How Banking Panics Led to Central Banking

The Bank of England was chartered in 1694 not because anyone had a theory of central banking. It was chartered because the English government was broke and needed to borrow money to fight Louis XIV’s France. A syndicate of wealthy merchants agreed to lend the Crown £1.2 million at 8% interest, in exchange for a charter to operate a joint-stock bank with the right to issue banknotes. The arrangement was a fiscal expedient, not an institutional design. Nobody at the founding imagined that this convenient financial arrangement would evolve, over the next two centuries, into the model for a new type of institution — the lender of last resort — that every major economy would eventually adopt as a structural necessity. They were just trying to fund a war.

This origin story is typical of central banking. The institution was not designed from first principles to solve a known problem. It was improvised in response to a series of crises, each exposing a flaw, each improvised response creating a precedent that hardened into permanent institutional design. The history of central banking is an evolutionary history — a sequence of crises and adaptations that slowly converged on an understanding of what the banking system requires to remain stable, purchased at enormous cost in panics, failures, and depressions.

The inherent instability of fractional reserve banking is not a defect that bad bankers created. It is a structural property of the system. A bank takes deposits from a large number of depositors who expect to be able to withdraw their funds on demand. The bank lends most of those funds to borrowers on longer-term contracts. The gap between the maturity of the bank’s liabilities (short-term deposits, withdrawable on demand) and the maturity of the bank’s assets (longer-term loans, not immediately convertible to cash) is called maturity transformation, and it is both the economic function and the Achilles heel of banking. The function: the bank creates liquidity, converting long-term illiquid investments (loans) into short-term liquid claims (deposits) that the economy needs. The Achilles heel: if enough depositors simultaneously demand their money back, the bank cannot pay them all, because the loans it made cannot be immediately recovered. This is a run, and it can destroy a perfectly solvent bank — a bank that holds good loans worth more than its deposits — if the run happens fast enough.

The Bank of England discovered its lender-of-last-resort function not through theoretical reasoning but through repeated confrontations with the practical consequences of bank runs. The financial crises of 1720 (the South Sea Bubble), 1763, 1772, 1793, 1825, 1847, 1857, and 1866 each exposed the Bank to the same dynamic: a shock causes doubts about the solvency of some institutions, doubts cause depositors to withdraw funds from all institutions as a precaution, withdrawal demands cannot be met across the banking system simultaneously, solvent banks fail alongside insolvent ones, and the contraction of credit causes a severe economic downturn.

The Bank’s response in each crisis evolved over time. In the early crises, the Bank often contracted credit during a panic, trying to protect its own reserve position by lending less at exactly the moment the market needed more lending. This contractionary response accelerated the panic: when the lender that the market depended on withdrew, the shortage of liquidity became catastrophic. The lesson, learned slowly and painfully, was that the institution with the deepest reserves needed to do the opposite — to lend aggressively when everyone else was withdrawing, to be the buyer of last resort when everyone else was selling.

Walter Bagehot articulated this understanding in “Lombard Street” (1873), the foundational text of central banking doctrine. His prescription was precise: in a panic, the central bank should lend freely, at penalty rates, against good collateral. Lending freely prevents solvent institutions from being destroyed by liquidity shortages. Penalty rates prevent the facility from subsidizing institutions that should fail. Requiring good collateral prevents it from becoming a bailout for the genuinely insolvent. Bagehot was systematizing a practice discovered through experience rather than derived from theory, and the doctrine spread as every financial system that lacked a lender of last resort experienced the same destructive runs that had plagued Britain.

The United States resisted this institutional development longer than any other major economy, with predictably costly consequences. The American political tradition was deeply hostile to centralized financial power. The First Bank of the United States (1791-1811) was allowed to lapse when its charter expired, after the political coalition that included small farmers and state bank interests successfully argued that a central bank was unconstitutional and served the interests of financial elites at the expense of ordinary citizens. The Second Bank of the United States (1816-1836) was destroyed by Andrew Jackson, who vetoed its recharter in 1832 in one of the most consequential acts of deliberate financial institutional vandalism in American history.

What followed was the era of “free banking” — a period in which banking was lightly regulated at the state level, with no federal supervision and no lender of last resort. The banking panics that resulted were not occasional misfortunes; they were a regular feature of the American economic landscape. Major banking panics occurred in 1837, 1857, 1873, 1884, 1893, and 1907. Each panic followed the same pattern: a shock (typically the failure of a large institution or a decline in commodity prices) triggered depositor withdrawals, which spread through the system, collapsing banks that were fundamentally solvent but illiquid, contracting credit sharply, and producing a recession or depression.

The 1873 panic is instructive for understanding the severity of the institutional gap. The failure of Jay Cooke & Company, a major investment banking house that had overextended itself in Northern Pacific Railroad financing, triggered a cascade of bank failures and a contraction so severe that the New York Stock Exchange closed for ten days — the first closure in its history. The contraction of credit that followed produced a depression that lasted, with varying intensity, until 1879. There was no institution that could step in as lender of last resort, because none existed. The market had to clear through bank failures, business bankruptcies, and a grinding contraction of economic activity.

