Adequacy of the Global Financial Safety Net––Proposal for a New Policy Coordination Instrument

The global financial safety net (GFSN) has become larger and more decentralized, creating a need for greater coordination.


The global financial safety net (GFSN) has become larger and more decentralized, creating a need for greater coordination.


1. The global financial safety net (GFSN) has evolved over time with reforms taking place at the regional and international level. Following the global financial crisis there was a renewed push to reform the GFSN to strengthen crisis prevention, provide adequate financing when a crisis hits, and incentivize sound macroeconomic policies. The size of the GFSN expanded significantly, the number of participants providing financing increased, and new tools for crisis prevention and adjustment were introduced or made available to a wider set of countries. The result is a GFSN which is larger, more decentralized, and consists of numerous instruments designed to address a wider variety of shocks.

2. As the GFSN has become more decentralized, there is a strong consensus on the need to improve cooperation between its various layers. As part of the Fund’s work on strengthening the international monetary system (IMS), an assessment of the adequacy of the GFSN identified important gaps in the current architecture.1 In particular, many countries need to access multiple layers of the GFSN to cover financing needs, resulting in coordination challenges in the event of a shock. In addition, incentives to implement sound policies for crisis prevention or to enhance macroeconomic stability could be strengthened. Specifically, the Executive Board called for further work on an instrument “that could help countries better coordinate their access to multiple layers of the GFSN and signal their commitment to a policy reform agenda”.

3. A new Policy Coordination Instrument (PCI) would help signal commitment to reforms and catalyze financing. The proposed new non-financing instrument is designed to assist countries to formulate and implement a macroeconomic policy program, signal commitment to an economic reform agenda, and possibly help unlock financing from official creditors (e.g., Regional Financing Arrangements (RFAs)) or private investors. It would enable closer cooperation with the Fund, where countries commit to a coherent and credible policy framework and more intensive monitoring of policies, without the need for Fund financial resources. By assisting countries that do not need Fund resources but may seek financing from other sources, the PCI facilitates a more efficient allocation of global resources, helps to reduce moral hazard by incentivizing stronger policies, and reduces problems associated with “facility shopping”, including from MDBs. It would thus also help advance the G20 Principles for ensuring effective coordination between the IMF and MDBs in case of countries requesting financing while facing macroeconomic vulnerabilities.

4. The proposed introduction of the PCI is part of a broader set of proposals to reform Fund facilities and instruments to address gaps in the GFSN. These proposals include establishing a new backstop for “swap-like” liquidity support to enhance the Fund’s crisis prevention toolkit. Stronger IMF-RFA cooperation is also an important aspect of strengthening the GFSN. In addition, a comprehensive review of LIC facilities, including a review of the Policy Support Instrument (PSI), is planned for 2018.

5. Consultations with RFAs indicate that there is broad support for the PCI. RFAs have developed in Asia, Europe, Latin America, the Middle East, and Africa, as well as in cross-regional groupings such as the recently established BRICS contingent reserve arrangement. The objectives and operations of RFAs vary depending on their origins. Some explicitly require Fund involvement in their lending programs while others do not. In a series of consultations with RFAs, there was broad support for the PCI as a Fund policy monitoring instrument, although it was noted that there would continue to be cases in which Fund financing would be necessary. Broader issues relating to IMF-RFA interactions—including those arising from differences in governing structures, instruments, and activities—will be covered in a separate paper. Operational guidelines resulting from that work will inform the modalities for using the PCI in the context of IMF-RFA cooperation.


6. The PCI is designed to enable closer policy dialogue with countries, regular monitoring of economic developments and policies, and Board endorsement of those policies. The overall objective would be to support countries in designing and implementing policies through a fully-fledged macroeconomic program to (i) prevent crises and build buffers, (ii) enhance macroeconomic stability, or (iii) address macroeconomic imbalances. The PCI would be available to all member countries except those that need Fund financial support at the time of PCI approval or have overdue obligations to the Fund.

