Relatore
Alessandro Cardinali
University of Plymouth
Abstract
In this presentation we illustrate a novel inferential approach to estimate time-varying parameters of (multiple) locally stationary time series. This approach is based on costationary combinations, that is, time-varying deterministic linear combinations of locally stationary time series that are second-order stationary. We first review the theory of costationarity and formalize a Generalised Method of Moments estimator for the coefficient vectors. We then use this new framework to derive an estimator for the (time-varying) covariance of locally stationary time series. We show that the new covariance estimator is more efficient than classical estimators exclusively based on the evolutionary cross-periodogram through a simulation experiment. We then present a new analysis of financial log-returns showing that our new estimator is capable to highlight well known economic shocks. As a second example of our aproach we finally discuss forecasting of locally stationary time series based on costationary combinations.
Organizzazione
Alessandra Luati