Abstract
This paper focuses on a specific set of issues -- qualification criteria, monitoring structure, access levels and terms, links to other Fund facilities, and links to the private sector. This paper is a complement and follow-up to an earlier paper, which itself considered the broader issues involved in creating a new instrument. The paper does not put forward a concrete proposal to Directors, but seeks to achieve a further convergence of views on these key issues.
I. Introduction 1
1. In August 2006, the Executive Board discussed in a seminar a possible new liquidity instrument for market access countries. Consideration of such an instrument was called for in the Medium-Term Strategy to strengthen the Fund’s support for members’ efforts to reduce the risk of capital account crises. The new instrument would aim to benefit members and the financial system as a whole by reinforcing strong policies, reducing the likelihood of crises, and by offering an alternative to costly self-insurance (Box 1). It would respond to the view of emerging market members, and others, who believe that the expiration of the Contingent Credit Lines (CCL) in late 2003 left a void in the Fund’s crisis prevention framework. Annex I provides references to relevant documents.
RAL and Risk Pooling
Many countries have chosen to insure themselves against future shocks by building liquidity buffers through the accumulation of international reserves, although in some cases the decision to amass reserves also reflects other policy objectives. Large stocks of reserves are expensive to maintain, but these costs can be mitigated through the pooling of reserves among countries. Some countries have decided to complement their participation in the Fund with regional pooling arrangements. Regional arrangements may also reflect an interest in enhanced regional integration and political cooperation, and members’ dissatisfaction with the Fund’s past lending decisions, the perceived burden of conditionality, and uncertainty about the future availability of Fund resources.1 The RAL could offer an alternative to costly self-insurance, and would allow countries to capture efficiencies through global risk sharing, while providing the benefits of crisis prevention.
1 See “Country Insurance: The Role of Domestic Policies,” (June 19, 2006).2. Directors recognized the potential benefits of such an instrument and the design challenges involved. Most recognized that a well-designed liquidity instrument could provide benefits by supporting members’ crisis prevention efforts, although some questioned whether the risks could be adequately controlled. They also recognized the design challenges inherent in seeking to provide effective support to borrowing members while ensuring safeguards to the Fund and minimizing moral hazard. Finally, they saw the staff paper’s Reserve Augmentation Line (RAL), which includes both strong qualification and ongoing monitoring features, as a useful starting point for further consideration.
3. The present paper addresses design issues raised in the August seminar. It
focuses on a specific set of issues—qualification criteria, monitoring structure, access levels and terms, links to other Fund facilities, and links to the private sector. This paper is a complement and follow-up to the August paper, which itself considered the broader issues involved in creating a new instrument. The paper does not put forward a concrete proposal to Directors, but seeks to achieve a further convergence of views on these key issues. It would be important for a new instrument to have broad support across the membership, recognizing that its design inevitably will involve compromises and tradeoffs.
II. Outreach
4. Fund Management and staff have discussed a new liquidity instrument with official sector representatives and market participants in a number of settings over the past months. These include a seminar in Singapore in July with members from various regions; a regional seminar in August in Korea with members from Asia; seminars in December with members from North and South America (in Chile), and Africa, Europe and the Middle East (in Italy). Fund staff from the Monetary and Capital Markets Department and the Policy and Development Review Department have also held meetings with financial market participants in New York and London.
5. Many official sector representatives saw merit in a well-designed instrument that would strengthen crisis prevention, while ensuring adequate safeguards of Fund resources. The following are the main points that were expressed:
Objectives: Many, but not all, country officials saw merit in a new Fund instrument aimed at preventing rather than containing crises. Many considered the RAL proposal a valuable step forward and believed views were converging. They noted that a RAL should complement but not substitute for countries’ own efforts to prevent crises. Country officials underscored that the design of the instrument would need to provide for the immediate drawing of a meaningful amount of resources. Some officials emphasized the need for strong safeguards to the Fund and to minimize moral hazard risks.
Demand: Though the current favorable global market environment may dampen demand for the instrument, most considered this a good time to discuss a facility aimed at prevention. Emerging market and other member countries saw considerable merit in having such an instrument in the Fund’s toolkit. Some were concerned that if a new instrument were never used, it would have reputational costs to the Fund.
Terms: Country officials considered that the new facility would need to be differentiated from existing Fund instruments. Emerging market countries called for a longer duration and less frequent reviews to reduce uncertainties regarding the availability of resources. A number thought that access at 300 percent of quota was on the low side, and suggested that the CCL’s 300-500 percent range be considered. Some expressed concern that the standard commitment fee might reduce demand for the facility. Others emphasized the need for a strong monitoring framework so that if a drawing were made it would support an appropriate policy response.
6. Financial market participants had mixed views about a new instrument.2 The participants made the following points.
Objectives: While many were neutral or positive, several doubted the instrument was needed given the changes in markets and stronger country policies. Many emphasized the RAL’s commitment and signaling framework as its most important element. Many also believed the financing to be important, but thought that 300 percent of quota, while giving a strong signal that the Fund was putting its financing on the line, was low in respect to countries’ reserves or rollover needs. A few worried that a new liquidity instrument could, in a period of already abundant liquidity and narrow spreads, lead to undue further compression of spreads and overborrowing. Some considered the RAL particularly useful for newly emerging markets and doubted the signaling benefit for countries that have established a strong reputation. Most considered that a request for a RAL where a crisis was imminent would send a negative signal.
Terms: Market representatives stressed that automaticity would be critical if the instrument were to provide a credible signal. While many countries could increase reserves by further borrowing, or secure a bank credit line, private sector representatives were not aware of other instruments that provided contingent financing like the RAL. On pricing, there was a mix of views on the current commitment fee structure, which some saw as reasonable and in the range of pricing of credit lines, while others argued that fees may need to be lower to make the instrument attractive.