7. The PCI would be based on the design of Fund financial arrangements, but have important differences from currently available tools. The design, request, and review process is based on Fund financial arrangements, with several design features drawing on the Policy Support Instrument (PSI).2 There are some mechanisms in the Fund’s toolkit that can play the role of providing policy monitoring and signaling, but important gaps remain. Compared with mechanisms in the current toolkit, the key differences of the PCI are:

  1. a.Eligibility for all member countries. While Fund arrangements treated as precautionary (e.g., under the SCF or SBA) can provide closer monitoring and a clear signal, they are not available to members that have no present, potential, or prospective balance of payments need to draw on Fund resources, for instance because of substantial market access and/or access to other official sources of financing.3 The PSI provides such support but its availability is restricted to a subset of PRGT-eligible countries that have a policy framework focused on consolidating macroeconomic stability while deepening structural reforms in areas where growth and poverty reduction is constrained. The PCI would have no qualification criteria, but countries would have to commit to policies strong enough to constitute the basis for a Fund-supported program (i.e., upper-credit tranche (UCT) quality policies).
  2. b.Executive Board endorsement of policies and interim updates. Within the current toolkit, assessment letters are generally not published, and neither the assessment letters, nor staff monitored programs (SMPs) involve Board endorsement of policies. The PCI would have regular reviews by the Board, and interim updates by staff to the Board if reviews cannot be completed within a limited buffer period.
  3. c.A clear approach for assessing policies. While surveillance through Article IV consultations do provide some form of policy monitoring and signaling, there are generally no clear targets or economic policies which have been agreed ex-ante and are then explicitly monitored and reported. In addition, Article IV consultations are typically conducted once a year, whereas there are cases when more frequent monitoring would be warranted. As noted above, the proposed design of the PCI would ensure that programs monitored under the instrument meet the UCT standard. The PCI differs from the SMP, which can only be used to allow countries to build a track record, and thereby potentially pave the way for a Fund-supported program or Fund emergency assistance, or for the resumption of an off-track program.

A. Eligibility

8. There would be no qualification or eligibility criteria. The instrument would be available to all member countries except those that need Fund financial support at the time of PCI approval, or those with overdue obligations to the Fund. PCIs for countries that seek financing from other sources would require assurances that such externally-supported programs are fully financed in line with the Fund’s policy on financing assurances. Financial assistance could be from other official partners, such as RFAs, multilateral development banks, bilateral creditors or donors. Legally, the PCI is a form of technical assistance (TA) that is voluntary for both the member and the Fund.4

9. The unrestricted eligibility of the PCI is the main reason for staff’s proposal to maintain the PSI alongside it. Removing the PSI would leave PRGT countries without the signaling benefit provided by its qualification requirements. Eliminating the PSI could therefore take away an important positive differentiating factor for a subset of the membership that may need it most as PRGT-eligible countries are typically monitored less extensively by rating agencies, global financial institutions, and other private sector agents than more advanced economies (see also ¶¶23-24).

B. Program Standard

10. Countries would commit to upper credit tranche (UCT) quality policies.5 In the context of Fund financial support, the standard of UCT quality policies aims at ensuring that: (i) policies will allow the country to solve its balance of payments difficulties within the program period in a manner consistent with the Articles of Agreement; (ii) the likelihood the Fund gets repaid is high; and (iii) the Fund gets repaid relatively quickly, consistent with the revolving nature of its resources. For the PCI, this standard can be interpreted as signifying policies strong enough to constitute the basis for a Fund-supported program should a user request Fund financial assistance, which is identical to the standard set for the PSI.

