عنوان مقاله [English]
نویسندگان [English]چکیده [English]
One of the most important concerns of investors in the capital market is the selection of a portfolio that is optimal in terms of profitability. Therefore, variability in portfolio selection method in investment and complexity in decision making has been greatly expanded in recent decades. Traditional methods in portfolio selection and optimization do not offer enough efficiency that is why the use of innovative methods is preferred. The purpose of this study is to model portfolio selection problem using different risk measures, including variance, semi-variance, value at risk, as well as conditional value at risk and optimization of its value using one of these algorithms, i.e., the ant colony algorithm. In this regard, a comparison of the efficient frontier of different risk methods and a comparison of different models in terms of the CPU time has been made. To determine the optimal portfolio, financial data of listed companies at the Tehran Stock Exchange were used for the period of 2001-2008 (1380-1387). Furthermore, for designing the graphs of the study in order that the comparison of different risk methods is rendered in terms of CPU time, the Excel2007 software was used and to compare different efficient frontier the Matlab7.1 software was utilized.
The results of the review indicated that the Mean-CVAR method can show higher levels of return with the least conditional value at risk. On the other hand, run time of the algorithm for the four methods in different sizes of the portfolio was compared. According to the presented results, the time spent on the implementation of the Mean-Variance method assigned the lowest, and the time spent for the implementation of the Mean-CVAR method appropriated the highest time respectively.
As a result, it can be said that although the CVAR offers better efficient frontier it is not a good measure for the big sizes of portfolio. That is why, in many cases variance due to its calculation simplicity is still used as risk criteria by many investors