III. The Reserve Augmentation Line
A. The RAL Framework
7. The key features of the RAL framework (as set out in the August paper) are described in Box 2. The main differences from the CCL are also noted. The RAL combines elements from models for contingent financing that rely on up-front qualification standards as the primary safeguard and models that focus on monitoring of ongoing performance. It recognizes that well-defined qualification criteria are important to ensure selective use of the instrument. At the same time, the framework would include ongoing monitoring to help further safeguard Fund resources and thus mitigate the need to design overly restrictive qualification criteria, which could unduly limit the set of members that could make use of the instrument.
Key Features of the RAL (as set out in the August paper)
Objectives. The RAL would be intended to benefit members and the financial system as a whole by reinforcing strong policies and reducing the likelihood of capital account crises. The new instrument would also allow members to reduce their reliance on costly self-insurance. The RAL’s key features include:
Qualification. RAL would be available to members with strong macroeconomic policies, a sustainable debt, transparent reporting, and proven credibility in policy implementation that are still faced with, but making progress to reduce, vulnerabilities to capital account crises (see Box 4). Frontloaded access under the RAL would be justified by the strength of the member’s existing position and the credibility of its forward looking commitments. Qualifying members would be expected to be in a position where a significant macroeconomic policy adjustment is not needed. Qualification would be assessed at the request of the member (i.e., no automatic assessments).
The Four Qualification Criteria under the RAL
1—No need. At the time of approval, the member is not expected to need to use Fund resources. This criterion is intended to ensure that the RAL is not used where the financing need is immediate, and other Fund facilities would be more appropriate. A fortiori, members using other exceptional balance of payments financing would not qualify.
2—Good policies. The member has pursued and remains committed to strong macroeconomic management and to policies directed at reducing remaining vulnerabilities, including as they relate to balance sheets and the financial sector, giving confidence that the member will react appropriately in the event of crisis. These policies are described in a forward-looking economic and financial program prepared by the member that would include a quantified framework for the period covered by the RAL.
3—Sustainable debt. The third criterion requires that a rigorous debt sustainability analysis indicates that there is a high probability that the debt will remain sustainable. Although this criterion is linked intrinsically to the previous criterion, especially for those countries for which debt is a key vulnerability, it is intended to ensure that a rigorous forward-looking debt sustainability analysis—for public and external debt—is conducted in all cases. The assessment would cover both the evolution of the level of debt and the rollover and financing requirements under various scenarios and stress tests, and would need to provide reasonable assurances that the debt will remain sustainable even in the event of a crisis of the sort that could lead to a drawing under the RAL.
4—Transparency. The fourth criterion requires that the member has demonstrated a commitment to transparent reporting of economic data. The member should have subscribed to the Special Data Dissemination Standard and be judged to be making satisfactory progress toward meeting its requirements. The member should also be working to meet the standards set by the codes of transparency in the areas of fiscal and monetary and financial policies. Finally, members requesting the RAL should also have indicated their intention to publish promptly all related reports, along with their forward-looking economic and financial program, including a quantified framework for the period covered.
Monitoring. The member would present a policy framework, including quarterly quantitative indicators, that meets the standards required for drawings in the upper credit tranches. There would be no performance criteria except for the standard continuous performance criteria on exchange and trade restrictions. The program would focus on policies to maintain macroeconomic stability and reduce capital account vulnerabilities. Performance would be monitored by staff, with a formal semi-annual Board review whose related documents would be published. Qualification would be interrupted after a fixed date if this review could not be completed.
Financing. Financing would be 300 percent of quota with the full amount available at the outset if a capital account crisis emerged and the program was on track. As an exception, financing could be lower for members exiting from large use of Fund resources; in such cases, the Fund should be open to members’ requests to allow the size of financing to decline over time when the member wishes to avoid a sharp drop in financing upon exit.
Terms. The RAL would be established as part of the Supplemental Reserve Facility (SRF), as it is intended to address the same type of special balance of payments need. Resources committed under the RAL would thus have the same charges and maturities structure as the SRF (including the feature that credit tranche terms would apply to the portion of access within the credit tranche access limits and SRF terms on access beyond this). The arrangement would be one year in duration and a series of several RALs would not be discouraged.
Comparison with the CCL. The RAL would provide an immediate drawing of 300 percent of quota for countries that have met program monitoring conditions. Under a CCL a limited “activation” review by the Executive Board was required before drawing and the CCL involved smaller upfront drawings—normally one-third of access in the expected range of 300-500 percent of quota. The use of the CCL was limited to a narrow set of members with very strong fundamentals, while the RAL would be intended for countries with strong fundamentals but some remaining vulnerability. The RAL would not include the CCL’s requirement that drawings could only be made in a contagion-related crisis.
B. Qualification
8. The qualification criteria should provide for selective use of the RAL. Carefully designed qualification criteria are important for a transparent and disciplined assessment. They are also essential to ensure that the member’s current macroeconomic position is strong and that its commitment to reducing remaining vulnerabilities is credible, and this would justify front-loaded access and help limit the risk of moral hazard (Box 3). The criteria would need to be defined so that the instrument is useful to members with strong policy frameworks but with some remaining vulnerability to capital account crisis.
Moral Hazard
Concerns. While it has long been recognized that the prospect of Fund lending may distort incentives—leading members to adopt excessively risky policies (debtor moral hazard) and private creditors to underprice lending risks (creditor moral hazard)—such concerns came to the fore in the context of capital account crises and exceptional access Fund arrangements. Moral hazard may be especially relevant in cases where the systemic and political importance of the member means that it is perceived as being “too big to fail.”
