A multi-criteria approach in a financial portfolio selection framework

We proposed a portfolio composition method within a multicriteria framework. The procedure is based on a cluster analysis of the time-varying betas, estimated by using P-spline, so that each resulting group contains different level of systemic risk. Then we computed some risk adjusted performance measures for the stocks of the recognized clusters. Finally, we use these indexes as input of the ELECTRE III method to obtain a stock ranking useful for the asset selection phase