Evaluating the quality of solutions in project portfolio selection

Abstract

Assessing the quality of decisions while selecting project portfolios becomes an inherent part of the decision-making process when the project parameters are inaccurate or uncertain. Small adjustments to the initial parameters can lead to situations where the preferred portfolio no longer reflects the investor’s requirements. The paper studies the post-optimal analysis of the Pareto optimal portfolios chosen by Savage’s risk criteria. Stability characteristic, such as the stability function, is considered. Using the stability function, we evaluate the quality of feasible portfolios. This function indicates the robustness of portfolios to any changes in the initial data. Using the stability function the formula for calculating the optimality threshold is obtained, which determines the level of risk reduction when the selected Pareto optimal portfolio can obtain optimal properties. The performances of the stability function and the optimality threshold are shown in the case study using global risk assessments for projects participating in the Belt and Road Initiative. The computation results demonstrate the ability through the stability function to evaluate the quality and optimal properties of feasible project portfolios.