Business Cycle Analysis: Comparison of Approaches to Cycle Measurement

In the context of recent economic recession public attention was focused on the business cycle analysis and possibility of forecasting the cycle movements. One of the methods used for the business cycle analysis is based on the study of composite indicators which combine several individual economic indexes. The indexes can be divided into groups of leading, coincident and lagging ones with regard to the reference time series (usually GDP or industrial production index). The dependency between cycles of the individual indicators and the reference series can be determined with cross correlations and it is essential to adjust the time series first, so we get rid of the trend and seasonal component. The traditional approach (used by OECD and other organizations) is using algorithms like Hodrick-Prescott filter for the decomposition of the time series. We compare this methodology with smoothing the series of growth rates and with direct usage of the year-to-year growth rates. We discuss how these methods affect the cross correlations and the selection of the individual indexes for the composite indicators.