DocumentCode :
2224774
Title :
Better trade exits for foreign exchange currency trading using FXGP
Author :
Loginov, Alexander ; Wilson, Garnett ; Heywood, Malcolm
Author_Institution :
Dalhousie University, Faculty of Computer Science, 6050 University Avenue Halifax, NS, Canada
fYear :
2015
fDate :
25-28 May 2015
Firstpage :
2510
Lastpage :
2517
Abstract :
Retracement is the tendency of markets to move between upper ‘resistance’ and lower ‘support’ price levels. Human traders frequently make use of visual tools to help identify these resistance and support levels so that they can by used in their trading decisions. These decision can be put into trading strategies composed of rules designed to mitigate losses after a trade is started, often called ‘stop loss’ orders, or to take profit at a near optimal time, often called ‘take profit’ orders. However, identifying such resistance and support levels is notoriously difficult given market volatility. Indeed, the levels need recalculating on a continuous basis, and only hold to an approximate degree. In this work we describe an approach for evolving buy-stay-sell currency trading rules using genetic programming. These rules are explicitly linked to technical indicators that incorporate features characterizing retracement. Benchmarking is then performed using the most recent three years of data from the EURUSD foreign exchange market with three different methods of identifying retracement based on moving average, pivot points and Fibonacci ratios. Investment strategies employing Fibonacci ratios and found to provide superior performance among the strategies examined.
Keywords :
Decision trees; Immune system; Market research; Resistance; Sociology; Statistics; Training;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Evolutionary Computation (CEC), 2015 IEEE Congress on
Conference_Location :
Sendai, Japan
Type :
conf
DOI :
10.1109/CEC.2015.7257197
Filename :
7257197
Link To Document :
بازگشت