Chaos based portfolio selection: A nonlinear dynamics approach

作者：

Highlights：

• We propose a portfolio selection method based on chaos and dynamical systems theory.

• We build Constant Chaoticity Portfolios (CCPs) via nonlinear time series forecasting.

• The CCP overwhelms several competing methods in terms of risk-return profiles.

• A sensitivity analysis on model configurations confirms the CCP’s overperformance.

摘要

•We propose a portfolio selection method based on chaos and dynamical systems theory.•We build Constant Chaoticity Portfolios (CCPs) via nonlinear time series forecasting.•The CCP overwhelms several competing methods in terms of risk-return profiles.•A sensitivity analysis on model configurations confirms the CCP’s overperformance.

论文关键词：Time series forecasting,Financial markets,Portfolio strategies,Chaos theory,Statistical mechanics

论文评审过程：Received 1 February 2021, Revised 21 September 2021, Accepted 5 October 2021, Available online 13 October 2021, Version of Record 23 October 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116055