Controlling the risk of moral hazard. Experience has shown that Fund financing will not always be forthcoming, and—even in large, and politically sensitive cases—the Fund has exercised selectivity and has interrupted its support when it believed program success was unlikely. Creditors have sustained important losses in some crises, encouraging careful risk assessments. Furthermore, the enormous cost of a crisis to the member, even when the Fund provides financing, is a strong deterrent to imprudent policies. The Fund’s surveillance activities are also intended to ensure that members pursue sound policies.
Moral hazard and the RAL. In the context of the RAL, selectivity is important, including through the inclusion of substantive qualification criteria, to ensure that the instrument would be available only to members that have been pursuing sound policies aimed at reducing remaining vulnerability, and remain committed to them. In comparison to traditional Fund lending, which provides financing conditional on the member adopting corrective measures after a crisis, the RAL’s ex ante selectivity serves to encourage sound policies before a crisis—and hence to prevent crises from occurring (and helping to limit moral hazard). This underscores the importance of the RAL’s qualification criteria, and of the Fund ending its contingent support at the time of mid-year reviews if policies deteriorate or fail to respond appropriately to changing circumstances. Under the RAL, creditors and debtors would have assurance that a pre-specified amount of financing would be available for six months, if needed, as they do for shorter periods in a precautionary stand-by arrangement.
Further work on moral hazard. A survey of the topic and updates empirical tests for creditor moral hazard is provided in “Fund Financial Support and Moral Hazard: Analytics and Empirics”.
9. In the August seminar, Directors noted the benefits of objective standards for qualification together with discretion to reflect the diversity of country situations. There was little support for automatic assessment (before a member has expressed interest in a RAL). Quantitative criteria and benchmarks strengthen the transparency, predictability, and even-handedness of the decision-making process, reducing the scope for political influence. At the same time, it would be extremely difficult to define specific thresholds to test performance in light of the wide disparities among member’s circumstances and across time, the complexity of policy reaction functions, and the difficulty of capturing sensitivities to changes in underlying global conditions (e.g., global interest rates, commodity prices). The approach suggested below is based on consideration of a member’s performance against a standard set of quantitative indicators, showing comparators for emerging markets and all market access countries.
10. Directors considered the four qualification criteria outlined in the August report a good starting point but asked that they be further specified (Box 4). In the August discussion, Directors stressed that qualification should give strong weight to the member’s current position and the credibility of the member’s commitment to implementing strong policies to strengthen the country’s resilience to shocks. The objective of the qualification criteria should be to help form a careful assessment of both (a) the degree of robustness or vulnerability to capital account shocks; and (b) the quality of policies in reducing vulnerability. This would allow for a transparent recommendation and approval of a member’s request for a RAL that takes into account both factors. For example, a country with low vulnerability and strong policies would qualify for the RAL, while a country with low vulnerabilities but poor policies would not be expected to qualify. Conversely, a country with some vulnerabilities but that is pursuing strong policies to address them could qualify.
11. Elaboration of the second criterion is needed to guide decisions. To qualify for the RAL, a member should be in a position where a significant macroeconomic policy adjustment is not required and the structural reform agenda to reduce vulnerabilities would be well under way. The staffs assessment, which would draw on the Fund’s surveillance, including the last Article IV consultation, would consider the member’s vulnerabilities, the strength of its policies and policy reactions, and the credibility of its commitment to continued strong policies. The staffs assessment would consist of the following elements:
An assessment of the macroeconomic situation and vulnerabilities. The assessment would be analogous to the staffs vulnerability exercise. It would comment on the member’s performance with reference to key quantitative indicators, showing comparators for emerging markets and all market access countries. The staff assessment would also consider the prevailing market sentiment vis-à-vis the member, as reflected in, for example, bond spreads and credit default swaps. While each country’s position would be evaluated based on its particular structures, indicators could systematically include:
➢ Real economy: real GDP growth, inflation;
➢ Public sector: overall fiscal balance, gross public debt, short-term public debt, and foreign-currency linked debt, all as a proportion of GDP;
➢ Financial sector: non-performing loans, foreign currency loans, capital-adequacy ratios and credit-to-GDP ratios. The strength of the banking system and bank supervision would also be assessed, and would take into account adherence to the Basle Core Principles for Banking Supervision and participation in the FSAP;
➢ Corporate sector: Indicators of corporate sector balance sheet vulnerabilities should be included to the extent that they are available; and
➢ External sector: reserve coverage of short-term debt, gross external debt, gross financing requirements, current account deficit, and indicators of competitiveness including the real exchange rate. The evaluation would take into account the member’s choice of exchange rate regime.
An assessment of policies and policy reactions. To reinforce the rigor and transparency of the judgments involved in this assessment, the staff report recommending the RAL could frame its qualitative assessment of the member’s policies around the following three policy aspects:
➢ First, the country should be assessed as having responded appropriately to recent (last 2-3 years) adverse shocks, such as the turbulence in emerging markets in the spring of 2006;
➢ Second, the country’s macroeconomic and structural policies should be considered as credible to achieving a reduction in key vulnerabilities; and
➢ Third, looking ahead, the country should be considered to have adequate institutional and political capacity to implement strong policies.
12. A member in any of the following economic circumstances would not be expected to qualify for the RAL (though this list is not exhaustive):
The debt is unsustainable or the member has a need for exceptional financing;
The exchange rate is pegged and considered significantly overvalued;
The member has a need for substantial macroeconomic policy adjustment; and
The member has a substantial unfinished reform agenda to address vulnerabilities.
This framework acknowledges that, ultimately, an assessment of vulnerabilities and economic policies will require judgments taking into account the circumstances of each member—as is the case for all Fund lending decisions. The procedures for early Board involvement (see below) would ensure that the Board is informed about the qualification assessment at an early stage.
