100929 IDA17 Mid-Term Review Enhancing IDA’s Financial Support in IDA17 Proposal for a Scale-up Facility IDA Resource Mobilization Department (DFIRM) Development Finance (DFi) November 2, 2015 ACRONYMS AND ABBREVIATIONS ABCDQ Actions on Bunching, Commitments, Disbursements, and Quality AFF Agriculture, Fishing and Forestry AOB Absence of Objection CCSA Cross Cutting Solution Area CPIA Country Policy and Institutional Assessment CPF Country Partnership Framework CRW Crisis Response Window DEC Development Economics Vice Presidency DFi Development Finance Vice Presidency DFIRM Development Finance IDA Resource Mobilization DLP Debt Limits Policy DSA Debt Sustainability Analysis DSF Debt Sustainability Framework EDU Education E&M Energy and Mining ESIA Environmental and Social Impact Assessment FCS Fragile and Conflict-Affected State FIN Finance GNI Gross National Income GP Global Practice HIPC Heavily Indebted Poor Country HSS Health and Social Services I&T Industry and trade IBRD International Bank for Reconstruction and Development I&C Information & Communications IFL IBRD Flexible Loan IDA International Development Association IMD Investment Management Department IMF International Monetary Fund LDC Least Developed Country LIBOR London Interbank Offered Rate MDB Multilateral Development Bank MDRI Multilateral Debt Relief Initiative MFM Macro Fiscal Management Global Practice NCBP Non-Concessional Borrowing Policy OPCS Operations and Policy Services PER Public Expenditure Review PBA Performance-Based Allocation PDO Project Development Objectives PRGF Poverty Reduction and Growth Facility ROC Regional Operations Committee SCD Systematic Country Diagnostic SCL Single Currency Lending SDR Special Drawing Rights TRA Transportation TRE Treasury WSF Water, Sanitation and Flood Protection WBG World Bank Group Table of Contents EXECUTIVE SUMMARY ............................................................................................................................. i I. Introduction ............................................................................................................................................. 1 II. Implementation Arrangements for the Facility ....................................................................................... 3 A. Preliminary Demand Analysis .................................................................................................... 3 B. Implementation Arrangements .................................................................................................... 6 C. Illustrative Allocation and Prioritization Simulation ................................................................ 14 III. Financial Framework of the Facility ..................................................................................................... 17 A. Proposed Adjustment to Liquidity Management ...................................................................... 17 B. Proposed Lending Terms for the Facility.................................................................................. 19 C. Asset Liability Management Considerations ............................................................................ 23 IV. Issues for Discussion............................................................................................................................. 26 Annexes Annex 1. Indicative Demand for IDA17 Scale-up Facility: Projects Submitted by the Regions in July 2015 ........................................................................ 27 Annex 2. Scale-Up Facility Financial Risk Management .......................................................................... 29 Annex 3. Proposed Pricing Methodology and Indicative Rates ................................................................ 37 Boxes Box 1. NCBP Guidance on Non-Concessional Borrowing Decisions ................................................... 13 Box 2. Background on IDA’s liquidity management framework .......................................................... 17 Box 3. Proposed Lending Terms............................................................................................................ 21 Tables Table 1. IDA17 Scale-up Facility: Summary of Initial Regional Submissions ......................................... 3 Table 2. Regional allocations ..................................................................................................................... 9 Table 3. Regional allocations corrected for demand ................................................................................ 14 Table 4. Difference in Reflows between Non-concessional and Concessional Lending ......................... 23 Figures Figure 1. IDA17 Scale-up Facility: Distribution of Single-Country Projects............................................. 4 Figure 2. IDA17 Scale-up Facility: Sectoral Distribution of Single-Country Projects ............................... 4 Figure 3. IDA17 Scale-up Facility: Distribution of Single-Country Projects ............................................. 5 Figure 4A. IDA17 Scale-up Facility: Mapping of Potential Country Demand: Demand Volume, Policy and Instutional Environment and Risk of Debt Distress .................................................. 6 Figure 4B. IDA17 Scale-up Facility: Mapping of Potential Country Demand Demand Volume, Policy and Institutional Environment and Country Size ................................ 6 Figure 5. IDA17 Scale-Up Facility: Proposed Implementation Arrangements .......................................... 7 Figure 6A. IDA17 Scale-up Facility: Mapping of Possible Use of Facility By Income, Borrowing Category, FCS and LDC status ............................................................................... 15 Figure 6B. IDA17 Scale-up Facility: Mapping of Country Demand: Demand Volume, Policy and Institutional Environment and Risk of Debt Distress .............................................. 16 Figure 6C. IDA17 Scale-up Facility: Mapping of Country Demand: Demand Volume, Policy and Institutional Environment and Country Size ........................................................... 16 Figure 7. IDA's Projected Liquidity Buffer to 2030 ................................................................................. 18 Figure 8. Funding Needs from Tranche 1 under Liquidity Stress Scenario .............................................. 19 Figure 9. IDA17 Scale-up Facility Lending Costs vs. Estimated Non-concessional Loan Revenue........ 22 EXECUTIVE SUMMARY i. The World Bank Group (WBG) is exploring options to enhance its ability to respond to the ambition of the 2030 agenda for sustainable development. This echoes calls from the international community for MDBs to continue providing both concessional and non-concessional stable, long-term development finance, making optimal use of their resources and balance sheets. Moreover, the international community noted that addressing the infrastructure gap – estimated for developing countries at US$1.0 to US$1.5 trillion annually – and other strategically important investments will be key to overcome growth and poverty reduction bottlenecks. In this context, the IDA (International Development Association) partners affirmed the need for innovation and welcomed further exploration of options that best support IDA’s evolving client base in the context of the new more ambitious development paradigm. Recent developments point to very constrained financing for IDA clients during the IDA17 period – underscoring the urgency to identify additional innovative financing options. ii. To respond to these immediate pressures, this paper lays out a proposal to establish a facility to scale-up IDA’s financing for the remainder of the IDA17 period . The proposed facility would make available – on a one-off basis – up to US$5.0 billion to IDA countries. Of this amount, a contingent amount of up to US$900 million would be used to replenish the Crisis Response Window (CRW), which has already been depleted in response to the Ebola epidemic and the Nepal earthquake. The remaining US$4,100 million – plus any unused portion of the CRW resources – would be made available to IDA countries on non-concessional terms (similar to those from the International Bank for Reconstruction and Development – IBRD). These resources would be additional to the regular concessional allocations (core and non-core) to IDA-eligible countries in FY17. Submissions from the Bank’s regional units show that the indicative demand from IDA clients for projects that could be financed with non-concessional resources from the facility far exceeds its expected volume. iii. In addition to responding to IDA country demand, the facility may help provide valuable insights to the design features of a non-concessional facility financed by leveraging IDA’s balance sheet. However the nature of the proposed facility, in particular its funding, makes it a one-off facility whose design is not dependent upon the leveraging proposals being considered. iv. An adjustment to IDA’s liquidity management could provide an efficient and prudent approach to increase commitments by US$5.0 billion to fund the facility. This proposal would expand the current eligibility of liquid assets that can count towards IDA’s minimum liquidity buffer. A sensitivity analysis confirmed that as a result, IDA can release US$5.0 billion in internal resources that have been set aside in the investment portfolio and are currently uncommitted while meeting IDA’s liquidity objectives even under emergency liquidity scenarios. v. Non-concessional terms are proposed to ensure IDA’s financial sustainability and equitable treatment among IDA recipients, while maximizing the resources available to IDA countries. For IDA’s financial sustainability, the use of the facility’s resources should generate sufficient income to cover the associated costs, primarily from administrative fees and forgone investment income. In addition, through the facility, IDA would be able to reallocate internal resources held in its investment portfolio for the benefit of those few countries that are eligible. Charging non-concessional terms ensures that IDA earns reflows at levels similar to the investment income it would have earned through its investment portfolio thereby securing internal resources - ii - for the benefit of all IDA clients. In doing so, IDA can maximize available funding for the facility and for future replenishments. vi. The proposed implementation arrangements for the non-concessional resources from the facility are as follows:  Purpose. To provide non-concessional financing for projects with strong development impact, which could include investment project financing for both infrastructure and non- infrastructure projects as well as development policy and program for results operations and guarantees. The facility would reflect due consideration of individual countries’ debt situation, while ensuring consistency with the Non-Concessional Borrowing Policy (NCBP) and the IMF Debt Limits Policy (DLP).  Eligibility. Eligibility for support from the facility will be limited to projects in all current IDA clients that would be approved by the Board of Executive Directors before the end of the IDA17 period.  Allocation framework. While the nature of the non-concessional resources requires a flexible allocation framework, ideally it would draw as much as possible on links with the Performance-Based Allocation (PBA) system to ensure consistency with its performance and poverty orientation. The allocation framework would also need to focus on equity, performance, poverty and debt sustainability. Under the proposed framework, resources would be first allocated at the regional level. Within these regional allocations, resources would be allocated at the project level based on a process that would involve: a) prioritization of projects by the regions (using the criteria outlined in the next bullet); and b) the allocation of resources by DFi to prioritized projects confirming the proposed caps to address possible concentration and debt issues and ensuring consistency with the prioritization criteria). The calculation of the regional allocations and caps is detailed below. ○ Regional allocations: Each region will benefit from a share of the facility corresponding to its share of core IDA in FY16. Should a region’s demand for the facility be less than its allocated share, the available financing would be reallocated pro-rata to the regions with excess demand. ○ Country caps: Within the above regional allocations, the caps would help: (i) Strengthen the focus on performance, poverty and absorptive capacity. Assuming scalability of the projects, a country’s overall support will have a notional cap equal to 100 percent of a country’s FY16 allocation of core IDA. Greater flexibility could be considered for small countries given their relatively small core IDA allocations and their greater vulnerability to climate change risks. (ii) Address debt sustainability considerations. First, access to support from the facility will be such that when incorporated in the country’s debt sustainability analysis (DSA) it would not lead to a deterioration of the country’s risk of debt distress. Second, in countries that are not at low risk of debt distress, consideration will be given to limit access to support from the facility to a level that would still ensure that the country’s overall IDA envelope remains concessional (as defined under the NCBP). - iii - (iii) Additional considerations. For countries subject to IDA’s NCBP and the IMF DLP, consistency between the projects financed by the facility and these policies will be ensured (including with respect to possible established non-concessional borrowing ceilings).  Prioritization. Once regional allocations are determined, additional elements will help regions to prioritize the projects proposed for the facility. ○ Debt sustainability: Low risk countries would be the highest priority and moderate risk countries would be the next priority. In principle, high risk countries would not be eligible. However, since resources would be allocated at the project level, prioritization across projects will take into account both country-specific and loan specific factors. Furthermore, as for any non-concessional borrowing, consistency with the NCBP and the IMF DPL will be required for all countries subject to these policies, which includes IDA-only non-gap countries at moderate or high risk, as well as post-MDRI countries at low risk. Projects in high risk countries would not be eligible for the facility unless the projects are deemed consistent with the NCBP on the basis of high social and economic returns. Based on the regional submissions as part of the demand assessment and the proposed prioritization criteria, high risk countries are not expected to benefit from the facility. ○ Capacity: Countries with the capacity to absorb the resources well (as measured by a Country Policy and Institutional Assessment (CPIA) score above the median – i.e. 3.4 – for IDA countries or a combination of the CPIA along with portfolio performance) will be prioritized. ○ Consistency with IDA17 policy priorities and the WBG goals: Regions will prioritize projects that are aligned with the WBG goals and best promote the IDA17 policy priorities (e.g., climate change, fragility and conflict, regional integration).  