The Multilateral Investment Guarantee Agency’s Experience with Non-Honoring of Sovereign, Sub- Sovereign, and State- Owned Enterprise Financial Obligation Guarantees Meso-Evaluation © 2022 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org ATTRIBUTION Please cite the report as: World Bank. 2022. The Multilateral Investment Guarantee Agency’s Experience with Non-Honoring of Sovereign, Sub-Sovereign, and State-Owned Enterprise Financial Obligation Guarantees. Meso-Evaluation. Independent Evaluation Group. Washington, DC: World Bank. COVER PHOTO Shutterstock/Lightspring EDITING AND PRODUCTION Amanda O’Brien GRAPHIC DESIGN Luísa Ulhoa This work is a product of the staff of The World Bank with external contributions. The findings, inter- pretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundar- ies, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorse- ment or acceptance of such boundaries. RIGHTS AND PERMISSIONS The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522- 2625; e-mail: pubrights@worldbank.org. The Multilateral Investment Guarantee Agency’s Experience with Non-Honoring of Sovereign, Sub-Sovereign, and State-Owned Enterprise Financial Obligation Guarantees Meso-Evaluation May 19, 2022 1 Contents Abbreviations .................................................................................................................................................... iv Acknowledgments ............................................................................................................................................ v Overview ............................................................................................................................................................vii Multilateral Investment Guarantee Agency Management Comments ......................................xiv 1. Introduction ...................................................................................................................................................1 2. Multilateral Investment Guarantee Agency’s Non-Honoring of Financial Obligations Guarantee Products .........................................................................................................................................2 Background ....................................................................................................................................................3 Business Rationale for Multilateral Investment Guarantee Agency Non-Honoring of Financial Obligations Products ................................................................................................................5 Development Rationale for Multilateral Investment Guarantee Agency Non-Honoring of Financial Obligations Products ................................................................................................................6 3. Evaluation Approach...................................................................................................................................8 Approach and Data Collection Methods .............................................................................................8 Definition of Terms ......................................................................................................................................8 4. Evaluation Findings .....................................................................................................................................9 How Did the Non-Honoring Products Enhance the Multilateral Investment Guarantee Agency’s Additionality and Development Effectiveness? ...............................................................9 How Well Did the Multilateral Investment Guarantee Agency Manage the Deployment of its Non-Honoring Products? ................................................................................................................. 25 5. Summary and Implications for Future Strategy: What’s Next for Multilateral Investment Guarantee Agency’s Non-Honoring Products .................................................................................... 33 How Did Multilateral Investment Guarantee Agency Enhance Its Development Effectiveness and Additionality with the New Non-Honoring Products?............................... 33 How Well Did Multilateral Investment Guarantee Agency Manage the Deployment of its Non-Honoring Products? ....................................................................................................................... 34 Implications for Multilateral Investment Guarantee Agency’s 2021–23 Strategy ................ 35 Bibliography..................................................................................................................................................... 45 Figures Figure 4.1. Non-Honoring Products Helped Grow MIGA’s Business, but Political Risk Insurance Has Contributed More to Volume Growth since 2017 .................................................... 10 ii Contents Figure 4.2. Multilateral Investment Guarantee Agency Accounts for a Small Share of State Obligation Insurance among Berne Union Members but Exceeded Other Multilaterals and ECAs ..................................................................................................................................................................... 12 Figure 4.3. An Overwhelming Majority of NH Projects are in High- and Upper-Middle- Income Countries ........................................................................................................................................... 20 Figure 4.4. Country Risk Ratings Profile of Multilateral Investment Guarantee Agency Non- Honoring and PRI Using S&P, Fitch and/or Moody’s Ratings ......................................................... 22 Figure 4.5. ECAs and Other Multilateral Insurers Provided State Obligation Insurance in Countries Rated Low Credit Risk ............................................................................................................... 24 Figure 4.6. To Manage Non-Honoring Risk to its Portfolio, Multilateral Investment Guarantee Agency Retained a Third of Overall Exposure—a Significant Proportion was Reinsured, Mostly to Private Insurers. ...................................................................................................... 32 Tables Table 2.1. Differences between Multilateral Investment Guarantee Agency Non-Honoring Guarantees and Political Risk Insurance.....................................................................................................2 Table 4.1. Multilateral Investment Guarantee Agency Non-Honoring and PRI Issued, by Strategic Priority Areas, 2011–19 ................................................................................................................. 18 Appendixes Appendix A. Non-Honoring of Sovereign Financial Obligations and Non-Honoring of Financial Obligations by a State-Owned Enterprise Projects, Fiscal Years 2011–19 ............ 49 Appendix B. Glossary of Terms Used ..................................................................................................... 51 Appendix C. Methodology ......................................................................................................................... 54 Appendix D. Profile of Projects Insured by MIGA’s Non-Honoring of Financial Obligations Products ............................................................................................................................................................ 59 Appendix E. Comparison of MIGA Non-Honoring of Financial Obligations Products with Non-Payment Guarantee Products of Selected Institutions ......................................................... 80 Appendix F. Comparison of Attributes of World Bank Group Guarantee Instruments....... 95 Appendix G. Sovereign Defaults and Credit Ratings, Various Countries, 1990–2018 ......... 98 iii Abbreviations BU Berne Union CPF Country Partnership Framework E&S environmental and social ECA export credit agency EDC Export Development Corporation EPC engineering, procurement, and construction FCS fragile and conflict-affected situations FY fiscal year IBRD International Bank for Reconstruction and Development IDA International Development Association IEG Independent Evaluation Group IFC International Finance Corporation KfW Kreditanstalt für Wiederaufbau MIGA Multilateral Investment Guarantee Agency MLT medium- and long-term NH Non-Honoring insurance products, composed of NHFO-SOE and NHSFO NHFO-SOE Non-Honoring of Financial Obligations by a State-Owned Enterprise NHSFO Non-Honoring of Sovereign and Sub-Sovereign Financial Obligations PRI political risk insurance SOE state-owned enterprise All dollar amounts are US dollars and all years in World Bank Group fiscal year, unless otherwise indicated. iv Acknowledgments This report was prepared by a team in the Finance and Private Sector Department of the Independent Evaluation Group (IEG) led by Aurora Medina Siy, including Paul Mudde, Mitko Grigorov, and Doruk Yarin Kiroglu. Andrew Stone and Stephan Wegner gave valuable advice. Emelda Cudilla, Marylou Kam-Cheong, and Rosemarie Peña provided critical administrative support. Xiaoxiao Peng and Ma. Roanna Sayes helped improve the graphs and figures used in this report. The team greatly appreciates the cooperation of the management and staff of the Multilateral Investment Guarantee Agency, the International Finance Corporation, and the World Bank, and that of the private insurers, brokers, and representatives of international banks and other stakeholders who generously gave their time and shared their valuable knowledge and insights. IEG is especially grateful for the expert advice, guidance, and support of peer reviewers Christophe Bellinger, Julie Martin, Mark Roland Thomas, and Daniel Villar. Their insightful questions sparked critical thinking. The team appreciates their availability to answer questions and clarify issues throughout the evaluation exercise. The report was conducted under the encouragement, guidance, and support of Stoyan Tenev (senior manager), Marialisa Motta (manager), José Carbajo Martínez (former IEGSP director), Carmen Nonay (IEGSP director) and Alison Evans (Director-General, IEG). v Overview The introduction by the Multilateral Investment Guarantee Agency (MIGA) of its credit enhancement products, the first in 2009 and the second in 2013, was an important development in the institution’s evolution —a step needed to grow MIGA’s business and catch up with the rest of the credit and investment insurance market. In assessing MIGA’s 10-year experience with its Non-Honoring of Sovereign, Sub-sovereign and State-Owned Enterprises Financial Obligation (as the credit enhancement products are called) insurance products, the Independent Evaluation Group (IEG) reviewed the extent to which the availability of these products has enhanced MIGA’s development impact and additionality. IEG also reviewed the extent to which MIGA achieved its business objectives in deploying the non-honoring (NH) products. Mindful of the provision in MIGA’s Convention that its operations must be self-sustaining, the evaluation also considered the risk and sustainability implications of the NH products to MIGA’s operations. This report provides the first independent evaluation of MIGA’s experience with the NH products against the expectations set at the time of their introduction and considering their subsequent evolution. Background In 2009 and 2013, MIGA introduced two new insurance products offering credit enhancement through coverage against non-honoring or nonpayment of financial obligations of public sector borrowers, mainly from debt financing, linked directly or indirectly to public sector investment projects. The first product aimed at insuring international banks or lenders against the Non-Honoring of Sovereign and Sub- Sovereign Financial Obligations (NHSFO). The second product sought to deal with the risk of the Non-Honoring of Financial Obligations by a State-Owned Enterprise (NHFO- SOE). These NH products were expected to boost support to MIGA’s strategic focus areas; namely, investments in International Development Association (IDA) countries and Africa, postconflict countries, South-South investments, and the financing of complex projects. MIGA’s additionality with the new products was expected to be the strongest in high- risk environments and in poorer countries. By reinsuring a significant share of its NH risk exposure to private insurance companies, MIGA would also crowd-in insurance capacity from private insurers to support investments in riskier markets—while managing country and project limits. vii Overview Additionally, the NHFO-SOE coverage was expected to encourage state-owned enterprises (SOEs) to adopt good practices in corporate governance and international best practices in environmental and social sustainability. A defining feature of MIGA’s two NH products is that they do not require an arbitral award or court judgment for payment of claim. This feature, which carries a risk for MIGA higher than its political risk insurance (PRI) business line, was expected to incentivize international banks to finance public sector undertakings by government entities that are less known or do not have adequate access to medium- and long-term financing from international financial markets. As a result, it would help public sector borrowers diversify their financing sources and improve their financing conditions through longer loan tenors and lower cost of borrowing. Before the introduction of the NH products, demand for MIGA’s PRI business had been stagnant because of several factors: (i) the 1997 Asian financial crisis, (ii) new capital adequacy rules on credit risk applied after 2004, and (iii) increasing demand by cross- border investors and lenders for comprehensive cover, which were already available from private insurers, and later a range of multilateral insurers, began offering. Most public sector insurers and export credit agencies already offered comprehensive cover— including political and commercial risks— and state obligation insurance (the other terms used by the credit and investment insurance market for NH insurance) before MIGA sought approval of the NH insurance in 2009. Since MIGA issued its first NH cover in 2011, the amount of insurance issued reached $11.3 billion in 2019 for 34 projects in 14 countries. The NH insurance comprises over a third of MIGA’s outstanding gross exposure and a third of MIGA’s guarantee amounts issued. How Did MIGA Enhance Its Development Effectiveness and Additionality with the New Non-Honoring Insurance Products? Development Outcomes Six of seven evaluated NH projects were rated satisfactory for their development outcomes—measured in terms of project business success, economic sustainability, environmental and social effects and contribution to private sector development. The positive outcome ratings were primarily owing to attainment of project objectives and the environmental and social effects at the project level, which confirm MIGA’s role and contribution to improved environmental and social sustainability. Economic sustainability had the weakest rating among the development outcome indicators. The dependence on government subsidies, the lack of ex post information on the actual public finance implications, debt servicing and debt sustainability effects are viii Overview several reasons for this weak rating. Understanding the actual development effectiveness of NH products would be strengthened by assessing the effects on the ultimate project beneficiaries and poverty reduction, in the case of the NH-insured commercial loans to public sector borrowers. Thus, the current methodology used in evaluating development outcomes of NH projects requires recalibration suitable to the public sector nature of NH projects. Benefits to Borrowers The NH products have helped sovereign, sub-sovereign, and SOE borrowers meet their need for long-term financing. The products allowed public sector borrowers to augment and diversify public and private financing sources to fund their priority needs. The NH products have enabled public borrowers to secure repayment periods for commercial loans that are longer than those available without MIGA’s NH insurance. This evaluation could not determine—because of lack of sufficient evidence—whether the NH insurance products have led to lower interest rates for public sector borrowers. This reflects the difficulty in quantifying and attributing the extent to which commercial banks pass on the benefits of the NH products to their public sector borrowers. In cases in which public sector borrowers have negotiating leverage, and when overt or covert competitive bidding has occurred among international lenders, it is possible that some of the capital enhancement benefits of MIGA’s NH insurance lowered the interest rate passed on to the public sector borrowers. The timing of the issuance of NH insurance also matters, especially in instances when MIGA is brought into the project to cover additional work after lending terms have already been negotiated and MIGA has no option to influence them. All-in cost to the sovereign, sub-sovereign and SOE borrowers also includes the MIGA premium. The absence of clear evidence made it difficult to assess whether or not MIGA’s NHFO- SOE insurance influenced public sector borrowers to increase transparency and disclosure or follow good corporate governance, as per expectations set out in the 2013 NHFO-SOE document. MIGA conducts intensive integrity checks on the management of the project enterprise as part of its due diligence process, and ensures that such practices are standard requirements in MIGA guarantee contracts and in the loan agreements between the group of lenders and the public sector borrower. However, the evaluation could not confirm if and how MIGA improved the corporate practices of the SOEs involved through its NHFO-SOE product. Benefits to Lenders Lenders could offer longer repayment periods to public sector borrowers because of the zero-risk weighting of the credit risk of MIGA-backed loans using Basel II and Basel III ix Overview capital adequacy rules. Because MIGA is recognized as a highly rated multilateral development bank under the Basel II and Basel III bank regulations, banks do not have to set aside risk capital for the loan amount insured by MIGA’s NH. This benefit lowers the banks’ cost of lending. Additionally, MIGA-insured banks package and syndicate the MIGA-insured loans to other participating banks or in the secondary market, thereby lowering the banks’ loan portfolio risk while earning additional revenues from selling the packaged loans. Although the guarantee holders of record are all MIGA repeat clients, the NH products have also brought in midtier international banks as first-time MIGA clients, opening new business opportunities for MIGA, and for some of the new clients. According to the feedback provided by lenders, they value the unique strength that MIGA derives as a member of the World Bank Group, which provides an umbrella of deterrence against government actions that could disrupt projects. Contribution to the Multilateral Investment Guarantee Agency Growth and Strategic Priority Areas The NH insurance products have also contributed to MIGA’s growth, which was helpful as demand for traditional PRI tapered off after 1997. From 2011–19, MIGA’s NH insurance products accounted for over a third of its total guarantee business. MIGA’s NH insurance enabled the gross volume of MIGA’s guarantee business to grow by an average of 11 percent annually between 2011 (the year in which MIGA issued its first Non-Honoring of Sovereign Financial Obligations or NHSFO contract) and 2019. The NH insurance products crowded-in an estimated $7.6 billion in private sector financing and insurance capacity from private insurers to support public sector projects. MIGA applied its NH insurance products to support several projects in areas of strategic focus in Africa or IDA countries, fragile and conflict-affected situations (FCSs), and other subsequent priority areas. But PRI is the instrument that has been deployed more to MIGA’s focus areas, especially in IDA and FCS countries and more recently, to climate change adaptation and risk mitigation. The NH projects have generally taken place in upper-middle- and high-income economies (86 percent by total amounts issued), which have better access to financing from international and domestic commercial sources, international capital markets, and funding from bilateral and multilateral development institutions. The NH insurance has facilitated relatively large projects in infrastructure and financial markets. It has supported several projects with innovative features for MIGA that have potential demonstration effects. x Overview How Well Did the Multilateral Investment Guarantee Agency Manage the Deployment of Its Non-Honoring Insurance Products? MIGA ensured that its NH products would be attractive to international lenders and help grow the business. MIGA expanded the investments eligible for coverage. It extensively used provisions in the MIGA Convention to get Board authorization to cover potential types of investments so that it could respond quickly to business opportunities. It marketed the NH products to large and sophisticated banks that were former PRI repeat clients and to midtier international banks. It expanded its business development efforts to large engineering, procurement, and construction companies involved in public sector projects and the lenders financing them and partnered with export credit agencies. MIGA coordinated and collaborated with the International Bank for Reconstruction and Development and the International Finance Corporation in finding opportunities for joint financing; received help with appraisal, assessment, and monitoring of some NH projects; and used an exposure exchange agreement to swap International Bank for Reconstruction and Development risk. Although the new NH products have increased the volume of MIGA guarantees, they also come with a higher risk profile. Specifically, the NH insurance cover deals with a comprehensive failure by a public borrower to meet its payment obligations, regardless of cause. In effect, MIGA assumes the credit risk of the sovereign, sub-sovereign, or SOE borrower. As such, the NH products, especially the NHFO-SOE, present a greater risk of potential financial loss, capital impairment, or both, than PRI does. In this context, MIGA proceeded cautiously in launching its NH products. The first internal guidance (2012) for MIGA’s incurring NH exposure required a minimum sovereign credit rating of B for obligors to become eligible for a MIGA guarantee. Obligors with lower credit ratings were required to provide additional risk mitigation to merit eligibility. Over time, the minimum sovereign credit rating required by MIGA to be eligible for NH coverage has edged upward. To manage risk, country, and project exposure limits and to avoid seeking additional capital from shareholders as its NH business grew, MIGA reinsured 73 percent of its NH risk exposure to private (and some public) credit and investment insurers. Such reinsurance also freed up additional insurance capacity for MIGA and mobilized private sector monies. So far, MIGA has not experienced any claims related to its NH cover and recovery of claims has not been tested. This higher risk profile of the NH business line needs to be xi Overview carefully considered, as MIGA grows the NH portfolio and if the institution considers expanding the NH coverage for cross-border investments in IDA and FCS countries. Implications for MIGA’s 2021–23 Strategy Experience to date indicates that MIGA NH insurance products are more suited for upper-middle and high-income countries and have been deployed less frequently in support of MIGA’s current strategic focus of facilitating investments in IDA and FCS countries. Legitimate concerns about minimizing the risk of financial loss, and capital impairment linked to the NH products, constrain MIGA from implementing such strategic intent. Lowering MIGA’s sovereign credit rating eligibility criteria by a notch (B+) or two (B) may not necessarily result in a significant expansion of NH coverage to IDA and FCS countries either. The IDA and FCS countries in those credit rating grades, according to external credit rating agencies include heavily indebted poor countries, which cannot readily service additional debt. Strict limits on commercial borrowing by low-income countries under the proposed International Monetary Fund-World Bank Debt Sustainability Framework hinders future commercial borrowings by IDA and FCS countries. In considering what is next for the NH products given experience and MIGA’s 2021–23 strategy’s focus on IDA and FCS countries, the findings from this evaluation invite MIGA’s management to address the following issues: • How might NH products be designed to better contribute to MIGA’s corporate commitment to IDA and FCS countries? • To what extent could MIGA’s current minimum credit rating eligibility criteria be modified to enhance the strategic usefulness of the NH products? • If MIGA concludes that the NH products cannot be made to work in IDA countries and FCSs, how should this operational stance be best reflected in MIGA’s business strategy, rules, and targets for future deployment of the instrument? • How can the evaluation methodology for assessing the development impact of NH projects fit their public sector nature and capture the public finance implications and the economic impact of NH commercial loans on the final beneficiaries and their poverty reduction impact, especially for projects in IDA and FCS? Addressing these questions provides MIGA’s management a draft blueprint to review the role of its NH products. The evaluative evidence and discussion presented in this xii Overview report will help define these roles—and the expectations attached to them—regarding their contribution to MIGA’s business activity, risk exposure, strategic focus, and development mandate. xiii Multilateral Investment Guarantee Agency Management Comments The Multilateral Investment Guarantee Agency (MIGA) welcomes the meso evaluation by Independent Evaluation Group (IEG), Multilateral Investment Guarantee Agency’s Experience with Non-Honoring of Sovereign, Sub-Sovereign, and State-Owned Enterprise Financial Obligation Guarantees. We wish to express appreciation to IEG for preparing this rich report. Management values many of the observations presented in the meso evaluation—and across several areas of engagement by MIGA. These will be noted and useful, especially as MIGA moves forward with its efforts to enhance and increase its development impact in its next three-year strategy cycle (fiscal years [FY]21–23) and beyond. In the following key areas raised in the meso evaluation, management wishes to share its perspectives. Applying the Non-Honoring Product Considering Debt Sustainability Concerns and Multilateral Investment Guarantee Agency’s Financial Sustainability The meso evaluation raises the issue of the applicability of the Non-Honoring (NH) product to International Development Association (IDA) and fragile and conflict- affected situation (FCS) countries, two focus areas of MIGA’s current strategy with the other being climate finance—areas that will continue to be a focus in MIGA’s FY21–23 strategy update. As the meso evaluation acknowledges, there are challenges to deploying the NH products in IDA and FCS countries—specifically, whether it is appropriate for IDA and FCS countries to take on nonconcessional debt to fund public sector projects. Given their level of development, typically, these sovereigns, sub- sovereigns, and state-owned enterprises (SOEs) would be better served by the development community if they had access to highly concessional financing at long maturities—that is IDA-type financing—for funding public projects. IDA financing, by design, is aligned with the debt servicing capacity and limited resilience of these economies in the face of potential shocks. In fact, IDA was established in 1960, 16 years after the creation of International Bank for Reconstruction and Development, to address concerns that the poorest countries could not afford to borrow at the institution’s nearer commercial terms. Early in the practical application of the NH product, MIGA recognized this very challenge. MIGA also acknowledged that, owing to the higher probability of default at the riskier credit rating categories of most IDA and FCS countries, the use of NH guarantees in support of nonconcessional debt instruments to these countries was not aligned with MIGA’s own financial sustainability. As a result, MIGA carefully reviewed the extent to which guaranteeing commercial financing xiv Multilateral Investment Guarantee Agency Management Comments provided to IDA and FCS countries was appropriate both from a development perspective at the country level and in view of MIGA’s own financial sustainability. After careful analysis, in 2015, MIGA reset the benchmark rating threshold below which it would not provide NH guarantees to a sovereign, sub-sovereign, or SOE. It is important to note that this does not preclude IDA or FCS countries from benefitting from the NH product; instead, it ties availability to a credit rating that reflects an adequate ability and willingness to service this external debt on commercial terms. Considering the current concern of the development community about the excessively high debt levels in low-income and developing countries—for example, the latest data by the International Monetary Fund indicates that 45 percent of these countries in Sub- Saharan Africa are in debt distress or at high risk of debt distress (IMF 2019, 6)—MIGA’s decision seems reasonable if not prescient.1, 2 Communication to Shareholders on the Evolution of the Non-Honoring Product The meso evaluation’s finding that the NH product was not typically deployed in IDA or FCS countries was by MIGA management design and was also communicated clearly to MIGA’s shareholders and even to its Council of Governors as this approach evolved.3 MIGA communicated how it was approaching the product’s application to IDA and FCS countries, and countries where there may be credit concerns. Early in the history of the NH product MIGA provided to the Board of Executive Directors its plan to use the political risk insurance (PRI) product rather than the NH product to serve IDA and FCS countries (see figure MR.1). It is particularly noteworthy that in December 2016 the periodic review of MIGA’s activities to its Council of Governors—a review mandated by the MIGA convention—MIGA put forward clearly what it saw as the benefits of the NH product and indicated that the NH product would support investments in middle- income countries. 1 Note that the 45 percent excludes high- and middle-income countries in Sub-Saharan Africa. 