Preparing data functions for the 2017 stress tests

In November 2016, the Bank of England (BoE) published the results of its 2016 banking stress tests which measured the resilience of UK’s major banks’ balance sheet in adverse scenarios. These incorporated a synchronised UK and global recession with associated shocks to financial market prices, and an independent stress of misconduct costs. The stress tests also represented the BoE’s first annual cyclical scenario (ACS), a new approach to stress testing, which examines the resilience of the system to a more severe stress than in previous years.

In 2017, the BoE is expected to extend stress testing even further by including a biennial exploratory scenario which will test the resilience of banks to risks that may not be directly linked to the financial cycle. At a UK level, the 2017 stress test scenario also includes a severe level of stress, with substantial impact on UK residential and commercial property, UK GDP and unemployment. However, the impact could be even more severe if the economic and political challenges currently facing the EU and Eurozone were to be incorporated, such as high-debt levels, security concerns and Brexit.

The introduction of an ACS and the exploratory scenario demonstrates the BoE’s ongoing effort to ensure that stress tests become more rigorous, complex and broader in scope. Banks in the UK need to demonstrate that they have sufficient liquidity and capital positions in place to withstand major disruptions and understand and capture the complexities of these risks and their inter-dependencies across the global economy.

In order to prepare for the upcoming stress tests, banks should ensure they have an effective analysis of balance sheet vulnerabilities in place and adopt an integrated approach to their risk management functions. An important part of this is to build optimal granular data infrastructure and risk data aggregation as per BCBS 239 requirements. These principles, which were published by the Basel Committee on Banking Supervision, came into force in January 2016 and aim to increase the ability of global systemically important banks to aggregate and report risk data and cover a range of topics including data architecture and IT infrastructure, the completeness of data, as well as the timeliness and frequency of reporting.

In addition to the core data, the BoE will continue to make scenario-specific data requests as appropriate and the requested information will give the central bank the flexibility to gain deeper insight into the way banks have taken account of specific features of such scenarios in their projections. This will allow the BoE to examine areas of balance sheets that are likely to be affected in a given scenario. Consequently, the provision of core data will need to become more automated over time and participants will have more time to provide scenario-specific data.

The BoE is not expected to release details about its biennial exploratory scenario before spring this year; however, banks need to ensure that they have appropriate resources in place to run two or even more scenarios simultaneously. In order to do so, they should start identifying key data gaps and limitations that could be exposed through these new scenarios.

Questions remain as to whether banks are prepared to fully capture risks or whether the established BoE stress test templates, such as the firm data submission framework (FDSF), are the most effective method to determine balance sheet vulnerabilities.

The Prudential Regulation Authority has developed FDSF to provide quantitative, forward-looking assessments of the capital adequacy of the UK banking system and individual institutions within it. FDSF is likely to be managed by risk teams, who have up until now focused on internal reporting and who will now need to upgrade their technical infrastructure to comply with external disclosure requirements. In order to do so, they will need robust calculation engines to run the stress tests and the ability to reconcile the results with data reported as part of other regulatory requirements.

Finally, banks need to ensure that they have the flexibility to model their own scenarios, not just comply with regulatory requirements, but also as an internal capital management exercise. It is imperative that calculations, regulatory and internal reports are all consistent and banks have to be able to trace back results and calculations all the way from regulatory returns to originations, with the ability to view the entire data lineage.

In short, in order to be prepared for this year’s stress tests, over and above the methodology, design and implementation of models, banks need to ensure that they have the appropriate controls and a comprehensive governance process.

Sufyan Khan, 

Product Manager, EMEA, 

AxiomSL