Reflections on the 2023 banking turmoil

Speech by Pablo Hernández de Cos, Chair of the Basel Committee on Banking Supervision and Governor of the Bank of Spain, at the Eurofi Financial Forum 2023, Santiago de Compostela, 14 September 2023.

Introduction

Good evening, and thank you for inviting me to speak at our dinner tonight.

I should start by wishing you all "una gran bienvenida" to Spain. And, in the event that some of you came to Santiago de Compostela by completing the Camino, let me say "felicidades" and "Ultreia et Suseia"!

A common expression in Spain is that "el Camino da más de lo que recibe" – the Camino gives more than it receives. While I cannot claim to offer you any more ecclesiastical insights this evening, I will be reflecting on the recent banking turmoil and the implications for the global banking system and the Basel Committee.¹ 

For some of you, the turmoil may seem like a distant memory. Since the frenzied months of March to May, many banks have been reporting bumper financial results on the wave of rising interest rates. A cursory look at financial markets since that period would also suggest that the worst may be behind us. So why do I plan to look back at what may be regarded as some as a historical event?

Put simply, the banking turmoil that started in March is the most significant system-wide banking stress since the Great Financial Crisis (GFC) in terms of scale and scope. Over the span of 11 days – from 8 to 19 March 2023 – four banks with total assets of about $900 billion were shut down, put into receivership or rescued. This was followed by the failure of a fifth bank with roughly $230 billion in assets on 1 May 2023. To give you a sense of the order of magnitude, the total value of these banks' assets is roughly equivalent to Spain's annual GDP (leaving aside the stock versus flow nature of these numbers).

The distress of these individual banks, while having largely distinct causes, triggered an assessment of the resilience of the broader banking system. In response, large-scale public support measures were deployed by some jurisdictions to mitigate the impact of the stress, including significant central bank liquidity provision to banks, the activation of FX swap lines, government backstops or guarantees, and, in certain cases, an extension of deposit guarantee schemes. In many respects, today's stabilisation of the banking system is due to a combination of public support measures and the increased resilience provided by post-GFC regulatory reforms, most notably Basel III. We had hoped that we would not need to rely on the former so frequently.  

Against that backdrop, the Basel Committee undertook a review of this period and conducted a stocktake of the regulatory and supervisory implications of these developments, with a view to learning lessons. I am pleased to inform you that, as recently announced by the Group of Governors and Heads of Supervision, good progress has been made with this work.² I will focus my remarks tonight by offering my personal views on some of the main takeaways and identifying some issues that may warrant further reflection.

Risk management and governance

There is perhaps a near universal agreement that one of the main lessons from the turmoil is the importance of banks' risk management practices and governance arrangements as the first and most important source of financial and operational resilience. The boards and management of banks should be the first port of call in managing and overseeing risks; these functions cannot be outsourced to supervisors. Jumping straight to discussions about the regulatory and supervisory implications of recent events is akin to forgiving banks for not fulfilling their primary responsibilities and likewise shareholders for not exercising due diligence.³

Yet the banking turmoil highlighted a series of weaknesses by some banks in this area, including:

• fundamental shortcomings in (basic) risk management of traditional banking risks (such as interest rate risk and liquidity risk, and various forms of concentration risk);
• a failure to appreciate how various risks that were building up were interrelated and could compound one another;
• inadequate and unsustainable business models, including an excessive focus on growth and short-term profitability (fuelled by remuneration policies), at the expense of appropriate risk management;
• a poor risk culture and ineffective senior management and board oversight; and
• a failure to adequately respond to supervisory feedback and recommendations.

Many of these elements may appear obvious and quite basic in nature. So it is of deep concern to see that, in 2023, some banks' boards and senior management failed in their most elementary responsibilities of overseeing and challenging a bank's strategy and risk tolerance. More is clearly needed to shore up such responsibilities.

Consider the following historical anecdote.⁴ In 1800, a French chemist by the name of Éleuthère Irénée du Pont set up a gunpowder factory in Delaware. He quickly realised that gunpowder factories have an undesirable property: they tend to explode frequently. In response, du Pont took two initiatives. First, he required that the director (himself) live inside the factory with his family, putting his life on the line – what you could view as "skin in the game". Second, he established a rule that every new piece of machinery had to be operated for the first time by the factory's senior management. If the machine blew up, the manager would suffer the consequences. Needless to say, the safety of the plant increased overnight.

I don't think I need to draw out explicitly the comparisons with today's banking system. But it is clear that the turmoil raises some fundamental questions about the current banking system.

