Comparing Performance of M.V, E.G.P and M.V.S Based on Genetic Algorithm in Iranian Capital Market

Sanati, Ali | 2013

564 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 44397 (44)
  4. University: Sharif University of Technology
  5. Department: Management and Economics
  6. Advisor(s): Bahramgiri, Mohsen
  7. Abstract:
  8. The portfolio selection problem is always one of the most important problems of finance and investments due to its great implication and vital role in financial institutions. Many of researches in this area are based on the mean-variance model, originally proposed by Markoitz. In the last two decades, however, researchers and investors have attracted to some new models that import some new factors other than mean and variance in the portfolio decision problem, such as different risk measures, etc. In this research we compare performances of mean-variance, Elton-Gruber-Padberg (EGP) and mean-variance-skewness based on genetic algorithm in Tehran Stock Exchange. Moreover, in order to find the most efficient weight for skewness in the mean-variance-skewness model we compare performance of this model with three different weights. As performance measures we have chosen Sharpe ration and future returns. Results shows that with regard to Sharpe ratio as performance measure, mean-variance model has been in the first place, followed by EGP and mean-variance-skewness, which seem to have same performance. However, with regard to future returns as performance measure, EGP has had the highest performance and the two other models, which performed same as each other, have been in the second place. Furthermore, this research shows that the most efficient weight for skewness in the so called model is 0.01 for Iranian capital market
  9. Keywords:
  10. Genetic Algorithm ; Portfolio Optimization ; Mean-Variance Method ; Elton-Gruber-Padberg Model ; Mean-Variance-Skewness Model

 Digital Object List