Hedge Funds, Financial Stability and Market Integrity 3. Public-Policy Responses

The risks that some hedge funds pose to the stability of the financial system and the integrity of markets create a strong ‘in principle’ case for a public-policy response. The issue is what form of response is appropriate.

There are three broad options:

• the direct regulation of hedge funds;
• the development of more comprehensive disclosure requirements; and
• improving bank supervision and the supervisory framework.

These options are discussed below.

3.1 The Direct Regulation of Hedge Funds

In essence, a hedge fund is a form of mutual fund. As such, there is an argument that hedge funds should be subject to the same form of regulation as mutual funds. While the form of this regulation varies across countries, it often involves licensing and the setting of standards for the issuing of prospectuses. Importantly, in many countries there are also restrictions on the use of derivatives and debt by mutual funds.

The primary reason for this type of regulation is to protect the interests of investors. It is generally argued that such regulation is appropriate for mutual funds, since small investors do not have the ability and resources to fully understand and monitor the risks being incurred by the fund. This argument, however, does not apply to investors in hedge funds. As was argued above, these investors have both the ability and resources to monitor risk, and it would be inappropriate to extend regulation based on the need to protect investors to the hedge fund industry.

An alternative is to design regulations that apply specifically to hedge funds and that are designed to enhance the stability of the financial system and the integrity of markets. Despite the ‘in principle’ appeal of this approach, there are considerable practical difficulties.

The first is that hedge funds may be able to circumvent regulation by basing their activities in non-regulated offshore jurisdictions. This possibility would be reduced if governments and others put pressure on offshore centres to implement internationally agreed standards, and bank supervisors required significantly higher capital charges on exposures to institutions operating from these centres. Ultimately, however, action on either of front requires a high-level political commitment on behalf of the major countries.

Another difficulty is that hedge-fund-specific regulation is likely to lead to other institutions developing outside the expanded regulatory net, with the end result being little change in the degree of systemic risk and the integrity of markets. A further complication is that the case for regulating hedge funds arises largely from their ability to gain large positions in markets. As was discussed above, not all hedge funds have such positions, and so the case for regulation of all hedge funds is weak. It might, however, be problematic to base regulation on the size of an institution's balance sheet, or positions in financial markets.

A more promising approach is to address the conditions and practices that have allowed hedge funds to gain large positions in markets.

3.2 Improving Disclosure

Improving disclosure should make a contribution to both market integrity and financial stability. More comprehensive disclosure would make it easier to identify institutions with market power, and ultimately sanction institutions that abused that power. In addition, through contributing to better credit assessment, it would reduce the probability of credit-extension practices that contribute to financial instability. Disclosure can also help limit market panic in times of stress, by reducing uncertainty about the extent of exposures.

Despite the advantages of disclosure, recent events highlight that the market has failed to ensure that adequate information is available to counterparties, regulators and the general investment community. In part, this market failure reflects the fact that in determining private disclosure arrangements, institutions do not take into account the public benefits that accrue from the release of information.

There is also a coordination problem. Even though some institutions recognise that the system would work better if standards of disclosure were enhanced, they are unwilling to impose tougher standards on their counterparties than those imposed by their competitors. Moreover, individual institutions are sometimes reluctant to disclose to the market extensive details of their own risk profile for fear that such an action would be misinterpreted. To overcome these coordination problems and the apparent divergence in private and public gains from better disclosure some public-sector intervention is required.

Over recent years, the argument for improved disclosure has often been applied to the release of information by governments and central banks, but it applies equally to private-sector participants. In markets which consist entirely of private-sector buyers and sellers, and where coordination problems have been addressed through the development of regulated exchanges (such as share markets and futures markets), strict disclosure standards have evolved to ensure that the market works efficiently and fairly. In markets such as foreign exchange, however, standards of disclosure are much weaker or non existent. This may reflect the difficulties of coordination and of collecting information in over-the-counter markets, as opposed to exchange-traded markets. Also important is that historically these markets have tended to be dominated by governments and central banks, with individual private-sector players having little, or no, market power. But with large investors now taking on foreign exchange positions which rival, or exceed, those of the authorities, it is important that disclosure standards be improved, particularly in the foreign exchange market.

Disclosure requirements should apply to as broad a set of institutions as is possible. If requirements apply to just a subset of institutions (say currently regulated institutions and hedge funds), other types of institutions would evolve outside the regulatory net. In cases in which universal requirements create practical difficulties, specific institutional types could be exempted on the basis that regulation was not needed for market integrity and financial stability reasons. Any exemption could be reviewed from time to time.

