Investment decisions and portfolios classificationbased on robust methods of estimation

In the process of assets selection and their allocation to the investment portfolio the most important factor issue thing is the accurate evaluation of the volatility of the return rate. In order to achieve stable and accurate estimates of parameters for contaminated multivariate normal distributions the robust estimators are required. In this paper we used some of the robust estimators to selection the optimal investment portfolios. The main goal of this paper was the comparative analysis of generated investment portfolios with respect to chosen robust estimation methods.