2The meso evaluation makes this very point (23). It maintains that even if MIGA lowered its sovereign credit rating criteria, thereby expanding NH coverage to a greater number of IDA and FCS countries, this might not be feasible because (i) supporting additional commercial lending to heavily indebted poor countries would not be aligned with efforts by the International Monetary Fund and World Bank to limit these countries’ external borrowings, and (ii) increasing exposure to IDA and FCS countries creates more risk and costs for MIGA. 3For example, “but PRI is the instrument that has been deployed more to MIGA’s focus areas, especially in IDA and FCS countries” (x). xv Multilateral Investment Guarantee Agency Management Comments Figure 1.1. Multilateral Investment Guarantee Agency’s Strategic Choice by Product Source: MIGA 2014. Improving the Non-Honoring Product’s Impact on Client Countries As MIGA enters an intense period of discussions with shareholders and stakeholders on its upcoming three-year strategy update (FY21–23), MIGA management welcomes IEG’s invitation to assess how the application of the NH product can be made more impactful and how MIGA can continue to increase its development impact using the NH product—including in areas that are most likely to help in attaining the twin goals and the Sustainable Development Goals. As noted in the meso evaluation, MIGA has a track record of pioneering new applications of the NH product, and MIGA expects to continue its leadership in product innovation while retaining a focus on developing countries’ debt servicing capacity and MIGA’s financial health. Review of the Non-Honoring Product and Implications for its Future Deployment The rating threshold for application of the NH product need not be seen as a constraint for MIGA in achieving its strategic objectives; instead, MIGA’s deployment of its two products is complementary to the strategy, including in the current areas of strategic focus, and represent a prudent and responsible approach to achieving development impact across MIGA’s full mandate. MIGA’s present strategy update (FY18–20) has three areas of priority focus: IDA, FCS, and climate. Aligned with the Forward Look and xvi Multilateral Investment Guarantee Agency Management Comments its convention, MIGA is also committed to serving all clients and to ensuring its own financial sustainability—without which it cannot serve the development community effectively. The current application of MIGA’s two product lines, the NH suite of products and PRI, provide coverage across MIGA’s client countries. MIGA’s PRI product is available to private investors operating in almost all of MIGA’s client countries, although the type of coverage offered by this product—transfer restriction and currency inconvertibility, expropriation, breach of contract, and war and civil disturbance—are most sought after by private investors in countries where there is a history of political violence; where there is lack of a track record of economic, political and legal reforms that support private cross-border investments; and where institutional capacity is low to implement such reforms. The NH product, as currently made available, is applicable to countries with an appropriate level of creditworthiness and, thus, is well suited to support many middle-income countries. MIGA is also endeavoring to target the use of the NH product to support its efforts to increase its climate finance work—an important strategic priority, including in middle-income countries. Objectives of the Non-Honoring Product at Inception The track record of the NH product to date provides credible and tangible evidence of the value of the product across the original objectives communicated to MIGA’s Board. The 2012 paper submitted to the Board set forth the rationale for management’s recommendation of the NH product and specified three distinct areas of MIGA’s contribution that the product would enhance: (i) the product would bring positive development results; (ii) it would do this especially in MIGA’s strategic focus areas, of which six were specifically listed; and (iii) it would be able to do more business and strengthen its position as a financially sustainable institution (World Bank 2012, ii; 6). The strategic areas of focus provided in the 2012 Board paper were IDA countries and Africa in particular, frontier markets, postconflict countries, South-South investments, and complex projects. It is important to note that the 2012 Board paper encompassed additional reforms to MIGA’s operations, some of which were needed to support MIGA’s introduction of the NH product, while these and other changes were also intended to support the growth of MIGA’s PRI business.4 Together, the introduction of the NH product and reforms that buttressed the growth of the PRI product would support the three distinct objectives. 4Additional reforms included (i) enhanced breach of contract coverage; (ii) enhanced war and civil disturbance coverage; (iii) clarifying aspects related to project implementation and acquisition of existing assets; and (iv) addressing eligibility and other issues. xvii Multilateral Investment Guarantee Agency Management Comments MIGA’s assessment of the performance against each of these three objectives for the NH product alone is as follows: i. Positive development results: Of the seven NH projects validated by IEG, six of these (or 87 percent) were evaluated to have had development outcomes assessed as successful by IEG. This provides the most granular assessment of the effectiveness of the NH product to deliver positive development results. ii. Projects in strategic focus areas: MIGA’s analysis indicates that 68 percent of the 34 NH guarantee projects considered in the report were aligned with at least one of the original six strategic focus areas, with 35 percent of the 34 projects aligned with two or more strategic areas.5 iii. More business: MIGA’s NH guarantees enabled the gross volume of MIGA’s guarantee business to grow by an average of 11 percent annually between 2011 (the year in which MIGA issued its first NH contract) and 2019. These guarantees succeeded in directly mobilizing private finance of $7.6 billion. The NH product was never meant to be targeted only to IDA or FCS countries as might be inferred from portions of the meso evaluation. MIGA does not agree with the meso evaluation’s conclusion that “managing institutional risk brought about by the NH products became MIGA’s main concern and the driver of MIGA’s overall approach” (23), and the evaluation presents no evidence to support it. In contrast, the positive development results achieved (as indicated in the previous paragraph) provide evidence that factors other than “institutional risk” were important inputs into MIGA’s decision to support NH projects. Evaluation Methodology for Non-Honoring Guarantee Projects MIGA appreciates the invitation from IEG to recalibrate the joint IEG-MIGA evaluation guidelines for NH projects.6 MIGA views these guidelines as a living document and so would find the incorporation of additional learnings and insights from the meso 5If Climate Finance were to be considered a strategic priority area, the percentage would be even higher at 71 percent. 6The meso evaluation asks “how can the evaluation methodology for assessing the development impact of NH projects fit their public sector nature and capture the public finance, debt servicing and debt sustainability implications and the economic impact of NH-insured commercial loans on the final beneficiaries and their poverty reduction impact, especially for projects in IDA and FCS? How can the ex post development effectiveness evaluation methodology capture the debt sustainability implications of projects that have a series of NH guarantee contracts with the same obligor, project enterprise or host country?” (33) xviii Multilateral Investment Guarantee Agency Management Comments evaluation, or any other source, a beneficial exercise. Given this connection, MIGA would like to share that i. The current joint IEG-MIGA evaluation methodology for NH guarantee projects was based on extensive IEG-MIGA discussions that explicitly recognized the public sector nature of NH projects and resulted in the issuance of a Policy and Procedure Framework document in January 2016. The joint IEG-MIGA Working Group discussed the applicability of the guidance methodology for private sector investments (like PRI projects) to NH guarantee projects given their typically public sector nature. The working group noted that NH guarantee projects, notwithstanding their public sector nature, have to be financially sustainable with sufficient funds either from user charges or budget resources. ii. The joint IEG-MIGA evaluation guidelines for both PRI and NH products were updated very recently and issued as a Policy and Procedure Framework document in July 2019 after long dialogue and discussion with IEG, with no suggestion of the need to amend the guidelines during this comprehensive review. iii. IEG had not voiced the need for revising the evaluation guidelines for NH guarantee projects during the evaluation and validation of the NH guarantee projects evaluated to date. With regard to the specific recommendation on debt sustainability, MIGA would welcome capturing the debt sustainability implications of NH projects in its ex post evaluation framework. As IEG is aware, MIGA does extensive work on assessing the ability of the beneficiary of the MIGA-guaranteed loan to service its existing debt and the additional debt envisaged in the NH project prior to moving ahead with a NH transaction. Debt sustainability issues have been analyzed in all MIGA NH project documents. Ex post information on debt servicing requirements would typically be incorporated in countries’ projected debt service forecasts and, as such, is an indicator that MIGA carefully monitors its frequent review of a country’s NH credit rating. It is also noteworthy that loans supported by MIGA’s NH products are typically small relative to total general government debt. Except for one large export-earning NH guarantee project, MIGA-guaranteed debt did not account for more than 2.5 percent of a country’s total general government debt. Limitations of the Meso Evaluation’s Findings and the Purpose of a Meso Evaluation IEG did not interview “sovereign, sub-sovereign and SOE borrowers and counterparties regarding their experience with MIGA’s NH products” (54). This was contrary to what xix Multilateral Investment Guarantee Agency Management Comments was indicated in the Concept Note prepared for the meso evaluation, which included an outline of the evaluation scope, approach, and data collection methods for the meso evaluation. The Concept Note indicated that IEG intended to interview “at least half of the sovereign guarantee, sovereign, and state-owned enterprise borrowers.”7 Interviews with sovereigns, sub-sovereigns, and SOE borrowers may have helped provide evidence for some of the questions posed by IEG for which it indicates it lacked sufficient evidence—specifically whether the NH insurance products led to lower interest rates, whether MIGA’s Non-Honoring of Financial Obligations by a State-Owned Enterprise (NHFO-SOE) insurance influenced public sector borrowers to increase transparency and disclosure and follow good corporate governance, and whether projects would have happened without MIGA’s NH guarantees. In addition, IEG does not seem to have interviewed other groups or subgroups of stakeholders identified in the Concept Note, or to the extent originally envisaged. For example, guarantee holders were interviewed—but fewer than one-third of the interviews IEG originally indicated in its Concept Note were completed. The use of interviews was especially important to this meso evaluation given that only seven evaluations of NH projects had been completed to date. The interviews were intended to be a primary data source informing three of the four main evaluation questions posed by IEG and a secondary data source in the fourth question. Another limitation is that the list of the specific guarantee holders contacted by IEG was not shared with MIGA, as indicated in IEG’s Concept Note.8 IEG had advised that meetings with guarantee holders would be channeled through MIGA. The failure to do so limited MIGA’s ability to cross check the information provided and to determine whether the information provided to IEG was appropriately calibrated by its evaluation team. IEG indicates that meso evaluations use “a nimble processing protocol to maximize responsiveness and value added for the World Bank Group.”9 The flexibility of the meso evaluation protocol facilitated IEG’s ability to proceed with the meso evaluation even though only 7 of the 34 NH guarantee projects had been evaluated with the result that detailed and granular assessments were available only for a fraction of the NH projects. However, the knowledge gaps that inevitably result from a processing protocol of this kind should be made clear in the text of the meso evaluation. For example, the meso 7 Concept Note for the meso evaluation, November 2018, page 9. 8In the Concept Note for the meso evaluation, IEG states that it would “channel its request for meetings with guarantee holders…to MIGA” (9). 9 See IEG’s website https://ieg.worldbankgroup.org/evaluations. xx Multilateral Investment Guarantee Agency Management Comments evaluation notes dependence on government subsidies as a reason for weak ratings for some evaluation dimensions (viii). However, this claim applied to only one of the seven projects evaluated; moreover, the subsidy applied to official support for export credits that was fully consistent with relevant Development Assistance Committee rules and practices. More generally, MIGA notes that the meso evaluation’s conclusions are often asserted without supporting evidence or are based on generalizations from very limited information. It may well be that this is the nature of meso evaluations, which aim to be work distilled by IEG to be “just in time.” However, readers may not be sufficiently aware of the nature of meso evaluations and may confuse the approach taken here with IEG’s other evaluation approaches, in which findings are more typically based on a large pool of completed project-level evaluations. As such, MIGA finds it odd that the term “meso evaluation” appears only once in IEG’s meso evaluation and that the nature of this type of evaluation and its emphasis on learning is not explained in the main text of the document or in the appendix on methodology. As such, readers may not understand how the meso evaluation product differs from other IEG evaluation products, including its evidentiary basis, and may not then put its conclusions and findings in the appropriate perspective. Meso evaluations are intended to have a strong focus on learning.10 However, because the meso evaluation took as its starting point the assessment of “MIGA’s additionality, development effectiveness and experience in deploying the NH product against expectations presented to the Board in 2009 and 2013” (51), MIGA believes that stakeholders may have missed out on the opportunity to benefit from additional learning that may have materialized had a wider lens than an “accountability lens” been applied. This is especially so, considering MIGA’s active and deliberate development of the NH product over time across many dimensions—risk management, strategic priorities, development impact, and application across beneficiary countries. MIGA management is of the view that a broader methodological and theoretical lens would have resulted in more useful findings for input into MIGA’s strategic discussion. Areas of the Report Needing Greater Clarification Claim of ample private capacity in the credit insurance market, including for sovereign credits rated MIGA’s minimum credit risk rating criteria. In the context of exploring MIGA’s additionality, the meso evaluation states that there is ample private capacity in the credit insurance market to insure bankable projects in developing countries, including those where the sovereign credit rating is lower than MIGA’s 10 See IEG’s website https://ieg.worldbankgroup.org/evaluations. xxi Multilateral Investment Guarantee Agency Management Comments minimum credit risk criteria. The meso evaluation cites an amount of $900 billion to be the estimated combined capacity of private insurers. MIGA notes several concerns with the meso evaluation’s interpretation of what appears to be a single brokerage report and questions whether the information cited is useful in framing conclusions about MIGA’s additionality. The assessment of private capacity in the insurance market in the brokerage report appears to be based on the maximum amounts of insurance per project that insurers offer and misses the fact that this is typically an amount available only to the highest-rated credits, that is, those in the A ratings bracket. As indicated in appendix D of IEG’s meso evaluation, the sovereign credit rating by the external credit rating agencies at time of issuance for MIGA projects never exceeded BBB. The meso evaluation provides no evidence of ample private insurance capacity in the single B ratings range. Comparison of country risk ratings of MIGA-supported projects by NH and PRI. Figure 4.4 in the meso evaluation provides a comparison of the rating agencies’ sovereign credit ratings of the host countries of MIGA projects by product—the sovereign credit ratings of countries where MIGA’s PRI projects are located is compared with the sovereign credit rating for MIGA’s NH projects. It is worth noting that sovereign credit ratings assess a different risk (creditworthiness) than that covered under PRI (political risk) and, as a result, sovereign credit ratings are not an adequate proxy for PRI ratings. In addition, even within PRI ratings, there are four different applicable ratings for each of the separate points of cover—and these ratings may differ significantly, reflecting the different risks of each PRI cover type. MIGA may have covered one, two, three, or all four of the points in any PRI project. For these reasons, any conclusions drawn from this graph, including in the adjoining text of the meso evaluation, are based on an unreliable framework of analysis. Direct mobilization figures for MIGA’s NH product. The meso evaluation states that MIGA’s NH product crowded in $7.6 billion in private monies through direct mobilization or reinsurance. The meso evaluation obtains this number by summing up $3.8 billion in private direct mobilization between 2016 and 2018 and $3.7 billion mobilized in 2011–15, including the amount of NH exposure reinsured. However, MIGA estimates the correct amount for private direct mobilization between 2011 and 2018 at $7.6 billion of which reinsurance is not included as per the methodology agreed by the multilateral development banks for calculating private direct mobilization. (It would appear to be a coincidence that the overall amount from IEG is the same as the actual private direct mobilization amount.) If reinsurance were included for the period 2011– 18, MIGA estimates the applicable amount at $14.4 billion (composed of $7.6 billion in private direct mobilization plus $6.8 billion of reinsurance). xxii Multilateral Investment Guarantee Agency Management Comments Evidence of whether MIGA’s NHFO-SOE product influenced public sector borrowers with respect to transparency or good corporate governance. The meso evaluation observes an absence of clear evidence whether MIGA’s NHFO-SOE insurance influenced public sector borrowers to increase transparency or follow good corporate governance. MIGA notes that in the case of the only two NHFO-SOE guarantee projects that have been evaluated by IEG to date, evidence regarding improvements in corporate governance have been confirmed. In one guarantee project, the evaluation found that the SOE achieved high standards of corporate governance and encouraged innovation, conveying additionality beyond financing. In the other guarantee project, the IEG assessment found that the SOE has continuously supported knowledge transfer to staff through training in different areas to enhance good governance, ethics, and values. MIGA also notes that the Agency has been very selective in providing the NHFO-SOE coverage and has declined projects where compliance with MIGA’s integrity standards was an issue. xxiii 1. Introduction 1.1 This report presents the first independent evaluation of the two most recent products of the Multilateral Investment Guarantee Agency (MIGA): the Non-Honoring of Sovereign and Sub-Sovereign Financial Obligations (NHSFO) and the Non-Honoring of Financial Obligations by a State-Owned Enterprise (NHFO-SOE). These two products were approved by the MIGA Board of Executive Directors in 2009 and 2013, respectively. Their introduction was supported by significant operational policy changes that made it possible for MIGA to offer the non-honoring of financial obligations (NH) guarantees or insurance. MIGA considered the introduction of these NH products crucial for enhancing its development impact, financial sustainability and growth, and relevance in the global credit and investment insurance market. 1.2 Since their introduction, MIGA’s NH products have become a substantial share of MIGA’s overall business. As of 2019, they make up over a third of MIGA’s outstanding gross exposure (22 percent in outstanding net exposure amounts), and a third of the amount of MIGA guarantees issued since 2011, supporting 34 projects in 14 countries, primarily in the finance and infrastructure sectors (appendix A presents the list of NH projects from fiscal year 2011 to 2019). 1.3 The review by the Independent Evaluation Group (IEG) of MIGA’s experience to date with the NH products seeks to provide feedback, draw lessons, and help inform MIGA’s 2021–23 strategy. The evaluation findings also provide insights on how credit enhancement products can mobilize private capital to finance development projects, which is at the core of the World Bank Group’s Maximizing Finance for Development agenda and Forward Look 2030 Strategy. This review also aims to provide information about the proposed role for MIGA in the October 2018 report, “Making the Global Financial System Work for All” by the G-20 Eminent Persons Group on Global Financial Governance.1 1 2. Multilateral Investment Guarantee Agency’s Non-Honoring of Financial Obligations Guarantee Products 2.1 MIGA offers two kinds of risk mitigation products: political risk insurance (PRI) and credit enhancement cover. MIGA’s PRI covers the following four risks: (i) transfer restriction and currency inconvertibility, (ii) expropriation, (iii) war and civil disturbance, and (iv) breach of contract. MIGA’s credit enhancement cover provides insurance against the risk of Non-Honoring of Sovereign Financial Obligations (NHSFO) or NHFO-SOE. The NHSFO and NHFO-SOE differ from MIGA’s PRI regarding several key characteristics of these products (table 2.1). Table 2.1. Differences between Multilateral Investment Guarantee Agency Non- Honoring Guarantees and Political Risk Insurance Characteristics NHSFO and NHFO-SOE Political Risk Insurance Purpose Credit enhancement Political risk mitigation Nature of project Purely public sector undertakings or Mainly private sector; project finance projects Ownership of project National or subnational governments; Mainly private companies; may also include state- company state-owned enterprises and banks owned enterprises and banks; public insurers Eligible investments Debt (nonshareholder loans); quasi- Equity; debt (shareholder and nonshareholder covered equity; swapsa loans); loan guarantees; quasi-equity; swapsa; other nonequity direct investmentsb MIGA-insured party (or Lenders: international banks; bilateral Equity investors (private and public), lenders guarantee holder) development institutions; public (private and public; company or individuals); insurers; state-owned banks and bilateral development institutions; public insurers; enterprises state-owned enterprises Covered risk Nonpayment risk of unconditional and Currency inconvertibility and transfer restriction; irrevocable obligation by public expropriation; war, terrorism, and civil borrowers disturbance; breach of contract MIGA country risk With minimum credit ratingc Not applicable rating eligibility criteria 2 Chapter 2 The Multilateral Investment Guarantee Agency’s Non-Honoring of Financial Obligations Guarantees Characteristics NHSFO and NHFO-SOE Political Risk Insurance Claim resolution 180 day waiting period; only on the For breach of contract, requires an arbitral mechanism scheduled principal and interest process, an arbitral award rendered, and host payment government failure to comply with the arbitral decision Sectors (based on Finance and Capital Markets; All sectors projects supported as Infrastructure of FY19) Sources: MIGA Convention, Operational Policies, and MIGA directives and guidance for underwriting NHSFO, NHFO-SOE, and political risk insurance. Note: FY = fiscal year; MIGA = Multilateral Investment Guarantee Agency; NHFO-SOE = Non-Honoring of Financial Obligations by a State-Owned Enterprise; NHFSO = Non-Honoring of Sovereign and Sub-Sovereign Financial Obligations; PRI = political risk insurance. a. Swaps and other hedging transactions entered in connection with interest rate and currency risks with a tenor of more than a year are eligible for MIGA coverage if such transactions are related to an underlying investment covered by MIGA or a swap transaction on a stand-alone basis (that is, without MIGA coverage of the underlying transaction). See Article 1.10(b) of MIGA Operational Policies, effective January 6, 2015. b. Other nonequity direct investments eligible for MIGA coverage include production-sharing contracts; profit-sharing contracts; management contracts; franchising agreements; licensing agreements; operating leases; engineering, procurement, and construction contracts, turnkey contracts, and related performance bonds. See MIGA’s Operational Policies, effective January 6, 2015. c. Based on MIGA’s NHSFO and NHFO-SOE credit risk rating. Background 2.2 When MIGA was established in 1988, its Convention explicitly authorized coverage only to cross-border investors against risk of currency inconvertibility and transfer restriction;1 expropriation; war, terrorism, and civil disturbance; and breach of contract.2 Collectively, this was called PRI or noncommercial risk coverage. MIGA’s PRI product primarily supports cross-border or foreign investments by private sector entities into developing countries. The NHSFO cover was approved by the Board in 2009 as part of the amendments to MIGA’s Operational Regulations.3 It provides protection against losses resulting from the failure of a sovereign or sub-sovereign (that is, city, municipality, or region) to make a payment when due under an unconditional and irrevocable financial payment obligation or guarantee related to a MIGA-insured investment.4, 5 The NHSFO product supports purely public sector undertakings or projects. 2.3 An example of NHSFO is MIGA coverage of an unconditional and irrevocable loan to the government of Panama from three banks led by Citibank (as the facility agent and the MIGA guarantee holder of record) to finance an engineering, procurement, and construction contract for the construction, supply, and installation of Panama Metro’s Line One. Metro Secretariat/Metro de Panama SA is the implementing agency,6 but payment of the syndicated loan is guaranteed by Panama (acting through the Ministry of Economy and Finance). With the NHSFO product, if the Ministry of Economy and Finance cannot pay the scheduled payment to the banks, the banks can file a claim for 3 Chapter 2 The Multilateral Investment Guarantee Agency’s Non-Honoring of Financial Obligations Guarantees the scheduled due payments from MIGA, which it is bound to pay after an agreed period. 2.4 Breach of Contract, a PRI product offered by MIGA, also insures against contractual breach, including payment default by the government or a state-owned enterprise (SOE); but if the government breaks its contractual obligation, claiming compensation from MIGA involves several steps. It requires that (i) an arbitral process is initiated, and an arbitral award rendered (arbitral award) and (ii) the government fails to comply or delays the arbitral award’s enforcement (see appendix B for a glossary of these insurance terms).7 In the previous example of the Panama Metro, if the project is structured as a public-private partnership with conditionalities regarding the payment of the loan, and the lenders purchased Breach of Contract coverage instead, the banks would have to go through the entire arbitration process and get a judgment in their favor to be paid. Only if the government does not comply, or there is considerable delay in enforcing the arbitral award decision, can MIGA pay the claim. In contrast with the Breach of Contract coverage, the NHFO does not require the investor to obtain an arbitral award to file a claim for compensation to MIGA. 2.5 The NHSFO product applies to situations when a financial payment obligation by the sovereign or sub-sovereign to a lender is unconditional and irrevocable, and not subject to defenses other than the nonpayment of the underlying debt, whether because of inability or unwillingness to pay.8 In effect, MIGA assumes the credit risk of the sovereign and sub-sovereign. MIGA assumes the credit default risk of the sovereign and sub-sovereign. Therefore, through the NHSFO product, MIGA is providing coverage against noncommercial risks, which includes default risk. 2.6 The NHSFO cover was introduced in 2009 along with other amendments to MIGA’s Operational Regulations. Four years later, in June 2013, MIGA sought Board approval for a new coverage, the NHFO-SOE. The NHFO-SOE coverage relates to an unconditional and irrevocable financial obligation or guarantee provided by a SOE. The NHFO-SOE is a hybrid between MIGA’s NHSFO and Breach of Contract-SOE risk covers.9 Both NHSFO and NHFO-SOE provide coverage or insurance against risk of nonpayment (or non-honoring) of unconditional and irrevocable financial obligations to a lender, typically an international bank or financial institution. It does not require an arbitral award. 4 Chapter 2 The Multilateral Investment Guarantee Agency’s Non-Honoring of Financial Obligations Guarantees Business Rationale for Multilateral Investment Guarantee Agency Non-Honoring of Financial Obligations Products 2.7 Before the NHSFO product was introduced, MIGA’s PRI business had been stagnant for years. Global demand for PRI had dropped significantly after the Asian currency crisis in 1997 and the introduction of a new capital adequacy framework in 2004 under the Basel II accord. Since the 1997 Asian currency crisis, private and public insurers and other multilateral insurers increased their offering of comprehensive cover—insurance of commercial, credit and political risks. -Export credit agencies (ECAs) have been providing both insurance and financing for cross-border medium- and long-term (MLT) trade-related transactions to sovereigns since these were established. Demand from debt financiers for PRI, particularly from international banks, had been declining as the type of financial instruments and investments needs were changing. MIGA, established in 1988 to provide a monoline product (PRI as the core business),10 faced declining demand and relevance of its PRI product. To grow and remain relevant, MIGA needed to expand its product offerings to catch up with private and public insurers that were offering a broader range of risk mitigation products. This was necessary for MIGA to continue delivering its development mandate. 