The Scariest Stock Market Crashes of All Time – and What We Can Learn From Them

Just in time for Halloween, things are getting scary today: we're looking at the worst stock market crashes in history. An index that loses 89%. A single day with −22.6%. An entire country whose stock market was under water for 34 years. And a DAX company whose billions probably never existed. But don't worry: in the end, it's mainly about what we can learn from these crashes – and there are some pretty good studies on that.

One note up front: some of the best-known crash stories aren't actually true. We'll look at that too.

As always, the most important thing first: this is not investment advice and not a buy recommendation, just my personal opinion. Past performance is no guarantee of future results, and stocks can stay in the red for years. All index levels in this article are closing prices, nominal and not adjusted for inflation. Careful when comparing: the DAX is a performance index (dividends are notionally reinvested), while the Dow Jones, S&P 500, Nasdaq and Nikkei are pure price indices. Links marked with * are partner links: if you use them, I get a small commission, at no extra cost to you.

Prologue: the crash myths everyone knows

Tulip mania 1637 – the bubble that was mostly a story

Tulip mania is considered the first speculative bubble in history. In the Netherlands, prices for rare, patterned tulip bulbs rose sharply from around November 1636, and some of the trading took place via futures contracts in taverns. In early February 1637, prices collapsed. The economist Peter Garber recalculated this for a common variety: on 5 February 1637, at the peak, the Switser bulb cost 0.17 guilders per aas (one twentieth of a gram); a few days later it was only 0.11 guilders – around −35% (Journal of Economic Perspectives, 1990).

According to the historian Anne Goldgar, the most expensive bulbs cost around 5,000 guilders – the price of a well-appointed house. And now comes the myth: mass bankruptcies, ruined families, people drowning themselves in the canals. According to modern research, none of that is documented. Goldgar went through the archives in Haarlem, Amsterdam, Alkmaar and Enkhuizen (The Conversation, 2018):

“I found not a single bankrupt in these years who could be identified. […] No one drowned themselves in canals.”

Anne Goldgar, historian

Only 37 people can be shown to have spent more than 300 guilders on bulbs, and the Dutch economy was left unaffected. Much of what the writer Charles Mackay wrote about tulip mania in 1841 goes back to satirical songs and pamphlets from 1637. The famous story of the sailor who mistook a valuable bulb for an onion and ate it isn't documented in any records either.

The lesson: the biggest bubble here was the story itself. If you believe crash stories without checking the source, you fall for a myth that's almost 390 years old.

The South Sea Bubble 1720 – and Isaac Newton's costly mistake

The South Sea Bubble in England was much more real. In 1720, the South Sea Company took over a large part of the British national debt in exchange for its own shares. According to Garber, the share rose from about £120 in January 1720 to around £950 on 1 July – and then crashed from about £775 on 31 August to about £290 on 1 October. In just one month, around £103 million of market value evaporated, more than twice the national debt the company had taken on.

One of the losers: Isaac Newton. The mathematician Andrew Odlyzko reconstructed his finances (Notes and Records of the Royal Society, 2019). Newton first sold at a profit, but then got back in “almost at the very top” and ended up with essentially his entire fortune in South Sea securities. Out of a net worth of somewhat over £30,000, he lost around £10,000 – about a third. He wasn't ruined, but had he simply kept his early profits, he would have come out of it worth over £50,000.

You may know the famous quote in which Newton supposedly said he could calculate the motions of the heavenly bodies, but not the madness of people. This quote is only attributed to him. The oldest trace is a second-hand anecdote written down decades later – and it doesn't mention heavenly bodies at all. What is documented is what Newton himself wrote in a letter in 1724:

“I lost very much by the South Sea company which makes my pockets empty, & my mind averse from dealing in these matters.”

Isaac Newton, letter of 14 September 1724

The lesson: even one of the greatest geniuses in history sold at the right time – and then, out of fear of missing out, bought back in at the top. Intelligence doesn't protect you; a plan that you stick to does.

