics 112 (3): 663–691. https://doi.org/10.1162/003355397555316.
- International Monetary Fund. 2026. “Transforming Financial Access and Use: Nonbanks and the Rise of Digital Finance.” IMF Data Brief, October 30. https://data.imf.org/en/news/imf%20data%20brief%20october%2028.
- International Monetary Fund. 2026. Global Financial Stability Report, October 2026: Resilience, Risk, and Transformation. https://www.imf.org/en/publications/gfsr/issues/2026/10/13/global-financial-stability-report-october-2026/.
- Kashyap, Anil K., Raghuram G. Rajan, and Jeremy C. Stein. 2002. “Banks as Liquidity Providers: An Explanation for the Coexistence of Lending and Deposit-Taking.” Journal of Finance 57 (1): 33–73. https://doi.org/10.1111/1540-6261.00415.
- Laeven, Luc, and Fabian Valencia. 2026. “Systemic Banking Crises Database: 1970–2025.” IMF Working Paper 2026/094, May 15. https://doi.org/10.5089/9798229045971.001.
- McLeay, Michael, Amar Radia, and Ryland Thomas. 2014. “Money Creation in the Modern Economy.” Bank of England Quarterly Bulletin 54 (1): 14–27. https://www.bankofengland.co.uk/quarterly-bulletin/2014/q1/money-creation-in-the-modern-economy.
- Oosthuizen, Dick, and Ryan Zalla. 2024. “Funding Deposit Insurance.” Journal of Financial Stability 75: 101342. https://doi.org/10.1016/j.jfs.2024.101342.
- Stiglitz, Joseph E., and Andrew Weiss. 1981. “Credit Rationing in Markets with Imperfect Information.” American Economic Review 71 (3): 393–410. https://ideas.repec.org/a/aea/aecrev/v71y1981i3p393-410.html.
- World Bank. 2025. The Global Findex Database 2025: Connectivity and Financial Inclusion in the Digital Economy. Report and survey documentation. https://www.worldbank.org/en/publication/globalfindex/report?sf227617306=1.