4.2.2 Central bank policy perspective

While any of the various settlement mechanisms described above could potentially eliminate FX settlement exposures, each has particular strengths and weaknesses that should be considered. For instance, some arrangements might, if they are not well designed, increase certain risks while reducing others. Set out below are some of the major factors that should be considered at the design stage to ensure that a potential multi-currency settlement mechanism would achieve an appropriate balance.

Reduced FX settlement risk. Any sound multi-currency settlement mechanism - whether a guaranteed receipt or a guaranteed refund system, or a system that settled simultaneously or sequentially - could go directly to the heart of the problem and completely eliminate FX settlement exposures when settling individual trades. By removing the potential for settlement losses and the associated liquidity pressures, such arrangements would remove a major source of systemic risk that can arise in the settlement of FX trades.

The risk-reducing benefits of any multi-currency settlement mechanism could be magnified when combined with bilateral or multilateral obligation netting of the underlying FX trades. Legally valid obligation netting (whether arranged bilaterally between pairs of individual banks or multilaterally with the aid of an industry utility) would reduce the number and size of settlement flows and, hence, could reduce the intraday liquidity needs for settling these trades through a multi-currency settlement mechanism. Depending on the circumstances, however, problems in either the netting arrangement or the settlement mechanism could potentially impair the other. Accordingly, any guarantee underlying a payment/receipt relationship and the safety of any accompanying netting scheme must both be sufficiently strong to ensure that the combined arrangement does not create greater problems than it solves.

Reduced sources of systemic risk. Strong guaranteed receipt systems could have benefits beyond the potential elimination of FX settlement risk. With sufficient resources behind their guarantees, such systems would encourage banks to honour their settlement obligations even in the face of sudden concerns about their counterparties. Accordingly, they could be stabilising factors for money markets, especially at times of market stress. While no system's guarantee will prove fail-safe in every conceivable situation, the Lamfalussy framework could be used to assess the adequacy of the risk controls that stand behind a system's assurance of settlement.

Possible new sources of systemic risk. Despite its risk-reducing potential, a multi-currency settlement mechanism might also create a new source of systemic risk: a disruption in the settlement of one currency could disrupt the settlement of all other linked currencies. This concern could be acute in guaranteed refund systems. If, for instance, a system returned or cancelled conditional payments in the event of a default, the remaining system participants could face substantial, unexpected liquidity and replacement risks. These problems could be exacerbated if system participants were depending on their expected FX settlement receipts to extinguish their payment obligations under transactions in the FX, securities or other markets to counterparties within and outside the arrangement. Depending on the system's particular structure, the number of participants and the magnitude and pattern of the FX trades being settled, a failure to settle certain trades in a single currency might lead to major liquidity pressures in an unpredictable number of financial markets.

The possibility of not receiving the currencies they purchased on time could lead participants in guaranteed refund systems to hold back their payments at times of market stress, thereby increasing the total number of failed settlements. Even though the system could eliminate its participants' concerns regarding their FX settlement exposures, they would still face liquidity risks if their trades failed to settle. In such circumstances, participants might rationally choose to avoid delivering through the settlement system valuable resources (i.e. funds that they would use to settle their FX obligations) that might ultimately be returned to them if, instead, they could use these funds outside the system to avoid potential liquidity shortfalls. An individual bank would be likely to base its decision on the direct costs it might face (e.g. late-payment penalties, currency swap rates, late-day funding rates) without full consideration of the potential systemic impact of not paying its FX settlement obligations on time.

Possible liquidity pressures. Possible liquidity pressures could increase if, under any multi-currency mechanism, FX settlements were to shift to a less liquid time for the market. For instance, some potential simultaneous settlement systems might require final payment in central bank balances at different times than today. Such a settlement would be susceptible to liquidity problems if it occurred at a time when the market for immediately deliverable central bank balances in any one of the currencies being settled was not sufficiently deep and liquid. This, in turn, could lead to an unacceptable reliance on central bank credit and liquidity facilities, at least until the local market adapted to the new settlement pattern.

