Herfindahl-Hirschman Index for Portfolios: Quantifying Weight vs Risk Concentration

Adapting Industrial Concentration Metrics to Portfolios

Originally formulated by economist Orris Herfindahl (1950) and Albert Hirschman to evaluate industrial market concentration and antitrust policy, the Herfindahl-Hirschman Index (HHI) serves as a vital diagnostic in quantitative portfolio risk management.

In traditional industrial economics, HHI sums the squared market shares of competing firms. In investment management, applying HHI to asset weights provides a clear mathematical measure of capital concentration:

$\text{HHI}_{\text{weight}} = \sum_{i=1}^N w_i^2$

Where $w_i$ represents the portfolio weight of holding $i$ ($\sum w_i = 1$).

The inverse of weight HHI defines the Effective Number of Constituents (ENC):

$\text{ENC} = \frac{1}{\text{HHI}_{\text{weight}}} = \frac{1}{\sum_{i=1}^N w_i^2}$

In a portfolio of $N$ equally weighted assets ($w_i = 1/N$), $\text{HHI} = 1/N$ and $\text{ENC} = N$. In contrast, as capital concentrates into top holdings—such as mega-cap tech stocks in market-cap weighted benchmarks—HHI rises and ENC drops significantly below the total holding count.

Capital-Weight HHI vs Risk-Contribution HHI

While capital-weight HHI is widely used in fiduciary reporting, it measures only dollar allocations. It completely ignores differences in asset volatility and correlation structure. A 2% allocation to an ultra-volatile biotechnology stock generates vastly more portfolio variance than a 2% allocation to a stable consumer staples utility.

To evaluate genuine portfolio concentration, quantitative risk managers calculate Risk-Contribution HHI ($\text{HHI}_{\text{risk}}$).

Let $RC_i$ represent the percentage marginal contribution to total portfolio risk from asset $i$:

$RC_i = \frac{w_i (\Sigma w)_i}{\sigma_p^2}, \quad \text{where } \sum_{i=1}^N RC_i = 1$

Risk-Contribution HHI is the sum of squared risk contributions:

$\text{HHI}_{\text{risk}} = \sum_{i=1}^N RC_i^2$

The Effective Number of Risk Contributors ($\text{ENRC}$) is the reciprocal:

$\text{ENRC} = \frac{1}{\text{HHI}_{\text{risk}}}$

In portfolios where a small subset of high-beta or highly correlated assets dominates total portfolio variance, $\text{HHI}_{\text{risk}}$ far exceeds $\text{HHI}_{\text{weight}}$, and $\text{ENRC}$ falls well below $\text{ENC}$. This divergence exposes stealth concentration: capital appears evenly spread, but risk is heavily clustered.

Fiduciary Auditing and Benchmark-Relative Tracking

Auditing both capital and risk HHI protects institutional allocators from benchmark-hugging traps and unintended factor concentration. When evaluating external managers, fiduciaries use HHI divergence to identify whether active risk is driven by broad stock-picking skill or an unhedged bet on a single volatile sector.

Portfolio managers can audit weight and risk HHI, uncover concentration anomalies, and stress-test factor exposures using the Parallax Portfolio Analyzer, backed by the institutional frameworks detailed in Parallax Features.