In designing enhanced disclosure requirements there are at least three central issues:

• what information should be disclosed;
• to whom should it be disclosed; and
• how should disclosure be enforced.

What information should be disclosed?

There is a need for enhanced information in at least three areas. These include: information concerning market concentration; information that promotes sound credit assessment; and information that allows market participants and regulators to assess the health and stability of markets.

One way of addressing the market concentration issue is to require some form of large position reporting, where large positions are defined in terms of the relevant market. For example, institutions could be required to disclose positions that account for more than some percentage of a market's turnover or outstanding contracts. The benchmark levels for disclosure could be determined by national regulatory agencies, or through international agreement, perhaps through the BIS. At one extreme, institutions might be required to notify the authorities of the specific details of positions that exceeded the relevant benchmark. A less intrusive approach would be to require some form of public reporting in which institutions periodically disclose whether they had ‘large’ positions in certain markets, but were not required to disclose specific details of those positions. Other alternatives are also possible, although one disadvantage of any disclosure based on size of positions is that it increases the incentive for institutions to spread positions across a number of related cash and derivatives markets in such a way that each position is small enough to escape the reporting requirement. One way of overcoming this problem is to disclose statistics which should, in principle, be highly correlated with the size of exposures relative to the market.

The disclosure of information on market concentration should also be helpful from the point of view of making sound credit assessments, as large market positions can involve considerable liquidity risk. Another particularly important element in the credit assessment process is the ‘value-at-risk’ (VaR) of an institution's portfolio and various subsets of the portfolio. One way of providing this data is for institutions to disclose a matrix of risk exposures by country and individual asset type – for example, institutions might be required to disclose the VaR of their foreign exchange positions in both major markets and emerging markets, as well as the VaR on positions in equity and securities markets. Some aggregation across specific markets is likely to be necessary to avoid the reporting requirements from becoming excessively onerous.

While the VaR is a useful summary measure, it can hide a variety of risks. For example, an institution might hold extremely large positions, but report a small VaR if the positions are assumed to be to very tightly negatively correlated (for example, a long position in one bond, and a short position in another bond with closely matched, but not identical, characteristics). This assumption might be valid in normal times, but in times of stress it may fail to hold and liquidity problems may make it impossible to unwind the positions. In such cases, the VaR calculation might seriously underestimate the amount of risk being incurred. An implication of this is that disclosure of the VaR needs to be supplemented with additional information.

One option is for institutions to disclose the assumptions underlying the VaR calculation. While this should be done as a matter of good practice, the complex and detailed nature of these assumptions in many cases reduces the usefulness of this information.

A second option is for institutions to report the results of stress tests that incorporate large movements in market prices and liquidity problems. Once again the difficulty here is developing a methodology that is sufficiently standardised that the results are meaningful. Supervisors in some countries, such as Australia, have made progress in this area and now require the results of standardised tests to be reported to the supervisory authorities. These tests could form the basis of a broader disclosure requirement.

A third, and perhaps more useful, option is to require disclosure of ex post measures of VaR performance. These include the number of days during the reporting period on which losses exceeded the VaR estimate, together with the maximum daily loss. An alternative would be to disclose a histogram of the ratio of daily changes in the value of the portfolio to the daily VaR. These measures of risk could be judged against the institution's risk policy. In addition, institutions could disclose summary measures of risk-adjusted returns – for example, the ratio of the portfolio's volatility to market volatility. Such a measure could provide an indication that an institution had large positions relative to the market, or was highly leveraged. Alternatively, institutions could disclose the standard deviation of changes in the value of the portfolio (scaled by the volatility of the market) or other measures of the distribution of returns, for example, the lower 5 per cent and upper 95 per cent. If these measures were to be disclosed, accepted benchmark measures of market volatility would need to be developed.

In a number of the above areas, the reporting of end-of-period data is of limited value due to the ability of institutions to window-dress their portfolios on disclosure dates. It is therefore important that end-of-period reporting is supplemented with intra-period data (for example, on high and low values).

In addition to the disclosure by individual institutions there is a strong case for additional aggregate data to be made available, as such data are necessary for regulators and market participants to assess the health and stability of markets. The current BIS banking and derivatives statistics are a useful starting point, but these collections need expanding to include the international exposures of investment banks, hedge funds and other institutional investors. A related possibility is to develop some form of international credit registry along the lines of the registries that are currently in operation in a number of countries. This approach was canvassed in the G22 Working Group on Strengthening Financial Systems. This Group advocated work to improve the efficiency of the existing systems, particularly in the area of cross-border exchanges of information.

