Hierarchical risk parity using security selection based on peripheral assets of correlation-based minimum spanning trees

Abstract

This study proposes hierarchical risk parity portfolios using a new correlation matrix and security selection. We suggest a global motion subtracted correlation matrix, which eliminates the global motion in the cross-correlation matrix. Also, we suggest utilizing the peripheral assets of a correlation-based minimum spanning tree for security selection. The proposed portfolio strategies with security selection outperform benchmarks, showing their nature as smart beta strategies. Specifically, the full correlation with a small number and global motion subtracted correlation with a relatively large number of selected assets exhibit decent performances during the post-crisis bull markets and crisis-induced bear markets, respectively.