Is it simply inevitable that there will always be "outlier" banks with serious governance and risk management shortcomings? Is this a "feature" of a banking model that combines leverage and maturity transformation with a focus on short-term gains? Have we optimised the alignment of incentives between banks' boards and senior management and broader financial stability objectives? I don't have the answers to all of these questions, but I think they certainly merit further reflection.

Strong and effective supervision

The banking turmoil also highlighted the importance of strong and effective supervision across various dimensions. These include recurrent issues that we've seen in previous banking crises in addition to newer elements. Either way, they raise important takeaways for supervision, which I've grouped into six categories.

First, the turmoil underlined the importance of supervisors developing a thorough understanding of the viability/sustainability of banks' business models as part of their supervisory process, including identifying any areas in which a bank is an outlier, so they can assess and take action to address any weaknesses at an early stage. This may all seem obvious to you, but there are clearly outstanding challenges for supervision, including: (i) how best to assess the viability of business models in a holistic manner (eg relying on a broad set of quantitative and qualitative indicators); (ii) how to proactively engage with outlier banks without "crossing the line" and "co-owning" a bank's business strategy; and (iii) how to monitor medium-term structural changes to better identify their impact on different business models.

Second, a core element of supervisory work is ensuring that banks have effective and robust governance and risk management. This includes, but is not limited to, the composition of the board and the extent to which its members have relevant experience, including banking and financial expertise; the board's ability to effectively challenge the bank's senior management, oversee the bank's risk profile and steer its strategy; the independence and empowerment of the risk management and internal audit functions; the enterprise-wide risk culture, including how embedded it is in corporate and business processes; and the incentives provided by senior management compensation schemes.

Third, the turmoil highlighted clear challenges in overseeing banks' liquidity risk. These challenges relate to: the speed and volume of deposit outflows and changes in banks' funding profile; the importance of banks being operationally prepared for liquidity stress scenarios (eg by having credible and tested contingency funding plans, and operational readiness to access central bank liquidity facilities); and the role of social media and the digitalisation of finance in hastening the speed and impact of a bank's distress. These developments, in turn, prompt considerations for supervisors around, among other issues, whether (i) their monitoring of banks' liquidity risk profile provides the relevant information in a timely manner; (ii) the frequency of monitoring can be increased, both during times of stress and business as usual; (iii) supervisory monitoring can leverage different sources of information and high-frequency data; and (iv) monitoring of concentration risks is warranted.

Fourth, we've been reminded once again that supervisory judgment is a critical element to ensure that the intent, as well as the letter, of regulation is addressed. A rules-based approach on its own is unlikely to appropriately identify, assess and allow the timely mitigation of key risks to a bank's safety and soundness and broader financial stability. This does not diminish the role of a rules-based approach in setting minimum standards. Rather, it prompts considerations for supervisors around how they can effectively complement such standards by exercising judgment – and therefore intervene proactively even when specific rules have not been breached – to make bank supervision dynamic and adapted to a bank's specific business model and operations, and the risks that they present.

Fifth, it is important to reflect on the role and scope of existing supervisory toolkits as complements to minimum global standards and to ensure they are sufficient to drive concrete action at banks, including in the light of any legislative/regulatory constraints on how or when they might be applied. A recent paper by staff at the International Monetary Fund finds that, while the importance of a sound institutional setting for effective bank supervision is widely accepted, many jurisdictions do not equip bank supervisors with the necessary powers and conditions for their work.⁵ Supervisory authorities could also review whether the guidance and processes given to individual supervisory teams appropriately incentivise a willingness to act early, accompanied by a clarity of process on how to do so.

Sixth, while there were several positive elements of cross-border supervisory cooperation during the turmoil – including at the Committee level – consideration could be given as to whether broader information-sharing protocols at a cross-border level are necessary. Any such protocols would, of course, have to take into account constraints on authorities' ability to share confidential information, existing information-sharing arrangements and resource implications.

Robust regulation

Moving to regulatory reflections, let me be clear upfront: the regulatory imperative for the Basel Committee at this stage is to implement all aspects of the Basel III framework in full, consistently, and as soon as possible. Nevertheless, there are issues directly or indirectly related to the turmoil that I think would merit further analysis and reflection.

My starting point is that prudential regulation – and Basel III more specifically – is not calibrated to produce "zero failures", but seeks to reduce the likelihood and impact of banking stress, while facilitating financial intermediation and economic growth.

Moreover, most of the banks that failed were not subject to the Basel III framework in full.

