��

��

��

��

��

 I INTRODUCTION

A successful financial market must be both sound and competitive. A financial system that is competitive, but not sound, cannot be sustained. The lack of soundness will cause the system at some stage to disintegrate, causing massive dislocation not just within the financial system but throughout the entire economy.

A financial market that is sound, but not competitive, will decrease efficiency both within the financial system itself and throughout the economy. The financial system is the sector which ensures that resources flow to their most productive uses within the economy. An uncompetitive financial system can stifle both the efficient allocation of credit and/or equity and reduce available savings, so reducing the economy's growth potential and hence the living standards of the economy's population.

The task of constructing a financial system that is both sound and competitive is the focus of this paper. Our starting point is the set of competition principles being developed in the APEC context (both by APEC and PECC). These are being formulated at a general level to cover all industries. With these principles as a background we turn to the financial sector to analyse what competition policy implies for the regulatory and industry structure of the financial sector.

Here we encounter a major issue confronting financial regulation. In many markets, the process of enhancing competition is largely one of removing state-imposed regulations that had previously inhibited new entry to the market. In financial markets, however, the reputation and soundness of the system is regarded by most policy-makers as a sine qua non of appropriate policy-making. Is this approach anti-competitive? The literature in this regard is examined and it is concluded that while prudential restrictions can be anti-competitive, some limited level of prudential regulation can promote competitive forces.

After outlining an approach that can lead to the construction of both a sound and competitive financial system, we discuss particular policy aspects pertaining to economies suffering financial crisis. Sequencing of prudential measures in this situation is important, and we discuss the likely ramifications for implementing our preferred system. We also discuss the possibility of implementing additional, possibly temporary, measures to enhance confidence in the financial system. An example of such a measure - which may be implementable in a relatively short time-frame and is designed to be consistent with both soundness and competition principles - is the mandatory adoption of minimum subordinated debt holdings by regulated institutions.

II COMPETITION PRINCIPLES: GENERAL THEMES

The enhancement of competition in a previously over-regulated industry normally involves a process that enables greater working of market mechanisms unfettered by government regulations. However, it is also generally accepted (even by most market liberals) that some markets may, on occasion, be fettered by their own structure (e.g. natural monopoly, or cartelisation). In these rare cases some government intervention may be required to achieve welfare-enhancing outcomes. In this paper, we adopt this approach.

We focus on a policy process designed to enhance competition between and within markets. Normally, this will take the form of removing previously imposed regulations which inhibit market forces from operating fully. At (rare) times, however, they may also take the form of imposing regulations that enhance market competition where unfettered markets would otherwise result in restrictive practices or in other non-competitive outcomes. Our starting point is therefore to determine the principles that characterise competitive markets.

One highly relevant project which can assist in determining general principles is the PECC Competition Principles Project. The Pacific Economic Co-operation Council (PECC) is working to establish a set of competition principles (as opposed to rules or minimum standards) that might guide the development of an international competition policy framework. The intention is that these principles will provide a framework for all industries (goods and services) including investment and cross-border transactions. They provide a coherent and sensible set of competition principles which we may apply to the financial sector.

The PECC initiative has been taken within the context of APEC's stated competition policy objective, as follows:

APEC economies will enhance the competitive environment in the Asia-Pacific Region by introducing and maintaining effective or adequate competition policy and/or laws and associated enforcement policies, ensuring the transparency of the above, and promoting co-operation among APEC economies, thereby maximising, inter alia, the efficient operation of markets, competition among producers and traders, and consumer benefits.

The Principles recognise that competition policy is designed to protect (or at least provide "minimum distortion" to) the competitive process, not to protect competitors/ producers. Thus sound competition principles are concerned with facilitating total welfare, not just producer welfare. This principle will be important in the application to financial markets.

The Principles include a number of common themes. An over-arching theme is:

The openness of markets to contest from all sources of supply.

Within this over-arching theme are a number of sub-themes. The most relevant for our purposes can be summarised as follows:

The primary importance �� for enhancing the contestability of markets and the competitive process �� of generally reducing, and where possible eliminating, government barriers to market entry (including �� but certainly not confined to �� barriers to cross-border trade and investment in goods and services).

The associated importance of ensuring that the erection of private anti-competitive barriers (or a hybrid of government/private barriers) does not work against other measures designed to achieve more open, contestable and competitive markets.

