A tweet and a bunch of confused computers was all it took to take down Wall Street for about 10 minutes on a weekday.
Last Tuesday at 1:08 p.m., a tweet had been sent via The Associated Press’ Twitter account that stated the following: “Breaking: Two Explosions in the White House and Barack Obama is injured.”
The next two minutes turned Wall Street into a financial nightmare as investors rushed to sell off their stock at the mere mention of such panic-inducing bad news. The Dow fell 143 points.
Then the AP’s Media Relations team (@AP_CorpComm) tweeted the following: “That is a bogus @AP tweet.” The Associated Press’ Twitter account had been hacked and the White House Bombing tweet had been false.
By 1:18 p.m., soon after the AP said the tweet was false, the plunge in the stock market that began 10 minutes earlier had rebounded and the Dow had regained its losses.
A group known as The Syrian Electronic Army has stepped forward to claim responsibility for Tuesday’s hack-tweet, though that claim has not been proven.
The next day, The Associated Press came out with its official timeline of Tuesday’sTwitter incident and an investigation into Wall Street’s resultant plunge. They came to an interesting conclusion: Tuesday’s plunge wasn’t really due to a bunch of anxiety-ridden, eager-to-sell human investors. It was largely the fault of supercomputers pre-programmed to buy and sell stock automatically based on, among other things, buzzwords they find in news items.
This use of such computers and special algorithms to trade stocks at an exceptionally accelerated pace is, according to CNBC, the core of what is known as high-frequency trading (HFT). This automated approach to investing allows algorithm-directed computers to make their own decisions on when and which stocks to buy or sell within fractions of a second. As a result, stocks are traded thousands of times a day and as, Business Insider points out, with the widespread use of HFT, the “holding period” (the amount of time an investor keeps their stock in a company), has “fallen from eight years in the 1960s to around five days today.”
So as a financial trend, how widespread is HFT? The likes of CNBC and The New York Times have reported that, as of 2012, approximately 50 percent of all equity (stock) trades are made by HFT computers. And to give you a visual of how quickly the use of HFT on Wall Street and elsewhere has grown, here’s an animated graph GIF from Nanex that depicts “the rise of HFT [algorithm] machines from 2007 to 2012.” (Nanex is an American data feed company. It is also the creator of NxCore, software that delivers and databases stock quotes to computers.)
The whole point of high frequency trading is speed and so even the technology side of HFT reflects that. Everything from the algorithms used to the exact placement of the supercomputers that control HFT is constantly tweaked in an effort to complete trades faster. In HFT, it’s not the amount of money that’s moved within a micro-second trade, it’s the amount of trades you can make within a second that’s important.
As Mother Jones reported in their January/February 2013 issue, trades are made “in less than a half a millionth of a second—more than a million times faster than the human mind can make a decision” and in those tiny, split-seconds, HFT algorithms are designed to analyze long-term trends and tiny momentary blips in trading to make quick trading decisions over what initially seems like worthless fractions of a dollar. But when you’re completing trades at an average of 10,000 a second and each trade is worth a profit of 0.0001 cents, the money adds up rapidly to a total profit of $3,600 in just one hour.
(Want to hear just how fast HFT trades are made? Check out this song that was made by attaching a musical note to every trade made for a single stock. The song was composed by Eric Hunsader of Nanex.)
But there’s a darker, more competitive side to HFT trading, as Mother Jones points out: the supercomputers are also programmed to sabotage each other. Trade orders are often sent and cancelled just to throw off other automated investors. Not every order sent is a legitimate trade. Some are just made to flood the market to trick other computers into thinking that there is a meaningful trading trend here and that a decision to buy or sell should be made based on it.
And so if it is that easy for these HFT supercomputers to fool each other, it should come as no surprise that these computers could be fooled by a human with a fake tweet. They’re programmed to make quick decisions about numbers and to react to positive or negative-sounding words, not analyze the exact significance (or verify the credibility) of a news tweet.
“The events last Tuesday were likely caused by the news-reacting algorithms that are designed to electronically read and interpret machine-readable news,” said in an emailed response by Irene Aldridge, a hedge fund consultant on algorithms and author of High-Frequency Trading: A Practical Guide to Algorithmic Strategies and Trading Systems. “Most of the web content is machine-readable, so lots of algos are built on reading and reacting to news such as the Twitter hack… Clearly in the Twitter case, trading machines did not ascertain credibility of the tweet,”
So how does the future of HFT look in light of Tuesday’s AP hack tweet incident?
Aldridge said in terms of the technology, “going forward, many algorithm designers will take into account the Tuesday situation, and will build ever more sophisticated approaches.”
However, as Mother Jones’ (and even Nasdaq) points out, when it comes to the technology of HFT computers, speed is still considered the highest priority as even the locations of these computers have recently become the focus of the advancement of HFT technology; as computers “are strategically placed to shave microseconds from the time the order is sent to the time it reaches the exchange” and as $300 million-worth of underwater cables are being built in the Atlantic Ocean to facilitate faster trades between Wall Street and the London Stock Exchange.
And while the future of high frequency trading (and that of your portfolios) remains uncertain, it seems like a pretty safe bet that with a continued emphasis on speed and without improvements in the accuracy of HFT trades that human investors could witness another plunge in the market not unlike the one that struck fear in the hearts of many last Tuesday.