In addition, moving from a sequential settlement to a simultaneous one could shift risk and increase certain participant costs. For instance, settling a currency earlier or later would require a bank either to cover its settlement obligations sooner or to fund its anticipated settlement receipts for longer. Unless banks had available idle balances during the relevant time periods, these transactions (whether explicit or implicit) would shift exposures from FX counterparties to other entities (possibly including, if permitted, central banks) and could impose additional funding costs on the participants. Although these funding costs would not be considered a direct cost of participating in the system, depending on the circumstances they might nonetheless represent a significant ongoing indirect cost to a bank when using a simultaneous settlement system.

Although interbank settlements generally involve transfers of balances at central banks, settlement could also involve transfers of balances at another financial intermediary such as an existing or specially created private bank. For instance, participants in such an arrangement could establish accounts denominated in different currencies at a private bank and settle their mutual obligations by debiting and crediting these accounts. If liquidity at settlement time were limited to the system participants' existing balances at a private bank, several problems could arise. For instance, in some possible versions of these systems, participants might be forced to act as reciprocal providers of liquidity to one another, potentially creating undesirable liquidity interdependencies, a concentration of liquidity risk and large and unforeseeable credit exposures. Other versions of these systems could be designed to address these problems.

Impact on monetary policy. The establishment of a multi-currency settlement mechanism would be likely to have a measurable impact on the flows of payments through the national large-value payments systems of the currencies concerned and might also raise potentially significant monetary policy concerns for some central banks. This impact might depend, at least in part, on whether settlement would involve transfers of balances in accounts at central banks or at another financial intermediary such as a private bank. In particular, in countries where the settlement of FX transactions represents a large share of domestic payments, effective use of some arrangements involving a private bank might require a large and lengthy daily transfer of central bank balances to the bank's account. This could seriously limit the remaining amount of privately held central bank balances that could be used to support other domestic payments. Depending on the timing, size, distribution and predictability of settlements, this could have a major impact on domestic money markets and on the volatility of demand for central bank balances and credit and liquidity facilities. Accordingly, the developers of a multi-currency settlement mechanism, and the central banks of issue, will need to ensure that the mechanism would not create problems for the availability of intraday liquidity in the payments systems concerned, or reduce their ability to handle, in a timely and efficient manner, the remaining non-FX-related payment traffic in the respective currencies.

International interdependencies. By making the settlement of each currency directly dependent on the settlement of every other currency, multi-currency settlement mechanisms might both increase the risk of undesirable global payments gridlock and constrain the ability of central banks of issue to respond in a relatively independent manner to all home-currency settlement problems. These concerns would be greatest in a system providing a guaranteed refund with simultaneous settlement since the failure of one currency to settle could immediately trigger a series of unexpected settlement failures in all other currencies. Today, the possible failure of any home-currency system to settle on time would most likely be linked to a liquidity problem in that currency of one or more of the system's participants, even though a participant's problems could originate elsewhere. If deemed desirable, in such circumstances the central bank of issue might be in a position to mitigate these problems directly. In contrast, the operation of a multi-currency settlement mechanism would institutionalise currently informal interdependencies. Under such circumstances, if a possible home-currency settlement failure was related to a liquidity problem in another country and currency, the resolution of the underlying liquidity problem and, hence, of the potential home-currency settlement failure might depend more directly on the liquidity of money markets in other currencies and on the actions of other central banks.

Included currencies. To combine the benefits of a multi-currency settlement system with the benefits of obligation netting, a multi-currency settlement system would need to create a direct relationship between transfers in all of the currencies included in the netting agreement. In this light, a settlement mechanism that, for example, only linked transfers in two currencies (i.e. the basic PVP arrangement as seen by the market) would be incompatible with the use of broad obligation netting arrangements covering trades in many currencies. For instance, under an obligation netting arrangement covering trades in all of the major currencies, a bank might, from time to time, be either a net receiver or a net payer in any particular pair of currencies. In such circumstances, creating a relationship between the transfers of only that pair of currencies would offer little or no protection against FX settlement risk. Accordingly, a bank might choose either not to use the dual-currency settlement system, or to exclude its trades in the two currencies from the broader netting arrangement and settle them individually over the dual-currency system.

