Nyberg: Issues in Supervision of Banks - European Experience
Background
First of all, I want to thank you for giving me this opportunity to discuss with you some questions concerning the supervision of banks. The issue, I understand, is quite timely, since Sri Lanka has embarked on an ambitious project of developing the financial sector, opening the banking system to international influence and modernising it accordingly.
A sound and well functioning financial sector is a key to economic development. This statement may sound uncontroversial in your ears, but it is not always accepted by the general public and by the politicians. The contributions of the financial sector are sometimes hard to explain in simple terms and people tend to look at banks with suspicion. To create credibility and respect for the financial sector in general and most importantly for banks, a proper and well-developed supervision is essential. Bank of Sri Lanka has realised this and initiated a comprehensive upgrading of bank supervision. In my view this is exactly the right point to start developing the financial sector.
Supervision, however, is developing quickly. Supervising a traditional highly regulated banking system is far different from supervising a deregulated system open to foreign competition. Furthermore, going from a state of strict regulations to one of deregulation implies risks for undermining banking stability. Many countries, not least my own, have had bitter and expensive experiences from this.
Of course, all countries have specific national features in their banking systems. The main risks are common all over the world, however, and the same challenges to supervision seem to appear everywhere. I will focus on some important issues seen from my perspective, which includes the experiences we gained during the Swedish banking crisis in the early nineties.
Deregulation
Why deregulate if the consequences may be so dangerous? The obvious answer is that regulations tend to create inefficiency and prevent the domestic banking system from developing in response to international competition. The long run cost to society of extensive regulations is a poorly functioning financial sector.
While regulations are usually put in place to decrease systemic risk they may indeed in the long run do quite the opposite. Strong regulations mean that banks have low incentives to adopt good credit policies and risk management techniques. Supervision tends to be excessively formal, focusing on formalities rather than on actual risks. Nor are the regulations and techniques of supervision updated often enough to foster the development of new activities and management of risks. Regulations shelter the domestic banking industry from outside influence, which slowly tends to decrease efficiency and increase systemic risk.
Furthermore, in a regulated environment the likelihood is large that some credits may be subsidised, perhaps even channelled to politically desirable rather than economically sound projects, leading to misallocation of resources.
Albeit many people today acknowledge that the traditional kind of regulations do not produce the desired results, we tend to keep them too long. Let me give you an example. In Sweden, we had foreign exchange controls in place until 1989. Even though it was obvious that the controls were easy to circumvent and quite inefficient, the authorities were worried by what might happen if they were abolished. Perhaps there would be a huge capital drain? Hundreds of pages were written in attempts to map the consequences, but no comfort was provided this way. Eventually the Central Bank decided to go ahead in spite of the remaining uncertainty. To some surprise and some disappointment, nothing particular happened in the market. The move had been anticipated and the controls inefficient enough to allow the capital flows to adjust prior to the move.
Credit controls and interest rate regulation had been abolished in Sweden four years earlier, in 1985, again probably far too late. Some people argue that the banking crisis we had in the early nineties might have been much less severe if the banking system had been forced into a competitive environment by deregulation a decade earlier.
Keeping regulations too long is a danger, getting rid of them too quickly is another. This may sound peculiar, but let me again give you an illustration. When, in 1985, credit controls were abolished in Sweden, they had been in operation for more than forty years. Since the authorities had decided on how much each bank could increase its credit book every single month, there was really no need to appraise credit applications in a particularly serious manner. Mortgage credits took most of the available credit volume and the remainder was allocated to safe projects. Credit losses in the banking system was only a fraction of a percentage point of the credit stock each year. As a consequence, the ability to appraise credit risk had slowly dwindled out of the banks. So, unfortunately, was the case for the supervisors.
When credit controls were finally removed the volume of credit virtually exploded. All banks wanted to increase their market shares and there was plenty of pent-up demand. Many loans were given to customers who, as it appeared later, were far from credit worthy and should not have passed a proper credit appraisal. The regulations had been abolished at the wrong time and without realising what an extended period of regulation had done to the ability of banks and supervisors to handle credit risk.
