Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

14.1 Introduction

The financial crisis that started in 2007 was not the first and will certainly not be the last that financial supervisors have to face. One important question is whether the toolkit currently available to supervisors is sufficient to allow them to recognise the build-up of vulnerabilities, even during good times when there are no signs of problems, and to take action to manage the risks in good times. What else can supervisors do? And what can society expect of them in the future?

14.2 Supervision in the Limelight

The next decade should be an incredibly important decade for supervisors because the global financial crisis has shone a spotlight on the craft of supervision, separate and apart from the craft of writing new regulations. Regulation is about setting speed limits and mandating the use of airbags. It is about rule-making, such as Basel III capital rules, liquidity rules, and leverage rules – an extremely important activity. Supervision is about oversight of financial institutions’ implementation of these rules. It is about putting up yellow flags to slow things down and trying to ensure that banking is carried out safely without putting depositors and taxpayers at risk. It is about determining whether there could be a breakdown in risk management controls at an institution, and whether the culture of the institution and its appetite for risk will create dangers that could lead to the bank running off the road (i.e., becoming insolvent).

To explain further, supervisory oversight is about the kind of attention financial institutions receive from supervisors on a regular basis. It is about the questions we ask, what we say to institutions, how we say it, the type of information we request, the people we ask to meet, how we deal with push back, what we do when we go on-site or otherwise deal with an institution, and the extent to which we tick boxes or think about the core risks and how they are being managed.

In short, supervisors are the people on the front lines who seek to identify weak risk management systems at individual institutions and decide what to do about them. Supervisors must decide, for example, whether to tell an institution to stop growing a business until its problems are fixed or require an institution to raise more capital to absorb unexpected losses. It is the supervisor who may require an institution to do more stress testing, or require an institution to hire expertise in a particular area, or to spend money on data systems so that risks can be more readily and accurately aggregated and assessed. Such supervisory actions are costly to the institution, but are intended to help make institutions safer and limit losses.

The combination of rules, and supervisory oversight and judgment, are critical.

The link between regulation and supervision is clear when one talks about capital. A bank’s reported capital number is only as good as parties that oversee these numbers. That includes a bank’s senior management, the internal and external auditors, and bank supervisors. Bank supervisors must be on the lookout for practices that inflate the capital position, such as: banks that avoid downgrading bad loans; banks that choose to interpret some capital rules by the letter of the rule versus the spirit; banks that place mechanical reliance on models; banks that write business based on what the capital rules require versus what the real risks are as products and circumstances change (the banks bulk up on business that may be risky but for which capital rules erroneously assign low risk weights); and banks that assume that risk weights assigned under the capital rules are the end of the analysis (for example, banks that assume that sovereign debt has zero risk).

Further, if regulations rather than supervision become the focus, a system with more risk may be created. Rules often have unintended consequences, which can take quite some time to see (see the contribution of Nouy in Chap. 4). Our record in getting rules right is not perfect.

Also, globally, there are many more people involved in supervision than in writing the rules, which suggests we have a vested interest in determining what makes supervision effective. At the same time, supervision is difficult to assess as it is typically carried out behind the scenes. Importantly, it is much more time consuming to change supervision or build supervisory capacity than it is to change a rule. There is no quick fix if a supervisory function is weak.

Supervisors should seize the opportunity provided by the global financial crisis and take advantage of the attention and recognition being paid by the Financial Stability Board (FSB) and others to the craft of supervision and the drivers of an effective supervisory system.

14.3 What Is Effective Supervision?

Under the auspices of the FSB’s Supervisory Intensity and Effectiveness Group (SIE), there have been global discussions about different supervisory approaches around the world, and about how to strengthen supervision.

Despite its importance, there is surprisingly little information on what constitutes an effective supervisory regime. Perhaps this explains why there are some different approaches to supervision around the world. As noted later in this chapter, different approaches are not all bad – the nature of banking systems varies greatly and some trial and error can also be good. But over time, there should be some convergence on what constitutes good practices and what does not.