Review process. The proposed review process aims at ensuring consistency of the projects financed by the facility with the WBG goals, the IDA17 priorities, the beneficiary’s debt sustainability and the non-concessional borrowing policies in place. It will entail three stages focused primarily on the projects in countries that are not at low risk of debt distress: (i) a high level senior management overview meeting to take place shortly after the Board of Executive Directors’ approval of the facility to provide an early forum to discuss each region’s selected project pipeline; (ii) a project review stage (concept stage and ROC/OC stage) following regular Bank project review processes, to ensure that funding from the facility is granted only to projects where there is a clear rationale and where it will not exacerbate the country’s debt outlook; and (iii) a review meeting in the first quarter of FY17 and subsequent periodic reviews of implementation progress as needed which could result in the reallocation of funding to different projects.  Role of the Board of Executive Directors. Upon agreement by the Deputies, implementation of the facility would require approval by the Board of Executive Directors. In addition, the facility will provide for a strong oversight role by the Executive Directors during the implementation period. Upon the Board’s approval of the changes to IDA’s commitment authority and the terms being made available to recipients to establish the facility, projects financed with the facility’s resources would be sent to the Board of - iv - Executive Directors for approval. In addition, Management will report on the implementation experience shortly after the conclusion of the IDA17 period. vii. The implementation arrangements for the resources transferred to the CRW will be the same as those agreed in the context of the IDA17 replenishment discussions. In addition, the portion of the resources transferred to the CRW that would remain unused by end-December 2016 would be reprogramed to stand-by projects on non-concessional terms using the allocation and prioritization criteria outlined above. viii. If supported by Deputies at the IDA17 MTR, the facility would be launched as soon as possible after Board approval. I. Introduction 1. The international community has called upon the multilateral development banks (MDBs) to explore innovative approaches to support the 2030 agenda for sustainable development.1 Participants at the Addis Ababa Financing for Development Conference invited the MDBs to “continue providing both concessional and non-concessional stable, long-term development finance,” making “optimal use of their resources and balance sheets.” 2 They also urged the MDBs to “update and develop their policies in support of the post-2015 development agenda including the sustainable development goals.” Moreover, they noted that “development banks can play a particularly important role in alleviating constraints on infrastructure investment.” Addressing the infrastructure gap – estimated for developing countries at US$1.0 to US$1.5 trillion annually – and other strategically important investments will be key to overcome growth and poverty reduction bottlenecks. In response, the World Bank Group (WBG) is exploring innovations to enhance its capacity to meet the growing client demand for greater volumes of all types of finance, including concessional and non-concessional financing, equity investments and guarantees. In discussions over 2015, International Development Association (IDA) partners affirmed the need for innovation and welcomed further exploration of options to best support IDA’s evolving client base in the context of the new more ambitious development paradigm. 2. As the main instrument for achieving the WBG goals in the poorest countries, IDA is well-suited for making a significant and catalytic contribution. IDA’s operational and financial model – shaped over time by a global coalition of partners and clients – provides a critical platform to enhance aid effectiveness and is uniquely positioned to help clients maximize the impact of global knowledge, financing and partnerships. As detailed in the bullets below, recent developments point to very constrained financing during the IDA17 period; underscoring the urgency to identify additional innovative financing options as soon as possible.  On the demand side, current client demand for IDA financing is stronger than ever: IDA funding approved over FY15 reached over US$19 billion, a record for a first year of a replenishment. Looking ahead, when asked to identify possible demand for additional non- concessional resources over the last half of IDA17, regions responded with requests for financing of about $16 billion. Of the requests, over 90 percent were for infrastructure projects.  On the supply side, IDA’s ability to support its clients is constrained. Less than 60 percent of core IDA17 remain for the final two years of the replenishment, and several set-asides are under severe pressure.3 In addition, on the non-concessional front, constrained access to MDB funding leaves IDA clients trying to finance large infrastructure gaps with difficult choices, such as drawing down limited concessional resources (with significant 1 See IDA/SecM2014-0071, From Billions to Trillions – Transforming Development Finance. Post-2015 Financing for Development – Multilateral Development Finance. Development Committee Meeting, April 18, 2015. 2 Outcome Document of the Third International Conference on Financing for Development: Addis Ababa Action Agenda, July 15, 2015 (Addis Outcome Document), paragraphs 70 and 75. This priority issue has consistently been highlighted by leaders of developing countries and the G20. 3 The Ebola epidemic and the Nepal earthquake have already exhausted the IDA17 Crisis Response Window (CRW). Heavy demand for the Regional Program has already utilized 60 percent of the available funds in the Africa region, and over 50 percent of funding for all regions in FY15 alone. Regional funding is expected to be almost fully exhausted by the end of FY16. -2- opportunity costs for financing priority health or education investments) or seeking more expensive private commercial borrowing. 3. Given these pressures, Management has been exploring options to enhance IDA’s financing capacity. These options include the leveraging of IDA’s balance sheet in the longer term to create additional lending capacity for the 2030 agenda, as well as options to more efficiently use IDA’s liquidity to free up additional resources in the short term. These options were discussed in the context of the IDA17 Working Group on IDA’s Long term Vision and Financial sustainability. At the 2015 Spring Meetings, participants in this working group expressed broad support for further exploring a longer term leveraging facility. In addition, there was support for Management to present a proposal for a shorter-term facility financed through IDA’s liquidity to address immediate financing pressures. 4. In line with the above guidance and that of the IDA Executive Board, this paper lays out a proposal to establish a short term facility to scale-up financing for the remainder of the IDA17 period. The proposed facility would make available – on a one-off basis – up to US$5.0 billion to IDA countries on non-concessional terms (similar to those from the International Bank for Reconstruction and Development – IBRD). Its design would provide insights for the design of the longer term leveraging proposal being considered for IDA18, but is not dependent on it. As detailed in the paper, these resources would be freed-up through a more efficient use of IDA’s liquidity. These resources would be in addition to the regular concessional resources (core and non-core) that countries would receive in IDA17, and would be channeled to projects with strong returns and, as detailed below, to replenish the CRW. 5. Given the depletion of CRW resources already in the first year of IDA17, a contingent allocation of up to US$900 million from the facility is proposed to be made available if needed, to replenish the CRW. Providing a contingent commitment of up to US$900 million from the facility for the CRW will help to ensure that IDA is able to provide rapid response in the event of further crises over the remainder of the IDA17 period, while not affecting core concessional IDA support for IDA borrowers in the last year of the replenishment.4 Given the unpredictable nature of the crises and the possibility that the amount may not be fully used, any unused resources by end-December 2016 could be reprogrammed to stand-by projects in non- concessional terms using the implementation framework detailed below. 6. Implementing the facility would provide valuable insights that could help inform the design and implementation of a longer-term non-concessional facility in IDA18. In particular, regional teams and beneficiaries would have the opportunity to gather experience with a non- concessional product from IDA. Implementing the facility would also provide valuable insights to improve understanding of the catalytic potential of such a non-concessional facility to attract other sources of finance (domestic, private) for sustainable development. 7. Structure of the paper. Informed by a careful assessment of potential demand, Section II of this paper covers the operational aspects of the facility, including purpose, eligibility and allocation criteria, prioritization criteria, and review and oversight modalities. Section III elaborates on the financial aspects of the facility; in particular it describes how the financing for 4 The use of these resources would be guided by the implementation arrangements agreed for the CRW in the context of the IDA17 replenishment discussions. -3- the facility would be mobilized and address risks and tradeoffs issues. Section IV presents issues for discussion. II. Implementation Arrangements for the Facility 8. This section sets out the implementation arrangements for the IDA17 Scale-up Facility. As detailed in Section III, changes to IDA’s liquidity management approach could free- up to US$5.0 billion. These resources would be lent to IDA countries on non-concessional terms during the second half of the IDA17 period, guided by the implementation arrangements set out below. The facility is designed to respond to client demand and to address the risks associated with providing non-concessional funding in the heterogeneous context of IDA countries and to ensure that these resources are used as efficiently as possible. To address these issues, this section presents: (i) an analysis of the potential demand for non-concessional financing from IDA for the second half of the IDA17 period; (ii) the proposed implementation arrangements for the facility; and (iii) an illustrative scenario for the allocation of financing from the facility based on the proposed implementation arrangements. A. Preliminary Demand Analysis5 9. The potential demand for non-concessional financing from the facility far exceeds its expected size (see Table 1 and Annex 1). Based on initial feedback from the Bank regional units, the demand for financing from the facility in IDA-eligible countries amounts to US$15.7 billion (i.e., more than three times the facility’s expected size). This includes 81 projects for a total of US$15.4 billion.6 Of this amount, US$14.1 billion correspond to single-country projects and US$1.3 billion would involve regional projects. At the project level, the financing ranges from US$5.0 million to US$1.0 billion. As detailed in the table below, in volume terms, almost three- quarters of the projects are in AFR, followed by SAR (19 percent) and EAP, ECA and LCR (each below 3 percent).7 Table 1. IDA17 Scale-up Facility: Summary of Initial Regional Submissions (project-specific requests for financing) Number of Volume Project Distribution Countries 2/ Count 1/ (US$ million) By projects By volume AFR 19 57 11,664 70% 76% EAP 1 1 50 1% 0% ECA 4 7 455 9% 3% LCR 3 9 270 11% 2% MNA --- --- --- --- --- SAR 2 7 3,000 9% 19% Overall total 29 81 15,439 1/ Country count reflects only single-country projects (regional projects are excluded). 2/ In some cases, the regional submissions bids included funding ranges. For such cases, the table reports the mid-point of the range. 5 The assessment reflects inputs from the Bank’s Regional Units. 6 This figure excludes US$0.3 billion of demand unassigned to specific projects. These figures are excluded from the analysis hereafter. 7 No projects were submitted by MNA, which only includes 2 IDA-eligible countries. -4- 10. The bulk of the requested financing is for projects that would benefit the poorest IDA countries and/or those that have limited financing options (see Figure 1). Almost three- quarters of the demand corresponds to projects in IDA-only countries, the vast majority of which would benefit non-gap countries (i.e., those with lower Gross National Income – GNI – per- capita).8 Only about a quarter of the project volume would benefit blend countries (i.e., with limited access to IBRD financing). In addition, about 20 percent of the requested financing would benefit Fragile and Conflict-affected States (FCSs), while the Least Developed Countries (LDCs) would benefit from 49 percent of the total financing. Figure 1. IDA17 Scale-up Facility: Distribution of Single-Country Projects (by volume) 8.0 52% 48% 49% 49% 6.0 (US$ billion) 4.0 25% 26% 19% 2.0 12 17 16 8 5 3 6 16 0.0 LICs LMICs Non-Gap Gap Blend FCSs LDCs By Income By IDA Borrowing Category Note: Figures inside the bars reflect the number of beneficiary countries in each category. 11. At the sectoral level, the data shows a strong preference for using the facility for infrastructure lending (see Figure 2). More than 90 percent of the demand for single-country projects is for projects in the energy and mining, transportation and water, sanitation and flood protection sectors. The assessment shows that there is also demand for financing in other sectors (including industry and trade, agriculture, fishing and forestry and finance). In aggregate terms, however, these sectors represent less than 10 percent of the overall volume. An even stronger focus on infrastructure was evidenced on the demand for financial support for regional projects. Figure 2. IDA17 Scale-up Facility: Sectoral Distribution of Single-Country Projects (by volume) 8.0 E&M: energy and mining 41% TRA: transportation 38% WSF: water, sanitation and flood (US$ billion) 6.0 protection 4.0 I&T: industry and trade AFF: agriculture, fishing and forestry 12% FIN: finance 2.0 3% 3% EDU: education 2% 1% 0% 0% 0.0 I&C: information and communications E&M TRA WSF I&T AFF FIN EDU I&C HSS HSS: health and social services (Sector) 12. Implementing the facility would require continued attention to debt sustainability considerations. Of the 29 countries for which projects were submitted, 12 are at low risk of debt distress; the remaining have been assessed at moderate (14) or high (3) risk of debt distress. 