11. The Fund’s policies on financing assurances, and arrears to external official and private creditors, would apply. The PCI would require the same financing assurances as Fund financial arrangements. This means that, where needed, creditors and donors have provided assurances that they will extend the necessary support to meet the program financing requirements. Specifically, the Fund would need to be satisfied that (i) firm commitments of financing are in place for the first 12 months following approval of the PCI, and, (ii) good prospects are in place that there will be adequate financing for the remaining period of the PCI beyond the first 12 months. Program reviews would require assurances of full financing for each successive 12-month period (or whatever period is left under the PCI). The Fund’s policy on sovereign external arrears to official multilateral creditors, in which the Fund generally calls for the resolution of existing arrears and non-accumulation of new arrears during the period of the program, would apply to the PCI by analogy.6 The Fund’s policies on arrears to official bilateral creditors and on Lending into Arrears (LIA) to private creditors would also apply by analogy in terms of enabling PCI engagement, on a case-by-case basis, in a situation of sovereign arrears to either group of external creditors.7 As with the PSI, a safeguards assessment is not required but would be encouraged on a voluntary basis as a way of enhancing the accountability, transparency, and institutional strength of central banks. It would also facilitate quick access to Fund resources in the event a need should materialize.

C. Modalities: Activation, Conditionality and Reviews

12. A PCI would typically be expected to have a duration of two to three years, similar to the duration of an SBA. However, there would be sufficient flexibility to meet the varying needs of countries’ reform agendas. In this regard, a PCI could be approved for a duration of six months to four years, and could be extended within the four year maximum. PCIs of less than one year would require at least one scheduled review. There would be no limit on the number of successor PCIs.

13. The process for requesting a PCI would be analogous to Fund financial arrangements. Countries would express an interest in a PCI, after which staff and country authorities would begin discussions on the scope and content of the program. Country authorities would present a Policy Statement setting out their program objectives and specific commitments in support of their request; it may be accompanied by a technical memorandum. The program would be underpinned by a quantified macroeconomic framework and debt sustainability analysis (DSA). Depending on country circumstances, policy objectives would typically include fiscal policies, monetary policies, exchange rate policies, financial sector policies, and macro-critical structural reforms. Progress against these objectives would be assessed and updated at each program review.

14. Monitorable measures and targets would be set to assess whether a country’s policies and implementation under PCIs are strong enough to warrant continued endorsement by the Fund. The program would be monitored by quantitative targets (QTs) and/or reform targets (RTs), which would be analogous to performance criteria (but monitored with a review-based approach as discussed below) and structural benchmarks in the context of Fund financial arrangements.8 Prior actions (PAs) before approval or completion of a review would be applied in situations where it is critical for the successful implementation of the program; standard continuous targets would also be included.9 Conditionality (PAs, QTs, RTs and standard continuous targets) would cover macroeconomic variables and structural measures that are reasonably within the member’s direct or indirect control and that are, generally, either (i) of critical importance for achieving the goals of the program10 or for assessing program implementation, or, (ii) analogous to the UFR context, necessary for the implementation of specific provisions of the IMF Articles of Agreement or policies adopted thereunder. In some cases, conditions may be outside the Fund’s core area of expertise, in which case a more detailed explanation of their critical importance would be required. The design and modalities for these conditions should be consistent with the Fund’s Guidelines on Conditionality.11

15. Reviews would follow the same general process as under financial arrangements meeting the standard of UCT quality policies. A country’s past performance under the PCI, including performance against conditions such as RTs, QTs and standard continuous targets, would be assessed in the context of periodic reviews, and policy understandings for the future, including updates to program design and conditionality in cases of changed conditions or prospects, would need to be reached in the context of reviews. Test dates for QTs and RTs would normally be set on a semi-annual basis corresponding to scheduled reviews. At approval of a PCI or at the completion of a review, QTs and/or RTs would be set for the subsequent two reviews (i.e., for a 12-month period in the case of semi-annual reviews) or through the end of the PCI if it is scheduled to expire in less than 12 months. Completion of a review by the Board would signify the Fund’s assessment that policies and commitments are on track to meet program objectives, in accordance with the standard of UCT quality policies.

Fund Experience with Fixed Review Schedules

Unlike Fund financial arrangements that allow flexibility in review schedules, the PSI was created with a rigidly fixed schedule of reviews.1 Under the PSI, reviews are normally conducted on a semi-annual basis irrespective of the status of program implementation or prospects. If performance is not sufficiently strong, or if an understanding on the forward-looking elements of the policy program cannot be reached, a staff report is still circulated to the Board, recommending that the review not be completed at that time. After this “off” signal, the member could get back on track by completing the review before the test date for the quantitative assessment criteria linked to the next scheduled review, or by completing the next scheduled review. This is in contrast with the practice under Fund financial arrangements which allows flexibility to delay reviews without circulating a staff report to the Board when circumstances for the completion of the review are not in place.

