A0920
Title: Parametric Whittle estimation of cyclically integrated time series
Authors: Liudas Giraitis - Queen Mary University of London (United Kingdom)
Edward Hill - Queen Mary University of London (United Kingdom) [presenting]
Abstract: Macroeconomic and climatic time series may contain unknown or hidden periodicities characterized by a singularity in the spectral density function at a non-zero frequency. Recently, a study introduced a parametric ARCIMA model for the modelling of cyclically integrated time series. The focus is on the estimation of parametric ARCIMA models. The inference is performed in the frequency domain using a modified Whittle likelihood function, where the location of the singularity in the spectral density has been estimated using a separate procedure developed in a recent study. The parametric rate of consistency is derived, and the estimator's asymptotic normality is established. Simulations confirm the good performance of the two-stage estimation procedure in finite samples.
