2022 · Economics

Bank runs: how a rumour can topple a healthy bank

Q: What does it mean to say a bank performs 'maturity transformation'?

It turns short-term, withdraw-anytime deposits into long-term loans. A bank takes in deposits that can be pulled out at any moment and uses them to fund loans that take years to repay. Bridging that mismatch in timing is what maturity transformation means, and it is both the bank's core service and the root of its fragility.

Q: In the Diamond-Dybvig model, why is a bank run described as self-fulfilling?

Because the expectation that others will withdraw makes withdrawing your own money the rational choice, which drains the bank. The model has two equilibria. If you believe everyone else will run, you should run too, since the bank pays people in order until the cash is gone. Everyone acting on that belief is exactly what empties a solvent bank. The fear creates the very outcome that was feared.

Q: What did Bernanke's research show about the bank failures of the early 1930s?

They destroyed banks' knowledge of borrowers, raising the cost of credit and deepening the Depression. Bernanke argued the purely monetary story was too small to explain so long a slump. When banks collapsed, the hard-won information about which borrowers were creditworthy was lost, credit became scarce and costly, and a recession was driven into a decade-long depression.

Awarded to Ben Bernanke, Douglas Diamond and Philip Dybvig “for research on banks and financial crises”.

What was the 2022 Nobel Prize in Economics awarded for?

The 2022 Economics prize honours the work that explained what banks really do and why they can fail so suddenly. Douglas Diamond and Philip Dybvig built a simple model showing that a bank funds long-term loans with deposits people can pull out at any moment, which is useful but leaves it open to a self-fulfilling panic. Ben Bernanke showed, using the Great Depression, that when banks collapse the damage spreads to the whole economy and turns a downturn into a catastrophe.

Predict first

A bank has made good loans and is perfectly solvent. Not one loan has gone bad. Yet on Monday morning it collapses. What could have destroyed it overnight?

A bank run. The bank's loans are sound but tied up for years, while its deposits can be withdrawn instantly. If depositors come to believe everyone else is about to pull their money out, each one races to withdraw first, and the rush itself empties the bank. The panic, not any bad loan, is what causes the collapse. This is the fragility Diamond and Dybvig modelled.

Predict first

If a bank run can destroy even a healthy bank, why do rich countries almost never see them today?

Deposit insurance. When the government guarantees your savings, you have no reason to rush to the bank at the first rumour, so the panic never gets started. The guarantee works precisely by rarely needing to be used. That single policy idea, drawn from Diamond and Dybvig's 1983 model, is why bank runs went from common to rare.

A bank borrows short and lends long. That is useful, but if every depositor fears the others will withdraw, all rush at once and even a sound bank fails. A government guarantee removes the reason to rush, so the run never starts.

Imagine a town where everyone keeps their savings at one bank. It does not leave that money sitting idle. It lends most of it to families buying homes and shops that want to grow, and those loans take years to pay back.

Here is the catch. You can take your money out whenever you want, but the bank's money is tied up in long loans it cannot call back quickly. If only a few people withdraw at a time, all is well. The danger is everyone trying to pull their money out at once.

Why a run feeds itself

Fear becomes the cause

If you hear a rumour that the bank is in trouble, the smart move is to grab your money before it runs out. But everyone thinking that way is what empties the bank. The panic makes itself come true, even if the bank was healthy.

The fix is simple. If the government promises your money back no matter what, you have no reason to rush, so the panic never starts. That promise is called deposit insurance, and it is one big reason bank runs are far rarer today.

Banks perform a trick economists call maturity transformation. Savers want their money to stay liquid, ready to withdraw on demand. Borrowers want the opposite: long loans they can repay slowly over years. A bank sits between the two, taking in short-term deposits and turning them into long-term loans. Diamond and Dybvig's 1983 paper showed that an institution shaped exactly like a bank is the efficient way to satisfy both sides at once.

This service is valuable, but it builds a fragility into the bank's very structure. Its assets, the loans, are illiquid and locked up for years, while its liabilities, the deposits, can be demanded back instantly. The bank never holds enough cash to pay everyone at the same moment. It does not need to, as long as withdrawals stay normal.

Two equilibria

The run is a self-fulfilling prophecy

Diamond and Dybvig showed the model has two stable outcomes. In the good one, only depositors who genuinely need cash withdraw, and the bank works normally. In the bad one, depositors who fear a run rush to withdraw first, which drains the bank and forces it to dump loans at a loss. Because each depositor's best move depends on what the others do, the panic can be set off by rumour alone. The expectation of a run is what causes the run.

What turns a worry into a collapse

A rumour spreads that the bank may not be able to pay everyone.
Depositors know the bank serves people in order, first come first served, so latecomers may get nothing.
Each saver races to withdraw before the cash runs out.
The bank is forced to sell long-term loans early and at a loss, and it fails even though it was solvent.

Diamond and Dybvig also showed how to break the cycle. Government deposit insurance guarantees savers their money back, which removes any reason to join a run. In theory the guarantee is so effective that it rarely has to be paid out: knowing it exists is enough to keep everyone calm. This is why nearly every country now insures bank deposits, and it is a central part of modern bank regulation.

The Diamond-Dybvig model formalises maturity transformation as an insurance problem. Depositors face private, idiosyncratic liquidity shocks: some will need to consume early, some late, and none knows in advance which type they will be. A long investment pays well if held to maturity but little if interrupted. A bank offering a demand-deposit contract can pool this risk, paying early withdrawers more than the liquidation value of the asset and late withdrawers less than the full maturity return. Every depositor then gets better risk sharing than they could obtain by investing alone, which is the precise sense in which banks create liquidity.

