Mapping the Global Business Cycle Network

Abstract

We build a weighted network of business cycle similarities across countries and assess its main quantitative properties. Business cycle similarity is measured at the annual frequency using the Euclidean distance. Network analysis is well suited to map the full set of pairwise similarities at an annual frequency. We find that the global business cycle network has become more dense and more homogeneous over time, reflecting global trends such as rising trade and financial integration. At the same time, similarity exhibits a jagged pattern, underscoring the importance of also taking into account short-term factors to explain the dynamics of global business cycle interdependence. Unlike earlier studies focused on aggregate measures of similarity, our empirical approach is able to uncover and assess both the long-term trend rise and the short-term pattern of business cycle similarity.