Skip to main content

A Twitter page is displayed on an Apple iPhone.


If you need one more reason to follow Paris Hilton or William Shatner on Twitter, here it is: Their tweets might make you a better investor.

Sure, each tweet limits its author - big star or regular Joe - to 140 characters. That isn't a lot of room to opine on stock valuations, earnings guesses or hot rumours. However, collectively, tweets might be telling us something about the mood of investors - even if they have nothing to say about investing in particular.

Johan Bollen and Huina Mao, professors at the School of Informatics and Computing at Indiana University, and Xiao-Jun Zeng, a professor at the School of Computer Science at the University of Manchester, looked into the matter and found that the collective mood among tweeters has a remarkable accuracy rate of 87.6 per cent in predicting the daily up and down changes in the Dow Jones industrial average.

Story continues below advertisement

Given that Google searches can reflect consumer spending and disease infection rates, and Twitter-feeds on movies can predict box office receipts, making the leap to stock market prediction might not be such a big one.

"Although news most certainly influences stock market prices, public mood states or sentiment may play an equally important role," the authors said in their paper. "We know from psychological research that emotions, in addition to information, play a significant role in human decision-making.

For their study, the authors examined more than 9.8 million tweets generated by 2.7 million users between February 28, 2008, and December 19, 2008, grabbing key words and statements that clearly reflected the authors' moods. Calm, upbeat moods were strongly predictive of higher stock market closes, and vice versa.

The end of 2008 was a tumultuous time for stocks, with the Dow falling more than 30 per cent over this period as the financial crisis shook investor confidence in the global economy in general, and Wall Street in particular. Presumably, tweeters were like the rest of the investing world and feeling pretty glum.

However, the Dow didn't descend in a straight line, but rather zigzagged throughout most of the 10-month period. As well, the authors point out that the sharp declines in stock prices in the fall of 2008 - after Lehman Brothers went bust - actually make stock market prediction more difficult than in other periods.

The authors don't uphold their study as a get-rich scheme, but rather as a potential tool to measure sentiment. There are, of course, other tools widely available to investors, including various consumer sentiment indexes and bullish or bearish readings on newsletter writers.

Some of these surveys, though, are either expensive or time-consuming.

Story continues below advertisement

"Public mood analysis from Twitter-feeds on the other hand offers an automatic, fast, free and large-scale addition to this toolkit that may in addition be optimized to measure a variety of dimensions of the public mood state," the authors said.

They acknowledge that there is more work to be done. Twitter-feeds are largely written in English and the writers - at least in their study - are mostly based in the United States. Investors, though, form a global community.

And while the research suggests that there is a correlation between Twitter-feeds and the Dow's direction, it didn't examine the reasons behind this connection.

"One could speculate that the general public is presently as strongly invested in the DJIA as financial experts, and that therefore their mood states will directly affect their investment decisions and thus stock market values," they said. "But this too remains an area of future research."

Report an error Editorial code of conduct Licensing Options
As of December 20, 2017, we have temporarily removed commenting from our articles. We hope to have this resolved by the end of January 2018. Thank you for your patience. If you are looking to give feedback on our new site, please send it along to If you want to write a letter to the editor, please forward to