FINANCE FINANCE EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT Credit Worthy: ESG Factors and Sovereign Credit Ratings Ekaterina M. Gratcheva, Bryan Gurhy, Andrius Skarnulis, Fiona E. Stewart, and Dieter Wang © 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 Some rights reserved. This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, 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 boundaries, 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 endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO), http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Ekaterina M. Gratcheva, Bryan Gurhy, Andrius Skarnulis, Fiona E. Stewart, and Dieter Wang. 2021. “Credit Worthy: ESG Factors and Sovereign Credit Ratings” EFI Insight-Finance. Washington, DC: World Bank. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third- party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. Cover photo: Giordano Aita/Shutterstock.com This paper forms part of a series of publications under the (FCI) Global Practice (GP) in collaboration with World Bank Global Program on Sustainability (GPS). The series is a Treasury (TRE), Development Economics Vice Presidency knowledge product of GPS Pillar 3 with the objective to promote (DEC), and other GPs. Focusing on ESG issues in sovereign the use of high-quality data and analysis of sustainability to investing the series disseminates practical, evidence-based better inform decisions made by governments, the private recommendations for market participants, including institutional sector, and financial institutions. GPS Pillar 3 is led by the investors, sovereign issuers, credit rating agencies, ESG data World Bank’s Finance, Competitiveness, and Innovation and service providers, among others. “A New Dawn - Rethinking Sovereign “Demystifying Sovereign ESG” ESG” proposes improvements to the focuses on comparing sovereign ESG sovereign ESG framework and builds on methodologies of leading sovereign findings and recommendations discussed ESG providers and presents structural in other papers in the series. challenges with the current sovereign ESG framework. “Riding the Wave: Navigating the ESG “Paving the Path: Lessons from Chile’s Landscape for Sovereign Debt Managers” Experiences as a Sovereign Issuer for provides a thorough discussion of Sustainable Finance Action” provides sovereign ESG from a debt management a focused study of Chile’s ESG focused office perspective. issuances to date and relevant lessons. “Spatial Finance: Challenges and “Natural Capital and Sovereign Bonds” Opportunities in a Changing World” builds the case that countries’ level of in partnership with WWF discussed development dominates ESG-related challenges with the E data, including at metrics. “1% Growth in Natural Capital: the sovereign level, and explores the use Why It Matters for Sovereign Bonds” of satellite data to address the quality and quantifies materiality of natural capital availability of E data. and its impact on sovereign bonds by adjusting for ingrained income bias. The chapter “Natural Allies: Wealth “Natural Capital and Sovereign Bonds” and Sovereign ESG” from the book introduces the concept of ingrained “The Changing Wealth of Nations income bias and presents evidence that 2021: Managing Assets for the Future” sovereign bond yields reflect a country’s focuses on challenges in ESG data, and various types of natural capital. discusses solutions with the application of the World Bank Wealth data. Contents Abbreviations 4 Acknowledgments 5 Executive Summary 7 A Changing Financial Sector Ecosphere 8 Current Sovereign ESG Methodologies 8 Sovereign Credit Ratings and Sovereign ESG Scores 10 A More Holistic View of a Country’s Long-Term Sustainability 11 Key Conclusions 12 1. A Changing Financial Sector Ecosphere 15 2. ESG Factors and Sovereign Credit Ratings 23 3. Sovereign Credit Ratings and Sovereign ESG Scores 29 4. Sovereign Creditworthiness and a Country’s Wealth 35 The Low-Carbon Transition and Stranded Assets 5.  in Sovereign Credit Ratings 43 6. Conclusions 51 References 53 Additional Material 54 Appendix A 56 Extended History of Credit Rating Agencies 56 Integration of ESG as a Separate Component into the Sovereign Credit Methodology 57 Wealth Data 58 Appendix B: Additional Tables 59 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 3 Abbreviations CRA credit rating agency DMO debt management office DOJ U.S. Department of Justice EMDE emerging market and developing economy ESG environmental, social, and governance ESMA European Securities and Markets Authority EU European Union PRI Principles for Responsible Investment SDGs Sustainable Development Goals V.E Vigeo Eiris, a company acquired by Moody’s Corporation in 2019 WAVES Wealth Accounting and the Valuation of Ecosystem Services 4 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Acknowledgments This publication was prepared by a team consisting of Ekaterina M. Gratcheva from the Climate Investment Funds and Bryan Gurhy, Andrius Skarnulis, Fiona E. Stewart, and Dieter Wang under the supervision of Anderson Silva, all from the Finance, Competitiveness and Innovation (FCI) Global Practice of the World Bank Group (WBG). The authors would like to thank those who provided the comments received during the formal peer review process, including Swee Ee Ang, (Senior Financial Sector Specialist, FCI); Ana Maria Aviles (Senior Financial Sector Economist, FCI); Heike Reichelt (Manager, World Bank treasury), the WBG’s Credit risk office team represented by Casey Elizabeth Reckman (Senior Economist), Girum Dagnachew Abate (Economist) and Smriti Seth (Economist), Fergus McCormack (Director of Sovereign Research at the Emerging Markets Investors Alliance), Coziana Ciurea (Product Manager, V.E- part of Moody’s ESG solutions) and Ju Shik Ha (Senior Economist, FCI). Also, a special thank you to Jean Pesme, Global Director of the Finance, Competitiveness and Innovation Global Practice of the World Bank Group and Elizabeth Price (Senior External Affairs Officer, FCI communications) for their stellar support and guidance. The authors would like to thank the credit rating agency community, ESG providers, asset management industry, and wider financial sector for the continued discussions on this quickly evolving topic. The views expressed herein are solely the authors’ and should not be attributed to the WBG. The report was edited by Kathryn McCarthy and Marcy Gessel, Publications Professionals LLC. Hanna Chang led the creative design and formatting of the publication. We thank them all. