Strong and modern supervision in a fast-moving world

Introduction

Your Excellencies, Governors and Heads of Supervision (past and present), distinguished members of the Basel Committee (past and present), ladies and gentlemen: good morning, and welcome to the 24th International Conference of Banking Supervisors (ICBS). 

Let me start by thanking Governor Destry Damayanti and Bank Indonesia, and Chair Friderica Widyasari Dewi and the Indonesian Financial Services Authority, for their warm hospitality and for bringing together supervisors from across the world. It is a pleasure to be in Bali this week. 

When I arrived in Bali earlier this week, I couldn’t help but notice that most of the people around me appeared to be on holiday. It quickly became clear that they were heading for very different conversations and activities than the ones taking place at this conference. While they were discussing beaches and cocktails, we are here to discuss capital buffers and supervisory effectiveness. 

At first glance, it may seem that one group drew the better deal. Yet there is perhaps something fitting about discussing supervision here. Bali has spent centuries navigating powerful currents, shifting conditions and unexpected storms. And in many ways, that is precisely the challenge facing supervisors today. 

Indeed, Indonesia’s history offers a useful lesson for all of us gathered here today. For centuries, it sat at the crossroads of some of the world’s most important routes. Goods, people, ideas and technologies flowed through these islands, linking Asia, the Middle East, Africa and Europe.¹  Success depended not on resisting change, but on navigating it. Not on standing still, but on adapting while preserving stability. 

That challenge feels remarkably familiar today. The world around us is moving faster than at any point in recent memory. Technology is reshaping financial services. Artificial intelligence (AI) is transforming how banks operate and how risks emerge. Cyber threats continue to evolve. New forms of financial intermediation are changing the structure of the financial system, and with it the distribution of risk. Geopolitical tensions are increasingly clouding the outlook. And the evolving political economy landscape is questioning the role of independent central banks and supervisory authorities. 

In such a world, the existential question for supervisors is: how do we remain effective when the environment around us is changing so rapidly? 

My answer is as follows: we need supervision to be both strong and modern. 

Too often, these terms are set against each other. To its critiques, “strong” supervision is often interpreted as a rigid, backward-looking approach that represses innovation and a risk-taking culture. Equally, “modernisation” is at times euphemistically used to hint towards a lax, bank-friendly mindset that prioritises short-term objectives such as economic growth over resilience and stability. 

Yet my message today is that strong supervision and modern supervision are not competing objectives. They are mutually reinforcing. And they will be essential if we are to preserve trust, resilience and economic prosperity. 

Regulation: necessary but insufficient 

The Basel Committee is often associated with regulation. The first thing that often comes to mind is “Basel III” or whatever trendy domestic variant labelling is preferred (which range from decimal tenths, including Basel 3.1 and Basel 3.5, to Basel IV or even apparent Hollywood blockbuster references).²  Online searches for the Committee’s regulatory outputs exceed those of its supervisory ones by over 2,500%.³

That is perhaps understandable. Over the past 16 years, the Committee’s work has rightly focused on shoring up banks’ resilience by fixing the regulatory fault lines exposed by the Great Financial Crisis.⁴  The Basel III reforms have unquestionably strengthened banks’ ability to absorb shocks and continue to lend to households and businesses. Let me mention just a few examples:

• Banks have significantly increased their loss-absorbing capital positions. Since 2011, banks’ Common Equity Tier 1 (CET1) risk-based ratio has increased by over 40% and now stands at 14%.⁵
• Basel III introduced complementary capital metrics, in the form of the leverage ratio and the output floor. These aim to offset some of the shortcomings of the risk-based capital framework, and provide a healthy check on bank leverage. Indeed, at a system-wide level, the amount of debt in the global banking system relative to loss-absorbing capital has almost halved, from 30x to 17x.⁶  
• Liquidity positions have also strengthened. The share of high-quality liquid assets held by banks stands at over 20%, and banks rely significantly more on stabler sources of funding.⁷

These are not merely improvements on a spreadsheet. Stronger capital and liquidity positions helped the banking system withstand a global pandemic, the sharpest tightening of monetary policy in decades, a series of geopolitical shocks and the banking turmoil of 2023. Basel III helped avoid a repeat of the Great Financial Crisis in each of these instances. It is therefore crucial that we implement all outstanding Basel III standards in full and consistently, and as soon as possible.⁸

But it would be a mistake to conclude that strong regulation has made strong supervision less important.

