Pearson correlation coefficients.

This work describes the design of a novel financial multiplex network composed of three layers obtained by applying the MST-based cross-correlation network, using the data from 465 companies listed on the US market. The study employs a combined approach of complex multiplex networks, to examine the statistical properties of asset interdependence within the financial market. In addition, it performs an extensive analysis of both the similarities and the differences between this financial multiplex network, its individual layers, and the commonly studied stock return network. The results highlight the importance of the financial multiplex network, demonstrating that its network layers offer unique information within the multiplex dataset. Empirical analysis reveals dissimilarities between the financial multiplex network and the stock return monoplex network, indicating that the two networks provide distinct insights into the structure of the stock market. Furthermore, the financial multiplex network outperforms the singleplex network of stock returns because it has a structure that better determines the future Sharpe ratio. These findings add substantially to our understanding of the financial market system in which multiple types of relationship among financial assets play an important role.本研究阐述了一种新型金融复层网络（financial multiplex network）的构建方案：该网络基于最小生成树（Minimum Spanning Tree, MST）的互相关网络构建得到三层结构，所用数据集覆盖美国市场465家上市公司。本研究采用复杂复层网络的组合分析方法，探究金融市场内金融资产间相互依存关系的统计特性。此外，本研究还针对该金融复层网络、其各独立子层，以及经典研究中广泛采用的股票收益网络（stock return network），展开了全面的异同对比分析。研究结果凸显了金融复层网络的重要价值，证明其各网络子层在复层数据集中蕴含独特的信息维度。实证分析显示，金融复层网络与股票收益单层网络（monoplex network）存在显著差异，表明两类网络对股票市场结构的解读视角各不相同。此外，金融复层网络的表现优于股票收益单层网络，因其结构能够更精准地预测未来夏普比率（Sharpe Ratio）。上述研究发现极大地深化了我们对金融市场体系的认知——金融资产间的多重关联类型在该体系中扮演着关键角色。