HKMA Guidance on Interest Rate Risk Management

Our Ref.: B1/15C 12 February 2025 The Chief Executive All Authorized Institutions Dear Sir/Madam, Interest rate risk management I am writing to provide further guidance for authorized institutions (AIs) to strengthen the management of interest rate risk in view of the changing operating environment1 . It also takes into account observations by the Hong Kong Monetary Authority (HKMA) in its recent review of AIs’ practices, especially with regard to disclosure of capital adequacy ratios (CARs) adjusted for unrealised losses on debt securities investment and use of behavioural models for measuring interest rate risk in the banking book (IRRBB), as set out below. (i) Disclosure of CARs adjusted for unrealised losses on debt securities investment Recognising the benefits of enhanced transparency at times of market stresses, AIs should closely monitor the unrealised losses on their investment in debt securities which are intended to be held to maturity (HTM debt securities), and should consider disclosing their CARs adjusted for these losses (adjusted CARs) when they are significant. In this connection, AIs are expected to put in place a robust system and internal controls to compute and monitor the unrealised losses on their HTM debt securities. If the unrealised losses exceed 10% of Common Equity Tier 1 capital amount, the AI concerned should notify and seek guidance from the HKMA as soon as reasonably practicable on its proposed timing and manner of the disclosure. In addition, it is crucial for AIs to communicate clearly with the public the intention of any disclosure with a view to facilitating a true and fair interpretation and avoiding unfounded market concerns. Practical guidance on the calculation of the adjusted CARs for disclosure is provided in Annex 1. It should be emphasised that following the above guidance alone does not discharge AIs’ responsibility for establishing and maintaining an effective framework for managing the risks associated with their debt securities investment. AIs should, amongst others, continue to incorporate into their internal processes the potential impact of incurring losses from selling HTM debt securities. For the avoidance of doubt, the above guidance pertains to additional disclosure and does not affect the implementation of the existing capital adequacy framework, including all the three pillars thereof.

1 This circular should be read in conjunction with the one titled “Lessons drawn on the Banking Turmoil in the US and Europe” issued by the HKMA in December 2023.

• 2 - (ii) Use of behavioural models for measuring the IRRBB Having reviewed AIs’ behavioural models with respect to their non-maturity deposits (NMDs), and following further discussion with the reviewed AIs, the HKMA considers that adequate model governance is the most important element contributing to robust measurements of IRRBB and warrants additional guidance. Specifically, AIs using behavioural models should put in place proper model governance framework comprising sound policies and procedures, documentation, independent model validation, ongoing monitoring, reporting on model outputs and performance, as well as senior management oversight. In particular, senior management of AIs should approve the behavioural models before adoption, any significant changes made to the models and overlays applied to the model outputs. On model validation, AIs should backtest the model outputs, including both the proportion of core deposits in NMDs and the corresponding behavioural maturity, before model implementation as well as at least once a year after implementation, against the actual outcomes. Given the use of behavioural models may significantly affect the IRRBB measurements, AIs should promptly notify the HKMA of any intention to start adopting such models or make major changes to the existing models. The HKMA also observed a number of good industry practices in its recent review, in areas including model governance and senior management oversight, modelling approaches, performance monitoring and validation, use of model outputs, third-party vendor or group-level models. Details are shared in Annex 2. AIs, whether currently using behavioural models for measuring IRRBB or planning to do so in the future, are expected to give due consideration to these good practices and strengthen their approaches as appropriate. AIs should review respective risk management framework and practices, and take any necessary steps to strengthen the measurement and management of interest rate risk having regard to the HKMA guidance and good industry practices. Should you have any questions, please feel free to contact us at bsdmarketrisk@hkma.gov.hk. Yours faithfully, Carmen Chu Executive Director (Banking Supervision) Encl.