A Hybrid Model for Portfolio Optimization Based on Stock Clustering and Different Investment Strategies
In today's dynamic business environment, in order to compete in the market, financial institutes are trying to find the best portfolio policy that in turn leads to an increase in the return and a decrease in the risk for the investors. The objective of this study is to develop a portfolio considering the behavior of investors in risk taking. This research aims to support investors, experts and intermediate managers in establishing optimized portfolio of stocks according to investment strategy. The proposed model has used the five indexes of risk, return, skewness, liquidity and current ratio of 66 companies that enlisted in Tehran Stock Exchange Market and then clustered different companies using the hybrid method of clustering algorithm. After that, the clusters ranked using Topsis method. Ultimately, using genetic algorithm, the portfolio is established for different classes of investors with respect to their risk-taking level. The results show that the proposed model in comparison to general index, the industry index and the index of 50 more active companies are better in Tehran Stock Exchange.
Keywords: portfolio optimization, clustering, neural network, genetic algorithm
JEL Classifications: C880, C610