To whom should information be disclosed?

If institutions are to disclose information such as large positions relative to the market, their VaR, the results of stress tests and VaR model performance, an important issue is to whom this information should be disclosed. One option is for this information to be provided only to an institution's counterparties. This could be achieved by regulators of banks and securities firms requiring that institutions obtain the above information from any other institution with which they are dealing. This approach, while worthy of consideration, has two significant disadvantages. First, it is only partial in nature. In many cases, institutions should be seeking much more detailed information from their counterparties than that outlined above, and regulators cannot hope to prescribe all the relevant information. Second, while disclosure to counterparties might improve credit assessment, it is unlikely to make a significant contribution to improving the integrity of markets.

The alternative to setting disclosure standards that apply to counterparties is to establish standards of public disclosure that apply to all institutions active in financial markets. The advantage of this approach is that the information is available to the market as a whole, including regulators, investors and counterparties. It would ensure that there was a basic minimum amount of information in the public domain concerning institutions operating in financial markets. This information should form one of the building blocks of good credit assessment, but institutions would still be expected to obtain additional information from institutions with which they are dealing. The approach would also allow for more effective monitoring of large positions.

In addition to public disclosure, there is a need for more detailed information to be disclosed to statistical or regulatory authorities so that relevant aggregate data can be published.

How should disclosure be enforced?

A crucial issue is how enhanced disclosure arrangements might be implemented.

One option is for regulatory agencies in each country to mandate minimum standards of public disclosure. This could be done by the regulators of currently regulated institutions (including banks and securities firms) requiring disclosure along the lines discussed above as part of their licensing requirements. For other market participants some form of legislation may be needed. This could involve considerable practical difficulties, especially in ensuring compliance if institutions and investors were not also subject to some form of registration and monitoring. Registering all investors could create administrative difficulties, drive institutions offshore, and ultimately create moral hazard problems for the authorities if registration was equated with ‘government oversight’.

An alternative approach is to work through the institutions that are already subject to prudential regulation. One way of doing this is to use the current Basle capital arrangements to create an incentive for institutions to disclose relevant information. If a participant in financial markets complied with a designated set of disclosure requirements, the standard capital charges on exposures to that participant would apply. A set of penalty capital charges would then apply to non-complying participants. For example, if a hedge fund was unwilling to disclose to the market the relevant information, the risk weight that applied to any derivatives exposures to that institution might be double or triple the weight that applied to exposures to complying institutions. One advantage of such an approach is that it recognises that exposures to institutions which fail to comply with basic disclosure requirements are inherently more risky, and therefore require more capital.

Finally, it is worth recalling that improved disclosure is no panacea for the problems that have plagued financial markets over recent years. There is even a risk that in some cases it could exacerbate problems. Improved disclosure could reinforce the strong reputations of some market participants by highlighting their successful performance. While the reputation might be dimmed by the release of comprehensive information about the risks that the institution is taking, experience shows that on some occasions such information is ignored. The end result might be greater herding in the market, and potentially, the creation of additional market power. Further, the disclosure of large positions relative to the market is no guarantee that institutions will not attempt to develop and use market power.

Notwithstanding these qualifications, improved public disclosure concerning large positions, the risks that institutions are taking, and the overall health of markets, should promote the resilience of the financial system and the integrity of the markets within the system.

3.3 Improving Bank Supervision and the Supervisory Framework

The third possible public-policy response is for bank supervisors to improve the way that they supervise currently regulated institutions. If, over recent years, regulated institutions had paid greater attention to the risks that they were taking, some of the recent problems almost surely would have been avoided.

While, ultimately, the measurement and monitoring of credit and market risk must remain the responsibility of bank management, supervisors have a role to play in ensuring that this responsibility is being met, and that the regulatory arrangements are not encouraging inappropriate risk taking. Possible supervisory responses fall into one of three categories:

• initiatives to improve the credit assessment process within institutions;
• changes to the Basle capital requirements; and
• active use of supervisory instruments by supervisors.

Each of these is discussed below.

3.3.1 Improving the Credit Assessment Process

One of the lessons from recent events is that the credit assessment processes within a number of major international financial institutions were deficient. All too often institutions have put too favourable a gloss on both credit and market risks. In part, this reflects the lack of appropriate internal systems for the measurement and management of these risks. But even where these systems exist, they have often been overruled in an effort to build market share and to preserve and strengthen trading relationships.

