Four Crashes. Recovery Times of Two Years, Fifteen Years, and Five Months. No Pattern You Could Have Traded.

10 Aug 2026 · MARKET HISTORY

Everything in this piece is general market history — publicly documented, widely reported facts about past crashes — not a claim tested against our own data the way our backtests are. We're saying that plainly because this site's whole approach is to separate "we verified this against our own archive" from "this is what's publicly known," and history doesn't get to skip that distinction just because it's a good story.

With that said: four crashes, four different shapes, and the same lesson showing up in all of them from a different angle.

Black Monday, October 19, 1987

The Dow Jones Industrial Average fell 22.6% in a single session — still the largest one-day percentage decline in its history, before or since. There was no single dominant news catalyst identified afterward that fully explained a move that size; portfolio-insurance trading strategies amplifying the decline through automated selling are widely cited as a major contributor. The market recovered its Black Monday losses within about two years.

The lesson isn't "it always comes back fast" — it's the opposite. Nobody watching that day, in real time, could have told you it would take roughly two years rather than five or ten. The recovery timeline is only obvious in hindsight; in the moment, a 22.6% single-day drop looked exactly as terrifying as a genuine multi-year bear market would have looked on day one.

The dot-com bust, 2000–2002

The Nasdaq Composite fell roughly 78% from its March 2000 peak to its October 2002 trough — a decline that erased essentially all the gains of the late-1990s internet boom. It took the Nasdaq until 2015 to reclaim its 2000 high in nominal terms, about fifteen years.

Companies with no revenue and no path to profitability had traded at valuations implying decades of dominance that never arrived. The lesson repeated in nearly every retrospective is the same one: a rising price is not evidence a business model works, and the market can sustain a wrong valuation for far longer than seems possible, right up until it can't.

The 2008 financial crisis

The S&P 500 fell roughly 57% peak-to-trough between October 2007 and March 2009 — the worst bear market since the Great Depression. Unlike 1987 or 2000, this one had a clearly identifiable structural cause: a housing and mortgage-lending bubble that, once it burst, took the banking system down with it, because it wasn't just stock prices that were wrong, it was the credit underneath them.

This is the crash most often cited for a specific, narrower lesson: leverage turns a bad outcome into a catastrophic one. Firms and households that had borrowed heavily against inflated asset values couldn't absorb the decline; ones with less leverage generally could.

The COVID crash, February–March 2020

The S&P 500 fell about 34% in roughly a month — one of the fastest bear markets on record by a wide margin, driven by a real-world shock (a global pandemic and the lockdowns that followed) rather than by anything wrong with the market's own internal plumbing. It was also one of the fastest recoveries: the index reclaimed its prior high by August 2020, in months rather than years.

This is the one that breaks the naive lesson "crashes take years to recover from." Sometimes they don't — and the difference wasn't something you could have priced in real time on day one of the decline. Nobody watching in March 2020 could reliably distinguish "this recovers in five months" from "this is 2008 again."

The actual lesson, stated once

Four crashes, four different causes, four different recovery timelines — two years, fifteen years, years, months. If there were a reliable way to know in real time which kind you were in, that would be worth more than almost anything else in investing. Nobody has reliably had that in any of these four cases. What every one of them rewarded instead was the boring stuff: not being leveraged past what you could survive, having a plan for how much you were willing to lose before you needed to act on it, and not needing to guess correctly about the recovery shape to avoid being wiped out waiting for it.

Scores in this piece come from the ZonedInOut model. How the score works →

Frequently asked

How long did it take the stock market to recover from Black Monday in 1987?
The Dow Jones Industrial Average fell 22.6% on October 19, 1987 — still the largest single-day percentage decline in its history. The market recovered those losses within approximately two years. At the time, there was no way to know in advance that the recovery would take roughly two years rather than much longer; the recovery timeline is only clear in hindsight.
How long did the dot-com crash take to recover?
The Nasdaq Composite fell approximately 78% from its March 2000 peak to its October 2002 trough. It did not reclaim its March 2000 high in nominal terms until 2015 — roughly fifteen years later, making it one of the slowest major recoveries in modern market history.
Was the 2020 COVID crash the fastest stock market recovery in history?
The S&P 500 fell about 34% between February and March 2020 in roughly a month, one of the fastest bear markets on record, and then reclaimed its prior all-time high by August 2020 — a recovery measured in months rather than years, in sharp contrast to the multi-year recoveries after 1987, the dot-com bust, and 2008. It illustrates that crash recovery timelines vary enormously and are not reliably predictable in real time from the shape of the decline itself.
What do past stock market crashes have in common?
Beyond the shared experience of a sharp decline, the major modern crashes (1987, the dot-com bust, 2008, and the 2020 COVID crash) had different causes and dramatically different recovery timelines, ranging from months to roughly fifteen years. No single "crashes always recover within X" rule fits all of them. The pattern that does hold across all of them is that leverage and lack of a predefined risk plan turned an otherwise survivable decline into a much worse outcome for those affected.

Get the Sunday Brief

One email a week, before the open: what the board did, which calls worked, which didn't, and what changed. Free. No spam, unsubscribe anytime.