Bank runs

A bank promises you can withdraw your money any time. Fractional-reserve banking means it doesn’t actually keep enough to make good on that promise for everyone at once. Usually that’s fine — until the day everyone asks at the same time. Then you get a bank run.

The mechanism

Recall the goldsmith who lent out most of the gold in the vault. If depositors hear a rumour the bank is in trouble, the rational move for each of them is to rush and withdraw before the reserves run out. But that rush is precisely what empties the reserves. A run can topple a perfectly healthy bank, because no fractional-reserve bank can pay all depositors simultaneously — the money isn’t there; it’s been lent out.

“It is not enough to be solvent. A bank must also be liquid — and a run turns a solvency problem into a liquidity problem in an afternoon.”

A self-fulfilling panic

Bank runs are a coordination trap. If you believe others will withdraw, you must withdraw too, even if you think the bank is sound — because the last in line gets nothing. Fear alone can bankrupt an institution. In the Great Depression, thousands of American banks failed this way, wiping out ordinary people’s savings and deepening the collapse.

How the modern system answers it

• Deposit insurance — the state guarantees your deposits up to a limit, so you’ve no reason to run.
• A lender of last resort — the central bank stands ready to print and lend to any solvent bank facing a run.

These backstops work — runs are rarer now — but look closely at the cost. Both rely on the power to create money without limit: to guarantee everyone and to lend infinitely, the central bank must be able to conjure currency at will. The cure for the fragility of fractional reserves is more fiat.