Flash Crashes: What They Are, How They Work

Remember when a tulip bulb was worth a year’s wages? Let’s take a quick (but important) trip through history. 👇

In 1634, the tulip was introduced to the Dutch Republic and it quickly became a hit. Speculative investments exploded and the prices by late 1636 were reported as reaching the equivalent of tens of thousands of modern dollars. 💵

In early 1637 the entire thing completely collapsed and the tulip was humbled. While this tulipmania is often considered the first real financial bubble in history, it bears certain similarities to flash crashes as well—and their cousins: flash spikes.

Considering that life was far slower in the 17th century, the speed at which the entire thing unfolded was lightning fast, and, much like many of the modern flash crises, it didn’t really have an impact on the overall Dutch economy at the time.

Furthermore, very similar to modern flash crashes, the initial dramatic price change didn’t have much relation to any realistic factors—they were a novelty and the interest was driven solely by sentiment.

While it is true that the price collapse was more of a price correction than a flash crash, this isn’t that dissimilar to many historic flash crashes. Indeed, while the age of the modern flash crash came about in 2010, it was decades in the making and encompasses events like the 1962 Kennedy slide, and the 1987 black Monday before finally arriving into the 21st century and the digital era—changing its pace from months to minutes.

So, let’s find out what causes these crises nowadays, what they are, and how you can be like the Dutch Republic—and survive them unscathed.

What Exactly is a Flash Crash? 👨‍🏫

A flash crash is a deceptively simple concept. In a nutshell, sometimes prices of certain assets face a sharp and sudden drop—from hundreds of dollars to mere cents in some extreme cases—before recovering just as quickly.

Most flash crashes happen within minutes leaving little to no long-lasting effects. This is where the deceptive part comes in. The swiftness with which this phenomenon happens makes it fairly hard to study and understand.

Indeed, while there are multiple theories as to what causes them—some of which appear very reasonable and reasonably well-founded—the general consensus appears to be that nobody is quite sure how a flash crash happens.

On the other hand, we are completely certain that they happen a lot. CNN reported already in 2013 that there are about 12 mini flash crashes each day, and by 2017 there were an estimated 20,000 flash crashes since 2006.

What Causes a Flash Crash? 📚

While no flash crash appears to have caused serious damage, both the fact that they are common, and that they aren’t very well-understood makes studying them more than worthwhile. So, what exactly causes a flash crash?

Generally speaking, a flash crash occurs when the number of sell orders vastly overwhelms the buy orders. Obviously, the fact that a lot of assets are being sold while not as many are being bought makes them abundant and drives the prices way down—and when the volume of stocks being sold is suddenly balloons, bad things tend to happen.

It is generally believed that there are two main possible originators of any particular flash crash—and they read a bit like a sci-fi novel. One is human—and can be both intentional and accidental—and the other stems from the machines that do much of the trading.

Human Causes 👨‍💻

As is blatantly obvious from other dubious practices like naked short selling, some trades aren’t exactly made in good faith. Sometimes, flash crashes are initiated by investors and traders attempting to spoof the market.

Spoofing is a form of market manipulation where a malicious actor places a large number of orders in hopes they will tank the prices of an asset—making it a nice cheap buy for them—only to cancel those orders after the prices start to dip.

Without proper fail-safes in place, spoofing can be a disturbingly effective tactic—emotions play an important role in determining stock prices, and panic is a very powerful emotion.

On the other hand, not all man-made flash crashes are intentional. A proposed reason for some of these events is what is called the fat finger phenomenon. Basically, this is just another term for human error.

The idea is that somebody selling a part of their portfolio isn’t very precise with the keyboard and types some extra zeroes—you know, the everyday phenomenon of wanting to sell 1,000 shares and typing 1,000,000—and actually having 1,000,000 shares to sell.

Yeah, as you’ve probably guessed by now, most flash crashes aren’t the fault of individual investors but larger institutions—with numerous studies examining the link between high-frequency traders and this phenomenon being made and published over the last decade.

This link can be seen both in human, and machine-caused crashes.

Computer Causes 💻

Arguably, computers play a bigger and more interesting role in flash crashes—so much so that the very first semi-proper flash crash in 1987 coincided with the proliferation of trading software, though there were many factors at play.

Machines tend to play three major roles in any such crash—they can cause it, they can enable and worsen it, and they can, somewhat surprisingly, resolve it.

Kicking off, computers are prone to errors. Glitches happen all the time, and they can send faulty orders, misread the pricing of assets, and even completely lose contact with key servers of a market. All of these have the potential to mess up the values and cause a flash crash.

What is more interesting—and perhaps more symptomatic as to why flash crashes appear to be an illness of the digital age—is when well-functioning systems facilitate these occurrences. 

Essentially, most institutional and many individual traders have multiple stock orders in place at the same time. These are often intended to help an investor avoid losses—or seize a great opportunity—without having to look at the ticker 24/7.

Say that there is a stock valued at $20 that you own. You can set an order for that stock to be sold if it reaches $22 as that can garner you a nice profit, or if it falls to $18 to help you avoid serious losses.

