In a matter of minutes, something like a trillion dollars in value vanished and then mostly came back. Some shares changed hands for a penny; others briefly went for a hundred thousand dollars.
On the afternoon of May 6, 2010, the American stock market did something it had never quite done before. Over a handful of minutes, the Dow Jones Industrial Average dropped by roughly a thousand points, about nine percent, and then clawed back most of that loss within roughly twenty minutes. Something on the order of a trillion dollars in value disappeared and largely returned in less time than it takes to eat lunch.
The details on individual stocks are stranger than the headline number suggests. Shares of well-established companies traded at a single penny. Others, briefly, traded near a hundred thousand dollars. These weren't typos or glitches in a data feed. They were real transactions, in the shares of real companies whose underlying businesses hadn't changed one bit during those few minutes.
What caused it is still argued over in the details, but the broad mechanism is well understood. A large automated sell order hit a market that was already jittery. Automated trading firms, which normally stand ready to buy and sell and give investors the continuous availability of a counterparty they take for granted, responded to the chaos by pulling back. Some stopped trading altogether. Others kept quoting prices, but ones so far removed from anything reasonable that they were effectively out of the market too.
Here's the essential fact of the whole episode: with the usual buyers and sellers gone, orders arriving in the market found almost nobody on the other side. An order to sell at whatever price is available gets executed at whatever price is available, and if the only bid left standing happens to be a penny, that's the price the trade prints at. Those numbers weren't judgments about what anything was worth. They were simply whatever was left in the order book once everyone else had cleared out.
This exposes something about prices most investors never have reason to think about. A price isn't a fixed property of a stock the way mass is a property of an object. A price is the result of a transaction, and a transaction needs someone on the other side willing to take it. Under normal conditions there's always a crowd of willing counterparties, and that abundance is so dependable that investors start treating "there is always a price" as a law of nature. The Flash Crash showed it isn't a law of nature. It's a service, provided voluntarily by participants who can walk away from it.
That withdrawal wasn't irrational, and it's worth understanding rather than resenting it. A firm whose business is standing ready to trade does so on the assumption that it can manage the risk of doing so. When conditions get disorderly enough that nobody can tell what anything is worth, the sensible move is to stop, and every such firm tends to reach that conclusion at roughly the same moment, for roughly the same reasons. Liquidity doesn't drain away gradually. It disappears all at once, collectively.
There's a direct, practical implication buried in this about the mechanics of placing an order, one of the few genuinely useful pieces of market plumbing an ordinary investor needs to know. An instruction to trade at whatever price is available will, in fact, be executed at whatever price is available, and during those minutes that meant prices no sane person would have accepted. An instruction that sets a limit, a price beyond which you refuse to transact, simply doesn't get filled under those conditions. That's a considerably better outcome.
Many of the most extreme trades from that afternoon were later canceled by the exchanges as clearly erroneous. That offers less comfort than it sounds like. The cancellation was a discretionary call, made after the fact, applying thresholds set after the fact, and an investor whose trade happened to fall just outside those thresholds got no relief at all. Counting on the chance that someone else will reverse a bad outcome on your behalf is not a risk management strategy.
The deeper lesson is about the relationship between price and value, and this episode separates the two with unusual clarity. During those few minutes, the companies whose shares traded at a penny were worth exactly what they'd been worth an hour before. Nothing about them had changed. The price moved wildly; the value sat still; the two simply went their separate ways for a while. Ordinarily this gap is invisible because it's small. The Flash Crash blew it up large enough to see.
An investor who has absorbed this treats a falling price with more caution and considerably less reverence. The price is telling you what somebody was willing to pay at one particular instant, under whatever conditions happened to prevail at that instant, and those conditions may have had nothing to do with the business itself. It's information. It just isn't necessarily information about the company.
There's a further point worth making about how quickly this kind of understanding fades. The episode is now sixteen years in the past, and the safeguards introduced afterward have worked well enough that most investors alive today have never seen anything like it. That's the usual pattern: a rare event happens, prompts a round of soul-searching and reform, and then fades from memory precisely because the reforms seem to have worked. The risk is mistaking the absence of a repeat for proof that the underlying vulnerability is gone, when it may simply mean the right conditions haven't come together again. Liquidity is still a favor, not a right. Nothing compels anyone to stand ready to buy what you want to sell. The mechanism that produced that afternoon was never repealed.
At VESTFY™, the Flash Crash is taught as the clearest available proof that a market is a mechanism, not an oracle. It works beautifully almost all the time, and its reliability is easy to mistake for a fact of nature. It isn't one, and an investor who understands that will be a lot less inclined to take whatever the mechanism happens to print as a verdict on what they own.