Financial Stability Review – September 2004 3. Developments in the Financial System Infrastructure
A prerequisite for a stable financial system is sound financial infrastructure – the regulatory, accounting and legal framework that supports the day-to-day operations of financial intermediaries and markets. Over the past six months, agreement has been reached internationally on the implementation of two important initiatives to reinforce this infrastructure: the new Basel Capital Framework, which will enhance the prudential oversight of the international banking system; and the International Financial Reporting Standards, which will improve the transparency and comparability of financial statements across countries. Taken together, these initiatives have the potential to provide market participants with better measures of risks within the financial system and the scope to better manage those risks. Nonetheless, they are complex changes and, as discussed below, raise some challenging issues.
3.1 The New Basel Capital Framework
The new Framework, developed by the Basel Committee on Banking Supervision, was finalised in June and will be available for implementation internationally from the end of 2006. It represents a major advance on the current Accord, established in 1988, providing a significantly more risk-sensitive approach to the setting of regulatory capital requirements for banks – a part of the new Framework that is usually referred to as ‘Pillar 1’. The new Framework also aims to reinforce the supervisory review process – the dialogue between prudential supervisors and banks on the adequacy of their capital and on their overall approach to risk management (‘Pillar 2’) – and to strengthen market discipline by enhancing the transparency of banks' financial reporting (‘Pillar 3’).[4] The three Pillars are intended to be mutually reinforcing. At this stage in the implementation process, however, it is the new arrangements for calculating capital requirements that are commanding most attention.
Under the new Framework, a bank must hold capital equal to at least 8 per cent of its risk-weighted assets – a core requirement that remains unchanged from the current Accord. However, the means by which various assets are risk weighted has been overhauled with potentially far-reaching implications for the regulatory capital requirements of individual institutions and, by extension, for the financial system as a whole.
In calculating capital requirements under the new Framework, a bank must have regard to three business risks – credit risk, market risk and operational risk. The approach to measuring market risk – the risk of trading losses – is unchanged from the current Accord, while that for operational risk – covering losses resulting from events such as fraud and technology failure – has been formally incorporated into the capital adequacy framework for the first time. The measurement of credit risk – the risk of losses arising from default by customers or counterparties and by far the largest risk for most banks – has been considerably reworked to provide more risk sensitivity.
The new Framework provides banks with three options for measuring credit risk, with the choice depending largely on the sophistication of a bank's risk management systems. The simplest option is the standardised approach, which is conceptually similar to the current Accord: a bank allocates a risk weight to each of its assets (and off-balance sheet positions) to produce a sum of risk-weighted values against which it must hold capital. Under the current Accord there are, however, only a very limited number of risk weights that apply and these depend largely on the type of borrower (i.e. sovereigns, banks and corporates). An advantage of the standardised approach under the new Framework is that it will allow banks to refine these weights by reference to the risk assessments available from external credit rating agencies. For example, while the current Accord provides only a single risk weight of 100 per cent for corporate lending, the standardised approach will see corporate borrowers assigned to one of four categories based on their credit ratings, with risk weights of 20, 50, 100, and 150 per cent. Of course, not all borrowers have a credit rating, in which case the existing 100 per cent weight will continue to apply.
The more innovative part of Pillar 1 is the provision of two internal ratings-based (IRB) options for calculating regulatory capital. These will allow banks to rely, for the first time, on their own estimates of credit risk to determine the amount of capital they are required to hold. In the case of the ‘foundation’ IRB approach, banks will need to estimate the probability of default for each borrower with other determinants of credit risk, including the likely loss given default, provided by the supervisor. In the ‘advanced’ IRB approach, banks can use their own estimates of both probability of default and loss given default to determine the capital requirement.
Implications of the New Framework for the Financial System
The new Framework will obviously influence the behaviour of individual banks, but it will also have some important implications for the financial system as a whole. A welcome initiative, for example, is the incentive created by the new Framework for banks to invest in advanced risk management systems. Capital requirements under Pillar 1 are such that banks adopting an IRB approach to measuring their credit risk should, for a given risk profile, generate some capital savings over those that adopt the standardised approach. Ultimately, an improvement in risk management should promote the stability of both individual institutions and the system as a whole.
