How Random Is Random? by Clifford J. Sherry, Ph.D.
Some fundamentalists and most technicians would probably agree that tick-to-tick or maybe even day-to-day price changes aren't completely random. Most fundamentalists would argue that price changes over longer time periods are affected by different factors, the impact of which makes price changes random. Technicians, on the other hand, argue that past prices affect current prices. It's difficult for them to believe price changes are completely random. I originally proposed a test for divergence from randomness about seven years ago, which was published in the Journal of Electrophysiological Technique . I used this test to demonstrate that the changes in Standard & Poor's 500 stock index are apparently random. I am not aware of any other simple tests that detect divergence from randomness, nor of any adverse commentary on this test since it was published.
Now, a new question has occurred to me: Just how random is random? Might prices appear random when examined individually but, if aggregated, may not be?