Some of the differences include:

  • The extent to which supervisors focus on compliance versus risk identifycation and mitigation (the latter is more difficult);

  • The role of the supervisor in corporate governance (some supervisors have more of a hands-off approach; some routinely question and probe directors in an attempt to assess effectiveness; some send observers to sit in on board meetings, and some put potential directors through intense interviews prior to their appointment);

  • The extent to which work is outsourced to consultants or external auditors, versus being performed by qualified in-house staff, which can have positive or negative impacts on quality of work and corporate memory at the supervisory agency;

  • Different styles of oversight (this covers a range of practices, from working behind closed doors with institutions to get problems fixed; working via public admonishment to get problems fixed; working via relationships that are too cozy or, at the other extreme, too toxic);

  • Different levels of attention being paid to issues such as succession planning; oversight of models; oversight of operational risk (some supervisors focus largely on capital while other supervisors are also very active determining whether institutions have processes in place to ensure operational risk is constantly considered); and oversight of capital markets activity; and

  • The attention paid to the risk culture at an institution; and the willingness of supervisors to assess and question the business model of an institution, its planned source of future profits and to engage with the board and management in a review of the adequacy of the institution’s long-term strategy.

14.4 The Fundamentals: Independence, Resources and Mandates

If regulatory and supervisory agencies fail, it may be due to factors including inadequate mandates, inadequate powers, and inadequate independence. A necessary condition for success is having supervisors who have the ability to exercise strong independent judgement. Also, to be an effective supervisor, adequate resources and appropriate mandates are necessary. But FSAPs have shown that these three areas – independence, resources and mandates – are often the areas with the weakest assessment results.

The next decade will not be promising in terms of achieving intensive and effective supervision if these three fundamentals are not in place. Fundamentals, such as independence and properly constructed mandates, target supervisory incentives, which drive supervisory actions and judgements.

Independence

Many supervisory decisions can cost institutions money in the short term (for example, when supervisors require banks to invest in IT systems, or place limits on bank growth). That is why supervisory independence and freedom to make unpopular decisions are important. Independence has long been recognized in central bank/monetary policy literature, where various reports over many years emphasize institution and incentive design as a key ingredient for success. In the case of supervisory agencies, there is a less literature and less attention in practice to this issue, and thus less of a foundation for independence for supervision, although these considerations are also very important for sound supervision

Mandates

Mandates that are geared toward active early intervention can drive behaviour and accountability. As noted by the SIE,Footnote 1 whenever supervisors take an early intervention approach, there are often no tangible risk indicators (i.e. how does one measure the absence of losses) to confirm that this intervention was needed. This makes it difficult to convince firms and their boards that such measures are appropriate to deal proactively with emerging areas of risk in a systemically important financial institution (SIFI). Mandates that create the expectation that supervisors will act early help to set the stage for a healthy tension that ought to exist between the supervisor and the industry. By contrast, mandates that suggest that supervisors ought to promote development of the country as a financial centre, or that place more emphasis on promoting competition within the industry versus prudence, may adversely affect incentives and the ultimate system of regulation/supervision.

Resources – Budgets

From a budgetary perspective, resources includes resources for salaries to obtain the necessary skills and number of staff, resources for IT systems to analyse information from financial institutions, resources for travel and training, and resources for physical premises (a professional looking office). All are important if bright and energetic individuals are to consider making the function of supervision a lifetime career.

As has been noted by the SIE, a model based on industry fees versus one derived from government budgets seems preferable as it helps guarantee a more stable funding source over the cycle and shields supervisory agencies from fiscal vacillations.

While high quality resources in sufficient numbers are needed to conduct intensive and effective supervision, inadequate resources continue to be a problem in many countries. This is despite the obvious cost to national economies caused by weak supervision. It is also despite the costs placed on institutions if they have to spend too much time dealing with supervisors who do not have adequate skills or a good grounding in their business or on the issues.

Some supervisors are much further ahead than others in having the resources they need; even then they often say that they are barely keeping their heads above water as new demands on supervisors are imposed. The necessary resource level is a moving target and it is difficult to put a number on the level of resources needed for new approaches to supervision and for new initiatives such as living wills.

