Banking – Q3 2022

List of topics covered:

1. Winding down 'synthetic' sterling LIBOR and US dollar LIBOR – Outcomes of the FCA’s consultation
2. Fundamental Review of the Trading Book – Letter to Firms
3. DP4/22 – the PRA’s future approach to Policy
4. Vicky Saporta’s speech on Liquidity Buffers
5. DP 5/22 – Artificial Intelligence and Machine Learning
6. CP 12/22 – Risk from contingent leverage
7. Strong and Simple Regime – Implications on Pillar 1 Capital Requirements

Winding down 'synthetic' sterling LIBOR and US dollar LIBOR – Outcomes of the FCA’s consultation

The FCA previously provided notice that they would continue publication of 1-, 3- and 6-month sterling LIBOR post-closing of panel bank submissions for LIBOR for sterling, yen, franc and euro on 31 December 2021. This synthetic sterling LIBOR benchmark would be a temporary fix to reduce market disruption during transition of SONIA.

Following consultation, the FCA now expects that both 1- and 6-month synthetic sterling LIBOR will be closedby March 2023. However, there is less clarity on 3-month synthetic sterling LIBOR due to its use in mortgages. The regulator and market participants have identified the fact that mortgage conversion may take place and, in doing so, require consent from retail customers. Therefore, the FCA is taking an open ended approach and will, ‘in due course, provide further information onwhen it will be possible for the 3-month sterling LIBOR setting to cease in an orderly fashion’.

The FCA is also continuing to assess whether a synthetic US dollar LIBOR rate will be appropriate ahead of the current benchmark closing on 30 June 2023. This should be of interest to a wide range of market participants including regulated users of synthetic sterling LIBOR and US dollar LIBOR. The FCA has sought to understand the size and exposures of market participants to US dollar LIBOR and any challenges associated with publication of a synthetic benchmark.

We will provide an update on LIBOR in the next quarterly newsletter.

Fundamental Review of the Trading Book – Letter to Firms

This is a subset of the wider Basel 3.1 reform package which is scheduled to be introduced on 1 January 2025. The PRA is currently in the process of consulting on the wider reform package as part of a consultation paper (CP) due this quarter. There will be a significant focus on publication of this CP given the importance of Basel 3.1 in shaping the landscape for prudential requirements from 2025 onwards.

The new Basel 3.1 framework overhauls the IMA market risk methodologies with an entirely new approach. Existing model permissions will become redundant, and firms will move to the standardised approach upon implementation of the new rules. The only exception to this is instances when firms had received an IRB permission, based on the new approach, from the PRA.

As a consequence, the PRA is engaging with firms ahead of FRTB implementation, as they have been doing so since 2018. The final date for firms IMA submissions, to ensure that they avoid going onto the standardised approach, is 1 January 2024. The planned schedule for review is as follows:

• Q4-2022 – Default Risk Charge (DRC)
• Q1-2023 – Risk factor eligibility test (RFET)
• Q2-2023 – Non-modellable risk factors (NMRF)
• Q3-2023 – P&L attribution test (PLAT) and back-testing

It is important for firms to bear these deadlines in mind in prioritising their work and engaging with the PRA.

DP4/22 – the PRA’s future approach to Policy

This Discussion Paper (DP) describes how the PRA will change its policy making approach as the organisation takes on a greater degree of rulemaking powers through application of the Financial Services and Markets Bill. It is directly applicable for all PRA-regulated firms.

The most important aspect of the bill is to introduce a new secondary objective for the PRA; seek to facilitate the international competitiveness of the UK economy while still aligning with international standards. This will give the PRA scope to review elements of the policy landscape that previously had been fixed by UK legislation. Some areas that may benefit firms in the medium and long term include a push to make the regulatory framework more accessible and user-friendly, as well as reducing the complexity and fragmentation of the current regime.

There are 4 main elements to the PRA’s overarching proposal.

1) Strong standards

Re-emphasis on the fact that the strong regulatory standards put in place by the Bank of England following the Financial Crisis of 2007-08, support the stability of firms and wider market. This will continue to be important in underpinning competitiveness internationally as set out in the new secondary objective.

