Twitter Predicts DOW
arxiv.org
arxiv.org
I think there's a mistake with the early discussion of the efficient markets hypothesis though.
"There are two problems with EMH. First, numerous studies show that stock market prices do not follow a random walk and can indeed to some degree be predicted [5], [6], [7], [8]."
Any bets 5, 6, 7 & 8 have now been factored into financial trading models cancelling out the predictive advantage?
I think it only makes sense to think of EMH as a dynamic rule. With new information (/new ways of understanding old information) disrupting the equilibrium price which then converges to a long term equilibrium; at least until the next shock.
I'd apply this on every level from the technology arms race to new sources of info & analytical models (such as twitter + NLP).