Gold analysis part 3, can Blogs or Twitter predict the stock market?
Journalists and scientists are extremely excited about methods for predicting the future based on the collective mood from Twitter, Blogs, or message boards. Should they be? Rebecca Greenfield in the Atlantic Wire recently reported that the Twitter hedge fund beat the market and other hedge funds. One such method was self-published on arxiv.org and is supposedly a basis for the hedge fund. Another method by Eric Gilbert and Karrie Karahalios was peer reviewed for ICSWM 2010 and widely praised in the academic community.
I like Eric and the other authors, and their work is interesting. But the model we discussed in Part 2 on gold pricing (developed by Christophe Faugère at the University of Albany) is about 20 to 100 times more effective at predicting changes in gold pricing than the the best method we can build from either of the techniques discussed in the HCI/Crowdsourcing academic literature. And this is the naive implementation of the Faugère model!
Although the naive methods for predicting stock market price changes (in this discussion, gold is a proxy for market reactions and moods, as was outlined in Part 1 of our discussion), it may be possible to use these different methods as signals that pricing shifts will occur. It’s just not obvious how “improvements” from the HCI / Crowdsourcing literature are immediately adding to this body of practice.
While we’re waiting for social scientists to comment on the model (generated from finance and economics), I’ll leave you with this question: how should journalists or the individual reading about these new inventions interpret these advances?
