Integrating analysts’ forecasts in the security screening process: empirical evidence from the Eurostoxx 50

Abstract

In this article we attempt to expand the limited framework of single-objective optimization and broaden Markowitz's market standard, within which the portfolio selection problem is conventionally confronted. The typical theory's fundamental principle is that investment decisions are generally made using two criteria, corresponding to the first two moments of return distributions, namely the portfolio's expected return and variance. One heavy criticism over this pattern, which has often been addressed by both practitioners and academics, is that it fails to incorporate the whole spectrum of investors’ criteria, thus realistically express real-world investment policy statements. The proposed approach constitutes an innovative methodological framework for dealing with the security selection or screening phase, one of the most crucial stages in the portfolio management process. Our aim is to assist portfolio managers in formulating successful investment strategies, by providing them with an effective decision-making tool in order to obtain the so-called approved or authorized lists of most attractive stocks. More precisely, we exploit the decision technology of linguistic variables and we apply a state-of-the-art linguistic method. The validity of the proposed approach is finally tested through an illustrative application in one of the most popular European stock market indices, the Eurostoxx 50. The results obtained are characterized as very encouraging, since an adequate number of efficient portfolios produced by the model, appear to possess superior out-of-sample returns with respect to the underlying benchmark.