Economic Versus Statistical Clustering in Multi-Asset Multi-Factor Strategies

Abstract

Maximizing for diversification in the multi-asset multi-factor universe, the literature advances diversified risk parity strategies across economic clusters. For handling overly complex correlation matrices, hierarchical clustering techniques have recently been put forward to guide risk parity allocations. Indeed, such statistical clusters might be considered natural portfolio building blocks given that they automatically pick up the dependence structure and thus form meaningful ingredients to aid portfolio diversification. We explain the intuition and nature of hierarchical clustering techniques in the context of multi-asset multi-factor investing vis-à-vis the use of economic factors in diversified risk-based allocation paradigms such as 1/N, minimum-variance and diversified risk parity.

Keywords: Multi-asset multi-factor investing, diversification, hierarchical risk parity

JEL Classification: G11, D81