The history of banking crises, including the 2023 turmoil, tells us why. Time and again, we saw that failed or distressed banks met their minimum capital requirements. Some also met applicable liquidity requirements. Yet beneath apparently satisfactory regulatory ratios sat deep weaknesses in governance, risk management and business models.⁹

Those qualitative weaknesses eventually became quantitative shocks. Poor interest rate risk management produced losses. Concentrated and unstable funding structures produced deposit runs. Weak governance allowed vulnerabilities to grow unchecked.

To be clear, regulatory ratios did not cease to matter. But they did not tell the whole story. A ratio tells us where a bank stands at a particular moment; it is a snapshot. But supervision is the film. Supervision asks how it got there, where it may be heading and how it is likely to respond when conditions change.

The distinction matters because risks rarely arrive in neat regulatory categories. They often appear first in incentives, behaviours and concentrations. They emerge in the quality of governance, the credibility of internal challenge and the assumptions embedded in a bank’s strategy. Only later do they show up as losses, outflows and distress.

The absence of a regulatory breach is therefore not evidence of the absence of risk. Nor is compliance with minimum standards a substitute for sound management. Regulation establishes the guardrails. Supervision seeks to keep the vehicle on the road.

This brings me back to the full name of the Basel Committee. We are not the Basel Committee on Banking Regulation. We are the Basel Committee on Banking Supervision. Our mandate is to strengthen the regulation, supervision and practices of banks worldwide. So we need both effective regulation and supervision to safeguard global financial stability. 

What makes supervision effective?

My question for you today is whether our supervisory systems are equipped to deliver in the world now taking shape.

That requires us to be clear about what we mean by effective supervision. The Committee’s own work in this area defines supervisory effectiveness in outcome-oriented terms. Effective supervision means promptly assessing prudential risks, identifying material shortcomings and using the available toolkit to ensure that banks and the banking system more generally address those shortcomings in a timely manner.¹⁰

Three verbs matter here: see, say and act. Supervisors must see risks early. They must say clearly what needs to change. And they must act when change does not happen. Each step is necessary. None is sufficient on its own.

A supervisor may identify a risk accurately but communicate it too cautiously. A finding may be communicated clearly but followed by no meaningful escalation. A formal action may eventually be taken, but only after the vulnerability has become too large or too costly to correct.

Timeliness is therefore not an administrative detail. It is part of the policy. An unresolved supervisory finding is not an outcome. It is a record of unfinished business.¹¹

The “act” part is perhaps the hardest. Supervisors need sufficient judgment to intervene before risks have fully crystallised, but that judgment must be anchored in a clear mandate, evidence and due process. Too vague an approach risks inconsistency and legal challenge. Yet too mechanical an approach risks the opposite problem: box ticking, excessive caution and an inaction bias in which supervisors wait for a rule to be breached before acting.

So effective supervision sits between these extremes. It uses judgment, but disciplines it. It provides discretion, but makes it accountable. And, above all, it gives supervisors the confidence to act while action can still make a difference. 

This is one of the clearest lessons from past bank failures. Warning signs are frequently visible before failure. The problem is often not an absolute absence of information, but a failure to connect that information, draw sufficiently strong conclusions and secure timely remediation.¹²

Yet supervision is an unusual public function. When it fails, the results can be spectacular. When it succeeds, the results are often invisible. In financial stability, success is an orphan.

This creates an awkward asymmetry. Supervisory failures are visible and measurable. Supervisory successes are frequently counterfactual. It is difficult to prove that an intervention prevented an adverse outcome..¹³

This asymmetry can also distort the public debate. After a banking crisis, supervision is criticised for doing too little. Between crises (and it is only a question of “when”, not “if”, a future banking crisis occurs), it is often criticised for doing too much. When memories of instability are fresh, people demand vigilance. When memories fade, vigilance is recast as unnecessary interference and a hindrance.

Effective supervision therefore depends on more than technical skills. It requires the will to act when the evidence is incomplete, the benefits are difficult to observe and the immediate costs are highly visible.

That is not easy. But it is the essence of the job.

Judgment is not a dirty word

This brings me to supervisory judgment.

Judgment has become an uncomfortable word in some policy discussions. It can be associated with subjectivity, inconsistency and opacity. Rules appear precise and predictable. Judgments can appear harder to explain and easier to contest.

