Coventry Building Society – Written evidence (SCG0054)
Introduction
Coventry Building Society welcomes the opportunity to respond to the Committee’s inquiry into the Financial Conduct Authority (FCA) and Prudential Regulation Authority’s (PRA) secondary objectives and to comment on how the regulatory landscape can better support UK growth and competitiveness.
This response supplements the answers given by our Chief Executive, Steve Hughes, and others from the sector to the Enquiry session on 11 December 2024, providing answers to the additional questions and information discussed around capital regulations and the risk weighting of mortgages and the competitive impact on the large Building Societies and similar ‘large, strong and simple’ firms.
In November 2013 the Chancellor of the Exchequer asked the Financial Policy Committee to look at introducing a leverage ratio requirement on UK firms, consistent with the emerging intentions from the Basel Committee on Banking Supervision for globally significant firms. This was intended to act as a backstop to both IRB and the Standardised risk-based approach to capital requirements.
Current PRA Policy Statement (PS21/21) requires firms with >£50bn of deposits (or >£10bn of non-UK assets) to maintain a minimum leverage ratio of 3.25%, a countercyclical leverage buffer (currently 0.70%) and an additional leverage buffer for systemic importance of between 0.35% and 0.70%.
Smaller firms have a supervisory expectation that their leverage ratio will not fall below 3.25% - but are generally assessed by investors and others on their leverage ratios relative to large firms.
For most UK Banks this leverage ‘backstop’ is not binding as they are capital constrained on a risk weighted basis, needing to maintain a minimum Common Equity Tier 1 Ratio (CET1) which already generates a higher leverage ratio than needed to comply with the leverage policy above.
From published accounts and the 2023 PRA stress test we can see that Nationwide and parts of HSBC UK are also Leverage constrained. We understand Nationwide have issued Core Capital Deferred shares and Additional Tier 1 Capital during the last decade in part to meet these requirements.
Coventry Building Society also issued Additional Tier 1 Capital in 2014, at a net premium to other sources of funding of c. £15m per year, to support its leverage ratio relative to other firms. This raised the cost of funding and therefore impacted the Society’s performance and growth potential.
We note the PRA’s announcement on 10 September 2024 that the thresholds for the Leverage Framework are currently under review to ensure they remain consistent with the Bank of England’s concurrent stress testing framework. Should the thresholds for the Leverage Framework match the concurrent stress testing framework (c. 5% of lending or £105bn), then the Society would not expect to be captured by the Leverage Framework, though we would welcome clarity on the outcomes of the review as soon as practical to provide clarity and certainty over future growth and funding plans.
Whilst the policy is based on internationally agreed standards, the additional UK leverage buffers could be acting against the secondary growth and competitiveness objective by penalising already strong and stable firms on a risked weighted basis both domestically and internationally.
We note that the near final but now delayed Basel 3.1 regulations will impose an ‘output floor’ on IRB firms, meaning firms capital requirements will be at least 72.5% of the regulatory prescribed standardised approach. This inherently de-risks the use of the IRB approach in our view.
We encourage the PRA to consider whether a leverage backstop, above minimum international requirements, remains relevant for large, strong, and simple firms. This includes the largest building societies and potentially the ringfenced banks – particularly once Basel 3.1 backstop risk weights are in place that mitigate the IRB modelling risk.
The international Financial Stability Board introduced Total Loss Absorbing Capital (TLAC) standard for Globally Systemically Important Banks in 2015. This contained a leverage backstop of at least 6.75% plus any applicable buffers referred to as MREL. Firms can meet this requirement with a mixture of own funds, capital instruments, and external funding including Senior Non-Preferred (SNP) debt.
The intention of these requirements is to ensure that firms are not ‘too big to fail’ and are able to recapitalise in an orderly way which minimises impacts of financial stability, ensures the continuation of critical functions, and avoids exposing taxpayers to losses.
In the UK, the approach by the Bank of England is more cautious, with MREL requirements applied to Globally Systemically Important Banks, Domestically Systemically Important Banks, and firms with a bail-in resolution strategy. Firms with total assets exceeding between £15bn-£25bn (or 40,000 to 80,000 transactional accounts) are generally designated as a bail-in firm by the Bank of England.
This contrasts with the EU where MREL requirements typically only apply to systemic firms with >€100bn total assets (though national supervisory discretion applies for domestically systemic firms). In the USA MREL requirements only apply to Globally Systemically Important Banks.
We note the PRA’s consultation paper on MREL in October 2024, in particular the intention to increasing the threshold for a bail-in preferred resolution strategy to £20-£30bn. Whilst we welcome the proposed increase to these thresholds, we note that the UK still appears to be much more conservative than other jurisdictions with regards to MREL thresholds, potentially impacting international and domestic competitiveness.
If firms need to raise SNP to meet MREL requirements, this increases their cost of funding, impacting competitiveness, growth and pricing offered to consumers. The transactional account threshold may act as a barrier to growth, particular for smaller or less well-established firms.
In the UK, Domestically Systemically Important Banks and bail-in firms are required to maintain MREL resources more than:
Given the current Leverage Framework applies to firms with an excess of £50bn retail deposits, and leverage requirements are generally higher than risk-based requirements, the Society has had to both carefully manage growth and undertake an SNP issuance programme over the past 24 months to ensure that it can meet the increased MREL requirements upon reaching £50bn retail deposits.