The resources issue is not an issue that will be quickly or easily solved. It is affected by the lack of independence of some agencies to hire the resources they need, by head count and salary constraints that are imposed in some countries, and perhaps even by the tone “at the top” of agencies or central banks. While all of these issues can be solved if there is a will, resource quality is also affected by softer factors.

For example, even with flexibility to hire, it is not usually possible to hire a ready-made “supervisor”. Rather, agencies hire people from the industry or elsewhere and it takes time for such new hires to learn what it means to be a supervisor. On the other hand, some central banks with responsibility for supervision do not feel supervision is a “real” career and require rotation into central banks after 3–5 years.

Industry has suggested that the problem could be reduced by doing interchanges of employees between firms and supervisory agencies. But some financial institutions do not want to see the employees of competitors going to a supervisory agency to learn about the inner workings of all institutions in that country, and then returning to compete against them, knowing all of their practices. As well, some people would raise the revolving door issue, i.e. people moving between industry and regulatory agencies, and whether the judgements made by such supervisors would be affected by their goal of ultimately re-entering industry.

As noted by the SIE in its report of November 2010 (SIE 2010), some supervisors feel that hiring specialist skills from the market is critical, as such people have a perspective that cannot be obtained from being a career supervisor, while others feel that internally home-grown supervisors do the job better and have a more questioning attitude toward market “fads”. Some central banks with responsibility for supervision feel that the challenge of mixing people hired from industry with central bank PhDs is too difficult and opt for training instead. While high-quality training programs are important, they vary significantly from country to country.

Understanding the skills needed is critical. While most would say that a supervisory agency needs people with skills such as credit, market and operational risk, tenure and experience are also vital (such as people who have been through many financial cycles). Having access to financial historians can add value, because if people do not know history it is hard to avoid repeating history.

Even where resources are deemed broadly adequate, new areas are emerging where specialized resources may be in short supply globally, such as more resources to focus on models used by SIFIs (SIE 2010). Another issue is staff turnover; while some turnover is necessary, constantly changing supervisory teams can really affect knowledge of an institution and hamper risk assessment and early intervention.

Unless focus is placed on resources, supervisory intensity and effectiveness will suffer. If supervisory agencies are understaffed or staffed inappropriately, or there are high levels of turnover, and if training is inadequate, the implications for safety and soundness and financial stability can be profound.

Resources – Culture, Presence and Soft Skills

The previous section focused on important issues that many supervisory agencies are trying to address in the area of resources.

But the issue of culture, “presence” and soft skills, such as the ability to effectively communicate with CEOs and directors, deserves its own section. Having supervisors who can tell the chair of the board of a global bank that the board does not have the skills needed, or who can tell a CEO that the bank’s business model does not make sense under various scenarios – especially when there are no losses – requires a set of skills that may not be typically found in many agencies.

Also, while supervisors have a wide range of powers to force action, being able to convince an institution of the need for change (and therefore getting buy-in) leads to a far better outcome, as the institution gets behind the changes versus doing things reluctantly in order to pacify a supervisor.

14.5 Avoiding Supervisory Complacency Through Cycles

The basic expectations of supervisors are covered by the BCBS core principles on banking supervision (BCPs). The crisis revealed that in some cases many supervisors did not meet the BCPs. In other cases, supervisors met the BCPs and received very good ratings under FSAPs, but the problems uncovered by the crisis indicated otherwise. One explanation for this was the methodology for the FSAP rating itself. FSAP ratings were based on supervisors meeting essential criteria in BCPs but not additional criteria in BCPs. The additional criteria really spelled out what supervisors should have been doing, especially in regard to SIFIs.

Such problems have been addressed via stronger BCPs and new assessment methodologies for FSAP assessors. Importantly, more checks and balances are being put in place to monitor what supervisors are actually doing.

More checks and balances seem to be a clear need and hopefully will help address human nature, and the natural tendency to become complacent over time. Indeed, the material we are writing today about the importance of risk management, governance and quality supervision is really no different than what we wrote years ago. We all know these issues are important, but for some reason, the follow-through is weak and we do not get around to implementing what we know to be important. Measures that require us to constantly look over our shoulder – like rigorous and frequent FSAPs, FSB peer reviews, and BCBS peer reviews – make it more difficult for various players to not implement the things they say are important.