Let me now offer some personal reflections on four regulatory issues that I think would benefit from further analysis.

First, liquidity. While each of the banks that failed during the turmoil had idiosyncratic features, they all ultimately succumbed as a result of significant liquidity outflows and an inability to maintain sufficient stable funding. To date, most of the commentary has focused on the significant scale and speed of outflows experienced by these banks – up to 85% of deposits over the span of two days for one of them – and whether the Liquidity Coverage Ratio (LCR) and Net Stable Funding Ratio (NSFR) are miscalibrated as a result. It is helpful to take a step back and consider a broader set of questions about the Basel III liquidity standards:

• What exactly are the objectives of these standards? The LCR requires banks to hold sufficient liquid assets to meet a 30-day stress outflow period. So, before reviewing the "denominator" of this ratio (ie the assumed outflow rates), a more fundamental question is whether we still expect banks to be able to survive a liquidity stress for 30 days without some sort of public intervention/resolution/private sector solution. Should the LCR be more focused on buying enough time for authorities to address a liquidity stress? What is its role relative to other liquidity metrics, both quantitative and qualitative?
• A second fundamental question is with regard to the design of the LCR and NSFR. Unlike capital standards, there is no concept of a "hard" minimum requirement supplemented by a "buffer" requirement. In principle, banks should be able to dip into their stock of liquid assets in times of stress to meet outflows, while also submitting a satisfactory restoration plan to their supervisor. Yet it would appear that banks continue to be reluctant, or unable, to fully use their liquid assets in the manner envisaged. A number of potential factors behind such behaviour have been suggested, including the calibration of existing liquidity requirements, perceived stigma, market expectations and/or operational constraints.

There is also the more topical question about the role of digitalisation and social media on liquidity outflows. Through the modern history of finance, advances in communication technology have sped up the flow of information, affecting the nature and magnitude of banking crises. In the Panic of 1873, financial stress that began in Europe spread to North America, facilitated by the transatlantic telegraph cable completed in 1866. In the Black Monday global stock market crash in 1987, contagion spread across financial markets via electronic communications. In the present, rumours can spread through social media.

At the same time, innovation has made it faster and easier to move money, from the creation of the ATM to modern digital banking apps, alongside faster payments and reduced settlement windows. When combined with advances in communications technology, these developments have further reduced frictions and allowed for rapid inflows and outflows. As recently as 2008, depositors at IndyMac and Northern Rock still formed long lines outside bank branches; as we saw in the recent turmoil, withdrawals can now be initiated online in a matter of minutes if not seconds.

And while fingers have pointed at the role of social media, it is important to further unpack what this means. In practice, there is a wide spectrum of "social media" communication channels. This ranges from public platforms that target a broad audience and can amplify bank concerns (eg X/Twitter, Facebook, LinkedIn, Instagram), specialist (public) forums (eg Y-combinator, Reddit, Discord), encrypted messaging applications (eg WhatsApp, WeChat, Signal, Telegram), internal corporate messaging platforms (eg Slack and Circle) and even telephone calls. These platforms increase the global interconnections among clients, which could foment the risk of herd behaviour in times of stress. As a result, these developments may be relevant not only for regulators, but, as I previously mentioned, also raise important questions for supervisors as to how best to monitor and respond to social media, in both "peace" and "crisis" times.

Second, interest rate risk. A recurring theme related to the distress of some banks during the turmoil was the common and concentrated exposure to interest rate risk in the banking book (IRRBB). Again, these banks were not subject to the existing IRRBB standard, but these events have once again attracted attention towards the current regulatory treatment of IRRBB in the Basel Framework. Some areas that have been mentioned for further analysis and evaluation include whether the current Pillar 2/3 approach to addressing IRRBB is still appropriate? Are there ways to further strengthen it, by providing more stringent guidance and requiring further disclosures? Or is there a need to move towards a Pillar 1 capital framework for IRRBB to promote greater international consistency and comparability?

The third category of issues relates to two aspects of the definition of regulatory capital. First, unrealised interest rate losses on fixed income assets held at amortised cost were an important driver in the failure of several banks during the recent turmoil. If banks need to sell such securities before their maturity date to meet liquidity needs, unrealised losses on those securities become realised losses and would reduce both equity and regulatory capital. Moreover, the large-scale and ad hoc fire sales by some troubled banks to meet large-scale and simultaneous deposit withdrawals may also require reflection on how best to reflect the risks from second-round fire sales. This is an area where further analysis and evaluation could also be performed but, equally importantly, is of critical importance for supervision and banks' own risk management practices.   

