Statistical distances between countries and cluster structures in EU area according to macroeconomic indices fluctuations

Abstract

The paper applies some recent developments of network analysis in order to perform a comparative study of EU countries in relation with the fluctuations of some macroeconomic indicators. The statistical distances between countries, calculated for various moving average time windows, are mapped into the ultrametric subdominant space as in classical Minimal Spanning Tree methods. The novelty consists in applying the so-called Moving Average Minimal Length Path algorithm, which allows a decoupling of fluctuations with respect to the mass center of the system from the movement of the mass center itself. The present analysis pertains to the Gross Domestic Product and some of its sources for 15 EU countries between 1972 and 2004. It is found that the strongly correlated countries with respect to GDP fluctuations can be partitioned into stable clusters. The final structure proves to be stable against the fluctuations induced by the moving time window over the scanned time interval.