2.8 The NH products were expected to be most beneficial to international private banks that provide loans to sovereigns, sub-sovereigns, and SOEs. They offer several advantages. Foremost, loans with MIGA guarantees are accorded zero-risk weights under the Basel II and Basel III capital adequacy frameworks. Simply stated, private banks do not have to set aside and use capital to provision for the credit risk of the MIGA-insured loan. This regulatory relief allows banks to improve capital efficiency and frees up bank capital that otherwise would have to be set aside, which in turn lowers the cost of financing. And because MIGA typically insures 95 percent of the guarantee holder’s loan amount,11 the expectation is that banks will then provide better terms by way of lower loan pricing and/or longer tenors to public sector borrowers. Second, MIGA’s appetite for MLT risk allows banks to extend loan tenors offered to borrowers. Third, as a member of the Bank Group, MIGA’s involvement has an element of deterrence against sovereign risks that is valued by lenders and private insurers. This provides an extra layer of comfort to banks and private insurers that public sector borrowers will not renege on the MIGA-backed loans or, in the event of nonpayment, lenders and private insurers perceive that MIGA’s access to governments, as part of the Bank Group, will greatly facilitate recovery from the sovereign. Lenders and other private insurers interviewed by IEG considers the umbrella of deterrence offered by MIGA as a member of the Bank Group as its comparative advantage over the sovereign nonpayment product offered by other insurers. 5 Chapter 2 The Multilateral Investment Guarantee Agency’s Non-Honoring of Financial Obligations Guarantees Development Rationale for Multilateral Investment Guarantee Agency Non-Honoring of Financial Obligations Products 2.9 MIGA’s NH products offered potential developmental win-win solutions to public borrowers, the international lending banks, and MIGA. These products, which do not require an arbitral award judgment, are intended to enhance the bankability of public sector projects. This incentivizes international banks to finance public sector undertakings and enter into loan agreements with government entities in developing countries and emerging markets that are less known or do not have adequate access to MLT financing from international financial markets. Hence, the NH products increase development project opportunities for sovereign borrowers, sub-sovereign entities, and SOEs. The NH products were also expected to generate business opportunities for international banks, especially midtier international banks that want to expand their activities in developing countries and emerging markets. 2.10 MIGA’s NH products were expected to endow the host country with several benefits. These include (i) diversified financing sources made available to government entities; (ii) access to improved financing terms through longer loan tenors (maturity periods) and lower cost of funds that,12 if passed on to borrowers, would reduce the cost of borrowing for public sector entities in developing countries; and (iii) redirecting private resources to underserviced markets and to relatively unknown government entities without access to private financing and ready to borrow to finance their priority development projects. 2.11 For MIGA, the NH products, along with other operational policy changes, would boost support to its strategic focus areas (that is, International Development Association [IDA] countries and Africa, frontier markets, postconflict countries, South-South investments, and complex projects) and achieve positive development results. They would enhance MIGA’s development impact by supporting projects that are well designed; follow international best practices; are economically, environmentally, and socially sustainable; and are thus highly developmental. MIGA’s additionality was expected to be most helpful in high-risk environments and in poorer countries because of its development mandate, its ability to take higher risks compared with private insurers (thanks in part to its multilateral status), and because it is a member of the Bank Group. By reinsuring a significant share of its NH risk exposure to private insurance companies, MIGA would also crowd-in insurance capacity from private insurers to support investments in riskier areas. 2.12 Additionally, the NHFO-SOE coverage was intended to encourage SOEs to adopt improved standards of business conduct. These include increased transparency 6 Chapter 2 The Multilateral Investment Guarantee Agency’s Non-Honoring of Financial Obligations Guarantees and disclosure; good practices in corporate governance; anti-money laundering, anticorruption, antifraud, and international best practices in environmental and social (E&S) sustainability. 7 3. Evaluation Approach 3.1 This evaluation assessed the extent to which the introduction of the NH products helped MIGA enhance its additionality, development impact and achieve its business objectives as expressed in the 2009 and 2013 documents seeking MIGA Board authorization to offer the NHSFO and NHFO-SOEand considering their subsequent evolution (appendix D provides details of IEG’s evaluation approach and methodology).1 3.2 IEG reviewed all 34 NH projects issued by MIGA from FY11 (when MIGA issued its first NHSFO guarantee) to FY19. Of the 34 NHFO-insured projects, 8 have completed ex post evaluations, 7 of which have been rated.2 To situate the NH business line within the context of MIGA’s overall operations, the evaluation also reviewed the portfolio trends of its PRI business during the same period but did not conduct an in-depth review of the PRI projects. Approach and Data Collection Methods 3.3 In addition to the project portfolio analysis and the review and analysis of relevant policy and strategy documents, IEG carried out a benchmarking exercise to compare MIGA NH products with similar or complementary products by the International Bank for Reconstruction and Development (IBRD), IDA, International Finance Corporation (IFC), African Development Bank, Asian Development Bank, an export credit agency, and a private insurer. IEG also conducted semistructured interviews with key stakeholders to fill data and information gaps. These key stakeholders included MIGA directors and staff, select IFC and IBRD staff, and representatives of international banks, private insurers, brokers, and ECAs. Definition of Terms 3.4 Appendix B presents the definition of financial and technical terms used in this report. The report interchangeably uses the terms insurance, cover, coverage or guarantees and international banks, banks, and lenders. 8 4. Evaluation Findings How Did the Non-Honoring Products Enhance the Multilateral Investment Guarantee Agency’s Additionality and Development Effectiveness? 4.1 The NH products contributed to a marked increase in MIGA’s guarantee business. When MIGA first sought Board approval in 2009 of a set of policy changes and amendments to its Operational Regulations,1 including the NHSFO, the growth of its business had slowed for nearly a decade (figure 4.1). Between the issuance of the first NHSFO cover in 2011 and 2019, MIGA’s overall guarantee issuance grew 11 percent annually by volume (compounded annual growth rate) compared with the 2 percent annual growth rate nine years before the NH products were introduced. As of 2019, MIGA had issued NH coverage amounting to $11.3 billion for 34 public sector projects in 14 countries. Europe and Central Asia dominated with 42 percent of the volume of NH coverage issued, primarily in Turkey. With the insurance of a large project in Oman in FY19, the Middle East and North Africa region received 17 percent of MIGA NH amounts issued. Africa represented 20 percent of all NH guarantees amounts, with over half of the guarantees issued to SOE and state-owned bank projects in South Africa. 4.2 Underwriting of large investments contributed to the expansion of MIGA’s guarantee business (by gross exposure amount) in 2018 and 2019 for both NH and PRI products. Fiscal years 2016 and 2017 experienced the highest number of NH projects but in terms of amounts issued, fiscal years 2018 and 2019 outpaced the earlier years. The amounts of NH guarantees issued (on a gross basis) in 2019 doubled from the previous year, but this substantial increase is due to one large project in Oman.2 The potential of MIGA’s NH business to grow further as currently designed is an issue that requires careful consideration. Similarly, MIGA’s PRI business grew markedly in 2017, 2018 and 2019 because of its underwriting of large investments (over $1 billion in gross amount issued) in each of these years to complement smaller guarantee amounts issued. 9 Chapter 4 Findings Figure 4.1. Non-Honoring Products Helped Grow MIGA’s Business, but Political Risk Insurance Has Contributed More to Volume Growth since 2017 PRI NH $6,000 Exposure amounts issued $5,000 $4,000 ($, millions) $3,000 $2,000 $1,000 $0 Fiscal year Sources: Multilateral Investment Guarantee Agency Contracts Issued and Cancelations data; Multilateral Investment Guarantee Agency Audited Annual Reports, various years; and Multilateral Investment Guarantee Agency’s external website: www.miga.org/projects. Note: This graph presents the guarantee amounts issued by MIGA for each fiscal year, which includes new guarantees issued and additional exposure amounts. MIGA = Multilateral Investment Guarantee Agency; NH = non-honoring of sovereign, sub sovereign and state-owned enterprises financial obligations; PRI = political risk insurance.MIGA’s NH portfolio is highly concentrated by sectors, type of NH product, obligors, and guarantee holders of record. Infrastructure projects, including construction, oil and gas pipeline, power and transport, account for 67 percent of NH insurance amounts issued. The rest of the NH portfolio (33 percent) consists of insured commercial loans to state-owned banks primarily for on-lending to small and medium enterprises. By type of NH product, 27 were NHSFO contracts with guarantee from a sovereign obligor, usually the ministry of finance (19 projects), or with subnational entities (in six projects). The remaining seven projects were each insured by a NHFO-SOE contract with a SOE as the borrower. Four of these were with one state-owned export-import bank. With one exception, the NH guarantee holders of record were international private banks. Two private international banks, as lead arrangers of the syndicated loan insured by MIGA NH, received more than half (55 percent) of MIGA’s NH amounts (on a gross basis) issued. Aside from large international private banks, MIGA also provided NH coverage to midtier banks and several public financing institutions and insurers that participated in the syndicated loans covered by MIGA’s NH guarantees. 4.3 MIGA’s NH products crowded-in an estimated $7.6 billionin private monies to support priority public sector projects through private direct mobilization or reinsurance.3 Of that amount, $3.8 billion was mobilized as private direct mobilization for the period 2016–18. The other $3.6 billion mobilized included the amount of NH exposure reinsured for the period 2011–15, before MIGA adopted the multilateral development banks’ agreed methodology for estimating amount mobilized from the private sector. MIGA reinsured 73 percent of its NH gross exposure, mostly with private insurers, and it had not experienced any claims to date. 4.4 With the introduction of NH products MIGA caught up with other multilateral and private insurers, and ECAs. MIGA was a late adopter of the NH product. A PRI market study commissioned by MIGA in 2008 identified sovereign non-honoring insurance as one of the coverages that private insurers and ECAs were already offering. 10 Chapter 4 Findings Since the 1980s, private insurers and ECAs have already offered sovereign non-honoring coverage, although the Berne Union (BU) data only start in calendar year 2010. Figure 4.2. The Multilateral Investment Guarantee Agency Accounts for a Small Share of State Obligation Insurance among Berne Union Members Other multilaterals MIGA ECA Private Share of MIGA NH volume to total state obligation insurance volume 20 14 Share of MIGA NH volume (%) 12 15 New business amount 10 ($, billions) 8 10 6 4 5 2 0 0 2010a 2011 2012 2013 2014 2015 2016 2017 2018 Calendar year 4.5 , which presents only the state obligation insurance (as sovereign nonpayment coverage is called) new business by BU Investment Insurance Committee members, shows that MIGA has an 8 percent average share of this insurance since the institution issued its first NH guarantee contract in December 2010. This share is significant, especially if compared with other BU multilateral insurers (African Trade Insurance Agency and Islamic Corporation for the Insurance of Investment and Export Credit)4 that have limited geographical reach and therefore operate in smaller markets. Private insurers are among the top providers of this cover by the amount of new business issued, and still account for nearly 80 percent of the state obligation insurance issued by BU members in calendar years 2010 to 2018. 11 Chapter 4 Findings Figure 4.2. The Multilateral Investment Guarantee Agency Accounts for a Small Share of State Obligation Insurance among Berne Union Members Other multilaterals MIGA ECA Private Share of MIGA NH volume to total state obligation insurance volume 20 14 Share of MIGA NH volume (%) 12 15 New business amount 10 ($, billions) 8 10 6 4 5 2 0 0 2010a 2011 2012 2013 2014 2015 2016 2017 2018 Calendar year Source: Berne Union Investment Insurance data for calendar years 2010 to 2018 and IEG calculations. Note: Data relate to state obligation by Berne Union Investment Insurance Committee members only and presents a partial representation of the credit and investment insurance market. Data exclude state obligation insurance provided to Category One and some Category Two MIGA member countries that are nonborrowing members of the International Bank for Reconstruction and Development. The declining business volume of ECAs starting 2013 reflects the shift in the reporting of trade-related state obligation insurance from the Berne Union Investment Insurance Committee to its Export Credit Committee and is not due to decreasing business. ECAs = export credit agencies; MIGA = Multilateral Investment Guarantee Agency. a. Data for second half of calendar year 2010. 4.6 Some NH projects have pioneering or innovative features for MIGA, although their development outcome is sometimes difficult to ascertain. The Panama Metro project, for example, is the first metro in Central America and MIGA’s first NH project in Latin America. Another MIGA innovation relates to its NH coverage of the state-owned Land and Agricultural Bank of South Africa, where the US dollar denominated loan insured by MIGA was accompanied by a by a swap in local currency (South African rand).5 This is the first time that MIGA insured a local currency swap. TNHFSO coverage of the stand-alone Senegal US$/Euro Cross-Currency Swap is also considered a MIGA innovation because it is the first time that MIGA covered a stand-alone swap transaction that is not linked to an underlying investment or MIGA-insured loan. However, IEG’s evaluation of the Senegal swap revealed the difficulty in identifying and quantifying the economic benefits of the improved foreign currency risk management of public debt resulting from the swap. Assessing the development outcomes of stand- alone swaps is challenging because MIGA is unable to monitor the use of proceeds 12 Chapter 4 Findings resulting from the swap, irrespective of whether the stand-alone swap is done through a loan or bond format. 4.7 MIGA NH products have value-adding features to lenders and public sector borrowers, especially in helping them to meet and optimize their financing strategies. Stakeholders interviewed by IEG affirmed the value-added of MIGA’s NH products in terms of (i) benefits of the zero solvency requirements under Basel II and Basel III regulations, which allowed banks to use capital efficiently and freed up capital for additional lending; (ii) ability to offer longer loan tenors to government borrowers; and (iii) opening new opportunities for several midtier banks to participate in the syndicated loans covered by MIGA NH guarantees. According to the feedback provided by lenders and private insurers, they value the unique strength that MIGA derives as a member of the Bank Group, which provides an umbrella of deterrence against government actions that could disrupt projects. The insured banks could pool, package, and syndicate the MIGA-insured loans to other participating banks, or sell them in the secondary market, thereby managing their loan portfolio risk and earning revenues. 4.8 NH products helped public borrowers get financing for priority projects to supplement available funding sources. The syndicated loans insured by MIGA NH supplemented official development assistance and concessional financing from development finance institutions, borrowings from domestic and other international banks and proceeds from bond issuance in the international and (in a few countries), domestic capital markets. For example, the government of Bangladesh had prioritized the expansion and modernization of the country’s power generationand had drawn financing for this purpose from multiple sources such as loans from international commercial banks,6 including syndicated loans covered by MIGA NH guarantees to four government-owned power plants, and from development finance institutions such as the Asian Development Bank, IDA, IFC, and bilateral aid agencies. 4.9 Development outcomes of evaluated NH projects were rated satisfactory. Six of the seven evaluated NH projects had satisfactory development outcomes, which are measured in terms of project business performance, economic sustainability, E&S effects, and contribution to private sector development. The positive outcome ratings were primarily owing to attainment of project objectives and the E&S effects, which confirm MIGA’s role and contribution to improved environmental and social sustainability. Economic sustainability had the weakest rating among the development outcome indicators. The dependence on government subsidy, and the lack of ex post information on the actual public finance implications, debt servicing, and debt sustainability effects are several reasons for this weak rating. Understanding the actual development effectiveness of NH products would be strengthened by assessing the effects on the 13 Chapter 4 Findings ultimate project beneficiaries and poverty reduction impact, in the case of the NH- insured loans to public sector borrowers in IDA and fragile and conflict-affected situation (FCS) countries. Thus, the current methodology used in evaluating development outcomes of NH projects requires recalibration suitable to the purely public sector nature of NH projects. 4.10 Results from the evaluated NH projects confirmed MIGA’s positive role and contribution to improving environmental and social effects at the project level. Most of the evaluated NH projects were rated satisfactory in terms of their E&S effects. Compliance with MIGA’s E&S requirements is a requirement under MIGA’s contracts of guarantee, which specifies compliance with the Performance Standards, E&S requirements, and the required periodic reporting with which the lender and borrower must comply. In MIGA’s projects, the E&S metrics that must be complied with (some before the disbursement of the MIGA-insured loan) and monitored are listed in the MIGA guarantee contracts. These E&S requirements are often repeated verbatim in the facility agreements between the lenders (international banks) and the obligor and the intercreditor agreements among the different lenders (and ECAs) involved in the syndicated loan insured by MIGA. 4.11 Several examples indicate weak leverage by the MIGA guarantee holders (international banks) to enforce and follow-up on the reporting expected from the government counterparty in the MIGA contract of guarantee. The most challenging to track are the on-lending from the NH loans that finance the credit facilities of state- owned banks. Although this challenge is not unique to MIGA’s NH product, the ability of the lender or guarantee holder to enforce MIGA reporting requirements on a public entity can get complicated, especially if the obligor is not the implementing agency. 4.12 The absence of clear evidence made it difficult to assess whether or not MIGA’s NHFO-SOE coverage influenced public sector borrowers to increase transparency and disclosure or follow good corporate governance, as per expected additionality of the NHFO-SOE product. MIGA conducts in-depth integrity checks (“know-your-customer” due diligence) at underwriting and good corporate governance practices are covenanted in every MIGA contract of guarantee for both NH and PRI contracts, repeated verbatim in the loan agreements,7 and while the guarantee is active, monitored closely by a dedicated unit in MIGA. Nevertheless, determining attribution of SOE practices to MIGA’s NHFO-SOE insurance is difficult to ascertain and quite challenging to ensure, whether by MIGA or the lenders.8 At most, MIGA could cancel the contract of guarantee in case of noncompliance with the agreed covenants, which happened in an NHSFO- insured project when the borrower could not commit in writing to undertake all the 14 Chapter 4 Findings actions required by standard MIGA covenants related to auditing, reporting, and E&S matters. 4.13 Given the products’ risk profile, MIGA’s country and project risk assessments at underwriting are more extensive in NH projects with robust internal processes and specialized staff. MIGA conducts extensive credit risk assessment by assigning credit rating to each sovereign specifically for the risk that MIGA faces if it decides to support a project through its NH product. MIGA’s internal rating threshold, which it applies to all NH transactions, explicitly recognizes that MIGA supports only public sector borrowers that are creditworthy and therefore capable of servicing their underlying debt. MIGA’s Economics Unit, staffed by risk management officers, assesses MIGA’s own sovereign NH ratings on a continuous basis and reviews all NH risk ratings on a quarterly basis. Rigor is ensured in the NH rating assessments through MIGA’s internal processes, including voting by MIGA’s Credit Committee, quarterly presentation to MIGA management team, annual audits by MIGA’s external auditors and consultations with the World Bank and IFC. A new unit in MIGA was created to assess the creditworthiness of SOEs, including debt sustainability. The unit is led by an experienced credit risk professional under the Director of Finance and Risk. Nevertheless, from an underwriting perspective, PRI projects are usually more complex and document intensive. 4.14 For seven of the eight projects evaluated, the downstream effects of the MIGA- insured loan on the government’s finances, debt servicing and sustainability, were not assessed.9 This shortcoming is in part because the ex post evaluation framework is based on private sector operations rather than on public sector ones. Going forward, adequate assessment of the NH projects’ economic and public finance implications will be especially important in projects that have a series of NH contracts to the same obligor, enterprise, or country. 4.15 The intrinsic benefits of MIGA’s NH products were manifested mainly through the longer loan tenors that the international banks passed on to the sovereign, sub- sovereign, or SOE borrowers. MIGA’s NH products have mobilized longer-term loans from private banks t than would have been available without MIGA’s NH cover. The average tenor of the NH-insured loans and MIGA’s NH cover was 11 years. In one project, MIGA’s NH cover extended the loan repayment period to 12 years instead of 5 to 7 years repayment period without the MIGA cover. 4.16 The effect of MIGA’s NH coverage on lower loan pricing, and the pass-through of lower interest rates from the lenders to the sovereign, sub-sovereign, or SOE borrowers could not be confirmed. Apart from the challenge of determining a suitable counterfactual for MIGA’s NH cover, MIGA has not provided IEG with the information 15 Chapter 4 Findings nor the loan agreements between the banks and the borrowers. Therefore, IEG was unable to verify the effect of the MIGA NH cover on the various components of the loan pricing and effective interest rates to the borrowers. Still, there are some examples of lower pricing being passed on to the public sector borrowers. In the São Paulo Transport project, for example, IBRD’s involvement (apart from the MIGA NH cover)10 required competitive bidding by international banks for the commercial loan portion of the project so that the São Paulo municipality and government of Brazil received the best terms. In a NHFO-SOE project, the insured lender provided financing through its promotional loan program, which had an embedded concessional element within it. In other NH projects, the lower loan pricing resulted from the government leverage on the lenders to reduce bank margins or spreads. 4.17 MIGA’s ability to influence the loan pricing is low in the case of loans for supplemental works or when MIGA’s NH insurance is sought after the terms of the loan have already been agreed. For example, MIGA’s NH insurance covered a commercial bank’s loan for supplemental work that was ineligible for additional coverage by an ECA, which financed and insured a substantial portion of the commercial bank loan for the project. In this project, the terms of the loan insured by MIGA had to match the terms of the ECA-insured loan. Similarly, MIGA’s NH coverage of a subnational project became effective when the commercial bank loan was already in its third year of disbursement; thus, MIGA’s NH guarantee could not alter the loan pricing agreed between the municipality and the commercial bank.11 4.18 MIGA used the NH products to support cross-border investments in its strategic focus areas but more so with its PRI product. In seeking Board authorization in 2009 and 2013, MIGA intended the NH products, along with other changes, to support cross- border investments in Africa and IDA and FCS countries—in high-risk contexts where NHFO support can have the most value—and South-South investments and complex projects.12 Since 2011, the NH products added $2 billion to MIGA’s business in IDA countries, more than $2 billion in Africa (although mostly in an upper-middle-income country), and $500 million to a FCS country in FY13. But as seen in table 4.1, the percentage share of Africa and FCS and IDA countries in MIGA’s NH business is significantly lower than the corresponding shares in its PRI business line. Also, none of the NH projects involved South-South investments. MIGA support for IDA, FCS countries, and South-South investments was largely through its PRI product. 4.19 MIGA’s NH products supported complex projects in the energy, extractive, and other infrastructure sectors, although these were straightforward to underwrite. One in three NH projects were characterized as complex projects-–-a MIGA strategic focus area until 2017. However, MIGA’s definition of complex projects is somewhat tautological 16 Chapter 4 Findings and broad. It encompasses projects that involve project finance, structured finance, other multisourced deals, and those dealing with challenging E&S considerations, including projects in the infrastructure sectors and extractive industries.13 Such a definition gives MIGA ample latitude to classify any project as complex; however, from an underwriting perspective, the EPC transactions financed by the commercial loans insured by MIGA NH are straightforward to structure and underwrite. 4.20 Subsequent strategies shifted MIGA’s priority areas but support to foreign investments in IDA and FCS remained constant. PRI remains the instrument of choice to support MIGA’s priority focus areas, including climate change adaptation and mitigation. Table E.3 (appendix E) presents the evolution of MIGA’s priority areas based on its strategies since its FY09–11 Strategy, which was in place when the NHSFO product was approved in 2009. Notwithstanding shifts in focus areas, supporting investments to IDA and FCS countries remains a consistent priority, including in MIGA’s current FY18–20 strategy. Table 4.1 shows that MIGA has supported climate change adaptation and mitigation, its third priority area in the current strategy, through its PRI product. It should be noted that the climate change adaptation and mitigation priority area is a small percentage in terms of number of projects, but the gross amount issued is a sizable amount. 4.21 The NH projects are clustered in high and upper-middle-income economies, which continue to have access to financing from commercial and bilateral and multilateral development institutions. Sixty-eight percent of the NH projects (by project count) are in high and upper-middle-income countries at the time of issuance, compared with 42 percent of the PRI projects.14 The difference between MIGA’s two business lines is more pronounced in terms of the total amounts issued in each of the four country income levels ( 17 Chapter 4 Findings Figure 4.3. An Overwhelming Majority of Non-Honoring Projects are in High- and Upper-Middle-Income Countries Low income Lower-middle income Upper-middle income High income 100% 90% Share to total amount issued (%) 80% 70% 60% 50% 40% 30% 20% 10% 0% Non-honoring Political risk insurance 4.22 ). By amounts issued, an overwhelming majority of NH projects (86 percent) are in high-income and upper middle-income countries compared with half (53 percent) for the PRI projects. Projects in these more developed economies also tend to be larger (for both NH and PRI products). Table 4.1. Multilateral Investment Guarantee Agency Non-Honoring and PRI Issued, by Strategic Priority Areas, 2011–19 Total Gross Instruments and Priority Gross Amountc Amount Projects Areasa,b ($, millions) (%) (no.) NHSFO and NHFO-SOEd 11,266 34 Africa (2011 –17) d 2,276 20 9 Complex Projects 2,963 26 13 (2011d–17) FCS (2011d-Present) 512 5 1 IDA (2011 -Present) d 2,026 18 10 South-South 0 0f investments (2011d–17) Climate Change 2,309e 20 3 Adaptation and Mitigation (2017- Present) 18 Chapter 4 Findings Total Gross Instruments and Priority Gross Amountc Amount Projects Areasa,b ($, millions) (%) (no.) Partial (below 100%) 2,307f 20 2 Full (100%) 2 0 1 Not elsewhere classified 6,319 56 12 PRI 22,154 11,038 274 Africa (2011–17) 5,588 25 107 Complex Projects (2011– 4,210 19 36 17) FCS (2011-Present) 3,012 14 56 IDA (2011-Present) 8,828 40 131 South-South 1,884 9 38 investments (2011–17) Climate Change 2,086 9 44 Adaptation and Mitigation (2017- Present) Partial (below 100%) 307 1 3 Full (100%) 1,780 8 41 Not elsewhere classified 11,199 51 93 Sources: Multilateral Investment Guarantee Agency Contracts Issued database and Multilateral Investment Guarantee Agency strategy documents, various years. Note: Based on MIGA strategic focus areas mentioned in the 2009 Board document seeking approval of the NHSFO product and MIGA’s strategic focus areas in 2013, when MIGA sought Board approval of the NHFO-SOE product. FCS = fragile and conflict-affected situations; IDA = International Development Association; NHFO = Non-Honoring of Financial Obligations; NHFO-SOE = Non-Honoring of Financial Obligations by a State-Owned Enterprise; NHSFO = Non-Honoring of Sovereign and Sub-Sovereign Financial Obligations; PRI = political risk insurance. a. MIGA’s 2021–23 Strategy identified IDA, FCS, and Climate Change Adaptation and Mitigation as its three focus areas. b. Projects may be classified in more than one priority area, and the count is not mutually exclusive. Sum of amounts issued will exceed total NH and PRI issuance and percentages will exceed 100 percent because projects may be classified in more than one priority area. c. Gross amounts represent MIGA’s maximum aggregate liability at issuance.Net amount represents insured amount retained by MIGA after reinsurance. d. Although the non-honoring instrument was approved in FY09, the first project was not issued until FY11. e. One NH project was considered consistent with the Climate change adaptation and mitigation priority area based on 8.245 percent of the project meeting this MIGA priority area; the other NH was also considered consistent with this MIGA priority area because 15-percent of the project meets the criteria based on MIGA’s calculation. f. The large amount is because of a single project.. MIGA estimated 15.5 percent of the NH amount issued is consistent with Climate Adaptation, hence the large number. 