2010 Flash Crash: When Markets Broke in Minutes

Key Takeaways

• On May 6, 2010, the Dow fell about 1,000 points (roughly 9 percent) and rebounded within minutes.
• A $4.1 billion automated E-mini futures sale triggered a liquidity collapse across stocks and ETFs.
• Over 20,000 trades printed 60 percent or more off, some at a penny, and were later cancelled.
• It forced single-stock circuit breakers, limit up-limit down bands, and a stub-quote ban.

Background

By spring 2010, US stock trading had moved almost entirely onto fast electronic venues. Orders no longer flowed mainly through human specialists on one floor. They were split across dozens of exchanges and dark pools and matched in microseconds by algorithms, with high-frequency trading firms acting as a large share of the day-to-day liquidity.

These firms were not obligated market makers in the old sense. They quoted on both sides of the book because it was profitable, and they could withdraw whenever conditions turned against them. According to the CFTC study of the event, in the three days before May 6, high-frequency traders accounted for 34.22 percent of E-mini S&P 500 futures volume and designated market makers another 10.49 percent. A great deal of the market's apparent depth rested on participants who had no duty to keep standing there.

The mood on May 6 was already tense. European debt fears, centered on Greece, were spilling into global markets. US stocks opened weak and kept sliding through the early afternoon. Volatility was rising and quoted liquidity was thinning well before the crash itself began, which left the order book unusually shallow when the shock arrived.

The E-mini S&P 500 future sits at the center of US equity pricing. With the index near 1,000 points, each contract was worth roughly $50,000, so a large futures program could move enormous notional value quickly and pull cash stocks and exchange-traded funds along through index arbitrage.

What Happened

The acute phase lasted about half an hour. The joint CFTC-SEC report and the CFTC's economic study lay out a tight, dated sequence.

• 2:32 p.m. ET: A large fundamental trader, a mutual fund complex later identified in press reporting as Waddell & Reed, began selling 75,000 E-mini contracts, valued at approximately $4.1 billion, as a hedge against an existing equity position. The order went through an automated "Sell Algorithm" set to target 9 percent of the previous minute's trading volume, with no limit on price or time.
• 2:32 to 2:45 p.m.: As prices fell, the algorithm kept feeding orders. It sold about 35,000 contracts, roughly $1.9 billion of the intended 75,000, in just over thirteen minutes. High-frequency firms bought first, then resold to one another.
• 2:45:13 to 2:45:27 p.m. (Chicago time): At the most violent point, high-frequency traders exchanged more than 27,000 E-mini contracts, about 49 percent of total volume, while their net position changed by only about 200 contracts. This was selling passed back and forth, not real demand.
• 2:45:28 p.m.: Buy-side depth in the E-mini had collapsed to under 1,050 resting contracts, less than 1 percent of the morning's level. The CME's Stop Logic Functionality paused E-mini trading for five seconds to stop a further cascade.
• 2:45:33 p.m.: Trading resumed, prices stabilized, and the E-mini began to recover, with the SPY exchange-traded fund following.
• Roughly 2:40 to 3:00 p.m.: Even as futures recovered, sell orders in individual stocks hit empty order books. Over 20,000 trades across more than 300 securities executed at prices 60 percent or more away from their 2:40 p.m. levels, some as low as a penny and some as high as $100,000.
• By 3:08 p.m.: E-mini prices were back near pre-drop levels and most securities had reverted to prices reflecting true consensus values.

By the close, the Dow had recovered the bulk of an intraday swing the SEC-CFTC report measured at 1,010.14 points. The headline drop of about 998 points, near 9 percent, had unwound in less time than a lunch break.

Why It Happened

The crash was not caused by fresh catastrophic news. It was a structural failure inside the trading machinery, set off by one badly calibrated order meeting a thin, fast market.

The trigger was the Sell Algorithm's design. By targeting a fixed share of recent volume and ignoring price and time, it sold faster as activity rose. When high-frequency firms began churning contracts among themselves, that churn counted as volume, so the algorithm read it as a chance to sell more. The CFTC study notes that an identical 75,000-contract sale had been executed before over a whole day with little disruption; the difference on May 6 was that the same size was pushed out in about twenty minutes.

