Friday, 21 October 2011: 9:10 AM
We examine the cyclical components of the Euro area and Switzerland for the period 1995 – 2010 with quarterly real, non-seasonally adjusted, GDP time series. We use the Singular Spectrum Analysis, a relatively new, non-parametric approach, based on spectral decomposition of time series, as a device for separating the above time series into a set of unobserved components (trend, cycle, seasonals and noise). After this decomposition, the two cyclical components, containing frequencies in the typical business cycle band, are analyzed by means of frequency domain methods, in terms of spectral and cross-spectral densities, squared coherence, dynamic correlation and phase spectrum. The analysis indicates that there is evidence of some degree of synchronization between Euro are and Switzerland.