19 Chapter 4 Findings Figure 4.3. An Overwhelming Majority of Non-Honoring Projects are in High- and Upper-Middle-Income Countries Low income Lower-middle income Upper-middle income High income 100% 90% Share to total amount issued (%) 80% 70% 60% 50% 40% 30% 20% 10% 0% Non-honoring Political risk insurance Sources: Multilateral Investment Guarantee Agency Contracts Issued database, World Bank Country Income Classification, and Independent Evaluation Group staff calculations. Graph presents new business amount issued from fiscal years 2011 to 2019. 4.23 NH projects are concentrated in creditworthy countries that continue to have access to commercial and development financing, including concessional financing. All 14 countries—including 3 IDA countries—that MIGA supported with its NH cover had issued foreign currency bonds in international markets during the period 2009 to 2019. A NHSFO-insured state-owned bank’s creditors include five international finance institutions, among which are the World Bank and numerous international and domestic banks – in addition to the syndicated loans that MIGA insured under its NHFO-SOE. Using sovereign credit risk ratings by external credit rating agencies, 20 Chapter 4 Findings Figure 4.4. Country Risk Ratings Profile of Multilateral Investment Guarantee Agency Non-Honoring and PRI Using S&P, Fitch and/or Moody’s Ratings 4.24 illustrates the distribution of MIGA’s NH projects by sovereign credit risk ratings compared with the risk profile of MIGA’s PRI product at issuance. The risk profile of MIGA’s NH product shows a clustering in the investment grade (BBB−), noninvestment grade speculative and highly speculative range (BB− to B+). The reasons for the clustering of MIGA NH projects in less risky countries are discussed subsequently. 4.25 By contrast, the country risk profile of MIGA’s PRI business line is more dispersed, although projects with larger insured amounts are clustered about the investment-grade credit ratings. A significant number of PRI projects have low credit risk ratings (rated CCC, otherwise known as extremely speculative). On average, NH projects are much larger than PRI projects (3.5 times as large), but there are outliers in 21 Chapter 4 Findings the PRI chart in figure 4.4. In 2016, and 2017, and as of 2019, MIGA insured three projects under its PRI product, each with guarantee issued amounts of over $1 billion. Figure 4.4. Country Risk Ratings Profile of Multilateral Investment Guarantee Agency Non-Honoring and PRI Using S&P, Fitch and/or Moody’s Ratings Sources: Standard & Poor’s, Fitch, and Moody’s (for the country risk ratings), Multilateral Investment Guarantee Agency Contracts Issued database, and Independent Evaluation Group staff calculations. Note: Shade of the bubbles reflects MIGA’s maximum aggregate liability at issuance. NH = Non-Honoring of Sovereign, Sub-Sovereign, and State-Owned Enterprise Financial Obligations; PRI = political risk insurance; RD = restricted default; S&P = Standard & Poor’s. 4.26 Compared with private insurers offering similar insurance products (state obligation insurance), ECAs and multilateral insurers including MIGA, have supported investments in countries rated higher risk. Private insurers in the BU Investment Insurance Committee issued more state obligation insurance (by new business amounts) in high investment-grade countries compared with the multilateral insurers, including 22 Chapter 4 Findings MIGA. Approximately a quarter of the state obligation insurance new business of BU private insurers are in countries with an A+ sovereign credit risk rating or better ( Figure 4.5. ECAs and Other Multilateral Insurers Provided State Obligation Insurance in Countries Rated Low Credit Risk Above A− BBB+, BBB, BBB− BB+, BB, BB− B+, B, B− Below CCC+ 100% 90% New business amount (%) 80% 70% 60% 50% 40% 30% 20% 10% 0% Private insurers ECAs Multilaterals MIGA Other multilaterals Berne Union investment insurers 4.27 ) and a third are in the BBB+ to BBB- range. By contrast, MIGA’s NH business is primarily (65 percent) in the countries with sovereign credit ratings ranging from BB+ to BB− (investment grade but speculative) range. But 23 Chapter 4 Findings Figure 4.5. ECAs and Other Multilateral Insurers Provided State Obligation Insurance in Countries Rated Low Credit Risk Above A− BBB+, BBB, BBB− BB+, BB, BB− B+, B, B− Below CCC+ 100% 90% New business amount (%) 80% 70% 60% 50% 40% 30% 20% 10% 0% Private insurers ECAs Multilaterals MIGA Other multilaterals Berne Union investment insurers 4.28 also reveals the presence of private insurers, ECAs, and other multilateral insurers in the same space where MIGA has NH exposure, and even in countries rated lower credit risk (B− to B+) where MIGA has minimal NH exposure. Figure 4.5. ECAs and Other Multilateral Insurers Provided State Obligation Insurance in Countries Rated Low Credit Risk Above A− BBB+, BBB, BBB− BB+, BB, BB− B+, B, B− Below CCC+ 100% 90% New business amount (%) 80% 70% 60% 50% 40% 30% 20% 10% 0% Private insurers ECAs Multilaterals MIGA Other multilaterals Berne Union investment insurers Source: Berne Union state obligation insurance data from calendar years 2010 to 2018; Standard & Poor’s, Fitch, Moody’s data, and Independent Evaluation Group staff calculations. 24 Chapter 4 Findings Note: Data relate to state obligation insurance by Berne Union Investment Insurance Committee members only and present only a snapshot of the credit and investment insurance market. ECAs = export credit agencies; MIGA = Multilateral Investment Guarantee Agency. 4.29 The evaluation could not consistently verify the extent to which projects would have happened without the MIGA NH insurance. The additionality of MIGA NH products, based on counterfactuals, is difficult to ascertain – that is, to find out whether other insurers were willing to insure the bank loans albeit at shorter tenors or higher interest rates. An alternative counterfactual is whether the public sector projects could be funded through a concessional loan or perhaps from the public budget. A review of the NH project documents and interview responses point to some instances when MIGA stepped in because other insurers faced country limits or capacity constraints. In one project, for example, MIGA provided the NH cover to the EPC contractor after the export credit agency withdrew, thus allowing the contractor to proceed and complete the delivery of the newly built equipment to the SOE operator. In a power project, MIGA insured the portion of the commercial loan that an ECA could not cover. The São Paulo Transport project is another example in which IBRD sought MIGA’s NH coverage for the commercial loan portion of the project because it had reached its country lending limits for Brazil. 4.30 Though estimates vary, there is ample private capacity in the credit insurance market to insure bankable projects in developing countries, including those where the sovereign credit rating is lower than MIGA’s minimum credit risk criterion. The tenor may be shorter, the uninsured portion higher,15 or the premium rates higher. Private insurers have combined (theoretical) capacity to offer 15 to 20 years tenor for nonpayment public obligor insurance with a maximum per-risk capacity of approximately $3.0 billion per transaction, albeit applied selectively.16 This suggests that MIGA’s additionality, based on a counterfactual, is highly context-specific. In many instances, there are alternatives to MIGA NH products, including from ECAs, multilateral and regional development institutions,17 and private insurers as mentioned in the previous paragraphs (refer to appendix E for a comparison of NH products offered by selected development institutions and appendix F for a comparison of Bank Group guarantee products). How Well Did the Multilateral Investment Guarantee Agency Manage the Deployment of its Non-Honoring Products? 4.31 The NH products increased MIGA’s guarantee business but also exposed it to bigger and different types of risks, especially with its NHFO-SOE product. The NH guarantee is a radically different product and a new business line for MIGA. It goes beyond the PRI for private sector projects or public-private partnerships and offers 25 Chapter 4 Findings comprehensive failure-to-pay insurance to international banks for their loans to public entities to finance purely public sector undertakings or projects. By covering failure-to- pay risk by the public borrower, MIGA assumes the credit risk of the sovereign, sub- sovereign, or SOE, thus exposing MIGA to higher risk of potential financial loss and capital impairment.18 4.32 The NHFO-SOE product presents an even higher risk to MIGA because the product itself implies a conditional guarantee from the government—only to the extent that government has control of the SOE.19 As an example, should the government decide to privatize the SOE, either the NHFO-SOE--covered loan is prepaid in full or MIGA pays the claim. Moreover, SOEs are generally created to operate as separate corporate entities and many are established to operate on a commercial basis thus potentially giving governments the legal basis to avoid SOEs’ liabilities. MIGA has created an SOE unit within its Finance and Risk Management Group to assess the creditworthiness of SOEs and develop an analytical framework, policy and operating guidelines, and reporting and monitoring frameworks for the credit risks assumed by MIGA under this product. In addition, MIGA sets aside higher economic capital consumption for NHFO- SOE projects than for the NHSFO products and covers only state-owned banks that are fully controlled by the government. 4.33 Managing institutional risk brought about by the NH products became MIGA’s main concern and the driver of MIGA’s overall approach. In addition to its development mandate, MIGA is required by its convention to carry out its activities “with sound business and prudent financial management practices” (Article 25 of MIGA Convention).20 The Convention’s preamble counsels MIGA that it “should, to the extent possible, meet its obligations without resort to its callable capital and that such an objective would be served by continued improvement in investment conditions.”21 Second, as an insurer, MIGA has a paid-in capital of $0.4 million and its callable capital from member countries stands at $1.55 billion as of 2019.22 So far, MIGA’s operating capital (total capital available to undertake MIGA’s current and future business activities) suggests adequate capital to meet the guarantee business and economic capital requirements of operational and investment risks. 4.34 MIGA took a deliberate “learning by doing” approach in developing eligibility criteria, policies, and procedures, and adjusted them as staff gained experience with the NH products. It took MIGA a year between the Board authorization of the NHSFO cover in December 2009 and the issuance of the first NHSFO contract of guarantee in December 2010. Meanwhile, MIGA developed its underwriting guidelines and internal risk management policies. Guidelines for the NHFO-SOE took less time to develop because of MIGA’s experience with the NHSFO. Since then, MIGA has regularly 26 Chapter 4 Findings adjusted its eligibility, rating methodology guidelines, and procedures as staff have continued to gain experience and a greater understanding of the implications of the NH products on MIGA’s operations and institutional risk exposure. 4.35 MIGA had initially set the eligibility criteria for NH coverage at a level that opened the potential for insuring commercial loans to public sector projects in IDA and FCS countries. MIGA’s initial underwriting guidance to staff in 2012 required that the obligor have a minimum rating to be eligible. Obligors that have lower credit ratings required additional risk mitigants to be eligible for NHSFO coverage.23 Subsequent guidance to staff in 2015 tightened the minimum credit rating, which meant that only creditworthy countries, mostly upper-middle- and high-income countries, would become eligible for NH coverage. Based on current guidelines, MIGA can insure NHSFO projects and SOEs in countries, one notch lower than MIGA’s minimum rating but the latter must have an explicit government guarantee, and additional strong risk mitigants must be in place. As Figure 4.5. ECAs and Other Multilateral Insurers Provided State Obligation Insurance in Countries Rated Low Credit Risk Above A− BBB+, BBB, BBB− BB+, BB, BB− B+, B, B− Below CCC+ 100% 90% New business amount (%) 80% 70% 60% 50% 40% 30% 20% 10% 0% Private insurers ECAs Multilaterals MIGA Other multilaterals Berne Union investment insurers 4.36 illustrates, MIGA’s revealed risks tolerance has been largely consistent, whereas the guidelines have been applied flexibly. Even before 2015, when the country credit risk threshold was lower, most of the NH projects were in countries above its current minimum credit rating criteria. There are more exceptions to the 2015 guidelines since it was adopted than before its adoption. The sovereign credit ratings below MIGA’s minimum credit rating threshold, from 2015 onward, reflected the country rating downgrades when NH contracts were issued to cover subsequent additional loans to 27 Chapter 4 Findings three projects—two of which are in the Africa region and one in the Europe and Central Asia region. 4.37 MIGA applied its NH products to support a few projects in areas of strategic focus in Africa or IDA countries, FCS countries, and other subsequent priority areas. But PRI is the product that has been deployed to a large extent in MIGA’s focus areas, especially in IDA and FCS countries. MIGA’s concern to minimize risk of financial losses and capital impairment overrode this strategic intent. Even if MIGA lowers its sovereign credit rating eligibility criterion by a notch or two, expanding NH coverage into IDA and FCS countries may not be feasible for two reasons. First, IDA and FCS countries in those risk categories are already heavily indebted poor countries (see appendix G). Supporting additional commercial lending to heavily indebted poor countries contravenes efforts by the International Monetary Fund and the World Bank to limit these countries’ external borrowings. Secondly, increasing exposure to IDA and FCS countries creates more risk and higher costs for MIGA. The ratio of projects in IDA and FCS with preclaims to active projects has increased significantly in the past five years, with MIGA’s cost of managing these preclaims increasing as it takes more risk.24 4.38 From a business standpoint, MIGA successfully encouraged the uptake of NH products. It expanded the types of investments eligible for MIGA coverage in response to new types of investments and increased MIGA’s net exposure limits. MIGA introduced new types of eligible investments including (i) coverage of stand-alone debt;25 (ii) extending the definition of nonshareholder loans to include other financial transactions and forms of debt instruments; (iii) coverage of investments in existing assets, EPC contracts, turnkey contracts, and related performance bonds; (iv) coverage of stand-alone swaps; and (v) redefining MLT investments from a minimum of three years to investments that have tenor of more than one year.26 In 2015, MIGA increased single- country net exposure limits from $720 million to $820 million, single-project net exposure limits were raised to $250 million from $220 million to allow for coverage of larger-size projects. The net exposure limits that took effect July 1, 2019, increased the single-country net exposure limit to $1 billion and $300 million for single-project net exposure limit. These changes equipped MIGA to respond to new types and underwrite larger-size investments. 4.39 In cases in which the proposed investment is not explicitly stated in its policies, MIGA secured Board authorization, on a project-by-project basis, to issue NH contracts. Thus, MIGA was able to push the boundaries of its Convention and Operational Policies to support new types of investments through the NH projects. The current MIGA Convention (ratified in November 2010) did not mention the NHSFO and NHFO-SOE among the eligible risks for MIGA coverage,27 although these risks are listed in the 2015 28 Chapter 4 Findings Operational Policies.28 To insure new types of investments and address policy gaps, MIGA sought Board authorization under Article 11(b) and Article 12(b) of the Convention. With this flexibility, MIGA was able to provide NHSFO coverage to investments in the form of stand-alone swaps, capital market transactions and a NHFO- SOE coverage of a DFI loan to a state-owned development bank. 4.40 To develop the NH business, MIGA marketed the NH products to large and sophisticated commercial banks, mostly to repeat clients that had previous PRI coverage from MIGA. MIGA NH guarantee holders of record are repeat clients who have purchased MIGA PRI in the past. Standard Chartered Bank, for example, had PRI contracts from MIGA for its investments in China, Djibouti, Senegal, and Uganda before its first NHSFO guarantee for the Senegal US$/Euro Cross-Currency Swap and Sirajganj II Power Plant (Bangladesh) in 2016.29 MIGA has since issued several NH and PRI contracts for Standard Chartered Bank’s subsequent investments. Other former clients had not purchased MIGA cover for nearly a decade—because of lack of demand for PRI products—before their first NH contract. Citibank’s NH-insured loan for the Panama Metro, for example, was its first contract with MIGA since a PRI coverage of its investment in Brazil in 2001. Standard Bank’s NH contract for the cross-currency swap in Senegal in 2012 was its first contract with MIGA after a PRI contract in 2007 for its investment in Mozambique. The NH contract was followed by nine successive PRI contracts for its investments in Eastern Africa. Banco Santander’s late involvement in the Panama Metro project, and its first stand-alone NH contract with MIGA for the São Paulo Transport project in 2015, were the client’s new engagements with MIGA since its last PRI contract in 2001. The NH contracts were subsequently followed by several large investments covered (for capital optimization) by six PRI contracts on behalf of the bank’s subsidiaries. The return of past clients demonstrates the latent demand for MIGA’s NH products. 4.41 MIGA promoted the NH product to large international EPC contractors that participate in government projects and to the lenders financing them that require MLT sovereign nonpayment insurance to cover procurement of equipment and services. The demand for MLT trade-related nonpayment coverage of commercial loans by public borrowers is evident in the infrastructure and extractive industry transactions covered by MIGA NH guarantees. Several NH-guaranteed loans in the energy, gas, and transport sectors financed payments by sovereign, sub-sovereign and state-owned enterprises to experienced and large international EPC or supply contractors.30 Some examples include Constructora Norberto Odebrecht SA (Cambambe Hydropower II,31 Panama Metro),32 Daewoo Shipbuilding and Marine Engineering (Tunisia Passenger– Car Ferry TANIT),33 Mitsubishi Corporation (Istanbul Üsküdar-Ümraniye-Çekmeköy Metro),34 and China National Machinery Import and Export Corporation (CMC) and 29 Chapter 4 Findings Alstom Switzerland Ltd. Baden (Ghorasal 3rd Unit Repowering Project).35 These contractors and suppliers also received trade financing and insurance from ECAs. MIGA’s NH-insured loans for power projects covered the portion of EPC contracts that could not get financing or insurance from the ECAs because such an insured portion did not meet the domestic content requirements under the Organisation for Economic Co- operation and Development Arrangement on Officially Supported Export Credits. 4.42 By insuring the commercial loans by sovereigns, sub-sovereigns and SOEs for MLT trade-related investments in public sector projects, MIGA expanded its NH business into the domain of ECAs, signaling complementarity and opportunity for partnerships. Partnerships with ECAs were an untapped market for MIGA, although its mandate allows it to “cooperate and complement the operations of national entities of MIGA members” (Article 19 of MIGA Convention).36 Among the ECAs involved in NH projects are Compañía Española de Seguros de Crédito a la Exportación,; EDC, Hermes, Office National du Ducroire/Nationale Delcrederedienst, Swiss Export Risk Insurance, and Sinosure. These ECAs are also providing MLT export credit financing and insurance to their domestic EPC companies or equipment suppliers providing construction services or equipment to public sector projects. 4.43 MIGA coordinated and collaborated with the World Bank and IFC in finding opportunities for joint financing and in the appraisal and assessment of NH projects. Of the 34 NH projects, 7 have support from IBRD or IFC participation, either through cofinancing, or parallel or sequential joint projects (see table E.5 in appendix E).37 For example, in the Izmir Tramway project (Turkey), IFC’s loan commitment,38 equal in amount to the MIGA NH-insured commercial bank loan, came a few months before the signing of MIGA’s NHSFO contract. In the Istanbul Kadikoy-Kartal-Kaynarca Metro project (Turkey), IFC’s loan was committed two years before MIGA’s NH guarantee.39 In these two projects, MIGA arranged for IFC to conduct monitoring of E&S requirements under the NH guarantee contracts. In the Izmir Light Rail project, IFC and MIGA jointly sought Board authorization to provide loan and guarantee, respectively, to the municipality of Izmir. MIGA issued the NHSFO contract to the commercial lender and IFC has a loan commitment under its account to the municipal government.40 IFC’s and MIGA’s joint support for the Istanbul and Izmir municipalities transport projects are part of the Bank Group Turkey Sustainable Cities framework. As mentioned previously, IBRD and MIGA have joint commitments to the São Paulo Transport project in Brazil (IBRD loan, NHSFO) and the Southern Gas Corridor/TANAP project in Azerbaijan (IBRD loan, IBRD partial risk guarantee, NHSFO).41 4.44 Nearly all NH projects are aligned with Bank Group country strategies and priorities. The three Istanbul and three Izmir projects were envisaged in the Bank 30 Chapter 4 Findings Group’s FY12–15 Country Partnership Strategy and in the FY18–21 Country Partnership Framework (CPF) for Turkey.42 These projects were also reflected in the Bank Group’s Turkey Sustainable Cities Joint Implementation Plan, a road map for an integrated Bank Group approach to providing comprehensive public-private solutions to the pressures brought about by Turkey’s rapid urbanization. MIGA’s support for the Bangladesh power projects is consistent with the CPF objective of increasing the country’s power generation capacity and access to clean energy.43 The NH projects, Ghorasal 3, Sirajganj II, and Sirajganj III, are listed in Bangladesh’s CPF results framework as MIGA activities that would contribute to the achievement of this CPF objective. Two NH-insured projects are in countries where the World Bank does not have a country program and engagement is limited. Among the evaluated NH projects, except for one project, the other seven evaluated NH projects were rated positively for their consistency with Bank Group country strategies. 4.45 There has been greater collaboration with IBRD than with IFC, given the public finance nature of the projects and the sovereign nature of the obligors. Although IFC (through its Sub-National Lending and Cities Programs) and MIGA have cofinanced projects that used the NH guarantees, and the IFC-MIGA joint business arrangement continues to generate potential deal flows, business development in the NH space is increasingly challenging. First, IFC does not provide loans, equity investments, or risk- sharing facilities to sovereigns. To invest in sub-nationals and SOEs, IFC requires that these entities be creditworthy or operate commercially; therefore, IFC does not require a sovereign counter guarantee. This eligibility criterion limits the scope for cooperation. Nevertheless, IFC and MIGA cooperation continues to grow through the IFC-MIGA joint business agreement–but primarily with MIGA’s PRI product. 4.46 IBRD and MIGA devised an innovative exposure exchange agreement that expanded cooperation beyond cofinancing or co-guaranteeing of sovereign projects. IBRD and MIGA entered into an Exposure Exchange Agreement in December 2013 to exchange exposure of existing IBRD loans against MIGA exposure under its NHSFO coverage. The agreement allows IBRD and MIGA to exchange exposures in situations when one of the institutions is facing country lending or guarantee capacity limits and exchange this with the other institution by taking exposure in a country or countries where it has available capacity.44 In effect, MIGA swapped IBRD’s exposure and vice versa. At present, the agreement is limited to projects where the counterparty is the sovereign (as opposed to a sub-sovereign, which typically is guaranteed by MIGA without a sovereign guarantee). 4.47 As the NH business volume grew, MIGA reinsured a significant portion of its NH exposure, primarily with private insurers, to ease capacity constraints, manage 31 Chapter 4 Findings concentration risk, and avoid requesting additional capital from shareholders. MIGA ceded 73 percent of its NH gross exposure to a group (syndicate) of reinsurers to manage potential risk of financial loss and capital impairment. Of the $11.3 billion NH guarantees issued since 2011, MIGA retained $3.0 billion in its books. Error! Reference source not found. shows the share of NH risk that MIGA retains in its portfolio after reinsurance from 2011 to 2019; the rest is held by reinsurers. The reinsured portion reached as high as 90 percent of MIGA’s maximum aggregate liability in some projects, although MIGA had capacity and did not reinsure its exposure in six NH projects. Figure 4.6. To Manage Non-Honoring Risk to its Portfolio, Multilateral Investment Guarantee Agency Retained a Third of Overall Exposure—a Significant Proportion was Reinsured, Mostly to Private Insurers Gross amounts issued MIGA net retention rate (%) 2,000 Net exposure amount Amount issued or exposure amount 1,800 1,600 1,400 ($, millions) 1,200 15 14 1,000 37 800 19 600 71 55 21 21 400 41 200 0 2011 2012 2013 2014 2015 2016 2017 2018 2019 Fiscal year Sources: Multilateral Investment Guarantee Agency Contracts Issued database; Independent Evaluation Group Staff Calculations. Note: Net retention rate is calculated as the percentage of MIGA’s net exposure amount after reinsurance divided by the NH guarantee amounts issued by MIGA before reinsurance. It represents the share of NH risk that MIGA holds in its portfolio after reinsurance. MIGA = Multilateral Investment Guarantee Agency; NH = non-honoring of sovereign, sub- sovereign and SOE financial obligations. 4.48 The NH products involve risks for MIGA in cases of claim payout resulting from a sovereign default. So far, MIGA has not experienced any claims related to its NH cover and to date, it has remained untested. Nonetheless, the NH product involved a higher risk profile than PRI products because it can be triggered by a default by the sovereign entity in breach of its lending obligations, no matter the cause. This risk needs to be carefully considered as MIGA continues to expand its NH business and considers expanding in IDA and FCS countries. 32 5. Summary and Implications for Future Strategy: What’s Next for Multilateral Investment Guarantee Agency’s Non-Honoring Products 5.1 The introduction of the NH products was an important development in the evolution of MIGA’s business lines, a needed step to grow MIGA’s business and catch up with the rest of the credit and investment insurance market. In reviewing MIGA’s 10- year experience with its NH products, this evaluation has assessed the extent to which the availability of NH cover has enhanced MIGA’s development impact, additionality, and achieved its business objectives. How Did Multilateral Investment Guarantee Agency Enhance Its Development Effectiveness and Additionality with the New Non- Honoring Products? 5.2 MIGA enhanced its development effectiveness and additionality with the NH products. The NH products helped sovereign, sub-sovereign, and SOE borrowers with their long-term financing to supplement concessional financing from development finance institutions, domestic banks, or international capital markets. Lenders also benefited from MIGA’s NH guarantees. With the MIGA NH cover, the cost of lending was reduced, and banks could provide a longer repayment period thanks to the zero- risk weighting of MIGA-backed loans under the Basel II and Basel III capital adequacy regulations. According to the feedback provided by lenders and private insurers, they value the unique strength that MIGA derives as a member of the Bank Group, which provides an umbrella of deterrence against government actions that could disrupt projects. They could pool, package, and syndicate the MIGA-insured loans to other participating banks or sell these in the secondary market thereby managing their loan portfolio risk and earning revenues. Although the MIGA guarantee holders of record are repeat clients, the NH products opened new business opportunities for midtier international banks (as new clients) that participated in the syndicated loan insured by MIGA. 5.3 The NH products contributed to the growth of MIGA’s guarantee business, responded to the risk mitigation needs of international banks, and kept pace with the credit and investment insurance market. The volume of MIGA’s guarantee business grew by 11 percent annual average (between 2011 and 2019) compared with a 2 percent annual average growth rate before the NH introduction (between 2002 and 2010). MIGA also insured some innovative and pioneering projects. Development outcomes from the evaluated projects were positive and confirmed MIGA’s positive role and contribution 33 Chapter 5 Summary and Implications for Future Strategy in improving E&S effects at the project level. The NH products crowded-in an estimated $7.6 billion in private sector financing and insurance capacity to support priority public sector undertakings. They also helped in optimizing the development financing strategy of the public borrowers and the international lenders. 