Loading field map
Financial intermediation is the activity of channeling funds between economic units with surplus savings and those with productive investment needs. It is the core business of banks, insurance companies, mutual funds, pension funds, and a range of other institutions that stand between ultimate lenders and ultimate borrowers. The subfield of financial economics that studies this activity asks why such institutions exist at all, what functions they perform, how they are structured, and what consequences their behavior has for the broader economy.
To understand financial intermediation, begin with a world without intermediaries. Savers with excess funds would need to find borrowers directly, evaluate the borrowers' creditworthiness, negotiate loan terms, monitor the borrowers' behavior over time, and bear the risk that borrowers default. Borrowers would need to locate savers willing to accept their particular risk, maturity, and liquidity preferences. This direct matching is costly and often infeasible. Information is dispersed: a saver cannot easily know whether a small business owner will use borrowed funds productively or abscond with them. Risks are hard to diversify: a single saver lending to a single borrower bears that borrower's full default risk. Maturities mismatch: savers typically want liquidity—the ability to withdraw on short notice—while borrowers typically want long-term, stable funding.
Intermediaries arise to address these frictions. A bank, for instance, pools deposits from many savers, lends to many borrowers, and thereby diversifies away much of the idiosyncratic default risk that any single lender would face. It offers depositors demandable claims—accounts they can withdraw from at any time—while holding longer-term, illiquid loans. It develops expertise in screening loan applicants and monitoring borrowers, expertise that individual savers could not economically replicate. In short, intermediaries transform the characteristics of primary securities (loans, bonds, equities) into different characteristics that savers prefer: more liquidity, more diversification, lower monitoring costs.
This perspective—that intermediaries exist to overcome market frictions—is the organizing insight of the modern field. It frames the central questions: What exactly are the frictions that make direct finance costly? Which frictions explain which institutions? And what happens when intermediaries themselves fail to perform these functions?
Early thinking about financial intermediation did not treat it as a distinct subject. Classical and neoclassical economists generally assumed that financial markets were perfect—that savers and borrowers could transact directly at no cost—and therefore had little to say about why banks existed. Banks appeared in monetary theory as creators of money, not as economic institutions with a distinctive role in allocating capital.
The modern field took shape in the 1960s and 1970s, when economists began to model the frictions that make intermediation valuable. A crucial early contribution came from George Akerlof's 1970 analysis of markets with asymmetric information, which showed that when sellers know more than buyers about product quality, markets can break down entirely. Applied to credit, the insight was immediate: a borrower knows more about his own honesty and project quality than a lender does, so lenders must either charge high rates that drive out good borrowers or find ways to screen them. Intermediaries, the argument went, are institutions that specialize in overcoming such information problems.
Around the same time, work on the economics of uncertainty and liquidity provided a complementary rationale. Douglas Diamond and Philip Dybvig's 1983 model of bank runs showed how banks' provision of liquidity—demandable deposits that allow savers to consume early if needed—creates a vulnerability: if depositors fear that others will withdraw, each has an incentive to withdraw early, and the bank can fail even though its assets are sound. This model remains central to understanding both why banks exist and why they are fragile.
A third strand emphasized monitoring. If lenders must oversee borrowers to ensure they do not take excessive risk or shirk, then having many small lenders each monitor the same borrower is wasteful duplication. A bank that collects deposits from many savers and lends to many borrowers can monitor each loan once, on behalf of all depositors. This "delegated monitoring" rationale, formalized by Diamond in 1984, explains why banks—rather than direct lending—dominate in environments where information is hard to verify.
These information-based theories displaced an older view that intermediaries were simply passive transformers of maturities or risk. They also connected the field to broader questions in corporate finance, monetary economics, and macroeconomics. By the 1990s, financial intermediation had become a mature subfield with a well-defined research agenda: model the frictions, derive the optimal intermediary structure, and analyze the consequences for the real economy.
The field is organized less by rival schools than by complementary explanations of why intermediaries exist. These explanations are not mutually exclusive; real intermediaries perform multiple functions simultaneously, and different theories emphasize different functions.