Link to the Fund’s Financial Transactions Plan
13. Approval of a RAL arrangement should not preclude a member’s participation in the Financial Transactions Plan (FTP), or vice versa.3 Selection of members for the FTP is based on a judgment about of the strength of members’ balance of payments and reserves positions. Under current practice, a member is excluded from consideration for the FTP if it has a drawing arrangement, whereas a precautionary arrangement in the credit tranches is not grounds for exclusion. In line with the treatment of precautionary arrangements, an arrangement under the RAL would also not be grounds for exclusion from the FTP; indeed, the RAL may well be a suitable instrument for some current FTP participants.
Timing of request for a RAL
14. A member could request a RAL at any time that it meets the qualification criteria. Ideally, if the member is interested, the qualification assessments and RAL-related travel could be conducted at the time of the Article IV consultation. This would help to keep the member’s exploration of a RAL low key. Another option would be to conduct the semiannual RAL review at the time of the Article IV consultation. However, there would be no requirement to link RAL requests or reviews to Article IV consultations.
Link to the Fund’s Exceptional Access Framework
15. A request for a RAL would be subject to the requirements under the exceptional access framework.4
Exceptional access criteria. Any request for a RAL would need to be justified in light of the four exceptional access criteria (Box 5). It should be noted that these criteria were written assuming the member is already in the throes of a capital account crisis—experiencing both an existing balance of payments need and loss of market access. The RAL is designed explicitly for members that are not experiencing balance of payments pressure at the time of the request and have access to private markets. Thus, as for any case where a member is not in a capital account crisis, the first and the third criteria would not be relevant for a member qualifying for the RAL.
The Four Criteria in the Exceptional Access Framework
The exceptional access framework includes four substantive criteria that a member in a capital account crisis would need to meet. For a member that is not in the midst of a capital account crisis, a request for exceptional access would still need to be “justified in light” of the four criteria. The four criteria are:
Criterion 1—The member is experiencing exceptional balance of payments pressures on the capital account resulting in a need for Fund financing that cannot be met within the normal limits;
Criterion 2—A rigorous and systematic analysis indicates that there is a high probability that the debt will remain sustainable;
Criterion 3—The member has good prospects of regaining access to private markets within the time Fund resources would be outstanding; and
Criterion 4—The policy program of the member provides a reasonably strong prospect of success, including not only the member’s adjustment plans but also its institutional and political capacity to deliver that adjustment.
Procedures and reporting requirements. Any request for an arrangement under the RAL would follow the procedures, including for early Board involvement and increased reporting, set out in the exceptional access framework.
C. Length of Arrangement
16. The RAL framework defined in the August report focused on a one-year renewable arrangement. Limiting the length of an arrangement under the RAL to one year would contain the risk of a material change in policies or circumstances during the arrangement and could provide for a more graceful exit. A longer period of coverage could be achieved by having a series of successor RAL arrangements, so long as the member continues to meet the qualification criteria. Some Directors, however, as well as participants in the outreach seminars, felt that the option of a longer RAL arrangement, for example two to three years, should be left open. A longer initial arrangement might increase interest in the
RAL, especially when the risk of a crisis seems further off, and might reduce the cost of repeated entry. Although staff would not favor a very long arrangement given the greater uncertainties involved, some flexibility, perhaps up to 18 months or two years may be considered. Such a modality for RAL resources would be different from that currently governing the SRF, where financing can be committed only for a period of up to one year. The difference could be justified by the precautionary nature of the RAL (i.e., unlike the SRF, members would not have an actual short-term balance of payments need at the time of request for a RAL).
17. Another option would be to allow for the one-year arrangement under the RAL to be extended at the time of the mid-year review so as to cover twelve months ahead.
Under this structure, the Board would initially approve an arrangement under the RAL covering a period of twelve months. At the time of the semi-annual review, the member could request that the RAL be extended for six months so that the arrangement would continue to cover twelve months ahead (with another semi-annual review scheduled six months ahead). Under this option, the future period over which Fund financing could be available would be lengthened to a minimum of six months, giving greater confidence to the country and market participants. A request for an extension would need to be accompanied with an updated policy document with quarterly targets for the coming twelve months. There would need to be a maximum limit for a RAL arrangement of three years, after which a new RAL would need to be requested. This limit is intended to ensure a fresh look at the member’s policies and vulnerabilities.
D. Monitoring Structure
18. The August paper discussed a monitoring structure consisting of semi-annual reviews based on the authorities’ forward-looking economic program. The reviews would focus on verifying that macroeconomic policies remain appropriate and that the member is making progress in reducing vulnerabilities. The member would maintain the right to draw immediately so long as reviews were completed by the scheduled date. There would be a presumption that a review would be completed where there are no material deviations from the member’s policy framework and a forward-looking assessment indicates that policies remain supportive of the underlying objectives of macroeconomic stability and reducing vulnerabilities to a capital account crisis.
19. The unique structure of the monitoring framework in the RAL is designed to reflect the qualifying members’ strong track record of performance. The requirements for completing the review would need to be sufficiently clear to provide assurances to the member of the conditions under which drawings could be made, and when the Fund would not be able to complete the reviews. The member’s quantitative targets would help reinforce objectivity, and the members’ policy document would need to describe explicitly the issues to be considered during the review. The economic program could also include monitorable structural milestones to the extent these are needed for addressing key vulnerabilities.
