A Portfolio Selection Methodology Based on Data Envelopment Analysis

This paper presents a four-step methodology based on Data Envelopment Analysis for portfolio selection of decision-making units (DMUs) which can be stocks or other financial assets. Along the steps of the methodology, DMUs efficiency ratios are first computed, and then, the generation of a portfolio is carried out by a mathematical model which optimizes the weighted sum of the DMUs' efficiency ratios included in this portfolio, which is optimal for the Decision Maker's system of preferences. The methodology is illustrated through a portfolio of stocks where the results are compared to those of Ben Abdelaziz et al. (2007) model which has used the same example. The proposed methodology was capable of reproducing similar results with the model of Ben Abdelaziz et al. (2007) while being easier to apply and to understand for practitioners and decision makers.