Fixed review schedules provide regular “on/off” signals, but create certain risks for program implementation. For instance, there could be pressure on the Fund to complete a review by the scheduled date even when there are delays/slippages in policies to meet program objectives, given the potential for negative consequences to both the Fund and the member of an “off” signal. Greater flexibility in the review schedule could reduce such risks by allowing the member some additional time, when necessary, to take key policy actions and meet objectives. However, the downside is that the signals could become less regular and entail gray periods. For example, a delay in the review could create uncertainty as to whether the review would be completed soon (i.e., the member’s performance brought back on track) or it would never be completed (i.e., performance remains off track).

The rigidly fixed review cycle for the PSI may have put pressure on the completion of reviews. In ten years, there was a single non-completion of a review under the PSI, suggesting there may be a bias towards completion of PSI reviews. Experience also indicates that discussions with the authorities on the staff reports and MEFP are conducted until the very last minute before sending documents to Management or the Board in order to meet fixed review dates, indicating the desirability of adding some flexibility to the review schedule.

1 See The Chairman’s Summing Up—Implementation of the Policy Support Instrument, Executive Board Meeting 05/85, 10/05/05.

16. Reviews would be on a fixed schedule, with a limited time buffer to provide some flexibility (Figure 1). A review schedule will be set by the Board at the time a PCI is approved. Reviews by the Board would normally be on a semi-annual basis, although a more frequent review schedule (e.g., quarterly) could be applied in the case of urgent and wide-ranging reform agendas and/or when doing so could improve coordination with other financing partners. Unlike the PSI, where the reviews are based on a rigid schedule and are conducted irrespective of the status of program implementation or prospects (Box 1), it would be possible to delay PCI reviews to provide the authorities a short buffer to implement overdue policies, to take corrective measures, or to mobilize necessary financing to close financing gaps. If the review is not completed within three months (in the case of semiannual reviews) from the scheduled review date, an interim performance update by staff would be provided to the Board for information. An ad-hoc Board briefing could also be proposed if needed. A brief factual statement in the form of a press release stating the issuance of the performance update would be published, and the performance update report would be published subject to the member’s consent.12 The update on program performance would reduce uncertainty as to why the review has not been completed by providing an overview of program performance, indicating the sources of the delay, and highlighting the corrective actions needed in order to bring the program back on track. A review cannot be completed after the performance update has been triggered. The timing of the interim performance update is informed by recent experience under the GRA and PRGT, where about two-thirds of all reviews are completed with a maximum delay of three months (Figure 2). Non-completion of a review for a twelve-month period since the Board date of the most recently completed review would signify lack of Board endorsement on the member’s policies and result in an automatic termination of the PCI.* A factual statement in the form of press release stating the termination of the PCI would be published. For programs on a more frequent review schedule, the period after which a performance update is required would be adjusted downward (e.g., 1½ months for programs on a quarterly review schedule), but the twelve-month rule for the automatic termination of PCIs would be the same.

Figure 1.
Figure 1.

Illustrative Scenarios Under a Three-Year PCI

Citation: Policy Papers 2017, 041; 10.5089/9781498346610.007.A001

Figure 2.
Figure 2.

