Providing a Meta-Heuristic Method to Form the Portfolio (Stock Portfolio) Using TOPSIS Technique Based on K-Means Algorithm

The stock exchange is one of the major financial markets, which is an indicator of the economy of every country. Stagnation and boom of the stock exchange affect not only the national economy but also regional and global economy. On the one hand stock exchange market collects liquidity and savings of the private sector to finance investment projects and on the other hand it is an official and confident authority that holders of the stagnant savings can relatively seek an appropriate and safe place for investment and employs their funds to invest in companies. This study has discussed a meta-heuristic method to form the stock portfolio using TOPSIS technique based on K-Means algorithm (between 2009–2012). In the present study the companies that have been accepted in Tehran Stock Exchange cluster by K-Means algorithm method and then the stock portfolio will be formed. Two methods provided to form the stock portfolio named 1- clustered portfolio using the K-Means algorithm 2- non-clustered portfolio. After the stock portfolio is formed its optimization will be evaluated and the optimal portfolio selects among the portfolios that have been formed. The results of the study show that clustered portfolio is better than non-clustered portfolio and has better Sharpe ratio and investors can use the method to form the Portfolio.