Investing in banking supervision
Speech by Mr Agustín Carstens, General Manager of the BIS, at the European Banking Federation's International Banking Summit, Brussels, 1 June 2023.
Introduction
I would like to thank the European Banking Federation for inviting me to speak today.
Recent months have seen several episodes of banking stress, in both Europe and the United States.
Once again, central banks have stepped in to provide liquidity, restore confidence and prevent contagion. At the same time, they have been raising interest rates to bring inflation back to target.
In principle, central banks can separate their monetary policy actions from their financial stability function. But perceptions matter. And raising interest rates on the one hand, while easing financing conditions for banks on the other might create the impression that policy arms are pulling in different directions.
This wasn't meant to happen. The Basel III reforms aimed to ensure that banks kept ample liquidity on hand and remained well capitalised at all times. And, if systemically important institutions did fail, well planned resolution procedures were meant to prevent spillovers to other institutions.
What went wrong? Many have pointed fingers at rapid interest rate hikes, falling bond prices and flighty depositors. In some cases, these provided the trigger. But in my view, the main cause of recent bank crises was the failure of directors and senior managers to fulfil their responsibilities. Business models were poor, risk management procedures woefully inadequate and governance lacking.
These issues existed well before depositors ran and investors lost confidence.1 Many of the shortcomings could, and in my view should, have been identified and remedied ahead of time. This speaks to the crucial role of banking supervision. This will be the focus of my talk.
Banking supervision needs to up its game. It needs to identify weaknesses at an early stage and act forcefully to ensure that banks address them. To do this, supervisors will need to have operational independence, strengthen their forward-looking culture and adopt a more intrusive stance. They will also need to continuously seek to improve their capabilities. First, by accessing greater resources. And second, by enhancing their productivity with the aid of technology.
The crucial role of banking supervision
Rising interest rates have challenged some banks, central banks and banking supervisors. Some banks' business models have been exposed, particularly after a decade of exceptionally low interest rates led them to take on greater risks in search of yield.
The banks that have come under strain were similar in several respects. One was that they had poor governance and inadequate risk management. As the Chair of the Basel Committee recently noted,2 bank boards and senior managers are the first line of defence in managing and overseeing the risks posed by rising rates.
But there were also some important differences. Some of the banks that failed had a long track-record of financial underperformance. The weaknesses of their business models had been clear for some time. For others, financial difficulties emerged suddenly. I am referring here to banks who had significant exposures to long-term, fixed rate assets funded with less stable short-term funding. Of course, such liquidity transformation is inherent in the banking system – but the liquidity risks need to be managed.
The regulatory framework provides authorities with tools to address the risks posed by rapidly rising interest rates. In particular, well defined minimum capital and liquidity requirements – such as those in the Basel III standards – equip banks with a buffer to withstand adverse interest-rate movements and buy time to address more structural weaknesses. At the same time, we need to acknowledge that the current regulatory framework may not fully capture risks such as those posed by rising interest rates to exposures in the banking book, which were at the heart of some of the recent failures. As a result, recent developments may offer lessons that could justify adjustments to prudential rules – in specific jurisdictions or at the global level – to help them become even more effective.
However, regulation alone cannot address all channels through which higher interest rates affect a bank's viability. Minimum regulatory requirements are, by design, narrow in scope and are not tailored to each bank's risk profile. And there is simply no reasonable level of minimum capital and liquidity that can make a bank viable if it has an unsustainable business model or poor governance.3
This is why supervision is so important. In most jurisdictions, legal frameworks give supervisors a broad and flexible toolkit to identify, assess and, if warranted, address banks' exposure to heightened risks. Through the combined outputs of its various components, including onsite inspections, stress tests and business model analysis, supervisors can develop an informed, comprehensive assessment about the ability of banks to manage their main risk exposures and the sustainability of their business models.
These comprehensive assessments form the basis for supervisory actions to address problems pre-emptively, before risks crystallise. In the first instance, regular supervisory dialogue and moral suasion may be enough to resolve issues. Beyond that, supervisors often have the power to escalate matters and deploy formally binding requirements commensurate with the nature and severity of the identified deficiency. Available measures can include capital and liquidity add-ons, as well as qualitative measures aimed at correcting structural deficiencies in governance, risk management and even business models.4 With such a holistic and forward-looking approach, supervisors can prevent an identified vulnerability from evolving into a threat to the bank's safety and soundness.5
Early identification of vulnerabilities is even more relevant in the light of recent events. The combination of social media and technology seems to have amplified the speed at which bank runs can occur. Social media can spread concerns about a specific bank among its depositors even more quickly. And technology such as a mobile banking app lets customers open and close accounts and transfer their deposits in a matter of minutes.6 Therefore, an intrusive and forward-looking supervision, which takes forceful action at an early stage when necessary, is more important than ever.
The ability and will of supervisors to take early and forceful action when needed are predicated on supervisors having operational independence, appropriate powers and legal protection. An organisational culture that empowers supervisors to take such actions even when faced with uncertainty is also essential. These are some of the pre-conditions that allow supervisors to properly discharge their responsibilities with autonomy while remaining accountable for their actions and decisions. However, while necessary, these elements may not be sufficient if supervisors do not have the resources or the ability to use them efficiently.
Stepping up supervision part 1: resources
After the Great Financial Crisis, most supervisors increased the intensity of their supervision of systemically important banks following a risk-based approach. This meant prioritising resources where they were most needed. In some cases, however, it also meant greater reliance on automated processes and fewer resources devoted to dealing with less systemic financial institutions.