Access to services. It is unlikely that all FX market participants will have direct access to every settlement system. In addition, not all traded currencies are likely to be included in every arrangement. As a result, it may be impossible to settle all trades with all counterparties through a single settlement mechanism. This might leave a significant level of FX settlement exposure outside the system. It might also create competitive distortions among the different included and excluded currencies and FX market participants. To address these concerns, complementary action by individual banks to improve their practices for settling trades in all currencies with all counterparties may be needed.

Private sector versus public sector provision of services. The G-10 central banks share the view that multi-currency services would best be provided by the private sector rather than the public sector. One factor behind this view is that no inherent barriers to the private sector provision of such services have been discovered. Indeed, as discussed in Section 3, some risk-reducing private sector services are already available and others are under development or study. These existing and prospective multi-currency services include settlement mechanisms with guaranteed receipts, guaranteed refunds, simultaneous settlement, sequential settlement and different combinations of these features. Some services offer netting of obligations at the transaction stage, netting of obligations at the settlement stage, the inclusion of forward transactions in the system, and the use of a special clearing house bank or other private financial intermediary to facilitate netting or settlement. Overall, market demand for these services should increase as individual banks see the need to control their FX settlement exposures.

Another factor is that market-place competition in providing multi-currency services would bring important benefits. For instance, a privately operated multi-currency settlement system could draw on successful private sector methods for controlling risk. Private sector groups in some countries currently operate a variety of systems for transferring funds and securities with the aid of risk controls such as bilateral credit limits, multilateral credit limits and explicit liquidity-sharing and loss-allocation rules. This private sector approach to clarifying risks and giving participants the tools and incentives to control them should be a valuable addition to the operations of a multi-currency settlement system. Beyond this, the test of the market-place could promote competition and ongoing innovation and would make use of continual market pressure to provide cost-effective arrangements. This could be particularly useful in shaping an efficient array of complementary and competing services; indeed, it is not yet clear whether a market-wide reduction in risk would best be achieved with a small or a large number of service providers. Market participants themselves can make appropriate choices once the costs of alternative services are properly identified and allocated, and if consistent regulatory requirements are met in full by every service provider.

While G-10 central banks believe that multi-currency services would best be provided by the private sector, they also recognise that the successful creation of multi-currency settlement mechanisms would require cooperation between market participants and central banks, since all these parties need to be concerned with the safety and soundness, as well as the economic viability, of any multi-currency settlement system.

4.3 Action by central banks to induce rapid private sector progress

The G-10 central banks believe that private sector institutions can adequately address the systemic risk inherent in current practices for settling FX transactions. Indeed, some major banks are already concerned about the sizable FX settlement risks they face and are actively pursuing ways to improve their own settlement practices and to collectively develop risk-reducing multi-currency services. Nevertheless, despite their considerable capacity to reduce FX settlement risk through individual and collective action, many of these efforts are at an early stage of development, and timely progress across the market cannot be guaranteed. Among the impediments at the individual bank level is a belief held by some banks that the probability of an actual settlement loss is too low to justify the cost of reducing exposures. At the industry level, doubts may persist as to the optimal form and economic viability of risk-reducing multi-currency services.

Recognising the potential barriers to success, each central bank, in cooperation, where appropriate, with the relevant supervisory authorities, will choose the most effective steps to stimulate satisfactory private sector action over the next two years in its domestic market. In addition, where appropriate and feasible, central banks will make or seek to achieve certain key enhancements to national payments systems and will consider other steps to facilitate private sector risk reduction efforts. This reflects the likely need for public authorities to encourage action by individual banks and industry groups, and to cooperate with these groups, to bring about timely, market-wide progress.

