Document Type : Research Paper
Author
Assistant Professor of Economics, Nahavand Higher Education Complex, Bu-Ali Sina University, Hamedan, Iran,
Abstract
This paper introduces a framework for fuzzy clustering of financial time series based on directional volatility spillovers. Detecting clusters in the volatility spillovers between financial time series provides deep insight into market structure, which can be useful to portfolio managers as well as policymakers. For this purpose, firstly, the directional volatility spillovers - “From” and “To” others - are measured based on the generalized forecast-error variance decomposition (GFEVD) approach. Moreover, to measure the dissimilarity, the improved weighted Euclidean distance based on the directional spillovers is used. Then, by adopting a fuzzy approach of partitioning around medoids, namely, (VS-FCMdd) volatility spillover-based fuzzy C-Medoids Clustering Model, optimal weights and degrees of membership are estimated. In addition, to reduce the impact of outliers in the clustering process, an exponential transformation of the weighted dissimilarity measure is also considered in the fuzzy clustering model, as (VS-E-FCMdd) volatility spillover-based exponential fuzzy C-Medoids Clustering Model. In this research, for the first time, these two fuzzy clustering models are employed to analyze the directional volatility spillovers in the Tehran Securities Exchange. For this purpose, a sample comprising 30 stock picks for 28 industries, from 1387 to 1402 is considered. The value of the Xie-Beni index reaches the minimum value at cluster 2. The comparison of results shows that the VS-E-FCMdd produces more fuzzy partitions and is effective in clustering the stock volatility spillovers of the Tehran Stock Exchange. The stocks listed on the Tehran Stock Exchange are differentiated mostly in terms of “From” volatility spillover, i.e., the amount of volatility spillover that each stock receives from all other stocks. Particularly, cluster 1 includes the stocks which, in comparison to others, receive more “From” spillovers, but transfer fewer “To” spillovers.
Keywords
- Fuzzy Partitioning Around Medoids
- Directional Volatility Spillovers
- Weighted Dissimilarity Measure
- Exponential Distance
- Tehran Securities Exchange
Main Subjects
)