What can be learnt from this? Quite obviously, deregulating does not mean the abolishing of all rules according to which banks should behave. There are good and well documented reasons for regulating banks. Deregulation means the adoption of a new legal structure that is consistent with a competitive market economy. The dilemma of regulation is to balance between having too little and too much. Too much will stifle the banks, too little will lead to unsound risk taking. The traditional form of regulation was to set very strict rules or to forbid banks from conducting certain activities. The modern form is to be more flexible. Regulations set the boundaries while allowing the banks to operate within these boundaries if they can prove that they have the skills and resources to do it in a sound manner. Modern regulation puts much more responsibility on the banks themselves and their owners and management.
In Sweden, the former Banking Act actually prohibited the banks to engage in any activity that was not explicitly permitted in the law. The new Banking Act starts at the other end, permitting the banks to do whatever they find suitable except for certain activities that are explicitly forbidden.
Today’s financial institutions, instruments and markets are, of course, much more sophisticated than earlier. There is also much broader diversity in the sector. For example, the needs of customers are satisfied in many different ways and by many different actors, certainly not only by banks.
As a consequence, modern regulation needs to be more flexible than in earlier days, when the market was more homogeneous. The regulation should encourage competition and innovation, but at the same time discourage unsound risk taking and generally unbecoming or even illegal behaviour. Only when there are strong arguments should the regulator outright restrict or prohibit certain activities. Instead, regulations should provide incentives for banks and others to behave in a sound manner. For instance, capital requirements do not prohibit risky exposures, but they make them expensive to the bank.
Before moving somewhat deeper into the questions of supervision, there are some things to say about risk management.
Riskmanagement
Risk management systems are at the core of modern banks. Banks must have consistent systems for identifying, monitoring, controlling and managing risks. These systems must be endorsed and understood not only by credit officers, but also by the board and the management.
In a deregulated banking sector, the division of control and responsibilities in a bank is of particular importance. Owners must take an active interest in the bank by setting broad strategic guidelines and policies for risk management and internal controls. Bank boards must have access to timely and adequate information and act on it when necessary. There must be a division of labour making it clear to all members of management and staff what specific responsibilities they have.
From the Swedish banking crisis I would single out a few types of risk.
Credit risk in general, linked to the issue of collateral, was - as should be expected - the main problem. It was our experience that banks coming from the sheltered world of strict regulation did not have good and consistent systems for credit evaluation and documentation. Today, banks must have credit policies, stating what kind of credits they wish to grant. Usually, small and simple loans are provided at branch office level, but larger and complicated loans are provided at successively higher levels. Banks have also adopted internal risk evaluation systems in which the individual loans are graded according to their risk. Loan documentation is improved to facilitate monitoring of the loan but also to ensure that the bank can collect on the loan in the event of a judicial process.
In Sweden, concentration of credit risk was a major problem during the crises. Some two thirds of the losses were linked to the sharp price declines in real estate.
A loan and its collateral must be evaluated separately, something that we had forgotten during the many years of regulation. If the expected returns from the investment made by the loan are uncertain or not adequate the loan should in most cases not be given, irrespective of the collateral. The reason is that the collateral, in a crisis situation, often proves to be worth far less than expected. This is particularly true when it is tied to the borrower – such as when a company has put up its own building as collateral.
The linkage between credit risk and foreign exchange risk was large during the Swedish banking crises. This was seen also in the Asian crises in 1998. For example: Swedish borrowers wanted loans in yen and German marks because the interest rates in these currencies were low. When the Swedish currency depreciated, the borrowers could not repay the higher loan values because their revenue stream was in local currency. The banks believed themselves to be hedged against the foreign exchange risk since they were funded in the currency of the loans given, but they failed to see the corresponding increased credit risk.
Other risks were less prominent in the Swedish crisis but, in my opinion, will be of increasing importance to banks all over the world. Among those are operational risks. Several recent financial debacles, such as the Baring crisis, were due to operational risk. Even though the Y2K problem was successfully managed, the huge computer systems of banks could cause large losses if technical or other failures occur.
Another risk, which in my mind will become increasingly important, may be called "strategic risk". The financial market in most countries is highly competitive. In order to survive and prosper, banks must take bold decisions on new activities, new establishments, mergers, technological investments and so on. Decisions like these have to be taken, but should they go wrong, banks may suffer large losses and may even go bankrupt.
To handle these and other risks, internal controls and internal audit systems are becoming increasingly important. The huge mass of information in modern banking makes it impossible for supervisors to monitor individual transactions. They must rely on banks to maintain reliable internal control functions, which not only verify the accuracy of transactions and operations but also take a broader view of the bank’s compliance with its own rules and policies as well as with external rules and policies.