One can see this phenomenon in the 5 years since the financial crisis began. The pendulum swung far to one side post 2008, and then back somewhat as banks and even some governments pushed back on capital requirements, and has now swung back again given events in 2012 (“whale” trades, LIBOR manipulation, and US AML violations). The BCPs, and rules such as Basel III, are designed to be a set of core principles or rules for all times (based on the knowledge we have today). BCBS and FSB peer reviews, and FSAPs, are designed to constantly test adherence and implementation. The added emphasis on looking over the shoulders of players should help deal with the swinging of the pendulum and human nature.

Notably, some supervisory agencies which have done well through the crisis have also decided to hire supervisors from countries where the opposite occurred, to ensure that the agencies do not become too complacent about their abilities to recognize risks.

14.6 Supervisors and the Link to Economic Research

Major efforts are currently being put into research – research about adequate capital levels; research about how risks are transmitted across financial institutions and markets, and the feedback loop between the financial system and the economy; research into optimal size of financial systems as a percentage of domestic GDP; etc. Supervisors need to be aware of and contribute to this work from their own perspectives.

Some of this research could make the job of supervisors easier. If early-warning indicators, such as credit growth, provide leading information about banking crises, or improve our assessment of risks to financial systems, not only would supervisors be better informed, but they would have even more evidence to back up unpopular supervisory decisions (such as explaining why capital needs to be increased).

But supervisors also need to be cautious about these developments. Before the crisis there were many debates about optimal regulation. For example, before the crisis, some academics and regulators promoted the efficient markets theory, which led to “light touch” regulation in some countries, which subsequently led to major problems in these countries’ financial systems.

Supervisors need to be sceptical and ask questions about everything we are told – not only by banks but also by others, including academics and researchers.

Financial system modelling is in its infancy, as is research on the build-up and bursting of bubbles. We clearly need to advance this thinking, while recognizing that the real world is ever-changing, dynamic, and innovative, with no complete understanding on the part of macroeconomists and regulators and supervisors of how the parts acting on their own will affect the system.

In an effort to promote stability, academics and researchers try to model the real world but they have to make many assumptions about behaviours under stress. Given complexities, some may assume away a lot of things – like taxes, bankruptcy costs, agency costs, and asymmetric information. Some may assume away some of the complexities that arise due to the fact that real people run firms – people with egos, people with all sorts of incentives, and people who do things that are not always consistent with standard economic theory. This is where supervisors come in – they can see what is happening on a day-to-day basis within firms and observe various behaviours. Here, again, in practice the idea is to have different perspectives and different skill sets informing decision-making.

Supervisors should work to add value to these efforts by informing themselves of the research being done and providing their views on what goes on day-to-day in banks. Above all we need to be cautious as theories can be proven to be incorrect.

At the same time, supervisors cannot go it alone and need to better leverage other talent pools. A blended approach involving front line supervisors, economists, and quants is now often seen, because bringing different perspectives to the table results in more accurate problem identification and more thoughtful solutions to problems. To be successful, however, such teams will need to bring matters to conclusion quickly so that supervisors can take action early to respond to issues. It should be noted that managing such teams successfully can be challenging.

14.7 New Vulnerabilities Will Arise from Solutions Advocated Today

Some suggest that new measures agreed to by BCBS and FSB have fixed the problems that led to the global financial crisis. On the other hand, some suggest that Basel III is too complex and needs to be greatly simplified.

What we can agree on is that new vulnerabilities are likely to arise as a result of the changes we are making to the system today. We must constantly be on our guard to identify these vulnerabilities.

An example is Centralized derivatives clearing. This is a critical initiative, but also one that poses risks if central counterparties are not appropriately risk proofed. Thus, risk-proofing must be a focus of efforts on all fronts and we must be vigilant in detecting activity that has resulted from our decisions which may require a further response from regulators and supervisors.