1929: the crash that became the Great Depression

In the 1920s, Wall Street only knew one direction. The Dow Jones rose six-fold from 63 points in August 1921 to 381 points in September 1929. According to Federal Reserve History, many buyers put down only a fraction of the price themselves, “typically 10 percent”, and borrowed the rest.

On 3 September 1929, the Dow closed at 381.17 points. Then came October: on Black Monday, 28 October 1929, the index fell almost 13%, and on Black Tuesday, 29 October, almost another 12%. By mid-November it had lost almost half its value. And it kept going down – until 8 July 1932. On that day the Dow closed at 41.22 points, its lowest level of the 20th century and 89% below its peak. $10,000 had turned into $1,100.

The Dow only regained its old high on 23 November 1954 – after just over 25 years. You hear that number a lot, but it's only half the truth: the Dow is a price index without dividends, and prices fell sharply during the Great Depression. According to an analysis by Mark Hulbert (New York Times, 2009), a broadly diversified investor who bought at the 1929 peak was, with reinvested dividends and taking deflation into account, back to break-even by late 1936 – just over 7 years later. In purely nominal terms, it still took 25 years.

The myth: the images of stockbrokers jumping out of windows one after another are among the best known in economic history. But there was no such wave. The economist John Kenneth Galbraith writes in The Great Crash 1929 that the suicide wave is a legend. According to figures from New York's chief medical examiner at the time, Manhattan saw 44 suicides in the four weeks after the crash – compared with 53 in the same period a year earlier. There were tragic individual cases, but no wave.

The lesson: leverage turned the crash into a disaster. If only 10% of the money is your own, you have to sell when prices fall – no matter at what price.

Black Monday 1987: −22.6% in a single day

On 19 October 1987, the Dow Jones lost 508 points, or 22.6% – according to Federal Reserve History, the largest one-day percentage drop in stock market history. On the New York Stock Exchange, 604 million shares changed hands that day, more than three times the average daily volume at the time. Markets fell worldwide, most severely in New Zealand at −60%.

One contributing factor was a then-new hedging strategy called portfolio insurance: computer programs automatically sold futures when prices fell, which pushed prices down further and triggered new sales. According to the US government's investigation report (the Brady Report, 1988), portfolio insurers sold futures worth around $4 billion on 19 October. But they weren't solely to blame.

And now the really interesting part: anyone who sold that day locked in their loss. Anyone who held on was soon back on top. In the following two trading sessions alone, the Dow regained 57% of the day's loss. From the peak on 25 August 1987 (2,722.42 points) to the new record on 24 August 1989, only around 2 years passed.

By the way: the DAX didn't even exist on Black Monday – it was only calculated from July 1988. Its largest real one-day loss came on 16 October 1989 at −12.81% (Deutsche Börse). October again.

The lesson: a single day of −22.6% is terrible. But those who lost most were the ones who sold. And automatic “protection” can even speed up a crash.

Japan: 34 years under water

Now it gets really scary. In the late 1980s, Japan experienced an enormous stock and real estate bubble. Economists later estimated that the grounds of the Imperial Palace in Tokyo were worth more than all of California. On 29 December 1989, the Nikkei 225 closed at 38,915.87 points.

Then the bubble burst. By August 1990 the index had roughly halved – and that was only the beginning. Its lowest close came on 10 March 2009 at 7,054.98 points: −81.9%, almost two decades after the peak (Nikkei Inc.). A new record high only came on 22 February 2024 at 39,098.68 points (Bloomberg). That's just over 34 years. Anyone who bought at the peak at 30 was 64 by the time it recovered.

To be fair: with dividends it went faster. According to Dimensional, a broad Japanese stock index including dividends (MSCI Japan, in US dollars) was back above its 1989 high from May 2017. But that's still more than 27 years.

The lesson: an entire country can be under water for decades – even one that was considered unbeatable at the time. More on that in the lessons.

The dotcom bubble and the Neuer Markt: Germany's trauma

In the late 1990s, the internet was the big story of the future. And it was right. But the prices weren't: according to one fund manager, many growth companies were trading at 200 to 300 times the next year's expected earnings.

