Dynamic portfolio optimization with inverse covariance clustering
作者：
Highlights：
• Each period exists a supreme market regime with persistently high log-likelihood.
• Optimizing only on this supreme regime yields superior portfolio performance.
• Market regime clustering and dynamic portfolio optimization process are integrated.
• The performance is significant, and agnostic to markets and optimization methods.
• Inverse covariance sparsification improves both clustering and optimization.
摘要
•Each period exists a supreme market regime with persistently high log-likelihood.•Optimizing only on this supreme regime yields superior portfolio performance.•Market regime clustering and dynamic portfolio optimization process are integrated.•The performance is significant, and agnostic to markets and optimization methods.•Inverse covariance sparsification improves both clustering and optimization.
论文关键词：Dynamic portfolio optimization,Portfolio management,Financial market states,Market regimes,Temporal clustering,Information filtering networks,Covariance structure
论文评审过程：Received 4 April 2022, Revised 20 June 2022, Accepted 29 August 2022, Available online 13 September 2022, Version of Record 26 September 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.118739
