A0691
Title: Estimation and inference in large dimensional threshold factor models with weaker loadings
Authors: Daniele Massacci - Kings College London (United Kingdom) [presenting]
Abstract: Estimation and inference are studied in large-dimensional threshold factor models in which (some of) the eigenvalues of the covariance matrix of the data diverge at a rate that is slower than the cross-sectional dimension N. The convergence rate of the concentrated least squares estimator is derived for the threshold parameter, and the asymptotic distribution of the principal components estimator is obtained for factors and loadings within each regime. Finally, the relevance of these theoretical findings is illustrated for conditional asset pricing.
