Hierarchical Equal Risk Contribution for Portfolio Allocation

Summary

The article describes Hierarchical Equal Risk Contribution (HERC), a portfolio allocation method that combines hierarchical clustering with cluster-aware capital allocation and risk balancing. It first groups assets from their return correlations, selects a suitable number of clusters, allocates capital across clusters through top-down recursive bisection, and applies risk parity within clusters. Unlike standard Hierarchical Risk Parity, HERC aims to respect the dendrogram’s cluster structure and allows linkage methods beyond single linkage.

The article motivates these design choices by discussing HRP’s deep, chain-like trees, potential overfitting when trees grow fully, and bisections that may split natural clusters. It also describes alternatives to variance, including expected shortfall and conditional drawdown at risk. A portfolio comparison illustrates that changing the risk measure or allocation procedure can alter weights, but the author characterizes it as rudimentary and calls for statistically adjusted analysis such as resampling. Claims of improved out-of-sample risk-adjusted performance are presented as motivation, not established by the brief comparison.

Key ideas

• HERC combines hierarchical clustering, cluster selection, recursive allocation, and risk parity within clusters.
• It seeks to preserve dendrogram structure when dividing portfolio weights.
• The method supports linkage choices beyond the single linkage used by HRP.
• Risk allocation can use downside measures such as expected shortfall or conditional drawdown at risk.
• The article’s portfolio comparison is illustrative and does not establish robust comparative performance.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.