In a similar vein, more relevant and timely disclosure of information about the risks that financial institutions are incurring should promote more effective market discipline. To the extent that problems are identified and dealt with at an earlier stage by both the market and prudential supervisors, improved disclosure can assist in ensuring that problems do not develop to the point where serious difficulties are inevitable.
In addition, by aligning regulatory capital more closely with underlying risks, the new Framework should enhance not only the stability of the financial system, but also its efficiency. In particular, it should reduce the incentives for financial institutions to develop structures specifically to arbitrage differences between regulatory capital requirements and their own internal assessment of the appropriate level of capital. Risk-based capital requirements may also prove helpful in encouraging more risk-based pricing.
Other implications of the new Framework are less clear from a stability perspective. Two issues in particular have stood out in recent debates. The first is the prospect of some quite sizeable falls in regulatory capital for some lines of business, notably lending for housing. And the second is the issue of ‘procyclicality’ – the potential for the new Framework to amplify swings in the economy by making bank credit more easily available in economic expansions, and more difficult to obtain in contractions.
Potential to Lower the Amount of Regulatory Capital
The intention of the Basel Committee is that the new Framework should not lower the overall amount of capital in the banking system. The Framework will, however, significantly change the capital requirements on specific lines of business. In the Australian context, changes in the capital requirements on residential property loans are particularly important, given that these loans account for over half of banks' total loans.
Under the standardised approach, it is proposed that the risk weight applying to residential property loans will drop from 50 per cent to 35 per cent; or in other words, the minimum capital requirement on housing loans will drop from 4 cents in the dollar to 2.8 cents in the dollar (Graph 50). Under the IRB approach, the fall can be significantly larger, although the outcome will depend upon the estimates of probability of default and the loss given default. According to data from the Insurance Council of Australia, the average annual claim rate over the past couple of decades on insured mortgages has been less than 0.2 per cent (see Box C in the previous chapter). Under the foundation IRB approach, a residential mortgage with a probability of default of 0.2 per cent would generate a capital requirement of just 0.6 cents in the dollar (assuming a loss given default of 25 per cent), which is less than one sixth of the current requirement.
These calculations suggest there is scope for a considerable reduction in minimum capital requirements on residential mortgages. Any reduction, however, may be offset by higher requirements elsewhere. Banks will need to allocate capital against operational risk under the new Framework and there is scope under Pillar 2 for supervisors to require capital to be held against other risks, such as interest-rate risk.
Two other factors are likely to limit any decline in the amount of capital that banks hold. The first factor is the assessment of banks by rating agencies, and financial markets more generally. One reason that Australian banks enjoy relatively high credit ratings is that they are well capitalised. If they were to reduce the amount of capital they hold, or expand their assets without a commensurate increase in capital, their ratings would likely suffer, irrespective of how their regulatory capital ratios evolve. The second factor is that the new Framework is to be phased in gradually. Until at least 2010 there will be a floor under any capital reductions. In 2008 for example, the capital requirements for banks that use either IRB approach cannot be lower than 90 per cent of the requirement calculated under the current Accord.
Procyclicality
One aspect of the new Framework that has been the subject of considerable discussion among central banks and banking regulators is its potential to influence the amount of bank lending through the course of an economic cycle – a characteristic usually described as ‘procyclicality’. At issue is the potential for capital requirements to fall in good economic times and to increase in bad times, and how any such changes are likely to affect the evolution of the economy.
If during an economic downturn banks reassess the probability of default in an upward direction, capital requirements will inevitably increase. Under the foundation IRB approach, for example, if a corporate borrower was downgraded from A to A− (which is equivalent to an increase in the probability of default from, say, 0.18 to 0.31 per cent) the capital requirement on loans to that borrower would increase from 3.2 cents in the dollar to 4.4 cents in the dollar (Graph 51).[5] The standardised approach would deliver a similar outcome. For residential mortgages, increases in probabilities of default would also lead to an increase in capital requirements, but of a smaller magnitude than for business loans.
