A New Mathematical Model Based Approach to Multi-Criteria Stock Selection Problem

We propose stock selection models that guide the decision maker (DM) by integrating multiple financial evaluation criteria into a linear programming framework. We provide a robust framework for stock selection and offer practical insights for investment decision-making. The models aim to achieve the most coherent sequence of alternatives (stocks) with a DM s ranking by maximizing Kendall s Tau score. We assume an underlying value function that represents the DM s preferences. We then develop two mixed integer linear programming models; the first model assumes an underlying linear value function, while the second assumes a Tchebycheff value function. We conduct experiments on stocks listed in the Standard and Poor s 500 index and compare the performance of our models against three benchmark methods from the literature, using various performance metrics. The results demonstrate that our models achieve high-quality solutions, outperforming benchmark methods in terms of ranking coherence and overall performance.