20. One advantage of the review-based monitoring structure is that it avoids the “black-out” periods for access inherent in stand-by arrangements. Black-out periods result from structures where the availability of financing is suspended for the time necessary for data to become available to confirm observance of the performance criteria for the most recent test date. For example, under Brazil’s (2003) extension of its stand-by arrangement that it treated as precautionary, the availability of financing was lost every quarter for about five weeks, until data became available to confirm observance of the targets for the most recent quarterly test date (Figure 1). Such black-out periods are especially awkward for a liquidity instrument aiming at capital account crisis prevention. Under the review-based monitoring structure proposed for the RAL, the member would have continuous access to Fund financing up until the date at which the semi-annual review is scheduled to be completed. If the review is completed on time, then there would be effectively no interruption in access over the period of the arrangement. The review should be scheduled for a time when data is expected to be available on the outcome of the mid-year (quarterly) indicators.
Black-out periods under RAL and SBA
Citation: Policy Papers 2007, 004; 10.5089/9781498333788.007.A001
Alternative monitoring structures
21. Alternative monitoring structures could be considered. Alternatives could be based on (i) different time frames for reviews, (ii) incorporation of an activation review at the time of a request for a purchase (as under the CCL), (iii) monitoring against performance criteria, or (iv) a structure for ad hoc review, which could trigger termination of access under certain conditions.
Different frequency of reviews. Periodic reviews should provide for monitoring that is sufficiently frequent to meet safeguard needs, without introducing too high a degree of uncertainty about the availability of financing or creating possible negative exit signals. Extensive gaps between reviews could introduce excess risks to Fund resources.
Activation review. A structure where the request for purchase needs to be approved by the Board, such as the activation review used under the CCL, would ensure a discussion of the appropriate policy response to the shock and provide additional safeguards to the Fund. The uncertainty created by an activation review would reduce the perception that financing was readily available, however, and this could make the RAL less useful for liquidity support or as a signal.
Performance criteria. As set out in the August paper, the RAL did not include performance criteria other than the standard continuous criteria related to exchange and trade issues. As an alternative, however, consideration could be given to having the RAL include performance criteria (quarterly or semi-annually) against which the member’s performance would be assessed. Performance criteria would in some respects perform a similar function to the quantified indicators in the RAL. An important difference is that a failure to meet a performance criterion would block drawings and would require a waiver from the Executive Board to restore drawing rights. In contrast, under the RAL design, a deviation from the quantified indicators would only be taken into account in the overall context of the semi-annual review. A structure with performance criteria would be less differentiated from precautionary stand-by arrangements in the credit tranches. It would also introduce problems of black-out periods discussed above.
Ad hoc review triggered by material deviations. A possible alternative structure would be to provide for ad hoc reviews, triggered by material deviations from the member’s program. Such a structure could serve as a complement to semi-annual (or less frequent) reviews. This structure would allow for the suspension of the RAL in case of a flagrant departure from the policy framework. This structure would strengthen protection for the Fund in the face of large, unexpected, and rapid changes in the member’s circumstances. Its value would depend on the possible frequency of such shocks. The cost of ad hoc reviews would be in increasing the uncertainty for the member about whether financing would be available in a crisis. It would also pose difficulties in terms of articulation of what would constitute a “flagrant departure” in a way that is robust across members’ varying circumstances and over time, and a decision-making framework would need to be designed.
22. On balance, staff tends to favor the original RAL structure with automatic access after approval and a monitoring structure based on semi-annual reviews. This structure finds a reasonable balance, in the staffs view, between strong safeguards and confidence that resources would be available.
Post-drawing review
23. A drawing under the RAL would prompt an immediate post-drawing review.
This review would consider the circumstances of the drawing, and the authorities’ policies going forward. While the review could in principle also consider a request from the member for further financial support, what is essential is that the review take place shortly after the drawing. Since the request for further assistance might require more time (to assess the shock and available financing, and to design the policy adjustment), it would likely be the case that any such request would be better handled in a separate, subsequent Board discussion.
24. Any additional financing would be provided under the SRF (most likely) or a stand-by arrangement in the credit tranches, with phased drawings and a conditionality structure similar to that under existing arrangements. At that stage, the right combination of financing and policy adjustment to the shock would need to be considered, taking into account also financing from other official and private sources.
E. Structure: Access, Terms, and Other Issues
Access
25. A key feature of the RAL design is the provision of large access up front. Such frontloading is important—given the nature of capital-account-driven balance of payments pressures—to reinforce confidence that resources would be available on a timely basis.
26. Determining the right level of access is challenging, however. While noting the difficulty of assessing the scale of potential need in the context of capital account vulnerabilities, staff had proposed a uniform level of access of 300 percent of quota.5 The August paper provided an extensive discussion of the basis for the proposed 300 percent of quota. Another option that was favored by some Directors—and also by some participants in the outreach seminars—is to provide for a range of 300 to 500 percent. This structure would provide more flexibility and would be consistent with traditional Fund practices, linking access to the specific (and differing) circumstances of members. At the same time, given the uncertainties involved in establishing potential balance of payments need for market access countries, justifying a particular level of access within this range would not be straightforward, and it is unclear what signals different access levels would send. For these reasons, staff would favor a uniform level of access at 300 percent of quota. Such a structure would be consistent with the access norm under the CCL (albeit in the case of the RAL, access would be significantly more frontloaded).
27. In the event of a crisis, after the first purchase has been drawn, any further Fund financing would be provided in the context of a new specified policy program.
Staff sees two broad options for how access could be structured:
Provide for a second purchase under the RAL. One option would be for the commitment of resources under a RAL arrangement to include a second purchase, perhaps 100-200 percent of quota. The first purchase (of 300 percent of quota) would still be available automatically so long as the program is on track and the member is hit by a capital account shock of the type for which the RAL resources are intended. The first purchase would then prompt a Board review to consider the appropriateness of the policy response to the shock, set conditionality, and decide on the release of the second purchase, and the level and phasing of any subsequent purchases. One advantage of this structure is that it would identify a larger amount of access from the start. The prospect of additional financing could encourage more continued dialogue with the Fund during the crisis. However, the availability of 100-200 percent of quota may not be a decisive factor in countries’ decisions about such engagement.
