Genetically Programmed Trading Rules for the Foreign Exchange Market

Abstract

The identification of price patterns and trends, and the formation of rules to generate market signals have a long history in foreign exchange rate markets. Recent studies, however, question the profitability of the simple rules that have been shown to yield abnormal profits in previous decades. Rather than assuming a fixed set of rules, in this paper we employ genetic programming to identify rules in the Euro US Dollar daily exchange rate series over the period 1/1/1999 to 30/12/2005. Preliminary experimental results suggest that genetic programming is capable of generating profitable rules but for a limited time into the future.