We investigate the optimal portfolio selection problem incorporating fundamentals into higher order moments on a new approach to a portfolio of assets. The returns frequent skewed and excess kurtosis behavior along with investors’ preferences set a new basis of discussion. The higher order moments than the kurtosis will add extended information on investors, than the standard model. Hence a more analytical problem arises, of higher flexibility, non-convexity, in unlimited scale fitted into financial management. We discuss the model introducing three hybrid neuro-genetic models of numerous topologies and one regression. Firstly the Radial Basis Function Networks are thoroughly detected in 40 various hybrid formations and 10 RBF Neural Nets whilst results are compared to past results of 50 similar topologies of Time-Lag Recurrent Network Hybrids respectively, 10 topologies on the MLP Neural Nets, and from econometrics the Bayesian Logistic Regression, to define highly competitive methods in asset allocation and corporate evaluation. Novel solutions are offered under specific hybrids whilst acceptance is either evolutionary or intelligent.