What is a bank run?

The short answer

A bank run occurs when many depositors try to withdraw their money at once because they fear the bank may fail. The fear can be self-fulfilling: the bank may hold long-term, illiquid assets, so it cannot convert them into cash quickly enough to pay everyone.

A run can happen to a solvent bank. The problem is liquidity, not necessarily solvency.

The coordination problem

Diamond and Dybvig's classic model shows why. Depositors have claims that are liquid, but the bank's assets are illiquid. If everyone leaves their money in the bank, the bank can hold the assets to maturity and pay everyone. If everyone tries to withdraw, the bank must sell assets at fire-sale prices, and some depositors may not be paid in full.

Each depositor's best response depends on what they expect others to do. If you believe others will run, you run first. The expectation becomes the outcome.^[1]

Liquidity versus solvency

• Solvency means assets are worth more than liabilities.
• Liquidity means the bank can meet obligations as they fall due.

A bank can be solvent but illiquid. It may hold good loans that will be repaid, but it cannot sell them quickly without a loss. A run can force it into failure even though its assets would have covered its liabilities over time.

A bank can also be insolvent but still open, if losses have not been recognised. That is why supervision and transparent accounting matter.

What stops runs

• Deposit insurance. If depositors know their insured balances are safe, they have less reason to run.
• Lender of last resort. The central bank can lend against good collateral to a solvent but illiquid bank.
• Capital and liquidity rules. Buffers reduce the probability that a shock becomes a run.
• Resolution regimes. Clear rules for winding down a failing bank can reduce uncertainty.
• Communication and coordination. Deposit guarantees and credible backstops can change expectations.

These tools reduce run risk, but they can create moral hazard. If creditors expect to be protected, they may demand less compensation for risk. That is why insurance and liquidity support are usually paired with supervision and capital requirements.^[2]

Modern runs

Runs do not only happen at branch counters. Wholesale funding can run quickly: money market funds, repo lenders, and commercial paper investors can refuse to roll over funding when they are uncertain about collateral or solvency. Digital banking can also accelerate retail withdrawals.

The Federal Reserve's financial stability work monitors funding risks and bank resilience, but the underlying mismatch between liquid liabilities and illiquid assets never disappears entirely.^[3]

The bottom line

A bank run is a coordination failure. It is not simply a sign that a bank is badly run, although poor management can increase the risk. It is a structural feature of a system that finances long-term assets with short-term liabilities. Deposit insurance and central-bank liquidity are attempts to make that structure safer without eliminating it.