8 IDA-only gap countries are those for which their GNI per-capita has remained above the IDA operational cutoff for at least three consecutive years but are not creditworthy for IBRD borrowing. -5- Projects for countries at low risk are more numerous and larger; while the opposite is the case in high risk countries. As a result, in volume terms, more than half of the demand corresponds to countries at low risk; while those at high risk represent under one percent (see Figure 3). While most of the demand is from countries at low risk of debt distress, a cautious approach would be needed for the provision of financing from the facility as a number of the potential beneficiaries have in the past benefitted from debt relief and/or debt restructurings. Finally, of the 29 countries, 17 are subject to the Non Concessional Borrowing Policy (NCBP) – including one to which NCBP remedies have been applied – and/or the International Monetary Fund (IMF) Debt Limits Policy (DLP). These countries account for US$9.1 billion of the demand. Figure 3. IDA17 Scale-up Facility: Distribution of Potential Demand for Single-Country Projects (by risk of debt distress) 8.0 51% 48% 6.0 (US$ billion) 4.0 2.0 1% 0.0 Low Moderate High Risk of Debt Distress 13. The analysis of potential demand also suggests considering implementation capacity as part of the facility’s operational arrangements. For several countries, including most of those at moderate or high risk of debt distress, the aggregate demand is larger than the country’s FY16 core IDA allocation (see Figure 4A). It should be noted, however, that about two-thirds of the countries for which the facility demand is multiple times (i.e., twice or more) the FY16 core IDA allocation have relatively smaller populations (see Figure 4B).9 While core IDA allocations to these countries is relatively high in per-capita terms (reflecting the minimum base allocation), the overall core IDA support is constrained by the small population base.10 Also, 10 out of the 29 countries for which projects were submitted have a relatively weak policy and institutional environment.11 Of these 10 countries, more than two-thirds have a share of problem projects above the Bank average or an undisbursed balance ratio below the Bank average. 9 For the purpose of this analysis, a country with small population is defined as a country with a population below 6.5 million (corresponding to the first quartile limit for IDA countries excluding small states – i.e., countries with a population below 1.5 million). Countries with a medium population are those with less than 26.6 million (the third quartile limit). The remaining are considered countries with large population. 10 Based on the definition above, the FY16 allocation for medium-sized countries was about 5 times that for small countries and that for large countries 5 times that for medium-sized countries. 11 Defined, for the purpose of this analysis, as having a CPIA score lower than the median for IDA countries. Nine additional countries have a CPIA score in the third quartile. Only 10 of the 29 countries are among the top IDA performers (i.e., have a CPIA score in the fourth quartile). -6- Figure 4A. IDA17 Scale-up Facility: Mapping of Potential Country Demand Demand Volume, Policy and Institutional Environment and Risk of Debt Distress 100 Demand as share of FY16 core IDA allocation Risk of debt distress: Low Moderate High 10 (logarithmic scale) 1 0.1 2.0 2.5 3.0 3.5 4.0 4.5 (CPIA score) Figure 4B. IDA17 Scale-up Facility: Mapping of Potential Country Demand Demand Volume, Policy and Institutional Environment and Country Size 100 Demand as share of FY16 core IDA allocation Country size: Small Medium Large 10 (logarithmic scale) 1 0.1 2.0 2.5 3.0 3.5 4.0 4.5 (CPIA score) Note: Markers below the solid black line represent countries for which the demand is below the FY16 core IDA allocation. Markers above the dotted black line correspond to countries for which the demand is more than twice the FY16 core IDA allocation. The solid red line represent the median CPIA score for IDA countries (3.4). The size of the bubbles represent the volume of the demand. B. Implementation Arrangements 14. This sub-section outlines the proposed implementation arrangements for the facility. The proposed arrangements aim to strike a balance between rules and flexibility which, coupled with robust review process, would ensure the optimal use of resources from the facility and follows IDA’s country-driven model, developmental and debt sustainability perspective. The proposed implementation arrangements (see Figure 5) include the facility’s: (i) purpose; (ii) eligibility criteria; (iii) resource allocation framework; (iv) project prioritization criteria; and (v) review process. As detailed in the following paragraphs, the allocation of resources will be managed at the corporate level by DFi – in collaboration with OPCS and MFM, prioritization of projects will -7- be decentralized to the Bank regional units and financing decisions will be subject to review and oversight by Management and the Board of Executive Directors. Figure 5. IDA17 Scale-Up Facility: Proposed Implementation Arrangements 15. Purpose. The purpose of the facility will be to provide non-concessional financing for projects with strong development impact, which could include investment project financing – for both infrastructure and non-infrastructure projects – as well as development policy and program for results financing (DPO and PforR, respectively) and guarantees.12 The facility would reflect due consideration of individual countries’ debt situation, while ensuring consistency with the NCBP and the IMF DLP.  With respect to the targeting of investment financing to specific sector(s), the recommendation is that resources are made available for projects in any sector. This recommendation acknowledges that both infrastructure and non-infrastructure projects can have strong returns to investment, development impact and growth dividends.13 With respect to infrastructure projects, well-planned infrastructure investments could unlock 12 Given the short time frame of the facility, scaling up successful interventions could be an effective use of the facility for some countries. 13 Similar to an enclave-like approach, using the facility for infrastructure financing would help align consistency with the non- concessional nature of the resources while not being as restrictive as the criteria for an IBRD enclave project. -8- bottlenecks to development and might be expected to have strong returns to investment and growth dividend. With respect to non-infrastructure projects, investments in Early Childhood Development – for example – can have significant rates of return of around 7 to 10 percent per year, and the returns to these and other social investments can make present and future investments in other sectors more productive due to dynamic complementarities.14 Furthermore, while infrastructure projects account for the lion’s share of bids, there was also demand for a number of other sectors.15 Leaving the facility open to infrastructure and non-infrastructure projects alike would enable the flexibility in the use of the resources to maximize the impact of financing, while still resulting in overall financing that is skewed to infrastructure. The alternative of limiting the resources available to infrastructure projects only was dropped in light of the above considerations.  With respect to the recommendation for including DPOs, PforRs and guarantees, while the preliminary demand received from regions was focused on infrastructure investments, all instruments could have strong returns. For instance, a guarantee is an increasingly important instrument for crowding in private sector investments (as in the case of Dasu hydropower project in Pakistan and the Sankofa gas project in Ghana).16 DPOs, and more recently PforRs, have been powerful instruments for Bank support of development results. For instance DPOs have successfully been deployed to support environment and natural resource management programs in Africa, which had a number of impacts including improved mining and forestry sector revenue collection, management and transparency. This inclusive approach is consistent with the emphasis of the 2030 agenda on the role of various instruments to be mutually interdependent in their design in order to maximize development impact. 16. Eligibility criteria. Given the significant demand, and the limited volume of financing from the facility, eligibility would be limited to projects in all current IDA clients that would be approved by the Board of Executive Directors before the end of the IDA17 period.17 While there was also demand from recent IDA graduates, this project facility will be geared towards current IDA clients who tend to have more limited options for non-concessional financing than graduates, and who partners have committed to support under the IDA17 framework. While all current IDA clients would be eligible, the actual selection of projects will be subject to allocation and prioritization criteria detailed below. 17. Allocation framework. Allocation decisions for the facility need to take into account the non-concessional nature of the available resources. This requires a flexible allocation, while ideally drawing as much as possible on links with the Performance-Based Allocation (PBA) to ensure consistency with the PBA’s performance and poverty orientation. The resource allocation framework will incorporate two stages. As described below, in the first stage, allocations will be 14 See “The Rate of Return to the High Scope Perry Preschool Program”, Heckman, J. J., Moon S. H., Pinto R., Savelyev P. A., and Yavitz A. Q. (February, 2010) and “Investing in Our Young People”, Cunha, F., Heckman, J. J. (July, 2010). 15 Projects in the finance, industry and trade, agriculture, fishing and forestry, and education sectors accounted for about 10 percent of project demand volumes. 16 The Dasu Hydropower Stage I was approved in May 2014 to expand electricity supply of hydropower in Pakistan. For this project, IDA provided a US$460 million public sector loan guarantee to back US$2.4 billion in private financing. The Sankofa Gas Project was approved in July 2015 to generate a new supply of natural gas for 1000 megawatts of clean, reliable and affordable electricity for domestic use in Ghana. IDA provided a US$500 million payment guarantee, complementing a US$200 million IBRD Loan Guarantee; these guarantees backed a total US$7.9 billion in private financing. 17 This excludes India which receives financing from IDA on an exceptional basis in IDA17 through transitional support. -9- determined at the regional level. In the second stage, within these regional allocations, resources will be allocated at the project level. The allocations per region outlined below as well as the concentration limits and limits to address debt sustainability considerations will be managed at the corporate level by DFi in collaboration with OPCS and MFM.  Regional allocations: Given significant excess demand, and in line with benefitting all countries, a share of the Facility will be allocated to each region. Each region’s share will correspond to its share of core IDA in FY16. Table 2 shows the corresponding regional shares scaled to the expected size of the facility (US$5.0 billion). Should a region’s demand for the Facility be less than their allocated share, the available financing would be reallocated pro-rata to the regions with excess demand. Table 2. Regional allocations (US$ billion) Region Based on FY16 allocation AFR 2.81 EAP 0.75 ECA 0.17 LCR 0.10 MNA 0.06 SAR 1.13 Total 5.00  Country caps: As detailed below, within the above regional allocations, additional criteria would help: (i) strengthen the focus of the facility on country performance, poverty and absorptive capacity; and (ii) address debt sustainability considerations. The allocation of the resources to specific projects will made by DFi based on the list of prioritized projects prepared by the regions (see paragraph 18) after application of the caps in consultation with OPCS and MFM and confirmation of consistency with the prioritization criteria. ○ Focus on performance, poverty and absorptive capacity. Assuming scalability of the prioritized projects, a country’s overall support from the facility will have a notional cap equal to 100 percent of a country’s allocation of core IDA (which reflects the country’s policy and institutional environment – as measured by the CPIA – as well as its poverty – as measured by the GNI per capita), or multiple thereof. 18 Greater flexibility could be considered for small countries given their relatively small core IDA allocations and greater vulnerability to climate change risks.19 18 Where projects are not scalable, additional flexibility could be discussed during the Senior Management Review, similar to other projects that are not fully in line with the prioritization criteria. As with all projects under the Scale-up Facility, the project would need to be consistent with the NCBP. 19 Small island countries are highly vulnerable to recurring natural disasters and at higher risk of climate change impact. As a result, most of these countries need significant financial support to create growth enabling infrastructure. In light of this, a notional cap of 200 percent could be considered for small countries while that for all other IDA countries could be 100 percent. - 10 - ○ Addressing debt sustainability considerations. Debt sustainability considerations will be incorporated in the allocation process in two steps. First, access to support from the facility will be such that, when incorporated in the country’s debt sustainability analysis (DSA), the project would not lead to a deterioration of the country’s risk of debt distress. Second, in countries that are not at low risk of debt distress, consideration will be given to limit access to support from the facility to a level that would still ensure that the country’s overall IDA envelope remains concessional (as defined in the context of the NCBP).20 ○ Additional considerations. For countries subject to IDA’s NCBP and the IMF DLP, consistency between the projects financed by the facility and these policies will need to be ensured (including with respect to possible established non-concessional borrowing ceilings). The set of criteria that is assessed in determining consistency with the NCBP is outlined in Box 1, and includes both country-specific and loan specific factors. In designing the above allocation framework, a number of options were considered but ultimately ruled out. One option considered was to take a first-come, first-served approach for individual projects, similar to the approach for access to CRW support. This approach was ruled out as it could potentially lead to a heavy concentration of lending in one region or in a few countries with high demand for non-concessional financing. Another option was to allocate the facility’s support to all IDA countries along PBA lines. Given the limited size of the facility, this approach would end up spreading the benefits of the facility thinly and to countries that may not have projects for which they would want to use non concessional resources at this time, or do not have the capacity to absorb non-concessional resources. 