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 5 Caption: School girls in Afghanistan | Source: pixabay.com >>> Executive Summary The increasing role of the financial sector in the move toward a more sustainable economic model continues apace. The COVID-19 “shock” shone a light on the need for all society to correct course, and the financial sector is responding. The pace of environmental, social, and governance (ESG) integration into investment decisions,1 which has become the prevalent form of sustainable finance, continues to accelerate.2 These developments reflect changing societal perspectives that challenge the traditionally ingrained investment approaches that have evolved over many decades. Against this backdrop, various financial sector stakeholders continue to evaluate how their role, products, and tools should adapt to this evolving landscape. This report forms part of a broader analytical report series under the auspices of the Global Program on Sustainability,3 which looks at pertinent ESG issues and how they apply to the sovereign fixed-income asset class. Our analysis to date highlights the “unique” nature of the sovereign debt asset class from an ESG perspective, largely driven by the complex nature of a sovereign nation, in particular compared to a corporate entity. Our publications are targeted at various financial sector stakeholders, including sovereign and sub-sovereign issuers, investors, credit rating agencies (CRAs), ESG providers, regulators, standard setters, and policy makers, as well as sustainable finance advocates. The series also aims to illuminate issues from various perspectives and to provide evidence to ensure that the market practices that are becoming embedded in the financial system are equitable and transparent to all. The increased focus on sustainable finance has resulted in contrasting views and sometimes heated debate4 between the key protagonists in the financial sector on how sustainability should be assessed and incorporated in financial decision-making. Investors are increasingly focused on a broader definition of sustainability from a longer-term perspective. For example, a recent survey from J.P. Morgan (JPM survey in chapter 2, Gratcheva et al. 2021) highlights the belief held by many investors that improving ESG fundamentals leads to lower sovereign credit risk. However, this view challenges traditional perceptions in the market regarding the key factors determining sovereign credit risk, such as debt and fiscal risks. Indeed, empirical evidence showing a causation between ESG factors and sovereign credit risk is mixed. The survey also indicates that many in the industry expect these two approaches to credit risk to converge. However, this expectation may not be well grounded, because an issuer’s creditworthiness and 1 ESG integration is the practice of incorporating ESG-related information into investment decisions to help enhance risk- adjusted returns, regardless of whether a strategy has a sustainable mandate. 2 See https://www.bloomberg.com/news/articles/2021-08-18/-35-trillion-in-sustainability-funds-does-it-do-any-good. 3 See the Global Program on Sustainability, https://www.worldbank.org/en/programs/global-program-on-sustainability. 4 See in particular Tariq Fancy, “Tariq Fancy on the Failure of Green Investing and the Need for State Action,” The Economist, November 4, 2021, https://www.economist.com/by-invitation/2021/11/04/tariq-fancy-on-the-failure-of-green-investing-and- the-need-for-state-action, and Robert G. Eccles, “A Critique of Tariq Fancy’s Critique of ESG Investing: An Interview with Clara Miller,” Forbes, October 1,2021, https://www.forbes.com/sites/bobeccles/2021/10/01/a-critique-of-tariq-fancys-critique- of-esg-investing-an-interview-with-clara-miller/?sh=60000c8c390d, among others. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 7 its ESG standing, or reputation may not be directly linked as related to so-called stranded assets like fossil fuel an issuer’s default risk could be independent of the entity’s resources—are manifested in sovereign credit rating favorable ESG characteristics (S&P 2021). assessments. This analysis is complemented by a network analysis of the publicly available sovereign CRA This paper focuses on sovereign credit ratings and methodologies, as well as information garnered from the empirically assesses how broad sovereign ESG factors— team’s discussions with the CRAs.5 as well as the ESG factors specific to a country’s national wealth and management of risks and opportunities A C HA N GIN G F I N A N C I A L S EC TO R ECO S PHERE Having evolved over nearly two centuries, CRAs have differences however, that affect how CRAs and ESG providers become a foundational part of the global financial have been responding to the demand for more sustainability- architecture.6 CRAs evolved over time to fulfill a specific based