Slippery slope

A recent study (1) that showed that one could create alpha by analyzing twitter feeds is a slippery slope. Academics are fickle animals – on the one hand they theorize about efficient markets and on the other, they try to pick pennies by proving markets are not efficient. They seem similar to those who have their heads in the hedges – those who do not understand efficient markets or portfolio theory. In either case, one thing is clear – showing alpha from short periods by ad-hoc analytics does not prove anything, except bolstering the alpha-maker's ego, that has a distinct half-life.

A more interesting question is why one finds correlations between brain dead twitter feeds and stock markets. Causality appears to be demonstrated by showing an alpha – but is it possible that one can find similar effects by randomly investing. Is the wisdom that is doled out in 140 constraining characters that powerful? What if a similar analysis is made based on the number of flights arriving on time in the US airports or the number of tomatoes in the grocery stores that went bad. Could one show alpha with such attributes? will one try? It seems less interesting, however.

It appears that the aura of stock market investing is pulling everybody in – one would have thought that it is time that the Wall street physicists returned to the careers they were actually trained in.

(1) Using Twitter to predict financial markets. Published: Monday, March 26, 2012 - 23:34 in Mathematics & Economics. Source: University of California - Riverside