Multi-objective Portfolio Optimization Model by Fruit Fly Optimization Algorithm

One of the most famous optimization problems in the field of financial engineering is portfolio selection problem. In its simplest form¡ while trying to minimize risk in the portfolio selection according to defined constraints such as budget and integer constraints it deals with selecting a basket of various assets. Generally¡ investors prefer to invest in some assets rather than investing in only one asset to reduce unsystematic risk by diversifying their investment. Complex computational models have been developed to solve this problem and there is not an optimal solution for many of them. In this paper¡ a new and innovative approach known as fruit fly optimization algorithm (FOA) is used for multi-objective problem solving based on mean-variance Markowitz problem with class and cardinality constraints. Fruit fly optimization algorithm is a new way to find the overall optimal solution based on the behavior of the fruit fly in finding food. So far¡ few studies have been done on this algorithm and almost none of them used this algorithm for portfolio optimization problem. The results indicated better comparative performance of the algorithm compared to genetic algorithm for data set of Tehran stock exchange.