products and services. Specifically, compared with purpose: to assess the capacity and willingness of an issuer CRA ratings, which are originated and paid for by issuers, to meet its bond payments on time and in full. In recent ESG scores/ratings/rankings are “unsolicited” as ESG years, with the increasing focus on ESG investing in various assessments are typically not requested by issuers and are asset classes, investors have sought a different type of paid for by investors. The ESG provider industry is nascent information—one that has a wider conceptual scope and is and is currently not regulated, while CRAs are regulated complementary to financial analysis. This development has and have a well-established mandate, methodologies, and resulted in the advent of ESG providers, who aim to give an terminologies. The age of the ESG industry, combined with extra-financial assessment of sustainability (not focused only access to new technologies and (big) data in contrast to the on debt and fiscal metrics) that is now perceived, by some, scarce information during the CRA industry’s formation, has to be affecting sovereign creditworthiness. There are notable led to a marketplace filled with a variety of actors. C UR R E N T SOV E R E I G N ES G METHO D O LO G IES CRAs argue that consideration of ESG credit factors highlights the fact that E and S factors play a limited role in (that is, factors that have a direct impact on an issuer’s the quantitative rating methodology of most CRAs. In some creditworthiness) should not be conflated with investment cases, CRAs may incorporate E and S factors in the qualitative strategies that target some positive ESG outcomes overlay process, which is also part of the process of deriving (nonfinancial impact as well as a financial return). The a final credit rating for most CRAs, but this process could be CRA industry underscores that factors that influence a deemed as quite subjective. sovereign’s creditworthiness over a particular time horizon Our analysis provides a mixed picture of how ESG factors (typically less than 10 years) and are visible (that is, their affect sovereign credit ratings: we show empirically that likelihood or effect can be assessed) are the only ones that there is scope to improve the current CRA approach by should be included in a sovereign credit rating assessment. better reflecting ESG factors in sovereign credit rating Credit factors with historical precedence, such as fiscal/debt methodologies. Materiality7 is a key concept in the discussion dynamics and governance factors, are explicitly included in around including credit-relevant ESG factors. Although on the credit assessments. Environmental (E) and social (S) factors, conceptual basis many ESG factors could affect a country’s however, are often not included, given the complexities of sustainability profile, the existing methods and/or data are evaluating them at a sovereign level. Further, the expected inadequate to properly assess their financial materiality over time horizon and greater uncertainty of their associated risks a particular horizon, and therefore these factors currently do going forward often make them harder to assess. Figure ES.1 not affect the “headline” credit assessment. However, figure provides a network analysis of the key credit factors currently ES.2 highlights the bind that many CRAs face—which is, in included in sovereign credit rating methodologies. The figure 5 We analyzed the methodologies used at the following companies: Fitch Ratings, Moody’s Investor Service, S&P Global Ratings, Kroll Bond Rating Agency (KBRA), DBRS Morningstar, Scope Ratings GmbH, and HR Ratings. 6 Credit ratings influence capital allocation, investment decisions, regulations, and even monetary policy collateral frameworks, and many investors see them as a precondition to inform the investment process. 7 Materiality is an accounting principle that defines which information is useful to an investor, given specific investment objectives. 8 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS FIGURE ES1 - Analysis of main credit factors in sovereign credit rating methodologies of seven key CRAs The network graphic provides a simple but insightful overview of how the seven CRAs differ in the way they include credit factors into their sovereign credit rating assessments. The figure shows the 100 most-mentioned terms in the CRAs’ methodology documents. The size of each term reflects how often it occurs, see Figure 2.2. Terms are associated with either environmental (3 terms), social (7), governance (24), economic (18) or financial (48) themes. The seven CRAs are located between the themes based on which terms occured in their methodology documents. The graphic highlights that (a) sovereign credit rating assessments are multifaceted in nature and (b) E and S factors currently play a limited role in a sovereign credit assessments. Source: World Bank staff illustration. Data from sovereign credit rating methodology by Fitch Ratings, Moody’s Investor Service, S&P Global Ratings, Kroll Bond Rating Agency KBRA, DBRS Morningstar, Scope Ratings GmbH, and HR Ratings Notes: - CRA = credit rating agency; E = environmental; Econ. = economic; Fin. = financial; G = governance; S = social. The three orange circles are used for the “Big Three” rating agencies (S&P Global Ratings, Moody’s Investor Service, and Fitch Group - Ratings) and dark red circles are used for the other four rating agencies (Kroll Bond Rating Agency KBRA, DBRS Morningstar, Scope Ratings GmbH, and HR Ratings) - The green circles are the 100 most-mentioned terms and their sizes reflect how often they have been mentioned (see Figure 2.2). - The five large bubbles depict the five themes each term can be associated with. - The term “climate change” is presented separately because it was the 129th most-mentioned term and therefore did not occur among the top 100. