Asymptotics of Empirical Eigen-Structure for Ultra-High-Dimensional Spiked Covariance Model

• Activity Number: 311
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 8:30 AM to 10:20 AM
• Sponsor: IMS
• Author(s): Weichen Wang* and Jianqing Fan
• Companies: Princeton University and Princeton University
• Keywords: Asymptotic distributions ; Principal component analysis ; Spiked covariance model ; Ultra-high dimension ; Diverging eigenvalues ; Approximate factor model

Abstract

We derived the asymptotic distributions of the spiked eigenvalues and eigenvectors under a generalized and unified asymptotic regime, which takes into account the spikeness of leading eigenvalues, sample size, and dimensionality. This new regime allows high dimensionality and diverging eigenvalue spikes and provides new insights on the roles the three key quantities played in PCA. The results are proven by a new technical device, which swaps the role of rows and columns and converts the high-dimensional problems into low-dimensional ones. Our results are a natural extension of those in Paul (2007) to more general setting with new insights and solve the rates of convergence problems in Shen et al. (2013). They also reveal the biases of the estimation of leading eigenvalues and eigenvectors by using PCA, and lead to a new covariance estimator for the approximate factor model, called shrinkage principal orthogonal complement thresholding (S-POET), which corrects the biases. Our results are successfully applied to outstanding problems in estimation of risks of large portfolios and false discovery proportions for dependent test statistics and are illustrated by simulation studies.