Submission to the Financial System Inquiry 5. Organisation of Prudential Supervision

Introduction

128. Calls have been made for a re-arrangement of supervisory responsibilities as a response to the perceived deficiencies of present arrangements. These deficiencies were assessed in the previous Chapter. International experience does not provide a conclusive guide to the ‘right’ way of organising prudential supervision. A variety of models exist, operating in the context of different countries' historical, political and cultural circumstances. This Chapter looks at some options for changes in Australia's arrangements and concludes that, in the RBA's view, there is no case for radical overhaul of the present system. Overseas arrangements are summarised in Appendix B.

A ‘Mega-regulator’?

129. Some commentators have suggested that all supervision be assigned to a single agency or ‘mega-regulator’. The justification usually appeals to the phenomena of blurring and inconsistencies in regulation, which were discussed in the previous Chapter. The RBA believes the supposed advantages of a ‘mega-regulator’ are illusory and that the approach poses significant dangers to the long-run stability of the system. The RBA would hold to this view, even if the proposal were to locate the ‘mega-regulator’ in the central bank.^[5]
130. The ‘mega-regulator’ is a difficult concept to pin down. Presumably, it would not be responsible for competition policy or consumer protection, but would be responsible for prudential supervision somehow defined. This model would still leave financial institutions answering to at least three regulatory bodies, and so would not achieve the simplicity inherent in the title ‘mega-regulator’. But this is not the main shortcoming of the proposal.
131. The main shortcoming is the belief that prudential supervision should be applied to such different activities as banking and insurance on the one hand, and funds management on the other. The analysis in Chapter 2 makes it clear that prudential supervision is an activity to be undertaken on an institutional basis, where the solvency of the institution is, at least potentially, at risk, ie where it will fail if it cannot meet its pre-determined commitments. It therefore makes sense to apply prudential supervision to banks and life offices. The type of regulation appropriate to funds managers and mutual funds is not prudential supervision because the solvency of the institution is not at risk – it is product disclosure and advice regulation. The two types of regulation are very different and the RBA does not believe there are any synergies between them. This is one reason why, in most countries, they are carried out in different agencies.
132. There are, in fact, strong reasons why they should not be carried out by the one agency. Prudential supervision necessarily carries with it some official ‘comfort’ for savers about the solvency of individual institutions and the maintenance of contractual claims. In Australia, the degree of comfort falls short of a guarantee. Support for an ailing institution may involve the use of public funds, but can also be achieved through the organisation of a takeover by a stronger institution or an officially-managed workout. There should be no expectation of such support in the case of managed funds. Yet investors might assume that some obligation for adequate investment performance is implicit if the same agency has responsibility for both types of institution, the more so when a capital-backed institution has an ownership linkage with the funds manager. Any arrangements which encourage the perception that managed funds are no more risky than deposits would distort the risk spectrum, increase moral hazard and the implicit ‘safety net’, and reduce the efficiency of the mechanism by which savings are allocated to investments of varying characteristics, including risk.
133. Proponents of a ‘mega-regulator’ argue that regulatory conflicts or overlaps would be eliminated, or at least significantly reduced, if individual regulators were absorbed within the one organisation and reported to the one chief executive. There may be some merit in this argument, but it is essentially one of managerial efficiency, rather than one based on economic principles. Large public sector organisations can suffer diseconomies of scale, and some disputes are better sorted out in public discourse between regulators than in private negotiation between departments of the same agency. In addition, the degree of overlap in the current prudential structure tends to be exaggerated. Even under the present arrangements, the statutory responsibilities of the existing prudential supervisors are clearly defined, and no single institution (as opposed to conglomerates) is subject to prudential supervision by more than one of them.
134. Another claim is that a single supervisor would mean more efficient supervision of financial conglomerates. The RBA accepts that conglomerates will generally be answerable to more than one agency because of the different jurisdictions to which constituent companies belong. However, any administrative economies from a single overseer will be limited as long as different supervisory standards apply to different parts of the group. The businesses of member companies and the types of risks involved remain sufficiently distinguishable to require different supervisory techniques and skills. For example, the capital adequacy rules applying to a bank will always be different to those applying to an insurance company and neither will bear any resemblance to the product disclosure rules for a funds manager.
135. It has also been suggested that a single supervisor would be better able to handle the related tasks of limiting the risk of contagion from one entity in a conglomerate to another, and of appraising the overall health of such a group. Again, as long as different prudential standards apply across the group, this claim is questionable, particularly if communication and co-ordination among the relevant agencies is effective.
136. An international consensus is emerging that, for most financial conglomerates, a lead regulator should be nominated to organise group-wide financial assessments, to exercise authority over special-purpose holding companies and to co-ordinate remedial action in a crisis. As noted in Chapter 3, this approach has been proposed by the CFS for conglomerates headed by special-purpose holding companies and is likely to be extended to others. The need for formal lead regulator arrangements in Australia has been lessened by the fact that in most cases hitherto one (supervised) institution has dominated each conglomerate.
137. Among OECD countries, only Sweden, Denmark and Norway have a stand-alone ‘mega-regulator’ which supervises banks, insurance companies and investment products. Japan is close to this model in that it has a ‘mega-regulator’ with these responsibilities, but it is part of the Ministry of Finance.
138. The BIS Annual Report of 1993 contained a survey of financial distress among its member countries over the previous decade (see also Llewellyn (1992)). The countries singled out for poor performance in terms of financial instability and large losses of taxpayers' funds in bank rescues were Sweden, Norway, Finland, Japan and the US. Three of the countries that employ the ‘mega-regulator’ model^[6] are included in this small group. While it is difficult to spell out the precise chain of causation, this correlation should not be ignored. Proponents of the ‘mega-regulator’ model need to explain why they are recommending an approach which has such a poor track record.^[7]

