GARCH Model-Based Fuzzy Clustering of Asian and Oceania Stock Markets

This paper applies fuzzy clustering based on partitioning around medoids algorithm to partition financial time series. To measure the similarity or dissimilarity between employed time series, GARCH parametric approach is exerted. Based on the fuzzy and crisp silhouette criteria, some main Asian and Oceania stock markets are partitioned in two (the best choice for number of) clusters. The results indicate that the stock market of Thailand, Taiwan, Singapore, Australia, India, Philippines, South korea, Hong kong, Indonesia and China have similar trend with high degree membership.