Ingves: Is there an optimal way to structure supervision?

Introduction
Thank you for inviting me to this 4th Summit Meeting of the Islamic Financial Services Board. The issues of regulation and supervision are generally the same in a system of Islamic financial services as in any other financial system. You need the same authorities and the same regulations to ensure that your financial institutions and markets are safe and sound.

My presentation is about finding the best way of structuring the financial regulatory authorities in a country. This issue has been hotly debated for a long time because of the important and political issues involved and the discussion is still going on. I will discuss the matter based on experiences from countries around the world, including situations in which I have been personally involved such as when a country reconsiders its institutional set-up after a crisis. First I will describe the alternatives for supervision that countries have chosen. Then I will discuss how certain key aspects for supervision, such as independence, accountability, transparency and efficiency can be accommodated under various supervisory structures. Third, I will debate the arguments for and against having supervision inside a Ministry or in a national central bank, which I will call NCB for short. I will also evaluate the arguments for and against combining various supervisory authorities.

My main conclusion is that there is no optimal solution for all cases; it depends on the circumstances in the individual country at the present moment. Ensuring operational independence, accountability and transparency for the supervisory agencies are more important than how they are structured. If your country has a well functioning banking supervision do not change its structure until you have clear arguments to do so. Do not make structural changes when the financial sector is weak or recovering from a crisis. On the other hand, developments in your financial sector such as an increasing role for other institutions than banks, argue for closer ties between the supervisory agencies.

There are two basic issues in the discussion on the organisation of supervision:

- Should supervision be inside or outside the NCB and the Ministries?
- Should supervision of different sectors be unified in the same authority?

These questions will be discussed later in my presentation. Let me first describe some of the present practices in organising supervision.

Present practices
Currently there is a great diversity of supervisory structures but the majority comes under one of the two main streams:

The first is the traditional one with bank supervision in the NCB and securities supervision in an independent standalone authority. Insurance supervision is in many countries also located in a separate authority, but in other cases it is a department of the Ministry of Commerce or some other ministry.

The second main stream is full unification of the supervision of banks, securities companies and insurance companies, all in one institution.

The trend since some 15 years is that bank supervision in a number of countries has been moved outside the NCB and has been consolidated with the supervision over securities and insurance companies. However, the most common structure is still the traditional one.(1)

But there are many variations. One is the combination of banking and securities supervision, which is logical due to the close relations between banking and securities activities. Another is the combination of securities and insurance supervision with the aim to create a stronger authority with more resources and expertise.

There are also some broader alternatives. The FSA in the UK performs supervision over a wide range of financial institutions, including brokers, pension funds and many more. Its objectives also include financial sector consumer issues and even the promotion of London as a financial market place. All kinds of supervision and other activities related to the financial sector are placed under one roof.

The Netherlands, Australia and others have followed an approach based on the purpose of the supervision. In these countries prudential supervision of banks and securities firms is conducted separately from the oversight of the conduct of the financial markets. The underlying rationale is that these activities are different. Ensuring the safety and stability of banks and securities firms requires different regulations and skills from monitoring market conduct and taking enforcement actions.

Ireland, France, Finland and others have retained bank supervision in an organisation closely related to the central bank but not in the central bank itself. In this way they benefit from the resources of the NCB and can make some cost savings, but avoid some of the arguments against locating supervision in the NCB itself.

Other countries, among them Germany, have a stand-alone authority for financial regulation but the central bank maintains a role in the conduct of supervision by conducting most of the on-site inspections of banks. In this way the Bundesbank benefits from its vast network of regional branch offices and its knowledge of the local economy.

Criteria for evaluating the organisation of supervision
The choice of structure for supervision is not an end in itself but a means to achieve certain objectives. We should therefore analyse the conditions that must be in place to ensure good supervision.

The first one is operational independence. There must be laws to prevent policy makers or financial market participants from trying to influence the operational decisions of the supervisor. Independence also implies sufficient and stable resources to attract skilled staff, to build good IT-systems, and to inspect supervised entities. The head of supervision should be independently appointed and should not be dismissed for other reasons than those provided in the law. The head and the staff should be protected against frivolous law suits, as long as they conduct supervision in good faith.

Traditionally, independence has been best protected when supervision is conducted in an NCB since the NCB itself is independent due to its role in monetary policy. The NCB also has its own revenues and budget. Outside the NCB, independence must be ensured through explicit legislation and through clear rules for obtaining the necessary resources, e.g. by levying fees on the supervised entities. Contrary to supervision located in the central banks, the supervisory sections which are entities of Ministries are by definition not operationally independent and their resources are dependent on budget appropriations.

