通过自适应多任务学习提升跨行业投资组合优化的因子模型精度。
Adaptive Multi-task Learning for Multi-sector Portfolio Optimization
- 基于因子建模框架，自动学习不同行业主时序子空间的相关性。
- 在罗素3000指数日度数据上显著改善因子模型估计与投资组合表现。
- 提出投影惩罚主成分分析法，实现高效可落地的多任务学习。
在涉及多个资产类别、数量庞大的多行业投资组合优化中，跨行业信息的有效迁移既重要又具挑战性。在因子建模框架下，本文提出一种新型数据自适应多任务学习方法，用于量化并学习多个研究行业间主时序子空间（由因子张成）之间的相关性。该方法不仅提升了多个因子模型的联合估计性能，还增强了对因子模型精确恢复高度依赖的多行业投资组合优化效果。此外，开发了一种新颖且易于实现的算法——投影惩罚主成分分析（projection-penalized principal component analysis），以完成多任务学习过程。通过多种模拟设计及在罗素3000指数日收益率数据上的实际应用，验证了该方法的优势。
原文摘要 · Abstract (English)
Accurate transfer of information across multiple sectors to enhance model estimation is both significant and challenging in multi-sector portfolio optimization involving a large number of assets in different classes. Within the framework of factor modeling, we propose a novel data-adaptive multi-task learning methodology that quantifies and learns the relatedness among the principal temporal subspaces (spanned by factors) across multiple sectors under study. This approach not only improves the simultaneous estimation of multiple factor models but also enhances multi-sector portfolio optimization, which heavily depends on the accurate recovery of these factor models. Additionally, a novel and easy-to-implement algorithm, termed projection-penalized principal component analysis, is developed to accomplish the multi-task learning procedure. Diverse simulation designs and practical application on daily return data from Russell 3000 index demonstrate the advantages of multi-task learning methodology.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。