PSO-Based Tuning of Murame Parameters for Creditworthiness Evaluation of Italian SMEs

Abstract

In this work we use a MultiCriteria Decision Analysis (MCDA) model to evaluate the creditworthiness of a sample of Italian Small and Medium-sized Enterprises (SMEs), on the basis of their balance sheet data provided by the AIDA database. Our methodology is able to consider simultaneously different factors affecting the firms’ solvency level, and can produce results in terms of scoring, classification into homogeneous rating classes and migration probabilities. In this contribution we compare the results obtained considering two scenarios. On one hand, we experience an exogenous specification of the parameters that describe the preference structure implicit in the used MCDA model. On the other hand, we consider the results obtained using a preference disaggregation method to endogenously determine some of the model parameters. Because of the complexity of the obtained mathematical programming problem, we use an heuristic methodology, namely Particle Swarm Optimization (PSO), which provides a reasonable compromise between the quality of the solution and the computational burden.