An Introduction to Maximum Entropy Method
(MEM) Technical Analysis
by Anthony W. Warren, PhD
Preface to Maximum Entropy Articles
A reasonable assumption concerning the motivation of subscribers to Technical Analysis of Stocks and
Commodities magazine might be that they are looking for practical methods to improve the 'bottom line'
or profit of their trading activities. It should be stated and emphasized that this article is one of several
planned articles leading to a conclusion which will indeed provide a practical tool intended to augment
that 'bottom line '. The reader will therefore be reassured that before the end of this series of articles, the
author will conclude with practical applications similar to those provided in previous issues for other
tools such as Fourier Analysis, ARIMA, Finite Impulse Response filters, TRIX, and Market Direction
Dr. Warren's article is a description of his research concerning the adaptation of an accepted
mathematical tool, Maximum Entropy, to the analysis of the data spectrum we traders call stock and
commodity prices. Furthermore, the article promises to provide a "method of adaptive filtering and data
forecasting': which is what a practical trader will examine for profit. Most 'technical analysts' feel a
need to understand the mathematical theory behind the use of the various tools at their disposal and this
article is the necessary first step to that understanding There is nothing in this first article which can be
plugged into your computer or your immediate decision making apparatus; however, be assured that
such is forthcoming.
A caution may be appropriate for some fraction of the readership. That caution is that no technique,
study or tool of technical analysis is perfect or promises success, by itself or in any combination with
other tools. An experienced, practical trader in this zero sum game seeks no more than an edge, say a
10% edge, in making his trading decisions. That edge combined with disciplined money and risk management will place this trader among the survivors ... no more! Often less!