Macro stress testing, which all supervisors are embracing, is another example. Macro stress testing should give both supervisors and financial institutions more information, which is always useful. But stress tests can also lull us into a false sense of security. This is because macro stress tests probably are unable to provide a realistic picture of the dynamics of distress, especially the adverse effects. Indeed, a shock is called a shock because the unexpected happens – the system does not behave the way you think it might, making it much more vulnerable than a stress test might suggest. Partly, this is because stress tests, by their nature, focus on tail risk and extreme events, which we don’t understand very well and are difficult to model. Further, even where we might have such data, tail-risk events are unique and vary over time, so history alone can be an unreliable guide. Importantly, stress tests generally focus on the first-round impacts, and models can have difficulty representing how behaviour might evolve after such initial impacts – and it is often those difficult-to-predict behavioural changes that cause the most trouble in real crises. Stress tests, especially those without sufficient regard for the preceding considerations, can also show the system to be highly resilient when it is not. So, as always in supervision, reliance on multiple information sources and an asymmetric regard for downside risks are important.

The key benefit of stress testing exercises is that the results provide a platform for supervisors to have critical discussions with banks, informed by stress test data. In other words, it is about the process, and the discussions supervisors have with the banks that matter the most – the numbers that come out of the exercise are of secondary importance given all the assumptions that go into the exercise and the unavoidable uncertainty that characterizes analysis of such tail-risk events.

Firms and supervisors should spend as much time on why the stress tests results might be wrong – resulting from things that might not play out as they expect – as they do in taking comfort from the results. In short, the limitations of stress testing must be understood or a false sense of invincibility, or a false sense of crisis, can arise.

14.8 Communication with Industry

The extent to which supervisors communicate with industry at senior levels about what is going on at the institution and in the industry is very important. Communication between knowledgeable senior supervisors and senior financial institution representatives can be a good way to challenge the views of both supervisors and institutions about current practices and risks. Supervisors have valuable knowledge about risk management practices across the industry and this is information that is harder for institutions to obtain. And institutions themselves see practices day to day in the industry or in their own institutions that can greatly increase a supervisor’s awareness. Conversations about such matters can be extremely rewarding when they lead to new insights.

Communication should not be confined to junior levels. Contact with industry may be ineffective if all contact is at junior levels and is seen as more of a box ticking exercise.

Different types of communication should also be sought. For example, while on-site reviews will include a necessary degree of cross-examination (i.e. posing questions and verifying answers), opportunities should be sought for different types of conversations (less exam focused) which, given their less formal nature, can pave the way for more wide ranging discussion, and thus a more complete understanding of the institution and its culture.

Contact with industry may be ineffective if the relationship between the supervisors and financial institutions is strained to the point of being toxic. While supervisors must be sceptical and tough, the willingness to seek out views of the industry and to listen are as important.

14.9 Internal Processes to Identify Risks

Having methods and processes in place to ferret out important information from the markets and industry and to digest such information and determine whether it is important pays dividends. Supervisory Committees like “Emerging Risk Committees” to exchange ideas and views about risk can be very helpful.

At the same time, prior to the crisis there was much discussion about emerging risks and many believed that the way the system had responded to various disturbances prior to that showed the system was resilient. Again, this seems to force supervisors back to the basics – nothing can replace basic “roll-up-the sleeves” old-fashioned supervisory processes.

Up to now, many supervisors have been trying to catch up to expectations that had existed all along (e.g. many did not review acquisitions even though this was a Basel Core Principle, many did not have the resources to carry out what BCPs suggested they should be doing). Beyond that a range of new demands has been placed on them such as preparing recovery plans and focusing on resolution. New activities are also being adopted by virtue of the enhanced knowledge and experience that oversight of a dynamic industry brings. In some countries, stress testing is becoming far more advanced, oversight of models is being treated far more seriously (including pillar 1 models and more recently models outside of pillar 1), and more use of horizontal reviews is being promoted. Oversight of operational risk is also changing as supervisors recognize its importance. Governance and succession planning processes are also moving to the forefront, as is risk appetite and risk culture.