The merits of an integrated approach to policy for the purposes of promoting competition, efficiency and welfare. Here it is noted that there is an inter-relationship between competition policy and other policies. Where other policies are required to implement particular objectives, the aim should be to minimise the regulatory distortion to the competitive process consistent with achieving the particular policy objective. [This will be an important consideration in the application to financial markets.]

The importance of competitive neutrality and comprehensiveness within economies �� meaning: where general competition disciplines are in place, the same disciplines apply generally to all commercial producers/suppliers, whether domestic or foreign, government or private, large or small. Here it is noted that this would not preclude industry-specific regulation if an economy judged this to be appropriate to address an industry-specific competition issue [again relevant to the financial sector].

The fundamental importance of transparency (to minimise discriminatory treatment) in respect of any competition disciplines or regulations, the criteria for application of any such disciplines or regulations, and discretionary decisions in relation to these.

These themes lead PECC to propose a set of fourteen Competition Principles, the key features of which include:

1. Adopt an integrated approach to the development and application of competition-driven policies across all areas.

2. Foster greater reliance upon market mechanisms and the role of competition in allocating scarce resources within and between markets.

3. Minimise exceptions from reliance upon market mechanisms. Where market intervention is deemed necessary, it is to be applied with minimum distortion to the competitive process, with net welfare gains being explicitly identified.

4. Ensure competitive neutrality through non-discriminatory application of the same competition principles across goods, services and direct investment provided by foreign or domestic suppliers from the private or public sectors.

5. Foster an efficiency-based approach to competition recognising that competition on the basis of economic merit (lower costs, competitive prices, improved product/service quality, innovation) is the relevant standard.

6. Minimise uncertainty for business through transparent and consistent application of rules, and avoidance of unclear roles, and procedures.

7. Facilitate the competitive process by progressively eliminating government regulations that create or maintain barriers to market entry.

8. Eliminate government influences that impede the ability of market players to compete through innovation and efficiency.

9. Minimise the risk that government efforts to make markets more competitive are replaced or impeded by anti-competitive business conduct, through adoption of appropriate competition disciplines on business conduct �� which may include a general competition law.

10. Design these competition disciplines on business conduct so that they are transparent and solely focused on the objective of promoting competition and efficiency. Where a comprehensive competition law is considered appropriate, it has the following characteristics:

• minimal exemptions or exceptions by sector or operation (e.g. government or private, domestic or foreign-owned);

• non-prescriptive in relation to types of business practices;

• enabling of a diversity of business transactions;

• is prohibitive of specific business conduct only where this is unambiguously harmful to economic efficiency and economic welfare.

11. Ensure that arrangements for enforcement implementation:

• provide for clear accountabilities;

• serve public not private interests, and hence serve total economic welfare;

• are alert to potential mis-use (by businesses) of enforcement procedures;

• serve to encourage self-enforcement;

• are independent of inappropriate government influence.

12. Have regard to any benefits that might be expected to flow from convergence of approaches to competition-based policy.

13. Be alert to the potential for benefits of co-operation among national competition agencies/authorities.

14. Provide for appropriate transitional features in relation to implementing new competition policies.

��

III SPECIFIC FINANCIAL SECTOR CONSIDERATIONS

Before applying these general competition principles to the financial sector, we turn our attention to specific financial market issues which may complicate the application of these principles.

Financial markets are in many ways like the markets for commodities (e.g. wool, bananas, etc) but in many ways they differ. They are alike (in an unrestricted market) in that many financial prices (e.g. money market rates, exchange rates) are set in a quasi-auction manner with many buyers and sellers. However they differ in two important respects:

1. crucial information (e.g. on credit-worthiness of borrowers and of banks themselves) is often asymmetrically distributed and is costly to obtain; and

2. failure of major financial institutions may impose considerable external costs on other agents.

These points of difference complicate the application of competition principles. We explore them further below.

The financial sector is one of the most heavily regulated sectors in almost all economies, in part because of the two factors cited above. A related cause is that banks (and other financial institutions) are particularly susceptible to "runs" by debt-holders (depositors) because the majority of their liabilities are in the form of demand deposits. Depositors, who find information about the solvency of a bank difficult to process or costly to obtain are individually best advised to withdraw their deposits from a bank which has any solvency doubts rather than acquire or process the relevant information. At a bank-wide level, this behaviour may lead to a run on a particular bank. The run may force the bank to liquidate assets quickly, and hence often at prices less than those which would be obtainable if more orderly disposal were allowed. In turn, this may lead to an otherwise solvent bank becoming insolvent.

