103. Stress test helps to improve the bank’s understanding of the vulnerabilities that it faces under exceptional, but plausible events, and provide the bank with an indication of how much capital might be needed to absorb losses if such events occur, which supplements other risk management approaches and measures. These events can be financial, operational, legal, or relate to any other risk that may have an economic impact on the bank. The results derived from stress tests can also assist the bank in determining the appropriate appetite for different types of risks and in estimating the amount of capital that should be set aside to cover them.

104. Each bank is required to implement a stress testing framework to address both the needs of the ICAAP and broader risk management. Stress tests and the stress test outcome analysis should not be confined to the ICAAP. It should be designed to support decision-making across the bank as explained in this section.

105. Each bank should perform an in-depth review of its potential vulnerabilities, capturing all material risks on a bank-wide basis that result from its business model and operating environment in the context of adverse events, stressed macroeconomic (e.g. economic cycle risk), and financial conditions.

106. As part of the ICAAP exercise, each bank should carry out integrated, regular, rigorous, and forward-looking stress tests that are appropriate to the nature of the bank’s business model and major sources of risk. The frequency should be annually and more frequently, when necessary, depending on the individual circumstances.

107. The Central Bank may challenge the key assumptions and their continuing relevance to ensure that there is sufficient capital to withstand the impact of possible adverse events and/ or changes in market conditions.

Governance

108. The Board is responsible for the effective implementation of the stress tests framework through appropriate delegation to Senior Management and subject-matter experts across the bank. This framework should be supported by robust governance, processes, methods, and systems with associated policies and procedures approved by the Board. The Board is not only responsible for the stress testing policies, but also for oversight of the stress testing execution. It is also responsible for the potential measures to mitigate stress scenario outcomes and the key decisions and actions taken based on the stress testing results, such as the consideration of stress testing outcomes in strategy and capital planning.

109. The stakeholders involved in a particular stress testing framework depend on the type of stress tests. The scenario design, quantification of impact and the identification of mitigating actions will involve a range of subject matter experts across the bank. Stress test-related activities are not the sole responsibility of the team in charge of preparing the ICAAP or in charge of the stress testing programme. Rather, the execution of stress tests is a collective exercise, whereby numerous stakeholders contribute to the design, measurement, reporting and analysis of stress tests. Stakeholders should include Senior Management and the Board.

Types of stress test exercises

110. Each bank is required to establish several distinct forms of stress test exercises as described hereunder, however for the purposes of an ICAAP the minimum expectation is to conduct internal enterprise-wide stress tests and portfolio-level stress tests. Regulatory stress tests are not acceptable as the only form of internal stress tests:

111. The frequency of stress test exercises should depend on their type, scope, depth, and on the wider economic context. Each bank should execute enterprise-wide stress tests based on a set of scenarios regularly at least quarterly is recommend. Each bank should execute portfolio-level stress tests more frequently depending on the needs of risk management and the business functions. Market risk stress tests in particular may have to be performed more frequently.

112. The capital impact results of these stress tests should be analysed, compared, incorporated, and presented in the ICAAP. One or several internal enterprise-wide stress test outcomes should be explicitly incorporated in the capital planning, and presented accordingly in the ICAAP capital planning section. The results from all types of stress tests exercise should be employed by Senior Management and the Board to assess the suitability of the bank’s capital.

Scenarios

113. Stress test scenarios should be designed to capture the risks and potential losses appropriately, in coherence with the characteristics of each bank’s risk profile and portfolio. The scope of these scenarios should cover all the risks identified as part of the identification process documented in the ICAAP. At a minimum, the scope of risks should cover strategic risk, credit risk, market risk, counterparty risk, operational risk, liquidity risk, IRRBB, credit concentration risk, funding risk, reputational risk, and climate risk.

114. Stress scenarios should lead to a reliable measurement of loss under extreme but plausible events. Such scenarios are essential tool to support risk quantification in providing impact on Pillar 1 and Pillar 2. Consequently, the scenario design should be supported by a clear choice of risk factors and associated shocks. Several types of design are commonly employed:

115. Each bank is expected to continuously monitor and identify new threats, vulnerabilities and changes in the environment to assess whether its stress testing scenarios remain appropriate at least quarterly and, if not, adapt them to the new circumstances. The impact of the scenarios is expected to be updated regularly (e.g. quarterly). In case of any material changes, the bank is expected to assess their potential impact on its capital adequacy over the course of the year.

116. If the bank forecasts the increase of its capital base (e.g. capital issuances, rights issues, reduction in the equity, etc.) and the capital planning reflects the proposed changes, the bank must perform an additional stress scenario. In these additional stress scenarios the bank should analyse the impact under the assumption that the capital increase does not materialise. The impact analysis should include management actions and formal trigger points.

Methodology

117. The process of stress tests should be supported by robust and documented methodologies. All models employed in the quantification of stress results should comply with the requirements presented in the Central Bank Model Management Standards and Guidance.

118. For the measurement of capital under stress, each bank should employ a dedicated financial model to forecast their financial statements under several economic conditions. Such projection should be constructed over a minimum of three (3) years, in coherence with the most recent capital plan and with the Central Bank regulatory exercise.

119. Stress scenarios may be derived from stochastic models or historical events, and can be developed with varying degrees of precision, depth, and severity, particularly the impact on asset quality, profitability, and capital adequacy. Each bank should consider three (3) to five (5) scenarios (each scenario can have multiple severity levels (i.e. Low, medium, and high). Although it is expected to consider the supervisory stress tests (“stress test exercise of the Financial Stability Department (FSD)”) as one scenario, it is the bank’s responsibility to define scenarios and sensitivities in the manner that best addresses its situation and to translate them into risk, loss, and capital figures.

Use Test

120. Stress tests should support decision-making throughout the bank effectively. Stress tests should be embedded in banks’ business decision-making and risk management process at several levels of the organisation. Senior Management and the Board should lead and approve all assumptions, the methodology framework and authorise the use of stress test results.

121. Stress tests do not stop with the production of results. Risk mitigations should always be considered in light of the stress severity and likelihood. If no action is deemed necessary, this should be documented and clearly justified.

Reverse Stress Test

122. In addition to normal stress testing, each bank is expected to conduct reverse stress tests. Reverse stress tests start with the identification of a pre-defined outcome where the bank’s business model becomes non-viable (e.g. through insolvency), or it breaches supervisory compliance minima, e.g. by breaching minimum capital requirements (i.e. the bank will breach the regulatory capital buffer and minimum capital requirements). The next step is to assess which scenarios and shocks lead to that identified outcome. Finally, the objective is to assess whether the likelihood of occurrence is realistic and the impact warrants mitigating actions. If a bank considers mitigation strategies, e.g. hedging strategies, the bank should consider if such strategies would be viable. For example, a market that is stressed at a financial system level may be characterised by a lack of market liquidity and increased counterparty credit risk.

123. Effective reverse stress testing is a challenging exercise that requires the involvement of all material risk areas across the bank’s subject matter experts, Senior Management, and the Board.

124. Each bank should conduct a reverse stress test at least once a year. A well-designed reverse stress test should also include enough diagnostic information to allow the identification of the sources of potential failure. This enables proactive risk management actions and implementation of an appropriate strategy for refined risk monitoring, prevention, and mitigation. The reverse stress test requires each bank to consider scenarios beyond normal business forecasts and aids identification of events linked to contagion with potential systemic implications. Reverse stress testing has important quantitative and qualitative uses, through informing Senior Management of vulnerabilities, and supporting measures to avoid them. (Please refer to Appendix 3.3).