The FBF welcomes the opportunity to share its comments on the EBA’s discussion paper on institution’s stress testing.
The FBF reiterates its support for a stable and resilient global financial system, while facilitating economic growth. To this end, while supporting the EBA’s initiative on institution’s stress testing, we believe that the proposed discussion paper (EBA-DP-2017-17) raises some concerns and requires some clarification.
Summary of key comments:
The application date of these Guidelines should be reconsidered in order to adopt a staged approach.
These Guidelines are very demanding for reverse stress tests in particular, second round/feedback effects or the interaction between solvency and liquidity shocks. It should be considered that internationally active institutions do not always have access to sufficiently granular data in order to perform stress tests at the level of business units in specific geographic regions, business sectors or business lines. The development of IT infrastructure is a long term and on-going process.
Requirements should be applicable first to the group, then to business units in specific regions and last but not least, to business sectors and business lines.
Third-country legal restrictions on bank secrecy should be considered.
If it is possible to assess the “overall severity” to complement the range of stress tests scenarios, it remains difficult to identify “plausible scenarios” for recovery purposes due to their extreme character. At least, we ask for the deletion of “plausible scenario” applicable for recovery stress tests and a clearer definition for “plausible scenario” applicable to institutions’ stress test.
Institutions should have enough time to adapt their functioning and to assess interconnections between risk types in a constructive process built with their competent authority.
Please find below our main comments and our detailed feedback to the EBA discussion paper.
2. Subject matter, scope and definitions
Paragraph n°10: The EBA should consider in the definition of the “date of application”, the complexity of large and internationally active banks, in order to adopt a staged approach. Indeed, these Guidelines are very demanding in particular for reverse stress tests (para. 84 to 101), second round or feedback effects (para. 78) or the interaction between solvency and liquidity shocks. A staged approach would offer more flexibility both for the industry and for competent authorities, in order to assess the implementation of the new Guidelines. The increasing interactions between solvency stress tests and liquidity stress tests require changes in institution processes and more collaboration with supervisors. Institutions need more guidance in order to revise their organisation and their processes.
4. Institution’s Stress Testing
1) Stress testing programme
Paragraph n°15: It should be clarified what scenarios the Guidelines are referring to by “respective scenarios”.
2) Governance aspects of stress testing
Paragraph n°25: We wonder why “external consultants responsible for stress testing” are mentioned in this paragraph. Whether institutions’ stress tests are conducted by internal teams or external consultants is irrelevant with regard to the subject of these Guidelines.
Paragraph n°31: The wording of this paragraph should be revised in order to better distinguish the expectations in terms of inputs and outputs of stress tests. It should be noted that the materiality in terms of current and potential risk is defined under the internal evaluation process of institutions’ risk appetite and limits. Subsequently, as part of the stress test exercise, the institution defines inputs in order to assess dedicated risks. Finally, the output of stress tests is the assessment of the impact (losses, capital and liquidity requirements).
3) Data infrastructure
Generally speaking, data infrastructure needs time, transitional periods and a comprehensive assessment by competent authorities.
4) Stress testing scope and coverage
Paragraph n°51: This paragraph is unclear. We kindly request more details and examples.
Paragraph n°54: The requirements of this article remain unclear and very ambitious. We request more details and examples on what is expected. In particular, the following phrase “(b) correlations, offsetting of individual exposures and concentrations may either lead to double counting of risks or to an underestimation of the impact of stressed risk factors” should be removed from this paragraph and developed in a new paragraph.
Paragraph n°55: A list of priorities should be given as the application of a stress testing programme to business lines and sectors requires an overhaul in information systems. In particular, it should be considered that, as of now, internationally active institutions and groups do not always have access to sufficiently granular data in order to perform stress tests at the level of business units in specific geographic regions, business sectors or business lines. The development of IT infrastructure is a long term and on-going process, strengthened by the recovery requirements which remain “under construction”. We request a comprehensive evaluation by competent authorities. At least, requirements should be applicable first to the group, then to business units in specific regions and last but not least, to business sectors and business lines. Third-country legal restrictions on bank secrecy should also be considered. The combination of requirements defined in paragraphs 21 and 55 should be measured, in particular for business units out of the scope of recovery and institutions’ stress tests as this may entail a high degree of complexity. For this reason, the responsibilities of business units exempted from stress tests, as defined in paragraph 21, should focus on modelling assumptions and not on processes.
