A century of crashes — from roughly −89% in 1929 to roughly −25% in 2022 — repeats the same four beats: a fall that feels endless, a bottom nobody calls, a recovery that begins while the news is still terrible, and a generation of investors who take away the wrong lesson. The depths differ enormously; the shape does not.
Every crash feels unprecedented while you are inside it. The cause is new, the speed is shocking, and a chorus of serious-sounding voices explains why this time the old rules no longer apply. Every drawdown figure in this article is approximate and rounded, in local-currency price terms for the relevant index — treat them as the scale of the thing, not as precise measurements.
I have personally invested through 2018, 2020, 2022, and the wobbles since. For everything before that, I rely on the historical record — and I think studying it is the single highest-return activity available to a long-term investor, because crashes are where most permanent damage to portfolios is actually done. Not by the market, but by our reactions to it.
This article walks through the major crashes of the past hundred years at approximate, rounded levels — the exact figures vary by index and source, so treat every number here as illustrative rather than precise — and then draws out ten lessons for European UCITS investors as of 2026.
Want the whole path in one place? — The Masterclass takes you from zero to a running, tax-aware portfolio — 55 text lessons, one weekend to read.
See the full curriculum — €147, 30-day guarantee →A century of crashes at a glance
All drawdowns below are approximate, in local-currency price terms for the relevant US or global index, and recovery times vary depending on whether you count dividends and inflation. The shape matters more than the decimals.
| Crash | Approximate depth | Speed | Defining feature |
|---|---|---|---|
| 1929 Great Crash | ~ -89% over several years | Slow grind, 1929–1932 | The outlier: depression, deflation, leverage |
| 1973–74 oil shock | ~ -48% | About two years | Inflation made real recovery take far longer |
| 1987 Black Monday | ~ -20% in one day | Hours | Pure market mechanics; economy barely noticed |
| 2000 dot-com bust | ~ -49% S&P 500, tech far worse | Roughly 2.5 years down | Valuation bubble deflating in slow motion |
| 2008 financial crisis | ~ -57% | About 17 months | Banking system itself was the epicentre |
| 2020 COVID crash | ~ -34% | Weeks down, months back | Fastest fall and fastest recovery on record |
| 2022 rate shock | ~ -25% stocks, bonds down too | A full year of grind | Stocks and bonds fell together — diversification’s bad year |
1929 to 1987: the classics
The 1929 crash is the one that haunts the imagination, and rightly — a fall of roughly 89% from peak to the 1932 trough, unwinding over years rather than weeks, amplified by margin debt, bank failures, and policy mistakes. Nominal recovery took until the 1950s, though with dividends reinvested and deflation accounted for, the real experience was meaningfully less apocalyptic than the headline number suggests. It remains the outlier: no crash since has come close, partly because deposit insurance, central bank lender-of-last-resort practice, and automatic fiscal stabilisers were invented largely in its aftermath.
1973–74 brought a roughly 48% fall as the oil embargo met an inflationary spiral. The lesson it left was subtle: prices recovered in nominal terms within a few years, but high inflation meant the real recovery took much of a decade. And 1987’s Black Monday — around 20% gone in a single day — proved a crash needs no economic cause at all; portfolio-insurance selling fed on itself, the economy shrugged, and the market regained the ground within about two years.
2000 and 2008: two halvings in one decade
The dot-com bust cut the S&P 500 roughly in half over two and a half grinding years, and technology-heavy indexes fell far more — the NASDAQ lost around three-quarters of its value and took about fifteen years to see its old nominal peak again. The 2008 financial crisis then took broad markets down roughly 57% in about seventeen months, this time with the banking system itself at the centre of the fire. An investor who retired into 2000 endured what is often called a lost decade for US equities.
Yet the same decade carries a quieter message: someone who kept buying monthly through both halvings, mechanically and without heroics, accumulated shares at deeply depressed prices and was rewarded handsomely in the 2010s. The decade was only “lost” for money invested once, at the top, and never added to.
2020 and 2022: speed, and the bond surprise
COVID compressed a full bear market into about five weeks — roughly a third of global market value gone — and then, propelled by unprecedented fiscal and monetary support, recovered within months. Anyone who sold in March 2020 and waited for the economy to look safe again missed the rebound almost entirely.
2022 taught the opposite tempo and a nastier lesson: as central banks raised rates to fight inflation, equities fell around a quarter while bonds — the traditional shock absorber — suffered one of their worst years in modern history. Stocks and bonds down together broke the assumption behind every 60/40 portfolio. I unpack what that means for portfolio construction in my 2022 bond meltdown lessons.
The ten lessons
- Every crash so far has been recovered. Broad, diversified indexes have eventually surpassed every previous peak, because aggregate corporate profits restart and grow. This is a statement about diversified markets, not individual stocks or single countries — plenty of companies never came back.
