Title :
GP-based optimisation of technical trading indicators and profitability in FX market
Author :
Lee, Cheng Song ; Loh, Khai Yam
Author_Institution :
Sch. of Bus. Syst., Monash Univ., Clayton, Vic., Australia
Abstract :
Some empirical evidence have suggested that it is possible to reap profit with one well chosen trading indicator that possesses embedded market timing adaptability, to trade either stocks or foreign currencies over medium term of 3 to 5 years. The profit is however attained with risk-taking embedded in the determination of a buy decision. To achieve more consistent profitability with a moderate risk, we propose a modified GP-based optimised trading rule, which involves the dynamic use of two out a finite number of pre-specified indicators. In this respect, the proposed rule is neither too risk-averse nor too risk-taking biased in the determination of a buy decision. Based on the minimum cash draw-down criterion and adopting momentum trading strategy (i.e., following the trend) the statistical test results suggest that consistent profit after accounting for transaction cost is achievable through extrapolating the trend.
Keywords :
foreign exchange trading; genetic algorithms; profitability; risk analysis; buy decision; foreign exchange market; genetic programming; optimisation; profitability; risk-averse; risk-taking; technical indicators; Computer networks; Economic forecasting; High performance computing; Humans; Information technology; Macroeconomics; Pattern analysis; Profitability; Robustness; Timing;
Conference_Titel :
Neural Information Processing, 2002. ICONIP '02. Proceedings of the 9th International Conference on
Print_ISBN :
981-04-7524-1
DOI :
10.1109/ICONIP.2002.1202803