Recent events have also highlighted the role of Additional Tier 1 (AT1) capital instruments in the capital framework. Investors and markets did not fully internalise the various trigger events that could lead to the loss participation of AT1 instruments, even though the Basel Framework contains explicit language on those trigger events and despite contractual documentation clearly highlighting the corresponding risk factors of such instruments. In addition, the fact that a distressed bank continued to make expensive replacement issuances and to pay substantial amounts of discretionary interest on these instruments (alongside dividend payments for common shares), despite reporting losses over several consecutive quarters, raises questions about the ability of such instruments to absorb losses on a going-concern basis. The Committee has previously evaluated the functioning of these instruments, but was unable to draw robust empirical conclusions regarding their loss-absorption capacity.⁶ Future analysis and evaluation would need to be considered as part of a more holistic assessment of the role of different regulatory capital instruments and their functioning in crisis times.

The fourth category of regulatory issues to reflect on pertains to the application of the Basel Framework. This includes the determination of what constitutes an "internationally active bank". The Basel Framework intentionally does not define this concept, given structural differences in banking systems across jurisdictions. Yet recent events have shown that the failure of a bank can have systemic implications through multiple channels, including first- and second-round propagation effects. Put differently, factors such as size and cross-border interconnections are important considerations when deciding on the appropriate scope of application of the Basel Framework.

The flip side of this issue is the role of proportionality for non-internationally active banks. As you know, jurisdictions may opt to apply the Basel Framework for non-internationally active banks, including smaller ones. In such cases, they can apply the framework in some proportionate manner, commensurate with the risk profile and systemic importance of banks. Member jurisdictions are wholly responsible for deciding on whether and how to apply and design proportionate frameworks, and the recent turmoil highlighted how the distress of banks subject to domestic proportionality regimes could have cross-border financial stability effects.

The turmoil also highlighted how the design of proportionality frameworks can impede effective supervision by reducing standards, increasing complexity and promoting a less assertive supervisory approach.

There may therefore be merit in members continuing to share their experiences in applying proportionality, monitoring the scope of banks subject to proportionate approaches, and in ensuring that these objectives are adequately met.

Conclusion

I started my remarks this evening with a Spanish expression about the Camino. Let me end with another one: "Nunca es demasiado tarde para encontrar el Camino" – it is never too late to find the Way.

So what is the way forward for the Committee with regard to the implications from the banking turmoil? I am pleased to note that there is broad agreement to prioritise further work to strengthen supervisory effectiveness, including identifying issues that could merit additional guidance at a global level. In addition, the Committee will pursue additional follow-up analytical work based on empirical evidence to assess whether specific features of the Basel Framework performed as intended during the turmoil, such as liquidity risk and interest rate risk in the banking book. And we will continue to coordinate with other global forums and standard-setting bodies on cross-cutting issues.

Importantly, the already-implemented Basel III reforms helped shield the global banking system and real economy from a more severe banking crisis. So there is also an equally broad agreement at the Committee level, reaffirmed by the Group of Governors and Heads and Supervision, on the critical importance of implementing all aspects of the Basel III framework in full, consistently, and as soon as possible. Put simply, none of the follow-up work to the turmoil should interrupt the imperative of implementing the outstanding Basel III standards. In this respect, the Committee will continue to monitor and assess the full and consistent implementation of Basel III.

Thank you.

References

Adrian, T, M Moretti, A Carvalho, H K Chon, K Seal, F Melo and J Surti (2023): "Good supervision: Lessons from the field", IMF Working Paper, WP/23/181, September.

Basel Committee on Banking Supervision (2022): "Evaluation of the impact and efficacy of the Basel III reforms", December.

--- (2023a): "Basel Committee discusses recent market developments, agrees to consult on Basel Core Principles, and advances work on cryptoassets", June.

--- (2023b): "Governors and Heads of Supervision endorse initiatives in response to the banking turmoil and reaffirm priority to implement Basel III", September.

Dellanna, L (2020): "Best practices for operational excellence", December.

Hernández de Cos, P (2023a): "Banking starts with banks: initial reflections on recent market stress episodes", April.

--- (2023b): "Where next for the Basel Committee?", April.

¹ See Hernández de Cos (2023a and 2023b) for initial reflections on the turmoil.
² BCBS (2023a and 2023b).
³ Hernández de Cos (2023a).
⁴ This paragraph is adapted from Dellanna (2020).
⁵ Adrian et al (2023).
⁶ BCBS (2022)