Implementing IFRS 9 by Adapting AIRB Models

Bogie Ozdemir

IFRS 9 implementation is a very significant effort for the banks globally. A key element of IFRS 9 is a forward-looking “expected credit loss” (ECL) impairment model, which represents a paradigm shift from the existing incurred loss model which is subject to a considerable degree of subjectivity and judgement. In this chapter, we will examine how to adapt the AIRB models in the estimation of ECL for IFRS 9 purposes. By leveraging the AIRB models, banks can lessen their modelling efforts significantly in estimating ECL under IFRS 9. We demonstrate the necessary model adaptations required and provide detailed examples.

Given the similarities between ECL and the credit risk measures under the AIRB, a pragmatic solution is for banks to leverage their well-established credit risk stress-testing models under AIRB to also satisfy the IFRS 9 modelling needs. This is what we will focus on this section. More specifically, we adapt the AIRB PD, LGD and EaD models for IFRS 9 use, and show how we can arrive at ECL by integrating the PD, LGD, and EaD parameters obtained from these models. To ensure the expected loss measure can reflect the current state of the economy and business cycle, the