The Nasdaq Composite reached 5,048.62 points on 10 March 2000. By 9 October 2002 it had fallen to 1,114.11 points – −77.9%. A new high only came on 23 April 2015, after just over 15 years. The DAX fell from 8,064.97 points on 7 March 2000 to 2,202.96 points on 12 March 2003 – −72.7%, almost three quarters (Deutsche Börse). It regained its old high in 2007, after just over 7 years – and that as a performance index with dividends.

But Germany was hit hardest on the Neuer Markt. The stock market segment for young tech companies launched in March 1997. According to a Bundesbank discussion paper, the NEMAX All Share rose 1,682% by 10 March 2000, and 229 companies were worth €243 billion combined. In total, 326 companies went public there. Then came the crash: the NEMAX 50 fell from around 9,700 points to a low of around 300 points in October 2002 – around −96%. On 5 June 2003, the Neuer Markt was closed.

And it wasn't just a bubble, but partly fraud as well. At the telematics provider Comroad, a special audit found that only 1.4% of the €93.6 million in revenue reported for 2001 was real. More than €90 million ran through a supposed customer in Hong Kong for whose existence the auditors found no evidence at all (heise online, 2002). Founder Bodo Schnabel was sentenced to seven years in prison in November 2002. At EM.TV, too, the figures had to be corrected in 2000, and the share fell to around 5% of its high.

The lesson: a right idea doesn't protect you from a wrong price. Hype without profits and a concentration in a single sector can wipe out almost everything. Many Germans have been afraid of stocks since the Neuer Markt – even though the crash was much milder for those who were broadly diversified.

The right account: Smartbroker+ with a €20 McDonald's share

If you take one thing away from these crashes – diversify broadly and stay in for the long term – you need a low-cost account for it. I use Smartbroker+, among others, and currently have my own partner promotion there: if you open a free account as a new customer through my link*, you get a McDonald's share worth €20 (WKN 856958) for free.

According to the promotion page (as of 26 October 2026), Smartbroker+ lets you trade stocks, ETFs and funds from €0 (from €500 order volume via gettex, plus usual market spreads, inducements and product costs), offers more than 4,600 savings plans with €0 order fees, and order processing runs through German institutions. How the promotion works and what to watch out for so the share actually reaches you is explained in my article on the Smartbroker+ promotion. You'll find the full terms of participation on the promotion page – best read them right before you sign up.

The financial crisis of 2008: when the banking system stood still

The financial crisis began in the US housing market and became the most severe crisis since the Great Depression. According to Federal Reserve History, house prices fell by around 30%, and US unemployment rose to 10% by October 2009.

The scariest day was 15 September 2008: after 158 years, the investment bank Lehman Brothers filed for bankruptcy – with total assets of $639 billion, the largest bankruptcy in US history. On the same day, Merrill Lynch was sold to Bank of America, the insurer AIG asked the central bank for help, and the Dow Jones fell 504 points (−4.4%).

The S&P 500 fell 56.8% from its peak on 9 October 2007 (1,565.15 points) to 9 March 2009 (676.53 points). Its new high came on 28 March 2013, after around 5.5 years. The DAX lost 54.8% from 16 July 2007 to 6 March 2009 and needed just under 6 years to come back. And the MSCI World? In euros with reinvested dividends (net return), it fell 53.4% and was back at a record on 28 February 2013 – after just over 5.5 years (my own analysis of MSCI data).

The lesson: leverage and opacity in the banking system can bring the world to a halt within days. And: anyone who kept saving in March 2009 bought at the cheapest prices of an entire generation.

Wirecard 2020: the DAX company whose billions were missing

For many German investors, this crash is the most personal one. Wirecard was a payment service provider and, from September 2018, even a member of the DAX – replacing Commerzbank, a founding member of the index. Critical reports, especially from the Financial Times, had been around for a long time. But instead of taking a closer look at Wirecard, the financial regulator BaFin banned new short sales of the stock on 18 February 2019 (BaFin) – the first time it used this EU rule to do so.