That in itself would not be cause for policy intervention, but the failure of a bank in these circumstances causes problems for other borrowers and depositors (as well as for shareholders in the bank). Also, the failure of a bank may (and, on a number of occasions, has) led to runs on other banks, for similar reasons (the failure of an institution sows the seed of doubt regarding the solvency of other, especially alike, institutions). The result is that existing and potential borrowers, even those who are credit-worthy in normal times, face difficulty in obtaining loans. This can have both micro-economic distortionary effects, and macro-economic effects resulting in recession or depression.

In what way do these issues impact on competition policy (and hence issues of liberalisation) within the financial sector? The key factor in this respect is the ability of a less sound financial institution (than others in the market) to compete against other institutions.

In most markets, quality is reasonably observable and is priced accordingly. Thus people are prepared to pay more for luxury cars than for a family sedan. According to the factors discussed above, this may be more difficult to ensure in the financial sector if there are no information disclosure or other requirements placed on institutions. Again this factor is not specific to the financial sector and many economies have consumer protection laws that prevent false claims being made by producers about their products. But in most sectors, there is no requirement on producers to reveal all potentially relevant information.

Interestingly, the food sector is one industry, where disclosure requirements - e.g. regarding ingredients or nutritional value - are mandated in a number of jurisdictions. These are normally justified by the difficulty of consumers verifying the relevant information and because of safety reasons relating to the importance of the industry to human life. The financial sector, where information is also difficult to verify and where the industry is of material importance to human well-being, may be regarded as an equally deserving case for disclosure or other regulations. This implies the potential for some industry-specific interventions, representing a departure from a pure free market approach to competition policy. These types of interventions are envisaged in the Competition Principles above (e.g. Principles III and IX). However the caveats that go with those Principles must also be considered.

A comprehensive examination of the reasons for regulatory intervention in the banking system has recently been provided by Dewatripont and Tirole (1994). They specifically refer to competition issues at a number of junctures.

Firstly, they note an upsurge in bank failures in the post-1970 period at a time when "there has been a sharp increase in competition [and] new and risky activities have been on the rise" (p.25). They argue that there has been both more competition among banks and more competition with other financial intermediaries. These developments have occurred partly because regulatory and cartel-type agreements have withered, and partly because of technological and other changes which have facilitated greater competition.

Secondly, financial intermediation is an industry where (theoretically at least) there are increasing returns to scale. This arises because the overall risk to the portfolio of a financial institution becomes smaller the greater the number of independent loans it has on its books (through standard principles of diversification). This leads to a reduction in the number of financial institutions relative to an industry without this increasing returns to scale feature. In turn, this may reduce potential competition and increase the possibility of abuse of market power. In practice, this feature does not appear to lead to a single bank emerging in most (or any) jurisdictions, with the implication being that there are limits to increasing returns to scale. If this is the case, then interventions (especially those cognisant of the efficiency considerations referred to in Principle V above), should ensure that institutions that can improve efficiency through the application of increasing returns to scale, should be allowed to do so. It is only the case of a dominant firm abusing its market power that should cause concern for competition policy.

Thirdly, competition may prevent efficient networks emerging. An example is networks for automatic teller machines (ATMs). At any point in time, efficiency normally requires a single ATM network to be in existence in any jurisdiction to maximise consumer choice. However, banks may have the incentive not to make their networks compatible so as to increase differentiation and so reduce competitive pressures. This, however, is a static notion of efficiency. Dynamic efficiency (which stresses the ability of competing networks to innovate, so engendering price and quality improvements over time �� as per the "creative destruction" of capitalism described by Joseph Schumpeter) may reverse the conclusion that only one network should exist at any one time.

Fourthly, and importantly from the current perspective, the opening up of competition through a lowering of barriers to entry without any prudential restrictions, may in some circumstances lead to a Pareto worsening in welfare. This comes about through an increase in risk taking by financial institutions faced with greater competition in light of some of the information issues discussed above.

��

��

IV COMPETITION PRINCIPLES: FINANCIAL SECTOR

The task now is to apply the general Competition Principles to the financial sector, taking into account the specific features pertaining to financial markets. Our approach is to list the features of intervention which may be required in financial markets to ensure efficient competition, and then to combine these with the general competition principles to form a list of proposed competition principles for the financial sector against which we can gauge progress on liberalisation.

