The theory of technical analysis suggests that future FOREX price development can be foretold by analyzing historical price ﬂuctuations and identifying repeti¬tive patterns. A computerized system, able to produce trade recommendations based on different aspects of this theory, has been implemented. The system utilizes trading agents, trained using machine learning techniques, capable of producing uniﬁed buy and sell signals. It has been evaluated using actual trade data from the FXDD FOREX broker over the period 2006 – 2009.
Compared to the simple strategy of buying and holding, some of the agents have proven to yield good results, both during years with extremely good FOREX market returns, as well as during times of recession.