Locally Stationary Wavelet Factor Model

Details

Supervisors

Van Bellegem, Sébastien

Faculty

Faculté des sciences économiques, sociales, politiques et de communication

Degree label

Master [120] en sciences économiques, orientation économétrie, à finalité approfondie

Abstract

In this paper we propose a new factor model based on Locally Stationary Wavelet processes of Nason et al. (2000). This peculiar decomposition is well-suited for analyzing nonlinear and non-stationary time series. The time-varying factor loadings are estimated by the eigenvectors of the Evolutionary Wavelet Spectrum, the wavelet analogue of the covariance matrix. We prove convergence under general identification conditions without directly solving the rotational indeterminacy inherent to factor models. We allow both time and cross-section dimensions go to infinity and pursue a simulation study to illustrate the convergences.