4.3.1 Description of recommended action

Facilitate private sector action. Central banks will seek to facilitate progress by increasing private sector understanding of what banks can do individually and collectively to reduce FX settlement exposure. For instance, this report is being published in order to explain the nature of FX settlement exposure and to offer suggestions that individual banks and industry groups could adopt to address the associated risks. In particular, this report is designed to:

• Increase market awareness and understanding of FX settlement risk
• Offer a clear definition of and guidelines for measuring FX settlement exposures
• Describe how banks can control their FX settlement exposures by improving their individual settlement procedures and practices, as well as existing market-wide systems and arrangements (including practices and arrangements at correspondent banks)

Central banks also plan to work cooperatively with industry groups seeking to develop well-constructed multi-currency services (e.g. multi-currency settlement mechanisms and bilateral or multilateral multi-currency obligation netting arrangements) that would be widely available and would help banks control their FX settlement exposures on a routine basis. Where appropriate and feasible, central banks will cooperate with industry groups in one or more of the following ways:

• Attend industry working groups as observers
• Work with industry groups to extend the operating hours of domestic payments systems
• Work with industry groups to clarify and, where possible, to resolve legal issues and cross-border collateral issues
• Consider granting access to settlement accounts to sound multi-currency settlement mechanisms or to their members
• Consider granting access, on appropriate terms, to central bank credit and liquidity facilities to sound multi-currency settlement mechanisms or to their members

In several areas, such as the extension of payments system operating hours, these efforts are well under way. In addition, the Lamfalussy framework of minimum standards and principles for cooperative oversight has been, and continues to be, a useful starting-point for discussions between central banks and industry groups. In this connection, central banks will review the way in which the Lamfalussy framework might apply to multi-currency settlement mechanisms that are not clearly designed as netting schemes.

Finally, central banks plan to facilitate private sector action by making or seeking the following key enhancements to national payments systems:

• Clarification of the times at which payment instructions become irrevocable and receipts become final in the settlement of FX transactions via home-currency payments systems or book-entry transfers on the accounts of correspondent banks
• Provision, where not now available in at least one large-value payments system, of an intraday final transfer capability or its equivalent
• Removal of obstacles (e.g. early cut-off times for third-party transfers) that inhibit payments system direct members from acting upon late-day customer payment instructions for same-day value
• Strengthening, as necessary, of the risk management arrangements of privately operated systems used to settle FX transactions

These enhancements could make it easier for individual banks to overcome at least some of the obstacles they might face when trying to measure and control their FX settlement exposures. Furthermore, to the extent that multi-currency settlement mechanisms would need to transfer funds over domestic payments systems, these enhancements might also make it easier for them to provide safe and effective risk controls. Indeed, many of the suggested key enhancements to national payments systems are already under way within the G-10 and EU community in response to domestic or regional policy objectives. Two notable examples are the development of domestic real-time gross settlement payments systems (or their equivalent) and the application of the Lamfalussy minimum standards to domestic netting schemes.

Domestic strategies. In addition to these measures, each central bank, in cooperation, where appropriate, with the relevant supervisory authorities, will choose the most effective overall strategy to stimulate satisfactory private sector action over the next two years in its domestic market. To different degrees in different countries, central banks may act as monetary authority, credit and liquidity provider, supervisor, regulator, payment and settlement service provider, payments system overseer and/or overseer of the financial system. In the light of these divergent roles, as well as other local circumstances, individual central banks may select different strategies to stimulate domestic action. Nevertheless, such strategies will be likely to include one or more of the following elements:

• Publicising this report and its recommendation that banks take immediate steps to apply an appropriate credit control process to their FX settlement exposures
• Using moral suasion to encourage banks to adopt this recommendation
• Seeking to reinforce this recommendation with supervisory measures
• Promoting this recommendation among relevant non-bank financial institutions

Individual central banks might publicise this report in their domestic markets with the intention of focusing, in particular, the attention of the senior management of banks on FX settlement exposure. This effort could encourage banks to consider, at top management level, the report's analysis of the nature of the problem and its possible solutions.

Some individual central banks might also use moral suasion to persuade domestic banks to follow the report's recommendation that banks take immediate steps to apply an appropriate credit control process to their FX settlement exposures. Depending on local market circumstances, central banks might choose to use moral suasion either directly or through industry groups (i.e. through peer pressure, codes of conduct, etc.).

