Abstract:
This research proposed 1) to compare the structure of the stocks in SET50 between the distance values based on Escoufiers RV coefficient and the distance values based on method of symbolization, using Minimum Spanning Tree (MST) and Hierarchical Tree (HT), with the time series of daily closing prices and volume trading from February 1, 2016 to February 2, 2017. 2) To apply Bootstrap Technique to show reliability of the relationship structure of stocks in MST and HT. The result showed that 1. The structure of MST based on the symbolization methods and based on the RV coefficient are similar and consisted five main clusters in business sector, banking, information technology & communication, petrochemicals & chemicals, energy & utilities, and finance & securities. In addition, the relationship structure of the stock has changed as the market situation changes. In during the time that the stock market fluctuates in a trendless, the stocks were included more clustering but grouping by business sector was less. The structure based on the symbolization methods was very sensitive to shows changes in the relative-stock structure faster than the structure based on the RV coefficient. 2. The HT shows the sequence of the connection of the stock within five main clusters in business sector, banking, information technology & communication, energy & utilities (Oil & Refinery), petrochemicals & chemicals, and finance & securities. 3. From the reliability of the MST and HT using the bootstrap value, we found that the connection of the stocks within business sector had high bootstrap value and short distances. It shows the opportunity to be connected very well with high reliability.