Islamic CAPM based on Cluster Analysis for Portfolio Selection

Abstract

This research combines portfolio management and cluster analysis as an alternative diversification strategy to reduce risk. The method used is a comparative approach to compare the performance of the formed portfolio. The object of this research was taken purposively from JII stocks for January 2018 to June 2023. Portfolio selection is based on two hierarchical clustering techniques, namely Ward’s algorithm and Complete Linkage. Meanwhile, the portfolio formation uses Islamic CAPM by replacing the risk-free rate with a sukuk coupon. The results of the analysis show that cluster analysis can be an effective tool for investors to select stocks. Each clustering technique produces 4 clusters, and then from each cluster, the most representative stock is selected so that the portfolio consists of 4 stocks each. Based on the Sharpe index, the Complete Linkage clustering technique is proven to outperform in portfolio formation compared to Ward’s algorithm.