Application of Gini coefficient and semivariance as estimators of risk in project selection

Abstract

This research has its origins in the growing need to understand and apply methodologies for project selection and ordering, framed within the efficient and effective utilization of productive resources of different organizations, which, being scarce, must be used in the best possible way. This led to the design and development of an optimization tool that allows to plan and manage a set of projects within an investment plan, so as to maximize the total profit generated by its implementation. Therefore, shown in detail the mathematical formulation of two theoretical models (Mean-Gini and Media-semivariance) used for the solution of the problem, obtaining an optimal portfolio. The first model in its objective function involves three levels of valuation: economic, financial and social, and uses the Gini coefficient as an estimator of portfolio risk, the second model is a bi-objective function, which is to maximize the net present value (VPN) portfolio and minimizing the risk of it, which usually are in conflict, as an optimization to the detriment of another, the estimate of risk in this model is the semivariance. With the aim of validating the models and methodological explanation, these were applied to a particular case adapted from a utility company in the region, on the basis of information provided by it, in terms of availability of capital, labor and machinery, and the same expectations about the maximum risk and the minimum return expected.INGENIERIA INDUSTRIALIngeniería y Competitividad, Volumen 13, No. 2, p. 61 - 81 (2011)Aplicación del coeficiente de Gini y la semivarianza como estimadores del riesgo en la selección de proyectosApplication of Gini coefficient and semivariance as estimators of risk in project selectionINDUSTRIAL ENGINEERING61