jttpbf

Generating a truly random series of data is a big task. It requires avoiding any possible slant which might be present in the method. A French trading system developer reported how he did it in his book and concluded the markets were random because the series of numbers he derived displayed some occasional trends, just like the ones produced by the markets. I believe he made a logic mistake. Two sytems leading to similar results (markets and randomness generators) are not necessarily identical. This does not minimize the value of the Artificial Inteligence and fuzzy based logic sytem he went on to develop to respond to the challenge of winning with the markets. I believe on the contrary that markets are not random but I have no formal proof of it, only arguments. The most important one is that a "random" generation by definition is one where any one event is totally independent from the previous ones. We all know that on the contrary all traders act wilfully in one of the fiollowing ways: to purchase or to sell quantities which surpass the quantities available at any one or several price levels (which moves prices) or to jump on the bandwagon when a trend develops or to take profits when the prices have reached their personal targets. These three behaivors (there might be more) seem to represent a sufficient number of events to conclude that the basic criterion of randomness is not met.