In the previous section, we listed four key competition issues isolated by Dewatripont and Tirole. The first is principally a description of what has happened, and does not require any intervention provided "unfair" competition (e.g. provision of misleading information) is not allowed (see below). The second and third competition issues can be covered by non-industry-specific legislation that includes:

• prevention of market dominance (where dominance is defined by the ability of a firm to consistently abuse market power through pricing that results in significant distortions from optimal resource allocation); and

• access of competing parties to networks.

If these features are not covered by general competition law, then specific provisions may be required for financial markets. However our starting point is that these are general issues and so should be covered by general rather than industry-specific legislation.

This leaves the fourth issue relating to the need for actual and potential bank clients to have access to sufficient information to prevent banks competing unfairly by posing as a higher-quality (sounder) institution than they actually are. Of the four issues, this is in many ways the most complicated and the most serious. Because it is a problem of information, the appropriate action is to ensure that appropriate information is made available to actual and potential clients, and their monitoring agents; there is nothing in the analysis above that suggests - from a competition angle at least - that regulation of bank activities is required (other than disclosure). This leads to the following interventions designed with competition aspects in mind:

1. Mandatory disclosure of all relevant information at the level of the individual bank (i.e. non-customer specific) required to assess the soundness of a bank. This information includes, inter alia:

the bank's liabilities broken down at least by duration and/or maturity structure;

the bank's assets broken down into duration and/or maturity structure and into credit categories - e.g. government risk, other bank risk, households, rated corporates by credit rating, unrated corporates, etc; and with the value and pricing of any equity holdings listed;

• connected lending exposures (to parents, affiliates, etc; which may have a different risk profile from other exposures);

• large exposures to individual parties (which pose greater risk than cumulated small exposures of the same credit class and of the same value);

• foreign exchange exposures (with measures of how bank capital would be affected by changes in exchange rates - particularly the home economy exchange rate);

• interest rate exposures (with measures of how bank capital would be affected by changes in interest rates of various terms);

• the type and nature of all derivative exposures (with measures of how bank capital would be affected by movements in variables influencing the value of those exposures);

• the bank's capital (using unweighted assets, although the Basle weighted asset measure may be provided additionally);

• nature of risk management systems, including attestations that adequate risk management systems are in place.

2. Mandated accounting treatments for all types of disclosed information. This may be mandated by regulators, or devolved to specified accounting bodies, or even to reputable financial industry bodies, provided that individual banks do not have the power to impose their own conventions.

Much of the information to be released is complex and beyond the ability of individual depositors to interpret. However specialised agents exist to process this information and inform depositors of their views. These agents include auditors; legislation is likely to require that disclosed information must be audited by an audit body that is independent of the disclosing institution. Other agents include financial advisers and credit rating agencies. The most assiduous (and knowledgeable) processors of the information are likely to be rival banks. In each case, the results of the analysis will flow through to depositors, either directly (e.g. from financial advisers to their clients) or through media channels via financial journalists.

These provisions therefore correct for the key information problems in the financial services industry which enable unfair competition to emerge. However they do not necessarily fix the contagion (or "runs") problem that can characterise the industry and result in the downfall of sound financial firms as a result of problems in an unrelated financial firm (however, provided the "sound" firm is indeed sound, the provision of reputable information should by itself reduce the risk of contagion).

Nor do they correct for contagion effects that may occur, not through explicit runs, but rather through a breakdown in the payments system caused by the failure of a single institution. In a "batched" (as opposed to "real time") payments system, failure of one institution can lead to failure of other institutions if expected payments from the failed institution are not made.

For this reason, authorities may require extra legislation to prevent runs and to ensure a sound (preferably real time) payments system. These measures are not strictly for competition purposes, but instead for soundness and stability purposes. This can be seen as an efficiency reason for intervention that complements the efficiency approach to competition policy already described in principle V above.

At the least, this factor means that two additional forms of intervention are required:

Firstly, the authorities must be active in ensuring that the economy's payments system(s) does not contain the feature that failure of one institution brings down other institutions by virtue of unintended intra-day inter-bank exposures through the payments system. The most efficient solution is to establish real-time settlement (at least for all large transactions). This measure, however, cannot be implemented quickly, and some form of central bank lender-of-last-resort facility (available only to institutions affected by a failed institution via intra-day exposures) may be required. A legally sound "netting" framework may also be of considerable use in reducing gross exposures of other institutions to a failed institution to (generally much smaller) net exposures. These are system-wide design features which supplement, rather than replace, the need for regulations to be applied to individual institutions.

