The portfolio selection process for projects involving the exploration and production of oil is a complex task, passing through different steps until and optimum solution is achieved, revealing the best portfolio. This paper presents a methodology for portfolio selection considering semi-standard deviation as a measure of risk. The methodology is applied to a set of six oil production projects located in the Campos Basin, where the risk minimization is subject to a certain return. Genetic algorithms are used as the optimization tool.. The portfolio selection was performed by maximization of the certainty equivalent, for different values of risk aversion coefficients. The results show that the semi-standard deviation presents better sensitivity in portfolio selection with projects that have a large magnitude of risk and return.
Semi-standard deviation; portfolio theory; preference theory; petroleum