Central banks or other public authorities might, after proper consultation, also choose to use supervisory measures to persuade individual banks to control their FX settlement exposures. If appropriate and feasible, one or more of the following measures could be taken:

• Supervisory guidelines for measuring FX settlement exposures in a manner consistent with the proposed methodology
• Regular confidential reporting of properly measured FX settlement exposures
• Regular public disclosure of properly measured FX settlement exposures
• Supervisory guidelines regarding the prudential management and control of properly measured FX settlement exposures
• Verification of compliance with the selected measures through bank examination and audit reports

If necessary, one or more of the following stronger supervisory measures might also be considered:

• The enforcement, by statute where available, of the use by individual banks of mechanisms to control their properly measured FX settlement exposures. This could include the setting of formal limits on those exposures
• Consideration, by agreement with banking regulators (G-10 and EU), of FX settlement risk in the set of risks subject to capital adequacy requirements
• The enforcement or imposition (by agreement with the relevant supervisors) of comparable measures applying to non-bank regulated financial institutions active in the FX market

4.3.2 Central bank policy perspective

Private sector inducements. It is recognised that, at present, some trading banks might conclude that there is no business case for their taking further steps to control their FX settlement exposures. Accordingly, the private sector needs, and should be given, encouragement and support to improve settlement practices in individual institutions and on a market-wide basis.

Inducements for individual banks. At the level of individual banks, inducements to action would best be pursued through the home central bank or, depending on national circumstances, through the appropriate supervisory authorities; the most effective combination of publicity, moral suasion and supervisory measures will be chosen to induce domestic banks and, where relevant and appropriate, non-bank financial institutions to improve their settlement practices within the next two years.

Many individual central banks might use publicity or moral suasion as part of a comprehensive strategy to reduce FX settlement risk. Indeed, in some markets either of these efforts might be sufficient to stimulate significant action by individual banks to improve their settlement practices. In some countries, however, publicity alone might not be sufficient to prompt private sector action. And, while banks in some countries might respond quickly if publicity was accompanied by central bank moral suasion, this might not be true in all markets. At the same time, policy or statutes might prevent some central banks from using moral suasion to call for improved practices by individual banks.

In some countries, central banks might find supervisory measures to be an effective market-wide strategy for bringing about desired improvements at the individual bank level. And, either through their direct supervisory role or through their relationship with other domestic authorities, some central banks might be able to initiate certain supervisory measures rather quickly. However, while supervisory measures might be useful in some markets, in others they might not be appropriate or feasible. Reaching international agreement on the amendment of existing regulations or the creation of new ones could be particularly time-consuming.

Some central banks might promote risk-reducing measures among relevant non-bank financial institutions so as to contain the systemic risk in FX settlement practices found outside the domestic banking sector. Some central banks might be able to influence these institutions directly with publicity and moral suasion, if not supervisory measures; other central banks would need to work through other public authorities.

Inducements for industry groups. At the industry group level, central banks plan to cooperate, where appropriate and feasible, with those existing and prospective private sector groups that would like to provide risk-reducing multi-currency services. Central banks are ready to work with the developers of such services to ensure that they satisfactorily address risk management and other key issues and meet, or, as necessary, surpass, applicable minimum standards or criteria. In this connection, the central banks concerned can oversee such service developments using an approach consistent with the Lamfalussy framework. When cooperating with industry groups, however, central banks must retain their ability to make their own judgements about the safety and soundness of specific services and practices.

When working with industry groups to bring about the identified key enhancements to national payments systems, central banks would need to guard against possible side-effects. For instance, removing obstacles to late-day domestic payments could be an important step in enhancing the efficiency of the FX settlement process, but in some markets this could have an adverse impact on the liquidity of individual banks, of correspondent banks, or of the money market as a whole if all payers sought to execute their payments as late in the day as possible. In such countries, central banks might need to accompany this change with an indication that "best practice" in their domestic markets should ensure the smooth flow of payments throughout the business day.

Monitoring progress. The G-10 central banks believe that the private sector can adequately address the systemic risk inherent in current practices for settling FX transactions. Some major banks, recognising the large size of their FX settlement exposures, are already taking steps to improve their procedures for settling FX trades, and industry groups are working to develop risk-reducing multi-currency services. However, there is no guarantee that the identified i