5.4 Assessed against the expected additionality and development objectives from NH products contained in the 2009 and 2013 Board approval documents and subsequent strategy documents, available evaluative evidence indicates that MIGA has not, to date, fulfilled some of those expectations and objectives or it is inconclusive about the impact of NH projects. First, few of the NH projects were in MIGA’s strategic focus areas. PRI continues to be the instrument of choice to support foreign investments in IDA and FCS countries. MIGA’s NH products supported projects in upper-middle- and high-income countries, which continue to have access to commercial and concessional financing. Second, the benefits of MIGA’s NH products were mainly through the extended loan tenors passed on to sovereign, sub-sovereign, and SOE borrowers. The evaluation could not confirm whether the NH guarantees have helped reduce the final interest rate that the public sector borrowers would pay on the commercial loans insured by MIGA. Third, little is known of whether and how the NH products influenced public sector borrowers to increase corporate transparency and disclosure, follow good corporate governance practices, and anti-money laundering, antifraud, and anticorruption practices, which were stated additionality of the NHFO-SOE product. Fourth, the actual effects of the NH-insured loans on debt servicing, debt sustainability, economic impact on beneficiaries, and poverty reduction are not known. NH projects are public sector undertakings, and the foreign currency-denominated commercial loan to the sovereign, sub-sovereign, or SOE must be understood in this broader public finance context. How Well Did Multilateral Investment Guarantee Agency Manage the Deployment of its Non-Honoring Products? 5.5 MIGA was successful in ensuring that its NH products would be attractive to international lenders and in growing the NH business. It expanded the investments eligible for MIGA coverage. It extensively used provisions in the MIGA Convention to get Board authorization to cover new types of investments and respond more quickly to business opportunities. It marketed the NH products to large and sophisticated banks who were former repeat clients that purchased PRI in the past. Through the NH products, MIGA expanded its client base to include EPC contractors and partnered with ECAs and new midtier bank clients that participated in the syndicated loan insured by MIGA’s NH product. MIGA coordinated and collaborated with the World Bank and IFC in finding opportunities for joint financing and in the appraisal, assessment, and monitoring of NH projects. Collaboration was greater between MIGA and IBRD given 34 Chapter 5 Summary and Implications for Future Strategy that the obligors are sovereigns and the projects are of a public finance nature. IBRD and MIGA entered into an Exposure Exchange Agreement in December 2013, in which MIGA swapped IBRD’s exposure and vice versa. 5.6 As MIGA expanded its NH guarantee business, it became more cautious in managing its operational risks, concentrating its NH exposure mainly in upper-middle- and high-income countries. To manage concentration risk, reduce the risk of financial loss and capital impairment from its NH exposure and avoid seeking for capital increase from its shareholders as the NH portfolio grew, MIGA relied on the reinsurance market. It reinsured 73 percent of its NH exposure, mainly with private insurers, compared with 49 percent ceded to reinsurers for PRI. Initial guidance in 2012 indicated that MIGA could insure projects with a minimum sovereign credit rating; however, the insured projects had sovereign credit ratings above that minimum threshold. Nevertheless, MIGA was able to insure several projects in IDA countries and an FCS country in 2012 and 2013. As MIGA gained experience with the NH product, it changed its minimum country or sovereign risk credit rating was set in 2015 and mainly linked to upper- middle- and high-income countries. The concentration of NH projects in these countries, which continue to have access to commercial and concessional sources of financing, reflects the extent of MIGA’s risk appetite. Implications for Multilateral Investment Guarantee Agency’s 2021–23 Strategy 5.7 With its current risk rating ceiling and considering MIGA’s legitimate concern about minimizing the risk of financial loss and capital impairment, experience to date indicates that MIGA’s NH products need revisiting to advance its strategic focus on facilitating investments in IDA countries and FCS. Even if MIGA eases its minimum sovereign credit rating threshold, the IDA countries and FCS in the new range are mostly heavily indebted poor countries that will find it difficult to increase public indebtedness. Finding bankable projects under these conditions will continue to pose a challenge. Additionally, the strict limits on commercial borrowing by low-income countries under the proposed Joint International Monetary Fund-World Bank Debt Sustainability Framework hinders future commercial borrowings by IDA and FCS countries. 5.8 MIGA has two types of guarantee products in its tool kit to fulfill its development mandate and meet its strategic objectives: the PRI and NH products. In considering what is next for the NH products, given experience and the focus of MIGA’s 2021–23 strategy and business outlook on IDA countries and FCS, the findings from this evaluation invite MIGA’s management to address the following issues: 35 Chapter 5 Summary and Implications for Future Strategy • How might NH products be designed to better contribute to MIGA’s corporate commitment to IDA and FCS countries? • To what extent could MIGA’s current minimum credit rating eligibility criteria be modified to enhance the strategic usefulness of the NH products? • If MIGA concludes that the NH products cannot be made to work in IDA countries and FCS, how should this operational stance be best reflected in MIGA’s business strategy, rules, and targets for future deployment of the instrument? • How can the evaluation methodology for assessing the development impact of NH projects fit their public sector nature and capture the public finance, debt servicing and debt sustainability implications and the economic impact of NH- insured commercial loans on the final beneficiaries and their poverty reduction impact, especially for projects in IDA and FCS? How can the ex post development effectiveness evaluation methodology capture the debt sustainability implications of projects that have a series of NH guarantee contracts with the same obligor, project enterprise or host country? Addressing these questions provides MIGA’s management a draft blueprint to review the role of its NH products. The evaluative evidence and discussion presented in this report will help define these roles, and the expectations attached to them—regarding their contribution to MIGA’s business activity, risk exposure, strategic focus, and development mandate. 36 Notes Chapter 1 1One of the proposals in the report by the G-20 Eminent Persons Group on Global Financial Governance (2018), “Making the Global Financial System Work For All,” referred to “develop systemwide political risk insurance and expand use of private reinsurance ma rkets” and, specifically, “the MDBs should, as a system, leverage on MIGA [the Multilateral Investment Guarantee Agency] as a global risk insurer in development finance” (proposal 4b). Chapter 2 1 “Convention on the Establishment of a Multilateral Investment Guarantee Agency.” 1988. 2Breach of Contract coverage was launched formally by the Multilateral Investment Guarantee Agency (MIGA) in 1999 as a separate cover in response to rising demand for private sector project finance of public-private partnerships in developing countries. Before 1999, MIGA- insured breach by the host government of its contractual obligation but under its Expropriation cover. 3 The 2009 Operational Regulation amendments include (i) offering of the Non-Honoring of Sovereign Foreign Obligation as a new area of coverage; (ii) enhancing the breach of contract coverage to state-owned enterprises (SOEs) and to all three types of denial of justice risks; (iii) enhancing the war and civil disturbance coverage to include temporary business interruption and violence against the investor’s government or nationality; ( iv) revising the definition of project implementation and expanding the coverage for the acquisition of existing assets to also include intangible assets or benefits; and v) addressing eligibility issues specifically expanding the definition of loan to include debt capital market securities without requirement of a related MIGA-insured equity investment, shortening minimum tenor of MIGA-insured loans from three years to one year, expanding MIGA coverage to jurisdictions where the project conducts essential operations, and eliminating minimum and maximum premium rates. 4 The investment insured by MIGA may be in the form of debt (loan), equity, quasi-equity, loan guarantee, bonds, and hedging and derivative instruments. 5With MIGA’s Operational Policies effective January 6, 2015, MIGA can insure swaps and other hedging instruments that are not linked to MIGA-covered investments. MIGA Operational Policies. 2015. https://www.miga.org/sites/default/files/archive/Documents/Operational- Policies.pdf 6 Environmental and Social Review Summary – Project Description, Panama Metro, https://www.miga.org/project/panama-metro; and Project Brief, Panama Metro Line One. https://www.miga.org/project/panama-metro-line-one. 37 Notes 7Thus, MIGA’s breach of contract covers as a political risk (i) the inability of the MIGA guarantee holder to have recourse to an arbitral or judicial forum for the claim or the decision of the arbitral, or (ii) judicial forum could not be enforced within a reasonable time or the foreign investor to enforce the arbitral award, or (iii) the arbitral decision could not be enforced. https://www.miga.org/product/breach-contract. 8 Meaning, the government or subnational government will not contest the merits of any dispute arising from the underlying transaction or investment insured by MIGA through its NHSFO insurance. MIGABrief—Non-Honoring of Financial Obligations. July 2013. https://www.miga.org/sites/default/files/2018-06/NHFObrief%20%281%29.pdf 9MIGA can and has supported (SOEs), either as the project enterprises or as guarantee holders through its political risk insurance coverages under different eligibility criteria. MIGA has provided coverage for currency inconvertibility and transfer restriction, expropriation, or war and civil disturbance risks to cross-border investments by (SOEs), government entities, and not- for-profit organizations (for example, including development finance institutions) in their capacity as guarantee holders. To be eligible for coverage, MIGA requires that these entities operate on a commercial basis or that the investment be made on a commercial basis. And if these SOEs or government entities, development finance institutions, or not-for-profit organizations have both commercial and noncommercial operation, MIGA is authorized to provide coverage to these entities only if the investment that MIGA will insure is proven at the outset to be of a commercial nature (that is, based on profit and income). Some past examples of public entities, SOEs, and development finance institutions that were insured by MIGA include Eskom (South Africa), FMO (The Netherlands), Industrial Development Corporation (South Africa), KfW (Germany), DEG (Germany), L’Office National des Telecommunications (Tunisia), Statkraft SF (Norway), and State Bank of India. MIGA had also reinsured political risk exposures of government or public export credit agencies such as Export Development Canada, Overseas Private Investment Corporation (United States), and Slovene Export and Development Corporation (Slovenia). The 2009 and 2013 changes in covering payment obligations of SOEs under Breach of Contract (2009) and Non-Honoring of Financial Obligations by a State-Owned Enterprise (NHFO-SOE) (2013) coverage involved payment obligations of developing member countries’ SOEs in their capacity as project enterprises. 10MIGA’s Convention also authorized it to carry out complementary activities, including technical assistance activities, dispute resolution, and knowledge services to support productive foreign direct investment. But MIGA’s core business remains the provision of political risk insurance. MIGA Convention. https://www.miga.org/sites/default/files/archive/Documents/MIGA%20Convention%20(April%2 02018).pdf 11MIGABrief—Non-Honoring of Financial Obligations. July 2013. https://www.miga.org/sites/default/files/2018-06/NHFObrief%20%281%29.pdf; and Standard Contracts of Guarantee – Non-Honoring of a Sovereign Financial Obligation (Government Borrower) in https://www.miga.org/terms-conditions and 38 Notes https://www.miga.org/sites/default/files/archive/Documents/Non- Honoring%20of%20a%20Sovereign%20Financial%20Obligation%20(Government%20Borrower).p df. 12MIGABrief—Non-Honoring of Financial Obligations. July 2013. https://www.miga.org/sites/default/files/2018-06/NHFObrief%20%281%29.pdf. Chapter 3 1 The Concept Note for this evaluation had four evaluation questions: i. To what extent have MIGA’s non-honoring insurance products been additional and effective in helping MIGA enhance its development mandate, particularly in its strategic focus areas and in underserviced markets? ii. To what extent have MIGA’s non-honoring insurance products expanded MIGA’s business? To what extent have the NHFO products helped international lenders risk mitigation needs and mobilize additional capital with better terms for priority public sector projects? iii. How well has MIGA managed the risks associated with its non-honoring insurance coverage and under what conditions do these products work best? iv. Based on the findings in (i) to (iii), what are the implications for MIGA’s 2021–23 Strategy? 2 The eight NHFO projects that have completed evaluations (MIGA self-evaluations and IEG validations) are (i) Istanbul Electricity Tram and Tunnel General Directorate (Otogar-Bağcilar- Ikitelli-Olimpic Village Metro); (ii) Hungary Eximbank I; (iii) Compagnie Tunisienne de Navigation SA; (iv) Metropolitan Municipality of Istanbul (Kadikoy-Kartal-Kaynarca Metro Project); (v) Senegal US$/EUR Cross-Currency Swap; (vi) Panama Metro/Metro de Panama; (vii) Turkey’s Eximbank I; and (viii) Financiera de Desarrollo Territorial SA The project Compagnie Tunisienne de Navigation SA was evaluated by IEG in the format of a Project Performance Assessment Report. Chapter 4 1MIGA first submitted the proposal to change its Operational Regulations on December 23, 2008. In 2009, MIGA submitted a revised document to the Board superseding the 2008 document. Sources: MIGA/R2008-0072. Proposed Changes to MIGA’s Policies and Operational Regulations. December 23, 2008; and MIGA/R2009-014 Proposed Changes to MIGA’s Policies and Operational Regulations. March 25, 2009. 39 Notes 2Project Brief—Special Economic Zone at Duqm. https://www.miga.org/project/special- economic-zone-duqm. 3 Mobilization amounts provided by MIGA to IEG as of June 30, 2019. 4The African Trade Insurance Agency offers guarantees and insurance to 14 Africa-member countries, while Islamic Corporation for the Insurance of Investment and Export Credit does so to 41 Islamic-member countries. Sources: African Trade Insurance Agency, http://www.ati- aca.org/investor-information/current-members/; and The Islamic Corporation for the Insurance of Investment and Export Credit, https://iciec.isdb.org/member#. 5Project Brief – Land and Agricultural Development Bank of South Africa in MIGA.org (MIGA’s external website), https://www.miga.org/project/land-and-agricultural-development-bank-south- africa 6World Bank Group. “Bangladesh Systematic Country Diagnostic.” October 25, 2015 and “Bangladesh—Country Partnership Framework, FY16–20.” March 8, 2016. 7Annex I in Part II of General Conditions of MIGA guarantee contracts defines and lists the coercive, corrupt, collusive, fraudulent, and obstructive practices in the context of MIGA’s operations. Source: Guide—Sample Contracts of Guarantee (June 16, 2018), Non-Honoring of a Sovereign Financial Obligation (Government Borrower) in MIGA.org (MIGA’s external website). https://www.miga.org/sites/default/files/2019-05/MIGA%20NHSFO%20- %20Govt%20Borrower%20Template%20-%20%5B2016%20FORMS%20- %20OCTOBER%202016%5D.pdf 8 For example, in an evaluated NHSFO-insured project, the project company did not have audited financial data for the full 2.5 years of operation, and did not share with the MIGA Project Evaluation Report team the updated financial model for the project. The MIGA guarantee holder is the lender but the obligor is the government acting through the Ministry of Economy and Finance, while the project operator is a state-owned company. 9The effect of the NH commercial loan on the government’s finances, debt servicing, and debt sustainability, and the economic impact on the beneficiaries and poverty reduction (especially in IDA and fragile and conflict-affected situation [FCS] countries) have not been considered in the analysis of economic contribution and sustainability results among the evaluated NH projects. Lessons from both self-evaluation and IEG validation have already identified this aspect as a shortcoming in the assessment of NH projects’ economic contributions and development results. 10Sources: São Paulo State Sustainable Transport Project, World Bank external website— https://projects.worldbank.org/en/projects-operations/project-detail/P127723?lang=en; and Project Brief—São Paulo Sustainable Transport, https://www.miga.org/project/sao-paulo-sustainable- transport. 40 Notes 11In addition, MIGA’s premium and fees are rolled into the NHFO-insured loan paid by the borrower, so the final loan pricing depends to some extent on the all-in price of MIGA’s NHFO guarantees. 12 Since the 2009 Board authorization of the NHSFO until 2017, MIGA’s strateg ies focused on four priority areas: (i) investments in IDA countries; (ii) investments in fragile and conflict-affected situations (FCS); (iii) investment in complex projects, mostly in infrastructure and the extractive industries; and (iv) South-South investments. In its 2018–20 strategy, MIGA identified three priority areas, for example, reaffirming its support for cross-border investments in IDA and FCS countries, and in climate change and adaptation. 13MIGA defines complex projects as those involving (i) project finance or structured finance or other multisourced deals; (ii) deals with challenging environmental considerations and social issues classified as ‘Category A’; (iii) infrastructure deals, including those in the transport, power, water and sanitation, telecom sectors; and (iv) extractive industries deals, including those in the oil, gas, mining sectors. 14IEG’s review of the NH issuance portfolio indicates that 74 percent of the $11.3 billion NH contract amounts issued during the period 2011 to 2019 were in middle income countries. Sixty- three percent of this amount went to support projects in five upper middle-income countries (for example, Azerbaijan, Brazil, Colombia, South Africa, and Turkey). 15 Except for two NHFO-insured loans, MIGA covered 95 percent of the guaranteed loan amount; the remaining 5 percent is considered the uninsured portion that the guarantee holder is responsible for. 16 BPL Global. “Market Insight 2019: Credit and Political Risk Insurance,” February 2, 2019 estimates the combined capacity of private insurers at $900 billion. MIGA’s NH products are similar to what is called contract frustration in the credit and political risk insurance market. Contract frustration risk is applicable where the counterparty risk insured is a government entity, or a majority state-owned commercial operation, and the underlying transaction is either trade- related or, if not, the obligation insured is secured by assets. However, the maximum amounts are only for countries with the very best risks (A rated countries, for example). 17There continues to be strong preference for concessional (cheaper) financing from multilateral development banks and other development finance institutions rather than borrowing from international commercial lenders. To a large extent, concessional financing also competes with MIGA’s NH product. Compounding this challenge is the (national and local) host governments’ lack of knowledge about the benefits and use of credit insurance to mobilize private sector financing for their development needs. 18By contrast, the breach of contract political risk insurance product requires that the insured party prove that it did not violate any conditionalities in the underlying contract and that the 41 Notes loan default was a result of government action or inaction. Additionally, an arbitral process and judgment are required. 19 Eligibility of a SOE for coverage under NHFO-SOE must meet three requirements: (i) the SOE must be subject to governmental supervision and control. MIGA views an entity as “controlled by the government” if: (i) the government owns an equity interest exceeding 50 percent in the SOE, or (ii) it is able to direct the affairs and/or control the composition of the board of directors or equivalent body of that entity; (ii) it must provide a public service or function, and (iii) it must be creditworthy as a stand-alone entity. Factors to look at include the SOE’s revenue base, funding sources, history of honoring financial commitments, applicable regulatory regime, level of direct and indirect government support, and others relating to the SOE’s financial viability. 20 Chapter IV: Financial Provisions, Article 25: Financial Management, MIGA Convention https://www.miga.org/sites/default/files/archive/Documents/MIGA%20Convention%20(April%2 02018).pdf. Refer also to “Commentary on the Convention Establishing the Multilateral Investment Guarantee Agency.” October 11, 1985. Paragraphs 53–55, pages 25 and 26, https://www.miga.org/sites/default/files/archive/Documents/commentary_convention_november _2010.pdf 21 Preamble, MIGA Convention https://www.miga.org/sites/default/files/archive/Documents/MIGA%20Convention%20(April%2 02018).pdf. 22 Managament’s Discussion & Analysis and Financial Statements. June 30, 2019 in MIGA external website: https://www.miga.org/sites/default/files/2019- 08/MIGA%20Financial%20Statements%20and%20Management%20Discussion%20%26%20Analy sis%20-%20June%2030%202019.pdf. 23Projects that had a rating of below the current minimum credit rating threshold were underwritten in the earlier years when MIGA was still drafting and testing its NHFO guidelines. However, MIGA required additional risk mitigants, such as opening of a debt service coverage escrow account overseas, upfront payment of MIGA premium that is then rolled into the insured loan, and a sovereign guarantee from the Ministry of Finance. MIGA Strategy and Business Outlook FY21–23. https://www.miga.org/sites/default/files/2020- 09/MIGA%20FY21-23%20Strategy%20%26%20Business%20Outlook%20%28Website%29.pdf 25 Before the MIGA Board-approved changes to its Convention and Operational Regulations in 2009, MIGA was not authorized to provide coverage to foreign investments in the form of stand- alone debt. Article 12(b) allows the Board, by special majority, to extend eligibility to other types of medium- to long-term investments “except that loans other than [shareholder loans] may be eligible only if they are related to a specific [equity] investment covered or to be covered by the Agency.” Thus, MIGA could insure nonshareholder loans only if there is a related equity investment(s) being covered by MIGA. The 2009 amendments also broadened the scope of MIGA’s insurance coverage for existing assets. 42 Notes 26For details, see MIGA Operational Policies (Part I: Guarantee Operations, Chapter 1: Eligibility Requirements), January 6, 2015. https://www.miga.org/sites/default/files/archive/Documents/Operational-Policies.pdf. 27Chapter III: Operations, Article 11: Covered Risks of the MIGA Convention (amended) only refers to Currency Transfer, Expropriation and Similar Measures, Breach of Contract, and War and Civil Disturbance. MIGA Convention. https://www.miga.org/sites/default/files/archive/Documents/MIGA%20Convention%20(April%2 02018).pdf 28MIGA Operational Policies, Section IV: Eligible Risks, Section 1.24(b)(ii) and (iii). January 6, 2015. https://www.miga.org/sites/default/files/archive/Documents/Operational-Policies.pdf. 29 Project Brief–Senegal US$/Euro Cross-Currency Swap, https://www.miga.org/project/senegal- cross-currency-swap; and Project Brief—Sirajganj II Power Plant Bangladesh, https://www.miga.org/project/sirajganj-225-mw-combined-cycle-power-plant-2nd-unit-dual-fuel. 30 NHFO-insured loans financed engineering, procurement, and construction contractors such as Construtora Norberto Odebrecht SA (Brazil) and its subsidiaries in the Cambambe Hydropower II project (Angola), the Panama Metro; a consortium of Chinese construction contractors in the Sirajganj II Power Project (Bangladesh); and the Supply Contractor, Daewoo Shipbuilding & Marine Engineering Co. Ltd., for the Tunisia ferry project are examples. 31 Environmental and Social Review Summary —Cambambe Hydroelectric Project-Phase II, https://www.miga.org/project/cambambe-hydroelectric-project-phase-ii-0; and Project Brief— Cambambe Hydroelectric Project-Phase II, https://www.miga.org/project/cambambe- hydroelectric-project-phase-ii. 32 Project Brief—Panama Metro Line One, https://www.miga.org/project/panama-metro-line-one 33 Project Brief—Passenger-Car Ferry TANIT (Tunisia), https://www.miga.org/project/passenger- car-ferry-tanit. 34Environmental and Social Review Summary —Istanbul Üsküdar-Ümraniye-Çekmeköy Metro Line, https://www.miga.org/project/uskudar-umraniye-cekmekoy-metro-line. 35Environmental and Social Review Summary —Ghorasal 3rd Unit Repowering Project, https://www.miga.org/project/ghorasal-3rd-unit-repowering-project?esrsid=130&pid=1369. 36 MIGA Convention. https://www.miga.org/sites/default/files/archive/Documents/MIGA%20Convention%20(April%2 02018).pdf. 37IEG classified joint World Bank Group projects into three types: cofinanced, parallel and sequential projects. IEG Learning Product, “World Bank Group Joint Projects: A Review of Two Decades of Experience—Lessons and Implications from Evaluation” 2017. 43 Notes 38IFC Summary of Investment Information—Izmir Tramway. https://disclosures.ifc.org/#/projectDetail/SII/34306. 39 IFC Summary of Proposed Investment—Istanbul Kadikoy-Kartal Metro. https://disclosures.ifc.org/#/projectDetail/SPI/27309. 40 IFC Summary of Proposed Investment—Izmir Railcars. https://disclosures.ifc.org/#/projectDetail/SII/35012; MIGA.org (MIGA external website). Project Brief, Izmir Metro Project, https://www.miga.org/project/izmir-metro-project. 41 The MIGA NHSFO contract of guarantee was canceled in September 2019. 42One of the strategic objectives of the FY12–15 Country Partnership Strategy for Turkey envisaged World Bank Group support for Turkey’s urbanization challenge through the Sustainable Cities Program, which became the foundation of the Turkey Sustainable Cities Joint Implementation Plan (JIP). The JIP’s objective was to devise a road map for an integrated Bank Group approach in providing comprehensive public-private solutions to the pressures brought about by Turkey’s rapid urbanization. The improved sustainability of Turkish cities was again emphasized as a Bank Group objective (Objective 8: Improved Sustainability and Resilience of Cities) in the Bank Group’s FY18–21 Country Partnership Framework for Turkey. Sources: World Bank Group FY12–15 Country Partnership Strategy for Turkey; World Bank Group FY18 –21 Country Partnership Framework for Turkey. 43 IDA/IFC/MIGA Country Partnership Framework for Bangladesh for the Period FY16 –20. March 9, 2016. 44Under the agreement, IBRD provided a guarantee on principal and interest pertaining to MIGA’s guarantee exposure under its Non-Honoring of Sovereign’s Financial Obligation in exchange for MIGA’s guarantee on IBRD's loan principal and interest exposure. IBRD would reimburse MIGA for payment that MIGA incurred on a particular NHSFO exposure, and MIGA would reimburse IBRD for nonpayment under one of IBRD’s loan exposures. The agreement applies only to projects in which the counterpart is the sovereign (as opposed to a sub-sovereign, which typically is guaranteed by MIGA without a sovereign guarantee). 44 Bibliography Asian Development Bank. 2017. Boosting ADB’s Mobilization Capacity: The Role of Credit Enhancement Products. Manila: Asian Development Bank. BPL Global. 2013. “Perspectives on Medium and Long-Term (MLT) Export Credit Insurance, Choice and the ‘Mixed Market’“ London: BPL Global. BPL Global. 2018. “Market Insight 2018: Credit and Political Risk Insurance.” London: BPL Global. BPL Global. 2019. “Market Insight 2019: Credit and Political Risk Insurance.” London: BPL Global. Braun, Alexander, and Marius Fischer. 2018 “Determinants of the Demand for Political Risk Insurance: Evidence from an International Survey.” The Geneva Papers on Risk and Insurance: Issues and Practice 43(3): 397–419. Economist Intelligence Unit. 2007. “World Investment Prospects to 2011.” New York: Economist G-20 Eminent Persons Group on Global Finance. 2018. “Introductory Guide to Infrastructure Guarantee Products from Multilateral Development Banks.” Technical Note no IDB-TN- 01611. https://publications.iadb.org/en/introductory-guide-infrastructure-guarantee- products-multilateral-development-banks G-20 Eminent Persons Group on Global Finance. 2018. “Making the Global Financial System Work for All.” Gallagher, Arthur J. 2016. Report and Market Update: Credit and Political Risk Insurance. New York: Arthur J. Gallagher & Co. Gallagher, Arthur J. 2018. Credit and Political Risk Insurance Report and Market Update. New York: Arthur J. Gallagher & Co. Global Impact Investing Network. 2013. Catalytic First Loss Capital. New York: Global Impact Investing Network. Gordon, Kathryn. 2008. “Investment Guarantees and Political Risk Insurance: Institutions, Incentives and Development.” In OECD Investment Policy Perspectives 2008, edited by OECD Investment Division of the Directorate for Financial and Enterprise Affairs, 91– 122. Paris: OECD Publications. International Finance Corporation. 2019. “Note on Implementation of IFC Investments in Sub- National Entities and Legal Note on the Power of IFC to Finance Sub-Sovereign Entities.” International Finance Corporation note, January 9. International Finance Corporation, Washington, DC. IMF (International Monetary Fund). 2019. Sub-Saharan Africa Regional Economic Outlook: Recovery Amid Elevated Uncertainty. Washington, DC: International Monetary Fund. https://www.imf.org/en/Publications/REO/SSA/Issues/2019/04/01/sreo0419 45 Bibliography Mandel, Benjamin H., Daniel Morgan, and Chenyang Wei. 2012. “The Role of Bank Credit Enhancements in Securitization.” Federal Reserve Bank of New York Economic Policy Review 18, (2): 35–46. Moody’s Investor Service. 