A topic in the Open Knowledge Graph — a free, open map of 15,495 topics and the order to learn them in.
Banks fund long-term, hard-to-sell loans with deposits that customers can withdraw at any time, so they keep only a fraction of deposits as reserves. This maturity transformation is useful, but it makes banks fragile: if many depositors withdraw at once, even a sound bank cannot pay everyone without selling assets at a loss. Because each depositor gains by withdrawing before others do, fear of a run can cause a run. Deposit insurance and central bank lending to solvent but illiquid banks remove the reason to run, at the cost of weakening depositors' incentive to watch how banks take risks.
Set up a small bank balance sheet with 10% reserves and 90% loans that sell at a discount if sold early. Ask what happens if 5% of depositors withdraw, then 40%. Then frame the depositor's choice as a game with two equilibria, "everyone waits" and "everyone runs," and show how a government guarantee removes the run equilibrium.
From money creation you know that banks keep only a fraction of deposits as reserves and lend the rest. That arrangement does something valuable. Savers get deposits they can withdraw whenever they like, and borrowers get loans that last years. The bank stands between them, turning short-term, liquid claims into long-term, illiquid ones. This is called maturity transformation.
The same arrangement creates a weakness. On a normal day only a few depositors withdraw, and reserves cover them. But loans cannot be sold quickly at full value. If many depositors come at once, the bank must sell loans at a discount, and it may run out of cash before everyone is paid. Depositors know this. If you think others are about to withdraw, your best move is to withdraw first. So fear of a run can cause one. Douglas Diamond and Philip Dybvig formalized this in a 1983 model with two equilibria: one where everyone waits and the bank is fine, and one where everyone runs and the bank fails. Nothing about the bank's loans needs to change to move from one to the other. Diamond and Dybvig shared the 2022 Nobel Memorial Prize in Economic Sciences with Ben Bernanke, who studied how bank failures deepened the Great Depression.
Runs can spread. When one bank fails, depositors at similar banks may fear the same, and banks selling assets at once push prices down for everyone. In the early 1930s, waves of runs closed thousands of American banks, which shrank the money supply and credit.
Two policies address the problem. The first is deposit insurance. In the United States, the Banking Act of 1933 created the Federal Deposit Insurance Corporation, which today insures deposits up to $250,000 per depositor, per insured bank, for each account ownership category. Insured depositors have no reason to rush, so the run equilibrium largely disappears. The second is a lender of last resort: a central bank that lends cash to a bank facing withdrawals, accepting its loans as collateral. Walter Bagehot's 1873 advice was to lend freely, at a high rate, against good collateral.
Both tools have costs. Guarantees reduce depositors' incentive to watch how banks invest, a form of moral hazard, so insured banks face capital rules and supervision. Uninsured deposits can still run. In March 2023, depositors at Silicon Valley Bank, most of whose deposits exceeded the insurance limit, tried to withdraw tens of billions of dollars in a single day, and regulators closed the bank.