These problems occurred despite a long-standing recognition that financial institutions' risk control procedures with respect to hedge funds needed improving. In 1994, Fed Governor LaWare testified to Congress (LaWare (1994) page 516):

Nevertheless, banks … need to carefully monitor their relationships with hedge funds. … Financial firms should continue to place the highest priority on reviewing, assessing, and improving their overall risk management practices. The Federal Reserve intends to continue to use its bank supervisory authority to make certain that further progress is made in this area and that risks are being adequately controlled.

Despite the Federal Reserve's advice to financial firms and its commitment to ensuring that further progress was made, risk management procedures did not keep pace with the changing environment. The practice of supervisors reminding and even imploring banks to be prudent had only limited effectiveness. An alternative approach needs to be found.

One promising alternative is the development of a set of standards, or sound practices, that institutions would be required to follow. These standards could effectively set some benchmarks against which institutions could measure their own internal procedures. They could also provide the basis for reporting exceptions to supervisors, and ultimately reporting exceptions to the market, through the institution's quarterly or annual reports. This second step is an important one. Mechanisms need to be developed to increase the incentives of institutions to comply with sound practices. Otherwise, faced with strong competitive pressures and confronted with strong market reputations, institutions are likely to ignore sound practices, just as they have done in the past.

The first element of this approach – the development of a set of standards for banks dealings with highly leveraged institutions (HLIs) – is currently being explored by the Basle Committee on Banking Supervision. The proposed standards cover the following areas: (i) the development of internal policies that govern banks' relationship with HLIs; (ii) the collection of information about the activities and credit risk of HLIs; (iii) the development of accurate measures of exposures resulting from trading and derivatives activities; (iv) the setting of meaningful overall exposure limits; (v) the appropriate use of collateralisation requirements; and (vi) the processes for monitoring credit exposures.

In developing standards in these areas two issues deserve particular attention: the use of stress tests and collateralisation requirements.

1. Stress Tests. Despite the considerable resources that some institutions have devoted to the modelling of market risk, the size of recent trading losses came as a major surprise; the widely used modelling techniques simply failed to capture the extent of the risk that institutions were incurring. The modelling approaches ignored the fact that in times of market stress, interrelationships between markets can change dramatically, liquidity can dry up, and assets need to be sold at distressed prices. These ‘facts’ were ignored by even the largest and most technically sophisticated institutions. One result of this was that they were prepared to grant huge trading lines and incurred unexpectedly high risks.

• One way that supervisors can contribute to better risk measurement is to require that regulated institutions undertake stress tests of their exposures. These tests should factor in the possibility of major market disruptions and liquidity problems. Particularly important is the need for banks to conduct stress tests on credit exposures arising from traded-markets activities. The extension of stress-testing techniques to the analysis of credit risk more broadly should also be pursued.
• Supervisors also need to work with institutions to develop appropriate reporting arrangements for the results of these tests. As noted earlier, the lack of a standardised methodology can make it difficult to interpret the results. One approach is for supervisors to require institutions to report tests based on standardised assumptions, but also to encourage institutions to conduct tailor-made tests that more closely focus on their own risk exposures.

2. Collateralisation Requirements. One of the important factors that contributed to the high leverage of some hedge funds was the absence of adequate collateralisation arrangements (see Appendix 2). While exposures generated through exchange-traded derivatives are typically collateralised, there are no standard arrangements for exposures generally in over-the-counter markets. In a number of cases, financial institutions, under pressure to retain trading business, were prepared to set very high threshold exposures before any collateral was required. Moreover, in cases in which exposures were required to be collateralised, collateral was not required to cover the potential for future increases in exposures arising from changes in market prices.

• A set of sound practices should include clear policies that link collateral arrangements to explicit assessments of risks. They might also include procedures for holding collateral against potential credit risk.

Quite apart from the quality of banks' credit assessment, senior bank management's direct investments in hedge funds can create a potential conflict of interest that may lead to the overriding of normal credit-risk management processes. This suggests a need for supervisors to encourage strengthened corporate governance within banks, including improved monitoring and disclosure of management's investments and remuneration schemes.

3.3.2 Changes to capital requirements

One of the factors that contributed to weak credit assessment by regulated institutions is the distorted incentives that are created by some aspects of the Basle capital framework. In a number of areas, the relative capital requirements do not bear a close relationship to the relative risks. In particular, recent events have highlighted the following:

• the concessional capital treatment of banks' derivatives exposures to non-banks;
• the absence of a capital charge on short-dated foreign exchange contracts;
• the simplistic capital requirements on future potential exposures;
• the concessional capital treatment of repurchase agreements; and
• the treatment of on-balance sheet exposures to hedge funds.

