Comparison of Value Risk Models and CoppolaCVaR in Portfolio Optimization in Tehran Stock Exchange

To optimize the investment portfolio, conditional value at risk is a new approach. To amend the non-normal distribution of return on assets, and the non-linear correlation between return, the modification of Copula-CVaR compound method has a better performance in measuring portfolio risk. In this study, it has been attempted to present a more efficient model for portfolio optimization that provides greater returns for investors, given the uncertainty investment conditions. The VaR model was compared variance-covariance approach and the Copula-CVaR model for their efficiency frontier. The research area entails of 2014 up 2018; The statistical population was the top 50 companies of TSE.The variance-covariance approach was used to estimate the VaR of the portfolio. Moreover to estimate the CopulaCVaR model we have used the ARIMA-GARCH time series disruption component of the asset return distribution; then the marginal distributions of the assets were estimated using the CAPA-Student function; finally through Monte Carlo simulation the return on assets and their CVaR for the 10-day period were calculated. The optimal portfolio composition was determined at 95% and 99% confidence levels for different levels of risk. The results of this study showed that the optimized portfolio formation using the compound model, the Copula-CVaR model, has performed better