The Planet Money team at NPR has built its own stock-picking bot to better understand Wall Street’s predominantly automated, computer-driven trading. There is one twist though: this bot’s trades are based purely on the Twitter feed of Donald Trump.
“Our bot is doing something seemingly simple: It looks at President Trump's Twitter feed, and when he tweets about a company, it trades stocks, with real money,” NPR said on their website.
@BOTUS, or Bot of the United States has $1,000 at its disposal to buy stocks based on Trump’s Twitter feed. If there is a positive mention of a publicly-traded company, a buy order is placed; if the tweet contains negative news, BOTUS knows to take a short position.
To afford BOTUS a certain competitive edge, however, the engineering team at NPR did incorporate VADER Sentiment Analysis to avoid any catastrophic trades that would bankrupt the bot.
“The Fine Print: This is for real. All decisions on buying and selling stocks are made automatically by BOTUS, a computer program,” NPR wrote in a disclaimer on their website.
One of the developers of the bot announced on Twitter that the bot will begin trading at market open on Monday.
As far back as 2010, almost 70 percent of all trades on Wall Street were performed by high frequency trading computer algorithms (algos).
Wall Street has its own algos that keep a close eye on social media and, in particular, Donald Trump’s Twitter feed, but BOTUS is the first algo with its own Twitter account to provide trade-by-trade updates to the wider world.
READ MORE: Trump tells American firms to look for China alternatives ahead of first meeting with Xi