Topologic Structure of Chilean Stock Market in 2009-2013
Keywords:
Correlation, euclidean distance, Kruskal algorithm, clusters, leader stockAbstract
This paper studies how to map a stock market. Using historical prices of most traded stocks on the Bolsa de Comercio de Santiago, a topological map of the Chilean stock market is designed for 2009 (based on 58 shares) 2010 (75) 2011 (66) 2012 (69) and 2013 (65), using a topological structure with the Kruskal algorithm based on graph theory. In 2009 a smaller distance between shares of companies in the same economic sector is observed, but clusters are also observed for shares in different sectors. This implies that, to diversify risk, not enough to invest in shares of various economic sectors, it is also necessary to analyze the correlation between stock returns. In 2013, the hub of the market would be the Banco Santander-Chile’s stock, which is situated in a central area, linking a large number of shares belonging to various sectors of the economy. The maps also show an apparent instability in the topological structure of the Chilean stock market, which is modified from one year to another, demanding greater monitoring efforts to identify clusters, leading stocks and followers. It is expected that the use of a topological structure contributes to a greater understanding of the mutual dependence on the stock’s behavior. Furthermore, the identification of clusters and leaders and followers stocks could be useful for designing trading strategies.