Harmonisation of the definition of default
EBA final guidance
Discover the key points and implications of the final report on the application of the definition of default!
Regulatory Newsflash | 11 October 2016
On 28 September, the EBA published the final report on the application of the definition of default under Article 178 of the Capital Requirements Regulation (CRR). A separate report was published for the results of the quantitative impact study (QIS).
The report follows a three-month consultation on the guidelines during which stakeholders raised a number of concerns associated with a stricter default definition, including potential conflicts with the requirement for a ‘use test’. While the EBA clarified certain aspects of the guidelines in light of those comments, it said it recognised the significant implementation challenge for firms, particularly for those that are on an IRB Approach, and set an application date of January 2021. The EBA also quantified the potential impact of the guidelines on firms’ capital requirements. It found that overall capital requirements would not significantly increase as a result of the proposed changes. The impact varies across firms and is generally greater for firms that are on an IRB Approach.
Changes to the definition of default
The guidelines provide detailed clarification of the definition of default and its application, including (but not limited to) the following aspects:
Past due criterion
Additional guidance is provided on how to apply the materiality threshold set out in the RTS on the materiality threshold. The guidelines also list a range of situations in which recognition of default arise from ‘technical’ – not credit-related – issues and where the exposures should not be treated as defaulted.
Technical defaults are broadly defined as:
- Data, system error or manual error
- Allocation of payment in the bank’s systems took too long
Application of the default definition for retail exposures
The guidelines clarify the scope of application for facility-level defaults for retail exposures, and clarify that banks should have processes to review situations where multiple facility-level defaults should give rise to customer-level default for retail customers.
The guidelines describe a ‘pulling effect’ (consistent with the ITS on reporting) for retail exposures with a threshold at 20% (i.e. when more than 20% of all exposures to retail obligors are in default, all exposures should be regarded as defaulted).
Identification of unlikeliness to pay
Clarification is provided for each indicator of unlikeness to pay:
- Specific Credit Risk Adjustments. The guidelines propose that all exposures classified as Stage 3 under IFRS9 (i.e. credit-impaired exposures) should be treated as defaulted with a few specified exceptions. Exposures classified in Stage 2 under IFRS 9 should not be regarded as defaulted unless other indications of unlikeliness to pay exist.
- Material Credit-related Loss. The guidelines clarify that sales of loans for reasons other than concern over credit-worthiness do not need to be captured as defaults. It also sets a threshold of 5% as the level of loss that is “material” when the loss does arise from sales driven by credit-related issues.
- Distressed Restructuring. The guidelines confirm the link between forbearance and default as a result of distressed restructuring, where there is a material impact on NPV. The materiality threshold is to be set by institutions, but should be subject to a cap of 1% which means that in practice only rounding differences are excluded.
- Bankruptcy. The guidelines propose to harmonise the concept of bankruptcy, making it clear that the concept includes all versions of insolvency proceedings listed in the Insolvency Regulation.
Return to non-defaulted status
The guidelines set out how to deal with exposures that return to non-defaulted status (also called cures):
- The guidelines specify that banks should have an assessment process before reclassification of a defaulted exposure to non-defaulted status.
- The guidelines set out a minimum probation period of 3 months (1 year for distressed restructuring) – once defaulted, the exposure cannot return to non-defaulted status until the probation period has expired.
Application of the definition of default in external data
The guidelines clarify that the requirements with regard to external data apply only to firms that use an IRB Approach and which use such data for the purposes of the estimation of IRB risk parameters.
The guidelines propose that banks maintain documentary evidence of the definition of default and how it is implemented in processes, systems and models. In addition banks are required to keep a register of all current and past versions of the default definition. If more than one definition of default is in use at any time, the scope of application of each definition should be clearly documented.
The guidelines note that the definition of default used in an IRB bank should be approved by the management body (or a sub-committee thereof) and IRB banks should use the definition of default consistently for the purpose of capital calculation and risk management processes – at least in the areas of monitoring of exposures and internal reporting to senior management and the management body.
The EBA recognises that significant efforts and resources will be required by firms to implement the harmonised default definition. The costs for firms will depend to a great extent on whether a firm is on the IRB Approach or the Standardised Approach.
Firms that are on the IRB Approach will be disproportionately affected as they have to adjust to the new definition of default by applying the following:
- Adjust the historical data used for the development and calibration of IRB models based on the new definition of default;
- Assess the materiality of impact on all risk parameters and capital requirements of the new definition of default, and compare to the previous definition after the relevant adjustments to historical data have been made;
- Include an additional margin of conservatism in their rating systems in order to account for the possible distortions resulting from the inconsistent definition of default in the historical data, until such time as the data series uses a consistent default definition throughout.
Firms that are on the Standardised Approach are expected to be less affected, as there will be no need to adjust historical data unless the firms decide to migrate to an IRB Approach. That said, depending on the profile of each firm, the costs of implementing the changes for firms on the Standardised Approach might still be significant and changes to processes and possibly IT systems will be required.
Impact on capital requirements
The EBA’s analysis found that bank capital requirements would not significantly increase as a result of harmonising the definition of default, although it observed a broad dispersion of the potential impact of some of the policy proposals across firms.
The changes would lead to a “modest increase” in capital charges, corresponding to an overall reduction in the average CET1 ratio of around 20 basis points for firms that are on an IRB Approach, and only a very limited decrease in CET1 ratio for firms on the Standardised Approach.
The main difference in the impact between the IRB Approach and the Standardised Approach stems from required changes in the materiality thresholds. Firms that have comparably higher materiality thresholds currently in place will need to apply lower materiality thresholds, which will in turn lead to higher defaulted exposures.
The final report notes that the biggest costs for firms are associated with redevelopment of internal models, rather than with higher capital requirements. The requirement to adjust the historical data based on the new default definition will present a particular challenge for IRB firms. The EBA said firms should apply an additional margin of conservatism in their ratings systems in order to account for possible distortions that may arise from the inconsistent default definition in the historical data. IRB firms will also have to receive approval from their supervisors for the changes to their internal models, so the guidelines represent a potentially significant additional workload for National Competent Authorities (NCAs).
The EBA appears to have struck a balance between the concerns raised by the industry regarding the operational burden of implementing the required changes and a sufficiently long time frame for banks to adjust their approaches to the new requirement. Specifically, clarity on the application date of January 2021 is helpful as will enable banks to also include forthcoming changes in the IRB framework introduced by the Basel Committee in their work to implement the EBA guidelines. Banks may also wish to assess overlaps between this work and their IFRS9 programmes. We note that as a result of the guidance nothing has changed in the EBA’s stance in favour of the continued use of the IRB Approach.
The guidelines will apply from January 2021, but the EBA encourages firms to begin preparing for compliance with the new default definition before then. The guidelines also allow NCAs to accelerate the implementation timeline.
For further information with respect to this subject, please contact:
Bernard De Meulemeester
Partner – Audit
Tel: + 32 2 800 24 64
Arno De Groote
Partner – Risk Advisory
Tel: + 32 2 800 24 73
Partner – Audit
Tel: + 32 2 800 21 42
Partner – Audit
Tel: + 32 2 800 20 45