Portfolio optimization by using the Copula Approach and multivariate conditional value at risk in Tehran Stock Exchange

One of the main problems of shareholders in the stock market is the discovery, quantification and calculation of market risk. In many studies, one-way distributions are used to estimate risk metrics that usually do not give credible results to the investor. Because the distribution of assets is generally a broad sequence, and the results of computations are not acceptable for the consideration of the univariate normal distribution and the use of parametric methods. In this paper, using the Coppola theory, we calculate risk-weighted value (VaR) and conditional value-at-risk (CVaR). After estimating the multivariate T- Copula and the normal distribution of multivariate, the Monte Carlo method is used to generate a scenario for calculating the variance of the portfolio as well as risk estimation. Also, the calculations performed using the loss function method are tested and the accuracy of the approximations is verified. Finally, the minimum value of the copula based on the variance of the portfolio as well as its CVaR value is considered as the function of the portfolio planning, and the optimal portfolio is obtained by considering the weight of each share index. In the calculation of the 1200 index, we consider a sample basket of different industries, by calculating VaR and CVaR with confidence levels of 95 and 99 percent. The results obtained from the efficiency and reliability of the Monte Carlo simulation by the Copula T-Student versus the normalized multivariate distribution.