Scoring Systems: Theoretical Aspects

Abstract

We embark here on one of the two core aspects of this book. In this chapter, we shall explain what rating systems are, why they are key elements in meeting the Basel 2 requirements for the IRB approaches, how to select an appropriate approach to building a rating model, what data to use, the common pitfalls to avoid, and how to validate the system. We concentrate here on a theoretical discussion. In Chapter 11, we shall illustrate this concretely with a case study. We shall then discuss how the scoring model can be implemented and articulated in a bank organization. We concentrate on a rating model for SME and corporate portfolios, although the basic principles can be applied to others. Rather than remaining at the level of general description and mathematical formulas, we shall try to give to readers clear examples on how each step can be implemented, using the real-life datasets that are given on the accompanying website. Our goal is that, having read this chapter, the interested reader will be able to begin their own research even without being “quants” (quantitative specialists). Succeeding in setting up and implementing an efficient rating system is not a matter of applying cutting-edge statistical techniques; it relies more on qualities such as a good critical sense, a minimum knowledge of what financial analysis is, a capacity to lead changes in an environment that is most likely to be (at least initially) hostile to the project, and finally a capacity to be sensible.