This work deals with a realistic portfolio allocation problem based on the Markowitz sharpe
optimization strategy. Many constraints were considered, aiming at taking into account realistic
conditions as maximum and minimum investments and chash outs, prot incomes and
minimum number of allocations. In order to nd local minimums of good quality, the heuristic
method Particle Swarm Optimization were applied to two formulations of the problem. Such
formulations were validated in a realistic portfolio.