Provide additional purchases under a new arrangement: As set out in August, another option would be for the RAL to only include the first purchase, and any further financing would be requested for a new program supported by an SRF (or by a stand-by arrangement in the credit tranches). The advantage of this option is that it keeps the RAL design simple, and focused on crisis prevention. Also, the magnitude of the additional balance of payments need, if any, would be more apparent after a crisis hits.
Fund liquidity and a global cap on commitments
28. A global cap on commitments under the RAL could be considered as a guideline for a review rather than a hard limit. Some Directors saw merit in consideration of a global cap, perhaps at SDR 50 billion, to ensure that sufficient resources remain available for support under traditional arrangements. Others expressed concern that such a cap might not be fair (“first come first served”) and in any event might be difficult to adhere to in practice. Moreover, it is important to consider the counterfactual: to the extent that the RAL reduces the likelihood of crisis, the demand for actual Fund drawings would be smaller than in a world without the RAL. With these considerations in mind, staff suggests that if commitments under the RAL reach a certain level, for example SDR 30 billion, the RAL would be reviewed. Such a review would consider the experience under the RAL, the design of the instrument (e.g., monitoring, terms, and access), whether it was being used appropriately, and how to proceed.
Establishing the RAL as part of the Supplemental Reserve Facility
29. Directors generally favored establishing the new instrument as part of the Supplemental Reserve Facility. The SRF Decision would need to be amended to allow for the approval of an arrangement under the RAL to address on a precautionary basis the SRF’s special balance of payments need. Establishing the RAL as part of the SRF would mean the following:
Circumstances for drawing. RAL resources could only be drawn for the special balance of payments need for which the SRF is designed, namely, “the member is experiencing exceptional balance of payments difficulties due to a large short-term financing need resulting from a sudden and disruptive loss of market confidence reflected in pressure on the capital account and the member’s reserves, if there is a reasonable expectation that the implementation of strong adjustment policies and adequate financing will result, within a short period of time, in an early correction of such difficulties.” From this “circumstances test” it follows that the RAL would only be a relevant instrument for members that are highly integrated, or in the process of integrating, into capital markets.
Charges and maturities. As for the SRF, financing under the RAL up to the exceptional access limit (up to 100 percent of quota, depending on whether the member has outstanding resources) would be made available with credit tranche rates of charge and repurchase periods. The remaining amounts would carry the same surcharge and maturity structure as the SRF. Currently, the SRF’s surcharge structure is time-based, starting at 300 basis points and rising to 500 basis points in six-monthly steps. And resources are expected to be repaid within 2-2% years, but this expectation can be extended by up to six months on request; resources are required to be repaid 21/2-3 years after a purchase. The upcoming follow-up paper on charges and maturities will examine the structure of surcharges for large-scale access, including under the SRF; if the RAL is established as part of the SRF, any changes from this review would also apply to the RAL.
Commitment fees
30. The commitment fee for the RAL is likely to be an important factor, particularly as the member is not expected to draw. The fee, which is set uniformly across all GRA arrangements, is 25 basis points for the first 100 percent of quota, and 10 basis points for access above that level. For a RAL arrangement with access of 300 percent of quota, this would give an average fee of 15 basis points per year. In principle, the commitment fee should reflect the Fund’s costs in providing GRA resources to the membership. Staffing and administrative costs for the approval and monitoring of an arrangement under a RAL would be higher than for a surveillance case, but probably similar to those of a precautionary stand-by arrangement. There is also a cost associated with the pre-commitment of Fund resources, but this is difficult to quantify. Commitment fees would provide some income for the Fund, but the fees should not be set with the aim of addressing the Fund’s current income situation.
31. Staff suggests, for the time being at least, to leave commitment fees unchanged at current levels. Commitment fees were reduced in the context of the 2000 Review of Fund Facilities, with the objective of making them more appropriate for the CCL. Given the unique role of the Fund and its financing, there is no satisfactory private-sector comparator, but available evidence suggests that the Fund’s commitment fees would be slightly lower than the 20-50 basis points charged for a typical bank credit line. They may also be substantially lower than the implied cost of holding reserves—although the overall cost to the member would also need to take into account the member’s perceived probability of drawing under the RAL.6 Staff suggests maintaining the existing structure of commitment fees, but with an expectation that this structure could be reviewed more generally as experience is gained.7
Sunset clause
32. A sunset clause for the RAL could be considered. A potential benefit is that it would help ensure that the Fund does not maintain instruments that are not used. At the same time, the Fund can decide at any time (as it has in the past) to eliminate instruments that are not being used. A disadvantage with a sunset clause is that the sunset could arrive before there was a sufficient period to test the new instrument’s potential merits, and that a forced termination could unnecessarily be seen as a failure. For these reasons, staff would not favor including a sunset clause. But if such a clause were considered useful, staff would recommend it be set far enough out, say, 8-10 years, to ensure sufficient experience with the cycles in market conditions. The CCL had a five-year sunset clause.
Voting majorities 8
33. Majority requirements for creating the RAL will depend on its terms. The Articles authorize the Fund to establish a special repurchase period for “special policies”; i.e. policies that are designed to address a special balance of payments problem (often referred to as “special facilities”). As set forth in Article V, Section 7(d), the establishment of a special facility with special repurchase periods requires an 85 percent majority of the total voting power (the SRF and EFF are examples of special facilities with special repurchase periods). The existence of a special balance of payments problem also provides the basis for establishing a special rate of charge, which requires a 70 percent majority of the total voting power (the SRF is an example here). Importantly, however, an amendment of a special facility with a special repurchase period and/or rate of charge does not require a qualified majority if: (a) the balance of payments problem being addressed under the special facility is unchanged by the amendment, (b) the amendment does not change those features that required the qualified majority (i.e., the repurchase period and/or rate of charge ), and (c) in the case where the special facility has a sunset provision, the amendment does not extend the duration of the facility. This reflects the general rule set forth in Article XII, Section 5(c) that “Except as otherwise specifically provided, all decisions of the Fund shall be made by a majority of votes cast.” Thus, for example, modifications to a special facility that change the level of access or the conditions regarding disbursement—but do not change the nature of the balance of payments problem being addressed by the facility—only require a decision taken by a majority of votes cast.
