Value at Risk as a tool for measuring market risk has become very popular in the recent past. In Pakistan, the reporting of Value at Risk forecast by listed firms and mutual funds is not practiced. This study highlights the importance of Value at Risk forecast for the regulator, so that, a framework of forecasting and reporting the Value at Risk measure by a valid model could be made for the listed firms and mutual funds. It will contribute to increased information disclosure on the part of listed firms and mutual funds. This information disclosure has the potential to solve the problem of adverse selection due to asymmetric information about the worst possible loss (Value at Risk), which investors with commitments (of certain payments in future) are facing. This study proposes the Variance-Covariance Method to be the best, using the volatility forecasts from valid GARCH type models (GARCH, GARCH-M, GJR), while accounting for the model risk. Variance-Covariance Method is compared with Historical Simulation and Monte Carlo Simulation for energy firms in Pakistan. Energy firms are selected from four sectors (Oil and Gas Exploration, Oil and Gas Marketing, Refinery, Power Generation and Distribution). Sum of Binary Loss Function and Quadratic Loss Function are used as evaluation criteria for comparing Value at Risk forecasts generated by these three methodologies. Supervisor:- Dr. Saud Ahmed Khan