Flash Crashes: What Causes Them & How to Protect Your Positions

What Is a Flash Crash?

A flash crash is a rapid, deep decline in securities prices followed by a quick recovery, typically unfolding in minutes rather than days or weeks. Unlike traditional stock market crashes that develop over weeks, flash crashes compress extreme volatility into an absurdly short window.

Flash crashes are a modern phenomenon, emerging as electronic and algorithmic trading replaced human market makers. When algorithms react to the same signals simultaneously, selling begets selling at machine speed. Human traders cannot react fast enough to provide a counterweight, and the result is a temporary collapse in orderly price discovery.

These events expose a fundamental vulnerability in modern markets: the liquidity that appears abundant under normal conditions can vanish in seconds when stress emerges.

The 2010 Flash Crash

The most notorious flash crash occurred on May 6, 2010. At 2:32 PM Eastern, the Dow Jones Industrial Average began plunging. Within 15 minutes, the Dow dropped approximately 998 points — nearly 9% — before rebounding almost entirely by 3:07 PM.

Individual stocks experienced even wilder moves. Accenture (ACN) traded down to $0.01 per share. Procter & Gamble (PG) dropped 37% in minutes. Some ETFs traded at prices that were 60% below their net asset value. Meanwhile, a handful of stocks briefly traded at absurd premiums.

The trigger was a large sell order from a Kansas City mutual fund company — a $4.1 billion E-mini S&P 500 futures contract executed via an algorithm that was indifferent to price and time. As this sell pressure hit the market, high-frequency trading firms detected the imbalance and pulled their bids simultaneously, creating a liquidity vacuum.

What Causes Flash Crashes

Flash crashes result from the interaction of several structural vulnerabilities in modern electronic markets.

Algorithmic herding. When thousands of algorithms use similar signals and strategies, they can all attempt to sell at the same time. This creates a one-sided order book with massive sell pressure and no buyers.

Liquidity withdrawal. High-frequency market makers provide the majority of quotes in modern markets. Unlike the old specialist system where designated market makers had an obligation to maintain orderly markets, today's electronic market makers can withdraw instantly when conditions deteriorate. When they pull bids simultaneously, the order book empties.

Stop-loss cascades. As prices drop, automated stop-loss orders trigger, adding sell pressure. These triggered stops become market orders, which execute at whatever price is available — potentially far below the stop price. Each triggered stop pushes prices lower, triggering more stops in a cascade.

Fat-finger errors and spoofing. Some flash crashes have been linked to erroneous orders or deliberate manipulation. Navinder Sarao was convicted of spoofing — placing and canceling thousands of large orders to manipulate prices — in connection with the 2010 crash.

Notable Flash Crashes Beyond 2010

August 24, 2015. The Dow dropped over 1,000 points at the open due to concerns about China's economy. Many ETFs traded at steep discounts to their underlying holdings. AAPL opened 8% below its previous close before recovering.

February 5, 2018. The VIX spiked from 17 to 50 in a single session, destroying several volatility-linked products. The XIV (inverse VIX ETN) lost 96% of its value overnight and was subsequently terminated. This event demonstrated how complex financial products can amplify flash crash dynamics.

March 2020 liquidity events. While not a single flash crash, the COVID sell-off included multiple intraday flash-crash-like moves. Circuit breakers were triggered four times in ten days — the first triggers since their implementation after the 1987 crash.

Individual stock flash crashes occur regularly and receive less attention. A single large algorithmic order can briefly crash a mid-cap stock 10-20% before recovery.

Circuit Breakers and Regulatory Responses

After the 2010 Flash Crash, regulators implemented several safeguards through circuit breakers.

Market-wide circuit breakers halt all trading when the S&P 500 declines 7% (Level 1), 13% (Level 2), or 20% (Level 3) from the prior close. Level 1 and 2 trigger 15-minute halts before 3:25 PM. Level 3 halts trading for the remainder of the day.

Limit Up-Limit Down (LULD) bands prevent individual stocks from trading outside a defined percentage range from their recent average price. If a stock hits the band, trading pauses for five minutes. This prevents the $0.01 trades that occurred in 2010.

Clearly erroneous trade rules allow exchanges to cancel trades that occur at prices substantially away from the prevailing market. After the 2010 crash, exchanges canceled trades that were more than 60% away from pre-crash prices.

How to Protect Your Portfolio

Flash crashes create both risk and opportunity. The key is preparation.

Use limit orders exclusively. Market orders during a flash crash can fill at catastrophic prices. A limit order ensures you only sell at or above your specified price, even if it means the order goes unfilled during the chaos.

Set stop-losses with care. Stop-loss orders convert to market orders when triggered, making them vulnerable to flash crash slippage. Consider using stop-limit orders instead, or use mental stops and execute manually. Position sizing through proper risk management is a more reliable protection than stops during extreme events.

Avoid trading during peak volatility. If circuit breakers are triggering or volatility is spiking, sitting on the sidelines is a valid strategy. Flash crashes resolve quickly — you lose little by waiting for conditions to normalize.

Be ready to buy. Flash crashes occasionally create irrational prices in fundamentally sound stocks. Having limit buy orders (good-till-canceled) at significant support levels or deep value prices allows you to capitalize on momentary panic.

Can flash crashes be predicted?

Not with precision, but conditions that increase flash crash risk are identifiable. Low liquidity environments (after-hours, holidays), elevated VIX, deteriorating market breadth, and high algorithmic trading volume all raise the probability. Events with binary outcomes — earnings, Fed decisions, geopolitical developments — can serve as triggers.

Are flash crashes becoming more or less common?

Individual stock flash crashes remain common, but market-wide events have become less frequent since the implementation of LULD bands and updated circuit breakers. The structural vulnerabilities remain, however. As algorithmic trading continues to grow as a percentage of total volume, the conditions for flash crashes persist.

Should I sell during a flash crash?

Almost never. Flash crashes reverse quickly, and selling during one virtually guarantees selling at or near the worst possible price. If you have no leverage and own fundamentally sound positions, the correct response is usually to hold — or to buy if you have available capital and the conviction to act during panic.