Can Twitter Sentiment Analysis Predict the Stock Market

writing, blogging, traffic, SEO

Ugly over at Uglychart.com just posted a link about research that sentiment mined over 10 million tweets from 2008 and was able to predict daily market behavior to an accuracy of 87.6%. While the post is vastly interesting from a text & sentiment mining perspective using social media, and the application of it to the stock market, I’m not 100% convinced its very viable.

Why? Well I tend to echo some of the comments left by readers at the bottom of the original post. For example, once this edge” is discovered by general market participants, it tends to get discounted and the edge goes away. So what we read here today is probably already discounted by the market and is just routine business as usual.”

Now, I certainly don’t mean we should abandon text & sentiment mining for the markets, but rather we should continue to use these tools to develop our own secret edges and evolve them as the market changes. Follow the advice of poker players and underarm deodorant manufacturers, never show your hand and never let them see you sweat.

3 thoughts on “Can Twitter Sentiment Analysis Predict the Stock Market”

  1. This is exactly what I gave up my fulltime job to develop a couple of years ago. When I got my head into it I thought the latency involved was to great. Its good to see some results on the subject. I stick to momentum these days – its the best chance for prediction looking at the same signals as everyone else.

  2. @Dave: Yes, latency is always an issue. The trick to following the heard is to get out first before they run off a cliff.

  3. Recorded Future is having a Webinar on Dec 17th. http://blog.recordedfuture….
    On Tuesday, December 7, we’re hosting a webcast to discuss quant finance applications for the Recorded Future news analytics API. Join us to see how our API data is used to support quantitative investment and trading strategies.

    We’ll illustrate an event study framework we are using with Recorded Future data as well as show the retrieval of our data live in an R analytic environment. Additionally, we’ll discuss how integrating data from our API into automated strategies can be used for activities ranging from construction of alpha-generating signals to regime change detectors.

Comments are closed.