Because banks lend out most deposited money, they are never in a position to return all deposits simultaneously. This is not fraud — it is the designed structure of fractional reserve banking. But it creates a specific vulnerability: the bank run.
A bank run occurs when a large number of depositors, fearing a bank may fail, all attempt to withdraw at the same time. The irony is that this fear is self-fulfilling. A bank that would otherwise be perfectly solvent — with good loan assets and solid long-term finances — can be destroyed within days simply because too many people demand cash simultaneously. The bank cannot liquidate its loan portfolio fast enough. It runs out of cash. It fails.
The most famous example is the bank runs of the Great Depression. Between 1930 and 1933, roughly 9,000 US banks failed, often in cascades: the failure of one bank sparked fears about others, triggering runs on previously healthy institutions. The economic consequences were catastrophic.
Modern economies have two main defences. First: deposit insurance. The UK's FSCS guarantees deposits up to £85,000 per person per bank. Since most ordinary depositors are fully covered, they have no reason to panic and withdraw — removing the trigger for the run.
Second: the lender of last resort function. If a bank faces a liquidity crisis — good assets but cannot convert them to cash quickly enough — the central bank can provide emergency loans. The Bank of England and Federal Reserve played this role extensively in 2008.
Even with these safeguards, bank runs still happen. Silicon Valley Bank's failure in 2023 was essentially a modern bank run, dramatically accelerated by social media and digital banking that allowed depositors to move money in seconds. Modern technology has made bank runs faster and harder to stop than ever before.