18. Prioritization. Once regional allocations are determined, additional elements would be needed to prioritize the projects financed under the facility. Within the overall regional envelope, each regional VPU would be responsible for prioritization of eligible projects based on country demand within its respective region, in close consultation with Global Practices (GPs) to maximize use of country knowledge and linkages, and guided by a set of criteria as outlined below. A central review process would help guide the regions’ decision. In prioritizing projects, regions would look to three key criteria: debt sustainability, capacity, and consistency with the WBG goals and IDA17 policy priorities.  Debt sustainability: The first criterion would be a country’s risk of debt distress, i.e., whether this is low, moderate, or high. Low risk countries would naturally be the highest priority and moderate risk countries would be the next priority. In principle, high risk countries would not be eligible. However, since this facility is not proposing to allocate resources at the country level, but rather at the project level, prioritization across projects will take into account both country-specific and loan specific factors. Box 1 presents the elements that have guided the assessments on the appropriateness of non-concessional borrowing performed under the NCBP; and have proven useful guidelines in the context of IDA countries, including those in high risk of debt distress. These criteria will be applied by the region in their prioritization process. Furthermore, as for any non-concessional borrowing, consistency with the IMF’s DPL and IDA’s NCBP will be required for all 20 For countries that are not at low risk of debt distress, IDA allocates grant resources (50 percent and 100 percent of allocations for moderate and high risk, respectively) to help countries maintain debt sustainability. - 11 - countries subject to these policies, which includes IDA-only non-gap countries at moderate or high risk, as well as post-MDRI countries at low risk. Projects in high risk countries would be excluded from eligibility for the facility unless the projects are deemed consistent with the NCBP on the basis of high social and economic returns. Paragraph 17 elaborates on the review process that will ensure that this filter is adequately reflected in project allocation decisions. Based on the regional submissions as part of the demand assessment and the proposed prioritization criteria, however, high risk countries are not expected to benefit from the facility.  Capacity: The second criterion would be a country’s capacity to absorb the non- concessional resources well. This will be measured by a CPIA score above the median level (3.4), or a combination of the CPIA along with portfolio performance. Placing a high priority on countries with high capacity is in line with aid effectiveness research that links capacity with the ability to use resources most efficiently. It also helps to align the facility with the performance-based nature of core IDA resources.  Consistency with the WBG goals and IDA17 policy priorities: this criterion calls for prioritizing the projects that are better aligned or best promote the IDA17 policy priorities including the overarching theme and the special themes. For example, priority would be given to projects with the largest climate change and disaster risk management co- benefits,21 or those that address the drivers or fragility and conflict, or have the potential to address regional integration or crowd-in additional public or private financing.22 Furthermore, the selected projects will need to be consistent with the Systematic Country Diagnostic (SCD) and the Country Partnership Framework (CPF), thus ensuring alignment with the promotion of the WBG goals. 19. Review process. A strong and clear review process is essential for ensuring robust and effective use of the facility. This would entail three stages focused primarily on projects in countries that are not at low risk of debt distress:  Overview meeting: First, a high level senior management meeting would be convened shortly after the Board of Executive Director’s approval of the facility. The list of projects being considered for support under the facility could be agreed at this meeting with a focus on those that are not at low risk of debt distress. A possible forum for this could be the regular ABCDQ meeting which has broad representation of GPs, Cross Cutting Solution Areas (CCSAs) and regional units. This meeting would provide an early forum to discuss each region’s selected pipeline of projects. At this step, the initial demand from the regions will be narrowed down based on the allocation and prioritization criteria described above.  Project Review: A second stage would be at the time of project review, in line with the due diligence done for all IDA projects under preparation. Where a project is being proposed for financing under the facility that is not fully in line with the prioritization filters (e.g., projects with high developmental potential but where the country’s risk of debt distress is not low), this would be flagged for close review at the concept note and decision 21 These benefits would be estimated by the Climate Change GP (CCG) using the methodology recently developed and implemented to measure a project’s co-benefits. 22 This last point would also ensure alignment with the Addis Ababa Action Agenda which highlights the role of MDBs to catalyze additional sources of financing. - 12 - note meetings of the ROC/OC, or in advance of ROC/OC stage through a committee such as the NCBP committee. As per the NCBP, where financing under the facility is proposed in a country subject to the policy, the NCBP committee would discuss the merits as per the country-specific and loan-specific criteria under the policy, and make a recommendation to management on consistency of the planned project with the NCBP. This same committee, which consists of members of DFi, MFM, OPCS, DEC, Legal, and the relevant regional unit, could also be convened to discuss the merits of a project even if the country is not subject to the NCBP policy, for instance in a moderate risk of debt distress country, with low capacity, where there would be an additional level of assessment to determine debt sustainability implications (see Box 1). At this stage of the review, other key factors would also be considered including consistency with the SCD and the CPF, and consistency with IDA17 policy priorities, including the special themes. In addition to the Bank’s usual due diligence, the objective of this stage for the facility is to ensure that funding is granted only to projects where there is a clear rationale and where it will not exacerbate the country’s debt outlook.  Reallocation decision meeting: A third stage could take place in the first quarter of FY17 and periodically as needed (e.g., if at end-December 2016 unused CRW resources were to be reprogrammed for project financing on non-concessional terms), to take stock of progress of all pending projects, and their likelihood of being presented to the Board of Executive Directors in FY17. Regions would ensure a buffer of stand-by projects that could be supported under the facility and could be substituted at the time of the reallocation decision meeting. 20. Role of the Board of Executive Directors. Upon agreement by the Deputies, implementation of the facility would require approval by the Board of Executive Directors. In addition, the facility will provide for a strong oversight role by the Executive Directors during the implementation period. Upon Board approval of the facility, as with all IDA projects, projects financed with resources from the facility would be sent to the IDA Executive Directors for approval. In addition, Management will report on the implementation experience of the Scale-up Facility shortly after the conclusion of the IDA17 period. 21. Financing terms. Section III below outlines a set of potential financing terms that would be available under the facility. The terms available range from fixed rate to floating rate loans, and the currencies being offered are SDR, USD, Euro, GBP and JPY. These aim to be simple but offer additional flexibility to benefit borrowers where possible. Given that a number of IDA clients have little experience with managing floating rate loans, the option of floating rate loans would be reserved for those clients at low risk of debt distress and with adequate debt management capacity - 13 - as measured through the recently updated joint Bank-Fund capacity assessments performed in the context of the NCBP and the IMF DLP.23 Box 1: NCBP Guidance on Non-Concessional Borrowing Decisions The NCBP has relied on the following country-specific and loan specific criteria to guide decisions under the policy. This set of guidelines will be drawn on in making decisions on the appropriateness of accessing resources from the facility for any projects proposed in countries subject to the NCBP as well as any countries that are not at low-risk of debt distress. Country-specific:  Overall borrowing plans of the country . A modest level of overall borrowing by the country on the basis of the DSA to accommodate a particular investment may warrant consideration. For such a consideration, clear reporting of overall borrowing plans is needed, and enhanced creditor coordination through the Debt Sustainability Framework (DSF) would facilitate this possibility.  Impact of borrowing on the macroeconomic framework. Whether or not the borrowing would have a deleterious effect on the macroeconomic framework would be a key consideration.  Impact on the risk of debt distress. The current risk classification and whether or not the loan is likely to lead to a higher risk of debt distress will be a key consideration. More flexibility is envisaged for low risk countries given the more benign debt outlook and generally better performance. In addition, moderate risk countries could benefit from somewhat greater (although still exceptional) flexibility than high risk ones. Strength of policies and institutions, especially public expenditure management and debt management. As the fiscal space Board paper makes clear, policies and institutions in particular those governing the efficiency of public investment are critical. Without these, even high return projects may fail to meet objectives. Loan-specific:  Development content and potential impact of the loan, i.e., investment will unlock a proven bottleneck to development as determined by analytical work such as a PER.  Estimated economic, financial and social returns to investment of the project, weighted by the probability that the project will succeed.  Lender equity stake in the project.  No additional costs associated with the loan, i.e., collateralization, hidden costs.  No other sources of more concessional financing are available.  Concessionality of the overall financing package for a particular investment. 23 See “Review and Update of IDA’s Non-Concessional Borrowing Policy (NCBP)”, World Bank, October 2015 and “Reform of the Policy on Public Debt Limits in Fund-Supported Programs”, IMF, December 2014. - 14 - C. Illustrative Allocation and Prioritization Simulation 22. This section outlines the potential resource distribution of an illustrative scenario assuming US$5.0 billion in non-concessional financing. This illustrative scenario should not be read as reflecting the final decisions as these would require careful prioritization by the regional units – in coordination with the GPs –based on their detailed knowledge of the countries and projects under consideration. Rather this indicates how the resources could be used along a particular decision path based on the proposed allocation framework, and prioritization decisions. 23. Allocation Framework for simulation:  Regional allocation. Initial regional allocations – based on the share of each region within the core IDA in FY16 – was determined. In a second step, for the purpose of this simulation, these initial allocations were adjusted (as described in paragraph 15) to reflect the inputs provided by the regions to the demand assessment (see Table 3).24 Table 3. Regional allocations corrected for demand (US$ billion) Region Based on FY16 allocation Adjusted AFR 2.81 3.30 EAP 0.75 0.10 ECA 0.17 0.20 LCR 0.10 0.10 MNA 0.06 0.00 SAR 1.13 1.30 Total 5.00 5.00  Performance, poverty and absorptive capacity. Assuming full scalability, for the purpose of this simulation, countries for which projects were submitted were assumed to access resources from the facility up to a maximum of 100 percent of the country’s FY16 allocation, with greater flexibility for small countries.25  Debt sustainability. For this illustrative scenario, limits on debt sustainability were not binding once the projects were filtered down using the performance, poverty and absorptive capacity filters in this simulation. 24. Prioritization rules for simulation:  Low risk of debt distress countries were prioritized, followed by moderate risk of debt distress. 24 The resulting adjusted figures in Table 3 are illustrative only. As discussed in the demand assessment, some regions showed limited or no demand for scale-up support at this point. The adjusted figures reflect the limited demand in some regions, relative to their core IDA-based shares, at this point. With greater clarity on the financing terms, additional demand may still be forthcoming in these regions, which will be reflected in the final allocation outcome. 25 For the purpose of this simulation, a small country would be one with a population below 6.5 million. The FY16 core IDA allocation to these countries was about US$33 million, on average. - 15 -  Within these debt distress ratings, countries with high capacity (CPIA above the median for IDA countries – 3.4) were prioritized; in addition, among low capacity countries, only those with low risk of debt distress were selected. 