6) Stress testing types
i. General requirements
Paragraph n°64: We note the introduction of Pre-Provision Net Revenue models (PPNR) while this element had not been a European requirement so far, but a US Comprehensive Capital Analysis and Review (CCAR) purpose. If expected by the regulators, such model should target aggregate according to European accounting framework (IFRS or Finrep like definitions) to ensure a consistent requirement across jurisdictions.
Paragraph n°64 and paragraph n°76: The relationship between the requirements of paragraph n°64 and paragraph n°76 requires some clarification due to a possible inconsistency. Paragraph n°64 promotes more interactions between liquidity and solvability shocks, whereas paragraph n°76 acknowledges that solvency and liquidity stress testing require different time horizons and scenarios. More detailed explanations are required on how to accommodate the requirements of articles 64 and 76, in particular on how shocks are spread across time.
ii. Sensitivity analysis
iii. Scenario analysis
Paragraph n°75(a): The definition of “material risk factor” should not be left unstressed or unconsidered. Should this “material risk factor” be identified under the internal risk cartography (with a materiality threshold, etc.)?
iv. Severity of scenarios
v. Reverse stress testing
Paragraph n°85: If it is possible to assess the “overall severity” to complement the range of stress tests scenarios, it remains difficult to identify “plausible scenarios” for recovery purposes due to their extreme character. At least, we ask for the deletion of “plausible scenario” applicable for recovery stress tests and a clearer definition for “plausible scenario” applicable to institutions’ stress test.
Moreover, this paragraph should distinguish between ICAAP reverse stress testing and Recovery Plan stress testing.
Paragraph n°97: This paragraph introduces the ‘near default’ scenarios for recovery planning purposes. We would welcome any detail about the rationale underpinning the introduction of this wording. We would appreciate if EBA could explicitly link the near default concept with the BRRD framework.
Paragraph n°98: The EBA should clarify its expectations with regard the relationships between the stress tests for ICAAP and recovery planning instead of stating that the two are not interlinked.
7) Individual risk areas
i. Credit and counterparty risk
iii. Market risk
Paragraph n°127: Clarification is required on what is requested regarding reserves related to portfolios (e.g. reserves for liquidity, for modelling uncertainties). Market risk reserves stress testing should be substantiated.
iv. Operational risk
v. Conduct related risk and associated litigation costs
Paragraph n°146: The paragraph should be revised as followed : “When an institution is unable to provide an estimate for an individual conduct related risk due to the extent of uncertainty, the institution should provide its assessment.” It remains impossible in many cases to provide an estimate for an individual conduct related risk. Therefore, it is not a rare case.
vi. Liquidity risk
Paragraph n°158: This paragraph should be clarified and a phase-in period should be defined in order to allow institutions to adapt their functioning and to assess interconnections between risk types in a constructive process built with their competent authority. Indeed, taking into account credit losses in the liquidity stress test scenario design is a complicated task.
An evaluation of best practices could be disclosed, in order to improve the assessment of stress events for “other risk types”. Also, a list of key indicators could be identified for this transversal requirement.
vii. Interest rate risk from non-trading activities
This part should be more developed, in order to better frame credit institutions in their institution’s stress tests on IRRBB. In particular, more guidance are expected about the requirements concerning NII and EVE measurement, each of the aiming at different stress testing framework.
viii. Concentration risk
Paragraphs n°171-178: We understand the need for identification of intra-risk type and inter-risk type but the two sets of risks types should be better distinguished. Moreover, it should be confirmed that intra-risk type is banned. However, inter-risk type should be authorised, once it is documented by credit institutions. This clarification is necessary for credit institution, in order to develop performing internal models.
ix. Foreign exchange lending risk
8) Application of stress testing programmes
i. Stress testing for ICAAP/ILAAP purposes
Paragraph n°187: Institutions request more time in order to improve their forward-looking view in risk management, strategic planning, capital planning and liquidity planning as part of the internal capital adequacy assessment process (ICAAP). This is an ongoing action that needs a constructive and long-term discussion between institutions and their competent authorities. If we support more flexibility, it should be clear that safeguard should be granted in order to prevent the level playing field.
Paragraph n°189(a) and paragraph 47: Institutions request a definition/clarification of “relevant structured entities” and “material risk categories”.
ii. Management actions
Paragraph n°194 and paragraph n°196: It should be clarified what a “credible management action” is. Do you consider “credible actions” as those defined in the recovery action plan? As a consequence, shall we use the same list of actions for recovery stress test and institution’s stress test?