The information-based approach treats intermediaries as solutions to problems of asymmetric information. Three distinct problems matter. Adverse selection occurs before a transaction: lenders cannot distinguish good borrowers from bad ones, so they must charge an average rate that drives out the good. Moral hazard occurs after a transaction: borrowers may take actions that reduce the probability of repayment, and lenders cannot perfectly observe those actions. Costly state verification occurs when a loan is due: the lender must spend resources to determine whether the borrower genuinely cannot pay or is strategically defaulting.
Intermediaries address these problems through specialization. A bank develops screening technology—credit scoring, relationship lending, collateral requirements—that reduces adverse selection. It writes covenants and monitors borrowers to mitigate moral hazard. It exploits economies of scale in verification: one bank verifying a loan on behalf of many depositors is cheaper than each depositor verifying independently. The delegated monitoring model formalizes this last point and shows that intermediaries arise naturally when verification costs are positive and diversification benefits are large.
The information-based approach has been enormously influential, but it has limits. It explains why intermediaries exist but says less about their internal organization—why banks are hierarchical, why they are regulated, why they sometimes fail. It also struggles to explain the coexistence of banks and markets: if intermediaries solve information problems so well, why do firms also issue bonds and equity directly to the public?
A second approach emphasizes the liquidity services that intermediaries provide. Savers face uncertainty about when they will need their funds. If they invest directly in long-term projects, they may be forced to sell early at a loss. A bank can pool many depositors and, by the law of large numbers, predict fairly accurately how many will withdraw at any given time. It can therefore offer demandable deposits—claims that are liquid for the depositor—while investing in illiquid long-term assets. This is maturity transformation: the bank borrows short and lends long.
The Diamond-Dybvig model formalizes this logic and reveals its dark side. Because deposits are demandable and bank assets are illiquid, a coordination failure among depositors can trigger a run: if each depositor believes others will withdraw, each rationally withdraws, and the bank must liquidate assets at fire-sale prices, making even patient depositors worse off. The model shows that deposit insurance or a lender of last resort can prevent runs, but it also shows that such interventions create moral hazard—depositors no longer monitor the bank, and banks may take excessive risk.
This approach complements the information-based view. Banks both solve information problems and provide liquidity; the two functions interact. A bank that monitors borrowers can make illiquid loans; a bank that offers demandable deposits can attract funds cheaply. But the combination also creates fragility, and much of the field's policy analysis concerns how to preserve the benefits of intermediation while limiting its vulnerabilities.
A third tradition, sometimes called the "financial intermediation approach to banking," treats banks not as special because they create money but as special because they intermediate credit. In this view, the distinctive feature of banks is not that their liabilities circulate as means of payment but that they hold a portfolio of loans that is costly for outsiders to evaluate. This approach, associated with the work of Eugene Fama and others in the 1980s, argues that the "specialness" of banks lies in their asset side—their ability to evaluate and monitor borrowers—rather than their liability side.
This perspective has important implications. It suggests that banks and other intermediaries are fundamentally similar: an insurance company that evaluates risks and holds a diversified portfolio is doing something analogous to a bank. It also suggests that the boundary between banks and markets is not fixed; as information technology improves, some intermediation moves from banks to markets, and some moves back. The approach has been criticized for understating the distinctive role of banks in the payments system and in monetary policy transmission.
A fourth approach, which gained prominence after the 2007–2009 global financial crisis, focuses on the aggregate consequences of intermediation. Rather than asking why intermediaries exist, it asks how their behavior affects the macroeconomy. This literature emphasizes that intermediaries are not neutral conduits; their balance sheets, leverage, and risk-taking respond to asset prices and in turn amplify economic fluctuations.
Key concepts include the financial accelerator: when asset prices fall, intermediaries' net worth falls, their ability to lend contracts, and the contraction in lending further depresses asset prices. This mechanism, formalized by Ben Bernanke, Mark Gertler, and Simon Gilchrist, explains how small shocks can have large macroeconomic effects. Related work examines systemic risk—the risk that the failure of one intermediary triggers failures of others through interconnected balance sheets, fire sales, or contagion.
