In this article, we concern ourselves with the modelling of macroeconomic time series within the unit-roots framework. We apply two approaches which seem to be well suited to model business cycles and long-run growth phenomena. First, we apply the Markov-switching model which is built around the idea that a variable may be associated with different regimes. We extend this approach to allow for asymmetries in business cycles and find that with this modification the model identifies regimes which cannot be associated with notions of the business cycle. Second, in a cointegration analysis we examine common stochastic trends and international transmission of macroeconomic shocks. The results show that transient shocks do not vanish, but have long persistent effects. Furthermore, we supplement the cointegration approach with an impulse response analysis and find that there exists a transmission of shocks between countries which indicates great international dependencies in economic activity.