These orders are fairly common and are cited as a major enabler of flash crashes. A computer error, a fat finger incident, or spoofing can cause the price of that stock to suddenly dip to $17, triggering hundreds, if not thousands of stop-loss orders to sell the same shares—driving the prices further into the ground.

Furthermore, many algorithms employed by institutional traders are very fast and very sensitive. They can detect buy and sell orders being placed, and act upon that information before the actual trades have been made—which makes spoofing easier.

Obviously, if such a price drop happens due to valid reasons—severely worsened economic climate, incredibly bad news about a company, or a horrible joke made by a CEO—it tends to herald a real crash. However, price dips preceding a flash crash tend not to reflect reality in any shape or form.

How Computers Help End a Flash Crash? 🤔

Automated stock orders are also often found to stop these occurrences and even help in a swift recovery. How? They are often set up to buy if the deal becomes great. 

If there are thousands of orders set up to sell if the prices fall to $17, there are likely just as many mandating a buy at $15—excellent value for otherwise $20 shares. These tend to quickly shift the balance and overwhelm the previously unsurmountable sell orders thus recovering the prices to around the normal level.

The 2010 Flash Crash 📉

May 6 2010 was a particularly traumatic day for many investors. The situation was already rather volatile—elections were upcoming in the UK, and Greek national debt was just downgraded to junk bonds.

Still, the day kicked off fairly normally. All of this changed abruptly at 2:30 pm when most of the major U.S. indices dropped significantly and suddenly—Dow Jones Industrial Average (DJIA) fell by 1000 points in just 10 minutes. Pretty much everything was affected—stocks, options, futures, and even funds like exchange-traded funds.

To make things even more bizarre, the apocalypse was apparently over in around half an hour—by 3 pm. While few of the affected assets made a full recovery on the same day, most were back up by over 50% by the closing hour.

The damage assessments of the event are somewhat inconclusive to this day. Most agree that about $1 trillion of wealth was destroyed in the crash, and many experts state that pretty much all of it was recovered. This appears to be correct—the complete disappearance of $1 trillion would most certainly be something we’d be feeling for decades to come.

What Caused the 2010 Crash? 📜

The truth behind the events of 2010 remained a mystery for years while a series of lengthy investigations were underway. In the meantime, speculation about the causes ran rampant. Everything from the aforementioned Greek financial crisis to human error was hypothesized.

Later in the same year, the SEC published the findings of its investigation—the general fragility of the market at the time was blamed for the crisis. An interesting note of the report regards a large sale order of E-mini S&P 500 contracts around 2:30 pm—a seemingly innocent event that simply appeared to have been the straw that broke the camel’s back.

However, this innocent event proved more sinister in 2015 when the British police arrested a London suburbanite trader Navinder Singh Sarao for attempting to spoof the market—you guessed it—on May 6 2010 by placing a number of disingenuous sell orders for E-mini S&P 500 contracts.

The saga of this spoofing finally entered its final chapter in 2020 when Singh Sarao was sentenced to one year of home detention.

The Aftermath of the Crash 🧮

In many ways, the flash crash of 2010 was a perfect storm. It highlighted how a single person could disrupt the stock market, how automated trading can worsen a crash, how automated trading can reverse it, and how hard it would be to get to the bottom of such an event. Somebody cynical could also point out it showcased how not to successfully spoof the prices.

This crisis had several key outcomes. The most obvious one is the implementation of various precautions intended to stop and control sudden price movements—both crashes and flash spikes. The second—and possibly more significant—was that it gave us a blueprint on what modern flash crashes can look like. While the Kennedy crash of 1061-1962 is often called a flash crash, it lasted far too long for the modern definition.

How to Prevent a Flash Crash ⚡

Considering that flash crashes tend to be lighting-fast, they can be avoided with some fairly simple measures. Generally, the most common ways of combating these occurrences are the so-called circuit breakers.

These temporarily halt trading in case of sharp and sudden price drops. A fall of 7% tends to stop trading for a few seconds or minutes. 13% will make an asset untradable for 15 minutes and 20% will put a stop to buying and selling for the rest of the day.

A market circuit breaker is intended to give the prices a respite in the wake of the surge in sales and it appears to be working. While it is very hard to tell just how common flash crashes are nowadays—even the 2013 estimate of 12 per day was just an estimate—they are yet to turn into large-scale catastrophic events.

In addition to circuit breakers, the SEC banned naked access to the stock market already in 2010. This means that technically you can only trade vicariously—you have to go through a licensed broker. However, this measure is reportedly compromised due to large investors using their connections to nonetheless access the market directly.

The basic idea behind both of these measures is to slow down the market if things turn strange. Since flash crashes tend to be started by singular but large-scale sales and facilitated by subsequent computer and human panic, giving the market time to update and adjust with everyone blocked from placing orders should theoretically keep it stable.

The caveat to all of this is that we aren’t sure if any of these fail-safes actually work. Since we don’t fully understand how flash crashes occur, we can’t accurately devise plans to combat them.

