Economic & Market Forecasts using Social Media

A recent article in the Bank of England’s Quarterly Bulletin, “Using internet search data as economic indicators“, 2011Q2 argues that  data on the volume of online searches through Google, etc can be used as indicators of economic activity. A recent example was the correlation between consumer confidence and searches on gold prices, as published in the Economist, “Behind the bald figures“, 27 August 2011.

According to the Financial Times, “Twitter research promises trading success“, 8 May 2011, researchers from Indiana University have designed an algorithm that looks at the correlation between mood statements on Twitter and the Dow Jones Industrial Average. What they discovered surprised everyone on the research team. Instead of a mood of negativity arising from a fall in the markets, researchers found the algorithm could predict a rise or fall in the DJIA with 87% accuracy three to four days beforehand.

For tweet trading in action have a look at:

Additional references:
Decoding Our Twitter Chatter, Wall Street Journal, 1 Oct 2011
Predicting the Present, Think Quarterly by Google, Q4 2011

Leave a Reply