摘要:The main objective of Markowitz work is seeking optimal allocation of wealth on a defined number of assets while minimizing risk and maximizing returns of expected portfolio. At the beginning, proposed models in this issue are resolved basing on quadratic programming. Unfortunately, the real state of financial markets makes these problems too complex. Metaheuristics are stochastic methods which aim to solve a large panel of NPhard problems without intervention of users. These methods are inspired from analogies with other fields such as physics, genetics, or ethologic. Already various Metaheuristics approaches have been proposed to solve asset allocation and portfolio optimization problems. In a first time, we survey some approaches on the topic, by categorizing them, describing results and involved techniques. Second part of this paper aims providing a good guide to the application of Metaheuristics to portfolio optimization and asset allocation problems