A novel backtesting methodology for clustering in mean–variance portfolio optimization
作者：
Highlights：
• Novel methodology for decision making in financial portfolio management.
• Incorporate cluster analysis into the mean–variance portfolio optimization model.
• Dynamically update clustering parameters prior to the optimization and backtesting.
• 6300 backtest results from DOW, NASDAQ and S&P indices at different temporal scales.
• Increased efficiency for different backtesting periods and rolling windows.
摘要
•Novel methodology for decision making in financial portfolio management.•Incorporate cluster analysis into the mean–variance portfolio optimization model.•Dynamically update clustering parameters prior to the optimization and backtesting.•6300 backtest results from DOW, NASDAQ and S&P indices at different temporal scales.•Increased efficiency for different backtesting periods and rolling windows.
论文关键词：Asset allocation,Portfolio selection,Unsupervised learning,Cardinality,Decision support system,Performance comparison
论文评审过程：Received 6 January 2020, Revised 14 July 2020, Accepted 4 September 2020, Available online 16 September 2020, Version of Record 29 September 2020.
论文官网地址：https://doi.org/10.1016/j.knosys.2020.106454
