This work shows the use of Value-at-Risk methodology, measuring market risk for Pension Funds fixed income investment funds, and the adoption of optimization methods for asset allocation. The first part presents different methodologies used to measure market risk (VaR). Among them, we can distinguish three approaches: parametric, Monte Carlo Simulation and Historical Simulation (called Full Simulation) - used to implement the models adopted in this work. The second part summarizes the most important concepts related to the Classic Theory of Portfolio Optimization, developed by Markowitz. Based on these concepts we have created different algorithms to be used in the active management - institution`s investment portfolio (in our case, Pension Funds). The last part focus in the results obtained and is concerned with their interpretation; in other words, it attempts to define how the institution could use this information to support its investment decision.