Multiple equilibria

Why one contract has both a calm state and a run state

The demand-deposit contract that delivers efficient risk sharing also admits a second Nash equilibrium. If a depositor expects everyone else to withdraw, the bank will be liquidated, so withdrawing now is the only way to recover anything. If she expects others to wait, waiting is best. Both expectations are self-confirming. The run equilibrium is driven by a shift in beliefs unconnected to the bank's fundamentals, so a solvent bank can be destroyed purely by a coordination failure among its own depositors.

Sequential service

First come, first served makes the race rational

What makes running individually rational is the sequential service constraint: the bank pays depositors in the order they arrive until its liquid resources are exhausted, so latecomers risk getting nothing. Suspending convertibility, freezing withdrawals once a threshold is hit, can halt a run, but Diamond and Dybvig showed it is not optimal when the true demand for cash is uncertain. Government-backed deposit insurance dominates it: by guaranteeing that the late withdrawer is made whole, it destroys the incentive to run in the first place and selects the good equilibrium.

Diamond and Dybvig explained why a single bank is fragile. Bernanke's 1983 paper, 'Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression', showed why a wave of bank failures wrecks the whole economy. The dominant account at the time, from Friedman and Schwartz, was monetary: bank failures shrank the money supply. Bernanke argued that this channel alone was too small and too short-lived to explain a downturn that ran from 1929 to 1933 and lingered for years afterward.

The credit channel

Banks are hard-to-replace information machines

Bernanke's insight was that a bank is not just a holder of deposits but a low-cost credit intermediary that gathers costly knowledge about which local borrowers are creditworthy. When thousands of banks collapsed, that information capital was destroyed and could not be rebuilt quickly. The real cost of channelling savings to sound borrowers rose, credit to farms and firms became scarce, and output fell further. Using statistical analysis and historical records, Bernanke showed that bank failures had real, nonmonetary effects that turned a severe recession into the longest depression in modern history.

Together the three laureates laid the foundation of modern bank regulation. The same logic reaches well beyond traditional banks: in the early 2000s, lightly regulated shadow banks performed maturity transformation outside the regulated sector, and runs on them sat at the centre of the 2008 to 2009 crisis. Diamond and Dybvig's model still describes those runs, and Bernanke, by then chair of the US Federal Reserve, acted on his own research to keep credit flowing through it.

Worth knowing

The economist who later had to use his own theory

Ben Bernanke spent the early 1980s proving that collapsing banks turned the slump of the 1930s into the Great Depression. In 2008, as chair of the US Federal Reserve, he faced almost exactly the crisis he had studied and used those lessons to keep credit flowing and stop a repeat. Few researchers ever get to apply their own findings on so large a stage.

Check yourself

What does it mean to say a bank performs 'maturity transformation'?

Why: A bank takes in deposits that can be pulled out at any moment and uses them to fund loans that take years to repay. Bridging that mismatch in timing is what maturity transformation means, and it is both the bank's core service and the root of its fragility.

In the Diamond-Dybvig model, why is a bank run described as self-fulfilling?

Why: The model has two equilibria. If you believe everyone else will run, you should run too, since the bank pays people in order until the cash is gone. Everyone acting on that belief is exactly what empties a solvent bank. The fear creates the very outcome that was feared.

What did Bernanke's research show about the bank failures of the early 1930s?

Why: Bernanke argued the purely monetary story was too small to explain so long a slump. When banks collapsed, the hard-won information about which borrowers were creditworthy was lost, credit became scarce and costly, and a recession was driven into a decade-long depression.

Key terms

Maturity transformation: Funding long-term loans with short-term deposits. The bank borrows money that can be withdrawn at any moment and lends it out for years, bridging the gap between what savers and borrowers want.
Bank run: A situation where many depositors try to withdraw their money at the same time, usually because they fear the bank will run out of cash. The rush can drain and topple even a solvent bank.
Self-fulfilling equilibrium: An outcome that happens only because people expect it to. In a bank run, the belief that others will withdraw makes withdrawing rational for everyone, so the feared collapse actually occurs.
Deposit insurance: A government guarantee that depositors will get their money back even if the bank fails. Because savers no longer need to rush to withdraw, the guarantee prevents runs and rarely has to be paid out.
Liquidity: How easily an asset can be turned into spendable cash without losing value. Deposits are highly liquid; long-term loans are not.
Credit intermediation: The work banks do to channel savings to borrowers, including judging who is creditworthy. Bernanke showed that when banks fail this service is lost, making credit scarce and costly.

The laureates

Ben Bernanke

The Brookings Institution, Washington, D.C., USA

Bernanke's 1983 study of the Great Depression used statistics and historical records to show that the wave of bank failures in the early 1930s was not merely a symptom of the slump but a cause of it. When banks collapsed, the knowledge they held about local borrowers was lost, credit dried up, and a sharp recession became a decade-long depression. Born 1953 in the USA.

Photo: United States Federal Reserve, Public domain (via Wikimedia Commons)

Douglas Diamond

University of Chicago, Chicago, IL, USA

With Philip Dybvig, Diamond built the 1983 model that explained why banks exist, why they are fragile, and how deposit insurance protects them. In a 1984 paper he also showed how banks serve society as delegated monitors, checking that borrowers honour their loans. Born 1953 in the USA.

Photo: Office of President Joe Biden, Public domain (via Wikimedia Commons)

Philip Dybvig

Washington University, St. Louis, MO, USA

With Douglas Diamond, Dybvig co-authored the 1983 paper 'Bank Runs, Deposit Insurance, and Liquidity', which showed how a bank run can be a self-fulfilling panic that destroys even a solvent bank, and how a government guarantee can stop it. Born 1955 in the USA.

Photo: US Embassy Sweden, CC BY 2.0 (via Wikimedia Commons)

Sources

Facts are pinned from the official Nobel Prize API. The explanations were written from these sources:

Your notessaved

← Back to all prizes