English

Estimation for Latent Factor Models for High-Dimensional Time Series

Statistics Theory 2010-06-15 v3 Methodology Statistics Theory

Abstract

This paper deals with the dimension reduction for high-dimensional time series based on common factors. In particular we allow the dimension of time series pp to be as large as, or even larger than, the sample size nn. The estimation for the factor loading matrix and the factor process itself is carried out via an eigenanalysis for a p×pp\times p non-negative definite matrix. We show that when all the factors are strong in the sense that the norm of each column in the factor loading matrix is of the order p1/2p^{1/2}, the estimator for the factor loading matrix, as well as the resulting estimator for the precision matrix of the original pp-variant time series, are weakly consistent in L2L_2-norm with the convergence rates independent of pp. This result exhibits clearly that the `curse' is canceled out by the `blessings' in dimensionality. We also establish the asymptotic properties of the estimation when not all factors are strong. For the latter case, a two-step estimation procedure is preferred accordingly to the asymptotic theory. The proposed methods together with their asymptotic properties are further illustrated in a simulation study. An application to a real data set is also reported.

Keywords

Cite

@article{arxiv.1004.2138,
  title  = {Estimation for Latent Factor Models for High-Dimensional Time Series},
  author = {Clifford Lam and Qiwei Yao and Neil Bathia},
  journal= {arXiv preprint arXiv:1004.2138},
  year   = {2010}
}

Comments

35 pages article, 4 figures

R2 v1 2026-06-21T15:09:44.686Z