Since raising capital during a downturn can be costly, banks faced with an increase in their capital requirements may choose to wind back their lending instead – a decision which may exacerbate the slowdown. In an economic upturn, the process has the potential to work in reverse.
One factor that may limit movements in capital requirements over the economic cycle is the use of through-the-cycle ratings. In this regard, the Basel Committee's preference is that ratings represent a bank's assessment of the borrower's ability to repay in weak (as well as strong) economic conditions. Such an assessment should not usually change a great deal over the course of a business cycle. Similarly, the standardised approach is predicated on the assumption that rating agencies take into account the riskiness of borrowers across a complete cycle, rather than at any particular point in time. Under Pillar 2, supervisors are also being encouraged to take into account the business cycle when assessing the adequacy of a bank's capital.
In addition, to the extent that banks operate with capital ratios above the regulatory minimum in good times, they will have scope to accommodate higher capital requirements in poorer times, without having to raise new capital or unnecessarily cut back their lending. Rating agencies will play some role in ensuring that such buffers are retained.
3.2 International Accounting Standards
The International Accounting Standards Board has been working for some years to develop a single set of global accounting standards to improve transparency and international comparability in financial statements. Australia will be one of the first countries to implement a full set of International Financial Reporting Standards, which will apply for reporting purposes from 1 January 2005.
The new standards encourage the use of ‘fair value’ – broadly speaking the use of market values net of transaction costs – for measuring assets and liabilities, particularly financial instruments. This is especially important given the extensive use of derivative contracts by financial intermediaries, which are sometimes recorded off-balance sheet at historical cost – an approach that has led to a misalignment in some countries between the information contained in financial statements and the risk profile of financial intermediaries. While, in principle, fair value could apply to all financial instruments, the new standards are essentially confined to instruments held for trading and assets available for sale.
Implications for the Financial System
The new standards have been generally welcomed as an important step towards a more resilient international financial system. Accounting standards that are mutually acceptable and compatible across borders should help to promote investor confidence in financial statements and, by doing so, foster deeper, more liquid markets in financial instruments.
Nevertheless, the introduction of the new standards is not without some challenges for financial intermediaries and their regulators. The new standards are likely to generate some additional volatility in financial statements as changes in fair value feed through to reported earnings and capital. The extent of this volatility may be mitigated to some extent by the application of hedge accounting rules within the standards, which cover the ability to use derivative transactions to hedge movements in the value of assets and liabilities. After much discussion, it was agreed that the standards should incorporate ‘macro hedging’ concepts under which assets or liabilities could be valued jointly, where the assets acted as a hedge against movements in the market price of liabilities or vice versa. However, the accounting standards do not permit hedge accounting involving demand deposits – a source of concern to those ADIs that tend to hedge the interest-rate risk on these liabilities.
During the development of the new standards, some argued that all financial instruments should be measured at fair value, including loans because, by doing so, financial reports would more accurately reflect changes in the credit quality of an ADI's loan portfolio. This approach attracted much debate with critics arguing that it was impractical, as few loans are actively traded. Accordingly, loans will continue to be reported at book value (less provisions) under the new standards, unless intended for sale by the ADI. While the new standards emphasise the use of ‘objective evidence’ in identifying loan impairment and thus provisions, they provide some scope for a forward-looking assessment using ‘experienced judgment’ where observable evidence is limited or irrelevant.