The second condition for good supervision is accountability. The authority must regularly provide information about its performance in various forms, such as Parliamentary hearings, speeches by the management, written reports and websites. On these occasions the authority describes and explains its actions and stands ready to receive criticism.

There is no specific structure of supervisory authorities that automatically leads to better accountability so clear rules must be set regardless of how the authority is structured.

The third condition is transparency. The general public and in particular the supervised entities need to have full information about regulation and supervision. The authority must consult with interested parties to ensure that efficient laws and regulations are drafted and after implementation they should cooperate with the market participants and others to ensure that the rules are well understood and implemented. Transparency is traditionally not a focus of supervisors, so it needs to be explicitly enacted, independently of how supervision is organised.

The fourth condition is the efficiency and efficacy of supervision. Does the authority perform its services at a reasonable cost? To a large extent this depends on the organisation of the work itself and not the overall supervisory structure. There is also an argument of economies of scale and scope, implying that some costs could be reduced if you combine the supervisory authorities.

A more important aspect of efficacy is whether the authority lives up to its stated objectives, for instance to ensure a safe and sound financial system. This has not so much to do with the organisation of the authority but more with its legal powers and independence. An even broader issue concerns the appropriate range of objectives for a supervisory authority, for instance if consumer protection, competition policy or promotion of the financial sector should be included together with the basic objective of financial stability and smooth operations in the financial sector and its institutions. My experience from many countries is that it is usually more efficient to separate these objectives between different authorities. For instance, handling consumer issues is resource intensive and may negatively affect confidence in the supervisory authority. Having too many objectives might also lead to conflicting goals for the authority. As an example, it is more difficult to propose needed regulatory measures at the same time as promoting the growth of the financial market.

To sum up: A country should be able to manage independence, accountability, transparency and efficiency satisfactorily in any kind of supervisory structure. For independence there might be some advantage with central bank supervision. But for cooperation leading to better efficiency there might be a slight advantage with a unified authority.

Arguments for and against various supervisory structures
After having discussed the conditions for good supervision let me now review some of the arguments for and against the various actual supervisory structures. I will first address the issues of having supervision inside or outside the ministries and the central banks. I will then turn to the arguments for and against unified authorities.

Supervision inside ministries may provide some benefits in the form of insight and staff expertise, maybe also access to resources. But there is a strong disadvantage in the lack of independence plus the risk to the public’s confidence in the government if failures should occur in the supervised sector. There is also a serious governance issue in those countries where the government is the owner of financial institutions and also supervises them.

Arguments for retaining supervision in the NCB
What about supervision inside the NCB? First, banks have a special importance in the financial system. Their activities are often complex. Therefore, they need high quality, resource-intensive supervision. The NCB normally has independence and revenues. NCBs can often recruit and retain good staff by offering attractive salaries and other advantages.

Second, Federal Reserve Board Chairman Bernanke(2) and others argue that information gathered in the conduct of supervision provides a necessary input for the monetary policy. Having updated knowledge about the financial system is also important if the central bank is requested to provide liquidity support in a crisis situation. Although such information can be summoned from a supervisory authority outside the NCB, the NCB will be in a better position if it is already familiar with the situation through its in-house supervision. A similar argument is that the central bank is often running or overseeing the large value payments system. Thus the central bank will be able to spot potential liquidity problems at an early stage and can take timely remedial action if it has the supervisory powers to do so.

Arguments against retaining bank supervision in the NCB

But there are also counterarguments to conducting supervision in the NCB. Some would reverse Bernanke’s argument and say that just because the NCB is responsible for monetary policy, payments system oversight and liquidity support in a crisis, it should not be responsible for daily supervision. An NCB with such responsibilities might run into a conflict between the objectives of monetary policy and of supervision. Simply put, the NCB might be tempted to keep interest rates excessively low in order to prevent that some weak banks run into acute problems. In fact, some research results(3) based on OECD countries have indicated that the inflation rate is higher and more volatile in countries where the central bank has the sole responsibility for banking supervision. A failure in a supervised entity might affect the reputation of the NCB to the extent that its capability to perform effective monetary policy is hampered. In a crisis, there is a management problem in that the solution of a crisis in the supervised entities will preoccupy the minds of the NCB management so that it will not have the time to focus on other important issues. The Finnish approach is an attempt to solve this dilemma by having a separate board for the supervisory decisions but still being a part of the central bank for its funding and sharing of administrative resources.

