Paperlane

no. 13 · Wednesday 11 November 2026

Spectral Characterisation of Covariance Existence in Markov-Switching Affine Recurrences

Wilson Tai

math.PR · v2

Abstract

This paper provides a finite-dimensional spectral characterisation of covariance existence for finite-state Markov-switching affine recurrences of the form \[ Z_{n+1}=A_{Θ_{n+1}}Z_n+ζ_{n+1}. \] While strict stationarity is typically governed by negativity of the top Lyapunov exponent of the random matrix products, this does not ensure finite variances, covariances, or Pearson correlations. We show that these second-order properties are governed instead by a Markov-switching Kronecker operator. Writing \(T_{ji}:=p_{ij}(A_j\otimes A_j)\) and denoting its spectral radius by \(ρ(T)\), we prove that, under a natural Perron-excitation condition on the innovation covariance, \(ρ(T)<1\) is necessary and sufficient for the stationary solution to be square-integrable. When this condition holds, the regime-weighted second moments solve a finite-dimensional linear system, yielding the covariance matrix explicitly. Thus, the mean-square spectral condition becomes an exact covariance-existence criterion for the stationary law. Examples illustrate the separation: the top Lyapunov exponent may be negative while the second-order spectral radius exceeds one, so strict stationarity can persist even when finite second moments, and hence population covariances and Pearson correlations, do not exist.

fields
Probability
pdf ↗ page arXiv:2607.09994 ↗