2019. Sovereign Default and Recovery Rat es, 1983–2018. Data Report. April 8. Mudde, Paul H. J. 2016. “Financing Gaps, Mobilisation and the Importance of Enhanced Cooperation between Development Financiers and Berne Union Members.” Berne Union Annual Report 2016. London: TXF News. Mudde, Paul H. J. 2018. “The Complementary Role of Official Development Finance: Some Observations and Recommendations.” In Berne Union Annual Report 2018. London: TXF News. MIGA (Multilateral Investment Guarantee Agency). 2014. MIGA Strategic Directions FY15–17 . Washington, DC: MIGA. Pereira dos Santos, Pablo. 2018. “Introductory Guide to Infrastructure Guarantee Products from Multilateral Development Banks.” Technical Note IDB-TN-01611, Inter-American Development Bank, Washington, DC. https://publications.iadb.org/en/introductory- guide-infrastructure-guarantee-products-multilateral-development-banks. Rega-Jones, Natasha. 2019. “Non-Payment Insurance Grows as Banks Shun Stuttering CDS Market.” Risk.net, January 24. https://www.risk.net/derivatives/6322621/non-payment- insurance-grows-as-banks-shun-stuttering-cds-market Schlegl, Matthias, Christoph Trebesch, and Mark L. J. Wright. May 2019. “The Seniority Structure of Sovereign Debt.” Working Paper 25793, National Bureau of Economic Research, Cambridge, MA. Shihata, Ibrahim F. I. 1988. MIGA and Foreign Investment: Origins, Operations, Policies, and Basic Documents of the Multilateral Investment Guarantee Agency. Boston, MA: Dordrecht and Martinus Nijhoff. World Bank. 2008. “Independent Evaluation Group—MIGA 2008 Annual Report. Approach Paper (English)”. Independent Evaluation Group (IEG) Approach Paper. Washington, DC: World Bank Group. World Bank. 2009. The World Bank Group Guarantee Instruments 1990–2007: An Independent Evaluation. Independent Evaluation Group. Washington, DC: World Bank. World Bank. 2012. Proposed Changes to MIGA's Policies and Operational Regulations. Washington DC: World Bank. https://imagebank2.worldbank.org/search/17007371 World Bank. 2016. World Bank Group Guarantee Products Guidance Note. Washington, DC: World Bank. 46 Appendix D Profile of MIGA Non-Honoring Projects World Bank. 2018. “A Joint World Bank-IMF Multipronged Approach for Addressing Emerging Debt Vulnerabilities” Prepared jointly by the World Bank and the International Monetary Fund. World Bank and International Monetary Fund, Washington, DC. 47 Appendix A. Non-Honoring of Sovereign Financial Obligations and Non-Honoring of Financial Obligations by a State-Owned Enterprise Projects, Fiscal Years 2011–19 FY Gross Amount Issued Issued Project Enterprise Name Country ($, millions) 2011 Istanbul Electricity Tram and Tunnel General Directorate Turkey 20 (Otogar-Bağcılar-Ikitelli-Olimpic Village Metro) 2011 Istanbul Municipality (Kadikoy-Kartal-Kaynarca Metro Line) Turkey 409 2012 Panama Metro Panama 320 2012 Compagnie Tunisienne de Navigation SA Tunisia 218 2012 Senegal US$/Euro Cross-Currency Swap Arrangement Senegal 99 2012 Takoradi III Power Plant Ghana 88 2013 Ashuganj Power Station Company Limited Bangladesh 251 2013 Cambambe Hydroelectric Power Dam – Phase II Angola 512 2013 Izmir Buyuksehir Belediyesi (Izmir Marine Transport)/ Izdeniz Turkey 66 A.S. (Izmir Ferry) 2014 The Municipality of Izmir (Izmir Railcars) Turkey 91 2014 Magyar Export-Import Bank Zrt. (Hungarian Export-Import Bank Hungary 575 Private Limited Company) 2014 Metro de Panama SA Panama 320 2014 BT20—Cuu Long Joint Stock Company Vietnam 500 2015 Financiera del Desarrollo S.A. (Findeter S.A.) Colombia 95 2015 Izmir Buyuksehir Belediyesi (Izmir Tramway-Light Rail; Izmir Turkey 33 Metro) 2015 The State of São Paulo (São Paulo Sustainable Transport Brazil 361 Project) 2015 Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank I) Turkey 333 2015 Istanbul Ulasim A.S. (Üsküdar-Ümraniye-Çekmiköy Metro) Turkey 193 2016 Cambambe Hydroelectric Power – Phase II Angola 15 2016 Eskom Holdings SOC Ltd South Africa 783 2016 VNeco Hoi Xuan Investment and Electricity Construction (Hoi Vietnam 240 Xuan Hydropower) 2016 Senegal Eurobond Cross-Currency Swap Senegal 100 2016 North-West Power Generation Company Limited (Sirajganj II) Bangladesh 69 2016 Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank I) Turkey 616 2017 Development Bank of South Africa South Africa 229 2017 Bangladesh Power Development Board (Ghorasal 3 Bangladesh 97 Repowering) 2017 Land and Agricultural Development Bank of South Africa South Africa 448 49 Appendix A NHFSO and NHFO–SOE Projects FY Gross Amount Issued Issued Project Enterprise Name Country ($, millions) 2017 Magyar Export-Import Bank Zrt. (Hungarian Export-Import Bank Hungary 312 Private Limited Company) 2017 North-West Power Generation Company Limited (Sirajganj 3) Bangladesh 70 2017 Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank II) Turkey 382 2018 Southern Gas Corridor Closed Joint Stock Company Azerbaijan 1,113 2019 Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank III) Turkey 606 2019 Cambambe Hydroelectric Power – Phase II (Swap)a Angola 1.6 2019 The Duqm Special Economic Zone Authority Oman 1,701 Source: www.miga.org/projects/. Note: FY = fiscal year; NH = Non-Honoring insurance products; NHSFO = Non-Honoring of Sovereign and Sub-Sovereign Financial Obligations; NHFO-SOE = Non-Honoring of Financial Obligations by a State-Owned Enterprise. a. This represents a NH contract issued for an interest swap transaction linked to the NH-insured syndicated loans in 2013 and 2016. 50 Appendix B. Glossary of Terms Used Arbitral award default: A process to settle a legal dispute in which a third party (an arbitrator) acts much like a judge, but in an out-of-court and less formal setting. Default occurs when a party does not comply with the arbitration ruling. Breach of contract: (i) Losses resulting from repudiation or arbitrary withdrawal of a party from its duties and responsibilities imposed by a contract. (ii) MIGA-specific definition: Coverage from payment default by the government or in certain circumstances, state-owned enterprises, which requires (i) that the guarantee holder request for arbitration or other dispute resolution mechanism set out in the contract of guarantee; (ii) the arbitration rules in favor of the guarantee holder; and (iii) either the government fails to comply with the terms of the arbitral award (nonpayment of an award) or the government interfered with the dispute resolution mechanism (denial of recourse). Comprehensive cover: Insurance of both commercial and credit and political risks. Credit enhancement: Improvement in the risk profile of a loan or credit, and reduction in a creditor’s risk of financial loss due to default of the obligor, through risk transfer or sharing with a third party. Credit enhancement products: Financial products aimed at (i) improving the credit rating or creditworthiness of an obligor or borrower to obtain more favorable financing terms, or (ii) eliminating or mitigating risks that pose as barriers to enable funding to occur. Examples of credit enhancement products include partial credit guarantees, partial risk guarantees, political risk insurance. Guarantee: The agreement by a guarantor to assume the responsibility for the performance of an action or obligation of another person or entity and to compensate the beneficiary in the event of nonperformance. Host country: The country where the MIGA project is located. Indemnity: Compensation for a loss. Insurance: A practice or arrangement by which a company provides a guarantee of compensation for a specified loss in return for payment of a premium. Maximum aggregate liability: The maximum aggregate amount of compensation payable by MIGA under the contract of guarantee over the guarantee period. Medium, Long Term: MIGA-specific definition: maturity or tenor of greater than one year. 51 Appendix B Glossary of Terms Non-Honoring of Sovereign and Sub-Sovereign Financial Obligation (NHSFO) guarantees: (i) Coverage or protection from payment defaults of a financial obligation or a guarantee by a sovereign or quasi-sovereign entity such as state-owned enterprises, a municipality, or a state, to a financier or lender or an investor. (ii) A financing operation without direct sovereign indemnity or any other form of direct recourse to a sovereign government as obligor. (iii) Failure of sovereign or sub-sovereign entities and state- owned enterprises to pay or satisfy direct debt obligations or guarantees. (iv) MIGA- specific definition: protection against losses resulting from a failure of a sovereign, sub- sovereign, or subnational (that is, city, municipality or region) to make a payment when due under an unconditional financial payment obligation or guarantee related to a MIGA-covered investment. Unlike breach of contract coverage of nonpayment obligation, non-honoring of sovereign financial obligation coverage does not require the investor to obtain an arbitral award to claim payment. NH coverage applies to situations when a financial payment obligation is unconditional and not subject to defenses (that is, not associated with any politically triggered event—such as expropriation—against the insured investment or project). Obligor: A person or legal entity that has contractually committed to perform an obligation (for example, the payment of principal or interest, or both, due under a loan). Political risk: The risk of loss when investing in a given country caused by changes in a country's political structure, policies, or stability, such as tax laws, tariffs, expropriation of assets, restriction in repatriation of profits, or political violence. Reinsurance: The process of a credit or political risk insurance provider issuing an insurance cover or guarantee with other insurance providers, which could be public (bilateral), private, or multilateral insurers, to reduce exposure by spreading the risk among the institutions involved. State-owned enterprise: (i) A legal entity created by a government with the purpose of participating in commercial activities on the government's behalf. It can be either wholly or partially owned by the government. (ii) MIGA-specific definition: The entity is performing a public service or fulfilling a governmental function in the host country and is under the control of the host government. Sub-sovereign: Political or territorial subdivisions of a state, including dependent territories. Syndicated loan: Financing offered by a group of lenders—referred to as a syndicate— who work together to provide funds for a single borrower. The borrower can be a corporation, a large project, or a sovereignty, such as a government. 52 Appendix B Glossary of Terms Unconditional financial obligation: MIGA-specific definition: a financial obligation or guarantee is considered unconditional when there are no defenses (meaning the host government will not contest the merits of any dispute arising from the underlying transaction or investment insured by MIGA) other than the nonpayment of the underlying debt because of inability or unwillingness to pay. 53 Appendix C. Methodology Theory of Change Underpinning this evaluation of Multilateral Investment Guarantee Agency’s non- honoring coverage is a theory of change based on MIGA’s rationale articulated in the 2009 and 2013 Board documents seeking authorization for the Non-Honoring of Sovereign Financial Obligations (NHSFO) and Non-Honoring of Financial Obligations by a State-Owned Enterprise (NHFO-SOE) coverages. The theory of change for MIGA’s NHSFO and NHFO-SOE coverages is expected to work in the following ways. MIGA. By insuring commercial lenders against nonpayment by government, sub- sovereign, and state-owned enterprises (SOEs) without requiring a final arbitral award judgment or a court decision to pay a claim, MIGA takes on more risks because it must pay the guarantee holder the insured amount if the government entity does not meet its payment obligation regardless of the cause. Without going through the arbitral process, claim payment occurs more quickly, which makes the value proposition of non- honoring insurance (NH) coverages attractive to commercial lenders. As more commercial lenders purchase its non-honoring products, MIGA’s business is expected to grow and, if the higher risks associated with these products are managed well, its sustainability strengthened. From a development impact perspective, the NH products would attract international lenders to finance public sector projects in underserviced areas. MIGA’s additionality is expected to be particularly manifest in high-risk environments and in poorer countries, partly because of its development mandate and its ability to take higher risks, partly because of its multilateral status and as a member of the World Bank Group. Moreover, through reinsurance of its exposure, MIGA can mobilize funds from private insurers. Host country. MIGA’s non-honoring coverage, which does not require an arbitral award judgment, incentivizes commercial lenders to enter into contracts with government entities that are less known or do not have adequate access to medium- and long-term (MLT) financing from international financial markets, thereby increasing business opportunities for sovereign, sub-sovereign, and SOE borrowers. Additionally, MIGA’s NH products can mobilize private resources for priority public sector development projects in underserviced areas. MIGA guarantee holders (insured party). By avoiding a prolonged arbitration process, and with the added benefit of being compliant with Basel II and Basel III rules on capital adequacy by purchasing MIGA’s non-honoring coverage, and the status of MIGA as a member of the Bank Group, commercial lenders can support risky projects in more 54 Appendix C Methodology difficult markets and offer better financing terms (for example, longer maturities or lower interest payment) to sovereign, sub-sovereign, and SOE borrowers. The evaluation assessed MIGA’s additionality, development effectiveness, and experience in deploying the NH products against expectations presented to the Board in 2009 and 2013, as illustrated in figure C.1. Figure C.1. Theory of Change for the Evaluation of Multilateral Investment Guarantee Agency’s NHSFO and NHFO-SOE Coverages DIRECT INTERMEDIATE LONG TERM INPUTS OUTPUTS OUTCOMES OUTCOMES OUTCOMES IMPACT Non-Honoring MIGA expands reach to new HOST COUNTRY/ MIGA delivers its Increase in private Sovereign and types of clients, investors, GOVERNMENT mandate more sector lending to Support Sub-Sovereign sponsors effectively and HG/SS/SOE priority Inclusive and Financial New entities receive insurance maintain future viability projects Sustainable Obligation Growth (NHSFO) Insurance issued and New types of projects financed products delivered Host Countries Increase in productive Non-Honoring receive financing with foreign investment in Sovereign - At lower cost better terms for host country*** INVESTORS/LENDERS Financial - With improved financial priority, development- Obligation terms New types of investors oriented, public-sector State-Owned projects Enterprises - Without arbitral award (NHFO-SOE) MIGA Priority public sector projects (in MIGA’s strategic MIGA reaches new clients No arbitral focus areas**) are financed PROJECT ASSUMPTIONS award required MIGA increases business volume • Are well-designed and follow best international practices MIGA projects are aligned with development priorities • Are well-implemented • Generate financial returns MIGA takes on greater risk • Have positive socio-economic impacts MIGA reinsures sizeable portion of the risk MIGA’s sustainability is strengthened *Targeted to commercial lenders. **Africa, IDA, post-conflict, frontier markets, South-South, complex projects. ***MIGA mandate Sources: (i) Multilateral Investment Guarantee Agency policies and guidance on the NHSFO and NHFO-SOE guarantees. (ii) Multilateral Investment Guarantee Agency Operational Policies effective January 6, 2015. (iii) Multilateral Investment Guarantee Agency Convention, April 2018. (iv) Multilateral Investment Guarantee Agency Strategy and Business Outlook, various years. Note: MIGA strategic focus areas include Africa, complex projects, frontier markets, IDA countries, postconflict countries, South-South investments, and starting FY16, climate change and adaptation. HG = host government or sovereign; SOE = state-owned enterprise; SS = sub-sovereign. Evaluation questions. Based on the theory of change, the following questions and subquestions guided this evaluation: 55 Appendix C Methodology 1. To what extent have MIGA’s non-honoring insurance products been additional and effective in helping MIGA enhance its development mandate, particularly in its strategic focus areas and in underserviced markets? 2. To what extent have MIGA’s non-honoring insurance products expanded its business? To what extent have MIGA’s non-honoring insurance products helped address cross-border investors’ (international lenders) risk mitigation needs and helped mobilize additional capital with better terms for priority public sector projects? 3. How well has MIGA managed the risks associated with its non-honoring insurance coverage and under what conditions do these products work best? 4. Based on the findings in (1) to (3), what are the implications for MIGA’s 2021–23 Strategy? Scope. The Independent Evaluation Group (IEG) reviewed all 34 NH projects from 2011 (when MIGA issued its first NHSFO cover) to the third quarter of 2019. Of 34 NH projects, 8 have completed evaluations; of these, 7 have been rated. To situate the NH business line within the context of MIGA’s overall operations, the evaluation also reviewed MIGA’s political risk insurance (PRI) portfolio trends but did not conduct an in-depth review of the PRI projects. Approach and data collection methods. In answering the evaluation questions, IEG applied counterfactual analysis using the following data collection methods: • Project portfolio review analysis was conducted at two levels. First, to place MIGA’s non-honoring business line in the context of its overall operations, IEG reviewed MIGA’s NH and PRI products and their respective portfolio characteristics (for example, types of investments, clients, exposure size, country, sector, region). Second, IEG conducted a desk review of available project documents of the 34 NH projects covering both evaluated projects and projects that are not yet mature for ex post evaluation. The desk review analyzed project context; objectives; client and project characteristics; nature of the underlying transaction and investments; expected and actual (for eight evaluated projects only) development outcomes; alignment with host government and MIGA’s strategic priorities; MIGA’s additionality, role, and contribution; and risk factors and mitigants. Data sources consisted of MIGA project documents including early decision documents, country risk assessment, Underwriting Paper, President’s Report, Contracts of Guarantee, financing agreements, and other relevant documents, such as Project Evaluation 56 Appendix C Methodology Reports, Project Evaluation Report Validation Notes, and Project Performance Assessment Reports for the evaluated projects.1 • Policy document review and analysis assessed MIGA’s objectives and rationale expressed in the 2009 and 2013 Board documents, and consistency with subsequent implementing guidelines, policies, and procedures and MIGA Convention and Operational Regulations. A review of changes that occurred since MIGA sought Board authorization for the non-honoring coverages and the reasons for such changes helped IEG assess MIGA’s considerations in deploying the NH products. IEG also reviewed past and current MIGA Convention and Operational Regulations and Policies; various MIGA strategy documents; client surveys and past PRI market studies commissioned by MIGA; and MIGA policies, procedures, and risk assessment guidance and guidelines for NHSFO and NHFO-SOE coverages. IEG also reviewed relevant literature and studies from journals and market studies prepared by credit and investment insurance market participants. However, documents and information pertaining to MIGA’s pricing and pricing considerations of NH products, project-level data on amounts mobilized by MIGA from the private sector through the NH products, and underlying financing agreements between the lenders and the borrowers were not provided to IEG. The lack of information limited this evaluation from verifying the actual benefits of MIGA’s NH products, especially any loan pricing benefits that can be passed on to the sovereign, sub-sovereign, and SOE borrowers. • Benchmarking study compared MIGA’s NHSFO and NHFO-SOE products with similar credit enhancement products (insurance and guarantees) or complementary products offered by African Development Bank, African Trade Insurance, Asian Development Bank, an export credit agency, International Bank for Reconstruction and Development (IBRD), International Development Association (IDA), International Finance Corporation (IFC), and a private insurer. A comparison of the types of product, eligibility criteria, types of projects, and other characteristics is presented in appendixes G and H. Also, a background paper on the global credit and investment insurance market was produced for this evaluation to provide a broader context. • Semistructured interviews of key stakeholders helped fill information gaps that could not be filled from the project portfolio and policy document reviews or because of lack of information provided to IEG. Questionnaires were developed for each interview group and sent in advance of the interviews. In selecting the number of interviewees, IEG tried to achieve sufficient diversity of perspectives 57 Appendix C Methodology while being mindful of budget and time implications. The key stakeholders interviewed by IEG are shown in table D.1. Limitations The evaluation had limitations and challenges. First, the evaluation team was unable to interview sovereign, sub-sovereign and SOE borrowers and counterparties regarding their experience with MIGA’s NH products, their financing and risk mitigation financing requirements, the factors influencing their choice of commercial lenders, and the benefits and challenges of MIGA’s non-honoring products. This was partly owing to time and budget constraints and partly owing to changes in the project counterparts. Second, actual development outcomes from evaluation are not available to the same extent as with more established MIGA PRI products because the NHSFO- and (especially) NHFO-SOE-insured projects became effective only in the past eight years. Last, for the evaluation to be useful for drawing meaningful lessons, full cooperation and support by MIGA management and staff are essential. Equally important is agile access to data and information, project and policy documents, and unimpeded access to MIGA staff requested by IEG for this evaluation. Notes 1IEG conducted one Project Performance Assessment Report of an NHSFO project in Tunisia in FY15. 58 Appendix D Profile of Insured Projects Appendix D. Profile of Projects Insured by MIGA’s Non-Honoring of Financial Obligations Products Amount of Non-Honoring of Financial Obligations (NH) guarantees issued and number of NH projects. From 2011, when the first NHSFO insurance contract was issued until 2019, the Multilateral Investment Guarantee Agency (MIGA) had provided $11.3 billion in NHSFO and NHFO-SOE guarantees for 34 projects. The NH products represent a third (34 percent) of all guarantees issued by MIGA during this period and a third of its gross outstanding portfolio as of 2019. MIGA’s political risk insurance (PRI) product accounts for the rest. Of the total $11.3 billion NH amounts issued since 2011, MIGA retained $3 billion of NH risk in its portfolio, equivalent to 27 percent of its total NH maximum aggregate liabilities. Figure D.1 presents MIGA’s NH year-on-year net retention rate. By contrast, MIGA’s average retention rate is higher for its PRI product (50 percent) than for the NH product. As of 2019, the largest NH gross exposure was the $1.7 billion NHSFO cover to Duqm special economic zone authority (Oman).1 After reinsurance by a syndicate of private and public insurers, MIGA’s net exposure amounted to $247 million, equivalent to 15 percent of its maximum aggregate liability in the project. Figure D.1. MIGA Non-Honoring of Financial Obligations Gross and Net Exposure Amounts, 2011–19 Gross amounts issued MIGA net retention rate (%) 2,000 Net exposure amount Amount issued or exposure amount 1,800 1,600 1,400 ($, millions) 1,200 1,000 37 15 14 800 19 600 71 55 21 21 400 41 200 - 2011 2012 2013 2014 2015 2016 2017 2018 2019 Fiscal year Sources: MIGA Contracts Issued Database and Independent Evaluation Group staff calculations. Note: Net retention rate is the percentage of MIGA’s net exposure to MIGA’s total maximum aggregate liability. MIGA’s guarantee business had grown by an average of 11 percent annually, by amounts issued, from 2011 to 2018. Before 2011, the total volume of guarantees issued 59 Appendix D Profile of Insured Projects (all PRI) averaged $1.2 billion annually. After MIGA issued its first NH cover in 2011 through 2018, the total volume of guarantees issued annually almost tripled to $3.5 billion (figure D.2). Nearly all the NH contracts of guarantee remain effective (or active status). Contracts of guarantee of two NHSFO projects were canceled early.2 Figure D.2. MIGA Guarantee Amounts Issued, by Product, 2011–19 PRI NH $6,000 Guarantee amounts issued $5,000 $4,000 ($, millions) $3,000 $2,000 $1,000 $0 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019* Fiscal year Sources: MIGA Contracts Issued Database; MIGA annual reports, various years; Project Briefs in www.miga.org/projects. Note: This graph presents the guarantee amounts issued by MIGA for each fiscal year, which includes new guarantees issued and additional exposure amounts. MIGA = Multilateral Investment Guarantee Agency; NH = non-honoring of sovereign, sub-sovereign and state-owned enterprises financial obligations; PRI = political risk insurance. Types of investments insured.3 Ninety percent of the NH guarantee amounts issued insured nonshareholder loans of debt financiers or lenders to sovereign, sub-sovereign entities, or state-owned enterprises (SOEs). Loan guarantees and swaps (currency and interest rate swaps) accounted for the rest. Two NH-insured stand-alone cross-currency swap transactions (a new eligible MIGA investment) involved the Senegal Cross- Currency Swap transactions MIGA’s NH product also covered the local currency swap transaction (another first for MIGA) linked to an NHFO-SOE insured commercial loan to the Land and Agriculture Bank of South Africa. MIGA also insured interest rate swaps linked to the NHSFO-insured loans from a syndicate of commercial lenders to the Cambambe Hydropower II project. Type of NH cover.4 Of the 34 projects with NH coverage, 27 were covered by NHSFO contracts involving either a sovereign obligor, usually the ministry of finance (19 projects) or a subnational entity (in six projects). The remaining seven projects were each covered by an NHFO-SOE contract with a SOE as the borrower. 60 Appendix D Profile of Insured Projects Four of these seven projects with NHFO-SOE contracts insured the international commercial bank loans to Turkey’s Eximbank. Although there are other SOE borrowers, almost all have sovereign guarantees, usually from the ministry of finance acting as the obligor. Examples of these SOEs include Hungary Eximbank, National Highway 20 (Vietnam), Metro de Panama, Compagnie Tunisienne de Navigation (Tunisia), Volta River Authority (Ghana), and the four Bangladesh power projects (Ashuganj, Sirajganj 2, Sirajganj 3, and Ghorasal 3).5 With one exception (Kreditanstalt für Wiederaufbau or KfW), the NH guarantee holders of record were international private banks. However, several public financing institutions and insurers were part of the syndicated loans covered by MIGA NH products. Profile of MIGA guarantee holder of record. By design, the NH product was intended to support international commercial banks and financial institutions that finance public sector undertakings. Except for three NH transactions), the NH products insured syndicated loans involving several international lenders, mostly international private banks, although there were public banks, insurers, and official export credit agencies (ECAs) in the mix. Typically, a lead bank acts as the Agent or Arranger of the syndicated loan and serves as the MIGA guarantee holder of record, although the other participating banks are also insured by MIGA. MIGA’s NH exposure is highly concentrated among a few clients. Four guarantee holders account for over 65 percent of total NH amounts issued and half the number of NH projects. Majority of the NH guarantee holders of record are former PRI clients. Standard Chartered Bank, Citibank, and HSBC have been involved in five NH projects, while ING has three (all in Izmir), and Banco Santander, BNP Paribas, and Sociétè Générale have two each.6 Sector profile.7 MIGA’s NH projects are concentrated in two sectors (figure D.3), namely: • Ten projects in the finance and capital markets account for 33 percent of the total amount of NH guarantees issued (24 percent of net exposure and 29 percent by number of projects). One NH project covered a bond issuance whereas the rest were commercial loans for on-lending facilities, mostly to small and medium enterprises. • Twenty-four infrastructure projects make up 67 percent of the total amount of NH guarantee contracts issued (76 percent by net exposure and 71 percent by number of projects). Projects in the transport and power sectors each account for nearly a fifth of the gross amount issued. Examples of NH transport projects include the respective metro, ferry, and light rail projects of Istanbul and Izmir 61 Appendix D Profile of Insured Projects municipalities, the Panama Metro, and the Tunisia ferry. MIGA also issued NHSFO contracts to a group of lenders for the National Highway 20 road project in Vietnam and for the São Paulo Sustainable Transport project.8 Of the 11 projects in the power sector (equivalent to $2.1 billion in gross amounts issued), 9 involved either construction of new generation plants or the rehabilitation or refurbishment of older generation plants. MIGA also supported the upgrading of South Africa’s transmission system and the construction of new substations. MIGA’s NH product also supported the construction of a gas pipeline (Azerbaijan), the special economic zone (Oman) and Senegal swap (savings from the swap will be earmarked for infrastructure projects). In comparison, MIGA’s PRI portfolio is less concentrated. Still, projects in finance and capital markets account for 38 percent of gross amounts issued (32 percent by net exposure and 27 percent by number of projects). PRI-insured projects in the infrastructure sector make up 40 percent of gross amounts issued (40 percent by net exposure and 33 percent by number of projects). These infrastructure sector projects include airport, telecom, transport, and wastewater treatment projects. PRI-insured projects in the agribusiness, manufacturing, and services sectors and in the oil, gas and mining sector each account for 13 percent of MIGA’s gross amounts issued during the 2011 to 2019 period. Although MIGA did not issue an NH cover directly to an agribusiness entity, it covered (through NHFO-SOE) a syndicated commercial loan to the Land and Agricultural Development Bank of South Africa in 2017 to on-lend to small and medium agribusiness enterprises. Figure D.3. MIGA Guarantees Amounts Issued, by Sector, 2011–19 a. Non-Honoring of Financial Obligation b. PRI Sources: MIGA Contracts Issued Database and Independent Evaluation Group staff calculations. Note: MIGA = Multilateral Investment Guarantee Agency *Includes Tourism and Retail Sectors 62 Appendix D Profile of Insured Projects Profile by region. Half of NH guarantee amounts issued by MIGA went to public sector projects in Europe and Central Asia, of which 60 percent were for 10 subnational and SOE projects in Turkey. Africa, one of MIGA’s past strategic focus area, received a quarter of the NH amounts issued, but over half of the amount covering nonpayment risks by three SOEs in South Africa (figure D.4). The Middle East and North Africa region accounts for 17 percent of total NH amounts issued mainly because of a large project (Oman). Africa’s share of business volume is the same for both NH and PRI products. However, compared with MIGA’s NH business, the regional distribution of MIGA’s PRI portfolio is less skewed if measured by the gross amount issued. Figure D.4. NH and PRI Share to Total MIGA Guarantee Amounts Issued, by Region, 2011–19 a. Non-Honoring of Financial Obligation b. PRI EAP, 7 EAP, SSA, 20 SSA, 14 25 SAR, 4 ECA, ECA, SAR, 3 42 MEN 28 MENA A, 17 ,6 LAC, LAC, 24 10 18% in IDA 40% in IDA Sources: MIGA Contracts Issued Database and Independent Evaluation Group staff calculations. Note: EAP = East Asia and Pacific; ECA = Europe and Central Asia; IDA = International Development Association; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. A closer look at MIGA’s gross amounts issued for each region (figure D.5) shows that its PRI business still forms a large share of MIGA’s business volume across all six regions. 