Hierarchical Clustering Based Asset Allocation

Abstract

A hierarchical clustering based asset allocation method, which uses graph theory and machine learning techniques, is proposed. Hierarchical clustering refers to the formation of a recursive clustering, suggested by the data, not defined a priori. Several hierarchical clustering methods are presented and tested. Once the assets are hierarchically clustered, a simple and efficient capital allocation within and across clusters of assets at multiple hierarchical levels is computed. The out-of-sample performances of hierarchical clustering based portfolios and more traditional risk-based portfolios are evaluated across three disparate datasets. To avoid data snooping, the comparison of profit measures is assessed using the bootstrap based model confidence set procedure. The empirical results indicate that hierarchical clustering based portfolios are robust, truly diversified and achieve statistically better risk-adjusted performances than commonly used portfolio optimization techniques.

Keywords: Hierarchical Clustering, Asset Allocation, Model Confidence Set, Portfolio Construction, Graph Theory, Financial Networks, Machine Learning, Risk Parity

JEL Classification: G00, G10, G11