Enhancing global portfolio optimization using genetic network programming

Financial risk has evolved from simple variability of returns in stock trading activities toward interconnected uncertainty factors in our economic systems. In this context, building global portfolios provides a natural mechanism to manage diversified risk between asset classes. This paper proposes a novel framework for the asset selection and allocation under global diversification principles using Genetic Network Program-ming(GNP) and Genetic Relation Algorithm(GRA). Asset classes such as stocks, bonds and currencies listed in relevant developed financial markets in USA, Europe and Asia are used. The comparison with conventional schemes in finance literature shows competitive advantages of the proposed approach.