Which institution should supervise banks?

139. The RBA believes that central banks are best placed to oversee the stability of the financial system. They are participants in that system – conducting a wide range of financial transactions in money, bond and foreign exchange markets, as well as often acting as banker to governments. And, they are the ultimate source of liquidity to the financial system. They may supply liquidity to the system generally (as in 1987 following the stock market crash), or they may be the ‘lender of last resort’ to individual financial institutions which are in difficulty. Last resort lending refers to a direct loan usually from a central bank to a commercial bank unable to raise liquidity from other, more usual, sources; this would typically be because the bank was experiencing a run on deposits due to a loss of confidence. Last resort lending would usually involve loans to a bank which was still solvent. It should be distinguished from financial support for an insolvent institution which, if it were provided, would usually involve the Government. Last resort lending has long been seen as a function of central banks. Indeed, the origin of many central banks was as ultimate backstop to the financial system, with the name ‘reserve bank’ indicating a role as custodian of the reserves of the commercial banking system.
140. While the RBA finds the case to remain the bank supervisor a convincing one, it is true that practice varies around the world. In some countries the central bank is the sole bank supervisor, in some countries it shares responsibility with another body and in some it has no supervisory responsibilities. In recent years, a number of countries have revisited the issue of who should supervise banks. In Norway, a parliamentary working party recommended that the bank supervisory authority be incorporated into the central bank, but was overruled by the full parliament who opted for the status quo. In South Africa, the Jacobs Commission made the opposite recommendation – namely, to take bank supervision away from the central bank – but again was overruled and no change has occurred. In the two cases where the recommendations were carried out – Finland and Hong Kong – the result was to move a formerly independent bank supervisor into the central bank (see Appendix B).
141. A number of arguments have been advanced against the central bank being the bank supervisor. The RBA remains unconvinced by these arguments and feels that, on balance, the arguments for having a dual role are much stronger. A summary of the relevant arguments is given below; a more detailed account is contained in Appendix C.
142. The most common argument for separation is that there is a conflict between responsibilities for bank supervision and monetary policy. That is, a central bank that is also the bank supervisor will place too much emphasis on financial stability, rather than concentrating single-mindedly on anti-inflationary monetary policy. This argument does not say that the combined institution will be a poor bank supervisor – it says that it is monetary policy that will suffer. But the argument is a curious one in another way, in that it implies that ignorance of the health of the banking sector is an advantage in conducting monetary policy. The RBA believes the opposite to be the case. A first-hand knowledge of whether the banking sector is robust or fragile is essential in order to understand whether a change in monetary policy will have a large or small effect on the economy. In essence, there are synergies between the two responsibilities, rather than conflicts.
143. A second argument for separation is that in the event of a ‘bail out’ of the banking sector, the cost will be borne by the taxpayer and so the decision should be made by the government (or a body reporting directly to it). This is a rather academic argument in the Australian context as neither the RBA nor the Federal Government has lost taxpayers' money through a bank rescue.^[8] Nevertheless, it is a possibility. However, if the argument is to have validity it would have to be established that the central bank was likely to extend lender-of-last-resort credit unwisely when a stand-alone supervisor would decide against exposing the taxpayers to loss.
