Hierarchical Clustering Approach to Portfolio Selection : The Case of Istanbul Stock Exchange

In this study, stocks traded in the Istanbul Stock Exchange-100 (ISE- 100) Index are classified according to a risk-return criterion using hierarchical clustering algorithms. Doing this, it is intended to show that how an informed investor can make more rational investments using cluster analysis. Stocks traded in the ISE-100 Index are sorted into 10 clusters and it was measured if the chosen first-period stock clusters should be chosen also in the second period in order to get profit. An investor using the information presented here can choose either to invest in high return stocks with the same risk level, or to invest in low risk stocks with the same return level; in order to maximize the profit gained