For regulators, there is considerable work to be done in reconciling various aspects of the new standards with current prudential requirements. For example, some hybrid capital instruments currently treated as equity by regulators will be classified as debt under the new standards and the tests for the accounting derecognition of assets sold by banks to securitisation vehicles will be more stringent than those currently applied by banking regulators. But there is also a more fundamental issue for regulators that arises from differences in the purposes of the new accounting standards and the objectives of prudential rules. Accounting standards measure the economic value of a bank as a going concern, with capital treated as the difference between assets and liabilities. In contrast, prudential supervisors are more concerned with values of assets and liabilities in stress scenarios and the ability of capital to meet the obligations of the bank in these situations. Reflecting this, the Basel Committee has recommended that some unrealised gains/losses arising from changes in market prices be excluded from regulatory capital. The Committee has also argued strongly that where liabilities are measured at fair value, only valuation changes due to general market movements should be taken into account. Doing so would lessen the possibility that a deterioration in an institution's credit rating would lead to a reduction in the value of its reported liabilities.
Like other regulators, APRA is giving very careful consideration to the reconciliation of prudential requirements and the new accounting standards. It has confirmed that no changes to Australian prudential requirements will be made before 1 July 2005 or without industry consultation.
3.3 Other Developments
Study of Financial System Guarantees
The HIH Royal Commission recommended that the Commonwealth Government establish a comprehensive scheme to support insurance policyholders against the failure of insurance companies. In response to this recommendation, the Government commissioned a technical study not only into the merits of this proposal, but the establishment of a system of guarantees for financial products more generally. The study, known as the Davis Report, was released in May. It covers issues including the economic rationale for explicit guarantees; criteria that could be used to determine which financial products are to be guaranteed; the cost of such a system; how it could be funded and priced; and what implications a guarantee scheme might have for the existing regulatory framework.
The Davis Report concluded that the costs and benefits of adopting a system of financial guarantees are ‘finely balanced’. Arguments in support of a guarantee scheme include consumer protection and greater certainty about the resolution of a failed financial institution. The case against explicit guarantees rests on the impact that such schemes can have on private incentives to manage risk. The Bank is of the view that for deposits, at least, there is merit from a crisis-management perspective in a limited guarantee in the form of deposit insurance. Australia is one of the few countries without such a scheme, instead providing depositors with first claim on the assets of a failed bank.
The Davis Report was intended to be used as a resource for parties interested in making submissions to Treasury on financial system guarantees. Treasury is currently considering the public submissions it has received.
Changes to Superannuation Arrangements
Choice-of-fund legislation, which has been on the regulatory agenda since the mid 1990s, was passed in June. From July 2005, this legislation will provide up to five million employees with the freedom to choose the superannuation fund into which they wish to place their compulsory superannuation contributions. The legislation will also affect around 700,000 employers, who will need to notify new and current employees of their rights in relation to fund choice before July 2005, and eventually allocate those that do not exercise choice to an eligible default fund. The new legislation is intended to increase competition and efficiency in the superannuation industry. It may also have implications for financial markets if significant shifts in the location of superannuation monies were to occur.
The Government also released a set of recommendations in early July for legislation designed to improve the protection of superannuation funds. The recommendations relate to the provision of financial assistance by the Government in the event of losses incurred by superannuation funds due to fraudulent conduct or theft.
Business Continuity Management for Financial Institutions
In early July, APRA released for public consultation a draft prudential standard on business-continuity management for ADIs, general insurers and life insurers. The standard is designed to ensure that financial institutions can continue to meet their obligations to customers in the event of a significant disruption to normal operating conditions.
As part of its crisis-management responsibilities, the Bank has also begun working with the financial industry to identify and address, where appropriate, infrastructure risks that could have systemic implications.
Crisis-management plans also have an international dimension. Cross-border co-ordination is particularly important between Australia and New Zealand, due to the size of the Australian banks' trans-Tasman business. New Zealand is the largest country exposure of Australian banks, representing 13 per cent of their consolidated assets, and the four major Australian banks account for 86 per cent of the New Zealand banking-system assets. Reflecting this deep integration, the two Governments established a working group earlier this year to explore options for more closely integrating the regulatory and crisis-management arrangements of both countries. The Bank participated in the working group along with officials from the Australian and New Zealand Treasuries, APRA and the Reserve Bank of New Zealand. The Report is now with the Australian Treasurer and the New Zealand Minister of Finance for their consideration.