An overarching argument is that the Parliament should not delegate too much power to a non publicly elected body such as the central bank. But I do not see this as a major problem. The Parliament should set the parameters and limits for the powers of the NCB and the NCB will have to account for its conduct on a regular basis. If supervision is located in another authority, there will be similar issues of delegation and accountability.

I noted earlier that the current trend is to move supervision out of the NCBs. However, some countries have gone the opposite way by moving securities and insurance supervision under the umbrella of the NCB. They have benefited from the independence and resources of the NCB. On the other hand, these sectors require partly different staff skills and methods and they include issues which are normally outside the mandate of a central bank such as investigations of market conduct and law enforcement actions against individual firms or persons. Failures in securities and insurance companies may affect overall confidence in the NCB, although these institutions are in most cases of limited interest to the NCB from the point of view of financial system stability. Such failures might even lead to what Professor Goodhart calls “a creep of the Central Bank safety net”(4) implying that exceptional liquidity assistance might be used for non-systematically important institutions. My conclusion is therefore that you should only bring securities and insurance supervision into the NCB if this is needed because they can not fulfil their tasks outside the NCB for some reason such as lack of independence or human resources.

After having discussed whether supervision should be inside or outside the Ministries and the NCB, I now turn to the issue of unified supervision. These are some often-heard arguments for the consolidation of supervisory authorities:

Consolidating supervision in one authority
Information sharing and coordination will be facilitated if the different sector supervisors are located in the same authority. In my experience this is mostly true, both for practical and legal reasons. But this requires that the unified supervisor is really integrated and not just a combination of separate branches. I have seen bad examples of unified supervisors where the staff are not allowed to discuss mutual issues since there are still legal barriers between the departments within the authority! I have also seen bad examples of turf battles within a unified authority, which has certainly not facilitated information sharing.

A unified authority will imply some cost savings because of economies of scale. However, this should not be exaggerated. Some administrative overhead costs may be reduced but other costs will remain. Most of the costs emanate from the supervisory operations and there is little scope for savings in this area. More substantial cost savings and more efficiency in supervision may be achieved through better streamlined cooperation between different sector supervisors but this can be achieved both when they are located separately and when they are within the same authority.

The authority conducting banking supervision is in most countries more advanced and has more resources than the other supervisory authorities. By combining bank supervisors with other supervisors it is hoped that the former can share resources and knowledge with the weaker sectors. This argument if often heard, but in practice it does not always work as planned. The risk is rather that banking supervision is weakened when resources are transferred and when skilled banking supervisors leave the authority.

Another argument is that a unified supervisory authority could become stronger and thus more independent than the individual authorities. I would argue that the issue of independence is not related to the unification of supervision. Independence could equally be given to a small standalone supervisory authority as to a unified large one.

Is there an urgent need to change your supervisory structure?

Let me turn to another aspect. Is there any urgency to change to unified supervision or could you wait? The main argument for changing is when the non-bank financial sector is expanding rapidly and is taking a significant market share. Consolidated supervision is then needed to ensure that all risks to the financial groups and markets are taken into account. We see some development in this direction in almost all countries. The financial sector is expanding and introducing new instruments and activities. The boundaries between banking, securities operations, and insurance business are becoming blurred. In many countries this is a slow process, but in others it moves faster.

There are arguments against changing the supervisory structures in the near term. The first one is the flip-side of the argument above. Also the banks are getting larger and more complex. In many countries they still dominate the financial sector and the payments system. Hence, society has a clear interest in strong banking supervision to avoid costly bank failures. We should take into account that banking supervision in many countries functions quite well, while supervision of the other financial sectors is much weaker. Major structural changes involving bank supervision will therefore hamper the conduct of such supervision for a long time due to the focus on the transition rather than on the operations. There is also an obvious risk that qualified staff will leave bank supervision due to the less attractive work prospects outside the NCB. For instance, government salaries and other benefits are usually not as good and it is less prestigious to work outside the NCB. The result would be weaker bank supervision without compensating improvements in other sectors.

There are other ways of meeting the challenge. Many countries have instituted arrangements for closer cooperation between the supervisors. Memorandas of Understanding have been drafted so that information may be shared without legal or other impediments. Joint forums have been established where representatives from the different supervisory authorities meet regularly. Cross directorships imply that high level supervisors sit on each others’ Boards and can inform themselves of developments.

