Why is financial history the best teacher?

The short answer

The discipline of financial history exists because financial systems exhibit recurring structural patterns despite changing technologies, regulations and instruments. From tulip mania in 1637 to the South Sea Bubble of 1720, from the Great Depression to the Global Financial Crisis, the build-ups share recognizable features: extended credit growth, narrative shifts that legitimize new asset valuations, complacency about tail risks, and reversal triggered by marginal events that would be ignored in calmer times.

The intuition behind taking history seriously is that human institutional behavior — including financial behavior — exhibits more pattern than novelty. Charles Kindleberger’s “Manias, Panics, and Crashes” (1978) and J.K. Galbraith’s “A Short History of Financial Euphoria” (1990) catalogued these patterns before Reinhart and Rogoff’s quantitative work formalized them.

The complication is that history teaches without prescribing timing. Knowing that bubbles pop does not tell you when. Knowing that “this time is different” claims usually fail does not tell you which contemporary claim of difference is the next failure.

→ New to crisis frameworks? Macro-financial regimes pillar

What the data shows

The empirical record of financial history has been compiled with increasing rigour over the past two decades.

The figures (Reinhart-Rogoff, Schularick-Taylor, BIS, 1800-2020):

• Reinhart and Rogoff’s database covers 70+ countries and roughly 800 years of debt crises, providing the longest quantitative record
• Schularick and Taylor (2012) document 14 advanced economies over 140 years and show that credit-driven booms predict subsequent recessions with statistical significance
• Banking crises in advanced economies have averaged roughly one per decade since 1800, but the gap between crises (1945-2007) was unusually long by historical standards
• The fall in real GDP during financial crises has averaged 9% peak-to-trough across the long sample, with the recovery to pre-crisis levels typically taking 5-7 years

The comforting finding for advanced economies is that the post-WW2 era saw substantially fewer banking crises than earlier periods, with regulation and deposit insurance plausibly contributing. The post-2008 reversion to higher crisis frequency raises the question of whether the post-WW2 period was the anomaly or the norm.

The exception worth flagging: emerging market crises have been more frequent throughout the post-WW2 period, with sudden-stop episodes (capital flight) playing a recurring role. The lesson that “advanced economies have de-risked” does not extend to all jurisdictions equally.

→ Dataset: US federal debt to GDP dataset

Why it happens — the macro mechanism

Three structural patterns make financial history a defensible guide to current dynamics.

Channel 1 — Credit cycle dynamics. Schularick and Taylor’s central finding is that excessive credit growth — credit-to-GDP rising materially above its trend — is the most consistent leading indicator of subsequent banking crisis across the long historical sample. The mechanism combines lender risk-taking that expands during good times, borrower behaviour that capitalises rising collateral values, and regulatory responses that typically lag the build-up.

Channel 2 — Pattern recurrence under structural change. This is the angle most overlooked. Each crisis involves a different specific instrument — railway shares in 1873, public utilities in 1929, dot-com equities in 2000, US subprime mortgages in 2007 — but the structural mechanism is similar: a new asset class with apparently different fundamentals attracts capital, creates positive feedback through credit, and reverses when a marginal event challenges the narrative. The instruments change; the architecture does not.

Channel 3 — Institutional learning lags. Regulators, banks and investors learn from the most recent crisis but tend to over-fit to its specific mechanism. Post-Great-Depression regulation was designed for bank-run dynamics; the 2008 crisis ran through capital-market intermediaries that the framework did not cover. Post-2008 regulation has tightened bank capital but pushed activity into non-bank channels (the FSB’s NBFI categorization captures this growth). The next crisis is unlikely to look exactly like 2008 because regulation has adapted to that exact form.

Synthesis by regime: in pre-WW2 history, frequent banking crises with limited central-bank backstops produced sharp but often brief recoveries; in the post-WW2 to 2007 era, deposit insurance and discretionary monetary policy reduced crisis frequency at the apparent cost of increased moral hazard accumulation; in the post-2008 era, tightened bank regulation has shifted risk to non-bank channels and central bank intervention has become a routine tool rather than an emergency lever. Three regimes, three institutional architectures of risk.

History does not repeat in form but rarely fails to repeat in mechanism — the new bubble of any era looks different on the surface and identical underneath, with credit growth as the most reliable shared feature.

→ Framework: Economic cycle phases pillar

What it means for different economic actors

Long-term investors can use the historical record as a calibration check. Periods of unusually high credit growth, unusually compressed risk premia and unusually narrow market leadership recur and tend to be followed by disappointing returns over multi-year horizons. None of these signals provides timing precision, but they provide perspective.

Risk managers face the practical implication that stress tests calibrated to recent history under-represent tails that the long sample reveals. The 2008 crisis exceeded most risk-management models’ “worst plausible scenario” because those models were calibrated to post-1980 data.

Citizens and policymakers can take comfort from the post-WW2 reduction in advanced-economy crisis frequency while taking warning from the post-2008 reversal. The institutional architecture matters; deregulating it has historically been followed by crisis frequency rising.

A common error is to treat history as either a reliable predictor (the “next crisis will look exactly like X”) or as irrelevant (the “structures have changed so previous patterns no longer apply”). The empirical record supports a position between: pattern repetition with surface variation, requiring careful translation between historical cases and current dynamics.

Practical observation

Go deeper

Related questions

• Why do bubbles feel rational at the time?
• What will the next financial crisis likely look like?
• How do narratives shape economic reality?
• How did the 2023 regional bank crisis differ from 2008?

Frequently asked questions

Hasn’t modern technology made historical analogies obsolete?

Specific instruments and transmission speeds have changed dramatically. The mechanisms — credit expansion against rising collateral, narrative coordination, regulatory lag, marginal-event triggers — have shown remarkable persistence. The 2008 subprime crisis used securitization technology that did not exist in 1929, but the underlying pattern of credit-fuelled asset boom followed by reversal would have been recognizable to 1873 observers. Technology accelerates and transforms; it does not obviously dissolve historical patterns.

Does the post-WW2 reduction in advanced-economy crises mean the regulators have “won”?

Partially, with caveats. The Great Depression-era reforms (deposit insurance, banking supervision, separation of investment and commercial banking) plausibly contributed to the long quiet from 1945 to 2007. But several elements unwound starting in the 1980s (Glass-Steagall repeal, expanded use of derivatives, growth of shadow banking), and the 2008 crisis was preceded by precisely the kind of regulatory laxity that pre-WW2 reformers had fought to prevent. The post-WW2 lull was real; whether it was a permanent regime shift or a temporary institutional achievement remains contested.

What is the most reliable lesson from financial history?

The most defensible single lesson is that excessive credit growth — credit-to-GDP rising materially above its trend for a sustained period — has been the most consistent precursor to subsequent banking crisis across the long historical sample. Schularick and Taylor’s quantitative work confirmed what Kindleberger had argued narratively. This does not provide timing precision, but it does provide a reliable signal for elevated probability over multi-year horizons.

Last updated — 30 July 2026