The Wall Street Journal ran a piece today discussing Elon Musk's use of Twitter and how it relates to Tesla share price. Max Margenot, Lead Data Scientist at Quantopian, constructed a sentiment model on Musk's tweets and used our new Self-Serve Data feature to upload the sentiment scores to Quantopian. He then constructed a momentum trading strategy in which positive sentiment swings cause buys, and negative sentiment swings cause sells. This strategy interestingly does make a little bit of money, although it's just an early stage piece of research and doesn't trade nearly enough to build great confidence in the effect being real/robust.
We're currently working on releasing the full analysis including the data for Self Serve, so you can experiment, but for now here's the Pyfolio tearsheet of the strategy.