Robust Optimization of Portfolio’s Value at Risk (VaR) by Using Random Uncertainty Series

Mojgan Fakhri


This study is conducted with the aim of robust optimization of value at risk (VaR) of portfolio by using random interval uncertainty series. Therefore, the present study is an applied study in terms of research method. Statistical population of this study is consisted of companies active in Tehran Stock Exchange during the time frame of 2005 to 2014, which have been selected by using Elimination method. Results of feedback test and calculation of value at risk related to portfolios during the examined time frame indicated that calculated value at risk by using daily returns is correct. The Genetic Algorithm used in this study is a Single-step algorithm and the used selection technique is Roulette wheel with 2000 generations and each generation is consisted of 20 individuals. Results indicated the existence of algorithm convergence. Also, testing algorithm’s stability indicated a small and insignificant difference between answers resulting from algorithm’s Iteration in multiple times. Results also indicated that application of robust optimization method improves portfolio performance in Real-world problems with uncertainty and that internal random series have a higher level of ability in modeling asymmetric uncertainties in financial fields. 


Robust optimization, value at risk and random internal uncertainty series.

Full Text:



  • There are currently no refbacks.