25. Indicative simulation results: Figures 6A, 6B and 6C show mappings of the illustrative allocations resulting from the strict application of the options outlined in paragraphs 21 and 22. In total, 19 countries would benefit from financing from the facility; of these, 12 are at low risk of debt distress and 7 at moderate risk of debt distress. Figure 6A shows that as illustrated in Figure 6B, the majority of the financing would be for low risk of debt distress countries with high capacity. It should be noted that financing for low risk countries is, in general, larger than that moderate risk ones and financing for high capacity countries larger than the few cases with low capacity. Finally, as illustrated in Figure 6C, only small countries (i.e., countries with small core IDA allocations) have large allocations relative to their PBA-based allocation; and all those cases have been assessed at low risk of debt distress. Simulation results indicate that the levels of non- concessional borrowing from the facility would only have a marginal impact on the debt outlook of the beneficiary countries under this illustrative scenario.26 Figure 6A. IDA17 Scale-up Facility: Mapping of Possible Use of Facility By Income, Borrowing Category, FCS and LDC status Note: figures inside the chart reflect indicative allocation and prioritization results from simulations, by country groupings 26 The impact was measured in terms of solvency (present value of debt-to-export and present value of debt-to-GDP ratios) and liquidity (debt service-to-revenue ratio). - 16 - Figure 6B. IDA17 Scale-up Facility: Mapping of Country Demand Demand Volume, Policy and Institutional Environment and Risk of Debt Distress 100 Demand as share of FY16 core IDA allocation Risk of debt distress: Low Moderate High (logarithmic scale) 10 1 0.1 2.0 2.5 3.0 3.5 4.0 4.5 (CPIA score) Figure 6C. IDA17 Scale-up Facility: Mapping of Country Demand Demand Volume, Policy and Institutional Environment and Country Size 100 Demand as share of FY16 core IDA allocation Country size: Small Medium Large 10 (logarithmic scale) 1 0.1 2.0 2.5 3.0 3.5 4.0 4.5 (CPIA score) Note: Markers below the solid black line represent countries for which the demand is below the FY16 core IDA allocation. Markers above the dotted black line correspond to countries for which the demand is more than twice the FY16 core IDA allocation. The solid red line represent the median CPIA score for IDA countries (3.4). The size of the bubbles represent the illustrative allocations. - 17 - III. Financial Framework of the Facility 26. IDA’s financial framework has been reviewed with an aim to increasing the efficient use of liquidity and respond to heightened demand from IDA clients. The analysis indicates that a change to IDA’s liquidity management approach can be accommodated, thereby freeing-up previously uncommitted resources for commitment in client countries, while maintaining a robust coverage of IDA’s liquidity needs and preserving IDA’s financial sustainability. Specifically, a change in the liquid assets that count towards IDA’s minimum liquidity buffer would permit a one- time increase in IDA’s commitment authority. 27. This section will describe the proposed adjustment to IDA’s liquidity management practice in order to fund the facility. It will then discuss the way in which the lending terms are structured as non-concessional so as to reduce the potential financial risks and costs to IDA. The section will then elaborate on these risks, costs and mitigating factors. Box 2. Background on IDA’s liquidity management framework The steady growth of liquid assets provides management more scope to use liquidity more efficiently while still fulfilling its liquidity objectives. Currently at US$27 billion, IDA’s liquid assets have increased over time as replenishments have grown and as partners continue to accelerate their contributions.* Furthermore, recent financial measures in IDA16 and IDA17 have changed the pattern of IDA’s cash flows and resulted in higher liquidity levels.** Liquidity management objectives: The primary objective of liquid assets is to ensure that IDA has sufficient funds to meet disbursements and unexpected financial demands. IDA’s financial framework focuses on matching cash inflows and outflows. As part of that framework, a minimum liquidity level is maintained that serves to accommodate expected and unexpected cash flow volatility, which is currently held in cash and shorter term investments (in “Tranche 2” and “Tranche 3”). In addition to Tranche 2 and 3, IDA holds liquid asset s that are comprised of accelerated encashments and payments invested in medium term assets (“Tranche 1”). The income from Tranche 1 investments historically has been used to contribute to the recovery of the discount provided by IDA to partners, and thereby maintain IDA’s commitment authority to recipient countries. Beyond meeting disbursements and financial demands, IDA also seeks to maximize returns, subject to pre-specified loss constraints, so as to generate additional investment income that can be used to increase IDA’s internal resources available for future replenishments. * Each replenishment includes a standard encashment requiring partners to provide their payments over typically 9 years. Many partners accelerate their IDA encashments, either to receive payment discounts, for budgetary purposes, or as a way of providing additional resources to IDA. ** Such measures include exercising the contractual prepayment clause in credit agreements for IDA graduating countries; shortening the lending maturity and grace periods in IDA16 and IDA17; and most recently borrowing from partners via concessional loans. In addition, a number of IDA graduates have voluntarily prepaid their outstanding IDA credits since IDA16. This additional liquidity resulting from these measures has been incorporated into commitment authority for IDA16 and 17. However, these commitments do not have an immediate impact on liquidity and are drawn according to project disbursement schedules typically over 11 years. A. Proposed Adjustment to Liquidity Management 28. By including Tranche 1 in the assets counting towards IDA’s minimum liquidity, IDA can release resources not previously committed that have been held in reserve. With this change, assets held in Tranche 1 – in addition to Tranches 2 and 3 – could be used to fulfill minimum liquidity requirements and drawn upon under exceptional circumstances. This approach would formally recognize the capacity for Tranche 1 to serve as additional emergency liquidity in two ways: primarily as collateral for short-term borrowing through the repurchase agreement - 18 - (repo) market and as a source of cash, if needed. This change is feasible because Tranche 1 assets are invested similarly to the rest of IDA’s portfolio – in high quality, liquid investments. As a result of the change, assets set aside and uncommitted in Tranches 2 and 3 for minimum liquidity requirements could be released to fund the Scale-up Facility as a one-time measure to increase commitments. 29. While adding an additional protective measure to address liquidity needs, this proposal would not change IDA’s overall liquidity management approach. The primary objective of IDA’s liquid assets will continue to be to ensure that IDA has sufficient funds to meet disbursements and unexpected financial demands. IDA has never breached its requirements for maintaining a minimum liquidity buffer. As such, IDA will continue to maintain minimum levels of liquid assets to serve as a primary layer of protection for IDA’s commitment authority. Furthermore, IDA’s current investment objectives and strategy would not need to change.27 30. To confirm the size of the facility, an analysis was conducted to ensure that IDA can fulfill its current commitments and those proposed under the facility. The average minimum liquidity used in financial projections is approximately US$5.0 billion. The analysis focused on the impact of the new commitments under the facility on IDA’s liquidity. As shown in Figure 7, IDA is projected to maintain a liquidity buffer even after disbursing funds for projects under the facility.28 Beyond this buffer, Tranche 1, currently at US$17.0 billion, would be available as temporary emergency liquidity; IDA would also have sufficient liquidity available for working capital requirements. Figure 7. IDA's Projected Liquidity Buffer to 2030 (After working capital requirements) 31. An additional analysis was conducted which showed that IDA can substantially manage its liquidity requirements even under extreme stress scenarios. Such scenarios stressed cash inflows (contribution levels, payments and borrower repayments) as well as outflows (unexpected calls on liquidity and faster disbursements). Analysis showed that in all cases, IDA would have sufficient overall liquidity balances to meet commitments. Only in the most extreme 27 IDA would continue to maintain liquidity for working capital purposes for disbursements as well as short-term expected cash flow volatility. 28 These projections assume a base case financial projections scenario, where IDA commitment authority grows according to inflation. - 19 - scenarios – e.g., in which all the possible draws on IDA’s liquidity happen simultaneously – would IDA need to draw on Tranche 1 assets to cover projected liquidity shortfalls.29 Figure 8 illustrates the impact on IDA’s projected liquidity after including commitments under the facility and after an extreme scenario in which unexpected liquidity draws happen in FY16. As a result of this drawdown, IDA is required to draw on Tranche 1 to satisfy commitments starting in FY22 to FY24. The funding needs are projected to be US$700 million per year, or less than 2 percent of IDA17’s commitment authority. Figure 8. Funding Needs from Tranche 1 under Liquidity Stress Scenario (After considering working capital needs) 32. The liquidity management process provides Management the capacity to adjust IDA’s commitment authority at the start of each replenishment and in response to unexpected liquidity needs. When commitment authority is determined and financial projections are updated, Management incorporates updated assumptions and data to hone estimates of liquidity needs to satisfy commitments over the long term. As part of the normal financial projection process, there is an opportunity to recast and refill liquidity needs when draws on liquidity are required through an adjustment of commitment authority.30 For example, in the stress scenario described above, there are two replenishments before IDA would need to draw on Tranche 1 liquidity – at these points the commitment authority would be adjusted in response to the unexpected, emergency draws on liquidity by the amount of the drawdown. Any situation requiring prolonged usage of Tranche 1 would require a careful assessment of the cause and duration of the usage, with the objective of ensuring that IDA’s disbursement needs continue to be supported in a robust manner. B. Proposed Lending Terms for the Facility 33. Non-concessional terms (aligned with IBRD terms) are proposed to ensure IDA’s financial sustainability and the equitable use of its resources among IDA clients, while maximizing the resources available to them. The proposed lending terms are detailed in Box 3. Through an adjustment to the liquidity framework and its use on non-concessional terms, the 29 These possible draws include spikes in demand for IDA’s assistance, increases in non -accruals, and calls on guarantees in addition to increased cash flow volatility. 30 At the start of IDA16, commitment authority was adjusted downwards to account for unexpected, long-term delays in partner contributions in IDA15. - 20 - facility would provide resources that are additional to the existing concessional commitment authority without altering the future concessional commitment authority. Increasing the concessional commitment authority for IDA17 would reduce future concessional commitment authority. See Annex 3 for proposed pricing methodology and indicative pricing. 34. Non-concessional terms would limit the adverse impact to IDA’s internal resources . With the facility, investment balances would be lower and hence IDA would forego investment income. This loss of investment income would not contribute to IDA’s internal resources and result in lower resources available for future commitments. The estimated foregone investment income to fund the facility, which is expected to draw on funds currently invested in Tranche 2 and 3, is approximately 0.40 percent above LIBOR on an annual basis. Over a fifteen year horizon, forgone investment income would translate to approximately $55 million per annum.31 35. Furthermore, lowering liquidity holdings would also increase the probability that IDA will have to incur additional financing costs by utilizing Tranche 1 should an unexpected liquidity need arise. As described above, extreme liquidity stress scenarios may require either liquidation of or borrowing against Tranche 1 resources. In the case of liquidation, given that these assets are invested over the medium term, Tranche 1’s foregone investment income would be higher.32 In the case of borrowing against Tranche 1 resources, financing costs may be incurred if repo markets are accessed for short-term liquidity (currently at 0.20 percent per year);33 in addition to these costs, availability of repo market counterparts would be a consideration. Non- concessional terms will help recoup these costs. 36. With the proposed non-concessional terms for the facility, IDA’s administrative costs and forgone investment income are expected to be covered. Figure 9 compares IDA’s administrative costs and forgone investment income with the estimated loan revenue under the facility. This revenue is based on an estimate of loan terms taken under the facility (terms will be fixed, however it is assumed that countries with low risk of debt distress and adequate debt management capacity would opt for variable rates).34 As indicated in the chart, IDA’s costs would just be offset through the proposed terms of the facility at current interest rates. In the unlikely case that IDA would need to draw on Tranche 1, IDA’s actual costs may differ and are represented in the chart as “Potential Financing Costs”. 31 This estimate is based on an average investment spread provided by Treasury and assumes outstanding balances using ten year disbursement and a 5/25 loan; expected return of IDA’s investments in Tranche 2 over the next 3 years assuming current market conditions prevail. 32 Investment horizon for Tranche 1 is 5 years. The historical five average return for Tranche 1 is 3.36 percent. 33 Repo borrowing costs and availability can vary according to market conditions. 