Abstract:Mean variance optimization model uses mean to describe expected return and variance to measure risk, and looks for the best asset allocation among many assets. In the classical mean variance model, the model and parameters are supposed to be known. However, in finance, a model is an approximation of the reality, and moreover within a model, the estimation problem is a difficult issue. In order to solve this problem, we study the robust mean variance optimization model under model uncertainty taking into account the ambiguity about the drift (mean return) and the correlation between different assets.