In fact, legislation might be necessary in order to ensure that the NCB can always obtain the information it needs to fulfil its responsibilities on financial stability, extraordinary liquidity assistance in a crisis, and monetary policy. In Sweden, the central bank concluded an MoU with our bank supervisors but we also have the legal power to ask the banks directly for information.

It is difficult to sum up all the arguments for and against. If you are contemplating change in your country, I would say that a broad and objective assessment of the present strengths and weaknesses will give good guidance as to which way to go. The optimal solution for the near term may not be the same as for the long term.

The aspects of cross-border supervision
Financial institutions increasingly operate across national borders. They establish branches and subsidiaries in other countries and they have business linkages with institutions in other countries. Is there any structure of the domestic regulatory authorities that is particularly suited to the efficient supervision of international financial operations? I do not think so. What matters is that there are arrangements between home and host supervisors which ensure information-sharing and close cooperation in particular in crisis situations. Such arrangements can be agreed on whether bank supervision is in a central bank or combined with securities and insurance supervision in a separate authority. That said, Roger Ferguson, a former Member of the Fed Board and Chairman of the FSF, argues that the well established global network among central banks, such as through the BIS, gives central banks a comparative advantage in informal information sharing and cooperation.(5)

Which role remains for the central bank?
Ensuring jobs for the central bank staff is not a valid reason for retaining bank supervision in the NCB. Hence, the question is: Which would be the primary tasks of the NCB if supervision is moved out?

Of course, the task of conducting monetary policy would remain with the NCB. The NCB will also run the large-value payments system, or at least exercise oversight over it. In addition, the NCB should always have the responsibility for the oversight of overall stability in the financial system. Financial stability is a necessary condition for efficient monetary operations and is in itself an important component in a country’s sustained macro economic growth. Since the NCB will still be the lender-of-last-resort the knowledge gained from its financial stability analysis will be crucial in a crisis situation. But the NCB cannot perform an efficient financial stability analysis, nor act as lender-of-last-resort without close cooperation with the supervisory authorities. The other authorities will also benefit from the stability work by the NCB.

Conclusions
Let me try to sum up. My first conclusion is to use the American expression “If it ain’t broke don’t fix it”. If bank supervision works well in your country, let it be as is and strengthen the other sector supervisors instead. Operational independence should be implemented and at the same time rules for accountability and transparency. This is more important than the structure of the supervisory authorities.

If you see good reasons in your country for changing the supervisory structure, you should follow a transparent process and be clear of the consequences. For obvious reasons, you must avoid making organisational changes when the financial sector is weak. Nor should you change during major developments in the financial sector, such as the introduction of the bank capital framework Basel II and the new accounting standard IAS 39. The structural change must be accompanied by measures to strengthen the capacity of the new authority. Otherwise the change is merely cosmetic.

In many countries there is a political pressure to change the supervisory structure to show expediency, in particular when there has been a financial crisis and irrespective of the supervisory agencies’ roles and behaviour before and during the crisis. As a result banking supervision may in such cases be moved out of the central bank or, at other times, into it. Also the unification of the supervisory authorities is sometimes seen as a political issue. In my experience, there is not much to gain from a political point of view. On the contrary, the politicians might be blamed for future failures of supervision.

Finally, the time has come to answer the question in the title of this presentation “Is there an optimal way to structure supervision”. My answer is “No”. The different supervisory structures reflect the specific situation of their countries, which may change over time, and there is no globally agreed best practice. As former Chinese Chairman Deng Xaioping once said: “The colour of the cat does not matter as long as it catches the mice”.

Notes:
1)Martin Čihák and Richard Podpiera: IMF Working Paper WP/06/57 from 2006.

2)Held at the Allied Social Science Association Annual Meeting, 9 January 2007. Link to the speech: http://www.federalreserve.gov/BoardDocs/speeches/2007/20070105/default.htm

3)Haubrich (1996) and Di Noia and Di Giorgi (2000). These and other issues related to the role, responsibilities and governance of central banks are discussed in the paper Governing the Governors: A clinical study of central banks, drafted by Lars Frisell, Kasper Roszbach and Giancarlo Spagnolo. A yet unpublished draft manuscript was issued in August 2006.

4)Charles Goodhart: The Organisational Structure of Supervision; FSI Occasional Papers No 1, Nov. 2000.

5)Ferguson, R (2000) Alternative Approaches to Financial Supervision and Regulation, Journal of Financial Services Research, 17(1), 297-303