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 9 FIGURE ES2 - The “Tragedy of the Horizon” of CRAs: Even though short-term factors are dependent on long-term issues, CRAs almost exclusively focus on the former. Striking the right balance will become paramount for CRAs CRA assessment horizon Materiality Forward-looking horizon GDP growth Human capital Inflation Trade balance Forest capital Current account Fossil fuels Protected areas international investment position Mineral wealth Short-term growth depends on long-term growth potential Source: World Bank staff illustration. Note: CRA = credit rating agency; GDP = gross domestic product. many ways, a “tragedy of the horizon.”8 Many factors related rating methodologies. Increasing the transparency of factors to a country’s long-term sustainability, such as mineral that are and are not included, as well as their weightings, wealth, fossil fuels, and forest capital, could be material for a would benefit the whole financial system. Although most ESG sovereign credit assessment but are currently not included. credit factors may not be seen as material now, the dynamic We argue that, because the ESG “lens” is now increasingly a nature of ESG credit factors could make them more important key part of the investment process, CRAs should also adapt in the future. Currently, only one CRA includes ESG factors their methodologies to explicitly include ESG factors, which as a separate explicit component in its sovereign credit rating predominantly reflect long-term issues, in their sovereign credit methodology (see appendix A). SOVERE IG N C R E DI T R AT I N G S AND S OVEREIG N ES G S CO RES By design sovereign credit ratings and sovereign ESG factors that come from different disciplines and are measured scores9 are fundamentally different measures—one in different units into a consolidated score. focused on an issuer’s creditworthiness and the other on We find divergences across country income groups an issuer’s sustainability. Comparing sovereign ESG scores when we investigate the relationship between sovereign and sovereign credit ratings is conceptually challenging, credit ratings and sovereign ESG scores. For high- although it is practiced by the industry (Bank of America 2021; income countries, ESG factors and credit ratings are highly Goldman Sachs 2020). First, these products were designed correlated, which implies that one could question the to measure different forms of materiality: creditworthiness additionality of sovereign ESG scores to current sovereign (or financial) versus some mix of financial and ESG (or credit ratings. For low-income countries, however, we see nonfinancial) materiality (Gratcheva et al. 2021). Second, no clear relationship suggesting that for these countries there is a fundamental problem with aggregating E, S, and G 8 Speech by Mark Carney, governor of the Bank of England and chairman of the Financial Stability Board, “Breaking the Tragedy of the Horizon: Climate Change and Financial Stability,” at Lloyd’s of London, London, September 29, 2015, https://www.bis.org/review/r151009a.pdf. 9 Throughout the paper we use “scores” to refer to sovereign ESG assessments by ESG providers to differentiate from “ratings” of sovereign credit rating agencies. For our analysis in this paper, we derive a single sovereign ESG score (per sovereign) across six ESG providers (including MSCI, Robeco, V.E, Sustainalytics, RepRisk, FTSE Russel) using principal components analysis, as these sovereign ESG scores are 90 percent correlated because of the ingrained income bias (Gratcheva, Emery and Wang 2021). Thus, “sovereign ESG scores” used in this paper represents a useful proxy for the sovereign ESG sector as a whole. 10 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS other criteria, such as macroeconomic, fiscal or debt factors, creditworthiness scale. Introducing a more granular rating generally dictate sovereign creditworthiness. This finding scale would allow ESG credit factors, among other credit is particularly relevant for lower-income countries, which factors, to be more visible to investors. are often assigned the same credit rating on the sovereign A MORE HOLISTIC VIEW OF A COUNTRY’S LONG-TERM SUSTAINABILITY We find that aspects of a country’s wealth—human and over 1995–2018). We deduce that if a credit assessment were produced capital10—are generally reflected in sovereign conducted over a longer time horizon, natural capital would credit rating assessments, but the portion of a country’s be more relevant for a thorough assessment of sovereign wealth that focuses on natural capital is generally not. creditworthiness. Although our finding may not be surprising for higher-income Finally, we compare sovereign credit ratings to a measure countries, the irrelevance of natural capital for lower-income of sovereign stranded-assets exposure and the country’s countries is unexpected, given their stronger reliance on institutional and economic resilience to a low-carbon natural resources (World Bank 2021). For this analysis we transition.12 This comparison enables us to explore use the World Bank wealth data set,11 which is uniquely suited whether risks and opportunities related to stranded to inform sovereign credit rating assessments because it (a) assets are captured in current sovereign credit rating estimates economic value of underlying factors and expresses assessments. First, we find that sovereign credit ratings do them in the same monetary unit, (b) has a forward-looking not necessarily reflect sovereign