Covariance spectra of random linear systems near instability
Abstract
We study how stationary variance is distributed among the principal components of a randomly coupled linear system near instability. The coupling matrix has entries N−1/2ξij, where the ξij are independent copies of a bounded, centered real random variable of variance one. As N→∞ and the excess damping δN→0, we prove that the covariance spectrum, with eigenvalues rescaled by δN2 and weighted by their shares of the total variance, converges in probability to an explicit beta-product law. The convergence holds throughout δN≥ N−1/2+ε for every fixed 0<ε<1/2. The law describes eigenvalues of order δN−2 at ranks of order NδN3/2. Its limiting rank profile decays as r−4/3 at large normalized rank and determines how many principal components retain a prescribed fraction of the stationary variance. The proof combines an exact computation of planar moments, uniform estimates for growing matrix words that accommodate nonsymmetric entry distributions, and resolvent bounds controlling the contribution of long times.
Related
- Linear Regression Beyond Covariance: Latent Slope Spectra from Higher-Order Cumulants
- Recursive Stochastic Linear-Quadratic Control with Indefinite Weights for Jump-Diffusion Systems with Random Coefficients
- A covariance density for a point process and a random field
- Linear Exponential Quadratic Gaussian Covariance Steering
- Score-based confidence intervals for variance-covariance parameters in linear mixed models
BibTeX
@misc{feng2026covariance,
title = {{Covariance spectra of random linear systems near instability}},
author = {Pengbin Feng},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.34920},
eprint = {2610.34920},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.34920},
}
RIS
TY - UNPB
TI - Covariance spectra of random linear systems near instability
AU - Pengbin Feng
PY - 2026
DA - 2026/10/26
PB - arXiv
DO - 10.48550/arxiv.2610.34920
UR - https://arxiv.org/abs/2610.34920
AB - We study how stationary variance is distributed among the principal components of a randomly coupled linear system near instability. The coupling matrix has entries N^−1/2ξ_ij, where the ξ_ij are independent copies of a bounded, centered real random variable of variance one. As N→∞ and the excess damping δ_N→0, we prove that the covariance spectrum, with eigenvalues rescaled by δ_N^2 and weighted by their shares of the total variance, converges in probability to an explicit beta-product law. The convergence holds throughout δ_N≥ N^−1/2+ε for every fixed 0<ε<1/2. The law describes eigenvalues of order δ_N^−2 at ranks of order Nδ_N^3/2. Its limiting rank profile decays as r^−4/3 at large normalized rank and determines how many principal components retain a prescribed fraction of the stationary variance. The proof combines an exact computation of planar moments, uniform estimates for growing matrix words that accommodate nonsymmetric entry distributions, and resolvent bounds controlling the contribution of long times.
ER -