Those concerns deserve to be taken seriously. Supervisory powers must be exercised responsibly. Decisions should be rooted in evidence, subject to appropriate governance and open to internal challenge. Banks should understand what is expected of them and why.

But we should not confuse judgment with arbitrariness. Nor should we assume that less judgment automatically produces better supervision. 

In a complex and changing financial system, the opposite may often be true. A supervisory system with no judgment would require an infinite rulebook. Every risk would need to be anticipated. Every possible combination of circumstances would need to be classified. Every appropriate supervisory response would need to be specified in advance.

Rules are necessarily general. Banks are necessarily specific. Two banks with identical capital and liquidity ratios may have very different vulnerabilities because of differences in their business models, depositor bases, governance arrangements, risk controls or operational dependencies.

The relevant choice is therefore not between judgment and no judgment. It is between good judgment and bad judgment.

Good supervisory judgment is not instinct dressed up as policy. It is a structured professional assessment. It combines quantitative evidence with qualitative information, considers plausible alternative explanations and asks what evidence might cause the initial conclusion to change. Indeed, recent empirical evidence suggests that judgment allows supervisors to incorporate forward-looking information that standardised indicators may miss.¹⁴

Good judgment should also be contestable within the supervisory authority. Important conclusions should be tested by people who did not form the original view. Difficult cases should be escalated. Supervisory teams should be asked not only what they know, but what they might have missed. Judgment improves when it meets resistance.

It must also be accompanied by institutional accountability. Decisions should be documented. The reasons for acting, or for not acting, should be clear. Authorities should examine after the event whether interventions achieved their intended outcomes. Judgment is not an escape from accountability. Accountability is what gives judgment legitimacy.

Capacity is policy

If judgment is central to effective supervision, supervisory capacity is central to good judgment. This is a key expectation set out in the Basel Core Principles.¹⁵

Capacity is sometimes treated as an administrative matter: headcount, budgets, recruitment, training and information technology. These issues may sound less consequential than capital standards or liquidity requirements.

They are not. Capacity is policy.¹⁶

A supervisory framework is only as strong as the institution applying it. An authority may have wide legal powers, detailed methodologies and access to large quantities of information. None of these guarantees that risks will be identified, communicated and addressed in time. 

The decisive factors are often human and organisational. Do supervisory teams have the expertise to understand a changing business model? Can they connect risks across traditional silos? Do concerns travel quickly through the institution, or are difficult messages softened as they move up the hierarchy? Above all, are supervisors willing and able to act?

Capacity should therefore be understood broadly. It includes people, but not only people. It includes specialist expertise, institutional memory, analytical tools, legal powers, operational independence and organisational culture.

It also includes time. A supervisor overloaded with procedures may have too little time left for supervision. A team consumed by collecting information may have too little capacity to interpret it. An authority that measures success by the number of reviews completed may overlook whether those reviews changed behaviour.

And yet assessments conducted against the Basel Core Principles repeatedly identify inadequate independence, resources and specialist skills as impediments to effective supervision. So what more can be done?

Investing in the supervisor of the future 

This is where modernisation kicks in. Modernisation should free supervisors to think, not simply enable them to process more. It should support supervisors in investing in the capabilities required as we deal with an ever changing banking system. 

To be sure, tomorrow’s supervisor will still need to understand capital, liquidity, credit risk and interest rate risk. These fundamentals do not become less important because technology is changing. In many cases, technology simply alters the speed with which familiar risks emerge and the form they take.

But traditional prudential expertise will no longer be enough on its own.¹⁷  Supervisors increasingly need to understand technology architecture, cloud dependencies, cyber security, data governance and AI. They need to follow the links between banks and non-bank financial intermediation. They must understand how vulnerabilities can travel across institutions, markets, service providers and borders. And they must be increasingly well versed in geopolitics, trade policy and supply chain resilience. 

The supervisor of the future will need to speak at least three languages: finance, technology and human behaviour. Finance reveals the balance sheet. Technology reveals the network. Behaviour reveals how both may respond under stress.

This does not mean turning every supervisor into a software engineer or behavioural scientist. It means building multidisciplinary teams and ensuring that specialist insights are integrated into the overall assessment of a bank. 

That integration is crucial because risks do not respect organisational charts. A cyber incident can become a liquidity event. A third-party disruption can become an operational and reputational crisis. Weak governance can amplify interest rate risk. A concentrated business model can become unstable when information spreads rapidly through social media.

