Unity and Plurality of the European Cycle

We apply uni- and multivariate unobserved components models to the study of European growth cycles. The multivariate dimension enables to search similar or, more strongly, common components among national GDP series (quarterly data from 1960 to 1999). Three successive ways to exhibit the European cycle satisfactorily converge: the direct decomposition of the aggregate European GDP; the aggregation of the member countries' national cycles; the search for common components between these national cycles. The European aggregate fluctuations reveal two distinct cyclical components, assimilated to the classical Juglar (decennial, related to investment) and Kitchin (triennial, related to inventories) cycles. The European Juglar cycle cannot be reduced to a single common component of the national cycles. It has at least a dimension of "three": it can be understood as the interference of three elementary and independent sequences of stochastic shocks, that correspond to the European geographical division. The euro-zone is not yet an optimal currency area, as the shocks generating the European cycles are not completely symmetrical. Studying the sequences of innovations extracted from the models shows that euro-zone vulnerability to strong shocks and asymmetry of these shocks tend to decrease during the last decades, but this trend is neither regular, nor irreversible.