Indeed, as supervisors around the world have set out to “up their game”, a variety of approaches are being tested and different supervisors are doing different things. For example, some supervisors may spend more time on assessing the people in key functions such as the CRO role, while others might spend more time on doing reviews of specific activities and coming to conclusions on the strength of the CRO in that fashion. Others may do a hybrid approach. Some supervisors may have permanent offices on site at banks while others may continue to prefer a model where offices are at a supervisory agency that is relatively close to the bank to facilitate interaction. Some may prefer to interview and approve new directors and senior management before they can assume their duties, while others would prefer to allow the institutions to make such decisions, and only act after the fact if performance issues arise. Some supervisors sit in on board meetings; others prefer to increase their interaction with board members via separate meetings with supervisors. Some supervisors do not think a separation of Board Chair and CEO is needed.

Another area receiving attention is the thinking about risk based supervision. Risk based supervision means supervisors are not looking at everything – they have to pick and choose their area of focus. This will continue to be the case, as supervisory teams are small in number relative to the number of bank employees in risk management, internal audit, and compliance areas. In the years preceding the crisis, most supervisors focused on the quality of control functions in banks, such as the quality of people in risk management and the depth of their work. This was aimed at ensuring that banks self-policed themselves. Post crisis even more efforts are being devoted to this as supervisors focus on boards of directors, and delve more deeply into areas such as succession planning within control functions. At the same time, supervisors are asking whether the focus on self-policing by control functions is enough, and whether some old-fashioned approaches such as basic financial analysis, and follow-the-money analysis is needed, especially for SIFIs. This will continue to be discussed.

Institutions and others many complain about diversity in supervisory approaches. Institutions would like convergence of supervisory approaches and rules. They would like colleges of supervisors for banks to agree on the issues. They do not want different messages from different supervisors.

But supervisors are not always going to agree. Perhaps this is not necessarily a bad thing. Just as financial institutions say that it is unwise for all firms to have the same view on risks and the same strategy, it could be unwise if all supervisors looked at things identically.

It is also important to emphasize that, while a variety of new approaches are being explored, the major priority needs to be placed on fully implementing all of the long-standing Basel Core Principles on Effective Supervision. It is impossible for supervisors to get their arms around a bank by focusing on one new thing, such as focusing on boards of directors, or on who the CRO is, or on risk appetite statements, or on models, or on building the supervisory process more around stress testing and economic analysis.

Looking forward to the next decade, the challenge will be what it has always been – determining when to intervene, and how to intervene. This can be hugely judgemental. Decisions on the exact supervisory approach to follow will also require judgement, and will vary country by country. To be successful, however, there should continue to be a sharing of views and experience.

Will supervisors be seen to have done enough? When the next crisis hits, inevitably there will be questions about whether supervisors were “asleep at the switch”. One of the challenges of the job is that the value that supervisors bring is not always measureable (one cannot measure losses that did not occur, or crises that did not occur). Further, there is very little public understanding of what a supervisor does (see the contribution of Adams in Chap. 7). The fact that supervisory teams are dwarfed by the teams at banks in risk management, audit and compliance, means that the supervisor will typically be in less of a position to spot risk than the bank itself, but the public often thinks otherwise. This is a challenge that is very difficult to overcome. More public discussion of our role might help.

14.10 Conclusion

Many changes are being made as a result of the global financial crisis. Various proposals have often come with language about how effectively the problems of the past have been, or will be, dealt with. But in reality there is no single element one can point to that will guarantee financial sector stability. Strong financial systems reflect many different factors, including multidimensional oversight of banks by multiple parties (e.g. oversight by regulators and supervisors via strong rules and robust daily supervision; oversight by the market (investors, analysts, rating agencies); oversight by bank management; oversight by bank boards; oversight by external auditors who provide audited financial statements; and policy setting action by central banks and governments in the form of sound macro and micro policies). Debates continue about whether the solutions developed by various bodies – including BCBS and the FSB – are adequate.

It is virtually impossible for one party to do the job – not the CEO, not the board, not the regulator, not the supervisor, not investors with money at stake, not analysts poring over disclosures, not central banks and not governments. All parties play a role and must carefully perform their critical functions. And they must avoid creating incentives that affect the performance of these roles.

Anything that sharpens incentives for the market to monitor financial institutions, for financial institutions to manage their risk and for supervisors to act early, is important. An openness to consider the ideas of all members of the supervisory community is also important, as are any measures that force action even when complacency sets in, as it inevitably will.