CLUSTERING IN EUROPEAN STOCK INDICES IN CRISIS AND NON-CRISIS PERIODS

Abstract

Grouping the major indices of stock markets based on their homogeneities may facilitate the selection period for investors especially today's information rich financial world. This paper attempts to detect and group the homogenous stock indices in Europe both throughout the crisis and non-crisis periods. The daily index returns of leading stock exchanges over the period 03.01.2007-09.04.2013 are considered, one of the hierarchical clustering techniques so-called Ward's Method is applied and similar cases are evaluated respectively. Then Wilcoxon signed rank test is employed for the same periods on daily index returns and meaningful differences are found.