Measuring portfolio Value at Risk: The application of copula approach

Subject Areas : Financial engineering

Esmaeil Pishbahar ¹ , sahar abedi ²

1 - Associate professor, Department of Agricultural Economics, Faculty of Agriculture, Tabriz University, Iran
2 - MSc. Graduated of Agricultural Economics, Faculty of Agriculture, Tabriz University, Iran

Received: 2016-09-07 Accepted : 2016-10-20 Published : 2017-04-21

Keywords: Multivariate value at risk, Evaluation tests, Copula approach, Food industry exchange,

Abstract :

Due to the fact that traditional univariate approach in portfolio value at risk measurements ignore the time varying correlation between its components, these models underestimate or overestimate value at risk. In addition, complex financial markets make it necessary to use effective approaches, such as multivariate risk measurement. Therefore, in this present study, we tried to evaluate four multivariate value at risk measurement approaches for two portfolios in food industry exchange. The result of Christoffersen, quadratic probability score and root mean squared error tests showed copula-based Monte Carlo approach has more reliable result in comparison with others. Hence, we applied this approach to investigate dependence structure and measure risk, and its result showed the maximum expected loss of dairy portfolio value over a week is 2.01 percent, while for sugar portfolio is 1.09 percent.