As noted in section 1, we welcome the PRAs publication on 10 September 2024 stating that the thresholds for the Leverage Framework are currently under review to ensure consistency with the Bank of England’s concurrent stress testing framework. Should the threshold for the Leverage Framework mirror the proposed concurrent stress testing framework (c. £105bn), this would result in significantly lower MREL requirements for the Society than previously anticipated – although the resulting surplus SNP funding already in place may act as a drag on performance until it matures.
We would encourage the PRA to be mindful of the additional costs arising from MREL requirements and consider whether a risk based only MREL threshold for large, strong, and simple domestic firms would both support the secondary objective and create a level playing field for those for whom Leverage rather than CET1 is currently the binding constraint.
In our response to the PRA Consultation Paper on Basel 3.1 (CP 16/22) In March 2023 we highlighted the following key impacts on a simple low risk firm:
3.1 A significant overall increase in capital requirements, with evidence that previous requirements exceeded observed historical losses. Under final Basel 3.1 rules, the Society would see a 67%/ £300m increase in the Pillar 1 retail credit risk under the risk-based approach requirement based on its 2022 loan book.
3.2 The application of standardised risk weight floors will disproportionately negatively impact low risk IRB lenders such as the Society and encourage higher LTV lending. For example, the average risk weight for a 65% LTV owner occupier customer doubles, whereas there is no increase in risk weights for 85%.
3.3 As a result, the Society will either need to re-assess returns on lower LTV lending and/or there will be an increase in price for lower risk customers.
Smaller firms could argue that the floors provide a more level playing field with IRB firms, and we welcome the strong and simple regime, changes to the rules on SME for example following industry feedback, and the implementation delay announced on 16th January to ensure international alignment. However, we still believe that the level of capital being held against the lowest risk mortgages could be acting against the secondary objective.
We would ask that both the standardised and Basel 3.1 floors for IRB are reviewed periodically to ensure the relative risk of different mortgage types, and mortgages in relation to other lending types are aligned to the observed risks through the economic cycle and stress testing outcomes.
The current calibration does seem to be encouraging greater risk taking and/ or too much capital being held against the lowest risk mortgage types which may be acting against the interest of both the secondary competitiveness and growth objective and primary stability outcomes.
Institutions (including building societies) looking to improve their risk management capability by applying for IRB permission are finding it difficult to obtain IRB permission, with significant delays by the PRA in approving IRB permission for new IRB firms.
These institutions are unable to benefit from lower lending risk weights which could encourage further growth and competition, notwithstanding the Basel 3.1 output floor will somewhat limit the benefits of IRB going forwards.
The Society (among other institutions) has been waiting for a significant period of time for the outcome of the PRA’s IRB model review. This creates uncertainty when developing the Society’s strategic and financial plans as the future capital position remains uncertain. Lenders may therefore be more cautious in their lending targets, to allow room for unexpected increases in risk-weights from PRA feedback on IRB models.
We would ask for additional PRA resource to be deployed to finalise the IRB model review and to expedite the approval of IRB permissions for new firms to support the secondary objective and create greater certainty about risk-weights.
APPENDIX – Coventry Building Society data points
Impairment losses | £92m
|
Profit before tax | £2.9bn |
Capital resources (June 2024) | £3.2bn |
Source: Published annual and half yearly results.
The Society therefore held reserves equivalent to 980 years of credit losses on 30 June 2024. Whist we acknowledge capital is held to cover tail risk events we believe that this level of reserves, compounded by additional MREL requirements, may be acting against the secondary objective.
Coventry issued £400m of Alternative Tier 1 to improve its leverage ratio in 2014 at a net cost to customers of c. £15m per annum relative to other sources of funding. This has added around 0.05% to the cost of funding a mortgage over this period or cumulatively prevented around
£3bn of additional mortgage lending over the last 10 years.
Coventry have issued an additional £1,500m of MREL compliant debt in total to meet the leverage backstop above the risk weighted requirement. This funding is at a c.2% premium to other sources of funding giving rise to a c.£30m per annum interest cost. This will add a further 0.05% to the cost of a mortgage or prevent around £400m of growth each year. This cost would not be borne by a firm of a similar size and complexity based in the EU or US.
Coventry does not expect to issue additional capital due to the Basel 3.1 IRB floors, although its reported CET1 ratio will fall from 29% to 22%. It is not yet clear whether this will impact the market perception of the Society’s capital strength and therefore cost of funding.
As outlined in section 3 of this paper, we expect the capital requirement for a 65% loan to value owner occupied mortgage to double based on the following table shared previously with the PRA. It is not yet clear how this could impact wider mortgage market pricing and pricing at higher loan to values.
Loan to value | 50% | 65% | 75% | 85% | 95% |
Basel 3.1 RWA | 15% | 21% | 25% | 29% | 31% |
IRB Model RWA | 3% | 11% | 19% | 32% | 46% |
RWA impact | +12% | +10% | +6% | -3% | -15% |
% impact | +400% | +90% | +30% | -10% | -33% |
Source: Society response to PRA CP16/22 - March 2023
The Society believes the increase in risk weights is not reflective of the Society’s low risk lending profile or the underlying tail risk of low-risk mortgage loans.
The Society, and other similar firms, will be disproportionately impacted compared to other firms who can generate better return on capital from other income streams where the gap between standardised floor and IRB risk weightings is not as pronounced.
17 January 2025