At 1:07 pm on Tuesday, April 23rd of last year, a single tweet from the Associated Press’s Twitter feed stopped the stock market in its tracks.
Breaking: Two Explosions in the White House and Barack Obama is injured.
In the next three minutes, the price of crude oil swooned, the yield on the 10-year Treasury note dropped as investors shifted out of risk, the Dow Jones Industrials dropped by 150 points and the S&P 500 lost more than $136 billion in capital value.
Within three minutes, an AP employee tweeted “We’ve been hacked.” That was followed quickly by official confirmation that the “news” was a hoax, the work of a hacker collective loyal to Syrian President Bashar al-Assad.
Within a few minutes the market had almost completely corrected, but the breathtaking speed of that sharp market U-turn brought home to portfolio managers and individual investors some important new facts of life on the trading floor: that Twitter, Facebook and other social media have become critical sources for market-moving information and they can move world markets in the blink of an eye.
“The AP fake tweet was the most active two minutes in stock market history,” says Eric Scott Hunsader, the founder of Nanex, which serves market data to trading houses. “It was one of those things where everything aligns. It was in the afternoon, nothing else was really going on, and also the news seemed credible. It was a perfect storm, and then suddenly it was gone.”
It was all over but the head-scratching. How could the market have reacted so swiftly, both to that single tweet and to the quickly following news that it was fake? Clearly social media had become a greater force in stock market movements than many previously understood, but how—and why?
One answer: large-volume trading driven by computer algorithms, otherwise known as high frequency trading (HFT). The product of Wall Street’s geekiest quants, HFT is a confection of advanced math, predictive analytics and machine learning. Its algorithms are built to evolve internally, “learning” from the results of their trades, which means they end up behaving in ways that even their creators can neither predict nor completely understand. But the AP tweet did make one thing clear: When Twitter talks, HFT listens.
According to Irene Aldridge, managing partner at ABLE Alpha Trading and author of a book on high frequency trading, HFT algorithms sift through lots of news sources—SEC filings, financial news sites, trade publications and, not incidentally, social media feeds—looking for and assigning values to specific words as well as words in close proximity. In the case of the AP tweet, says Aldridge, it is likely that the algorithms picked up on the word “explosions” in combination with “White House”, “Barack Obama” and “injured”, and deduced a major sell signal—or perhaps simply a signal to stop trading. Given the consistently high volume of HFT trades, that alone can cause prices to drop. The HFT algorithms also take cues from each other, which can redouble their market effect.
HFT algorithms and social media have been reinforcing each other for the last decade, with social media providing grist for the algorithms, which validate the significance of tweets and posts. In the process, Twitter and other social media began attracting a growing readership among human traders and investors.
That attention focused sharply after 2013, when the SEC began allowing companies to release earnings and other market-relevant news via social media, giving them the status of primary sources for critical investment intelligence—and making them irresistible as a platform for portfolio managers and analysts to post and tweet their own market insights. In April of that year, with a single post (“We currently have a large position in APPLE. We believe the company to be extremely undervalued”), Carl Icahn added $17 billion to the value of the company in the space of one hour.
If the result of the social-media congregation has been a sometimes bewildering array of data, opinion and advice on Twitter, Facebook, LinkedIn et al, it has also made them essential checkpoints for every active trader and investor.
The big investment banks and hedge funds have companies like Dataminr to sift for trading signals in the social media noise. Small players and individuals have a harder time of it. Even the biggest financial news organizations find themselves lagging behind the social-media speed curve. A crack journalist with the best set of Wall Street friends on Facebook and sources on TweetDeck is no match for an algorithm that sees it all, and all at once.
That said, every institutional and individual investor now stands to benefit by staying current with real-time social media data on the individual stocks they own or contemplate buying, which is easy as a hashtag search.
And every responsible investor can take another lesson from the fake AP tweet about that explosion at the White House: It pays to stay skeptical, about tweets and posts as well as the Wall Street Journal. As the author of Social Media Strategies for Investing, financial analyst Brian D. Degger, recently told Forbes: “You never want to act on information you receive from one source of insight…. Everything you receive has to be taken with a grain of salt.”
The third party content is being provided to you for educational purposes only. The content has been written by Atlantic Re:think, The Atlantic's creative marketing group, a third party not affiliated with E*TRADE Financial Corporation or any of its affiliates. E*TRADE Financial Corporation compensates The Atlantic for advertising and media related services. No information presented constitutes a recommendation by E*TRADE Financial or its affiliates to buy, sell or hold any security, financial product or instrument discussed therein or to engage in any specific investment strategy. The content neither is, nor should be construed as, an offer, or a solicitation of an offer, to buy, sell, or hold any securities by E*TRADE Financial Corporation or its affiliates.