Then came the "hot potato" effect. High-frequency traders bought the futures the algorithm dumped, but they did not want to hold a falling position, so they sold almost immediately, often to other high-frequency traders doing the same thing. Inventory passed around the group while the price fell and reported volume spiked. As the CFTC paper puts it, high-frequency traders "did not significantly change their total inventory even as the prices were rapidly falling."

The next stage was liquidity withdrawal. As volatility exploded, many automated firms widened their quotes or pulled out entirely to avoid being run over. With the largest providers gone, the futures order book hollowed out to under 1 percent of its morning depth, and there was almost nothing left to absorb sell orders.

The damage then spread from futures to everything else. Index arbitrage and cross-market hedging carried the E-mini decline into the SPY and into individual S&P 500 stocks. Retail and institutional stop-loss orders, designed as protection, triggered into a void. With real bids gone, those market orders matched against stub quotes, the placeholder prices market makers post at absurd levels, far from the market, only to satisfy a duty to quote on both sides. That is how a share could print at a penny while another printed at $100,000. The protection mechanism became the accelerant.

By the Numbers

• Intraday Dow drop: about 998 points, roughly 9 percent, with a measured intraday point swing of 1,010.14, the largest intraday point move to that date. (SEC-CFTC, Findings Regarding the Market Events of May 6, 2010)
• The trigger order: 75,000 E-mini S&P 500 contracts, about $4.1 billion, sold by a large fundamental trader as a hedge. (CFTC study; SEC-CFTC report)
• Algorithm setting: target 9 percent of the prior minute's volume, with no price or time limit. (CFTC study)
• First leg: about 35,000 contracts, roughly $1.9 billion, sold between 2:32 and 2:45 p.m. (CFTC study)
• Hot potato peak: over 27,000 contracts traded in about 14 seconds, near 49 percent of volume, on a net position change of only about 200 contracts. (CFTC study)
• Depth collapse: buy-side E-mini depth fell to under 1,050 contracts, below 1 percent of the morning level. (CFTC study)
• Pause: a five-second CME Stop Logic halt at 2:45:28 p.m. (CFTC study)
• Broken trades: over 20,000 trades across more than 300 securities executed 60 percent or more from 2:40 p.m. prices, later cancelled. (Federal Reserve Bank of New York; SEC-CFTC report)
• Price extremes: trades as low as one cent and as high as $100,000 against stub quotes. (Federal Reserve Bank of New York; SIFMA)
• Report: the joint SEC-CFTC report was published September 30, 2010. (SEC-CFTC report)

Aftermath

No firm went bankrupt and no rescue was needed, but the event exposed gaps that regulators moved to close. The first response was speed-limit infrastructure for stocks.

In June 2010, exchanges and the SEC introduced single-stock circuit breakers that paused trading in a name after a large, fast move. That pilot was later replaced by the Limit Up-Limit Down (LULD) plan, which prevents trades outside a price band set around a stock's recent average and pauses the stock if it cannot return inside the band. The aim was to stop any one security from printing at a penny again.

In November 2010, the SEC effectively banned stub quotes, requiring market-maker quotes to stay within a defined percentage of the national best bid and offer rather than sitting at placeholder levels far from the market. The market-wide circuit breakers were also rebuilt. The old framework keyed off Dow points; the revised system, phased in around 2012, halts the whole market on S&P 500 declines of 7 percent (Level 1), 13 percent (Level 2), and 20 percent (Level 3). Regulators also advanced the Consolidated Audit Trail, a system to record every order and quote across venues so a future event could be reconstructed quickly.

The legal aftermath took years and centered on one trader. In April 2015, the CFTC charged U.K. resident Navinder Singh Sarao and his firm Nav Sarao Futures Limited with manipulation, attempted manipulation, and spoofing of the E-mini S&P 500. The CFTC alleged he ran a "Layering Algorithm" that placed four to six large sell orders away from the best price and cancelled most before execution, and that on May 6 he used it "continuously, for over two hours," applying close to $200 million of downward pressure that "contributed to market conditions that led to the Flash Crash." The Department of Justice charged him in parallel.

Sarao pleaded guilty on November 9, 2016, before U.S. District Judge Virginia M. Kendall in the Northern District of Illinois, to one count of wire fraud and one count of spoofing. He admitted to making at least $12.8 million in illicit gains. On January 28, 2020, Judge Kendall sentenced him to time served, the roughly four months he had already spent in custody, plus one year of home confinement, citing his cooperation and his autism diagnosis. Sarao was not found to have single-handedly caused the crash; regulators were always clear it had multiple causes, and his spoofing was one contributing factor among the structural failures the joint report described.

