Robust technical trading strategies using GP for algorithmic portfolio selection
作者：
Highlights：
• GP is applied to learn trading rules that are used to automatically manage a portfolio of stocks.
• A new Random Sampling method is used to increase the robustness of the strategies evolved.
• The new Random Sampling method produces strategies able to withstand extreme market environments.
• The new Random Sampling method produces solutions that perform during out-of-sample testing similarly as during training.
• The results are based on testing a portfolio of 21 Spanish equities.
摘要
•GP is applied to learn trading rules that are used to automatically manage a portfolio of stocks.•A new Random Sampling method is used to increase the robustness of the strategies evolved.•The new Random Sampling method produces strategies able to withstand extreme market environments.•The new Random Sampling method produces solutions that perform during out-of-sample testing similarly as during training.•The results are based on testing a portfolio of 21 Spanish equities.
论文关键词：Genetic programming,Algorithmic trading,Portfolio management,Trading rule,Finance
论文评审过程：Received 6 March 2015, Revised 26 October 2015, Accepted 28 October 2015, Available online 2 November 2015, Version of Record 18 November 2015.
论文官网地址：https://doi.org/10.1016/j.eswa.2015.10.040