63 Appendix D Profile of Insured Projects Figure D.5. Guarantee Amounts Issued ($ millions), by MIGA Product, 2011–19 NH PRI East Asia and Pacific $740 $3,025 Europe and Central Asia $4,748 $6,141 Latin America and the Caribbean $1,096 $5,351 Middle East and North Africa $1,919 $1,359 South Asia $487 $689 Sub-Saharan Africa $2,276 $5,588 0% 20% 40% 60% 80% 100% Share to total guarantee amount issued (%) Sources: MIGA Contracts Issued Database and Independent Evaluation Group calculations. Note: NH = Non-Honoring of Financial Obligations; PRI = political risk insurance. Country concentration of NH portfolio. The $11.3 billion of commercial loans insured by MIGA NH financed public sector projects in 14 countries. Turkey was the number one country beneficiary of MIGA NH guarantees, by amount issued and by number of projects (see table D.1). Together, five countries accounted about three-quarters of the MIGA NH amount issued (on a gross basis). Three countries (Oman, Turkey, and South Africa) absorbed half of MIGA NH business. Figure D.12 further shows MIGA NH guarantees issued in each of the 14 countries compared with the state obligation insurance of other members of the Berne Union’s Investment Insurance Committee. 64 Appendix D Profile of Insured Projects Table D.1. Three Countries Account for Half of MIGA’s NH Amounts Issued Amounts Share of Total Issued Amounts Issued NH Projects Host Country ($, millions) (percent) (no.) Turkey 2,748 24 10 Oman 1,701 15 1 South Africa 1,460 13 3 Azerbaijan* 1,113 10 1 Hungary 887 8 2 Vietnam 740 7 2 Panama 640 6 2 Angola 528 5 3 Bangladesh 487 4 4 Brazil 361 3 1 Tunisia 218 2 1 Senegal 199 2 2 Colombia 95 1 1 Ghana 88 1 1 Total NH 11,266 34 Sources: MIGA Contracts Issued Database; Project Briefs disclosed in MIGA’s external website, https://www.miga.org/projects; and Independent Evaluation Group staff calculations. Note: NH = Non-Honoring of sovereign, sub-sovereign and SOE financial obligations. Data are from 2011 to 2019. *NHSFO guarantee was canceled in September 2019. Supporting MIGA’s strategic priority areas. When MIGA sought approval of the NHSFO product in 2009, its strategy focused on the following priority areas: Africa, complex projects, conflict-affected countries, frontier markets, International Development Association (IDA), and South-South investments. Thus, in the 2009 and in the 2013 Board documents, MIGA mentioned as development rationale for the approval of the NH products, support for foreign investments in these priority areas. Since then, MIGA’s priority areas changed but IDA and countries with FCS or remained as priority areas in MIGA strategies.9 Climate and energy efficiency was added as a priority area in MIGA’s FY15–17 strategy, and in its current strategy (FY18 to FY20), climate change adaptation and mitigation became MIGA’s third focus area in addition to supporting productive foreign investments in IDA and FCS countries. Table D.2. MIGA Priority Areas Changed Over Time but Support for IDA and FCS Countries Has Been Constant FY12– MIGA Priority Areas FY0911 14 FY15–17 FY18–20 Africa ✓ a a a Climate change and energy efficiency x x ✓ b 65 Appendix D Profile of Insured Projects FY12– MIGA Priority Areas FY0911 14 FY15–17 FY18–20 Climate change adaptation and mitigation x x ✓ ✓ Climate finance x x ✓ b Complex projects ✓ x x x Conflict-affected countries ✓ c c c IDA ✓ ✓ ✓ ✓ FCS ✓ ✓ ✓ Middle income countries x x ✓ x Postconflict countries ✓ ✓ c c South-South investments ✓ ✓ x x Transformational projects x x ✓ x Sources: MIGA Operational Directions FY09–11; MIGA FY12–14 Strategy: Achieving Value-Driven Volume; MIGA Strategic Directions FY15–17 Revision; MIGA Strategy and Business Outlook FY18–20. Note: FCS = Fragile and Conflict-affected Situations; IDA = International Development Association. (i) subsumed under IDA; (ii) merged into Climate Change Adaptation and Mitigation; (iii) merged into FCS; and (x) not identified as priority Though there were NH projects in these strategic focus areas, MIGA applied its PRI products more to meet its past and current strategic focus areas. As table D.3 indicates only one in five NH projects are in IDA countries compared with two in five PRI projects. There was only one NH project in an FCS country (although representing 5 percent of NH amounts issued) compared with 58 PRI projects (representing 13 percent of amounts issued for PRI) in fragile and conflict-affected areas. The breakdown of MIGA’s guarantee products issued to IDA and FCS countries shows that its PRI business still forms a substantial share of MIGA’s business volume across these two priority areas (figure E.6). It should be noted that the climate change adaptation and mitigation priority area is a small percentage in terms of number of projects, but the gross amount issued is sizable. 66 Appendix D Profile of Insured Projects Table D.3. MIGA Guarantees Issued, by Strategic Focus Areas, 2011–19 Gross Instruments and Amountc Total Gross Amount Projects Priority Areasa, b ($, millions) (percent) (no.) NHSFO and NHFO-SOEd 11,266 34 Africa (2011d–17) 2,276 20 9 Complex Projects (2011 –17)d 2,963 326 13 FCS (2011 -Present) d 512 5 1 IDA (2011 -Present) d 2,026 18 10 South-South investments (2011 –17) d 0 0f Climate Change Adaptation and Mitigation 2,309 e 20 3 (2017-Present) Partial (below 100%) 2,307g 20 2 Full (100%) 2 0 1 Not elsewhere classified 6,319 56 12 PRI 22,154 274 Africa (2011–17) 5,588 25 107 Complex Projects (2011–17) 4,210 19 36 FCS (2011-Present) 3,012 14 56 IDA (2011-Present) 8,828 40 131 South-South investments (2011–17) 1,884 9 38 Climate Change Adaptation and Mitigation 2,086 9 44 (2017-Present) Partial (below100%) 307 1 3 Full (100%) 1,780 8 41 Not elsewhere classified 11,199 51 93 Source: MIGA Contracts Issued Database and MIGA Strategy documents. Note: Based on MIGA strategic focus areas mentioned in the 2009 Board document seeking approval of the NHSFO product and MIGA’s strategic focus areas in 2013, when MIGA sought Board approval of the NHFO-SOE product. FCS = fragile and conflict-affected situation; IDA = International Development Association; NHSFO = Non-Honoring of Sovereign and Sub-Sovereign Financial Obligations; NHFO-SOE = Non-Honoring of Financial Obligations by State-Owned Enterprises; PRI = political risk insurance.. a. MIGA’s 2021–23 Strategy identified IDA, FCS, and Climate Change Adaptation and Mitigation as its three focus areas. b. Projects may be classified in more than one priority area, and the count is not mutually exclusive. Sum of amounts issued will exceed total NH and PRI issuance and percentages will exceed 100 percent because projects may be classified in more than one priority area. c. Gross amounts represent MIGA’s maximum aggregate liability at issuance. Net amount represents insured amount retained by MIGA after reinsurance. d. Although the non-honoring instrument was approved in FY09, the first project was not issued until FY11. e. f. One NH project was considered consistent with the Climate Change Adaptation and Mitigation priority area based on 8.245 percent of the project meeting this MIGA priority area; the other NH was also considered consistent with this MIGA priority area because 15 percent of the project meets the criteria based on MIGA’s calculation g. The large amount is because of a single project. MIGA estimated 15.5 percent of the guarantee amount issued is consistent with climate adaptation, hence the large number. 67 Appendix D Profile of Insured Projects Figure D.6. PRI Is the MIGA Product of Choice in IDA and FCS Countries a. Gross Exposure Issued in IDA Countries b. Share to Total Guarantees Issued in IDA Countries (percent) PRI PRI NHFO 2019 2019 2018 2018 2017 2017 Fiscal year Fiscal year 2016 2016 2015 2015 2014 2014 2013 2013 2012 2012 2011 2011 - 500 1,000 1,500 2,000 0% 50% 100% c. Gross Amount Issued in FCS d. Share to Total Amounts Issued in FCS (US$, millions) (percent) PRI NHFO PRI NHFO 2019 2019 2018 2018 2017 2017 2016 2016 Fiscal year Fiscal year 2015 2015 2014 2014 2013 2013 2012 2012 2011 2011 - 500 1,000 0% 50% 100% Source: MIGA Contracts Issued Database. Note: FCS = fragile and conflict-affected situation; IDA = International Development Association; NH = Non-Honoring of sovereign, sub-sovereign and SOE financial obligations; PRI = political risk insurance. 68 Appendix D Profile of Insured Projects Income level profile. A similar pattern emerges by country income level. Sixty-eight percent of the NH projects (by project count) are in upper-middle- and high-income countries, compared with 42 percent with its PRI product. The disparity is even more staggering in terms of amounts issued (figure D.7). Nine in 10 NH projects are in high- income and upper middle-income countries compared with over half for the PRI projects (53 percent). Gross amounts issued to projects in high-income and upper middle-income countries tend to be higher, whether insured by NH or PRI. Figure D.7. An Overwhelming Majority of MIGA Non-Honoring Projects are in High- Income and Upper-Middle Income Countries 100% Low income Lower middle Low income 90% income Share to total amount Issued (percent) 80% Lower middle 70% income 60% Upper middle 50% income 40% Upper middle 30% income 20% 10% High income High income 0% Non-Honoring Political Risk Insurance Sources: MIGA Contracts Issued Database, World Bank country income list, and Independent Evaluation Group staff calculations Country risk profile of MIGA NH projects. MIGA’s NH projects are clustered in countries that are rated low- or mid-risk (B+ to BB country risk rating) by Standard & Poor’s, Fitch, and Moody’s rating agencies—a trend that reflects MIGA’s application of its country risk rating eligibility criteria for NH guarantee. In comparison, MIGA’s PRI supported projects in countries with riskier ratings (B to CCC−). Figure D.8 plots the country risk rating of NH and PRI projects at issuance. Projects with large MIGA exposure amounts were in countries with investment-grade ratings (BBB− and above), whether covered by NH or PRI. 69 Appendix D Profile of Insured Projects Figure D.8. Sovereign Credit Risk Rating Profile at Issuance by External Credit Rating Agencies, 2011–19 Sources: Standard & Poor’s, Fitch, and Moody’s (for the country risk) 2011 to 2019, MIGA Contracts Issued Database, and Independent Evaluation Group staff calculations. Note: The shade of the circles reflects MIGA’s maximum aggregate liability at issuance. NH = Non-honoring of sovereign, sub-sovereign and SOE financial obligations insurance; PRI = political risk insurance. Although MIGA adhered to its minimum eligibility country risk rating criteria at issuance, recent country risk ratings of the NH projects increased as several host countries experienced rating downgrades. Figure D.9 shows the clustering of the NH and PRI projects based on sovereign credit risk ratings on June 30, 2019. In the 14 countries where the 34 NH projects were located, only Vietnam (from BB− to BB) and Hungary (BB+ to BBB−) had credit rating upgrades since the first NH cover to Hungary Eximbank–Magyar Eximbank Secured Funding was issued in 2014. 70 Appendix D Profile of Insured Projects Figure D.9. MIGA's NH and PRI Amounts Issued, by Recent Sovereign Credit Risk Ratings, 2011–19 Source: Standard & Poor’s, Fitch, and Moody’s (for the country risk), MIGA Contracts Issued Database, and Independent Evaluation Group staff calculations. Note: Shade of the bubbles reflect MIGA’s maximum aggregate liability at issuance. The credit ratings are Standard & Poor’s, Fitch, and Moody’s credit ratings as of April 2019. NH = Non-Honoring of Financial Obligations; PRI = political risk insurance. Seven of 13 countries where MIGA had NH exposure experienced sovereign credit rating downgrades after the initial NH contract was issued or became effective. Country risk rating changes for the NH projects are shown in table D.4. Table D.4. External Rating Agencies’ Risk Ratings of Countries with NH Projects Show Change in Risk Profile After MIGA Issuance External Most Recent Rating Agency External Host Fiscal Country Risk Rating Agency Project Name Country Year at Issuance* Country Risk** Cambambe Hydro Power II Angola 2013 BB- B- Cambambe Hydro Power II Angola 2016 B B- Cambambe Hydro Power II (Swap) Angola 2019 B B- Southern Gas Corridor Closed Joint Stock Azerbaijan 2018 BB+ BB+ Company Ashuganj Power Station Company Limited Banglades 2013 BB- BB- h Bangladesh Power Development Board Banglades 2017 BB- BB- h North-West Power Generation Company Limited I Banglades 2016 BB- BB- h North-West Power Generation Company Limited Banglades 2017 BB- BB- II h 71 Appendix D Profile of Insured Projects External Most Recent Rating Agency External Host Fiscal Country Risk Rating Agency Project Name Country Year at Issuance* Country Risk** São Paulo Sustainable Transport Project Brazil 2015 BB+ BB- Financiera del Desarrollo SA Colombia 2015 BBB BBB- Takoradi III Power Plant Ghana 2012 B+ B Hungarian Export-Import Bank Private Limited Company I Hungary 2014 BB+ BBB Hungarian Export-Import Bank Private Limited Company II Hungary 2017 BBB- BBB Panama Metro Panama 2012 BBB BBB+ Metro de Panama, SA Panama 2014 BBB BBB+ Senegal US$/Euro Cross-Currency Swap Senegal 2012 B+ B+ Senegal Eurobond Cross-Currency Swap Senegal 2016 B+ B+ South Eskom Holdings SOC Ltd Africa 2016 BBB- BB South Development Bank of Southern Africa Limited Africa 2017 BB BB Land and Agricultural Development Bank of South South Africa Africa 2017 BB BB Compagnie Tunisienne de Navigation SA Tunisia 2012 BBB- B Istanbul Electricity Tram and Tunnel General Directorate (Otogar-Bağcılar-Ikitelli-Olimpic Village Metro) Turkey 2011 BB B+ Istanbul Municipality (Kadikoy-Kartal-Kaynarca Metro Line) II Turkey 2011 BB B+ Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank I) / I Turkey 2015 BB+ B+ Izmir Buyuksehir Belediyesi (Izmir Marine Transport)/ Izdeniz A.S. (Izmir Ferry) / I Turkey 2013 BB+ B+ The Municipality of Izmir (Izmir Railcars) / II Turkey 2014 BB+ B+ Istanbul Ulasim A.S. (Üsküdar-Ümraniye- Çekmeköy Metro) / III Turkey 2015 BB+ B+ Izmir Buyuksehir Belediyesi (Izmir Tramway-Light Rail; Izmir Metro) / III Turkey 2015 BB+ B+ Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank I) / II Turkey 2016 BB B+ Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank II) / III Turkey 2017 BB B+ Türkiye Ihracat Kredi Bankasi A.S. (Turkish Eximbank III) / IV Turkey 2019 B+ B+ BT20—Cuu Long Joint Stock Company Vietnam 2014 BB- BB VNeco Hoi Xuan Investment & Electricity Construction (Hoi Xuan Hydropower) Vietnam 2016 BB- BB Duqm Special Economic Zone Authority Oman 2019 BB BB Sources: * *External Ratings as of June 30, 2019 and at issuance ratings collected from Standard & Poor’s, Fitch, and Moody’s where available. 72 Appendix D Profile of Insured Projects Note: NH = Non-Honoring of Sovereign, Sub-Sovereign and State-Owned Enterprises Financial Obligations; N.I. = not indicated. Consistency with World Bank Group Country Assistance Strategies. Nearly all NH projects were aligned with Bank Group country strategies and priorities. The six subnational projects in Turkey were envisaged in the Bank Group’s FY12–15 Country Partnership Strategy and in the FY18–21 Country Partnership Framework (CPF) for Turkey. These projects were also identified in the Bank Group’s Turkey Sustainable Cities Joint Implementation Plan, a road map for an integrated Bank Group approach to provide comprehensive public-private solutions in response to the government’s national urbanization framework. MIGA support for the Bangladesh power projects is consistent with the CPF objective of increasing the country’s power generation capacity and access to clean energy. The NH projects—Ghorasal 3, Sirajganj II and Sirajganj III— are listed in the CPF results framework as MIGA activities that would contribute to the achievement of this CPF objective. The construction of the Panama Metro was a priority in the Bank Group’s CPF for 2015–21. The rehabilitation of Vietnam’s National Highway 20 was also considered a priority infrastructure project in the Country Assistance Strategy. Except for one project, evaluated NH projects were consistent with Bank Group country strategies and the host country’s development priorities. Collaboration and cooperation with International Bank for Reconstruction and Development (IBRD) and International Finance Corporation (IFC). NH projects with the Bank Group’s Country Assistance Strategy and CPFs. IFC and MIGA support to the Istanbul and Izmir metro, tram, and light rail projects also reflects the Bank Group engagement under the Turkey Sustainable Cities program. In addition, several projects have either parallel, sequential, or cofinancing from either IBRD and IDA or IFC. Table D.5. MIGA Supported Several IFC and IBRD Projects through Its NH Products Fiscal MIGA Gross World Bank Commitment Year Exposure Group Fiscal Instrument Amount Project Name Issued ($, millions) Institution Year Type ($, millions) Kadikoy-Kartal-Kaynarca 2011 243.0 IFC 2009 Loan 67.9 Metro (Istanbul) Izmir Tramway 2014 55.0 IFC 2014 Loan 75.8 Izmir Light Rail 2015 23.5 IFC 2015 Loan 25.0 São Paulo Sustainable 2015 300.0 IBRD 2013 IPF 300.0 Transport Eskom Holdings SOC a 2016 470.0 IBRD 2010 IPF 3,750.0 Üsküdar-Ümraniye- 2016 134.0 IFC 2016 Loan 65.3 Çekmeköy Metro Line (Istanbul) 73 Appendix D Profile of Insured Projects Fiscal MIGA Gross World Bank Commitment Year Exposure Group Fiscal Instrument Amount Project Name Issued ($, millions) Institution Year Type ($, millions) Southern Gas Corridor/ 2018 215.0 IBRD 2017 IPF PRG 800.0 TANAP Source: Business Intelligence for MIGA and World Bank; IFC Commitment Database. Note: IBRD = International Bank for Reconstruction and Development; IFC = International Finance Corporation; IPF = Investment Project Finance; NH = Non-Honoring of Financial Obligations; PRG = partial risk guarantee. a. For the upgrading of Eskom’s transmission line only. IBRD’s $3.7 billion loan to Eskom for power generation, associated transmission and institutional strengthening was approved in FY10. MIGA’s non-Honoring product in a global context. The Berne Union (BU), the leading global association for the export credit and investment insurance industry, records non- honoring products as “state obligation insurance.” BU’s data set on state obligation insurance by its Investment Insurance Committee members shows that private insurers accounted for a significant amount (and percent share) of state obligation insurance new business of BU members.10 Figure D.10 shows MIGA’s small share, averaging approximately 8 percent for the past eight years. Compared with the other two BU member multilateral insurers (African Trade Insurance Agency, Islamic Corporation for the Insurance of Investment and Export Credit) that offer similar state obligation insurance, MIGA’s percent share of this business is sizable, though the other multilateral insurers’ operations have limited geographical reach. Private companies still account for most of the BU’s state obligation insurance new business in calendar years 2010 to 2018, followed by multilateral insurers, of which MIGA has a significant share. The decline in ECAs’ state obligation insurance business, starting in 2012, reflects the change in the reporting of trade-related state obligation insurance to the Export Credit Committee of the Berne Union and is not owing to the decline in their state obligation insurance business. 74 Appendix D Profile of Insured Projects Figure D.10. MIGA Has a Small Share of Overall State Obligation Insurance among Berne Union Members but Exceeds Other Multilateral Insurers Other Multilaterals MIGA ECA Private Share of MIGA NH Volume to Total State Obligation Insurance Volume (%) 20 14% New Business Amount Share of MIGA NH Volume (%) 12% (uS$ Billions) 15 10% 8% 10 6% 4% 5 2% - 0% 2010* 2011 2012 2013 2014 2015 2016 2017 2018 Calendar Year Sources: Berne Union Investment Insurance data for calendar years 2010 to 2018 and Independent Evaluation Group staff calculations. Note: *Data relate to state obligation insurance new business by Berne Union Investment Insurance Committee members only and presents a snapshot of the global commercial, credit and investment insurance market. Data exclude state obligation insurance provided to Category One (except South Africa) and some Category Two MIGA member countries that are nonborrowing members of the International Bank for Reconstruction and Development. The declining business volume of ECAs starting 2013 reflects the shift in the reporting of trade-related state obligation insurance from the Berne Union Investment Insurance Committee to its Export Credit Committee and is not due to decreasing business. ECA = export credit agency; MIGA = Multilateral Investment Guarantee Agency; NH = non-honoring insurance products. Compared with private insurers and ECAs offering state obligation insurance, MIGA and other multilateral insurers have higher risk tolerance. Private and public insurer members of the Berne Union issued more (by amount) state obligation insurance in high investment-grade countries (above A− and the BBB− country risk range) compared with multilateral insurers, including MIGA. Approximately half of state obligation insurance issued by BU private insurers was in countries with an A+ sovereign credit risk rating or better (high investment grade). A third of the sovereign nonpayment business issued by ECAs was also in these low-risk countries (figure D.11). By contrast, MIGA’s NH business was primarily in countries with sovereign credit ratings ranging from BB+ to BB−. But figure D.11 also reveals the presence of private, public, and other multilateral insurers in the same space where MIGA has NH exposure and even in higher-risk countries (B− to B+ ratings) where MIGA has minimal NH exposure. 75 Appendix D Profile of Insured Projects Figure D.11. Berne Union Members’ State Obligation Insurance Business Volume, by Country Risk Rating, CY11–18 Above A− BBB+, BBB, BBB− BB+, BB, BB− B+, B, B− Below CCC+ 100% 90% New business amount (%) 80% 70% 60% 50% 40% 30% 20% 10% 0% Private insurers ECAs Multilaterals MIGA Other multilaterals Berne Union Investment Insurers Sources: Berne Union Investment Insurance data, Standard & Poor’s, Fitch, and Moody’s data, and Independent Evaluation Group staff calculations. Note: Data relate to state obligation insurance by Berne Union Investment Insurance Committee members only and present only a snapshot of the credit and investment insurance market. Data presented are by calendar year. CY = calendar year; ECA = export credit agency; MIGA = Multilateral Investment Guarantee Agency. Figure D.12 shows the 14 countries and compares MIGA’s NH exposure against the state obligation insurance new business amount of Berne Union Investment Insurance Committee members—private, bilateral, and other multilateral insurers—from calendar year (CY)11 to CY18. It shows that MIGA’s gross exposure was higher in Hungary, South Africa, and Bangladesh when it issued the NH contracts of guarantee. However, the graphs also show that other Berne Union members have provided nonpayment of state obligation cover in the 14 countries before and after MIGA’s NH support. 76 Appendix D Profile of Insured Projects Figure D.12. State Obligation Insurance of MIGA Compared with Other Berne Union Insurers by Host Country, CY11–18 a. Turkey b. Panama 3,000 500 2,500 400 New Business Amount 2,000 New Business Amount 300 (US$ Millions) 1,500 (US$ Millions) 200 1,000 500 100 0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2010 2011 2012 2013 2014 2015 2016 2017 2018 Other Multilaterals MIGA Other Multilaterals MIGA c. Senegal d. Ghana 500 500 400 400 New Business Amount New Business Amount 300 300 (US$ Millions) (US$ Millions) 200 200 100 100 0 0 2011 2016 2010 2012 2013 2014 2015 2017 2018 2010 2011 2012 2013 2014 2015 2016 2017 2018 Other Multilaterals MIGA Other Multilaterals MIGA e. Angola f. Hungary 1,400 1,200 700 New Business Amount 1,000 New Business Amount 600 800 (US$ Millions) 500 (US$ Millions) 600 400 300 400 200 200 100 0 0 2013 2010 2011 2012 2014 2015 2016 2017 2018 2010 2011 2012 2013 2014 2015 2016 2017 2018 Other Multilaterals MIGA Other Multilaterals MIGA g. Colombia h. Brazil 77 Appendix D Profile of Insured Projects 250 200 1,000 New Business Amount New Business Amount 150 800 (US$ Millions) (US$ Millions) 600 100 400 50 200 0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2017 2010 2011 2012 2013 2014 2015 2016 2018 Other Multilaterals MIGA Other Multilaterals MIGA i. Azerbaijan j. South Africa 1,200 1,400 1,200 New Business Amount 1,000 New Business Amount 1,000 (US$ Millions) 800 (US$ Millions) 800 600 600 400 400 200 200 0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2010 2011 2012 2013 2014 2015 2016 2017 2018 Other Multilaterals MIGA Other Multilaterals MIGA k. Oman l. Vietnam 800 3,000 New Business Amount New Business Amount 2,500 600 (US$ Millions) (US$ Millions) 2,000 400 1,500 1,000 200 500 0 0 2011 2010 2012 2013 2014 2015 2016 2017 2018 2010 2011 2012 2013 2014 2015 2016 2017 2018 Other Multilaterals MIGA* Other Multilaterals MIGA Source: Berne Union Investment Insurance database calendar years 2010 to 2018 and Independent Evaluation Group staff calculations. Note: *The graph for Oman excludes MIGA’s and other multilateral insurers’ new business amounts for 2019, New business amount issued by other multilateral insurers in 2019 was not available at the time this report was written. 78 Appendix D Profile of Insured Projects Notes 1Project Brief—Special Economic Zone at Duqm. https://www.miga.org/project/special- economic-zone-duqm. 2 In 2019, another NHSFO guarantee contract was canceled for a gas pipeline in the ECA region was canceled and another NHSFO contract of guarantee expired. 3 Description of the NH-insured projects are available in the respective projects’ Environment and Social Review Summary (ESRS) and Project Briefs disclosed in MIGA’s external website, https://www.miga.org/projects. 4 Project details described in these two paragraphs are available in the respective projects’ Environment and Social Review Summary (ESRS) and Project Briefs disclosed in MIGA’s external website, https://www.miga.org/projects. 5 Information on these projects can be found in their respective Project Briefs disclosed in M IGA’s external website, https://www.miga.org/projects. 6 NH guarantee holders are identified in the respective projects’ Environment and Social Review Summaries (ESRS) and Project Briefs disclosed in MIGA’s ext ernal website, https://www.miga.org/projects 7 IEG used MIGA Sector Teams or Management Group Name for the sector breakdown. 8The São Paulo Sustainable Transport Project intends to rehabilitate and upgrade 650 kilometers of existing paved roads and the reconstruct two bridges in the São Paulo state. Environmental and Social Review Summary—São Paulo Sustainable Transport, https://www.miga.org/project/sao-paulo-sustainable-transport-0. 9 Later strategies identified MIGA’s priority areas as IDA-eligible, FCS, innovation, and Climate and Energy Efficiency (MIGA’s FY15–17 strategy) and IDA, FCS and Climate Change Adaptation and Mitigation in MIGA’s FY18–20 strategy (current). 