基于指数分层结构算法的风格资产聚类及动态配置有效性研究

An Effectiveness Study on Style Asset Categorizing and Dynamic Allocation from the Perspective of Index Hierarchical Structure Algorithm

运用A股市场21个概念板块指数2009～2014年的日对数收益率数据,依据彼此间的相关关系和拓扑结构特征,使用指数分层结构算法对其进行聚类,并据此动态构建风格资产组合。实证检验和对比分析表明,聚类结果具有稳定性和有效性,构建的资产组合具有相对较低的风险及较高的收益,能够为积极资产管理提供有益指导。

The index hierarchical structure algorithm,which is tested by the daily data of 21 Chinese style indices during 2009- 2014 according to their characteristics of correlativity and topological structure,can deduce a stabilized asset categorization. A further empirical test is conducted by using a group of stocks belonging to different style asset. The results show that asset categorization possesses stability and effectiveness,and that the constructed asset allocation helps decrease the total risk and increase the rate of return,which provides useful guidance for active asset management.