On 18 June 2020 came the shock: the auditor EY refused to sign off on the 2019 accounts because €1.9 billion in trust accounts could not be verified – around a quarter of total assets. The share lost almost 62% that day. On 22 June, Wirecard itself admitted that there was a prevailing likelihood that the €1.9 billion did not exist. On 25 June 2020, the company filed for insolvency – the first DAX company ever to do so.

The share fell from over €100 on 17 June to around €1 just over a week later – around −99%. In 2018 it had still been just under €200. For shareholders, it was practically a total loss.

And where do things stand today? Former CEO Markus Braun has been in pre-trial detention since July 2020. The criminal trial at the Munich I Regional Court has been running since December 2022; there is no verdict yet (as of the end of October 2026), and closing arguments are expected in November 2026 (LTO). Braun denies the allegations, and he is presumed innocent. Former board member Jan Marsalek has been on the run since June 2020 and is wanted internationally. The German Bundestag examined the case in its own committee of inquiry (Bundestag).

The lesson: neither the DAX, nor the regulator, nor a big auditor reliably protects you from a total loss on a single stock. In a broadly diversified index with a great many companies, a single collapse like this barely matters.

Covid 2020: the fastest crash in history

In February 2020, stock markets were at record highs. Then the coronavirus spread around the world. The S&P 500 fell 33.9% from 19 February (3,386.15 points) to 23 March 2020 (2,237.40 points) – in just 33 days. According to AP, it was the quickest bear market on record. The DAX lost as much as 38.8% from 19 February to 18 March, in less than four weeks.

And then? The recovery was just as fast. The S&P 500 was back at a record as early as 18 August 2020 (AP via PBS), just under 6 months after the peak. The DAX followed on 28 December 2020, and the MSCI World (net return in euros) on 7 January 2021. Particularly crazy: according to J.P. Morgan, the second-worst trading day of 2020, 12 March, was immediately followed by the second-best.

The lesson: anyone who panic-sold in March 2020 missed the fastest recovery in history.

All the crashes at a glance

Crash	Index	Peak	Low	Loss	Time to new high
1929	Dow Jones	381.17 (3 Sep 1929)	41.22 (8 Jul 1932)	−89%	25+ yrs (23 Nov 1954)
Oil crisis 1973/74	S&P 500	120.24 (11 Jan 1973)	62.28 (3 Oct 1974)	−48%	7.5 yrs (17 Jul 1980)
Black Monday	Dow Jones	2,722.42 (25 Aug 1987)	1,738.74 (19 Oct 1987)	−36%	2 yrs (24 Aug 1989)
Japan	Nikkei 225	38,915.87 (29 Dec 1989)	7,054.98 (10 Mar 2009)	−82%	34+ yrs (22 Feb 2024)
Dotcom	Nasdaq Comp.	5,048.62 (10 Mar 2000)	1,114.11 (9 Oct 2002)	−78%	15+ yrs (23 Apr 2015)
Dotcom	DAX	8,064.97 (7 Mar 2000)	2,202.96 (12 Mar 2003)	−73%	7+ yrs (2007)
Neuer Markt	NEMAX 50	around 9,700 (10 Mar 2000)	around 300 (Oct 2002)	around −96%	never (closed 2003)
Financial crisis	S&P 500	1,565.15 (9 Oct 2007)	676.53 (9 Mar 2009)	−57%	5.5 yrs (28 Mar 2013)
Financial crisis	DAX	8,105.69 (16 Jul 2007)	3,666.41 (6 Mar 2009)	−55%	almost 6 yrs (May 2013)
Financial crisis	MSCI World (net, EUR)	15 Jun 2007	9 Mar 2009	−53%	5.7 yrs (28 Feb 2013)
Wirecard	share	over €100 (17 Jun 2020)	around €1 (26 Jun 2020)	around −99%	never (insolvency)
Covid	S&P 500	3,386.15 (19 Feb 2020)	2,237.40 (23 Mar 2020)	−34%	6 months (18 Aug 2020)
Covid	DAX	13,789.00 (19 Feb 2020)	8,441.71 (18 Mar 2020)	−39%	10 months (28 Dec 2020)
Covid	MSCI World (net, EUR)	19 Feb 2020	23 Mar 2020	−34%	10.6 months (7 Jan 2021)

Closing prices, nominal, not adjusted for inflation; “time to new high” measured from the old peak. DAX and MSCI World Net include dividends, the other indices don't. Sources: Federal Reserve History, Deutsche Börse, Nikkei Inc., AP, Bloomberg, MSCI index data (my own analysis), partly recalculated from Yahoo Finance price data.