144. The question of which body – the central bank or the separate institution – would make the more prudent decision cannot be answered with any confidence. On the other hand, the question of which body could take the most timely action in the event of a crisis would have to be answered in favour of the central bank. It is transacting in the money, bond and foreign exchange markets virtually every day, and sits astride all the flows in the payments system. Whether it is the bank supervisor or not, it would have to be involved in any action designed to alleviate a financial disturbance or a crisis. It could do this either by adding liquidity to the market as a whole, in the course of its open market operations, or by extending lender-of-last-resort loans to a particular bank (or banks). If it felt that the problem in question was not a matter of liquidity but of solvency, it could arrange a merger or orderly closure. In principle, of course, it could do these things on the advice of an independent supervisor, but in practice this would be a clumsier arrangement. Time is of the essence in a crisis, and getting agreement between different institutions is often difficult in such circumstances. While it is usual to think of a bank failure as the event which precipitates a potential crisis, it is equally likely that the initiating event could be a disturbance in financial markets such as happened in the Penn Central case or following the share market crash of 1987 (see Appendix C). Thus, the skills required for crisis management may not only be those of the bank supervisor, but could also involve those of the financial market specialist within the central bank.
145. A third argument for separation is that a central bank could suffer reputational damage as a result of its bank supervision activities, which would reduce its credibility in conducting monetary policy. What gives this some force is that, even in an optimal system of bank supervision, an individual bank can (and must be allowed to) fail, but the public, the press and politicians invariably see this as evidence of incompetence on the part of the supervisor. This argument is widely used in the UK, where the reputation of the Bank of England has suffered as a result of the failure of Johnson Matthey Bankers, BCCI and, recently, Barings. Some commentators see this as affecting its ability to conduct monetary policy, but it has been hampered on this score at least as much by an Act which gives it virtually no independence.
146. This argument, like the first, does not imply that the central bank will be a poor bank supervisor – it says that it will be monetary policy that suffers. In the RBA's view, any merit in this argument is outweighed by other factors which point to advantages from combining bank supervision and monetary policy. For a start, there is always a risk that a stand-alone bank supervisor would see its role rather narrowly. This is because it would be penalised severely in the case of a bank failure, but receive little or no reward for other desirable outcomes such as the efficiency, innovative capacity and growth of the banking sector. A central bank, by necessity, has a wider set of objectives and responsibilities and is less likely to narrow its focus in this way. In similar fashion, a stand-alone supervisor is more likely to develop a rules-based culture, with lawyers and accountants predominant. In a central bank, the bank supervisors have to argue their case with colleagues who are predominantly economists and financial market specialists. This has the effect of softening the dependence on rules and introduces a higher degree of ‘market friendliness’.

Which institution should supervise insurance companies?

