The purpose of this seminar is to introduce you to the key methodologies to design, develop, calibrate and validate credit rating systems for corporate customers.
We start with an overview and discussion of the three main types of credit rating systems: the Early Warning systems, the Long-term Corporate (issuer) Ratings and 'Master Scale'-based Rating systems. Particular focus is put on the uses and misuses of each of the three system types, including their applicability to meet regulatory requirements, such as Basel III or IFRS 9, and their appropriateness to address business-related objectives, such as risk-adjusted pricing or operational risk management.
We then take a closer look at the 'Master Scale'-based Rating systems. 'Master scales' allocate a non-overlapping range of probabilities of default (PD) that are stable over time to each rating class. The rating methods for such systems need to produce accurate projections of the 1-year PD based on actually observed defaults. Starting from 'simpler' questions, such as what constitutes a default of a corporate customer, how to handle groups of legally or economically related entities from a data management perspective or how to build and maintain an appropriate historic record of defaults, we gradually dig into the core quantitative modelling methodology. Covered topics include statistical analysis of Corporate Balance Sheet KPIs, design and development of Integrated Rating Models based on quantitative factors and qualitative assessments, and model validation techniques. Throughout this part of the course we give practical advice and examples related to common challenges such as low default portfolios, missing / incomplete data and input data outliers.
After this, we turn to the Early Warning systems, which help the bank to identify reliably upcoming defaults or substantial credit risk increases of specific corporate customers on an ongoing basis. Customer account behaviour variables and expert opinion play a critically important role in the risk differentiation mechanics within such system, making the risk assessment a fully dynamic process. Building upon that, we present a more comprehensive framework to assess the impact of observable factors, such as market prices or macroeconomic indices, on the corporate customer's credit risk in a forward-looking manner.
Finally, we look at the Long-term Corporate (issuer) Ratings which express risk in relative rank order (i.e. they are ordinal measures of credit risk) and are not predictive of a specific frequency of default. The rating model development in this case needs to start from a specific industrial sector and only thereafter to combine the multiple specific models unto a unified rating scale. We briefly present an example of a Long-term Corporate Rating model within the healthcare sector and illustrate the process of mapping the model's results to a generic rating scale, such as S&P or Moody's.
Throughout the entire course, we revisit the topic of IFRS 9 multiple times, highlighting the best practices in applying the different types rating systems to comply with this new regulation. Covered topics include the development of IFRS 9 staging criteria, the estimation of lifetime PD and the calculation of impairment provisions.