10 Made up of private credit and investment insurance companies, bilateral or public insurers, and specialized multilateral credit and investment insurance companies such as MIGA. Appendix E. Comparison of MIGA Non-Honoring of Financial Obligations Products with Non-Payment Guarantee Products of Selected Institutions Table E.1. Comparison of MIGA’s Non-Honoring of Financial Obligations Products with Selected Institutions Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf Product name Non-Honoring of • Project-based • Partial Credit • Sovereign Political Risk Partial Credit Political Risk Political Financial Obligation by Guarantees: Guarantee Guarantees: Insurance (PRI) Guarantee Insurance of Risk Sovereign, Sub- Loan (PCG) Partial Credit product (PCG) Non-Honoring Insurance sovereign, State- Guarantees; • Full Credit Guarantee includes of a Sovereign product owned Enterprises o Public Guarantee (PCG)c coverage of Obligation (PRI- includes sector (FCG) Non-Sovereign nonpayment NHS) coverage projects Guarantees: risk by host Contract of (Partial African government Frustration nonpayme Credit Development and its agencies Insurance (CFI) nt risk by Guarantee Fund Partial foreign or PCG) Credit govern- Guarantee ments on o Private (ADF-PCG) cross- sector border projects loans or (Partial contracts. Risk PRI Guarantee products or PRG) include o Hybrid Customer guarantee Focused coverage Insurance; Payment Compre- Guarantees: hensive Direct Nonpaym Payment ent 80 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf Guarantee; Project World Bank- Finance supported Risk Letter of Insurance; Credit Structured Credit Insurance. Risks covered All payment risks Loan guarantees All credit risks— Partial Credit All payment risks Comprehensive PRI of NHS: Loan Failure of a provided for public sector generally no Guarantee covers risk coverage for losses resulting sovereign unconditional and projects cover “carve-outs.” Both a portion of a specified from nonpayment entity to irrevocable debt service PCG and FCG scheduled portion of a debt by a sovereign honor defaults under a provide an payments of service obligation. borrower or payment commercial loan irrevocable commercial debt ADB’s PCG covers guarantor. In obligations that is made by a promise by IFC to instruments nonpayment by some cases, the under a private lender to a pay all shortfalls against all risks or the borrower or ECA can consider promissory government in of principal specific events of issuer (for any insuring loans note, bond, the context of a and/or interest up defaults by reason) on the made to a state- loan, or public sector to a borrowers from guaranteed owned entity or guarantee; project. predetermined both public and portion of the to quasi- or sub- failure of a Previously known amount. PCG and private sectors. principal and sovereign sovereign as PCGs, they FCG cover ADF-PCG: partial interest due governments. entity to usually cover creditors guarantee of debt CFI: covers a honor defaults of debt irrespective of the service range of payment service payment cause of default. obligations of commercial and obligations by a public sector IFC’s PCG also has low-income political risks, under a borrower liquidity backstop countries (LICs) including buyer letter of regardless of the features, that is, if and well nonpayment; credit; failure cause of the debt a borrower faces performing SOEs buyer bankruptcy of a service default. temporary in LICs or insolvency; sovereign liquidity cancelation of entity to problems, IFC’s import or export honor its Loan guarantees hard- for private sector guarantee may be permits; currency drawn on to conversion and currency or projects cover prevent a default. transfer; a host local 81 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf debt service If the borrower government’s currency defaults on a repays IFC in a moratorium on payment commercial debt timely manner, debt; war, obligations for all or part of the guarantee revolution or under a the debt term, amount can be insurrection; and guarantee where such reinstated. contract agreement defaults are cancelation issued in caused by support of a government project; failure to meet its failure of the payment government obligations under guarantor, project contracts such as a to which it is a ministry of party and where finance, the specific cause central bank, of such payment government- default is covered owned bank, under the World or SOE, to Bank's loan honor its guarantee. payment Risks covered obligations may include may under a include risks such guarantee as currency issued in inconvertibility or support of a nontransferability; trade domestic political contract. force majeure risks such as expropriation, war and civil disturbance or material adverse 82 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf government actions; failure by the host government or an SOE to make contractually agreed payments; regulatory risk; and other specific risks that the government undertakes in a specific project. Payment guarantees cover payment defaults of non-loan- related government payment obligations (for example, recurring off- taker payments under a power purchase agreement or early termination payments under a concession agreement), to private entities or a foreign public entity where such 83 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf payment obligations (arising from contract, law, or regulation) require credit enhancement. Eligible projects Public sector Public sector and New investments PCGs can be used Private and public Private and public Public and private New and undertakings or private sector (including for both public sector projects. and private sector sector projects, existing projects. projects. expansion, sector and private PPP, projects. (limited recourse including PPP public sector Cross-border or Private sector privatization and sector investment financing) projects, financed and or foreign investments loan guarantees concession projects, projects. PPP by lending to the private only. may cover also transactions), or a especially in projects. ECA country sector Investments must be debt that pool of new infrastructure and company with projects; PPP located in MIGA refinances assets in a coverage to short assets or projects. member countries underlying project developing and medium- investments in an debt. member country. term instruments emerging market, Coverage of Mainly for trade such as or to emerging private sector, finance and commercial paper market related to PPP corporate finance issued by both governments that projects but only transactions, private and public purchase the ECA if the payment including support financial country’s goods obligations of the for small and institutions. and services. private project medium-sized sponsor is fully enterprises counter (SMEs). Rarely guaranteed by used for PPP the host transactions. government. Projects must be located in IBRD and IDA countries. 84 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf Eligible Commercial banks and Commercial Private lenders Private lenders Commercial Private financial Private sector Equity guarantee capital market debt lenders, corporate (loans) and and capital banks and capital institution or an companies; investors; holders investors bond issuers in investors (bonds) market investors market debt investor providing lenders. financial the bond market. Subnational investors debt financing to institutions (loan guarantees government a project that is and capital for public sector bodies/agencies eligible for ADB markets, and private sector financing. Public importers projects) institutions may and also benefit from exporters, an ADB PCG project provided that finance they operate on a lenders commercial basis. including ECAs and multilateral agencies. Eligible Creditworthy Government or Private lenders All AfDB Creditworthy PCGs can be Loans to Creditworthy obligors or sovereign, sub- sovereign; SOEs and investors borrowers. sovereign, sub- applied to loans emerging market (quality) borrowers sovereign, and SOEs ADF-PCG are sovereign, and or other debt government or, private and restricted to low SOEs. instruments under certain public sector income member issued by private conditions, to a borrowers countries and public sector state-owned (for example, projects (limited entity or quasi- public, sub- recourse sovereign sovereigns, financings), government to state-owned public-private finance purchase banks, partnerships, exports from the SOEs); corporates, and ECA country emerging (sub)sovereign market and entities OECD countries Eligible Debt or payment Project-based Debt (domestic or Loans (foreign Debt investments ADB provides Loans issued Cross-border underlying obligation; cross- guarantees: For international) and local PCGs to lenders directly to and 85 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf assets and currency and interest Loan guarantees, investments currency) and of most forms of sovereign domestic investments rate swaps; EPC and equity or debt. (either bonds or bonds, bonds to debt. These borrowers, semi- loans related performance For payment loans), issued by private sector include sovereign (principal, bonds; other forms of guarantees, IFC’s private PCGs can cover commercial bank obligors or fees and cross-border nonloan related sector clients. the principal for loans, shareholder guaranteed by interest nonequity investments government PCG can be bullet maturity of loans, loans sovereign rates), trade payment denominated in corporate bonds, guaranteed by borrowers. and other obligations (for either local or later maturity shareholders or CFI: applies to a investments example, currency (for principal third parties, single, specific transactions. recurring off- domestic payments of capital market export contract taker payments transactions) or amortizing debt instruments, for services, under a Power foreign currency syndicated loans. bonds, financial capital goods or Purchase (for cross-border leases, letters of projects. Agreement or transactions). credit, promissory early termination Local currency notes, and bills of payments under a FCGs may be exchange. concession offered in agreement), countries where World Bank- IFC does not supported Letters currently have the of Credit ability to provide local currency financing through the use of swap markets or other means. IFC’s FCG acts as a synthetic borrowing and on-lending for IFC providing the domestic lender with a AAA quality credit coverage for their 86 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf guaranteed loan, and the borrower with term financing in local currency. Other eligibility Debt investments only Project must have Subnationals Sovereigns, sub- Sovereigns, sub- Same country CFI: must benefit Creditworthy criteria (including capital a clear and must be sovereigns, and sovereigns, and creditworthiness ECA country; or quality market transactions defined creditworthy; SOEs must be SOEs must be and eligibility insured FI investors. and swaps that MIGA development SOEs must creditworthy, ADF creditworthy. criteria for ADB customer’s buyer classifies as debt) impact. operate on a countries are loans also apply must be With minimum credit Project must be of commercial basis. eligible for PCGs to guarantees. creditworthy; risk rating ceiling. For strong interest to Project must meet only if they are Projects and technical and those below credit and have the development classified as programs to be managerial ceiling, sovereign express objectives of the countries with supported by capabilities of the guarantee and commitment of host country and low risk of debt guarantees must insured FI additional risk the host benefit the local distress (green be consistent with customer; mitigants needed. government. economy. light countries ADB’s country contractual terms Project must be Projects must be Projects must be based on the strategy. and conditions; technically feasible, technically and technically, World Bank-IMF ADB must have a conditions and financially, and financially viable commercially, Debt direct or indirect economic outlook economically viable and sustainable in environmentally Sustainability participation in a in the foreign over the guarantee the short, and socially Framework traffic project or related buyer’s country. period. medium and sound. light country sector, through a long-term. classification) and loan, equity Project must deemed to have investment or contribute to the Project must be adequate debt technical development of the capable of management assistance. Any host country and meeting the Bank capacity. ADF- ADB financing consistent with the Group PCG is available instrument can declared development environmental, to SOEs in ADF satisfy the objectives of the host social, and countries with participation government. anticorruption low to moderate requirement, Project must meet guidelines. risk of debt provided that MIGA’s E&S standards distress (green 87 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf and policies and For private and yellow light both the MIGA’s integrity due projects, it must countries, instrument and diligence be the direct respectively, the guarantee requirements. beneficiary of an based on the share the same Project must comply obligation from World Bank-IMF development with host country’s the government, a Debt objective and laws and regulations. political Sustainability relate to the same For an SOE to be subdivision of Framework traffic sector in the eligible for NHFO– government, or a light country applicable SOE coverage, the SOE sovereign owned classification). country or must be 1) subject to entity, which can The country project. governmental be guaranteed by creditworthy and Guarantees may supervision and the World Bank. eligibility criteria be issued for controle; 2) provide a For loan applicable to projects and public service or guarantees, the loans will also borrowers/issuers function; and 3) debt must be apply for located in any creditworthy as a extended by guarantees. developing stand-alone entity. commercial Projects and member country entities. World programs to be of ADB. Bank guarantees supported by The ADB are not available guarantees must guaranteed to support be consistent with lender must bilateral debt or AfDB’s country ensure that the debt extended by strategy. borrower of the publicly owned guaranteed loan entities that must comply with operate under ADB’s policies, public law for including those public policy related to E&S purposes (for safeguards; example, DFIs, procurement; and ECAs). the prevention of corruption, and money 88 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf laundering, and financing of terrorist activities. Sovereign Yes, for SOEs and Yes. Counter No. IFC does not Yes. All PCGs to Yes, for sub- Yes. Sovereign only. Not Not guarantee subnationals rated guarantee and provide sovereign entities sovereigns and guarantee or indicated indicated required below minimum credit Indemnity guarantees to require a SOEs, but not for counterindemnity rating threshold. Agreement from entities that sovereign sovereigns is required under Host government the government require need counterindemnity the sovereign approval is required in favor of IBRD backing by the from the guarantee for all MIGA insurance or IDA as issuer of sovereign. government. This window. products the guarantee, However, IFC is not applicable Sovereign The Agreement requires to private sector guarantee may or requires acknowledgment projects may not be reimbursement to by the host required for IBRD or IDA by country private sector the host projects government in the event that the guarantee is called by the guaranteed party Percentage 95% of principal + IBRD/IDA No specific Determined on a 90%–95% of ADB sets the PRI of NHS: 90– Financial covered interest rate; retained guarantees cover percentage limit case-by-case principal +interest guaranteed 95% of principal, Institutions earnings, and so on. only some of the for PCG. PCG is basis. payment. percentage at the regular accrued and Capital On exceptional basis, project risks or structured to lowest level interest and Markets:- can cover up to part of the debt cover a portion of required to capitalized coverage is 99 percent of debt service hence are the guaranteed mobilize upfront non- available for amount. considered instrument’s total financing. honoring of 100 percent “partial.” debt service However, ADB is sovereign of private Percentage payment, subject not prevented by obligation sector determined on a to a maximum its policies from premiums. No transactions case-by-case cumulative enhancing 100% stated project and payout equal to of a lender’s limit. 95 percent 89 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf basis but will not the guarantee exposure to loss CFI: up to 90% of each reach 100%. amount. The under a the FI’s public sector guarantee guaranteed loan customer’s losses transaction; amount may be in certain limits of up expressed as a circumstances. to percentage of $150 million principal and per amortizes in transaction. proportion to the Equity bond or loan. investors: up to 100% of investment value; can indemnify up to $150 million per country. Importers or Exporters: up to 95% of the import/ export contract’s value; limits up to$150 millio n per country. Calls or claims 180 days (6 months) IBRD/IDA will only In the event that 6 months waiting PRI of NHS: mechanism waiting period. pay for any IFC pays a claim period Claims should be Claims payment by undisputed under the submitted MIGA would be based amount on guarantee made according to the on the outstanding receipt of a by the lender, a waiting period 90 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf principal and any demand under direct loan given in the accrued and unpaid the guarantee by between IFC and insurance policy, interest to the lender. the guaranteed the borrower is but no later than On payment of party for breach novated under six months from compensation under by the IFC’s Guarantee the date of the such claim: (i) MIGA is government party and Standby Loan loss. If the claim is subrogated to such of its covered Agreement accepted, the ECA rights or claims related obligations. If (GISLA) with the will pay it within to the guaranteed there is dispute borrower 10 days of investment as the between the notification of guarantee holder may government and claim payment have had against the the guaranteed decision. host country or other party over the CFI: claim can be obligors; and (ii) the amount owed, submitted in guarantee holder IBRD/IDA would respect of a debt transfers and assigns pay such amounts 120 days (four to MIGA, free and only once months) after the clear, all such rights, or applicable dispute risk event that claims resolution caused the loss. processes have However, if the been exhausted, insured buyer has or the amount either been owed otherwise placed into recognized by the receivership or parties or under has commenced provisional proceedings payment under bankruptcy mechanisms if or insolvency specifically laws, the claim provided for in may be submitted the guarantee immediately. To agreement. In the be considered, a event that IBRD or claim must be IDA makes a submitted no 91 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf payment under later than 12 the guarantee to months from the the guaranteed date of the beneficiary, occurrence of the IBRD/IDA will be risk event that legally caused the loss subrogated to the rights of the beneficiary under the guaranteed financing agreements, and the host government will be required to reimburse IBRD/IDA for the payments made under the guarantee Maximum tenor Generally, 15 years, Same maturity No formal limit, Up to twenty (20) 15–20 years Up to 15 years, or Up to 20 years Up to 15 may extend to 20 limits as IBRD but in practice, up years for in exceptional years, non- years. loans (35 years) to 10 years. sovereign cases, longer, cancelable and IDA credits guaranteed even provided policy. (40 years). borrowers. Up to that the tenor is fifteen (15) years justified and in for non‐sovereign line with ADB’s guaranteed risk policies. The borrowers. The guarantee period principal could match the repayment period full term of the of the financing guaranteed debt should match to instrument, or a the requirements portion, such as 92 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf of the project the back-end being financed. maturities. For structures with bullet repayments, the maximum period is limited to 15 years and an average life of 10 years. Experience with Since 2011a Established in the Since 2000 for Since 2001 Since 1998 for Since 1990. Since mid- nonpayment 1990s and has PCG; 2013 for PCGs. 1970s guarantees for expanded and ADF-PCG. sovereign, sub- enhanced since. sovereign, and In 2013, the SOEs World Bank moved away from offering a defined menu of project- based guarantee structures (PRG or PCG) to a customized structure (for example, differentiating project-based guarantees by the nature of the covered risks). 93 Appendix F Comparison of Guarantee Instruments Institution Attribute for MIGA IBRD and IDA IFC AfDB ATI ADB ECAd Private Comparison Insurerf NH(S)FO/ No known claims with No IBRD and IDA Unknown Unknown $21 million during Unknown $20 million during $22.5 million Sovereign Non- its NH products to guarantees were calendar year calendar year during payment/ State date called in FY18 and 2014 to 2018b 2016 to 2018b calendar Obligation FY19 year 2012 to Insurance 2018b Claims experience NH(S)FO/ $5.1 billion (FY19; $3.7 billion (FY19; $2.9 billion (FY19) No information $2.7 billion (CY18) $1.3 billion (CY18; $648 million $2.0 billion Sovereign NHSFO and NHFO- IBRD project- nontrade-related (CY18) (CY18) Nonpayment SOEy) based, policy- credit guarantees Insurance/ based and only) State enclave Obligation guarantees); and Insurance $2 billion (FY19; Outstanding IDA project-based exposure ($, and policy-based millions) guarantees) Sources: Asian Development Bank; African Development Bank; African Trade Insurance Agency; American International Group, Inc.; Export Development Canada; International Finance Corporation; Multilateral Investment Guarantee Agency; World Bank Group Products, May 2018; G20 Argentina 2018. “Introductory Guide to Infrastructure Guarantee Products from Multilateral Development Banks”; Overseas Development Institute, “Guarantees for Development: A Review of Multilateral Development Bank Operations,” December 2014; Asian Development Bank Partial Credit Guarantee Brochure; Asian Development Bank 2018 Management Financial Report; African Trade Insurance Agency Annual Report 2018; Export Development Canada Solutions Guide for Financial Institutions; Export Development Canada Annual Report 2018 and 2019; International Finance Corporation Annual Report 2019 Volume 2 – Financials; World Bank Annual Report 2019: Management's Discussion & Analysis and Financial Statements (Fiscal 2019). Note: (i) Appendix E compares MIGA’s NH product with providers of nonpayment guarantee products. MIGA also supports SOEs directly and indirectly through its Political Risk Insurance product but these are excluded from this table; (ii) Berne Union data; (iii) PCG can be used to support mobilization of private funds for project finance, financial intermediation and policy-based finance; (iv) based on Export Development Canada; (v) MIGA considers the SOE as an entity “controlled by the government” if: (i) the government owns an equity interest exceeding 50 percent in the SOE, or (ii) it is able to direct the affairs and/or control the composition of the board of directors or equivalent body of that entity. (vi) based on American International Group Inc’s published materials. ADB = Asian Development Bank; ADF = African Development Fund; AfDB = African Development Bank; ATI = African Trade Insurance Agency; CRI-SO = Credit Risk Insurance – Sovereign Obligation; CY = calendar year; ECA = export credit agency; FY = fiscal year; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; MLT =; NH = Non-Honoring of Sovereign, Sub-sovereign, and State-Owned Enterprise Financial Obligations; PCG = Partial Credit Guarantee; PPP = public-private partnerships; PRI = political risk insurance; SMEs = small and medium enterprises. 94 Appendix F Comparison of Guarantee Instruments Appendix F. Comparison of Attributes of World Bank Group Guarantee Instruments Table F.1. Attributes of World Bank Group Guarantee Instruments MIGA IBRD IFC IFC/World Bank SNF IFC SLG for PSW IDA IFC IFC Product Attributes PRI NH PBG PCG PRG RSF PCG Countries PRG PCG SCLG Risks covered Comprehensive risk (commercial/credit and + + + + + political) Commercial risks only + + + + + Political risks + + + + + Transfer and convertibility + + + + + + + + Breach of contract + + + + + + + + Expropriation + + + + + + + + Political violence + + + + + + + + Conventional terrorism + a + + + + + NHSFO; NHFO-SOE + + + + + + Non-honoring of arbitration award + + + + + + Wrongful calling of guarantee + + + + + + License cancelations + + + + + + + + Debt service default + + + + + + Sovereign default + + + + + + Beneficiary type Privately owned + + + + + + + + Government owned + b + + + + National +b + + + 95 Appendix F Comparison of Guarantee Instruments MIGA IBRD IFC IFC/World Bank SNF IFC SLG for PSW IDA IFC IFC Product Attributes PRI NH PBG PCG PRG RSF PCG Countries PRG PCG SCLG Subnational +b + + + + Parastatal or state-owned enterprise +b + + + SMEs + + Nonprofit organization + c Underlying assets or obligations Equity + Portfolio equity + Quasi-equity + + + + + + Subordinated loan + + + + + + + Senior loan + + + + + + + + Portfolio loan + Bonds + + + + + + Islamic bonds or kifala + First loss tranche + + + + Mezzanine tranche + + + Delivery of goods and services; engineering, + procurement, and construction Security or credit enhancement + + + + + + Short-term instruments + + + + Trade credit + + + + Portfolio of assets + + Production-sharing contracts + + + + Profit-sharing contracts + + + + Management contracts + + + + Franchising agreements + + + + 96 Appendix F Comparison of Guarantee Instruments MIGA IBRD IFC IFC/World Bank SNF IFC SLG for PSW IDA IFC IFC Product Attributes PRI NH PBG PCG PRG RSF PCG Countries PRG PCG SCLG Licensing agreements + + + + Turnkey contracts + + + + Operating leasing agreements + + + + Subordinated bonds + + + + + Guarantees or securities + + + + + Swaps or hedging Instruments + + + Guaranteed percentage Full + + + Partial + d + d + + + Others Government counterguarantee +e + + + + In-kind contribution f + Local currency + + + + + + Source: World Bank Group Products. Note: IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; IFC = International Finance Corporation; MIGA = Multilateral Investment Guarantee Agency; NH = Non-Honoring of Financial Obligations; PBG =; PCG = partial credit guarantee; PRG = partial risk guarantee; PRI = political risk insurance; PSW = Private Sector Window; RSF = risk-sharing facility; SCLG = single-currency loan guarantee; SLG = Small Loan Guarantee program; SMEs = small and medium enterprises; SNF = subnational finance. a. Covered under war and civil disturbance. b. Provided that the government-owned investor operates on a commercial basis. c. Provided that the specific investment for which the coverage is being sought will be carried out on a commercial basis. d. The portion of loss to be paid by MIGA in the event of a claim shall not exceed 99 percent for loans and 95 percent for all other instruments. e. For projects below the minimum credit rating threshold. f. Tangible or intangible assets that have monetary value, such as machinery, patents, processes, techniques, managerial know-how, trademarks, and marketing channels. 97 Appendix G. Sovereign Defaults and Credit Ratings, Various Countries, 1990–2018 Table G.1. Heavily Sovereign # of Years in # of Years in Default Claims Paid Most Recent # of Claims Country Lending Indebted Poor Rating Agency Default 1990- Default 1990- Default with Amount (US$ Amount (US$ Credit Rating 2010-2018 Countries 2018 2018 MDBs Millions) Millions) Bulgaria IBRD BBB- S&P Yes 22 19 40,786.76 - - Colombia IBRD BBB- S&P Yes 14 14 1,444.37 - - Croatia IBRD BBB- S&P Yes 5 - 8,549.00 - - Cyprus - BBB- S&P Yes 1 - 1,698.24 - - Hungary - BBB- Fitch Yes 1 1 9.20 - - India IBRD BBB- S&P Yes 2 2 4.10 2 7.30 Kazakhstan IBRD BBB- S&P Yes 6 5 950.13 - - Montserrat - BBB- S&P No - - - - - Morocco IBRD BBB- S&P Yes 20 19 16,176.47 - - Romania IBRD BBB- S&P Yes 20 20 1,418.32 1 0.30 Russia IBRD BBB- S&P Yes 22 18 580,559.25 1 0.20 San Marino - BBB- Fitch No - - - - - Uruguay IBRD BBB- Fitch Yes 3 - 8,195.00 - - Andorra - BBB S&P No - - - 1 1.50 Côte d'Ivoire* IDA HIPC BBB S&P Yes 27 22 143,035.28 - - Indonesia IBRD BBB Fitch Yes 16 11 33,217.95 - - Italy - BBB S&P No - - - - - Panama IBRD BBB S&P Yes 18 18 35,876.78 - - Philippines IBRD BBB S&P Yes 15 8 36,743.26 - - Portugal - BBB Fitch Yes 1 1 52,712.00 - - Aruba - BBB+ S&P No - - - - - Macao SAR, China - BBB+ S&P No - - - - - Mauritius IBRD BBB+ Moody's converted Yes 13 13 68.52 3 11.30 Mexico IBRD BBB+ S&P Yes 1 - 54,300.00 - - Peru IBRD BBB+ Fitch Yes 27 23 115,742.17 - - Thailand IBRD BBB+ S&P Yes 7 7 35.67 1 8.20 Trinidad and Tobago IBRD BBB+ S&P Yes 1 - 110.00 - - Turks and Caicos - BBB+ S&P No - - - - - Botswana IBRD A- S&P Yes 26 26 607.47 - - Malaysia IBRD A- S&P No - - - - - Malta - A- S&P No - - - - - Poland IBRD A- S&P Yes 13 8 114,826.70 - - Saudi Arabia - A- S&P No - - - - - Spain - A- S&P No - - - - - Iceland - A Fitch No - - - - - Latvia - A S&P No - - - 2 8.90 Lithuania - A S&P No - - - - - Bermuda - A+ S&P No - - - 1 0.50 Chile IBRD A+ S&P Yes 1 - 6,494.00 - - China IBRD A+ S&P No - - - 1 2.90 Ireland - A+ S&P Yes 1 1 88,290.00 - - Japan - A+ S&P No - - - - - Slovakia - A+ S&P No - - - - - Slovenia - A+ S&P Yes 5 - 3,619.33 - - Cayman Islands - AA- Moody's converted No - - - - - Czech Republic - AA- S&P No - - - - - Estonia - AA- Fitch No - - - 3 47.80 Guernsey - AA- S&P No - - - - - Israel - AA- S&P No - - - - - Jersey - AA- S&P No - - - - - Qatar - AA- S&P No - - - - - Taiwan - AA- S&P No - - - - - Belgium - AA S&P No - - - - - France - AA S&P No - - - - - Korea - AA S&P No - - - - - Kuwait - AA S&P No - - - - - New Zealand - AA S&P No - - - - - United Arab Emirates - AA Moody's converted No - - - - - United Kingdom - AA S&P No - - - - - Austria - AA+ S&P No - - - - - Finland - AA+ S&P No - - - - - Hong Kong - AA+ S&P No - - - - - United States of America - AA+ S&P No - - - 1 1.60 Australia - AAA S&P No - - - 2 5.30 Canada - AAA S&P No - - - - - Denmark - AAA S&P No - - - - - Germany - AAA S&P No - - - - - Liechtenstein - AAA S&P No - - - - - Luxembourg - AAA S&P No - - - - - Netherlands - AAA Fitch No - - - 1 3.30 Norway - AAA S&P No - - - 1 0.10 Singapore - AAA S&P No - - - - - Sweden - AAA S&P No - - - 3 2.70 Switzerland - AAA S&P No - - - - - 98 Appendix G Sovereign Defaults and Credit Ratings Heavily Sovereign # of Years in # of Years in Default Claims Paid Most Recent # of Claims Country Lending Indebted Poor Rating Agency Default 1990- Default 1990- Default with Amount (US$ Amount (US$ Credit Rating 2010-2018 Countries 2018 2018 MDBs Millions) Millions) Mozambique* IDA HIPC RD Fitch Yes 27 27 42,997.12 - - Barbados - SD S&P No - - - - - Iran (Islamic Republic) IBRD SD S&P Yes 14 10 22,958.08 1 1.40 Gabon IBRD NR S&P Yes 26 22 17,176.60 - - Grenada Blend NR S&P No - - - - - Kyrgyzstan IDA NR S&P Yes 24 23 1,715.77 - - Mali* IDA HIPC NR S&P Yes 27 26 10,444.28 - - Seychelles IBRD NR S&P Yes 25 23 2,466.60 1 2.80 Venezuela, RB IBRD NR S&P Yes 27 26 50,564.26 5 81.60 Turkmenistan IBRD - Moody's converted Yes 23 23 980.89 - - Sources: Bank of Canada and Bank of England Credit Rating Assessment Group for sovereign default data; David Beers and Jamshid Mavalwalla. Bank of Canada Staff Working Paper 2018–30. “The BoC-BoE Sovereign Default Database Revisited: What’s New in 2018?”; World Bank Databases for country and HIPC classifications; and Standard & Poor’s, Moody’s and Fitch for the credit ratings. Note: The Bank of Canada-Bank of England CRAG data set defines sovereign default as having occurred when debt service is not paid on the due date or within a specified grace period, when payments are not made within the time frame specified under a guarantee. Or, in the absence of an outright payment default, sovereign default may still occur where creditors incur material economic losses on the sovereign debt they hold because of agreements among governments and creditors that reduce interest rates and/or extend maturities on outstanding debt, government exchange offers to creditors where existing debt is swapped for new debt on less-economic terms, government redenomination of foreign currency debt into new local currency obligations on less-economic terms, retrospective taxes targeting sovereign debt service payments, and so on. * Indicates classification as fragile and conflict-affected situation. Countries with MIGA NH projects are highlighted in red font. The HIPC countries Afghanistan, Central African Republic, Chad, Comoros, Eritrea, The Gambia, Guinea, Guinea Bissau, Guyana, Haiti, Liberia, Madagascar, Malawi, Mauritania, Niger, São Tomé and Príncipe, Sierra Leone, Somali, Sudan, and Togo are not rated by any of the rating agencies and are not included in the list. HIPC = heavily indebted poor countries; IBRD = International Bank for Reconstruction and Development; IDA = International Development Association; MDB = multilateral development bank; NR =Not Rated; RD =Restricted Default; S&P = Standard & Poor; SD = Selective Default. 99 The World Bank 1818 H Street NW Washington, DC 20433