Supervision therefore needs to modernise to become more forward-looking without becoming speculative, more data-driven without becoming data-dependent, and faster without becoming careless.

That is a demanding combination. It will require sustained investment. There is also a more prosaic challenge: finding the people. Supervisory authorities are competing for many of the same skills sought by banks, technology firms, consultancies and other parts of the private sector. That competition is particularly acute in areas such as cyber security, data science, cloud architecture and AI, where specialist expertise is scarce and compensation gaps can be large. Public authorities also face budgetary constraints, lengthy recruitment processes and restrictions on how specialist expertise can be rewarded.

These constraints are real. But so are the costs of underinvestment. Well resourced supervision is not a burden placed on the financial system. It is part of the infrastructure on which a stable financial system depends. Supervisors must find ways to attract, develop and retain people with the skills needed to understand an increasingly complex financial system. 

The risks themselves are also becoming more difficult to assess. For example, AI can improve efficiency and decision-making, but it can also create new vulnerabilities around data quality, model risk, explainability, cyber security and accountability. Its use may deepen banks’ reliance on a relatively small number of third-party providers of cloud, models, software and data. The Committee’s recently finalised Principles for the sound management of third-party risk offer guidance on holistic third-party risk management for banks and supervisors.¹⁸

And there is a particular supervisory challenge for smaller banks. They have the most to gain from external AI tools, but the least in-house capacity to understand, challenge and govern them. The paradox is therefore that those with the least internal capacity may become the most dependent on external expertise. Supervisors will therefore need to assess not only what technology a bank is using, but whether its governance, controls and people are capable of using it safely. Proportionality matters, but proportionality cannot mean outsourcing responsibility. 

Three commitments for modern supervision

Let me suggest three commitments for the supervisory community.

First, we should invest in our institutions. That means protecting clear mandates and operational independence, maintaining adequate resources and ensuring that supervisors have the legal authority to intervene early. These foundations cannot be created during a crisis. They must be established beforehand. 

Second, we should invest in our people. Supervisors need the expertise to understand evolving risks, the time to form judgments, the confidence to challenge senior bankers and the institutional backing to escalate concerns. Tools matter. Culture matters more.

Third, we should invest in trust. Supervisory judgment must be evidence-based, structured and accountable. Authorities should explain how they operate, how decisions are governed and how they assess their own effectiveness. The legitimate need for confidentiality should not become a shield against accountability.

These commitments reinforce one another. Skilled supervisors cannot be effective without adequate powers. Strong powers cannot command trust without accountability. Technology cannot improve outcomes without capable people and sound governance.

That is what it means to be both strong and modern.

Global cooperation as supervisory capacity

There is a reason why we are all gathered here today. No authority can build all the necessary capabilities alone.

Global cooperation is itself a form of supervisory capacity. It enables authorities to pool knowledge, compare practices, identify emerging vulnerabilities and understand how risks are developing across borders.

That has been central to the Basel Committee since its creation. Since its inception, the Committee has been the main global forum for bringing supervisors together to exchange experiences and learn from what works – and what does not – across jurisdictions. As risks evolve more quickly and increasingly cut across borders, that practical cooperation can help build supervisory capacity, spread good practices and ensure that supervisors do not have to confront new challenges. Most recently, the Committee has developed practical supervisory tools on topical issues to foster a better understanding of diverse supervisory approaches and practices worldwide, benefiting both supervisors and banks.¹⁹  

Global cooperation is also central to the ICBS. The first ICBS was held in London in 1979. Its purpose was not to take votes or adopt formal resolutions. It was to build understanding among supervisors confronting common problems from different starting points.²⁰  That purpose remains as relevant today as it was then.

The value of the ICBS lies partly in its formal sessions. But it also lies in the conversations outside them: a supervisor learning that a problem thought to be local is emerging elsewhere; an authority discovering a different way to organise its work; a relationship established before a crisis makes cooperation much easier in times of stress.

This is supervisory infrastructure. We should continue to invest in it. Financial risks are becoming more interconnected at precisely the time when international cooperation is becoming more difficult.

Fragmentation in supervision would ultimately contribute to fragmentation in finance. Information gaps would widen. Opportunities for arbitrage would grow. Cross-border risks would become harder to identify and manage. The case for cooperation is therefore not based on institutional tradition. It is based on practical necessity.

