Loading field map
What do banks do, why do they exist, and why are they fragile? These questions have driven banking theory for more than two centuries. The answers have shifted dramatically, from classical debates over money supply determination to modern microfoundations that explain banks as liquidity providers, delegated monitors, and fragile institutions. Nine major frameworks have shaped this evolution, each responding to the limitations of its predecessors and often reviving older insights in new forms.
The earliest systematic framework, the Real Bills Doctrine (1776–1930), held that a bank could not overissue notes as long as it discounted only short-term commercial bills arising from real transactions. According to this view, the money supply would automatically adjust to the needs of trade, making active regulation unnecessary. The doctrine was influential for over a century, but it faced a fundamental challenge: it assumed that the demand for real bills was independent of the quantity of money, a circularity that critics exploited.
The Currency School (1837–1930) emerged in direct opposition to the Real Bills Doctrine. Its proponents argued that note issue could easily exceed the needs of trade, causing inflation and financial instability. They insisted that the money supply must be strictly tied to gold reserves through a fixed rule, such as the one embodied in the Bank Charter Act of 1844. For the Currency School, banking was a source of instability unless tightly constrained by metallic reserves.
The Banking School (1844–1930) countered that the Currency School’s mechanical link between reserves and notes was misguided. Banking School writers argued that credit is endogenous: banks create money in response to demand, and the volume of notes is determined by the needs of the economy, not by reserve ratios. They saw the Currency School’s rule as both unnecessary and harmful, because it would force contractions when gold flowed out, regardless of real economic conditions. The Banking School thus anticipated a key theme that would resurface a century later: the idea that banks are not passive intermediaries but active creators of money.
These three frameworks coexisted in a heated debate that defined nineteenth-century banking theory. The Real Bills Doctrine gradually lost credibility as its logical flaws became apparent, while the Currency School’s rule-based approach was implemented in Britain but never fully accepted. The Banking School’s endogenous credit view was marginalized after the classical gold standard era, but it never disappeared entirely.
By the mid-twentieth century, the Money-Multiplier View of Banking (1930–Present) had become the standard textbook model. In this framework, the central bank controls the monetary base (reserves and currency), and banks create deposits through a mechanical multiplier: an increase in reserves leads to a multiple expansion of loans and deposits. The money multiplier is treated as a stable, predictable ratio, and banks are seen as passive conduits that simply multiply the reserves injected by the central bank. This view fit neatly into the Neoclassical Synthesis of monetary economics, where money supply was exogenous and controllable.
However, the Money-Multiplier View came under sustained attack from the Endogenous Money and Credit Creation framework (1960–Present). Drawing on the Banking School’s earlier insights, endogenous money theorists—most notably Basil Moore and the post-Keynesian tradition—argued that loans create deposits, not the other way around. Banks first extend credit, then seek reserves later from the central bank or interbank market. The money supply is therefore demand-driven and endogenous, not a policy variable that the central bank can set precisely. This framework revived the Banking School’s core claim and gave it a modern analytical foundation. After the 2008 financial crisis, even mainstream central banks (such as the Bank of England in a 2014 quarterly bulletin) acknowledged that the money-multiplier model is a poor description of actual banking, lending support to the endogenous money view. Today, the Money-Multiplier View survives mainly in introductory textbooks, while endogenous money has become the working assumption of most monetary economists.
The 1980s brought a fundamental shift in banking theory. Instead of treating banks as a black box that creates money, theorists began asking why banks exist at all—what economic function they serve that cannot be performed by direct lending or securities markets. Three frameworks emerged in rapid succession, each offering a different answer.
The Bank Fragility and Bank Runs framework (1983–Present), launched by Diamond and Dybvig’s 1983 model, showed that banks’ core function—maturity transformation (borrowing short and lending long)—makes them inherently vulnerable to self-fulfilling runs. If depositors believe others will withdraw, it is rational to run, even if the bank is solvent. This framework explained why banks need deposit insurance or a lender of last resort to prevent panics. It narrowed the focus of banking theory to the coordination problem among depositors and the role of government safety nets.
Almost simultaneously, the Delegated Monitoring Theory of Banking (1984–Present), developed by Douglas Diamond in 1984, argued that banks exist because they reduce information costs. In a world where lenders cannot cheaply monitor borrowers, a bank acts as a delegated monitor for many small depositors, avoiding duplication of monitoring effort. This framework explained the existence of banks as intermediaries that solve an information asymmetry problem. It complemented the Bank Fragility view by focusing on the asset side of banking (monitoring loans) rather than the liability side (depositor coordination).
A third microfoundation, the Liquidity Creation Theory of Banking (1999–Present), synthesized insights from both earlier frameworks. Kashyap, Rajan, and Stein (1999) observed that banks provide liquidity on demand through both demand deposits and loan commitments (credit lines). These two activities are not separate; they are two manifestations of the same function: providing liquidity insurance. The theory explained why lending and deposit-taking are bundled in a single institution: both require holding liquid assets to meet unpredictable demands. This framework absorbed the insights of Bank Fragility (liquidity risk) and Delegated Monitoring (information-intensive lending) while offering a unified explanation for the coexistence of the two core banking activities.
These three microfoundations do not replace one another; they coexist as complementary explanations. Bank Fragility theory is best at explaining runs and the rationale for regulation. Delegated Monitoring theory is best at explaining why banks, rather than markets, finance opaque borrowers. Liquidity Creation theory is best at explaining the institutional structure of banks as both lenders and deposit-takers. Their assumptions overlap but also conflict: for example, Delegated Monitoring assumes banks are stable monitors, while Bank Fragility emphasizes their instability. The tension remains a live area of research.
Alongside these microfoundations, the Free Banking Theory (1984–Present) offered a radically different institutional prescription. Drawing on historical episodes of unregulated banking (e.g., in Scotland and Canada), free banking theorists such as George Selgin and Lawrence White argued that competitive note issue, without a central bank or deposit insurance, would produce a stable and efficient banking system. They claimed that banks would voluntarily hold adequate reserves to maintain convertibility, and that market discipline would prevent excessive risk-taking. Free Banking Theory directly challenged the regulatory conclusions of the Bank Fragility framework: if runs are a coordination problem, free bankers argued, they can be solved by private contracts (e.g., option clauses) rather than government intervention. This framework remains heterodox, but it has influenced debates on central banking and financial regulation, especially among libertarian and Austrian economists.
Today, banking theory is characterized by pluralism. The three microfoundations—Bank Fragility, Delegated Monitoring, and Liquidity Creation—are the leading frameworks for understanding why banks exist and what makes them fragile. They are often combined in empirical work and in macroprudential models. The Endogenous Money framework has largely won the debate over money creation: central banks and most academic economists now accept that banks create money through lending, not through a multiplier. The Money-Multiplier View is widely regarded as a pedagogical simplification rather than a realistic description. Free Banking Theory remains a minority position but continues to generate empirical research on historical banking systems.
What do these frameworks agree on? They all recognize that banks are special because they create liquidity and are exposed to runs. They agree that information frictions and coordination problems are central to understanding banking. They disagree on the proper policy response: the microfoundations generally justify some form of regulation (deposit insurance, capital requirements, lender of last resort), while Free Banking Theory argues that regulation is unnecessary or harmful. They also disagree on the nature of money creation: endogenous money theorists emphasize that banks are not constrained by reserves in normal times, while the older Money-Multiplier View assumed they were. The integration of banking theory with behavioral economics, network theory, and macroprudential regulation is an ongoing frontier, as the 2008 crisis showed that the microfoundations, while powerful, had not fully captured systemic risk and interconnectedness.