Secondly, a restriction is required which has the effect of giving authorities the power of suspension of convertibility. This means that authorities can impose on depositors the requirement that deposits must be left with a specified institution for a period of the authorities' choosing even if the contract between the institution and the depositor recognises that the deposit has matured. While it is difficult to argue that this restriction is purely for competition reasons, it is reasonable to conclude that the imposition or retention of such a restriction is compatible with a (sound and efficient) liberalised financial system.

Beyond these restrictions, it is difficult to substantiate reasons as to why any further restrictions are strictly required in normal circumstances (see below for discussion regarding extra responses that may be appropriate in times of crisis). In particular, if the above interventions are in place, there is little reason to insist on capital adequacy requirements, or limits on maximum exposures, or exchange rate, interest rate or derivatives exposures. The exposures in each of these cases would have to be transparently disclosed according to proper accounting practices and so would act as a strong ex ante market discipline on institutions wishing to attract business. If, for whatever reason, an institution did run into difficulties, the information released by sounder institutions should help protect them from contagion; and if the worst occurred and there was a run on sound institutions, the ability of authorities to suspend convertibility prevents those institutions from inefficiently having to liquidate assets quickly at below-market prices so protecting the value of their equity.

One exception is if there are banks that are regarded as "too big to fail"; i.e. cases where a single bank's failure would be so influential as to materially harm an economy's prospects. In this case, bank management, shareholders and depositors may count on the authorities to "bail out" depositors (and possibly even equity-holders) in the event of failure. Armed with this expectation, bank management may adopt a riskier asset management strategy to achieve higher expected returns, with depositors accepting this higher risk profile on the grounds that official intervention is expected to mitigate any down-side risks. In this circumstance, authorities may require a "too big to fail" institution to operate within certain prudential limits (including posting a minimum capital ratio sufficient to reduce probability of failure to very low levels).

It is possible to limit the scope of interventions by having more than one class of financial institution. Full disclosure and/or other restrictions would be required for one set of institutions and less than full (or even no mandated) disclosure for another set.

This menu of options approach is suggested by Dewatripont and Tirole (p.115) who contend that it may be better to allow institutions to self-select their class (and hence their level of regulation) rather than impose blanket regulations on all institutions (albeit with the potential to force "too big to fail" institutions into the more heavily regulated category). For instance, some institutions may not wish to disclose information (for whatever reason) and the menu of options approach enables them to protect their information, but choosing this option means foregoing inclusion as a bank (or whatever other name is used to denote the class of most heavily regulated financial institutions). This is, of course, on the proviso that depositors know the requirements laid down for each class of institutions. This approach enables competition between classes of financial intermediaries and so is preferable from a competition policy viewpoint to a blanket restrictions approach. It also serves as a very useful check on the over-use of regulatory interventions by banking authorities.

Given this analysis, we conclude that the general competition principles (I - XIV above) can all be retained for application to financial markets, but principle III can be extended to clarify the nature of desired interventions as follows:

III' Market mechanisms should be supplemented by restrictions pertaining to a class of financial institutions that:

require those institutions to make mandatory disclosures of relevant information specified by the authorities;

require them to meet accepted accounting requirements for all accounting purposes including disclosures;

- enable authorities to alter contracted terms in the face of contagion, by suspending convertibility of all deposits in specified institutions for an indefinite period; and

allow institutions to select whether they wish to be included in the class of institutions having these extra requirements imposed upon them or to be exposed only to general competition and other laws.

Principle XI of the general competition principles specifies that there should be clear accountabilities for enforcement of restrictions such as those specified above. This has two corollaries:

Firstly, there needs to be a single, clearly-specified agency which monitors (and implements any action relating to) the restrictions, including whether banks are adhering to the requirements. Any sanctions on banks must be carried out in a systematic manner without greater "forbearance" for some banks relative to others which have breached requirements.

Secondly, the banking legislation must make it clear who, within the bank, is legally accountable for ensuring that the restrictions are met, and the stated individuals must indeed be held legally accountable.

This second corollary relates to issues of corporate governance within banks. Given that the suggested approach includes the requirement that banks undergo greater regulatory control than other firms, it is necessary to clearly specify who is accountable within the bank for ensuring the bank complies with these regulations. Those people then have the incentive to ensure that governance procedures are such as to ensure compliance.