Supervision
Traditional supervision was of the type, which you may call "checking the ledger", that is verifying that all transactions are properly recorded and formally correct. To some extent this must remain, but the supervisor’s efforts in this sense can be reduced by applying modern statistical methods, e.g. for sampling, and by making sure that banks have adequate internal control systems.
Instead, supervision should be risk-oriented and focus on those activities, which pose significant risks to the individual bank (differing from bank to bank). Risk-oriented supervision also implies stricter supervision for large banks and for banks taking larger risks than others. Supervisors should focus on evaluating and monitoring risk management systems, internal controls, corporate governance, and information systems. This calls for large supervisory resources and for more sophisticated skills: lawyers, but also economists, financial analysts, statisticians, mathematicians and accountants. If the law permits, which indeed it should, the supervisor could call on outside experts to assist.
In order to be effective, supervision must in most cases be conducted on a consolidated basis, to include all parts of a financial conglomerate and even its non-financial affiliates. The increasing blurring of boundaries between different financial activities and the building up of large financial conglomerates make consolidated supervision crucial. Consolidation includes financial as well as non-financial entities within a country, but also world-wide. The co-operation and exchange of information between supervisors for different financial activities (banking, securities trading and insurance) is important, domestically as well as globally. Informal or formal agreements must be concluded to facilitate such co-operation. Supervisors should not license banks, or their affiliates, if their organisational structures prevent effective supervision. In the same vein, home supervisors should not allow their banks to establish affiliates abroad, if the host authorities are not willing to exchange important information regarding developments affecting this affiliate.
Transparency is key. Banks should be induced, by force and voluntarily, to properly disclose information of their activities to the general public. Information must also be timely, relevant and accurate. In this way, there is daily "supervision" of the bank by market analysts, depositors and others. It is understood that some bank activities are to be considered as strategic secrets which cannot be divulged, but in my view these are fewer than we tend to think.
Hence, a related subject of great importance is the issue of accounting. To create transparency, accounting rules, including valuation rules, must provide a clear picture and reflect the true situation of a bank. Let me give you an example. A lingering problem in many countries is that banks hold large amounts of non-performing loans to companies, even those that are state-owned or formerly state-owned. The loans are usually not valued according to their realistic, low, values. In my view, this is not helpful even if it seems to improve the financial statements of the banks. Nobody is fooled about the actual situation because these bad loans are well known in the market. The loans should preferably be transferred out of the books to some asset management company – priced at a realistic market value, of course.
Under all circumstances, supervisors should require that banks have consistent and true systems for the valuation of loans, performing and non-performing, with or without collateral. These systems should also take into account the setting aside of adequate provisions for weak loans. These provisions should reflect the expected losses to the bank. In calculating these losses, collateral may be taken into account, but as earlier mentioned, it is important that the collateral is conservatively valued because its value tends to decline sharply in a default situation.
For any country aiming for a credible financial system, auditing standards must be high and approaching the best international practices. Furthermore, the supervisors should maintain close and regular contacts with the major accounting and auditing firms in their country. This serves to provide and receive information on individual banks, but also to explain and to get feedback on the development of new rules and regulations and how they are implemented in the market.
Asset Managemnet Companies
Before proceeding, I would like to make a few points concerning the asset management companies (AMC:s) touched upon above, since these seem to be controversial in many countries. They certainly were in Sweden during the peak of the crisis.
Asset management companies are simply companies to which a major part of the non-performing assets of a bank can be transferred, given that these loans for some historic reason have grown to become a considerable part of the bank’s balance sheet. The purpose of the asset management company is to dispose of the assets as quickly and profitably as possible and then close down.
As I see it, there may be two main causes for such a transfer of assets. First, experience shows that bankers are not best equipped to handle the restructuring and work out of collateral taken in by the bank as a consequence of failing loans. Bankers may possibly be good at lending money and handling minor non-performing loans, but they usually have little experience in reorganising industrial activities and managing real estate companies to the extent required in an AMC. These activities require a different competence.
The trouble is that bankers often fail to see this, in particular if they have been heavily involved in handling the failing credits. They talk about long term relations with the customer. They argue that the loans could be kept in the book of the bank and that relevant competence could be hired. I strongly believe that this is not true. If a proper restructuring is to be done, professional restructuring people without historical relations to the customer should do it. Whenever I have been persuaded otherwise I have later been proved wrong.