Review Delays Under GRA, PRGT and PSI Arrangements

Citation: Policy Papers 2017, 041; 10.5089/9781498346610.007.A001

Source: IMF MONA Database

17. To help alleviate stigma and streamline the review process, program monitoring for PCIs would be fully review-based. The PCI would include a macroeconomic framework, along with a set of QTs and/or RTs that country authorities would commit to achieve by each test date (semi-annually, in most cases), as well as standard continuous targets. Under a review-based approach, QTs would differ from performance criteria (PCs) or assessment criteria (ACs) in that a failure to meet a QT would not trigger the need for a formal waiver. Rather, deviations from targets would serve as indicators that the PCI might no longer be on track for successful implementation (similar to indicative targets and structural benchmarks in current monitoring of Fund arrangements and PSIs). Completion of the review would require a judgment by the Board that there are compensating factors (e.g., that missed targets are assessed to be the result of minor or temporary factors, or corrective actions have been taken). In case of substantial deviations from target(s) and insufficient commitments to correct slippages, the review would not be completed.13

18. The Fund successfully moved to a review-based approach to structural conditionality in 2009 (Box 2). The elimination of structural performance criteria (and hence the elimination of the need for waivers) as a modality for monitoring performance under UFR arrangements has not led to a deterioration in implementation rates. Moreover, when this decision was adopted, it was explicitly recognized that program reviews provide a context for a more robust and “real-time” assessment of program implementation. Program reviews place the focus on the overall staff assessment, which provides a bottom-line appraisal as to whether policies are on track to meet the specified objectives, rather than on deviations from specific targets. In practice, the removal of waivers does not impact the quality of policies that are required for successful completion of the program.

Fund Experience with Review-Based Approach to Monitoring Conditionality

The IMF has already made a partial transition to review-based conditionality by eliminating structural PCs. In the 2009 conditionality review, staff proposed shifting to review-based approach to conditionality for all facilities and instruments in order to reduce stigma associated with the need to request waivers and modification of performance criteria, as well as to consolidate the central role of program reviews in program monitoring. The Board agreed to the discontinuation of structural PCs, effectively shifting to review-based approach for structural conditionality through the use of structural benchmarks. However, Directors decided to keep quantitative performance criteria, as they felt that their elimination would (i) reduce safeguards to the Fund’s resources, (ii) weaken the Fund’s positive signaling role, and (iii) diminish its effectiveness as a domestic commitment tool.


Share of Missed Conditionality in GRA Programs, percent

Citation: Policy Papers 2017, 041; 10.5089/9781498346610.007.A001

Source: IMF MONANote: Structural conditionality includes structural PCs (pre-2009 reform program reviews) and structural benchmarks, but excludes ex-ante conditionality (prior actions). Quantitative conditionality only includes performance criteria. Missed conditionality includes targets marked as “not met” or “waived” in MONA.

There is no indication that the review-based approach in respect of structural conditionality using structural benchmarks has eroded program performance. The partial shift to a review-based approach in 2009 (for structural, but not quantitative targets) provides a test of whether review-based monitoring weakens program performance. The left panel in the figure above shows that implementation of structural conditionality has not worsened in the wake of the elimination of structural PCs. In fact, the share of missed structural targets has decreased from pre-reform levels. The right panel in the same figure shows that this decline in missed conditionality has not extended to quantitative conditions (or at least it has not been as marked). So implementation of structural conditionality—after structural PCs no longer applied—has tended to improve relative to the implementation of quantitative targets. The experience suggests that a review-based approach does not dilute the signaling role of conditionality, nor does it diminish the effectiveness of conditionality as a commitment tool.

In practice, the removal of waivers does not impact the quality of policies that are required for successful completion of the program. In particular:

  • A review-based approach does not raise the importance of staff judgement in deciding the completion of the review. That decision remains with the Executive Board, and the staff report under review-based monitoring would continue to provide evidence and analysis to inform that decision (as is currently done in other UCT-quality programs).

  • The elimination of waivers does not affect the importance of targets. Their status would inform the staff appraisal and the Executive Board’s decision in the same manner as they do currently. For instance, while there are no structural PCs in UFR programs, reviews have been delayed due to slippages in the implementation of the structural reform agenda, even when quantitative PCs have been met.

  • A review-based approach to monitoring does not constitute a deviation from the UCT standard, as these are distinct concepts. The UCT standard refers to the substance of the member’s adjustment program and the associated conditionality (i.e., its ability to solve its balance of payments difficulties within the program period in a manner consistent with the Articles of Agreement), while the monitoring framework refers to the procedure used for evaluating its implementation.

