Manias, Panics, and Crashes — The Anatomy of a Bubble

Why Bubbles Keep Happening

If markets were efficient and investors rational, asset bubbles shouldn’t exist. Prices would never deviate far from fundamental value because any deviation would be immediately corrected by rational arbitrageurs. Yet bubbles are a consistent feature of financial history — tulips in 1637, South Sea Company in 1720, railroad mania in the 1840s, dot-com stocks in the 1990s, US housing in the 2000s — each with its own specific story and each following a recognizable pattern.

Charles Kindleberger’s Manias, Panics, and Crashes (1978) is the canonical account. Drawing on Hyman Minsky’s economic framework and documented financial history back to 1618, Kindleberger laid out a model of financial crises that reads like a template rather than a history. The reason the template fits so many episodes is that the underlying mechanism — how credit expansion and psychological feedback loops amplify price movements until reversal becomes inevitable — is driven by features of human psychology and financial systems that don’t change between episodes.

The Minsky-Kindleberger Model

The crisis cycle begins with a displacement — an exogenous change that creates new profit opportunities. It might be a technological innovation (railways, the internet), a policy shift (financial deregulation, a change in monetary policy), a new market opening, or a favorable macroeconomic shock. The displacement changes the profit expectations for some class of assets.

This triggers a boom. Investment pours into the favored sector. Credit expands — banks lend more, new credit instruments emerge to accommodate demand, collateral values rise as asset prices rise, enabling more borrowing, which pushes prices higher. The early investors make money, attracting more investors. The feedback loop between rising prices, expanding credit, and increasing investment is self-reinforcing.

Minsky’s key contribution was the financial instability hypothesis: over a boom, the character of borrowing shifts. Early in the expansion, borrowers are hedge financed — their expected cash flows cover both interest and principal. As the boom continues and expectations become more optimistic, borrowers shift to speculative finance — cash flows cover interest but not principal, so debts must be rolled over. Finally, at the peak, borrowers are Ponzi financed — cash flows don’t cover even interest; the borrower is counting on rising asset prices to cover the shortfall. The Ponzi position requires continued price appreciation just to stay solvent.

The shift toward Ponzi finance is invisible during the boom because rising prices keep everyone solvent. It only becomes visible when prices stop rising.

Euphoria, Overtrading, and the Madness

Kindleberger describes a characteristic phase of euphoria — a period when the normal rules seem suspended, skeptics are mocked as having missed the point, and asset prices have risen far beyond fundamental justifications. New valuation metrics emerge to justify the prices; old ones are dismissed as irrelevant to the new paradigm. Phrases like “this time is different” appear with regularity.

The dot-com peak is illustrative: companies with no revenue traded at market capitalizations of billions. Analysts justified valuations by discounting speculative future cash flows or by metrics like “eyeballs” that had no established relationship to earnings. Skeptics who pointed to negative cash flows were told they didn’t understand the internet economy. The justifications weren’t always irrational given the premise — if the internet were going to generate the revenues projected, many valuations would have been reasonable. The irrationality was in the premise.

Two dynamics extend the boom phase: overtrading (speculative activity well beyond what fundamentals justify) and swindles (fraud, which typically expands during booms because rising prices make fraud easy to conceal and punishment scarce). Ponzi schemes and fraudulent financial statements become more prevalent when everyone is making money and no one is looking too closely.

Distress, Revulsion, and Contagion

The reversal is triggered by some event — a bank failure, a fraud exposure, an interest rate rise, a piece of news that causes some participants to sell. The trigger is often minor; the crisis is caused by the underlying fragility built up during the boom, not by the trigger itself. This is why “what caused the crisis?” questions focused on the trigger are usually unsatisfying — the answer is the years of excess that made the system fragile.

Distress begins when early players try to exit. Prices fall, which makes leveraged positions insolvent, forcing more selling. Collateral values fall, triggering margin calls, forcing more selling. The credit expansion reverses: lenders call loans, new lending stops, the financial system contracts. What was a virtuous cycle during the boom becomes a vicious cycle in the crash.

Revulsion is the final phase — a psychological state that mirrors euphoria in the other direction. Investors who were enthusiastic buyers become determined sellers at any price. Assets that seemed desirable are now unsellable. Banks that were lending freely refuse credit to anyone. The same assets that represented opportunity are now tainted by association with the failed boom. Revulsion overshoot downward mirrors the euphoria’s overshoot upward.

Contagion spreads the crisis beyond the original sector. A crisis in one asset class affects related asset classes as investors sell what they can to meet redemptions in what they can’t. Banks exposed to one failing sector reduce lending across all sectors. International contagion operates through trade linkages, financial exposures, and confidence effects.

The Lender of Last Resort

Kindleberger documents the historical role of the lender of last resort — typically the central bank — in halting panics. The Bagehot rule (Walter Bagehot, 1873): in a panic, lend freely at penalty rates against good collateral. Provide liquidity to solvent institutions facing a temporary liquidity crisis; don’t bail out insolvent institutions.

The rule is harder to apply than to state. In a panic, it’s often unclear which institutions are merely illiquid (sound but facing a bank run) and which are genuinely insolvent (assets worth less than liabilities regardless of the run). Liquidity support for insolvent institutions creates moral hazard — it removes the consequences of excessive risk-taking and encourages future excess. But refusing support for illiquid-but-solvent institutions can turn a panic into a depression by cutting off credit to the productive economy.

The 2008 financial crisis was a real-time test of these principles. The Federal Reserve and Treasury applied Bagehot’s principles imperfectly and inconsistently — Lehman Brothers was allowed to fail (no lender of last resort) while Bear Stearns was rescued, and AIG was rescued because its failure would have cascaded through the global financial system. The crisis was eventually contained, but the inconsistency created uncertainty about which institutions would be rescued that amplified the panic.

The Pattern Persists

The Minsky-Kindleberger framework is not a predictive tool — it can’t tell you when the current boom will end or what the trigger will be. What it provides is a diagnostic framework: the signs of increasing financial fragility (rising leverage, shift toward Ponzi finance, new valuation metrics to justify high prices, fraud proliferating) that indicate the boom is approaching its unstable phase.

The pattern persists because the underlying mechanisms persist. Credit expansion amplifies price movements in both directions. Human psychology exhibits consistent tendencies toward extrapolation, herding, and overconfidence in good times and panic in bad times. Financial innovation recurrently creates instruments that allow leverage to expand beyond what prior experience would have permitted. And regulatory frameworks always lag the newest forms of excess.

The lesson is not that booms and busts can be eliminated — they appear to be intrinsic to the structure of financial systems and human psychology. The lesson is that they are predictable in structure if not in timing, that the damage they cause depends on how fragile the underlying financial system has become, and that the policy response in the crisis phase materially affects how bad the revulsion phase gets.