Identification and clustering of effective industrial at Tehran stock market based on minimum spanning tree of distance ultra-metric

Abstract

In essence, the stock market is a complex network. The price fluctuations among various stocks have complicated relationships. In this paper, the Minimum spanning tree (MST) is used in order to identify, categorize, and evaluate the relationship between different industries in Tehran Stock Exchange in 2012, 35 industries were addressed. Thus, the high degree repetition of cross-correlation among different pair indicators of industries investigated in Tehran stock market, the result is a hierarchical graph based-on metric distance between different pair industries. In following based-on ultra-metric function resulted from metric, a Minimum spanning tree exists from different industries, which showed relation among industries and their influences on each other. The present study in terms of aim of the research is applicable and in terms of the nature of the research, is Analysis –inference. In this paper, in order to analyze the data and obtain a Minimum spanning tree, Excel, Eviews and Matlab are used. Industries are divided to three categories according to the Minimum spanning tree obtained, based on the minimum distance and the effect, are: The first one are industries with five top-grade in the Minimum spanning tree, The second group of industries that have Vertices of degree four in the Minimum spanning tree and The third group of industries that have Vertex third degree in the Minimum spanning tree. The first category are the most effective industries that include Basic Metals, cement, lime and chalk.