34 This conclusion is based on an estimate of loan terms taken under the facility and guided by the implementation arrangements. Under the proposed arrangements, less than 50 percent of the expected loans could be taken under IFL with a Variable Rate. - 21 - Box 3. Proposed Lending Terms Several key principles informed the proposed non-concessional lending terms for the facility. The lending options have considered the needs and constraints of IDA clients (e.g., debt management capacity and debt sustainability). Further, the terms aim to be simple but offer additional flexibility to benefit borrowers, where possible. By incorporating more flexibility, the facility will provide important groundwork and data for understanding what may be needed for future IDA offerings (i.e., under IDA leveraging options). Currency The choice of currency will be similar to what is currently offered for IDA clients. Loans in Special Drawing Rights (SDR) will be available for the fixed rate loan option only. Single currency loans (in the constituent currencies of the SDR – i.e., USD, EUR, GBP, and JPY) will be available for the fixed rate and floating rate loan options. Interest Rate Loans will be offered at fixed rates, however countries with low risk of debt distress and adequate debt management capacity could opt for variable rates (see section III).  Fixed rates (SDR, USD, EUR, GBP, JPY). The single currency fixed rates will be calculated as the fixed rate equivalent of the IBRD Flexible Loan (IFL) with a Fixed Spread rate for the relevant currency and maturity option. The SDR fixed rates will be determined as the weighted average of the single-currency rates for the relevant maturity option. Fixed rates will be calculated and published quarterly. Credits approved in each quarter will be subject to the rates published at the beginning of that quarter.  Floating rates (USD, EUR, GBP, JPY). The pricing under the facility will be linked to IFL pricing for equivalent maturities. The floating rates will comprise LIBOR plus an interest rate spread; the spread will be either fixed (IFL with a Fixed Spread) or variable over the life of the loan (IFL with a Variable Spread). Tenor Three options for repayment terms will be available. Each of these choices will have fixed straight line amortization schedules. While these options will provide IDA borrowers with similar choice to what is currently provided under IDA’s blend terms (25 years maturity with 5 years grace period), the three options will have additional price differentiation to benefit borrowers by offering different repayment profiles and maturity premiums. Lending under the facility will not be subject to IDA’s acceleration repayment clause. Repayment period Grace period Average Maturity Grant Element* (years) (years) (years) 24 5 15 19.0% 27 8 18 18.3% 30 9 20 18.0% Fees Pricing will include the following fees: a 0.25 percent front-end fee; a 0.25 percent commitment fee on undisbursed balances, which will help offset the initial increase in budget necessary to implement these projects. *The grant element is determined using the NCBP grant element calculator, which is estimated based on the current parameters of the NCBP, which are subject to change over time. - 22 - Figure 9. Scale-up Facility Lending Costs vs. Estimated Non-concessional Loan Revenue (Spread above LIBOR in basis points)35, 36 As of October 2015, the implied LIBOR forwards project gradual increase in rates from the current level of 0.53 to 2.82 percent over the next 15 years, during which loans from the facility will be outstanding. 37. Non-concessional terms also enable IDA to maximize available funding for the facility and over the long-term. IDA would not be able to release the same amount of resources on concessional terms as is proposed under the facility, without impacting future commitment authority. This tradeoff occurs because lending terms directly impact future reflows and projected liquidity levels. For example, if financing from the facility is provided on concessional terms and especially if a portion of the additional funding is to be used for grant-only facilities, future reflows to IDA would be lower than if the funds are provided on non-concessional terms. Table 4 estimates the difference in reflows if the facility is lent on non-concessional versus IDA concessional terms. As shown, the difference between reflows from non-concessional and regular credits would be between US$300 million to US$800 million over 5 to 10 years, respectively for US$5.0 billion in commitments. The costs associated with the forgone investment income mentioned above are not included in this analysis, and would be in addition to these lost reflows. In addition, this analysis does not consider the impact on IDA’s future liquidity from lower reflows, which would necessitate a smaller facility. 35 Investment spread provided by IMD, based on Tranche 2 and Tranche 3’s expected return. The estimated opportunity cost is 40 bps above 6 month SDR Libor on an annual basis for Tranche 2, and 0 bps above 6 month SDR Libor for Tranche 3. Using a zero basis between USD and SDR LIBOR, there will be an equivalent spread for each Tranche over 6 month USD LIBOR. This estimate is based on a target duration of 2 years for Tranche 2 portfolio, and long term equilibrium interest rates scenario (over a 15 year period) provided by Oxford Economics model. The comparison shows the spread above LIBOR to illustrate the difference, as LIBOR would be included in costs and revenue. 36 Revenues include commitment and front-end fees, which will help to cover incremental budget increases at the start of the facility. - 23 - Table 4. Difference in Reflows between Non-concessional and Concessional Lending (for US$5 billion in Commitments in cumulative terms) Over 10 Over life of (in US$ millions) Over 5 years years the loan Non-concessional vs Grant 372 1,827 7,675 Non-concessional vs Regular 257 810 1,438 Non-concessional vs Blend 142 236 921 38. Allocating US$900 million of the funding to the CRW would impact internal resources only if they are needed to respond to crises. If the full US$900 million is used for crises and committed on concessional terms, reflows would be lower by $300 million over 10 years or $1.1 billion over the life of the loan.37 Because this allocation would be contingent, it is proposed to that any unused portion of the CRW resources by end-December 2016 would be made available to IDA countries on non-concessional terms to limit the impact on internal resources available for future replenishments. 39. Since the facility resources are limited, providing them on non-concessional terms helps address the issue of equitable treatment among IDA recipients. Through the facility, IDA is able to reallocate internal resources held in its investment portfolio for the benefit of those few countries that are eligible; as proposed, the facility would provide funding for specific projects. Charging non-concessional terms ensures that IDA earns reflows at levels similar to the investment income it would have otherwise earned through its investment portfolio, thereby securing internal resources for the benefit of all countries in future replenishments. 40. Finally, implementing a non-concessional financing model in IDA17 would provide insights for the design of a potential longer-term non-concessional facility in IDA18. Introducing non-concessional lending products in the facility would provide empirical knowledge of the needs and constraints of IDA clients for non-concessional financing. Prudently sized and funded by uncommitted resources, a non-concessional facility could yield important lessons learned, while limiting the financial risks to IDA. C. Asset Liability Management Considerations 41. As described in the proposed terms, the facility will offer single-currency and SDR fixed rate loans. Countries with low risk of debt distress and adequate debt management capacity could elect for floating rates. (SDR rates will only be offered at fixed rates). The implications on IDA’s financial framework of the proposed terms are as follows: Currency Risk 42. IDA employs an existing currency management framework that provides a portfolio hedge to mitigate currency risk.38 The approach ensures that the currency composition of IDA’s overall assets is aligned to the SDR basket.39 In addition to managing the currency exposure 37 This analysis is based on a historical breakdown of CRW support, i.e., 75 percent on credit terms, 25 percent on grant terms. 38 IDA does not hedge individual cash flow transactions to SDRs. 39 IDA’s overall asset composition is comprised of the liquid asset portfolio plus the present value of future “non -SDR” cash flows are, on a combined basis, in SDR composition. - 24 - relating to partner contributions, which are typically provided in local currencies, IDA provides recipients the opportunity to receive credits denominated in one of the SDR currencies under the Single-Currency Lending (SCL) Pilot program. IDA’s existing currency management framework adequately addresses the currency risk resulting from this program, which is limited to SDR3.0 billion (about US$4.5 billion). 43. To accommodate potential demand from the facility, the SCL would need to be increased to SDR7.0 billion (about US$10 billion). A larger SCL program can be managed within the existing currency framework, however would require measures to address the potential additional long-term impact on currency balances needed for working capital.40 Specifically, in addition to regular currency rebalancing, IDA may need to increase its use of market-based instruments (e.g., spot-forward trades and currency swaps) for currency management of working capital balances.41 These measures are currently permitted under the existing program and have been used in the past for currency management purposes, but would need to be increased to accommodate the facility (see Annex 2 for additional details). Management will also enhance its existing currency framework by expanding its operating guidelines. Use of these instruments, which will be executed by IBRD on behalf of IDA, will increase commercial counterparty risk exposure and may result in incrementally higher hedging cost per transaction due to the increased use of existing credit lines. Interest Rate Risk 44. For the facility, IDA is proposing to offer non-concessional credits with fixed rates. However, countries with low risk and adequate debt management capacity could opt for floating interest rates (LIBOR + fixed or variable spread). Management does not expect the floating rate option to impose any significant interest rate risks on IDA based on the following factors:  The facility carries no external funding costs to IDA. Instead, the facility will be financed through the portion of IDA’s liquidity which has not been committed.  Additionally, analysis described earlier indicates that IDA’s administrative costs will be covered by loan revenues. As administrative costs are the only fixed costs associated with the facility, there are no additional fixed costs to hedge.  From an overall balance sheet perspective, when measuring IDA’s interest rate exposure, adding floating interest rate credits to IDA’s portfolio of credits would reduce IDA’s equity interest rate exposure.  Moreover, the potential impact of including floating rate loans in IDA’s balance sheet is limited given the size of the facility.42  Finally, while floating interest income will add some variability to IDA’s internal resources (reflows) for future replenishments, it does not introduce any new risk in IDA’s financial 40 IDA’s liquidity is maintained in the four currencies of the SDR basket. As a result of the SCL, while overall liquidity may be sufficient for working capital purposes, the liquidity balances in one of the individual currencies may be too low, as a result of high demand for credits in one currency. 41 Alternatively, IDA could transfer cash from its medium duration tranche by liquidating investments and investing in futures to maintain the investment objectives of the tranche. 42 Analysis was conducted that assumed that the entire facility ($US5 billion) was lent on floating interest rate terms. A one percent shift in LIBOR would lead to a US$22 million change in average annual interest income over the life of the loan. This represents 0.51 percent of IDA’s current reflows. - 25 - framework. IDA’s investment income, which represents the opportunity cost for the facility, is subject to the same source of uncertainty. A shift of market interest rates would have the same effect on IDA’s investment income. Therefore, introducing floating rates for the facility does not introduce any new risks to IDA’s reflows which impacts future commitment authority. Given the proposed eligibility criteria for accessing floating rate credits from the facility, it is not expected that IDA’s exposure to floating interest loans or one single currency will be maximized to the full extent of the facility (US$5.0 billion). A mix of interest rate terms and currency composition taken by the client countries would largely reduce the potential risk IDA would face to currency and interest rate risks. Spread Risk 45. The facility will offer IFL with a Variable Spread as one of the alternatives. This product would be priced with a variable spread component that is based on IBRD’s funding costs.43 E.g., if IBRD’s funding costs fall, then the variable spread and overall interest rate will decrease as well. Because this product would be linked to IBRD funding costs, offering it for the facility may add risks to IDA’s reflows that need to be balanced by the potential benefits to clients. IBRD clients have benefitted from and prefer this option because it is currently lower than the fixed spread option.44 From the client perspective, an improvement in IBRD’s funding costs would benefit clients with lower lending rates, while the converse is true.45 Clear communication with IDA borrowers will be necessary to ensure they are aware of the way in which IFL with a Variable Spread are linked to IBRD funding risk.46 Additional implementation considerations 46. Given the non-concessional nature of the financing from the facility, outreach efforts will be undertaken to help borrowers make informed decisions about the new lending options. As Management looks to broaden lending options available to IDA clients not only through this facility but also through a leveraging facility, a long-term approach will be necessary to ensure IDA clients have ongoing engagement with Bank staff (Treasury’s Banking Team and DFIRM) in order to learn about new IDA lending options. A strategy for developing such an outreach program could draw from experience gained from introducing new IBRD loan terms and products as well as the terms proposed under the facility. 43 In addition, all IBRD loan prices include a commitment fee, upfront fee, contractual spread and maturity premium. 