1. Intraday momentum and breakouts on lower time frames have returned. They were more common before 10 years ago, but then died out for years. Breakout patterns were common and short term trend pullbacks were touted by the TA crowd at the turn of the century. Now, if I was a bot programmer looking for the current action, I would look again at these old patterns. The converse way of looking at it at a higher time would be a wide range.

The counter argument is that random price with drift, even under normal sample will produce many apparent trends, but are non predictable. However, a normal random sample with drift will produce more trend like structures. Here the simple rule moves to formation of complex structures. Wolfram in A New Kind of Science, posited that simple iterative rules will produce complex structures who underlying rule cannot be determined. Further, that such simple rules will result in symmetrical structures such as leaves, hands, bodies, symmetrical along one axis. Sample a distribution with fatter tails, add drift, and more structures should appear. The search for significant regularities using normal assumptions might be augmented with a search for basic simple rules whose application will reveal repetitive structures as a basic function. A random sample with drift is an example of a simple iterative rule that creates complex patterns with symmetry.

A long enough random series will produce a million identical digits in a row. An infinite random series will produce our entire history of natural existence and physical law. If existence is in fact random, the search for patterns would be nothing more than astrology.





Speak your mind


Resources & Links