IFRS 9 Model Risk

From Open Risk Manual

Definition

IFRS 9 Model Risk denotes in a broad sense the potential for error in the development and implementation of any of the quantitative elements underpinning the IFRS 9 accounting. This includes Significant Increase in Credit Risk and Expected Credit Loss models, and/or the application, interpretation and reporting of model results.

In turn IFRS 9 model errors can lead to a variety financial and / or reputational loss events. Undesirable consequences of this from a financial stability perspective may include investors’ reactions in the form of distrust in the accounting figures, gaining distorted images of the relative strength of banks in the system, and having a false sense of security similar to the one witnessed in the pre-crisis days, when banks fed their IRB formulas with over-optimistic internal estimates of PDs and LGDs. [1]

Context

The expected loss approach to impairment allowances under IFRS 9 requires a substantial modelling effort by the reporting entities and a corresponding effort by users, auditors and regulators to assess the quality of the models and interpret their outcomes (their comparability across entities, their volatility over time, etc.).

IFRS 9 entails a large degree of sophistication (e.g. when requesting expected losses to be computed as an average across several macroeconomic scenarios). This need poses challenges related to the lack of data or experience relevant to the required modelling as well as the role of managerial judgement and discretion in the modelling process.

Complexity, lack of experience, and managerial discretion in the modelling process leave room for investors and other users of the Financial Statements to receive richer, but at the same time noisier and potentially easier-to-manipulate, pieces of information.

References

  1. ESRB, Financial stability implications of IFRS 9, July 2017