D. Other Design Issues

19. An on-track PCI could expedite access to Fund financing in the event of a member’s BOP need, subject to normal Fund policies on use of Fund resources. Accordingly, the concurrent use of the PCI and a RCF/RFI or SCF/SBA would be possible. Similar to the PSI, concurrent use with arrangements that deal with medium-term adjustment, the ECF/EFF, would not be possible. For the duration of the concurrent use, program modalities (including conditionality and monitoring) under the PCI would match those established for the financing arrangement. The PCI need not be cancelled when Fund financing is requested, as the PCI could improve continuity of policy cooperation. However, a simultaneous request of a PCI and access to financial support, or a request for PCI while a Fund financial arrangement is already in place, would not be permissible. A PCI may be cancelled by the authorities at any time, including when the country experiences financing or adjustment needs that are expected to extend beyond the short term or have impacted the authorities’ objectives or their capacity or commitment to implement the PCI program.

20. A tailored misreporting framework would apply to the PCI. Although the PCI does not entail the use of Fund resources, and the Fund’s misreporting framework has not been applied in the past to targets that are not binding (in the sense that they would prevent completion of a review absent a Board waiver), the credibility of the data is critical for the monitoring and signaling objectives of the program. This calls for the use of a misreporting framework even in the context of review-based approach to monitoring conditionality. Analogous to the PSI, a misreporting framework would apply to prior actions, QTs and standard continuous targets.14 Only variables or measures that are objectively monitorable and critical for program implementation or monitoring will be established as QTs and be subject to the misreporting framework.

21. The misreporting framework would be analogous to that applied to the PSI.15 When evidence comes to the attention of the staff indicating that the country’s reporting of information on performance under QTs, standard continuous targets, or PA’s was inaccurate in relation to a PCI approved or a review completed within the preceding three years, the MD shall promptly inform the country concerned. If the MD finds, after consultation with the country, that the country had reported inaccurate information to the Fund, the MD shall promptly notify the member of this finding. The Board would then consider the misreporting at the time of the next scheduled review based on a combined misreporting/review staff report. The Board shall assess program performance in light of the revised information associated with the misreporting. Such a reassessment of past program performance in the light of a misreporting will not lead to the Board retroactively reversing its decision completing a review, which subsequently had become associated with a misreporting. If the Board determines misreporting has occurred (and it is not found to be de minimis), the member would be required to take corrective action as a condition for continued PCI support, and the relevant information would be published. In cases where Management determines that the possible misreporting is de minimis in nature, relevant notifications may be made by area departments. In cases of de minimis misreporting, the fact of misreporting shall not be published by the Fund.

22. The publication regime for PCI-related documents is proposed to be the same as for PSI documents under the Transparency Policy. The Board documents for both requests for the PCI and completion of reviews under the PCI would be sent to the Board for approval and, subject to the country’s consent, such documents, together with a press release containing the chairman’s statement, would be published pursuant to the Transparency Policy. Program performance updates prepared by staff would be issued to the Board for information before publication, and a brief factual statement to that effect would be published. While publication of PCI-related documents would be voluntary, they would be subject to the stronger publication presumption as applied to the PSI, so as to provide a clear and transparent signal.

Implications for the Fund

A. Toolkit and GFSN Cohesion

23. The introduction of the PCI would create significant overlap with the PSI, potentially creating scope for streamlining the toolkit. PCI users would benefit from a similar engagement with the Fund as that currently enjoyed by a narrow subset of PRGT-eligible countries. However, the PCI would be available to all member countries, and would improve on a number of design aspects, notably the review schedule and monitoring framework. Universal availability has one potential downside: as an instrument available to all, the PCI would not provide the same signal that the PSI currently provides as a consequence of its use being restricted to PRGT-eligible countries that have no present macroeconomic adjustment needs and where institutions are of sufficient quality to support continued good performance.