44 IFL with a Variable Spread accounts for 65 percent of IBRD’s total loan balance outstanding as of June 30, 2015. The IFL Variable Spread and Fixed Spread are priced to be equivalent on a risk-adjusted basis. 45 IBRD bears a similar exposure to improvement in its funding costs for the portion of loans funded by equity, and this risk has been viewed as acceptable given the ability for clients to benefit from IBRD’s improved funding performance. Conversely, should IBRD funding costs deteriorate, IDA clients would need to pay higher rates. 46 Offering IFL with a Variable Spread or any other option would be part of the testing nature of the Scale-up Facility and would not necessarily have to be offered in the future. If IDA changes its funding model, any lending product would need to be appropriate for IDA’s financial framework. - 26 - IV. Issues for Discussion 1. Do Deputies agree to establish, on a one-off basis, a facility of up to US$5.0 billion to scale-up IDA’s financing for the remainder of the IDA17 period? 2. Do Deputies agree with the proposed use of the resources from the facility? Specifically:  A contingent amount of up to US$900 million would be used to replenish the CRW.  The remaining US$4,100 million – plus any unused portion of the CRW resources by end-December 2016 – would be made available to IDA countries on non- concessional terms (as proposed in this paper). 3. Do Deputies agree with the proposed implementation arrangements for the facility, including with respect to the purpose, eligibility criteria, financing terms, allocation framework and project prioritization criteria for the non-concessional resources? - 27 - Annex 1. Indicative Demand for IDA17 Scale-up Facility: Projects Submitted by the Regions in July 2015 1 / Country Name of Project Bangladesh Financial Institutions Strengthening and Automation Investment Promotion and Financing Facility 2 Dhaka Sanitation Improvement Gas Sector Efficiency Improvement Private Sector Development Support Project 2 Cameroon Transport Sector Support Project Electricity Transmission Project Livestock Development Project Central African Republic Fiber-optic Backbone Cote d'Ivoire LNG Import Facility Power System Upgrade Greater Abidjan/Port Development Project: Phase 1 Improving Port of Abidjan Access and Productivity; Phase 2 Logistics Platform for Port of Abidjan at PK 25 Regional Integration and Trade Competitiveness - Abidjan-Ouaga Railway Corridor Rehabilitation Logistics Platform in Bouake Abidjan Urban Mobility Improvement Project - Tramway and Lagoon Transport Democratic Republic of Congo Inga 3 Ethiopia Adama-Awash Expressway (125km) Transmission and Distribution Network Reinforcement and Expansion Robe-Gasera-Ginir (120k) Sodere-Nuraera-Metehara (80km) Itaya-Robe-Seru (150km) Meiso-Dire Dawa (100km) Guinea Bissau Port Project Guyana Science and Technology Support Secondary Education Improvement Flood Risk Management Honduras Roads Project Kenya Water and Sanitation Service Improvement Project Transport Modernization Project Garissa-Mandera Road Climate Smart Agriculture Kosovo Power Project Education System Improvement Project Energy Efficiency and Renewable Energy Project Kyrgyz Republic Integrated Dairy Productivity Improvement Project Lao PDR Roads, including PPP for Road No.13 (North and South) Lesotho Highlands Water Project Phase II (LHWP2) Mali Urban Water Supply Irrigation Mauritania Banda Gas to Power Project Wind Power Project Free Zone Project Mozambique Water Service & Institutional Support II Agriculture and Natural Resource Management Landscape Project Feeder Roads Development Project Nicaragua Rural Roads Infrastructure Improvement Project Nigeria Power Distribution Project 1/ Preliminary information provided by Country Management Units in the World Bank in response to a query about potential client demand for non-concessional financing for projects in IDA17 with strong returns, consistent with a country’s debt sustainability and with NCBP requirements. List is indicative, and does not include DPOs or PforRs. - 28 - Annex 1. (Continued) Country Name of Project Pakistan Dasu Hydropower Stage I Project Tarbela 5th Hydropower Extension Proejct Republic of Congo Sounda Dam Project Rwanda Energy Infrastructure & Utility Modernization and Expansion Promoting Sustainable Economic Growth including Urban Development Mining and Supply Chain Development Sao Tome and Principe Power Sector Recovery Project Senegal Energy Sector Support Project Water Project Urban Project Solar Energy Project Tajikistan Winter Energy Project Tanzania Maritime Gateway Project Education and Skills for Productive Jobs - Program for Results Growth Poles Project Uganda First Oil Infrastructure Project Uzbekistan Electricity Distribution Systems Modernization Project Industrial Energy Efficiency Scale-up Program Zambia Improved Rural Connectivity Project Least-Cost Power Generation Capacity Expansion Africa Regional Lake Victoria Transport Program Gulu-Kisangani Road Tanzania-Zambia Interconnector Regional Transmission Development APL MZ-MW Interconnector Dakar Bamako Railway Improvement Project Northcore Interconnection (330 kV) Nigeria – Niger –Benin –Burkina Faso N’Djamena-Abéché-Sudan Corridor N’Djaména-Mao-Niger Road Ghana-Burkina Faso-Mali Interconnection LCR Regional Competitiveness Project Financial Sector Strengthening Regional Rebewable Energy Operation Human Development and Social Resilience Project Note: Graduates and Mongolia excluded. -29- Annex 2. IDA17 Scale-Up Facility Financial Risk Management In order to provide terms with additional flexibility under the Scale-up Facility, IDA will need to increase its Single-Currency Lending limit and address any impact on IDA’s financing framework from offering floating rate credits. This annex provides an overview of the risk implications, as well as a description of IDA’s approach to managing these risks. A. Currency Risk Background 1. IDA is inherently exposed to currency risk due to the nature of its cash inflows and outflows. While IDA credits and grants are provided in SDRs, IDA is exposed to the following “non-SDR” cash flows:  Partner contributions and MDRI compensation are typically provided in partners’ local currency.  World Bank Group transfers, IDA administrative expenses, HIPC costs, and legacy credit reflows1 are all denominated in US dollars.  Under IDA’s Single Currency Lending (SCL) Pilot program, credits are provided in one of USD, EUR, JPY, and GBP up to an equivalent of SDR 3 billion; the disbursements as well as the reflows and the service charges are all denominated in the currency of choice by the borrower.  Concessional Partner Loans (CPLs) are provided to IDA in one of USD, EUR, JPY, and GBP, and IDA pays interest on CPLs denominated in the currency of the loan. These cash flows generate IDA’s currency risk and are highlighted in the following figure. Figure 1: IDA Cash Flows 1 Legacy credit reflows are repayments on US denominated credits. Prior to 1980, all IDA credits were denominated in US dollars. - 30 - IDA’s Currency Risk Management Framework 2. IDA’s currency risk management framework provides a portfolio hedge to mitigate currency risk.2 The approach ensures that the currency composition of the liquid asset portfolio plus the currency composition of the present value of future “non-SDR” cash flows are, on a combined basis, in SDR composition (see Figure 2). Figure 2: Currency Risk Management Framework – Concept Highlight 3. The framework involves a three step process. First, at the start of each replenishment, FX forwards are executed to hedge the larger partner contributions to the four composite currencies of the SDR. The largest partner contributions in non-SDR currencies are converted through FX forwards, while the remaining non-SDR partner contributions are hedged using a currency correlation methodology.3 Second, overlay FX forwards are then used to align the converted partner contributions and other non-SDR currency flows to the SDR composition.4,5 Third, the currency composition of the liquid asset portfolio is adjusted on a regular basis to maintain the SDR composition of IDA’s net assets (i.e., liquid assets portfolio, future cash flows and FX forwards as highlighted in Figure 2). 4. Once aligned with the SDR composition, the “Combined Portfolio” does not deviate from the SDR equivalent when the exchange rates of the four SDR currencies change. However, there are a number of factors that can cause shifts in the currency composition of the combined portfolio, bringing the Combined Portfolio out of SDR alignment, and necessitate a rebalancing of IDA’s liquid asset portfolio or additional FX transactions to bring the Combined Portfolio back to the SDR composition. These factors are:  Changes to the liquid asset portfolio: Investment income, mark-to-market adjustments, and disbursements and repayments on credits;  Changes in present values of non-SDR cash flows: These changes are caused by changes in market interest rates in the SDR currencies; changes in partner payments schedules; changes in projected administrative expenses of IDA; the additional cash 2 IDA does not hedge individual cash flow transactions to SDRs. 3 “Use of Derivative Instruments for Management of IDA’s Currency Exposure through Intermediation by IBRD”, R2008 -0086, IDA/R2008-0101, April 29, 2008. 4 IDA hedges all WBG transfers, 3-years of MDRI compensation and HIPC costs, 2-years of rolling administrative expenses, and 10-years of legacy US dollar credit reflows. 5 All FX forwards are intermediated by IBRD. Occasionally, when required due to changes to partner payment schedules, or when a particular currency’s liquidity balance is temporarily too low, additional forwards are executed. - 31 - flows associated with each new IDA replenishment; and any cash flows resulting from new SCLs;  Changes in FX rates of non-SDR currencies not hedged through FX forwards; and  Changes to the SDR-basket Adjustments to IDA’s liquidity portfolio are made i) daily to adjust for IDA disbursement and repayments6 and for approvals of new single currency credits; and (ii) at least quarterly for the other above mentioned variables. Management of the SCL Credits and the Potential Impact of Increasing the SCL Limit 5. Single-currency credits add to IDA’s “non-SDR” cash flows and are managed by including their cash flows as part of the present value of future “non-SDR” cash flows in the above framework (see Figure 2).7 When new SCLs are approved, DFi and TRE are notified so that these loans can be incorporated into the “Combined Portfolio”. Specifically, DFi updates its future cash flow projections with new SCL project information. Based on this updated currency composition of IDA assets, DFi and TRE work together to determine whether any incremental adjustments to liquid assets are required. Because any currency mismatch is calculated based on the present value of all related future cash flows (disbursements and reflows), required adjustments to the liquid asset portfolio are typically small relative to the project size. 6. Including SCL cash flows as part of the present value of future non-SDR cash flows does not represent a “perfect” cash flow hedge and continuous adjustments to the liquid portfolio are needed. This is mainly because (i) the exact disbursement schedule of an IDA credit is not fully known at the outset; and (ii) the PVs change with fluctuations in interest rates. The DFi team uses an estimated disbursement schedule, which reflects the nature of the project (e.g., a fast disbursing development policy operation or a more slowly disbursing investment lending operation). The disbursement schedules included in the PVs are adjusted on a regular basis to reflect actual disbursements. Changes in the timing of payments (e.g., moving payments from one quarter to the next) typically cause only small changes in the PVs. Changes in interest rates are similarly managed by updating the PVs on a quarterly basis with revised interest rates. 7. The process described above ensures that the majority of the currency risk due to SCL credits is mitigated and that the PVs are updated continuously to reflect updated financial information. However, the size of IDA’s liquid asset portfolio, as well as currencies demanded by client countries ultimately determine the maximum potential size of the SCL program that can be managed through regular currency rebalancing of the liquid asset portfolio. Under the SCL pilot program, IDA is permitted to use additional market-based instruments or to adjust its liquid asset portfolio’s currency alignment away from the current SDR benchmark in order to ensure that liquidity balances are maintained at adequate levels.8 To date, however, only 6 Required given that disbursements and repayments are typically made in SDR equivalent amounts of US dollars or Euros. 7 Projected disbursements and reflows for the full duration of the credit (up to 40 years for IDA regular credits) would be incorporated in the PV calculation. 8 Even if the liquid asset portfolio is aligned away from the current SDR benchmark, as long as the “Combined Portfolio” is aligned with the SDR basket, IDA remains hedged to the SDR. - 32 - limited use of market-based instruments and liquid asset portfolio alignment adjustments has been required. 8. For the existing single currency program, which is limited to SDR3 billion (US$4.5 billion), the regular currency rebalancing approach meets the needs of the current program. Furthermore, IDA’s liquid asset portfolio has the capacity to manage any adjustments to the portfolio that would be required. 9. An SCL offering under the Scale-up Facility could increase IDA’s single currency lending portfolio to SDR7 billion (US$10 billion). Given that IDA has the ability to manage additional single currency exposure within its existing currency risk management framework, this proposed expansion of SCL is not expected to increase IDA’s currency risk significantly. 10. The framework, however, might require increased use of market-based instruments to ensure adequate working capital liquidity balances. An increase in the number of SCL projects might lead to a shortfall in one of the currency balances over time.9 This shortfall would occur particularly if all of the new SCL projects are approved in only one currency (such as USD or EUR). In the first few years of an SCL project, the disbursements will lead to a decrease in the balance of the chosen currency,10 which will be restored only when the project is fully repaid. This impact on the currency balance will not impose any additional currency risks on IDA because IDA’s total assets (liquid assets plus the present value of future receivables) will continue to be aligned to the SDR basket composition. However, during the disbursement periods of the project cycle, IDA may face a low liquidity balance in one of the currencies and this may impact IDA’s ability to make timely disbursements. The exact impact would depend on the actual usage of IDA’s SCL pilot program and the Scale-up Facility. Risk Management Approach 11. IDA has performed an analysis to estimate the potential impact on long-term currency balances. The analysis assumes an extreme scenario in which all of the SCL pilot program and the Scale-up Facility lending are used and approved in one currency, the US dollar. In this case, results shows a net maximum negative balance of US$7 billion in IDA’s portfolio during the disbursement period as described in the chart below. This negative balance would be gradually offset over the repayment period (25 years average is used in the analysis). While the USD balance of IDA’s entire liquid asset portfolio is large enough (US$14 billion) to cover this US$7 billion negative balance, only US$4 billion is invested in short term assets and available for working capital purposes. Hence, according to the results of the analysis, the level of working capital represents a constraint on the currency rebalancing approach. 