When the Markets Started Selling Without Knowing What to Sell
Some securities briefly traded for a penny. The Flash Crash was neither a simple outage nor the work of one person.
On 6 May 2010, for a few minutes, Wall Street stopped producing prices that made sense. Shares and exchange-traded funds changed hands for a penny. Then, almost as quickly, the screens returned to normal.
One cent. That is the price at which some stocks and ETFs briefly traded during the Flash Crash. Not because those companies had suddenly become worthless. Not because a human being had chosen to give them away. But because, at the moment the market was supposed to do its basic job — match buyers with sellers — the buyers able to stand their ground became scarce.
The scene was strange enough to look like a simple technical failure. It was not. It came from a market already under stress, an enormous automated sale, and algorithms reacting to the speed of a move rather than its meaning.
A session already shaking
On 6 May 2010, US markets did not begin the day calmly. Anxiety over Europe’s sovereign-debt crisis weighed on prices. Investors sold, prices fell, screens flickered. None of that yet looked historic. It was a bad trading day, one among many.
At about 2:32 p.m., however, a large mutual fund began a sell programme in the E-mini S&P 500 futures contract, a heavily traded product tied to the index. The order covered 75,000 contracts, worth more than $4 billion.
What mattered was not just its size. The algorithm executing it followed a mechanical rule: it would account for 9% of the volume traded in the previous minute. It did not slow because prices were falling. It did not stop to ask whether enough patient buyers were still on the other side. The more trading volume accelerated, the more it could keep feeding into the market.
In a calm market, that kind of rule can look neutral. In a nervous one, it can become a ramp.
Volume began feeding on itself
The first sales drew in other algorithms. Some bought and sold back quickly. Others followed the momentum. Market makers — the firms that normally post bids and offers and absorb part of a shock — saw orders arriving too fast and became cautious. Several reduced their activity or stepped away.
There was no single red button. There was a loop. Selling increased volume. More volume allowed the original algorithm to sell more. Other programmes saw the move and acted in turn, while the liquidity that made the market look solid withdrew just when it was needed most.
The joint SEC/CFTC report describes that interaction closely: the original sale did not by itself explain the size of the plunge, but it met a system able to turn strong pressure into a sudden rout.
Within minutes, the E-mini contract and the SPY ETF fell by about 5%. The Dow Jones then dropped nearly 1,000 points at its intraday low. That number travelled around the world. Yet inside the market, the most unsettling part was not only the fall in the index. It was what happened next.
When a price meant nothing
As markets began to recover, some shares and ETFs started showing trades that looked like the output of a broken test. Securities were executed at one cent, or at similarly absurd prices. Others traded at wildly high levels.
Those were not the “real” values of the companies involved. They were the traces of a moment when orders met in a fragmented market with too little depth and too few safeguards to stop certain executions from going through.
Exchanges later cancelled some of the clearly erroneous trades. That kept the episode from becoming permanently disastrous for everyone whose order had hit the wrong price. But the image remained: for a few minutes, one of the world’s most liquid markets was producing labels that described almost nothing.
A price exists only while somebody holds it
The Flash Crash exposed something markets usually hide very well. A price is not a truth stored somewhere. It is a proposal, held by buyers and sellers willing to stay in place long enough to meet one another. When the participants able to buy into a fall step back, orders may remain, but there is no longer enough depth to make them meaningful.
Market makers do not vanish because they have stopped believing in the economy. They cut back because they no longer know what information the move contains, or how fast it will continue. That response can be reasonable for each firm and still become dangerous for the system as a whole. The market does not run out of screens or trades. For a few moments, it runs out of people prepared to risk their own money to say, “at this price, I will buy.”
The tempting single villain
Simple explanations appeared immediately. Uncontrolled machines. A typo. A hacker. Later, Navinder Sarao’s name became central: in 2015, the CFTC charged him with using manipulation and spoofing techniques that allegedly contributed to the conditions leading to the Flash Crash.
The crucial word is “contributed.” The case did not turn the event into a one-man plot. Regulators had already described a broader chain: a fragile session, a large sale executed with no price or time limit, automated strategies reacting to one another, and liquidity providers unwilling to remain in front of the orders.
The Flash Crash was not a story about one machine going mad. It was a story about rules that looked reasonable in isolation and became dangerous when they met in the same move.
After 6 May, regulators strengthened trading pauses and mechanisms designed to stop a share or a whole market from crossing absurd price territory too quickly. Those safeguards do not abolish volatility. They acknowledge something more unsettling: in an electronic market, speed is not only an advantage. It can also take away the time a price needs to exist.
The minutes when nobody really knew how to buy
The Flash Crash did not permanently erase a trillion dollars. Prices largely rebounded, and the most extreme trades were corrected. But the danger that afternoon was not imaginary.
For a few minutes, markets did not need human panic to drift away from any readable value. They only needed orders responding to one another faster than buyers could decide what they were still willing to pay.
Leave a Reply