The Panic of 1907 finally broke the political resistance to creating a lender of last resort. The panic began in October 1907 with a failed attempt to corner the copper market that triggered runs on the Knickerbocker Trust Company and other trust companies in New York. Trust companies were a category of financial institution that sat outside the existing clearing house system and therefore had no access to the informal mutual support arrangements that clearing house member banks could provide during a crisis. When the Knickerbocker failed, the contagion spread rapidly.

The panic was arrested only because J.P. Morgan — the most powerful private banker in the United States, then 70 years old — personally organized a rescue operation. Morgan gathered the leading New York bankers in his library, assessed which institutions were solvent and which were not, organized the solvent ones to support each other, refused to provide funds to the insolvent ones, and effectively served as a private lender of last resort for the New York financial system. The country’s financial system survived the panic because one private individual happened to have the financial stature, the organizational capacity, and the personal authority to coordinate a response.

This was recognized at the time as both admirable and completely unsatisfactory as a permanent institutional arrangement. The financial stability of the United States could not be allowed to depend on the continued life and willingness to act of a single elderly private citizen. Morgan died in 1913. The Federal Reserve was established in 1913. The connection between these two facts is not coincidental. The Federal Reserve Act that created the Fed was the direct institutional response to the recognition that the 1907 panic exposed: a modern financial system requires a public institution performing the lender-of-last-resort function, because no private institution can perform it reliably and no private individual can perform it at all over time.

The Federal Reserve’s first major test came in 1930-1933, and it failed catastrophically. The Fed, which had been created to prevent bank panics, allowed one of the worst banking panics in American history to unfold without providing the lender-of-last-resort support that Bagehot’s doctrine prescribed. Between 1930 and 1933, roughly 9,000 banks failed. The money supply contracted by roughly a third. The resulting economic contraction was the Great Depression. Milton Friedman and Anna Schwartz’s “A Monetary History of the United States” made the case that the Depression’s severity was primarily a monetary phenomenon — the Fed’s failure to act as lender of last resort during the banking panics of 1930-1933 turned a recession into a catastrophe.

The institutional lessons from this failure were substantial. Deposit insurance — created by the Banking Act of 1933 — was the structural fix for the run problem: if depositors know their deposits are guaranteed by the government up to a specified amount, they have no reason to run on a bank even if they suspect it may have losses. A run by an individual depositor makes sense only if others will run first and clean out the bank before you get there; deposit insurance eliminates this first-mover incentive for small depositors. Bank regulatory supervision — the framework of capital requirements, examination, and resolution authority established through the 1930s — was designed to make bank failures less likely by requiring banks to hold more capital relative to their risk and by giving regulators the authority to identify and resolve troubled institutions before they collapsed.

The architecture of modern banking regulation — the combination of lender-of-last-resort central banking, deposit insurance, and prudential supervision — is a direct product of the accumulated experience of banking panics from the 18th through the 20th centuries. Each element addresses a specific failure mode identified in previous crises. The edifice is not elegant, and it contains internal tensions (deposit insurance creates moral hazard that prudential supervision is supposed to counteract; lender-of-last-resort lending creates moral hazard that penalty pricing is supposed to counteract). But it has succeeded in preventing the classic contagious bank run from producing Depression-level contractions in economies with well-developed versions of the system.

The 2008 financial crisis revealed both how far the regulatory architecture had evolved and how far it remained from addressing the inherent instability of the broader credit system. The classic bank run — depositors lining up to withdraw their savings — did not occur in 2008, because deposit insurance worked. What occurred was a modern equivalent: a run on wholesale funding markets, in which financial institutions that had borrowed short-term in repo markets and commercial paper markets suddenly found that their lenders would not roll over the loans. Shadow banking institutions — investment banks, money market funds, structured investment vehicles — performed the maturity transformation function of traditional banking without the regulatory protections that traditional banking carried.

The Federal Reserve’s response in 2008, under Ben Bernanke who had spent his academic career studying the Fed’s failures in 1930-1933, was to deploy the lender-of-last-resort function more aggressively and creatively than at any point in the institution’s history. The Fed lent to institutions it had no prior authority to lend to, against collateral it had never previously accepted, at rates it had never previously offered. The interventions were improvised under extreme time pressure. They were also largely effective at preventing the financial system collapse from producing a depression on the 1930s scale, even though the recession was the worst since the 1930s.

The recurring pattern across three centuries of banking history is that the inherent instability of fractional reserve banking eventually produces crises that private actors cannot resolve. When those crises occur, they create irresistible pressure for institutional innovation — for public institutions that can do what private institutions cannot: provide unlimited liquidity during a panic without exhausting their own resources, because their resources are ultimately backed by the government’s power to create money. The pattern does not suggest that better banking regulation will eventually eliminate crises. It suggests that the institutional response to crises is always lagged — new instruments create new vulnerabilities that the existing regulatory architecture doesn’t cover — and that the recurring cycle of crisis and institutional adaptation is a permanent feature of financial systems rather than a correctable defect. The next crisis will reveal a new dimension of the maturity transformation problem that the current architecture doesn’t adequately address. The institutional response to that crisis will become the next chapter in the history of central banking.