This approach does not replace the microeconomic theories; it builds on them. The financial accelerator requires that intermediaries face information frictions that make their lending depend on their net worth. Systemic risk requires that intermediaries hold illiquid assets and fund them with short-term liabilities, as the liquidity-transformation theories describe. The macroeconomic approach adds a layer of analysis: it takes the existence and fragility of intermediaries as given and studies the aggregate dynamics that result.
Contemporary financial intermediation research is characterized by several durable features. First, the field has become more empirical. The development of detailed datasets on bank balance sheets, loan-level data, and credit registry information has allowed researchers to test the predictions of the theoretical models. This empirical work has generally confirmed the importance of information frictions and liquidity provision, while also revealing that real intermediaries are more complex than the simple models suggest.
Second, the field has expanded beyond banks. The term "shadow banking" describes a system of credit intermediation that operates outside traditional banking regulation—money market funds, securitization vehicles, finance companies, and other entities that perform bank-like functions without bank-like oversight. Research on shadow banking asks whether these entities provide genuine diversification benefits or simply move risk to less transparent corners of the financial system. The 2007–2009 crisis demonstrated that shadow banking could be as fragile as traditional banking, and the field now treats the entire intermediation system—not just banks—as its object of study.
Third, the field has become more policy-oriented. The crisis prompted a wave of research on macroprudential regulation—policies aimed at limiting systemic risk rather than protecting individual institutions. This research asks how capital requirements, liquidity requirements, stress tests, and other tools affect intermediaries' behavior and the stability of the financial system. It draws on all the theoretical approaches described above, combining them into models that can be used for policy analysis.
Fourth, the field has engaged with technological change. Digital payment systems, peer-to-peer lending platforms, and decentralized finance raise questions about whether traditional intermediaries will be disintermediated or whether they will adapt by adopting new technologies. The information-based theories suggest that intermediaries exist because they solve information problems; if technology reduces those problems, the rationale for intermediation weakens. But technology also creates new information problems—cyber risk, algorithmic opacity, new forms of fraud—and it is not yet clear whether the net effect will be more or less intermediation.
Several questions continue to structure research in the field. One concerns the boundary between intermediaries and markets. When do firms borrow from banks rather than issuing bonds? When do savers hold deposits rather than mutual fund shares? The answer appears to depend on the severity of information frictions, the legal and regulatory environment, and the state of technology. But the field lacks a fully general theory that predicts the equilibrium mix of bank and market finance.
A second question concerns the optimal regulation of intermediaries. Banks are fragile because they provide liquidity; deposit insurance prevents runs but encourages risk-taking; capital requirements limit risk-taking but may reduce lending. The field has made progress in understanding these trade-offs, but the optimal regulatory design remains contested. Different models imply different answers, and the empirical evidence is not yet decisive.
A third question concerns the relationship between financial intermediation and economic growth. Do deeper financial systems cause faster growth, or do they merely accompany it? The evidence suggests a positive association, but the direction of causation is difficult to establish, and some research indicates that financial development beyond a certain point may be associated with slower growth or greater instability.
A fourth question concerns the stability of the intermediation system itself. The 2007–2009 crisis showed that the system could fail in ways that the standard models did not predict. Research since then has focused on understanding how risks build up in good times, how they propagate through interconnected balance sheets, and how regulation can make the system more resilient without stifling its economic functions.
Financial intermediation remains a vibrant subfield because its subject matter is both intellectually rich and practically important. The theories developed over the past half century—information economics, liquidity provision, delegated monitoring, financial amplification—provide a coherent framework for understanding why intermediaries exist and what they do. The empirical and policy work of the past two decades has tested and refined those theories, revealing both their power and their limits. For an educated newcomer, the field offers a clear central question—why do intermediaries exist, and what are the consequences?—and a set of complementary answers that together explain much of the structure of modern financial systems.