2) Accountability

The PRA continue to affirm their responsibility to engage HMT and government under both the old and new framework.

This is uncontroversial and indicates no change to the existing relationship between parliament and the regulator.

3) Responsive approach

Developing policy with a greater degree of flexibility for UK firms is highlighted, with the example of ‘Strong and Simple’ given. The increasing importance of regulatory data is also highlighted. PRA commits to publication of a framework for evaluation of the policies of the future.

4) Accessible rulebook

The intention is outlined to simplify the PRA Rulebook and improve its usability. This will be dependent on the government’s approach to repeal of relevant UK law.

Market participants will welcome this step. Given the regulators’ proclivity for increasing complexity since the Global Financial Crisis it is likely that the change in philosophy may take some time to generate tangible results.

The deadline for comments on the DP is Thursday 8 December.

Vicky Saporta’s speech on Liquidity Buffers

Vicky Saporta, the Bank of England’s Head of Prudential Policy, gave a wide-ranging speech to the Bank of International settlements which covered both capital and liquidity buffers.

She considered the lessons learnt from the recent Covid-19 stress test and what the implications of these could have on the structure of liquidity buffers going forward.

During the pandemic, there is evidence to suggest that banks were reluctant to utilise their liquidity buffers despite guidance from the regulator that they were usable. Saporta suggests there are 4 main reasons why this was the case:

• Regulatory environment – Surveys have consistently shown that firms are unwilling to utilise buffers due to potential regulatory responses. Firms in the UK also felt that there was a lack of clarity around how regulators would respond to firms entering into their LCR buffers. The regulator has, in the past, clarified that it is acceptable for firms to dip into buffers during times of stress (as buffers are designed specifically for that purpose), but this messaging has not necessarily triggered a market response.

• Market reaction – Concerns about stigma are equally applicable to financial markets. Leveraging liquidity reserves can give the impression to the market that a firm is in trouble. This might become a self-fulfilling prophecy when loss of short-term funding on this basis triggers severe financial distress or even failure. ‘Signalling’ has always been a powerful tool that firms have used to portray itself in the market; dividend-signalling being the most popular. However, it is a matter of fact that ratings agencies, exchanges and investors alike are looking at a number of different regulatory metrics to form opinions about firms’ creditworthiness; and this often counterbalances the influence of regulatory messaging on buffer-usability discussed above.

• Financial disclosures – A degree of openness in a firm’s disclosures, including liquidity adequacy and use of regulatory buffers, can lead to negative consequences. Firms often have an obligation to disclose to investors if they enter into buffers. Such disclosures to investors are not a regulatory requirement, but nonetheless, it is undeniable that investors see performance against regulatory metrics as a signal of firms’ health. This discourages firms from dipping into their buffers. There is perhaps an argument that such observed standards of discipline in firms’ adherence to capital requirements could one day translate into reduction in capital requirements.

• Uncertainty – A liquidity stress does not have a clear duration or severity. Rational actors may exercise, what in retrospect looks like, excessive caution because of a fear of what may be to come. As seen during Covid-19 market turbulence, belt-tightening is often unnecessary in the medium term given many shocks do not develop into more protracted market events.

Saporta suggests that regulatory authorities may have produced an ‘unintended equilibrium’ where a 100% LCR ratio is essentially baked into governance and oversight. Regulatory messaging is therefore, not enough, the regulators need to go further. The lack of usable minima in liquidity buffers is acting as a restriction because the only regulatory figure firms must base their prudential adequacy by is the 100% LCR ratio figure.

Interestingly, when discussing capital buffers in his Bufferati speech the head of the PRA, Sam Woods, spoke about a similar approach to capital.

DP 5/22 – Artificial Intelligence and Machine learning

This is an extremely wide-ranging paper jointly published by the PRA and FCA. The DP focuses on the regulation of AI and Machine learning (ML) in UK financial services. In many respects, the PRA and FCA appear to be utilising this DP as an opportunity for further discussion and dialogue on AI and ML rather than the development of anything concrete at this stage. However, there are a number of interesting insights into how a UK framework might develop over time. 