exposure to a low-carbon perspective, and (c) has a long history of curated data that are transition. Second, we find that institutional and economic comparable across 146 countries and over time (annual data FIGURE ES3 - Stranded assets and sovereign credit ratings Most relevant to sovereigns that are highly dependent on fossil fuels. Stranded-assets risk: Can create credit risks as well as opportunities, depending on economic diversification efforts and overall policy framework. No correlation between the level of a country’s exposure to a low carbon transition Exposure / Resilience to (index provided by Peszko et al (2020)) and its current sovereign credit. a low carbon transition versus High correlation between resilience to a low carbon transition and sovereign credit Sovereign credit ratings ratings, but lower for lower-income countries. Overall preparedness for a low- carbon transition is correlated with credit ratings in the high-income group only. Sovereign credit ratings may not fully capture stranded-asset risk implications, especially for lower-income sovereigns. In the current setting, other shorter-term factors, such as political and social instability, can have more impact on credit ratings. Source: World Bank staff illustration. 10 Human capital is computed as the present value of future earnings for the working population over their lifetimes. Produced capital includes the value of machinery, buildings, equipment, and residential and nonresidential urban land. See appendix A, Wealth Data section. 11 The concept of a country’s wealth is different from gross domestic product. A country’s wealth includes produced capital (buildings, machinery, and infrastructure); natural capital (such as agricultural land, forests, protected areas, minerals, and oil, coal, and gas reserves), human capital (broken down by gender and types of employment), and net foreign assets. Wealth accounting provides an estimate of the total wealth of nations by aggregating values of these different components of wealth. A change in wealth is an indicator that is used to assess a country’s potential to grow in the future. A fall in wealth indicates that a country is depleting its assets and may not be able to sustain its future gross domestic product growth. 12 We use sovereign exposure and resilience scores calculated by Peszko et al. (2020) for more than 90 countries. The exposure score consists of four factors that measure how heavily a country is exposed to low-carbon transition risks. The resilience score is a combination of 11 factors that measure a country’s ability and flexibility in adapting to structural changes caused by the low-carbon transition. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 11 resilience scores display a high and statistically significant credit assessments. Moreover, better assessing stranded- correlation with sovereign credit ratings. The relationship, asset risks and opportunities for lower-income countries in however, is notably lower for lower-income countries. These credit rating assessments would allow investors to better findings suggest that a country’s preparedness for a low- differentiate between EM sovereign credit risks (figure ES.3). carbon transition needs to be better reflected in their sovereign K EY C ON C LU S I O N S This paper contributes to the ongoing debate around the ESG landscape and its evolution at a time when the financial sector’s deeply embedded practices and methodologies, are adapting to a changing societal and political context. There is an inherent tension between the CRA industry Notwithstanding, the assessment of materiality of and the calls of sustainability advocates who are ESG factors for sovereign credit risk is evolving. insisting for ESG factors to be further reflected into Transparency around methodology is crucial and many sovereign credit rating assessments. The CRA mandate CRAs have already introduced additional insights into is clearly defined, and it is deeply rooted within the financial their credit assessments to clarify how ESG factors are sectors architecture: credit factors which are material included in the assessment. Additionally, many CRAs to a sovereign’s creditworthiness are included in credit have launched ESG-related products that aim to give assessments. However, the time horizon of events such as investors insight into broader ESG factors of an issuer. climate change pose two fundamental challenges to the calls This increased transparency is welcome, but there for the greater integration of ESG factors into the investment is room for improvement. Still, the current response process. First, limitations around modeling and data are of CRAs to the evolving market demand has not led significant and second, the disconnect between the current to market consensus around how sovereign ESG investment horizon ingrained in the financial industry and the factors affect sovereign creditworthiness and overall horizon that many ESG factors are expected to be material sustainability. from a credit worthiness perspective significantly curtails the possibility of integrating these factors into sovereign credit assessments. As a result, unless policy makers and regulators explicitly mandate CRAs to focus on a longer- term horizon, driven by changing societal perspectives on what constitutes investment “return”, it’s likely that the CRA focus will remain on the current shorter assessment horizon. 