Study: Twitter Predicts Stock Prices More Accurately Than Any Investment Tactic
Want to predict what will happen in the stock market tomorrow? Just use Twitter, say Californian researchers. A team at the University of California has created a computer model which allows them to ‘predict the future’ of the stock market by scanning the social network.
It has up to 11 per cent more accuracy than other computer models.
The software looks at volumes of Tweets containing references to companies or products, and how they link to Tweets on other subjects.
It can predict both the volume of trading and the value of a stock the next day.
The discovery could have a huge impact on investors.
The team from the University of California have built a model which they claim could have a huge impact on investors by using the data to help predict the traded volume and value of a stock the following day.
During a four-month test the model did better than the Dow Jones Industrial Average and outperformed other baseline strategies by between 1.4 percent and almost 11 percent.
Professor Vagelis Hristidis, who specialises in data mining research which focuses on discovering patterns in large data sets, worked on the model. He said: ‘These findings have the potential to have a big impact on market investors.
‘With so much data available from social media, many investors are looking to sort it out and profit from it.’
The team studied how activity in Twitter is correlated to stock prices and traded volume.
While past research has looked the sentiment, positive or negative, of tweets to predict stock price, little research has focused on the volume of tweets and the ways that tweets are linked to other tweets, topics or users.
Past work has also mostly studied the overall stock market indexes, and not individual stocks.
For the latest study the researchers obtained the daily closing price and the number of trades for 150 randomly selected companies in the S&P 500 Index for the first half of 2010.
Then, they developed filters to select only relevant tweets for those companies during that time period. For example, if they were looking at Apple, they excluded tweets that focused on the fruit.
They expected to find the number of trades was correlated with the number of tweets.
But the number of trades is slightly more correlated with the number of ‘connected components’ – the number of posts about distinct topics related to one company.
For example, using Apple, there might be separate networks of posts regarding Apple’s new CEO, a new product it released and its latest earnings report.
They also found stock price is slightly correlated with the number of connected components.
For the study, the researchers simulated a series of investments between March 1, 2010 and June 30, 2010 and analyzed performance using several investment strategies. During that time frame, the Dow Jones Industrial Average fell 4.2 percent.
The model the researchers developed using Twitter data lost on average 2.4 percent.
But Professor Hristidis noted several potential weaknesses in the study including that the trading strategy worked in a period when the Dow Jones dropped, but it may not produce the same results when the Dow Jones is rising.
There is also sensitivity related to the duration of the trading. For example, it took 30 days in the simulation to start outperforming the Dow Jones.