Semidefinite positive association measures and portfolio theory

Within the investment process, several suitable reward and risk measures might be considered. In case of portfolio optimization, besides the standard deviation and return of particular sources of randomness, also some sort of dependency among them has to be taken into account. In this paper, we formulate a portfolio optimization problem in terms of semidefinite positive association measures, potentially accompanied with dimensionality reduction process. The empirical analysis proves, that generally the best results are provided, when Kendall's type association measure is considered.