Questioning the efficient markets hypothesis: Big data evidence of non-random stock prices

The efficient markets hypothesis claims that stock prices fully reflect all available information, and that prediction of future changes in stock prices is impossible. For 50 companies listed on stock exchanges in the United States, this paper compares the real data with random data that follows a similar distribution as the real data, in order to ascertain how much usefully predictive information is in the real data. Surprisingly, it turns out that if one can tolerate a modest number of random false positives, around twelve percent of the time there is a modest amount of information.