Executive summary

On June 21, 2022, the Prudential Regulatory Authority (PRA) had published a consultation paper1 and a draft supervisory statement underscoring the expectations of the Bank of England on model risk management (MRM) at banks in the United Kingdom (UK). It also proposed a set of five principles to be followed for effective MRM.

The PRA is yet to publish the final supervisory statement following the consultation, which ended on October 21, 2022. The overarching framework with its five principles aims to address all stages of the model lifecycle. A simpler framework was proposed to make it easier for smaller2 firms to have a resilient MRM framework.

The long-awaited draft supervisory statement, in addition to bringing models such as accounting, or regulatory reporting, into the MRM ambit, also set unprecedented expectations from Boards and senior managements of banks on model risk oversight.

The proposed principles will complement existing MRM guidelines such as those for stress testing and algorithmic trading. It also lays special emphasis on oversight of new technologies such as artificial intelligence (AI) and machine learning (ML) in MRM.

The final supervisory statement will expectedly incorporate significant updates to the five principles compared with the draft. While already quite detailed, they could benefit from clarity and pragmatism on certain sub-principles.

It is imperative for banks to analyse, plan and act now as the 12-month window provided for it is inadequate.

This paper highlights the requirements, talks about expected updates to principles, and reviews how banks in the UK could benefit from the learnings and experiences of the global MRM landscape and establish effective and efficient MRM frameworks under the PRA regime.