AERB: developing AIRB PIT–TTC PD models using external ratings

ABSTRACT The use of credit rating agency (CRA) ratings in a credit institution's lending practices has been directly criticized since the financial crisis, as these ratings and observed risks have diverged. Some regulators do not allow the use of external ratings as direct inputs into a credit institution's internal probability of default (PD)/ratings model; hence, the capital planning process, as regulators generally prefer the use of internal assessments. However, regulators allow for the use of a CRA's long-run average default rates to benchmark an institution's internalPDmodel output.Arecent study we conducted shows, however, that the indiscriminate use of such benchmarks can introduce significant biases in credit lending. Even with these criticisms and new regulatory constraints, one can still make use of the rich CRA data available to create regulatorycompliant PD models. In this paper, we propose a class of agency replication-style models that make use of obligor information and CRA long-term DR information. Such models are extremely useful in cases where a credit institution has limited default samples, so a purely internal default-based model could be potentially erroneous, and where, in contrast, agencies have plenty of data supporting the development of robust models. In this paper, we show how one can use this class of models for modeling portfolios such as large corporates, banks and insurance companies. We discuss our experience developing approved, advanced internal ratings-based (AIRB) models, which we augment with external default data and, hence, colloquially call advanced external ratings-based (AERB) models. We show various simplifications of the formulation and show how they can be used in point-in-time/through-the-cycle-based credit rating systems. "AERB" is an acronym we started utilizing informally while developing PD models for regulatory approval under Basel II waivers. In contrast to AIRB models, AERB models are used when regulators, in our recent experience, require broader, external, long-run default calibrations to complement internal default data when internal default data is limited.