Diving into Other Examples of Flash Crashes

While the flash crash of 2010 is the most famous example, it was neither the first nor the only such event. Flash crashes can generally be divided into two major groups—digital and pre-digital.

Strictly speaking, most of the pre-digital ones don’t really fit the modern definition—they simply lasted too long—but are still worth a mention. The Kennedy slide of 1962 is often called a flash crash but it really isn’t clear if it indeed is one. Essentially, it represents a sharp decline in the market lasting from 1961 to 1962 which followed decades of growth that started with the end of the Great Depression.

Depending on the interpretation you believe, the Kennedy slide can be seen as the market naturally entering a bearish phase to correct its overgrowth from the 1950s. This was certainly the stance taken by many of the authorities of the time. However, the fact remains that the drop was both sudden and that the SEC never conducted a proper investigation.

1987 Black Monday was in many ways a transitory flash crash. On the one hand, its suddenness and sharpness, and the role computer algorithms played in it are akin to modern flash crashes. On the other hand, unlike such crises in the 21st century, the black Monday crash lasted far, far longer.

For this reason, the flash crash of 2010 is generally regarded as the first true example. However, there have been multiple proper modern flash crashes since then.

2017 Silver Futures Flash Crash 📊

In 2017 the silver futures market experienced a sudden drop in prices of around 11%. The crash happened overnight for American and European markets and it is widely believed it was caused by trading algorithms battling each other and exacerbated by thin trading from Asia.

This flash crash is interesting because the safety procedure of the affected market—CME group’s velocity logic—reportedly worked as intended. This can be taken as both a testament to the effectiveness of such precautions and as proof of their insufficiency.

Either way, the crash ended within hours and the market was mostly unaffected overall.

2017 Ethereum Flash Crash 🪙

The 2017 Ethereum flash crash happened on the since-closed GDAX market. It was, perhaps, the most dramatic in history. It occurred within seconds and saw the prices collapse from $319 to $0.10. Allegedly, this was caused by a single multi-million dollar sell order which triggered numerous other stop-loss orders.

This flash crash is peculiar as it was a nearly currency-wiping event that ended with Eth rebounding to nearly its previous value by the day’s end. When you think about it, this is really remarkable—there aren’t many assets that can boast they’ve made a $600 round trip within hours, especially coming so close to the brink.

2021 Gold Flash Crash 🗓

On a Monday in August 2021 gold experienced a flash crash during after-hours for Western traders. It was widely considered that it was a response by Asian traders to Friday’s news as it was their first opportunity to trade upon it. 

Ultimately, and as you’ll notice is a bit of a trend, this flash crash didn’t have much of an impact overall. However, the price of gold did keep staggering for the rest of 2021 hovering around $1,800 and it only shot up to over $1,900 in early 2022.

How Does a Flash Crash Affect Investors 🧐

A flash crash can really have any number of effects on an investor. However, the chances are that most can go through any number of these events unscathed. Generally, since flash crashes tend to happen within a few hours, the key thing is not to panic.

If you keep your head cool and your portfolio properly diversified, it should be pretty easy to weather any sudden price fluctuations—it would take a very unlikely series of events where most of your assets not only drop during such an event but also never really recover for your portfolio to be truly harmed.

This, obviously, means that it is generally wiser to stick to a long-term strategy. If you are just learning how to day trade, there is a higher likelihood for a flash crash to affect you as you are more sensitive to short-term market fluctuations.

However, this shouldn’t be taken as a call to inaction. Sometimes real crashes can start suddenly and it is important you don’t get caught sleeping during such an event. This means that getting acquainted with various methods of stock analysis is key.

You can often tell apart flash crashes from the beginnings of proper crises by how well the price changes reflect the overall state of the market and your assets. Fortunately, the existence of precise stock analysis software can make this process far easier, and help you keep pace with institutional investors.

For the more daring investors, flash crashes can create opportunities. Essentially, since they cause prices to tank way below realistic market value, they offer an opportunity to buy cheap and favorably enter a long position—or make decent profits quickly by selling as soon as the crash is over if you don’t have confidence in the long term performance of the targeted asset.

You should remember that such a strategy is inherently riskier. Dramatic market events can be unpredictable and you’d have to remain very vigilant throughout the entire process. There are many ways a trade in a very volatile period can go wrong.

Fortunately, the existence of reliable stock trading apps can help you keep your finger on the asset’s pulse, and react to changes without much, if any, delay.

Conclusion 🏁

So far, flash crashes tended to be stressful, yet largely inconsequential events. This means that they are probably best relegated to the back of your mind as it is important to be aware of them—so as not to panic—yet they are usually best ignored.

This inconsequential nature might be in for a change. If the crash of 2010 proved that a single trader moving a large number of contracts can disrupt the market. The Gamestop craze, on the other hand, demonstrated the power of online agoras to create even larger shockwaves.  It isn’t difficult to imagine that strange and sudden disturbances might grow far bigger and more dangerous in the future.

Get Started with a Stock Broker