A PROMETHEE-based approach to portfolio selection problems

In this paper, we study the use of PROMETHEE outranking methods for portfolio selection problems. Starting from a new formulation of the PROMETHEE V method, we develop several alternative approaches based on the concepts of boundary portfolios and c-optimal portfolios. The proposed methods are compared in an extensive computational study. Results of these experiments show that methods based on the concept of c-optimal portfolios provide a good approximation to the (often computationally untractable) PROMETHEE ranking of all portfolios, while requiring only small computational effort even for large problems. For smaller problems, a PROMETHEE ranking of all boundary portfolios can be performed and provides a close approximation of the total ranking.