Establishing the RAL as part of the SRF. Given the framework summarized above, only a majority of votes case would be required to establish the RAL as part of the SRF, as the RAL would address the same balance of payments problem and no changes are being proposed to the SRF’s existing special charges and repurchase period.9
Establishing the RAL for a different balance of payments problem. Conversely, were the RAL to be established as a new facility to address a different balance of payments problem (as was the case with the CCL where the balance of payment problem was defined as arising primarily from contagion), then a qualified majority would be required: (a) if the RAL were to have a special repurchase period (85 percent of total voting power), or (b) if the RAL were to have a special rate of charge (70 percent of total voting power).
F. RAL and the Private Sector
34. The private sector and member countries have not been able to find an instrument that provides reliable crisis-contingent financing. Any new Fund facility should complement and not compete with private sector instruments. However, the private sector has been reluctant to supply insurance to sovereigns that would provide sufficient and reliable resources to stem a crisis at rates that are of interest to countries during periods of stress (Box 6).10 Providing a fixed interest rate contingent financing that would be drawn when market rates are increasing would imply losses for banks which mark to market. To provide for these prospective losses, banks would have to charge ex ante commitment fees at levels that are unattractive to countries in comparison with the cost of self-insurance through the accumulation of reserves.11 Moreover, the private sector may offset its exposure from the contingent financing thereby limiting the benefits of such financing to the country.
Private Sector Contingent Credit Lines for Sovereigns1
Experience. Experience with private contingent credit lines for sovereigns is very limited. In three cases resources were insufficient to stem crises or were offered at rates unattractive to sovereigns.
Indonesia drew on a series of private contingent credit lines arranged with mostly Japanese banks during its crisis in 1997. The resources obtained under these credit lines proved too small to reduce effectively the impact of the crisis.
Mexico drew on its fixed rate credit lines in 1998. Banks that were marking to market were booking losses as the prevailing market rate increased. Consequently, they offered a variable interest rate on any subsequent facility—making it less attractive to Mexico.
Argentina’s central bank entered into repurchase agreements with international banks that would provide temporary liquidity against Argentine Euro bonds. It drew US$1.5 billion in September 2001, but the funds were insufficient in the face of a growing crisis.
Availability. Banks can provide (non-contingent) syndicated loans that the country might draw at any time. However, private providers of these lines hedge themselves against losses in a crisis by shorting their exposure, potentially exacerbating the crisis. By contrast, the RAL can act as a stabilizing force by providing fresh and unhedged resources at the time of a crisis.
1 See “Bailouts or Bail-Ins?” by Roubini and Setser, Institute for International Economics, August 2004.G. RAL’s Relation with Other Fund Instruments
35. It is important to consider the relationship between the RAL and other Fund instruments, especially precautionary stand-by arrangements in the credit tranches.
Precautionary stand-by arrangements have been a relatively popular form of engagement with the Fund, and they have had a long and successful track record. There is a concern that, if an instrument like the RAL is established, it could be more attractive than a precautionary arrangement, thereby muddying the latter’s signal if it is viewed as a “second-tier” product.
36. The RAL’s qualification structure is crucial to ensuring it is used appropriately, while preserving a clear role for precautionary arrangements in the credit tranches. The RAL’s qualification criteria are intended to ensure that the instrument is approved only for members in a clearly defined situation: those with strong macroeconomic policies and policy credibility, but with some remaining vulnerability to capital account crisis, and that have a very specific type of potential balance of payments need. Some countries that have used precautionary arrangements in similar circumstances might qualify for a RAL, and it is probably unavoidable that the RAL will be seen as providing the most positive signal for such members. However, well-defined entry criteria can help to prevent the use of the RAL where it is not the appropriate instrument. Also, some members may actually prefer the lower access (and the associated lower commitment fee) or the monitoring structure of a stand-by arrangement in the credit tranches.
IV. Next Steps
37. Following the Board discussion of this paper, a short paper setting out the steps required to establish a new instrument could be prepared soon. This paper would set out the modifications to existing decisions and policies that would be required to create the RAL. Depending on Board views and IMFC support, it might be possible to establish a new instrument as a next step. Creating the instrument could itself lead members meeting the criteria to request for it. However, there may be merit in waiting until there is more concrete interest in using the instrument. This would allow potential users to have an input on final design issues and could help in the Board’s decision-making process. It would also (obviously) reduce the risk that a new instrument is never used.
V. Issues for Discussion
38. Directors may wish to comment on how they would see a new liquidity instrument fitting within the Fund’s mandate for crisis prevention.
39. This paper focuses on specific design issues raised by Directors in August last year and by participants in the outreach meetings. In this context, do Directors:
See merit in the outlines of a qualification structure that is based on an analysis of specified objective indicators in a framework that allows for some judgment?
View a monitoring structure based on semi-annual reviews as providing for appropriate safeguards while preserving a strong degree of predictability?
See merit in a uniform access level of 300 percent of quota, or in a range (i.e., 300 to 500 percent of quota)?
See merit in establishing the RAL as part of the SRF?
Prefer a one-year timeframe for each arrangement, a rolling one-year structure, or a more flexible standard of up to two years, with successor arrangements under the RAL not to be discouraged?
