Network centrality and portfolio optimization using the genetic algorithm

This study aims to optimize the portfolio using the genetic operator and network centralization. The statistical population of the study is the top ۵۰ companies of Tehran Stock Exchange, in the first quarter of ۲۰۲۱, and to calculate the size of centrality, we used the difference in the overall performance of each company compared to all the top companies, based on a standard hybridization indicator. Then based on the companies’ performance in the capital market, the geometric mean of risk and return of efficient companies are determined, and given the real limitations of the budget, the requirements and expectations of the investors compared to the market’s performance and the risk-free investment, the problem of decision-making for the composition of the investment in the form of a multi-purpose paradigm is formulated. By using the modified optimization algorithm and the genetic algorithm with dual operators, we optimized the investment’s composition. Finally, we use the compound linear regression with data analysis approach to evaluate the effect of individual and systemic operators on determining the investment strategy, and the results represented the positive effect of these two operators.

Keywords:

Portfolio Optimization , Network centralization , Genetic Algorithm , Risk , Return volatility