24. Staff proposes that, for now, the toolkit should maintain both the PSI and the PCI. This would be in line with the current GRA/PRGT delineation of financial instruments and facilities: while LICs benefit from a tailored set of lending instruments and facilities, they can opt to access GRA resources on an equal footing with non-LICs. It is difficult to predict whether potential PSI qualifiers would prefer the additional signaling element offered currently by the PSI or the universality and the improved design aspects of the PCI. However, should the PCI attract a more diversified user base—both geographically and in terms of income level—current PSI users may perceive a benefit in switching to a PCI even if both instruments remain available. If so, the continued existence of the PSI could be reconsidered in the context of future reviews of the PSI, the next one of which is scheduled for 2018. The 2018 review could also consider the case for amending the PSI—if the decision is made to keep the instrument—to introduce some of the proposed elements of the PCI, such as providing for a more flexible review schedule and moving to review-based conditionality.

25. The PCI could result in a more efficient use of existing GFSN and Fund resources. Many of the recent SBAs treated as precautionary had low- to mid-level access. In these cases, supporting the authorities’ reform programs and signaling commitment to such reforms provided the main motivation for the program request. As the PCI is targeted at helping member countries achieve these objectives, it is expected that the new instrument could reduce demand for low-access arrangements treated as precautionary, including in cases where such arrangements are used to tap RFA resources and/or other layers of the GFSN. This shift would unlock Fund resources, thereby contributing to a key goal of ongoing GFSN reforms: a more efficient use of existing GFSN and Fund resources.

B. Resource Implications and Review of the PCI

26. Resource implications are uncertain but expected to be manageable. It is difficult to estimate the budget implications of the proposed instrument. Total cost would depend on potential demand, whereas as the marginal cost would depend on the counterfactual. For example, the marginal cost may vary between $0.5 and $1.2 million per year for a surveillance-only country seeking a PCI.16 However, for countries requesting a PCI instead of a low-access Fund arrangement treated as precautionary, the marginal cost is close to zero. In all, the resource implications should be manageable within the current budget envelope. However, given the uncertainty, it is proposed that the instrument be reviewed by the Board after the Board has approved 10 PCIs or after 5 years from the adoption of the PCI, whichever is triggered first.* Charging for PCI use will follow the relevant TA policy.17

C. Changes to Other Policies

27. The creation of the PCI has implications for other Fund policies. In line with the practice in program countries, it is proposed that the 24-month cycle for Article IV consultations be applied to members with a PCI, and the decision on Article IV consultation cycles would need to be amended accordingly. The Transparency and the Open Archives policies would also need to be amended to include the PCI. As with the PSI, the decision on post-program monitoring would be amended to exclude members with a PCI from PPM, and the decision on Lapse of Time Procedures for Completion of Program Reviews would likewise be applied to reviews under the PCI.18 Certain UFR policies, such as the Guidelines on Public Debt Conditionality and the policy on side letters, would also be applied to the PCI. The Decision on Strengthening the Effectiveness of Article VIII, Section 5 would be amended to subject the PCI to the same data reporting framework for the PSI.

Issues for Discussion

28. The PCI fills an important gap in the Fund’s toolkit at a time of heightened demand for improving coordination in the GFSN. The new instrument is designed to respond to demand from the membership to strengthen coordination in the GFSN and enable countries to signal commitment to reforms. The design of the instrument builds on the experience of existing tools to ensure it is appropriate for use by all countries, while also providing clear and transparent policy advice, conditionality, and monitoring. In responding to the staff’s proposal Directors may wish to address the questions below.

  • Do Directors support the introduction of a new Policy Coordination Instrument?

  • Do Directors agree with the proposed design and modalities of the PCI, including a fully review-based approach to monitoring conditionality?

  • Is there support for maintaining the PSI until a fuller assessment is made in the context of the review of LIC facilities in 2018?


See Strengthening the International Monetary System, IMF Policy Paper, February 2016 and Adequacy of the Global Financial Safety Net, IMF Policy Paper, March 2016.


The PSI is the Fund’s nonfinancial policy support instrument for a subset of PRGT-eligible countries. See Decision No. 13561-(05/85), October 5, 2005, as amended.