See merit in maintaining commitment fees at existing levels, at least initially?
See merit in not including a global cap on commitments under the RAL, but rather that a review should be triggered if access under the RAL exceeds SDR 30 billion?
See merit in excluding a sunset clause that might be counterproductive, or if needed, that it be set at 8-10 years?
Have views on the next steps?
Annex I. RAL Relevant Bibliography
The following documents issued by the International Monetary Fund are referred to in this paper or provide useful background.
“Consideration of a New Liquidity Instrument for Market Access Countries,” (August 4, 2006); see Public Information Notice No. 06/104 (September 13, 2006), available via the Internet at http://www.imf.org/external/np/sec/pn/2006/pn06104.htm . Referred to as the “August paper.”
“Communiqué of the International Monetary and Financial Committee of the Board of Governors of the International Monetary Fund,” (September 17, 2006), available via the Internet at http://www.imf.org/external/np/cm/2006/091706.htm .
“The Managing Directors Report on Implementing the Fund’s Medium-Term Strategy,” (April 5, 2006), available via the Internet at http://www.imf.org/external/np/exr/ib/2006/041806.htm .
“Fund Financial Support and Moral Hazard—Analytics and Empirics,” (March 2, 2007), available at http://www.imf.org/external/pp/longres.aspx?id=4043.
“Country Insurance: The Role of Domestic Policies,” (June 19, 2006), available via the Internet at http://www.imf.org/external/np/pp/eng/2006/061906.pdf .
“Fund-Supported Programs and Crisis Prevention,” (March 23, 2006), available via the Internet at http://www.imf.org/external/np/pp/eng/2006/032306.pdf .
“Precautionary Arrangements—Purposes and Performance,” (March 23, 2006), available via the Internet at http://www.imf.org/external/np/pp/eng/2006/032306p.pdf .
“Crisis Prevention and Precautionary Arrangements—Status Report,” (September 3, 2004), see Public Information Notice No. 04/117, available via the Internat at http://www.imf.org/external/np/sec/pn/2004/pn04117.htm .
“Adapting Precautionary Arrangements for Crisis Prevention,” (June 11, 2003), see Public Information Notice No. 03/146, available via the Internet at http://www.imf.org/external/np/sec/pn/2003/pn03146.htm .
“Review of Contingent Credit Lines,” (February 12, 2003), available via the Internet at http://www.imf.org/external/np/pdr/fac/2003/021103.htm, and Executive Board Assessment, available via the Internet at http://www.imf.org/external/np/sec/pn/2003/pn03146.htm#eba032003 .
“Completion of the Review of Contingent Credit Lines,” (November 12, 2003), see Public Information Notice No. 03/146, available via the Internet at http://www.imf.org/external/np/sec/pn/2003/pn03146.htm .
“Access Policy in Capital Account Crises,” (July 30, 2002), available via the Internet at http://www.imf.org/external/np/tre/access/2003/072902.htm.
“Access Policy in Capital Account Crises—Modifications to the Supplemental Reserve Facility and Follow-Up Issues Related to Exceptional Access Policy,” (January 14, 2003), available via the Internet at http://www.imf.org/external/np/tre/access/2003/011403.htm; see Public Information Notice No. 03/37, available via the Internet http://www.imf.org/external/np/sec/pn/2003/pn0337.htm .
“The Supplemental Reserve Facility,” Decision No. 11627-(97/123), December 17, 1997, available via the Internet at http://www.imf.org/external/pubs/ft/sd/index.asp?decision=11627-(97/123) .
1 This paper was prepared by Alan MacArthur, Ulric Erickson von Allmen, Michèle Shannon, Peter Breuer, and Brett House.
Market participants included hedge funds, asset managers, sell-side firms, a ratings agency, and a trade association.
The FTP is the mechanism through which the Fund selects members to provide currencies for use in the Fund's lending transactions.
See Annex I for references.
Not with standing the access level, the member's drawing could not exceed its actual balance of payments need at the time of the drawing.
The average opportunity cost of holding reserves over the last five years (2002-06) measured against the U.S. term premium (i.e., the difference between US Treasury average long and short rates) was about 180 basis points, while the opportunity cost estimated as the full difference between five-year average emerging market yields and U.S. short-term treasuries was about 440 basis points.
Any change to the commitment fee structure requires a decision taken by a 70 percent majority of the total voting power.
This portion of the paper was prepared by the Legal Department.
The fact that the RAL will be approved in a precautionary setting does not change the fact that the resources, if disbursed, would be used to address the same balance of payments need that is covered by the RAL. See Adapting Precautionary Arrangements to Crisis Prevention (June 11, 2003), where it was concluded that an amendment of the SRF to make it available on a precautionary basis would only require a majority of the votes cast.
Separately, a question was asked about whether a country could pledge the resources that could be drawn under the RAL to a private lender, thereby securing a private loan at a reduced interest rate. However, such an arrangement could be covered under some of the negative pledge clauses found in sovereign debt contracts.
In "Pricing and Hedging of Contingent Credit Lines," (forthcoming in the Journal of Derivatives), Loukoianova, Neftci, and Sharma discuss the analytics of contingent credit lines in the corporate sector. They describe various fees and rates associated with such credit lines—a participation fee of (typically) one basis points, annual fee on undrawn balance in the 8-13 basis points range (in 2001), and interest rate on drawn balances of 73 basis points in 2001 (53 in 2002) over LIBOR at 3.8 percent (2.2 percent). There are many similarities with the RAL in terms of benefits (signal, insurance) and costs, but one key difference is that corporates can draw on these credit lines at anytime only subject to a material circumstance clause that allows the bank to deny credit if the client's financial condition changes in a substantive way—e.g., if the borrower is downgraded.