147. The insurance industry provides a number of products such as life insurance, annuities and defined-benefit superannuation which involve binding nominal contracts. Failure to meet these contracts would cause the failure of the institution. They therefore fall clearly into the category that should be prudentially supervised. In which case, they are much closer to banks than to the managers of investment products (although they are also in this business). Is there a case for combining their supervisor with the bank supervisor?
148. In the RBA's view, there is no good reason to do so. It would not raise moral hazard issues but there is no positive case either. Synergies are hard to find because it would not be appropriate to apply common supervisory rules and techniques to those entities, notwithstanding the fact that they are all capital-backed. Particular expertise is required for the different types of financial business that they undertake. As noted in Chapter 2, the risks in the insurance and superannuation businesses require a different type of analysis than those in a deposit-taking operation. The application of actuarial techniques to the liabilities side of the balance sheet is the central element in the prudential supervision of a life office, while assessing the adequacy of a bank's capital, on the other hand, is a very different matter. Overseas experience is that where the supervision of such fundamentally different financial activities is the responsibility of a single agency, the work generally tends to be allocated to separate departments or divisions and different capital adequacy rules are applied. The major examples are the Scandinavian countries, where a single supervisor covers the banking, insurance and securities industries, and Canada, where banking and insurance fall under the one agency.
149. Another argument for combining insurance and banking supervision is that it would, in fact, help to produce a consistent set of capital requirements. The capital requirements on banks and life offices are indeed different but so is the nature of the risks they face. No-one has yet established that the requirements on one set of institutions are more onerous than the other, given these different risks. If it were so, more consistent capital requirements could be applied, but that could be achieved just as easily by discussions between separate regulators as between departments within one regulator. Banks and insurance companies wishing to operate in more than one country would, of course, also be constrained by international rules.

Other deposit-taking institutions

150. At present, building societies and credit unions are supervised by State agencies under the umbrella of AFIC, in arrangements which involve some duplication of resources and inflexibility. While those institutions are subject to contagion risk, their small size makes it doubtful that even a widespread run on them would have systemic consequences.
151. Nonetheless, there may be a case for the RBA to take on their supervision on efficiency grounds. The community sees building societies and credit unions as safe havens for a significant part of their savings, in much the same way as they do banks. Further, their business is functionally similar to that of banks, with relatively liquid liabilities and a significant part of their assets comprising long-term, non-marketable loans. Consequently, the supervisory arrangements for building societies and credit unions are based closely on the RBA's supervisory regime for banks. There is some potential for greater efficiency and cost savings if these bank-like entities were also supervised by the RBA. However, it would not be practical to attempt to force building societies and credit unions into a bank ‘mould’ as far as ownership rules, etc, are concerned.
152. If the RBA were to assume supervisory responsibility for both these groups, there would inevitably be some compression of the spectrum of perceived risk among financial intermediaries but, as noted, that would not be out of line with the community's assessment of these intermediaries. There could also be an extension of moral hazard risk. To mitigate this, it would need to be made clear that RBA supervision should not be regarded as a guarantee against institutional failure, even for a bank.
153. There is an anomaly in the present system in the treatment of merchant banks, which is a hangover from the days when banks were heavily regulated and foreign bank ownership was not permitted. Most of the firms in this sector are subsidiaries of foreign banks and their activities are difficult to distinguish from the wholesale banking activities of authorised banks. Indeed, many of the foreign-owned institutions which are now authorised banks converted from merchant bank status with little change to their operations.
154. The anomaly in these arrangements is that, although merchant banks are doing the same business as authorised banks, they are not supervised in the same way and are not subject to the NCD requirement. In other countries, it is normal for foreign banks to have to take out a banking licence if they wish to do banking business in another country's market. There is a case for imposing that requirement in Australia as well.
155. The continued existence of a significant merchant bank sector, with substantial foreign bank ownership, does not sit comfortably with international efforts to improve the supervision of the cross-border operations of international banks. In terms of the standards developed by the Basle Committee on Banking Supervision, there is an obligation on host authorities to exercise effective supervision of foreign banks operating in their territories. Host authorities are expected to be in a position to inform the home supervisor (who is charged with responsibility for exercising effective consolidated supervision of its banks) of any perceived shortcomings in the local activities of those banks. The RBA is not in this position with regard to merchant banks owned by foreign banks. It therefore favours a change in policy which would require such banks to become authorised under the Banking Act.
156. It is much less clear that finance companies should also be made subject to RBA supervision. This sector comprises about 100 entities with total assets of around $46 billion. Some two-thirds of their borrowings are from the wholesale markets. Some 20 finance companies borrow in the retail market, through the issue of debentures under the prospectus disclosure provisions of the Corporations Law and subject to the surveillance of the ASC. Many finance companies are equipment finance specialists, including some which support the sales activities of associated industrial concerns. Finance companies do not engage in maturity transformation to any significant degree as the bulk of their fund raising is in the form of term borrowings. Only about 5 per cent of their liabilities are at call, so the threat of a ‘run’ is not an issue. Nor is contagion a concern and finance companies would not seem to be a threat to financial system stability. Several have failed in the past 20 years without causing wider problems and the community accepts that they are further out on the risk/return frontier than banks.

