Evolutionary decision support system for stock market trading

Abstract

This paper proposes a decision support system for stock market trading, which is based on an evolution strategy algorithm applied to construct an efficient stock market trading expert built as a weighted average of a number of specific stock market trading rules analysing financial time series of recent price quotations. Although applying separately, such trading rules, which come from practictioner knowledge of financial analysts and market investors, give average results, combining them into one trading expert leads to a significant improvement and efficient investment strategies. Experiments on real data from the Paris Stock Exchange confirm the financial relevance of investment strategies based on such trading experts. © Springer-Verlag Berlin Heidelberg 2008.