Home › Stories › Market history

The flash crash of 2010: the Dow's 1,000-point drop in minutes

On the afternoon of 6 May 2010, the Dow dropped nearly 1,000 points in minutes, then won most of it back almost as fast. For a few strange moments, some big-company shares traded for one cent.

Story · 7-minute read · Updated 2026-10-21

The world before the crash

In spring 2010, markets were still nervous after the 2008 crisis. The big worry that week was Greece, which was struggling to pay its debts. Riots in Athens were on the news.

On 6 May, stocks were already down for the day by early afternoon. Traders were jumpy. Buyers were fewer than usual.

By 2010, most trading in the US was done by computers, many of them run by high-speed firms that bought and sold in fractions of a second.

What happened

The official joint report by the SEC and CFTC, released on 30 September 2010, laid out the key events.

At about 2:32 pm ET, a large fund company began selling 75,000 E-mini S&P 500 futures contracts, worth about $4.1 billion. Reports later named the firm as Waddell & Reed. The report said the computer program doing the selling was set to sell based on trading volume, without regard to price or time.

As prices fell, high-speed traders bought some of those contracts, then quickly sold them on to each other and back into the market. Trading volume soared. Because the selling program watched volume, it sold faster.

Buyers stepped back. With few buyers left, prices fell through the floor.

The moment everyone remembers

Between roughly 2:40 and 2:45 pm, the drop turned into free fall.

At its worst, the Dow was down nearly 1,000 points, about 9% on the day. Close to a trillion dollars of market value, by many estimates, disappeared on paper.

Then, at 2:45:28 pm, the futures exchange's own safety system paused E-mini trading for 5 seconds. That tiny break gave buyers time to come back.

Prices bounced almost as fast as they had fallen. By about 3:00 pm, most of the loss was gone.

But inside those minutes, things got very strange in single stocks:

• Shares of Accenture, a large consulting company, traded at $0.01.
• Some other shares traded at prices as high as $100,000.

This happened because some firms used "stub quotes": placeholder orders set at absurd prices, never meant to be filled. When real buyers vanished, those were the only orders left.

Exchanges later cancelled trades that were more than 60% away from earlier prices. The Dow closed down 347.80 points, about 3.2%.

What changed afterwards

• Single-stock circuit breakers were added in 2010, pausing a stock that moves too far too fast.
• These were later replaced by the Limit Up–Limit Down rules, which stop trades outside a set price band.
• Stub quotes at silly prices were banned.

In April 2015, US authorities charged a London trader, Navinder Sarao, with spoofing (placing large orders he meant to cancel) on that day and others. He pleaded guilty in 2016 to wire fraud and spoofing and was sentenced in 2020 to time served plus a year of home detention. How much he added to the crash is still debated.

What it teaches

1. A stop-loss can fill far from your price

A normal stop turns into a market order when it's hit. That means "sell at the best price available now". In a flash crash, the best price available can be terrible. Some stops on 6 May filled at prices that had nothing to do with the stock's real value.

Read stop-loss vs stop-limit to see the trade-off: a stop-limit won't fill below your limit, but it may not fill at all.

2. Market orders are risky when markets are thin

A limit order sets the worst price you'll accept. A market order doesn't. On calm days the difference is small. When buyers disappear, it can be huge. See market vs limit order.

3. Liquidity can vanish in seconds

Thousands of orders on screen look like a deep market. Many of them can be pulled in an instant. That's why liquidity is about who's willing to trade right now, not how busy the market looked a minute ago.

4. Size still decides how much it hurts

You can't stop a flash crash. You can keep your size small enough that a bad fill is annoying, not life-changing. Work out each trade with the position size calculator.

What it doesn't mean

• It doesn't mean computers are bad for markets. On most days, they help keep spreads tight. The problem was how a few systems interacted under stress.
• It doesn't mean one person caused it. The official report pointed to the large sell program in thin conditions. Sarao's role was charged later and is still argued about.
• It doesn't mean "buy the flash crash". The bounce on 6 May was fast. There's no rule that says the next one will be.
• It doesn't mean stops are useless. They still cap losses on almost every normal day. They just aren't a guarantee.

Timeline

Questions