QuantSeeker

QuantSeeker

Does Partial Risk Parity Improve Portfolio Diversification?

Testing a recent academic approach across equities, bonds, gold, commodities, and the US dollar

QuantSeeker's avatar
QuantSeeker
Nov 18, 2026
∙ Paid

Risk parity aims to diversify portfolios by distributing risk more evenly across assets. However, conventional risk parity relies on historical volatilities and correlations, which can be strongly influenced by common factors.

A recent paper by Anatoly B. Schmidt, Risk Parity in Terms of Partial Correlations, discusses using partial correlations to remove the influence of a common market factor.

The results are interesting and raise an important question: Does partial risk parity genuinely improve diversification, or can it introduce unintended portfolio exposures?

From Conventional to Partial Risk Parity

Traditional risk parity allocates portfolio weights so that each asset contributes equally to total portfolio risk, accounting for individual volatilities and correlations.

Partial risk parity instead removes the influence of a common market factor when estimating both asset variances and the relationships between assets.

Mathematically, the partial covariance matrix of the asset return vector (R), conditional on a common factor (M), is:

\Sigma_{\text{partial}}=\Sigma-\frac{\operatorname{Cov}(R,M)\operatorname{Cov}(M,R)}{\operatorname{Var}(M)}

where Sigma is the conventional covariance matrix.

Equivalently, it is the covariance matrix of residuals obtained by regressing each asset’s returns on the common factor.

The key distinction is that partial risk parity equalizes contributions to residual risk rather than total portfolio risk.

What Does the Academic Research Find?

Schmidt examines 12 US equity sector and industry ETFs over 2007–2025, using SPY as the conditioning market factor. He compares conventional and partial versions of risk parity, minimum variance, and maximum Sharpe, alongside equal weighting.

All three partial-correlation approaches achieve higher Sharpe ratios than their conventional counterparts. However, partial risk parity increases portfolio concentration, while partial minimum variance improves diversification.

Partial risk parity increases the allocation to technology (XLK) from approximately 8% to 26%, which Schmidt suggests is primarily due to XLK’s negative partial correlations with other sectors.

Partial minimum variance, by contrast, produces a more diversified portfolio overall, although it also increases the allocation to XLK.

These findings illustrate how partial correlations can substantially change portfolio allocations, with different implications for concentration depending on the optimization objective.

This observation motivated our current post.

What happens when we apply partial risk parity to a diversified multi-asset portfolio rather than a collection of equity sectors?

This post is for paid subscribers

Already a paid subscriber? Sign in
© 2026 QuantSeeker · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for writing, podcasting, and media founders.