9 IDA’s liquidity is maintained in the four currencies of the SDR basket. As a result of the SCL, while overall liquidity may b e sufficient for working capital purposes, the liquidity balances in one of the individual currencies may be too low, as a result of high demand for credits in one currency. 10 In contrast, the disbursement of a project approved on SDR terms will proportionally reduce the balances of the four SDR currencies in IDA’s liquid asset portfolio and therefore will not affect IDA’s liquidity balance to the same extent. - 33 - Figure 3: Net impact on currency balances with US$10 billion in approved SCL This chart indicates the single-currency cash flow impact on liquidity based on the case described above, where the full US$10 billion SCL is approved in USD (assuming liquidity starts at nil). 12. This suggests that in addition to regular currency rebalancing, IDA may need to use market-based instruments (e.g., spot-forward trades and currency swaps) for currency management.11 These measures are permitted under the current SCL program, but have not been used extensively to date and might be more frequently required with an expansion of the program. For example, IDA could employ spot-forward trades, in which it would purchase the currency it needs on the spot market and simultaneously sell it on the forward market, either directly with the market or through IBRD. By converting the currency of a future receivable, IDA can relieve the working capital constraints in that currency in the short term. The net impact of such a transaction on the currency balance would be minimal, as IDA would be trading the same currencies in both the spot and forward trades, and the SDR balance of IDA’s total assets will not be affected. IDA has implemented this type of currency transactions in the past for the same purposes. The Bank’s Treasury has confirmed that IBRD, which intermediates IDA’s FX trades with the market, has the capacity to use spot-forward currency swaps to cover the US$10 billion in potential currency rebalancing needs.12 Use of these instruments, which will be executed by IBRD on behalf of IDA, will increase commercial counterparty risk exposure and may result in incrementally higher hedging cost per transaction due to the increased use of existing credit lines. 13. Lessons learned from the SCL pilot could inform the extent to which additional risk measures that may need to be employed. To date, the utilization of the SCL pilot program has been modest; of the SDR 3 billion limit, SDR 1.8 billion still remains available as of September 2015.13 An effective system to incorporate the incremental currency impact of these additional “non-SDR” flows has been established and IDA has successfully managed any incremental currency risk. The uptake of the program so far indicates that, while there is some level of demand 11 Alternatively, IDA could occasionally invest a larger portion of the liquid asset portfolio in a composition that is “off” th e SDR basket to realign the currencies of the liquid asset portfolio. In doing so, it could convert securities in a needed currency into cash and transfer the funds into cash. 12 As indicated, this is the maximum IDA would require if all SCL was provided in one currency. 13 Of the SDR1.2 billion limit that has been used, about SDR1 billion represents the transitional support credits to India. - 34 - for single currency borrowing among IDA clients, the prevalent form of IDA borrowing is still in SDR denominated credits. Although there has been a significant increase particularly in the demand for USD-denominated credits, we have experienced some demand for EUR denominated credits as well. In the light of the above analysis and the lessons learned from the SCL pilot, the existing implementation guidelines, which contain the overall governance and decision making process, will be expanded to include the use of additional market-based instruments. B. Interest Rate Risk Background 14. IDA offers its concessional credits at fixed interest rates and charges a fixed rate service fee. With the recent extension of SCL pilot program, a floating rate option was offered for transitional support and hard-term lending credits. The rate for these credits are linked to IBRD’s IFL terms, and are priced at LIBOR plus a fixed spread. Because these credits have a fixed rate funding cost, IDA is required to hedge the floating interest rates on SCL transitional support and hard-term credits to fixed rate. For the Scale-up Facility, IDA is proposing to offer non-concessional credits with floating interest rates (LIBOR + fixed or variable spread), as well as fixed rates. Potential Impact and Risk Management Approach 15. Management does not expect the floating rate option to impose any significant interest rate risks on IDA based on the following factors:  Unlike the Transitional Support Window, the Scale-up Facility carries no external funding costs to IDA. Instead, the facility will be financed through the portion of IDA’s liquidity which has not been committed.  Additionally, analysis indicates that IDA’s administrative costs will be covered by loan revenues. As administrative costs are the only fixed costs associated with the facility, there are no additional fixed costs to hedge.  From an overall balance sheet perspective, when measuring IDA’s interest rate exposure using a net present value balance sheet approach, adding floating interest rate credits to IDA’s portfolio of credits would reduce IDA’s equity interest rate exposure.  Moreover, while floating interest income will add some uncertainty to IDA’s internal resources for future replenishments, it does not introduce any new risk in IDA’s financial framework. IDA’s investment income, which represents the opportunity cost for IDA17 Scale-up Facility, is subject to the same source of uncertainty. A shift in market interest rates would have the same effect on IDA’s investment income. Therefore, introducing floating rates for the Scale-up Facility does not introduce any new risks to IDA’s income.  The potential impact of including floating rate loans on IDA’s balance sheet is also limited given the size of the Scale-up Facility. To quantify this impact, analysis was conducted that assumed that the entire facility (US$5 billion) was lent on floating interest rate terms. A one percent shift in LIBOR would lead to a US$22 million change in average annual - 35 - interest income over the life of the loan.14 This represents 0.51 percent of IDA’s current reflows. Table 1: Change in Interest Income from a 1 percent Shift in Rates (US$ millions) Life of Loan Year 1-10 Year 11-20 Year 21-30 Total Average 284 310 71 665 22  Given the proposed eligibility criteria for accessing floating rate Scale-up Facility credits, it is not expected that IDA’s exposure to floating interest loans or one single currency will be maximized to the full extent of the facility (US$5 billion). A mix of interest rate terms and currency composition taken by the client countries would largely reduce the potential risk IDA would face to currency and interest rate risks. C. Spread risk 16. The proposed Scale-up Facility will include the option to obtain IBRD IFL terms with a Variable Spread, for countries with low risk of debt distress and adequate debt management capacity. This product would be priced with a variable spread component that is based on IBRD’s funding costs.15 (e.g., if IBRD’s funding costs fall, then the variable spread and overall interest rate paid by the borrower will decrease as well.) Because this product would be linked to IBRD funding costs, offering it for the facility may add risks to IDA’s income that need to be balanced by the potential benefits to recipient countries. 17. Financial sustainability. By offering IFL terms with a Variable Spread, the risk to IDA would rise if IBRD’s funding costs improve to a level where the adjusted IFL rate does not cover IDA’s costs over the life of the loan. For the Scale-up Facility, this cost is IDA’s administrative costs and forgone investment income. As presented earlier, IDA’s administrative costs and forgone investment income would be covered by the proposed terms of the Scale-up Facility, even with some variability in IFL with Variable terms.16 Historically, the variability of IBRD’s funding margins has been modest. Furthermore, the proposed implementation arrangements would limit access to floating interest rates to high capacity countries thereby mitigating the overall risk of offering this product. 18. Client preference. Although this product has unique characteristics that differ from IDA’s existing products, IBRD clients have benefitted from and strongly prefer this option because it is significantly cheaper than the fixed spread option.17 From the client perspective, an improvement 14 As a point of comparison, a 1 percent shift in interest rates is estimated to change IDA’s investment income by US$32 million using the current portfolio and data as of September 30, 2015. 15 In addition, all IBRD loan prices include a commitment fee, upfront fee, contractual spread and maturity premium. 16 This conclusion is based on an estimate of loan terms taken under the Scale-up Facility and guided by the implementation arrangements. Under the proposed arrangements, less than 50 percent of the expected loans could be taken under IFL Variable Spread. 17 IFL with a Variable Spread accounts for 65 percent of IBRD’s total loan balance outstanding as of June 30, 2015. Over the short-term IFL with a Variable Spread is currently lower than IFL with a Fixed Spread. However, over the long-term these rates are priced to be equivalent. - 30 - in IBRD’s funding costs would benefit clients with lower lending rates, while an increase in IBRD's funding cost would increase clients' lending rates.18 Clear communication with IDA borrowers will be necessary to ensure they are aware of the way in which IFL with a Variable Spread are linked to IBRD funding risk.19 18 IBRD bears a similar exposure to improvement in its funding costs for the portion of loans funded by equity, and this risk has been viewed as acceptable given the ability for clients to benefit from IBRD’s improved funding performance . 19 Offering IFL with a Variable Spread or any other option would be part of the testing nature of the Scale-up Facility and would not necessarily have to be offered in the future. If IDA changes its funding model, any lending product would need to be appropriate for IDA’s financial framework. -29- Annex 3. Proposed Pricing Methodology and Indicative Rates Loans will be offered at fixed rates, however countries with low risk and adequate debt management capacity could opt for variable rates. The following information describes the proposed pricing methodology and indicative rates based on this methodology and current interest rates. Fixed rates  Fixed rates will be available for both single currency and SDR denominations. They will be calculated and published quarterly by IDA. Credits approved in each quarter will be subject to the rates published in the beginning of that quarter.  The single currency fixed rates will be calculated as the fixed rate equivalent of the IBRD’s IFL with Fixed Spread rates for the relevant currency and maturity options. They will be determined by taking the average of IBRD’s fixed rate equivalent for the relevant currency and maturity over a 1 month period prior to the start of each quarter.  The SDR fixed rate will be determined as the weighted average of the single-currency fixed rates for the relevant maturity option. The currency weights in the SDR composition will be determined by taking the average of SDR currency weights over a 1 month period prior to the start of each quarter.  In addition, all fixed rates credit will carry a 0.25 percent front-end fee and a 0.25 percent commitment fee that is charged on undisbursed balances.  Lending under the Scale-up Facility will not be subject to IDA’s acceleration repayment clause. Table 1. Indicative Fixed Rates for Scale-up Facility Credits Based on current rates as of FY16 Q2 IBRD Maturity USD EUR JPY GBP SDR Bucket Option 1: 24/5 12-15 yrs 3.62% 2.39% 1.80% 3.17% 3.02% Option 2: 27/8 15-18 yrs 3.90% 2.70% 2.18% 3.42% 3.31% Option 3: 30/9 18-20 yrs 4.05% 2.84% 2.37% 3.53% 3.45% - 30 - Floating rates  Floating rates will be available for single currency denominations only. The pricing of the floating rate credits will be linked to the pricing of IBRD flexible loans with the same average maturity.  The floating rates will comprise LIBOR plus an interest rate spread; the spread will be either fixed (IFL with a Fixed Spread Loan) or variable over the life of the loan (IFL with a Variable Spread Loan).  Floating rate credits will carry a 0.25 percent front-end fee and a 0.25 percent commitment fee that is charged on undisbursed balances.  For the fixed spread option, the fixed spread will be determined at loan signing. For the variable spread option, the variable spread will be recalculated by IBRD every January 1 and July 1, or as needed.  Lending under the Scale-up Facility will not be subject to IDA’s acceleration repayment clause. Table 2. Indicative Floating Rates for Scale-up Facility Credits (based on current rates of FY16 Q21 IFL with Fixed Spread IBRD Maturity USD EUR JPY GBP Bucket Option 1: 24/5 12-15 yrs LIBOR+1.05% EURIBOR+1.00% LIBOR+0.90% LIBOR+1.05% Option 2: 27/8 15-18 yrs LIBOR+1.25% EURIBOR+1.20% LIBOR+1.10% LIBOR+1.25% Option 3: 30/9 18-20 yrs LIBOR+1.35% EURIBOR+1.30% LIBOR+1.20% LIBOR+1.35% IFL with a Variable Spread IBRD Maturity USD EUR JPY GBP Bucket Option 1: 24/5 12-15 yrs LIBOR+0.65% EURIBOR+0.65% LIBOR+0.65% LIBOR+0.65% Option 2: 27/8 15-18 yrs LIBOR+0.75% EURIBOR+0.75% LIBOR+0.75% LIBOR+0.75% Option 3: 30/9 18-20 yrs LIBOR+0.85% EURIBOR+0.85% LIBOR+0.85% LIBOR+0.85% 1 Over the short-term IFL with a Variable Spread is currently lower than IFL with a Fixed Spread. However, these products are priced to be equivalent over the long-term.