At 1:07pm EST on 23 April 2013, the official Associated Press twitter feed offered an unsettling announcement: “Breaking: Two Explosions in the White House and Barack Obama is injured.”
Clik here to view.
The Associated Press Twitter account was hacked and the above incorrect tweet affected financial markets.
Clik here to view.
The near instantaneous “twitter effect” on the Dow Jones Industrial Average. Image sourced from Google Finance.
No more than 3 minutes later, the tweet had been discredited by AP, whose Twitter account had been hacked and quickly suspended. But it was enough of a scare to send the Dow Jones Industrial Average falling over 140 points in a matter of seconds, while the broader S&P 500 index declined nearly a full percentage point.
Markets recovered from those losses within minutes, in fact ending the day up over 1%, but the event was another revealing episode in the ongoing drama of a market on technological steroids. Financial markets have always suffered from the natural instincts of investors to follow the herd, with irrational market selloffs as old as bank runs. But perhaps more worryingly, the extent and impact of algorithmic trading programs were once again revealed yesterday, leading to continued concerns over their viability.
Regulators and market participants are aware that a sizeable portion of global trading volume is now executed by “algos”, the computerised trading platforms which can execute trades (oftentimes thousands per second) without any human input or authorisation. Many such algos now incorporate “machine readable news” capabilities into their trading strategies, which allows computers to “read” specially coded news articles and execute trades based on specified criteria.
But yesterday’s crash was the result of a fake tweet, and the market reaction was nearly instantaneous; it begs the question of whether we’ve entered the age of “machine readable tweets”. Indeed, as early as 2011, US social media data aggregator Gnip began offering exactly such a service. This quiet and fascinating development has clearly gained enough traction with algorithmic traders proving to be able to move entire markets.
Australia regulators and market players, meanwhile, are grappling with their own issues around algorithmic trading. In February 2012, the ASX opened its Gore Hill Liquidity Centre, essentially a specialised venue for sophisticated market participants to execute algorithmic trading. Proponents claim that such platforms enhance liquidity by increasing trading volumes, but they also expose publicly listed companies to potentially wild fluctuations in their share price.
As communications professionals, these latest developments confront us with two serious concerns:
The first is what to do about Twitter. On the one hand, the micro blogging service was yesterday proven to be unreliable and unsecure, but also irrefutably relevant. We still recommend that companies have both a presence and an awareness of social media, albeit with a strong password.
Second, and more importantly, how can investor relations professionals, senior management, boards, employees, retail shareholders and other stakeholders understand and interpret share price movements in the age of machine readable tweets? Dramatic intraday blips in share prices may become more and more common as algos turn to social media for trading execution, although investors may take some comfort in knowing that shares typically pare back any losses before the market closes.
Perhaps, in the foreseeable future, Twitter feeds will become as ubiquitous as the Bloomberg terminal on the modern trading floor, and market participants must adapt to this new reality.
Image may be NSFW.
Clik here to view.