Lessons for Investors

1. Quoted depth is not the same as real liquidity. The futures book looked deep all morning, then fell below 1 percent of that depth in minutes once high-frequency firms stepped away. If your plan assumes you can always sell into a thick market, it will fail in the seconds that matter. Treat available liquidity as something that can vanish, not a constant.

2. Automation removes the human pause. The Sell Algorithm kept selling because it only watched volume, not price, and there was no person to ask whether something had gone wrong. Any system with feedback rules and risk-off triggers can spiral. Know what your own automated orders and stops will do when prices move violently against you.

3. Stop-loss orders are not free protection. Plain stop-loss orders became market orders that filled against stub quotes at a penny on May 6. A stop-limit order caps the price you accept but risks not executing at all. Understand that trade-off before relying on stops as your safety net in fast markets.

4. Crowded liquidity providers fail together. Because many high-frequency firms ran similar risk rules, they all pulled out at once, turning many small backstops into one big absence. A hedge or a liquidity source that everyone shares can disappear simultaneously, which is exactly when you need it.

5. Recovery does not erase the damage you locked in. Prices were back near normal by 3:08 p.m., but anyone whose order filled at a 30 percent discount that was not broken kept that loss. A market that round-trips in half an hour can still permanently harm a portfolio that was forced to transact in the gap.

Frequently Asked Questions

What was the 2010 Flash Crash in simple terms? The 2010 Flash Crash was May 6, 2010, when the Dow fell about 1,000 points, roughly 9 percent, in minutes and then recovered most of it. A large automated futures sale drained electronic liquidity, briefly sending some stocks to a penny.

Why did the 2010 Flash Crash happen? A $4.1 billion automated sale of E-mini S&P 500 futures was set to track trading volume without regard to price. As high-frequency firms churned contracts and then pulled their quotes, the order book emptied and prices collapsed before recovering.

How much money was lost in the 2010 Flash Crash? About $1 trillion in market value briefly evaporated at the lows, though most of it returned by the close. The lasting losses fell on investors whose orders filled at extreme prices that were not later cancelled, including fills as low as a penny.

Could the 2010 Flash Crash happen again today? A similar cascade is harder now. Limit up-limit down bands pause individual stocks, stub quotes are banned, and market-wide circuit breakers halt trading at 7, 13, and 20 percent. The underlying risk of automated, crowded liquidity withdrawing at once still exists.

What is the main lesson from the 2010 Flash Crash? Liquidity in fast electronic markets can disappear in seconds, and protective tools like stop-loss orders can fill at disastrous prices when it does. Size and structure your positions to survive a sudden void rather than assuming a deep market is always there.

Sources

1. U.S. Commodity Futures Trading Commission and U.S. Securities and Exchange Commission. Findings Regarding the Market Events of May 6, 2010 (joint report, September 30, 2010). https://www.sec.gov/files/marketevents-report.pdf
2. Kirilenko, A., Kyle, A. S., Samadi, M., and Tuzun, T. The Flash Crash: The Impact of High Frequency Trading on an Electronic Market. CFTC Office of the Chief Economist. https://www.cftc.gov/sites/default/files/idc/groups/public/@economicanalysis/documents/file/oce_flashcrash0314.pdf
3. U.S. Commodity Futures Trading Commission. Press Release 7156-15: CFTC Charges U.K. Resident Navinder Singh Sarao and His Company Nav Sarao Futures Limited PLC with Price Manipulation and Spoofing. https://www.cftc.gov/PressRoom/PressReleases/7156-15
4. Federal Reserve Bank of New York, Liberty Street Economics. The Flash Crash, Two Years On. https://libertystreeteconomics.newyorkfed.org/2012/05/the-flash-crash-two-years-on/
5. SIFMA. The 10th Anniversary of the Flash Crash. https://www.sifma.org/research/insights/10th-flash-crash-anniversary
6. CNBC. Flash crash trader Navinder Singh Sarao sentenced to home detention (January 29, 2020). https://www.cnbc.com/2020/01/29/flash-crash-trader-navinder-singh-sarao-sentenced-to-home-detention.html

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.