Comparison of Traditional DEA Approaches to Portfolio Selection by a New Proposed Algorithm with a Case Study of Stock Selection in Tehran Stock Exchange

Portfolio selection is one of the most important areas in financial decision-making; A portfolio of stocks that could bring the highest rate of return and the lowest risk investment for its owner simultaneously. However in choosing the most prefered portfolio just these factors are not decisive and according to the economic environment, many factors can affect this process which should be employed. Therefore, these diversity of factors, bring to the limelight the importance of multi-criteria decision-making approaches. Data Envelopment Analysis (DEA) is one of this approaches. The main purpose of this paper is comparing the traditional DEA approaches to a new proposed algorithm. In traditional approaches simply assumed that return to scale is constant or variable. This simplification may cause large errors. In the new algorithm by analyzing the behavior of return to scale, appropriate model will be used. As a case study, the models have been solved with real data belonging to Tehran stock exchange and the results have been analyzed.