Allocating resources following a risk-based approach makes sense. It helps supervisors to pursue their mandate and safeguard financial stability. But recent episodes remind us that banks that may not be considered systemic ex ante can create systemic distress, via contagion, when they fail. This indicates that resources should be sufficient to apply an adequate oversight to all institutions.
This is not just about numbers. It is also about having the right skills to keep up with relevant developments. Traditional financial disciplines such as accounting and risk management remain essential. But other, less traditional, skills are now needed as well. In particular, the large and far-reaching impact on banks of ongoing technological disruption means that supervisors must find and develop sufficient expertise in areas like cyber security, data analytics and artificial intelligence. That, of course, obliges authorities to face the challenges posed by the strong demand in all sectors for professionals with those qualifications.7
Needless to say, supervisory resources cost money. Budgets must rise, in some jurisdictions significantly so. A range of funding arrangements could facilitate these investments, including by creating or increasing industry-contributed supervisory fees. Some will no doubt complain. But this would be money well spent. Financial crises give rise to massive social and financial costs. By reducing their likelihood, investments in a more effective supervisory framework will certainly pay off and generate substantial returns for society as a whole.8
Stepping up supervision part 2: technology
Supervisory output and effectiveness may also be enhanced by means of productivity gains, which in turn could be obtained with the aid of technology. For example, tools may be developed to automate parts of the supervisory process, particularly repetitive tasks that do not require expert judgment, thus creating conditions for a more effective and efficient allocation of supervisory resources.
In addition, authorities may adopt innovative technology such as big data, artificial intelligence and machine learning to further enhance effectiveness and efficiency. Indeed, many authorities have followed this route, with the pandemic accelerating this trend. Travel restrictions and social distancing protocols forced supervisors to shift nearly all their on-site activities to an off-site surveillance mode. In response, several authorities developed new tools with a view to maintaining their capacity to assess the soundness of financial institutions even under such challenging conditions.
One area where significant progress has been made in the past few years is data analytics. These suptech tools, which use a vast amount of data, both qualitative and quantitative, have the potential to enhance different components of the supervisory process. For example, tools for text analysis, text summarisation and information classification allow for faster extraction of useful insights from lengthy documents produced by the supervised entities. Tools for sentiment analysis, network analysis and peer group identification can provide additional insights on the risks faced by banks and may therefore contribute to the difficult task of identifying deficiencies at an early stage.9
Concluding remarks
Let me conclude. The ultimate cause of recent bank failures lies with the institutions themselves, not with regulators or higher interest rates. There is no excuse for institutions to mismanage interest rate risk, or to fail to address long-term structural weaknesses in their business models.
But banking supervision needs to step up to ensure the safety and soundness of financial institutions under different macro-financial scenarios in the new technological environment. It is essential, in that regard, that supervisors are sufficiently forward-looking and intrusive. With sufficient resources and the aid of technology, supervisors will be able to identify more vulnerabilities at an early stage and to act on them before problems become too large and complex to handle. While this will not prevent all future bank failures, such investments will certainly reduce the likelihood and impact of failures affecting the stability of the financial system.
References
Beerman, K, J Prenio and R Zamil (2021): "Suptech tools for prudential supervision and their use during the pandemic", FSI Insights on policy implementation, no 37, December.
Borio C, J Contreras and F Zampolli (2020): "Assessing the fiscal implications of banking crises", BIS Working Papers, no 893, October.
Coelho, R, A Monteil, V Pozdyshev and J-P Svoronos (2022): "Supervisory practices for assessing the sustainability of banks' business models", FSI Insights on policy implementation, no 40, April.
Coelho, R, F Restoy and R Zamil (2023): "Rising interest rates and implications for banking supervision", FSI Briefs, no 19, May.
Crisanto J-C, J Prenio, M Singh and J Yong (2022): "Emerging sound practices on supervisory capacity development", FSI Insights on policy implementation, no 46, November.
Dahlgren S, R Himino, F Restoy and C Rogers (2023): "Assessment of the European Central Bank's Supervisory Review and Evaluation Process", Report by the Expert Group to the Chair of the Supervisory Board of the ECB.
Federal Deposit Insurance Corporation (2022): "Proposed 2023 FDIC Operating Budget", December.
Federal Reserve (2020): "Supervisory assessment fees", December.
Federal Reserve (2023): "Review of the Federal Reserve's supervision and regulation of Silicon Valley Bank", April.
Hernández de Cos, P (2023): "Banking starts with banks: initial reflections on recent market stress episodes", keynote speech at the Institute of International Finance Roundtable on the Shifting Risk Landscape, 12 April.
Office of the Comptroller of the Currency (2021): "Congressional budget justification and annual performance plan and report", December.
1 Coelho et al (2022).
2 Hernández de Cos (2023).
3 Dahlgren et al (2023).
4 Coelho et al (2022).
5 Coelho et al (2023).
6 Federal Reserve (2023).
7 Crisanto et al (2022).
8 For example, the budget allocated annually to banking regulation, supervision and resolution by all US federal agencies is about 0.03% of US GDP (Federal Reserve (2020), FDIC (2022), OCC (2021)) while the average fiscal cost of banking crises is around 20% of GDP (Borio et al (2020)), as estimated by the increase in public sector debt resulting from such crises. These numbers suggest that, even with a very small reduction of the expected fiscal cost of a crisis (say, 1.5% of the total cost or 0.3% of GDP) the financial impact of substantially enlarging the budget allocated to banking oversight (eg doubling the total budget over 10 years in this example) would pay off.
9 Beerman et al (2021).