The second argument for transferring the non-performing assets is related to this. When the management of a bank has a large number of bad credits to take care of, that is exactly what they do. All their capacity is used up taking care of problems and there is no time – and no money – to think about the future. This is disastrous for the whole organisation, which tends to become backward looking and defensive instead of forward looking and offensive. Only when the bad assets are out of the books of the bank, the minds of the management and thus of the employees can become aimed towards the future.
Having decided to transfer the bad assets, the next obstacle will be to decide on the prices. The bank usually considers the non-performing assets to be quite valuable, at least in the long run. Selling at what might be considered present market prices will create too big a hole in the balance sheet to make management comfortable. The management of the new AMC quite naturally has a different view. It has no wish to get assets into its balance sheet at values that later turn out to be excessively high.
According to my experience, it is important to listen to the AMC rather than to the bank if the purpose is to get the restructuring done. If the management and board of the AMC can make reasonable profits by selling its assets, which indeed is the purpose, this is what they will do. However, if they have aquired the assets at too high prices, every sale will show a loss and they may even have to approach the owners to get more capital. The owners will not like this, whether the company is owned by the bank which delivered the assets or some external financial investors. Hence the board will avoid showing losses and consequently little restructuring will be done.
Thus transferring assets at market prices is essential if the restructuring process is to work. The trouble is that by doing this you may be forced to write down the assets in the books of the bank, since they have not been valued correctly. This in turn will uncover the hole in the balance sheet and may require a capital injection. And putting more capital into the bank may not be feasible at the time, for political or for other reasons. At this point, you can always start a discussion of what market prices really are. This happened in Sweden and it has happened elsewhere. But a hole in the balance sheet is always a hole, whether you decide to see it or not. Either new capital must be put into the bank before the transfer of assets or it must be injected into the AMC when the real value of its assets becomes obvious to everyone. Putting the money straight into the bank speeds up the restructuring process.
Consolidated supervisory authorities
I mentioned that supervision should be conducted on a consolidated basis, covering activities in different financial and non-financial sectors. Should also the organisation of the supervisory authority be "consolidated"? Several countries, including the UK, Korea, Australia and others have recently moved in the direction of creating an "FSA" which is responsible for the supervision of banks, securities markets and insurance. In Sweden, we have had an FSA since 1990.
Is such a consolidation necessary and useful in all countries? In my view, the optimal structure of the supervisory authority depends on the structure of the individual country. In many countries, banks conduct the most important financial activities while securities markets are less important. Here at least bank supervision must be independent, have adequate resources and skilled staff. This is crucial. In many countries, bank supervision is a part of the central bank. This has benefited bank supervision because the central bank usually has legal independence, adequate resources and can attract qualified staff. Combining bank supervision with securities and insurance supervision will probably imply that supervision will move outside the central bank, since securities and insurance supervision is not normally a task of central banks. In this sense, bank supervision may become weaker. In my view, in markets where banks are dominating the financial sector and where the integration with securities and insurance markets are less pronounced, keeping the supervision separate and within the central bank makes good sense.
Having said this, I acknowledge that the present tendencies are for banks to merge with other financial institutions, and for financial instruments to look increasingly alike whether they are created in banks or in other institutions. These tendencies would indicate a corresponding need for supervision of different financial sectors to co-operate closer and closer and, perhaps, finally to merge. The trend is clearly seen in Europe, reinforced by the common currency, but is gradually developing also in other countries. As mentioned above, this is the way we have organised supervision in Sweden. But I do not see a need to consolidate the supervisory authorities in a specific country until the structure of its financial markets so requires. As an intermediate step, co-operation agreements between the supervisory authorities should be encouraged.
A strong rationale for locating bank supervision in a central bank is that bank stability is of crucial importance for the conduct of monetary policy. In addition, a central bank has the powers to provide liquidity – under normal as well as exceptional circumstances – to banks, for instance as a lender of last resort. To do this, it is obvious that the central bank must have full knowledge about the financial situation of a bank.