The recommended interventions above do not specifically recognise the Basle accord framework of the BIS which include the imposition of additional mandatory requirements on banks. Most prominent amongst these is the recommended minimum capital ratios for banks, including a minimum 4% of tier 1 capital relative to risk-weighted assets and a minimum 8% of tier 1 plus tier 2 capital as a ratio of risk-weighted assets.

Given the general competition principles XII and XIII above, and given that the Basle accord is an international framework (albeit not a mandatory one), each economy's authorities may well wish to impose the Basle standards on their banking systems, even if "first principles" analysis suggests that interventions of this nature are not strictly required (other than in the "too big to fail" case).

This is particularly the case in response to a crisis situation. In such times, measures which enhance confidence in, and the reputation of, the financial system are to be encouraged. Given that the Basle accord is well accepted and widely adopted internationally, adoption by a crisis-hit economy of the various BIS standards is likely to be helpful at least in the short term in re-establishing confidence in the economy's financial system. This includes adoption of large exposure limits and market-risk measures, and especially restrictions on connected lending (to parents, subsidiaries or affiliated organisations). Once the crisis is well past, some relaxation of these measures (if considered appropriate from both efficiency and soundness perspectives) may be considered, although retention of connected lending limits will almost certainly be required given the experience of poor lending practices to affiliates in the region.

If this course of action is adopted (consistent with principles XII and XIII) authorities should retain the menu of options approach suggested by Dewatripont and Tirole, so enabling institutions to choose whether they wish to be regulated as a "bank" in these terms, or as a less heavily regulated non-bank financial institution - again with the proviso that depositors are clearly informed of the difference, and that "too big to fail" institutions be treated as (more heavily regulated) banks.

Other than the above exceptions, the general competition principles should apply mutatis mutandi to financial institutions (even though application of some of these principles are frequently lacking in practice). This implies that within financial markets, from a competition policy standpoint, there should be:

open entry of financial institutions (that meet basic prudential requirements outlined above) to domestic financial markets, whether the institutions are domestic or foreign owned; and

freedom for (private or government) financial institutions to operate (within the restrictions listed above) without government interference in pricing (interest rates), loan approvals, and choice of services.

It is the case that in many economies (both developed and developing) the first of these principles is not observed. From an economic viewpoint, it is difficult to see any reason for imposing controls on foreign ownership in this sector. The only apparent reason seems to be if governments wish to direct loan and/or other decisions of banks in a non-transparent manner. They may find it easier to direct domestic than foreign owned enterprises, and/or easier to direct enterprises which are not exposed to market entry from more efficient (foreign or domestic) financial institutions.

However, if we take from the general competition principles that institutions should be free of government interventions in their business decisions (other than the prudential constraints outlined above), then these rationales for preventing foreign entry disappear. Provided there is open entry, normal business practices by financial institutions will ensure that credit flows to the most efficient potential borrowers (in terms of the risk/return trade-off from the investment), without the need for government directives and without fears of monopoly foreign control. Thus there is no more need for any differential treatment of the financial sector in this regard than there is for any other industry.

��

An economy undergoing financial crisis will wish to adopt measures that promote confidence in the system and delay measures which may reduce confidence in the short-term even where they are designed to increase confidence longer term. Already we have noted that this may encourage adoption of the full BIS recommended standards for banks, even if these are not necessarily thought to be fully justified long-term.

There are also implications for the introduction of mandatory disclosures. The adoption of mandatory disclosure is effective principally as an ex ante control measure. Bank management and directors will take considerably more care in their actions where they know that the resulting positions arising from those actions will have to be disclosed on a frequent (e.g. quarterly) basis. This is particularly the case where bank directors and/or management are held legally (and financially) accountable for the veracity of the disclosures. Thus mandatory disclosure and associated accountability arrangements can be expected to reduce risk-taking by banks where those risks would appear imprudent to depositors.

As an ex post measure, mandatory disclosure is of little use, and may even be detrimental. Once an institution is in trouble, disclosure can worsen the situation by promoting a run on the troubled bank. In the current situation within much of the Asia-Pacific region where many banks are not meeting required standards, the imposition of mandatory disclosure may therefore be detrimental.