What stands out in the table: the longest recoveries happened in individual countries or sectors (Japan, Nasdaq) and in single stocks that never came back at all. The broadly diversified MSCI World recovered comparatively quickly in the more recent crashes. To be completely honest, though: after the dotcom crash, a euro investor in the MSCI World stayed below the autumn 2000 peak for more than 13 years even with dividends, because a weak dollar came on top of the stock market crash.

The 5 lessons from stock market crashes – backed by studies

1. Crashes are normal – and they pass

According to Hartford Funds (data: Ned Davis Research), the S&P 500 has had 27 bear markets since 1928, meaning declines of at least 20%. On average the index fell 35%, and after an average of 289 days – less than ten months – the bear market was over (Hartford Funds). An average bull market, by contrast, delivered +112% and lasted 2.7 years. If you invest for 50 years, you should expect around 14 bear markets.

The longer you stay in, the narrower the range becomes: according to the J.P. Morgan Guide to Retirement 2026, single years in the S&P 500 since 1950 ranged from −37% to +52%. Over 20 years, however, there was not a single period with a loss – the worst still returned +6% a year. That applies to the US market in the past and is not a law of nature: the MSCI World in euros (price index), for example, was clearly in the red after ten years from the end of 1999 to the end of 2009.

2. Stay invested – the best days come in the middle of the crash

J.P. Morgan calculated what happened to $10,000 in the S&P 500 (with dividends) from 2006 to 2025:

$10,000 from 2006 to 2025	Final value	Return per year
always invested	$80,619	11.0%
missed the 10 best days	$35,866	6.6%
missed the 20 best days	$21,177	3.8%
missed the 30 best days	$13,826	1.6%

Source: J.P. Morgan Asset Management, Guide to Retirement 2026, S&P 500 Total Return, before costs.

Missing just 10 out of around 5,000 trading days left you with less than half at the end. And the nasty part: 6 of the 10 best days came within two weeks of the 10 worst. According to Hartford Funds, 42% of the strongest days of the past 20 years fell right in the middle of a bear market. To be honest, such calculations are hypothetical – anyone who misses the best days often misses bad ones too. The real message is a different one: getting out and back in at the right time practically never works, because good and bad days are so close together.

Morningstar shows that this isn't just theory. In its study Mind the Gap 2026, the funds themselves returned 9.9% a year on average over the ten years to the end of 2025 – but the investors in those funds only got 8.7%. The gap of 1.2 percentage points a year comes from the timing of purchases and sales. The more volatile the fund, the bigger the gap. Too much back and forth empties your pockets.

3. Diversify broadly – across countries and companies

Japan is the best example of why you shouldn't bet on a single country. At the end of 1989, Japan made up an estimated 40% or so of the global stock market – today, according to the UBS Global Investment Returns Yearbook, it's only just over 6%. Nobody knows in advance which country will be the next Japan. In my analysis of MSCI data (net return in euros, end of 2000 to end of 2025), the MSCI World returned +357% (6.3% a year), the MSCI Japan only +128% (3.4% a year).

With single stocks it's even more extreme. Finance professor Hendrik Bessembinder analysed all of the roughly 26,000 US stocks since 1926 (Journal of Financial Economics, 2018): 58% of stocks failed to beat even short-term US Treasury bills over their lifetime. The market's entire excess return of $35 trillion came from just 4.3% of companies. So the chance of picking exactly one of the few big winners with a single stock is small – with a broad index, you automatically own the winners.