A balance of payments need can be present (i.e., a need that exists in the current period), prospective (i.e., a need that is expected/projected to arise in the future, including during the implementation of a Fund-supported program), or potential (i.e., a need that may arise under an alternative, typically downside, macroeconomic scenario, but is not expected to arise based on baseline/program projections).


As with the PSI, the legal basis for the PCI would be Article V, Section 2(b) of the Fund’s Articles, which provides that “if requested, the Fund may decide to perform financial and technical services […] that are consistent with the purposes of the Fund.” This activity would therefore be voluntary for both the relevant member and the Fund.


The term upper credit tranche refers to the use of IMF credit beyond the first 25 percent of a member’s quota, against which conditionality is then applied.


Since the PCI does not involve Fund financing, the arrears policies apply by analogy, i.e., it applies in full unless there is a policy justification for a deviation, taking into account consistency and uniformity of treatment. For example, the requirement under the policy on arrears to official bilateral creditors or under the Lending into Arrears policy for private creditors that there is a need for prompt Fund financial assistance is not applicable.


For the policy on arrears to official bilateral creditors, see BUFF/15/107. The LIA policy (BUFF/02/42) applies to both sovereign arrears to external private creditors and non-sovereign arrears arising from the imposition of exchange controls. The Fund can lend only where: (i) prompt Fund support is essential for the successful implementation of the member’s adjustment program; and (ii) the member is pursuing appropriate policies and is making a good faith effort to reach a collaborative agreement with its private creditors (or to facilitate such an agreement between private debtors and their creditors, and good prospects exist for the removal of exchange controls).


The new terms “quantitative targets” and “reform targets” would help signal the shift to a review-based approach to monitoring of conditionality.


As in Fund financial arrangements, members with a PCI are expected not to: (i) impose or intensify restrictions on the making of payments and transfers for current international transactions; (ii) introduce or modify multiple currency practices; (iii) conclude bilateral payments agreements which are inconsistent with Article VIII; (iv) impose or intensify import restrictions for balance of payments reasons; and (v) accumulate external payments arrears.


The goals would be program-specific, but could include, inter alia, building up of external buffers, ensuring debt sustainability, and maintaining monetary and financial stability.


See the Guidelines on Conditionality, Decision No. 12864-(02/102), as amended and Revised Operational Guidance to IMF Staff on the 2002 Conditionality Guidelines, July 23, 2014.


If the report is published, the press release could also summarize staff views set out in the update.


This clause was revised to clarify circumstances that would trigger the automatic termination of the PCI as follows: “Non-completion of a review for a twelve-month period would signify lack of Board endorsement on the member’s policies and result in an automatic termination of the PCI.” This change was reflected in the revised decision.


Under UFR arrangements, a waiver of non-observance for a PC is granted if the Fund is satisfied that the program will nevertheless be successfully implemented—i.e., that it will achieve its goals—either because of the minor or temporary nature of the nonobservance, or because of remedial actions taken by the authorities to preserve program objectives. Under the PSI, a formal request from the member is not required, instead the waiver is “granted” if the conditions noted above are met.


There are no waivers for misreporting in the context of the PSI misreporting framework. If the nonobservance was de minimis in nature, the Executive Board is informed of the misreporting in a staff report on a review.


In addition, while the misreporting framework under Article VIII, Section 5 does not apply to information provided to the Fund solely for the purposes of a PCI, these procedures would apply in the context of a PCI when such information is otherwise subject to Article VIII, Section 5.


The main factor influencing the precise marginal cost is the decision to open/maintain a Resident Representative posting, which would be at the discretion of the corresponding Area Department. Other factors include cost of travel and additional TA required to achieve program objectives.


As reflected in the revised decision, this recommendation was amended to allow for an earlier review, if warranted.


According to current TA policy, self-financing is expected only in the case of use by an advanced economy.


SM/09/213, n.2.

Adequacy of the Global Financial Safety Net--Proposal for a New Policy Coordination Instrument
Author: International Monetary Fund