Designing a Multicriteria Decision Support System for Portfolio Selection and Management

Abstract

This paper presents a multicriteria decision support system for selection and management of stocks portfolio. Firstly the system allows a detailed financial and stock market analyses (i.e. common-size statements, financial ratios, ratios of the stock market) and to compute returns and betas of the corresponding stocks from theoretical models (Market Model and Capital Asset Pricing Model-CAPM). Secondly it gives graphs of the evolution of some of the financial and stock market variables. Then, univariate (correlation coefficient r, Kendall’s rank correlation coefficient τ, etc.) and multivariate (Principal Component Analysis) statistical techniques are utilized to aid the analysis of the beta’s stability and the identification of the most significant financial and stock market ratios for the portfolio selection. Finally, both the most important financial/stock market ratios and some qualitative criteria are integrated into two multicriteria decision making methods allowing first to select some attractive stocks and then to determine their proportions in the proposed portfolio. The capabilities of the system are illustrated with recent data provided by the Belgian Stock Exchange.