In a world where everyone’s opinion is publicly recorded, it should be theoretically possible to determine the most popular concept of the moment — be it a movie, a song, a feeling or even a stock. Certainly, we don’t have enough information at this point to record everyone’s fleeting thoughts, but Twitter is something that makes a good temporary approximation. And that’s what led University of California, Riverside professor Vagelis Hristidis to investigate whether Twitter sentiment could help predict the stock market.
Hristidis and his team have set out to look at how Twitter activity is related to stock prices and traded volume. Specifically they look at the volume of tweets and the link between between those tweets, other users’ tweets and specific stocks.
This is different than an experiment from March of 2011, where a group of researchers from Indiana University used the mood of the overall Twitter community to attempt to predict upcoming moves in the Dow Jones Industrial Average. They turned out an 87% success rate within four days of the sentiment analysis. In that analysis, we cited Brunswick Group’s study that only 4% were using social media to improve their financial decisions.
What Hristidis did was to analyze the number of tweets which had “connected components” related to a specific stock, and found a correlation between those tweets and the stock volume.
They expected to find the number of trades was correlated with the number of tweets. Surprisingly, the number of trades is slightly more correlated with the number of what they call “connected components.” That is the number of posts about distinct topics related to one company. For example, using Apple again, there might be separate networks of posts regarding Apple’s new CEO, a new product it released and its latest earnings report.
So the idea is that the number of distinct discussions related to a company is correlated to the amount of trades for a given day. This seems to line up with casual observation, as the more that people are being creative to actually analyze and discuss a topic, the more attention it has. The problem with just looking at pure volume is that it doesn’t really require a lot for people to simply retweet or make a comment like “I really like Apple”, but if they are discussing specific facets of the company, it means this is something important to them. And the more people who think Apple is actually important probably correlates better with the number of people who think it’s important enough to buy or sell the stock.
They also found that the stock price is “slightly correlated” with the number of connected components.
You can read more about their specific control tests and measurements over at UCRToday.