12 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS We propose five main policy takeaways to contribute to the current dialogue on this issue: 1 The CRA and ESG provider mandates need to become clearer and distinguishable. Both industries have an important role to play. CRAs are focused on their medium-term credit assessment from the credit and financial risk perspective. ESG providers need to focus on the longer-term sustainability measurement— and in particular to respond to the growing market demand to be able to assess the ESG impacts of investments (that is, ESG as an output, versus ESG as an input) (Boitreaud et al. 2020; Gratcheva et al. 2021). 2 Sovereign CRA methodologies could be strengthened by introducing an ESG pillar which would provide more transparency on the role that ESG factors play in sovereign credit assessments: Investors are increasingly focused on ESG issues. One view is that the ESG framework has been around for many decades in various other guises and the current framework is a “re-marketing” of an already established tenet of sovereign-debt analysis. We argue that the shift of the financial sector toward explicitly including ESG factors in the investment process requires that investment methodologies and the financial sector architecture shift as well. To this end, CRAs could consider an explicit ESG pillar in their methodology. Although E and S credit factors may not be seen as relevant or material for a sovereign credit rating assessment now, this status is likely to change in the future. Moreover, it’s clear that investors require more transparency on ESG issues. 3 Current credit rating scales could be made more granular to better reflect how ESG factors impact a credit assessment: There is room for sovereign CRA rating scales to become more detailed so they could capture differences related to credit-material ESG issues (among other credit factors), whether they be positive or negative. Although many ESG factors could be financially material, its impact may not be material enough to be captured on current ratings scales. Providing more granularity on the rating scale would allow investors to make more informed decisions and better distinguish between countries that are at the same credit-rating level but that may be performing better on E and S issues. While some CRAs have introduced additional guidance (both in their sovereign credit assessments and via additional products) as to the materiality of ESG factors to their sovereign credit decision, it would be significantly less cumbersome to include such additional ESG information within the existing sovereign credit rating framework. 4 World Bank research shows that wealth accounting data related to a country’s natural capital and information on a sovereign’s stranded-asset risks and opportunities represents additional ESG relevant information that could be better reflected in sovereign credit assessments. The wealth accounting approach is inherently forward-looking and captures determinants for long-term sustainable growth. Additionally transition risks and opportunities related to stranded assets are often not clearly taken into account in current sovereign CRA assessments- and CRAs could improve their analysis of these potentially relevant credit factors, going forward. 5 Finally, sovereign debt managers and policy makers have an important role to play in the changing financial sector ecosphere as economies transition to more sustainable economic models (Boitreaud et al. 2020). The debt management office (DMO) of a sovereign can play a key role in engaging with investors on the sovereign’s ESG credentials/climate performance. The provision of such information to market participants aids transparency and will be valued by market participants. This level of engagement is particularly important going forward as investors are increasingly looking to issuer-specific sustainability information beyond limiting their scrutiny to labeled sovereign issues as has been the early practice since the emergence of green and other thematic sovereign bonds. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 13 1. Caption: Shipping containers in Singapore | Source: Unsplash.com 14 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS >>> A Changing Financial Sector Ecosphere The COVID-19 pandemic has accelerated the financial sector’s embrace of sustainable finance as a key structural tenet of the investment world13 (figure 1.1). This increased acceptance is epitomized by the commitments of 128 global asset managers with US$43 trillion in assets under management who support the goal of net zero greenhouse gas emissions by 2050 or sooner,14 the extraordinary inflows of capital into environmental, social, and governance (ESG-)focused investment funds in the past year,15 and the seismic increase in thematic bond issuance by various types of issuers since the beginning of 2020.16 Some issuers, such as the World Bank, have been engaging investors to consider the issuer’s activities through a sustainability lens and measure against the progress toward the Sustainable Development Goals (SDGs) rather than limiting the focus to issue-specific considerations.17 Recent publications from the World Bank, under the auspices of the Global Program on Sustainability, document how the different parts of the financial system are adapting to these developments, the various challenges that are crystalizing, and their implication for emerging market and developing economies (EMDEs) (Boitreaud et al. 2020; Gratcheva, Emery, and Wang 2021; Gratcheva et al. 2021). Credit rating agencies (CRAs) have evolved over the past two centuries to become an integral part of the financial sector ecosphere and are seen as being a critical enabler for the transformation of the financial system’s architecture toward greater sustainability. As presented in detail in appendix A, CRAs have a long history, and their agility and ability to adapt to a changing world have been a key reason for their continued relevance in the global financial system architecture. The CRA industry is currently dominated by three large players—Moody’s Investor Service, S&P Global Ratings, and Fitch Ratings. The European Securities and Markets Authority (ESMA) estimates that, on a cumulative basis, these three CRAs account for 92.1 percent of the European Union (EU) market (ESMA 2019a). 