One of the most important methods employed to measure the market risk is value at risk calculation method. In this study, the value at risk of banks listed on the Tehran Stock Exchange and Over-the-counter (OTC) are calculated using parametric model, Monte Carlo simulation, historical simulation and Two-Sided Power (TSP) Distribution. The sample includes all listed banks in Iran. The results showed that the value at risk estimated by TSP and historical models is more accurate than the VaR estimated by Monte Carlo and GARCH models. TSP model and then historical model are more accurate than the other ones. Moreover, GARCH is the least accurate model. So far, no research has been conducted to investigate all four models of value at risk assessment. JEL Classification: E5, E58, J21