- Recovery times vary enormously — hence the five-year rule. 2020 took months; 1929 and 2000 took a decade or more in some terms. That spread is exactly why money needed within about five years does not belong in equities. The market owes you nothing on your schedule.
- The cause is always obvious afterwards and unforecastable before. Each crash produced confident retrospective explanations and essentially no successful advance predictions. The professional forecasting record, tracked year after year in the SPIVA scorecards on active management, should end any fantasy of dodging crashes on cue.
- Buying through crashes lowered average cost. For anyone mid-accumulation, every crash in the table was, in hindsight, a discount window. A standing monthly plan — see my dollar-cost averaging guide — buys more units precisely when prices are lowest, without requiring courage in the moment.
- Panic selling converted temporary falls into permanent losses. The market recovered every time; the investors who sold near bottoms did not. This is the mechanism by which crashes destroy wealth, and I’ve written a full piece on why investors panic sell and how to defuse it.
- Global diversification beat single-market bets. The cautionary tale is Japan: its market peaked in 1989 and took roughly three and a half decades to see that nominal level again (approximate, and dividends soften it). A Japanese investor holding only Japan lived a very different century than one holding the world. No single country — including today’s favourite — is exempt from that risk.
- Cash buffers turn crashes from threat into opportunity. Forced sellers set the bottom prices; buyers with reserves collect them. An adequate emergency fund, parked in deposits or money market ETFs, is what keeps you out of the first group and eligible for the second.
- Headlines reach peak certainty at bottoms. “Stocks may have further to fall,” “the end of equities,” “this time is different” — the most confident pessimism in every episode clustered near the lows. Media certainty is a contrarian indicator, not an action signal.
- Leverage is how crashes actually kill. Margin calls in 1929, over-levered banks in 2008, forced liquidations in every episode: the unleveraged investor can wait out any drawdown, while the leveraged one can be removed from the game at the bottom, permanently. Time heals unleveraged portfolios; it forecloses leveraged ones.
- A written rule beats a strong stomach. Nobody knows their true risk tolerance until they’ve watched six figures evaporate in a fortnight. Decide your crash behaviour now, on paper, while calm — my bear market checklist is exactly that pre-commitment device.
What I actually changed after living through three of these
Reading history is one thing; feeling your own portfolio drop is another. Three concrete changes came out of my own crash experience. First, I stopped holding any money in equities that I could plausibly need within five years — lesson two, learned viscerally in 2020 when I was briefly grateful not to need cash. Second, I automated everything: contributions leave my account the day after salary arrives, so no crash headline ever gets a vote. Third, I wrote my rules down, including the exact sentence I read to myself when markets fall 20%: “You have seen this before, and you know how it ends if you sell.”
What I deliberately did not change is the portfolio itself. A globally diversified UCITS equity core plus a cash-and-short-bond buffer survived every scenario in the table above — not painlessly, but survivably. As of 2026, that remains the whole strategy. The next crash will come, its cause will surprise everyone, its bottom will be called by no one, and the same ten lessons will apply. Verify current figures before acting on anything here; the history is approximate, but the behaviour it teaches is not.
Frequently asked questions
Q: How often do stock market crashes actually happen?
A: Falls of about 10% occur every year or two on average; bear markets of 20%+ arrive roughly every five to seven years, though the spacing is irregular. Declines of 50%+ have happened a handful of times in a century. Treat crashes as a recurring feature of equity investing, not an anomaly.
Q: Has the market really recovered from every crash?
A: Broad diversified indexes have eventually exceeded every prior peak, including 1929 — though recovery took from months (2020) to well over a decade (1929, 2000 in real terms). Individual stocks and single-country markets, like Japan after 1989, offer no such assurance. Diversification is what makes the pattern reliable.
Q: Should I wait for a crash before I start investing?
A: History argues no. Markets spend most of their time rising, and waiting for a crash has often meant buying at higher prices than when the wait began — if the crash even undercuts today’s level. A regular monthly plan sidesteps the timing question entirely and buys any future crash automatically.
Q: Was 1929 as bad for investors as the -89% figure suggests?
A: The roughly -89% figure is nominal, dividend-free, and peak-to-trough over about three years. With dividends reinvested and severe deflation raising the value of each dollar, diversified investors recovered in real terms considerably faster than the headline implies — though still painfully slowly. It remains the worst case on record.
Q: What is the single best crash preparation for a European ETF investor?
A: A written plan made in calm conditions: a globally diversified UCITS portfolio sized so you never need to sell within five years, an emergency fund covering several months of expenses, automated monthly contributions, and a pre-committed checklist for red days. Preparation beats prediction — no one reliably forecasts crashes.
For the other half of downside protection, see broker safety and investor protection.
European UCITS/ETF investor with 11+ years building a global dividend and index portfolio, and founder of Financial Expert Class. I write every guide here from first-hand experience investing across EU brokers and tax regimes — not theory. More about me →