Under current capital arrangements, if a bank incurs an exposure to a non-bank (for example, a hedge fund) by way of a direct loan, the capital requirement is 8 per cent of the exposure. In contrast, if the exposure is created through a derivatives transaction, the capital requirement is 4 per cent.^[3] Importantly, for foreign exchange contracts of less than 14-days maturity, there is no capital charge at all. Since these contracts account for the bulk of foreign exchange swaps, banks need to hold little, or no, capital against exposures to counterparties who are speculating in the foreign exchange market. This concessional capital treatment to short-dated foreign exchange transactions has contributed to very fine pricing on these contracts, and also reduced banks' incentive to limit exposures.

The capital concession to derivatives exposures to non-banks was originally granted on the grounds that only the best quality corporates had access to derivatives markets. This universal concession is no longer appropriate. As derivatives markets have expanded, both the range of participants and the systemic risks generated by these markets have increased. Similarly, the original concession on short-dated foreign exchange contracts was partly granted on the grounds that with a large well-diversified portfolio, the marginal contribution to overall risk from such contracts was relatively small. A review of both these concessions is required.

Another area for review is the capital requirements that apply to banks' future potential exposures. Currently, the capital charge on these exposures is determined by a simple formula (which includes the concessional risk weight), rather than by the approach used to calculate the capital charge for market risk on banks' trading portfolios (see Appendix 3). This latter approach would provide a more accurate measure of a bank's risk exposure which, if appropriately scaled, could allow a closer alignment of capital adequacy arrangements with desired incentives.

A related issue is the capital charges that apply to repurchase agreements. Under current arrangements, these agreements are treated as collateralised loans (and so the capital charge applies to the security, not the original exposure). This means that no capital needs to be held against a repurchase agreement involving a government security from an OECD country (provided that it is fully collateralised). This is despite the fact that a potential exposure exists if the counterparty defaults and market prices move adversely. Again, this concessional capital treatment has encouraged very fine pricing, and reduced the incentive for banks to limit their exposures.

Finally, the same capital charge (8 per cent) applies to a direct loan to a hedge fund, a direct investment in a hedge fund, and a direct loan to a high-quality industrial firm. This is despite large differences in the amount of risk being incurred. Consideration needs to be given to greater differentiation in the risk weights that are applied to various assets.

Any review of the current capital arrangements could be included within the Basle Committee's broader review of the Capital Accord, although this may well slow progress. An alternative approach is for a timely and focussed review of the capital arrangements that apply to derivatives exposures to non-banks and in the foreign exchange market.

3.3.3 Discretionary Supervisory Instruments

The third broad approach is for supervisors in individual countries to make more active use of their discretionary supervisory instruments.

If supervisors are of the opinion that systemic risk is increasing because regulated institutions are mispricing risk and permitting excessive leverage in financial markets, one possible response is to increase the minimum capital ratios that apply to institutions in the system (or the capital ratios of the institutions incurring the most risk). Alternatively, supervisors could use their legal authority to limit institutions' activities, for example by imposing limits on large exposures.

This approach has considerable merit, but faces practical difficulties. For supervisors to alter minimum capital ratios (or limits on large exposures) as the degree of systemic risk changes, they need to be able to measure systemic risk, and have a well-based understanding of which institutions are incurring the greatest risks. Even if supervisors have this information, they are likely to find it difficult to single out particular institutions for special treatment, except perhaps in the most egregious cases. They may also find it difficult to place significantly different requirements on internationally active banks to those placed on competitors in other countries. Not only would this distort the competitive landscape, but it would also encourage trading in a country's financial markets to shift offshore.

Moreover, this approach, even if it could be implemented, is unlikely to fully insulate a domestic financial system from turbulence in world financial markets. If LTCM had been allowed to fail, and Dr Greenspan's fears about the performance of financial markets had been realised, countries with sound banking systems would have felt a considerable impact, although perhaps somewhat less than countries with poorly regulated systems. Given the global nature of financial markets, a global response is needed.

The best approach is for individual country supervisors to ensure that their own banking systems comply with sound prudential standards, and for international agreement to be reached on disclosure arrangements and changes to the Basle capital arrangements. If such agreement is not possible, large countries in which hedge funds have major operations may be able to take the lead by unilaterally requiring enhanced disclosure and by changing the capital requirements that apply to banks in their own jurisdictions.

Footnote

If exposures are fully collateralised with high-quality assets, no capital charge is levied (see Appendix 3) [3]