Is Twitter Making the Stock Market Predictable?

Past research has shown that stock market prices can't be predicted with more than 50% accuracy, but new research shows that analyzing Twitter feeds could help predict the direction of the Dow.

This article published with permission from InvestmentU.com.

Imagine just how rich you could be if you had the uncanny ability to predict the daily movements of the stock market.

Instantly, you’d have a major advantage over other investors, knowing which way the majority of stocks would go each day.

Of course, this could never actually happen, right?

Past research has shown

such as Random Walk Theory and Efficient Market Hypothesis

that stock market prices can’t be predicted with more than 50% accuracy.

Wait, not so fast.

An academic study from Johan Bollen and Huina Mao at Indiana University revealed the stock market may be more predictable than originally thought.

In the middle of the Great Recession, they were able to predict the direction of the Dow Jones Industrial, two to six days in advance, with an amazing 87.6% accuracy.

Source: arXiv.org

By analyzing “useless” Twitter feeds.How’d they do it?

Hello, sentiment

Every minute, an estimated 236,000 tweets are fired off in the “Twitterverse.”

That’s over 339 million messages a day. Plenty of people have attempted to make profitable sense of this mountain of data in the past.

But Johan and Huina were the first ones who discovered one algorithm, called the Google-Profile of Mood States (GPOMS), can actually be used to financially exploit tweets.

GPOMS analyzes the level of six emotions on Twitter: calm, alert, sure, vital, kind and happy. Of these six states, there was a direct connection between calmness and the direction of the Dow.

The calmer people seemed on Twitter, the more likely it was that the Dow would rally. The more anxious, the more likely the Dow would drop.

From March to December 2008, they processed 9.7 million tweets to come up with their incredible findings.

The authors have since teamed up with DCM Capital to use Twitter sentiment as a predictive tool for specific stocks such as Apple (Nasdaq: AAPL), Facebook (Nasdaq: FB) and BAE Systems (LSX: BA).

Today, the company is considered the world’s first trading platform with built-in social media sentiment analysis.

In fact, it even released an app earlier this year that can be used to track sentiment for various stocks and indices.

A new tool for the tool belt

Whether or not calmness is the key ingredient to predicting stock market movements remains to be seen, in my opinion.

But stocks often do become oversold in the marketplace because investors let emotions

like anxiety and fear

get the best of them.

Having real-time access to at least a hint of that broad sentiment could be a useful tool for traders and investors alike… “Tool” being the keyword here.

Because as interesting and potentially lucrative as sentiment analysis may seem, finding solid companies that are selling at good prices, while consistently increasing their earnings and shareholder value, should always be at the core of your stock purchases.

But this sure is interesting stuff…

Mike Kapsch is part of the research team at InvestmentU.com. See more articles by Mike here.