What is AI?

Previously, the PRA and FCA have defined AI as ‘the theory and development of computer systems able to perform tasks which previously required human intelligence’. The term does not have a broad consensus definition, but it commonly relates to computer elements that input, output, process and store data.

Why is it important that regulators develop regulatory frameworks for AI?

In a highly regulated sector such as banking, firms have become increasingly used to operating between clearly defined regulatory boundaries. Lack of clarity from authorities about what banks should or shouldn’t do is now regarded as actually having the effect of stifling innovation in AI.

One example of this is in the absence of regulatory steer or guidance in the context of application of ML in advanced capital modelling (IRB models), which is expected to significantly increase in the short term. This is leading to banks struggling to approach the development of AI models with compliance confidence. This could cause the UK banking sector to lose ground on their international peers.

Potential regulatory approaches to AI

It is clear that the PRA will look to be at the forefront of developing responses to the rise of AI and ML.

However, there are far-reaching concerns associated with AI that will need to be considered by the regulators. AI does not just carry with it idiosyncratic risk, for example, the development of poor or misleading models; there are also significant financial stability concerns brought about by the potential amplification effects driven by the uniformity of models developed with AI architecture.

The PRA is already embedding their views within existing elements of the supervisory framework. For example, CP6/22 ‘Model Risk Management Principles for Banks’, incorporates consideration of the application of AI models as part of their revised supervisory principles. Firms should be mindful of this when engaging with regulatory consultation more generally.

Impact of a lack of skills and experience

Banks are experiencing the same issues around resourcing as the rest of the market, but particularly for roles that often see direct competition for talent with the technology sector. This paper also exposes the extent to which the regulatory authorities lack fundamental experience and resource in this space. This is compounded by the lack of focus in the drafting of the DP.

It is therefore highly likely they will be more open to material feedback in this area than normal in helping to guide or inform final policy.

The deadline for comments on the DP is Friday 10 February.

CP 12/22 – Risk from contingent leverage

What is contingent leverage? This can arise from different forms of financing (including derivatives, netting transactions, unsecured borrowing or lending of securities, collateral swaps and swaps more generally) which are considered capital efficient.

Why is capital needed? Contingent leverage risk arises when a firm can’t rely on these capital efficient types of financing, especially in stress. This risk has clearly crystallised across the industry during the pandemic. Capital is therefore required to be held against this type of financing, via the leverage ratio framework, to reduce the impact of this risk on the wider economy.

Who does this apply to? All LREQ firms - firms which sit under the Leverage Ratio regime either have non-UK assets in excess of £10bn or have UK retail deposits that are higher than £50bn.

What does the CP propose? Update to the PRA’s ICAAP guidance, including additional detail on how to assess risks arising from contingency leverage. Firms will be expected to cover how destressed selling caused by contingent leverage impacts business models and valuation of assets.

How will this be reported? There will be the introduction of a new reporting guidance for firms as part of wider leverage ratio reporting. Reporting will take place on a biannual basis, with first submissions expected on 31 December 2023. Firms will be expected to report data on trades that are most relevant to contingent leverage. This includes; collateral swaps, netted repos, cash and synthetic prime brokerage positions.

What are the intended outcomes? The anticipation is that the proposals would help market participants to manage contingent leverage risk and assist the PRA in taking action where required. The expectation is that by providing more guidance on contingent leverage for ICAAPs, the PRA will also be encouraging a level playing field in approaches to this risk as a leverage requirement.

PRA expects that this revised policy would take effect on 1 July 2023. The consultation period closes on 3 February 2023.

Strong and Simple – Implications on Pillar 1 Capital Requirements

For the past 24 months, senior figures at the Bank of England have been signalling their intention to develop a “strong and simple,” regime for less systemically important banks. A consultation paper was published in April 2022 which sought to define the definition of a ‘Simpler-regime Firm’. Further detail has been promised on how this will work in practice from a prudential (expected 2024) and non-prudential (expected 2023) perspective.