Investors always try to maximize their profits. One of the place where they can invest their money is stock market. But they should consider that every investment has its own risk. due to the kind of problems and our needs, we should apply the best and most related risk criteria. The main goal of this article is portfolio optimization using parametric and non-parametric models such as VaR, C-VaR Monte Carlo simulation. for estimating the level of risk. This is a procedure where the theoretical probability distribution of losses is known, but we do not know its parameters. We must then be suitably approximated with some theoretical probability distributions and thus used for the distribution of individual risk rate. Then, we make a portfolio with maximizing return and minimizing risk. Also, we build an efficient market frontier. the results simulate with Monte-Carlo simulation for portfolio optimization. The GARCH model is used for volatility forecasting. The statistical population under research contains five companies with symbols such as Fakhuz, Zagros, Khodro, Fameli and Kechad. The data are on daily basis from 2017 to 2019. Finally, we focused on conclusions and introduced the best risk criteria due to research.