Business and growth cycle analysis for the Euro area

This thesis revisits the issue of business cycle synchronisation in the Euro area by utilising time-series models that may overcome some of the drawbacks in the existing literature. Two major contributions are made to the existing literature of evaluating cycle synchronisation. First, instead of identifying turning points from individual macroeconomic timeseries, as carried out in most studies, this thesis obtains turning points from multivariate information. It is hoped that including more variables containing business cycle information in the dating process may produce more accurate turning points and, in turn, improve the accuracy of measuring cycle correlation. In doing so, both parametric and non-parametric business cycle dating procedures are used. These include the quarterly Bry–Boschan (BBQ) algorithm, a single dynamic factor model and the Markov switching dynamic factor model. Second, unlike the traditional approach that measures growth cycle synchronisation in the euro area by calculating pairwise cycle correlations, this thesis analyses the degree of growth cycle co-movement within a multivariate setting by using a VAR model with cointegration. [Continues.