What is a bank run, and how can it feed a crash?

What is a bank run in plain words?

A bank run happens when many depositors withdraw their money at once because they fear the bank will fail soon. The withdrawals themselves can cause the failure, so a run can become a self-fulfilling prophecy. A rumour is enough to start one, even when the bank is sound. Depositors in a bank run act on expectations, so the same bank can look safe on Monday and fragile on Tuesday.

The word run describes the behaviour: people hurry to be among the first to be paid. Early withdrawers receive their full balance, and late ones may receive less. That unequal outcome explains why a calm person may still join the queue, since waiting carries a cost.

Why can no bank pay every depositor in full on the same day?

Banks keep only a small share of deposits as cash and lend or invest the rest, often on longer terms than the deposits themselves. A home loan lasts years, while a deposit can leave in seconds. This design lets banks fund the economy, and it creates a gap between what customers may demand and what the bank holds in cash.

In normal times the gap stays invisible, because withdrawals and new deposits roughly balance. A run breaks the balance. When requests exceed cash, the bank must sell assets quickly, often at a loss, and each sale lowers its capital. The loss then adds to the fear that started the run. Supervisors therefore watch liquidity, meaning the cash and quickly saleable assets that a bank can reach in a crisis.

What starts a bank run?

Three triggers recur. The first is fear of insolvency, whether real or based on a false rumour. The second is imitation: once a run starts, depositors copy each other, and joining becomes rational. The third is contagion, where the collapse of one bank triggers runs on others.

Contagion explains why regulators treat a single failure as a system question. Depositors at healthy banks look at the failed bank, see a similar business, and move their money. The speed of the process has grown with technology. At Silicon Valley Bank, customers withdrew $42 billion in a single day, nearly a quarter of the deposits.

How does a bank run unfold, stage by stage?

A run follows a recognisable path. Doubt appears, first withdrawals arrive, and a queue forms, physical or digital. The bank sells assets to raise cash, and its capital shrinks. At the end of the path, a supervisor closes the bank or an acquirer takes it over, and depositors regain access. Each stage raises the pressure on the next one, so early action by a bank or a supervisor matters more than late action.

The final stage is usually quick and quiet. In the United States, the FDIC keeps takeovers confidential and reopens branches under new ownership on the next business day. The 2023 case followed this template: the FDIC announced that insured depositors would have full access by Monday morning, 13 March 2023.

How can a bank run feed a stock market crash?

A run turns a fall in prices into a shortage of money. Federal Reserve History describes banking panics from fall 1930 to winter 1933, during which the US money supply fell by nearly 30%. Investors facing a credit squeeze sell shares to raise cash, and the sales push the stock market lower.

The costs are large. For systemic banking crises between 1970 and 2007, the Wikipedia summary reports average fiscal costs of 13% of GDP and output losses of about 20% of GDP over the first 4 years. Not every run reaches that scale, but the numbers explain why authorities react fast. The wider picture of causes and history sits in the stock market crash hub.

What stops a bank run before it spreads?

Deposit insurance protects each depositor up to a set amount, which removes the reason to withdraw simply because others do. In the United States the FDIC standard amount is $250,000 per depositor, per insured bank, for each ownership category, and the FDIC has existed since 1933. Central bank lending and a fast takeover add further layers.

No safeguard is complete. Depositors may still fear losing access during a bank reorganisation, and deposits above the cap stay exposed. About 89% of deposit liabilities at Silicon Valley Bank exceeded the FDIC maximum, which helps explain the speed of that run. The EQOS Asia method page shows how a resilience score summarises such limits.

Which funding signals show that bank run risk is building?

Look at the funding side. A high share of uninsured deposits, a concentrated customer base, and large unrealised losses on assets all raise the odds that confidence breaks quickly. Check these items in a bank’s public reports before you choose where to hold operating cash. A short list of three banks and their funding profiles gives most of the insight for a small company.

Then size your own exposure. List balances by bank, mark the amounts above the insured limit, and set a ceiling per counterparty. The EQOS Asia pricing page lists the diagnostics that map such exposure for companies in Singapore, ASEAN, and Hong Kong. This page is general information, not investment advice.

Frequently asked questions

About the author

Stéphane Jambu

President, EQOS Forecast Asia

Stéphane Jambu leads EQOS Forecast Asia and works as a consultant in e-commerce, search visibility, and AI. LinkedIn profile