Can Twitter sentiment analysis guide stock market investment?

My job – when I’m coaching people in Voice and Presentation and Public Speaking, is to get speakers to make sure that their audience understands everything  they’re saying.  In most cases this means an in depth search inside themselves for the absolute clarity of the meaning of their statements, phrases and words.  “Deconstruction” is what my professor used to call it.  What are you trying to say?  What do you want your audience to understand about what you’re saying?

Every time I’ve coached someone in Banking and Investment I’ve learned and understood more about the stock market.  “What?”  I once asked a director of Steel Business Briefing, “makes the steel market fluctuate so much?”   They replied “sentiment”.  In a recent paper written by Johan Bollen, Huina Mao and Xiao-Jun Zeng  it states that Twitter sentiment analysis can guide stock market investment. In this technology review you can see how an analysis of almost 10 million tweets from 2008 shows  that they can be used to predict stock market movements up to 6 days in advance.

Not so long ago I heard a French person say that instead of voting for a President they should just count up the Twitters.

For individual coaching in Voice and Presentation and Public Speaking email:
For the Leadership – Communication in Management Course, book on line:


One thought on “Can Twitter sentiment analysis guide stock market investment?”

  1. My company: InfoTrie has developped FinSentS a Sentiment Analysis and News Analytics system. We process News, blogs and social media (mainly Twitter) for thousands of Stocks, FX and commodities. Our sentiment is then correlated with asset prices, and the results are quite good. It can be used as a technical indicator, or as a quantitative signal (market data) for algorithmic trading or downstream systems (risk, compliance, …)

    We provide Sentiment API. Feel free to have a look on!


Leave a Reply

Your email address will not be published. Required fields are marked *