Absorption Ratio Deep-Dive: PCA Eigenvalue Concentration as a Regime Signal
Publication Date: 18 May 2026 Classification: Observational Research — Not Investment Advice
1. What Is the Absorption Ratio?
The Absorption Ratio (AR) measures the fraction of total variance explained by the top 20% of eigenvalues (at least one) of the Spearman correlation matrix of the analysed assets:
AR = Σλ(top 20%) / Σλ(all)A high AR indicates that a small number of risk factors drive the majority of portfolio variance: the assets are moving together.
2. PCA Mechanics
Principal Component Analysis decomposes the correlation matrix C into eigenvectors and eigenvalues. MVIE uses correlation rather than covariance: on a covariance matrix the most volatile asset owns the top eigenvalue, and AR would end up measuring that asset's variance.
C = Q Λ QᵀWhere Λ is a diagonal matrix of eigenvalues sorted in descending order. When markets are diverse, eigenvalues are spread across many components. When correlations converge, eigenvalue mass concentrates in the top components, driving AR upward.
3. AR in the MVIE Composite
Within the composite:
MVIE = 0.40·s_AR + 0.35·s_IS + 0.25·s_PD, s = clip(50 + 50·z/3, 0, 100)AR carries the highest weight (40%) of the three components.
This document is produced by MVIE Pro for educational purposes only. Not investment advice.