4. Keep your savings plan running in a crash

How does a savings plan hold up in the worst case? Using MSCI's monthly data, I calculated how a euro investor in the MSCI World would have fared in the “lost decade” from 2000 to 2009 (price index in euros, without dividends, costs or taxes):

MSCI World in euros, 2000–2009	Paid in	Value end of 2009	Result
Lump sum at end of 1999	€12,000	€6,900	−42.5%
Savings plan €100/month, 2000–2009	€12,000	€10,321	−14.0%

My own calculation using MSCI index data (price index, EUR), purchases at each month-end, without costs, taxes or dividends. Not a real ETF.

To be completely honest: the savings plan was also in the red after ten years. But it reduced the damage compared with the lump sum by about two thirds – because it bought cheaply during the crashes. With dividends (net return, savings plan 2001 to 2009) it was only down 2.7%. And anyone who simply kept going turned €31,200 paid in from January 2000 to the end of 2025 into around €93,700 with the savings plan – without dividends.

It's even clearer in the financial crisis (net return in euros): a savings plan that started in June 2007, shortly after the peak, was back in the black from December 2009 – just over three years earlier than a lump sum invested at the peak, which only got back to break-even in February 2013.

5. Stay away from leverage

Almost all truly catastrophic crashes have one thing in common: borrowed money. In 1929, many buyers paid only 10% themselves. Before the financial crisis, the five big US investment banks were operating with leverage of up to 40 to 1, according to the commission of inquiry set up by the US Congress – so a drop in asset values of less than 3% could wipe out a firm (FCIC Final Report, 2011). And according to the US Securities and Exchange Commission (SEC), investor Bill Hwang grew his family office Archegos from around $1.5 billion to over $36 billion in capital within a year – holding positions of $160 billion through deals on credit (total return swaps). At the end of March 2021, everything collapsed within a few days. Hwang was sentenced to 18 years in prison.

If you invest without credit, you can sit out a crash. If you buy on credit, you have to sell in a crash – no matter at what price.

Halloween bonus: is October really a crash month?

1929, 1987, 2008 – many big crashes happened in October. So is the month cursed? In my analysis of the MSCI World (price index in US dollars, monthly returns 1997 to 2025), October really was the most volatile month and, with October 2008 (−19.0%), also had the worst single month. On average, though, it returned +1.2% and was positive in 62% of years. The weakest month on average was September at −1.0%. And many big crashes didn't start in October at all: the dotcom peak was in March 2000, Lehman went bust in September 2008, and the Covid crash came in February and March 2020. So the October curse is a myth with a grain of truth: the month is nervous, but not systematically bad.

Mark Twain summed it up as early as 1894 in Pudd'nhead Wilson:

“October. This is one of the peculiarly dangerous months to speculate in stocks in. The others are July, January, September, April, November, May, March, June, December, August, and February.”

Mark Twain, Pudd'nhead Wilson (1894)

And then there's the Halloween indicator: the observation that stocks do better on average from November to April than from May to October (“Sell in May and go away”). The economists Bouman and Jacobsen found this in 36 of 37 countries in 2002 (American Economic Review), and a large follow-up study in 2021 found the winter half-year ahead by around 4 percentage points. Statistically, Halloween is therefore more of a good day to get in than to get out. But be careful: the effect is disputed in research, and if you get out in summer, you pay costs and taxes and can miss rallies. It's not a reason for market timing.

Conclusion: the scariest crash is the one in which you sell

The history of stock market crashes really is scary: −89% in 1929, −22.6% in a single day in 1987, 34 years of waiting in Japan, a DAX company with missing billions. But it also shows a clear pattern: those who lost most were the ones who bought on credit, bet everything on one card or sold during the crash. Those who stayed in – broadly diversified, without leverage and with a running savings plan – were eventually back on top after most crashes, sometimes faster, sometimes much more slowly.

Warren Buffett put it this way in his shareholder letter for 1986:

“…we simply attempt to be fearful when others are greedy and to be greedy only when others are fearful.”

Warren Buffett, letter to Berkshire shareholders for 1986

Now I'm curious: which crash did you live through yourself – and what did you do back then? Feel free to write it in the comments under the video.

And if you're still looking for an account for your savings plan: at Smartbroker+ you currently get a McDonald's share worth €20 for free through my link*.

Links marked with * are partner links. If you open an account through them, you support my work – at no extra cost to you. Thank you!