Conclusion

In conclusion, a fast-moving world does not make the traditional foundations of banking supervision obsolete. It makes them more valuable. Capital matters. Liquidity matters. Sound governance matters. Sustainable business models matter. Early intervention matters.

But the way we deliver these outcomes must evolve. Supervisors must navigate not only financial cycles, but political and institutional ones. These require us to ensure that supervision is both strong and modern. 

Strong supervision means having the mandate, independence, powers and determination to act. Modern supervision means having the people, information, technology and methods to act intelligently.

We should be confident about the role of supervisory judgment, but never casual about its exercise. Judgment must be structured, evidence-based and accountable.

And we should treat investment in supervision as investment in financial stability. The quality of our people, institutions and cooperation will determine whether the standards we write work when they are most needed.

Indonesia’s history shows the value of navigating change without losing one’s anchor. The currents may shift. The weather may change. The destination must remain clear.

I cannot promise that the remainder of this conference will make capital buffers more attractive than Bali’s beaches.

But I am certain of this: in a fast-moving world, our strongest safeguard is moving together. Anchored by Basel standards, guided by our shared supervisory experiences and delivered through cooperation. 

Thank you.

References

Adrian, T, M Moretti, A Carvalho, H Kyong Chon, K Seal, F Melo and J Surti (2023): “Good supervision: Lessons from the field”, IMF Working Papers, no 2023/181. 

Anderson, B (2016): Imagined communities: reflections on the origin and spread of nationalism, Verso.

Badev, A, L Gutiérrez, S Lopes, K Duellmann, Y Endo, D Foos, L Rosello, S Palligkinis, R Oliveira, N Tanaka, F Vaghefi and T Vietin (2025): “Lessons on supervisory effectiveness – a literature review”, Basel Committee on Banking Supervision Working Paper, no 45. 

Basel Committee on Banking Supervision (2015): Report on the impact and accountability of banking supervision. 

——— (2023): Report on the 2023 banking turmoil.

——— (2024a): Core principles for effective banking supervision. 

——— (2024b): Digitalisation of finance. 

——— (2025a): Principles for the sound management of third-party risk. 

——— (2025b): Supervisory newsletter on supervisory issues.

——— (2026a): “Governors and Heads of Supervision welcome progress to implement Basel III and discuss elements of the Basel Committee’s work programme”, media release, 9 March. 

——— (2026b): “BCBS dashboards”, available online. 

Berthonnaud, P, E Cesati, M Drudi, K Jager, H Kick, M Lanciani, L Schneider, C Schwarz, V Siakoulis and R Vroege (2021): “Asset encumbrance in euro area banks: analysing trends, drivers and prediction properties for individual bank crises”, ECB Occasional Paper Series, no 261. 

Board of Governors of the Federal Reserve System (2023): Review of the Federal Reserve’s supervision and regulation of Silicon Valley Bank.

Bobeică, G and S Oprică (2026): “The role of judgment in supervisory scores and additional capital requirements assigned to banks”, European Central Bank Working Paper Series, no 3233. 

Borio, C, M Farag and N Tarashev (2020): “Post-crisis international financial regulatory reforms: a primer”, BIS Working Papers, no 859. 

FINMA (2023): “Lessons learned from the CS crisis”, FINMA Report, December. 

Hernández de Cos, P (2024): “Building on 50 years of global cooperation”, speech at the 23rd International Conference of Banking Supervisors, Basel, 24 April. 

House of Commons Treasury Committee (2008): The run on the Rock.

Reid, A (1988): Southeast Asia in the age of commerce, 1450-1680, volume one: the lands below the winds, Yale University Press. 

Ricklefs, M (2001): A history of modern Indonesia since c. 1300, Stanford University Press.

Thedéen, E (2024): “Charting the course: prudential regulation and supervision for smooth sailing”, keynote speech at the Institute of International Finance Annual Membership Meeting, Washington DC, 23 October. 

——— (2025a): “Resilient by design: why strong rules still matter”, speech at the International Business Forum / International Conference on Financing for Development, Sevilla, 1 July.

——— (2025b): “Resilience pays: the strategic value of regulation and supervision”, keynote speech at the Eurofi Financial Forum, Copenhagen, 19 September.

——— (2025c): “Implementation, innovation and interconnections”, keynote remarks at the 2025 Institute of International Finance Annual Membership Meeting, Washington DC, 15 October.