Regulation of investment products

157. Managed investments, whether sold directly to the public or to trustees of superannuation funds, should be subject to product regulation with the emphasis on disclosure requirements. There is no case to apply prudential supervision to institutions which manage investment products.
158. The authority responsible for the regulation of investment products should, of course, be separate from the authority responsible for prudential supervision of banks. Beyond that, the RBA has no strong views on institutional arrangements. At present, the ISC and the ASC regulate the products of the funds management industry, the superannuation industry as well as investment advisers. There have been reports that some firms have been able to shift products between the two regulatory jurisdictions in order to gain more favourable treatment. This is not necessarily a deficiency of the system; it only becomes so if it can be established that similar risks are receiving different regulatory requirements.
159. For the special reasons outlined in Chapter 3, a form of prudential supervision is exercised by the ISC over accumulation superannuation funds. This puts such funds at a different point on the risk spectrum compared with the generality of investment funds. Because of the importance of preserving a risk spectrum for investors, the RBA would be concerned at any prospect of superannuation-type oversight being extended across the whole managed funds industry. At the same time, it could be socially undesirable to lessen the present oversight of superannuation. It might, therefore, be appropriate to have superannuation funds overseen separately by the ISC, an arrangement which entails certain synergies with the supervision of the life insurance industry.

Summary

160. The RBA does not see a strong case for radical overhaul of present responsibilities for prudential supervision. In the interests of efficiency, responsibility for building societies and credit unions could be transferred to the RBA, although there are no pressing reasons in terms of financial stability for doing so. There is a stronger stability argument that the anomalous situation of international banks operating here as non-banks should be resolved by requiring them to become authorised banks.
161. The RBA supports the continuation of a specialist insurance supervisor and a separate product disclosure regulator of the managed funds industry. Oversight of superannuation could be located in either of these agencies, but there seems to be no strong case to relocate it from the ISC where there are some synergies with supervision of insurance companies.

Footnotes

In fact, the moral hazard problem could become more severe if the ‘mega-regulator’ was in the central bank. [5]

In the other two, Finland and the US, the banking system (or the part that incurred the losses – in the US case, the savings and loans institutions) was not supervised by the central bank. [6]

The bail-out of the banking systems in the Nordic countries was extremely expensive for taxpayers. In Norway and Sweden the cost was equivalent to nearly 5 per cent of GDP; in Finland it was 8 per cent of GDP. [7]

The RBA has never had the need to lend to a bank in liquidity difficulties, and therefore has never lost money through a last resort loan. In the early 1990s a number of banks in Australia made losses, but they were recapitalised by their owners. In most cases the owners were foreign banks, but in two cases the owners were State Governments which had guaranteed the liabilities of the banks they owned. The Commonwealth Government provided compensation to the Victorian and Tasmanian Governments for the tax equivalent revenue foregone following the sale to private banks of State-owned banks. This reflected government policy to encourage privatisation. The South Australian Government was also provided with assistance to help it reduce the debt burden associated with supporting the State Bank of South Australia (SBSA). This assistance was entirely discretionary and conditional on privatisation of SBSA; it occurred some two years after problems at SBSA were first revealed. [8]