However, there are also good arguments for keeping supervision separate from the central bank. There may be a conflict between the role as supervisor and the role of being a lender of last resort. If a bank would fail, being supervised by the central bank, the central bank may feel more inclined to provide liquidity support than is economically defendable. Bankers may know this and the case for moral hazard may therefore be stronger than with a separate supervisor. Rules for public administration of troubled banks outside the central bank may give some help in this respect.
There may also be a conflict of interest relating to monetary policy. If a bank fails while being supervised by the central bank, this may affect the central bank’s reputation and credibility. In my view, these are important arguments, because credibility of the central bank is basic not only to systemic stability but also to the conduct of monetary policy.
What is the role of the central bank in relation to financial stability when there is an independent supervisory authority? There are still remaining the increasingly important questions concerning the macroeconomic "oversight" or macro prudential supervision of the financial system. As a matter of fact, these questions often tend to be neglected when the ordinary supervision remains within the central bank, because they are in many ways different in character from the day to day supervision.
To effectuate its responsibilities of conducting monetary policy and ensuring a stable payment system, a central bank must monitor the risks in the financial sector. In this capacity, the central bank is mainly interested in systemic risks, not in the potential problems of individual banks unless these are large enough to affect the whole system. The macro prudential analysis aims at identifying risks which are common to a number of banks and other financial institutions or markets, including payment systems. Such risks often stem from the development of the macro economy. Volatile inflation developments or exchange rate fluctuations could lead to problems for a number of banks. Asset price "bubbles", excessive loan concentrations or excessive leverage ratios for borrowers might do the same. It is natural for the central bank to closely monitor such developments and take prompt action if it finds that financial stability is endangered. In some cases it may not have the powers to take direct action itself, but it may inform other parties, such as the political authorities or the supervisors, urging them to apply necessary measures.
I would like to emphasise that macro prudential analysis does not mean that the central bank is duplicating the work of the supervisory authority. We do not examine and audit the individual financial institutions, this is clearly the task of the supervisors. Most of our information is gathered from other sources, including from the supervisory authority itself.
An increasing number of central banks, e.g. those in England, Canada and Sweden have established a structured procedure to macro prudential analyses. In Sweden, our conclusions are published in a semi-annual Stability Report, where we analyse the development of risks in the banking system and where we state our overall view on systemic stability. The reports have increased the general knowledge about the stability situation of the Swedish financial sector and they constitute the basis for discussions with the financial institutions and with other authorities. Whether they will help us avoid the next financial crisis is still to be seen.
International standards
Financial systems and markets are increasingly interlinked across the country borders. In order to avoid international financial disturbances all countries that wishes to participate in cross-border transactions must adhere to a set of minimum rules for regulation. Also supervision must reach adequate levels in all countries, including co-operation and exchange of information between supervisors in different countries.
In 1997, the Basel Committee took the initiative of writing the Core Principles for Effective Banking Supervision. Complying with the Principles constitutes a minimum platform for any country that strives to have basically sound regulation and supervision of banks. If I may say so, there is nothing radically new with the Core Principles – all the underlying standards that countries are asked to fulfil are well established and internationally recognised. The main use of the principles is rather as a yardstick, whereby a country can evaluate itself in order to identify weaknesses in its present regulations and supervision and to take remedial measures to improve. There are national and international arguments for a country to adhere to the Core Principles.
The domestic argument is the one of avoiding or reducing the impact of bank problems, not least because they are costly to the macro economy. While the Principles are not intended to prevent single banks from failures, effective regulation and supervision should reduce potential risks and make them easier to identify at an early stage. The international argument for compliance is that other countries will require this before they open their financial markets. The fear of contagion, implying that problems of one country spread into others, remains strong.
Conducting an evaluation of compliance towards the Core Principles is especially helpful when a country is in a stage of modernising its financial sector, such as Sri Lanka. The result of the evaluation will identify weaknesses and can serve as a guide to reform measures. It is also the experience of other countries that the result of the evaluation may be a useful tool for the supervisory authority in convincing the politicians of needed legal and regulatory amendments.
Conclusions
I have touched upon a number of aspects of supervision, all in one way or another related to lessons learned during the last ten years. I shall make no attempt to summarise, just return to my first point that a high quality, risk-oriented and flexible supervision will create credibility for the financial sector and make it prosper. Inadequate and old-fashioned supervision based on rigid regulations will do right the opposite. And supervision is a dynamic concept. The tools and art of supervision must, in all countries, continuously improve along with the development of markets and market technology.
