The lender of last resort is the institutionalised answer to the bank run. Walter Bagehot’s 1873 formulation remains the reference: in a panic, lend freely, at a penalty rate, against collateral that would be good in normal times. The point is not to save any particular firm but to collapse the bad equilibrium — if cash is always available, there is no reason to queue for it.
The idea is far older than the Bank of England’s mastery of it. In AD 33, Tiberius answered a Roman credit seizure by distributing a hundred million sesterces through senatorial banks as interest-free loans — a liquidity injection recognisable to any modern central banker. The Panic of 1907 is the counter-example that built an institution: with no central bank, the United States depended on J.P. Morgan personally locking financiers in his library to arrange support. The Federal Reserve Act followed within six years.
Each element of Bagehot’s dictum does distinct work. Lending freely stops the panic; good collateral separates the illiquid from the insolvent; the penalty rate limits moral hazard by making last-resort money unattractive in ordinary times. In 2008 the doctrine was stretched to its modern limit — extended past banks to dealers, money funds, and commercial paper, against collateral whose “normal-times” value was precisely what the crisis had put in doubt.
The standing difficulty is that the illiquidity/insolvency line, crisp in the textbook, is unreadable in real time. Every lender of last resort eventually discovers it is also an appraiser of last resort.