Investors have been playing with the idea of Twitter as a stock market analysis and speculation tool for some time now.
After all, one of the first big hits in third-party Twitter apps was StockTwits, which features graphs and the ability to filter messages by company or stock exchange symbol. Startups like Chart.ly and Covestor operate in the same space, roughly.
But this is the first time we're seeing a literal and real application of cash to Twitter sentiment.
Derwent Capital Markets explained its scheme to The Next Web. Basically, as Twitter users reveal themselves to be more worried, upset, happy, content, etc., the market fluxuates with rough correlation, rising on calm days and falling on anxious ones. Currently, Twitter is about 87.6% accurate for predicting stock market fluctuations.