For this reason, it is recommended that economies only adopt this approach once banks are in a sound position. This means action to quickly write down asset values to true market worth, then re-capitalisation of institutions with insufficient capital prior to adoption of a mandatory disclosure regime.

The corollary is that once banks are re-capitalised (with appropriately valued assets) a mandatory disclosure regime is likely to enhance confidence in individual institutions and hence in the financial system, provided that investors can be confident that disclosures are accurate. Thus it is vital that a disclosure regime is implemented as part of a comprehensive package that includes accounting standards and clear responsibilities (and penalties) for bank directors and/or management regarding breaches of the regime.

There is one other measure, which does not naturally arise as part of a competition approach to policy that may be useful both in normal times and (more especially) in times of crisis. Our earlier analysis attributed the asymmetry and cost of information regarding banks and other financial institutions as a major problem with the financial sector. Individual small depositors do not have the incentive to monitor institutions, while government authorities may not be effective monitors and/or may suffer from regulatory forbearance. Equity holders may have the incentive to increase risk-taking following a "hit" to the balance sheet to restore capital (although this will be considerable lessened by frequent disclosures and accountabilities on directors regarding risk-controls).

One class of liability-holders which can assist monitoring in such circumstances are large subordinated debt holders. I have argued elsewhere (Grimes, 1996) that a disclosure regime may be enhanced by imposing on banks the requirement that they also hold a minimum level of subordinated debt, preferably with a minimum level on the largest holding of subordinated debt by a single debt-holder. More recently, United States commentators have raised a similar possibility. The advantage of imposing this restriction is that a subordinated debt holder does not face the same risk-taking incentives of an equity holder to "bet the bank" at times when capital is approaching zero, but at the same time has the incentive to monitor bank soundness much more closely than a standard depositor (both because they are "large" and because they are subordinated to other debt-holders).

This restriction could be adopted relatively quickly by most jurisdictions, it would not directly affect equity-holders, and (provided appropriate subordinated debt holders were found) would enhance the reputation of the banking systems. It does not cut across the competition principles adopted above in any material way, especially if the menu of options approach were retained, so that this requirement were imposed only on "banks". Like some of the BIS measures, the requirement could be relaxed over time if considered appropriate, although its adoption and retention even in normal times does warrant attention.

��

REFERENCES

Australian Competition and Consumer Commission (1997) "Westpac Banking Corporation/Bank of Melbourne Limited: Background to Decision on Merger Proposal".

Bernanke, Ben (1983) "Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression", American Economic Review, 73(3), June, 257-276.

Besanko, D. & A. Thakor (1993) "Relationship Banking, Deposit Insurance and Bank Portfolio Choice", in: C. Mayer & X. Vives (eds) Capital Markets and Financial Intermediation, Cambridge University Press, Cambridge, 292-318.

Dewatripont, Mathias and Jean Tirole (1994), The Prudential Regulation of Banks, The Walras-Pareto Lectures, at the Ecole des Hautes Etudes Commerciales, Universite de Lausanne, MIT Press, Cambridge, Mass.

Diamond, Douglas (1984), "Financial Intermediation and Delegated Monitoring", Review of Economic Studies, 59, 393-414.

Diamond, Douglas and P. Dybvig (1983) "Bank Runs, Deposit Insurance and Liquidity", Journal of Political Economy, 91, 401-419.

Grimes, Arthur (1995) "Review of M. Dewatripont & J. Tirole: The Prudential Regulation of Banks (MIT Press, Cambridge), Economic Record.

Grimes, Arthur (1996) "Public Standards and Private Monitoring: New Zealand's New Banking Supervision Regime", Agenda, 3(3), 277-286.

Grimes, Arthur (1998a) "The Asian Financial Crisis: How Long Will the Downturn Last?", New Zealand Economic Papers, 32(2), 215-224, December.

Grimes, Arthur (1998b) "The Role of Competition Policy and Financial Market Development: Australasian Experience", Paper presented to ADBI, ISEAS, PECC Workshop on Managing Asia's Financial Sector Recovery: The Role of Competition Policy and Corporate Governance, Singapore, November.

Hayek, Friedrich (1961) The Constitution of Liberty, Routledge.

Matutes, C. & J. Padilla (1991) "Shared ATM Networks and Banking Competition", mimeo, IAE, Barcelona and CEMFI, Madrid.

Yanelle, M.O. (1989), "The Strategic Analysis of Intermediation", European Economic Review, 33, 294-304.