Ranking portfolios from multiple articulated risk measures: an uncertainty decision approach

This paper proposes a performance ranking of efficient portfolios when risk is simultaneously evaluated by a set of measures articulated in an uncertainty decision model. After determining the mean-variance (E-V) efficient frontier of portfolios as a previous classic filter of selection, we introduce a new second stage in which the portfolio ranking is achieved by an uncertainty decision table designed for moderately pessimistic investors. In this table, the states of the world are defined in terms of the different risk measures so that the assessment of each portfolio row results from the articulation of composite profitability-safety performance indices depending on the states. Properties underpinning the rationale of the method are shown. A numerical application to a real world case (portfolio ranking of heuristics used by international brokers) is finally developed.