13 For example see recent COP26 announcements: https://ukcop26.org/cop26-goals/finance/ and https://www.unepfi.org/news/ industries/banking/net-zero-banking-alliance-at-cop-26/. 14 This number represents almost half of all global assets under management. Please see https://www.netzeroassetmanagers. org/ for more details. 15 As an example, the following article tells of a jump in fourth-quarter demand: Simon Jessup and Elizabeth Howcroft, “Sustainable Fund Assets Hit Record $1.7 trln in 2020: Morningstar,” Reuters, January 28, 2021, https://www.reuters.com/ article/us-global-funds-sustainable/sustainable-fund-assets-hit-record-1-7-trln-in-2020-morningstar-idUSKBN29X2NM. 16 For a 2020 report, see Liam Jones, “Record $700bn in Green, Social & Sustainability (GSS) Issuance in 2020: Global State of the Market Report,” Climate Bonds Initiative, April 23, 2021, https://www.climatebonds.net/2021/04/record-700bn-green- social-sustainability-gss-issuance-2020-global-state-market-report. 17 See the 2020 World Bank (International Bank for Reconstruction and Development) Sustainable Development Bonds & Green Bonds Impact Report: https://treasury.worldbank.org/en/about/unit/treasury/impact/impact-report. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 15 FIGURE 1.1 - The financial sector continues to adapt as sustainable finance becomes mainstream Financial Intermediation Sovereign End investors issuers Credit rating ESG rating ESG indexes Institutional Asset managers agencies providers providers investors All issuers that Firms that rate Firms that Firms taht Firms that Entities with Owners receive an ESG issuers provide raw construct ESG construct and fiduciary who bear ESG rating ESG data and indices market ESG responsabilities the ultimate composite funds, ETFs, to maanage reward and ESG scores etc. assets risks Rules & requirements Ethical standard setters (such as regulators, supervisors) (such as WBG, UN, OECD, CBI, ICMA) Source: Boitreaud et al. 2020. Note: CBI = Climate Bonds Initiative; ESG = environmental, social, and governance; ETF = exchange-trade fund; ICMA = International Capital Market Association; OECD = Organisation for Economic Co-operation and Development; UN = United Nations; WBG = World Bank Group. The 2008 global financial crisis was a seminal moment for the CRA industry.18 Post crisis, the CRA industry came under scrutiny regarding the quality of its credit rating assessments, as well as its rating actions during the crisis, which were viewed as amplifying financial shocks. This view was exacerbated by the perceived role by the industry in the euro-area sovereign debt crisis (2010–12).19 Against this backdrop, the investment industry looked for better data to inform investment decisions, and a new industry of specialized ESG providers burgeoned (Gratcheva, Emery, and Wang 2021) (see appendix A). The subsequent rise of ESG providers as a prominent information source for the investment community illustrated investors’ appetite for higher-quality information to inform their investment decisions; such information needed to be comparable, transparent, replicable, and thus credible. Since then, CRAs have been seen as playing a catch-up role in the ESG space, as some parts of the financial sector ecosphere were slower to adapt than others (box 1.1). Since 2016, CRAs have joined in the industry’s dialogue on ESG factors and how they could be relevant for credit assessments. 18 See this background article by the Council on Foreign Relations: https://www.cfr.org/backgrounder/credit-rating-controversy. 19 For background, see this 2012 article published by the European Central Bank: https://www.ecb.europa.eu/pub/pdf/scpwps/ecbwp1419.pdf. 16 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS FIGURE 1.2 - Credit rating definitions provided by the three biggest CRAs CRA Rating definition Credit ratings relating to issuers are an opinion on the relative ability of an entity to meet financial commitments. Fitch Credit ratings express risk in relative rank order, which is to say they are ordinal measures of credit risk and are not predictive of a specific frequency of default or loss. Fitch‘s opinions are forward looking and include Fitch‘s views of future performance. Ratings assigned on Moody’s global long-term and short-term rating scales are forward-looking opinions of the relative credit risks of financial obligations issued Moody’s by non-financial corporates, financial institutions, structured finance vehicles, project finance vehicles, and public sector entities. Issue credit rating is a forward-looking opinion about the creditworthiness of an obligor with respect to a specific financial obligation, a specific class of financial S&P obligations, or a specific financial program. Issuer credit rating is a forward-looking opinion about an obligor's overall creditworthiness. Sovereign credit ratings are forms of issuer credit ratings. Source: World Bank staff illustration. Note: CRA = credit rating agency. Caption: Port in Germany | Source: earth.google.com CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 17 > > > BOX 1.1 - CRA Industry Compared with ESG Providers Industry Although coexisting as two separate industries, credit rating agencies (CRAs) and environmental, social, and governance (ESG) providers seem to be on a path of convergence, but the substance of the “final” sustainable finance terrain remains unclear. This convergence is illustrated by the recent acquisition of ESG providers by CRAs and by market expectations that the financial system will be realigned with the mainstreaming of sustainable finance. While this dynamic is playing out, it is important to understand distinct stages of the two industries’ evolution and how various factors influence their respective products and services (figure B1.1.1). FIGURE B1.1.1 - Key differences between CRAs and ESG providers 1. Maturity of the industries 2. Mandates, regulation, and liabilities 3. Business model 4. Market view Source: World Bank staff. Maturity of the industries. The CRA industry is mature and has evolved over nearly 200 years. Over this period, CRAs have gradually come to play a central role in the financial architecture, and they have well-established mandates, methodologies, and terminologies. Having started with a large number of players, the current CRA industry is dominated by the “Big Three”— S&P Global, Moody’s Investor Service, and Fitch Ratings. The ESG providers’ industry, on the other hand, is young and does not yet have clear terminology associated with it. In their current form, most ESG providers emerged in the 2000sa and their sovereign ESG products have been introduced only recently. Although the ESG providers market has been consolidating and is currently dominated by a handful of players, more than 150 ESG heterogeneous data providers currently exist. (See Q 23, the JPM survey, Gratcheva et al. 2021.) The age of the ESG industry, combined with access to new technologies and (big) data in contrast to the scarce information during the CRA industry’s formation, has led to a marketplace filled with a variety of types of actors. Mandates, regulation, and liabilities. CRAs are regulated and mandated by law to publicly detail their methodological processes, while ESG providers are unregulated. CRAs have evolved over time to fulfill a specific purpose: the CRAs’ role is to assess the capacity and willingness of an issuer to meet its bond payments held by investors on time and in full. For investors, the ratings are one indication of how likely they are to receive their investments back on time and in full. The role of ESG providers, however, has been evolving as sustainable finance becomes mainstream. This growth resulted in a variety of product offerings and lack of clarity of what ESG scores are supposed to measure and how. ESG providers’ products include climate data, analytics, advisory services, corporate and country ESG research and scores, ESG portfolio monitoring, second opinions on compliance with bond principles, third-party assurance, certification and verification, and proxy- voting advisory services. Continue on next page 18 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Continued from previous page Business model: Issuer-paid for CRAs versus investors-paid for ESG providers. Compared with CRA ratings, which are originated and paid for by issuers, ESG scores/ratings/rankings are “unsolicited.” Assessments are typically not requested by issuers and are paid for by investors. The notable exception is the growth of third- party assessments requested by issuers of green and other labeled bonds. One of the most important differences between CRA ratings and ESG scores is the independence of ESG assessments from their subjects (that is, company, sovereign, and so on). In other words, the data sources used for respective ESG assessments are different—solicited versus unsolicited—ultimately affecting the type and quality of the data used for that purpose. In assessing corporate entities, CRAs use data from audited financial statements, according to internationally agreed accounting standards, but as of now no equivalent exists for ESG disclosures (note the recent formation of the International Financial Reporting Standards (IFRS) foundation for that purpose).b Moreover, CRAs have regular discussions with different stakeholders in the country (ministries, private sector, central banks, and so on). As a result, CRAs typically collect information that is not public, and this material often informs the discussions in credit committees and directly contributes to the qualitative adjustments of a credit rating assessment. In the case of ESG providers, the common data source is likely responsible for the high degree of correlation among various ESG providers for G and S scores (Gratcheva, Emery, and Wang 2021). Market view on ESG providers versus CRAs. In the sovereign ESG space, ESG scores are still predominately used to fill the perceived gap left by CRAs of not fully incorporating ESG factors in credit assessments. This fact is illustrated by the J.P. Morgan survey (JPM survey in chapter 2, Gratcheva et al. 2021), in which only 25 percent of survey respondents want sovereign ESG scores to capture the quantification of a country’s sustainability effort and 9 percent want quantification of a country’s sustainability profile, while 64 percent want to see quantification of material credit risks in sovereign ESG scores. Further, 78 percent believe that CRAs will have a larger role in emerging market sovereign ESG ratings compared to ESG providers, while 22 percent disagree. Also, 78 percent believe that improving sovereign ESG fundamentals will lead to lower sovereign credit risk, 16 percent are neutral, and 6 percent disagree somewhat. a. Although we settled on using “ESG providers” in our papers, there are many terms used in the industry and in various publications. b. The IFRS recently published a revised constitution and a feedback statement. See https://www.ifrs.org/projects/work-plan/sustainability-reporting/. In 2015, S&P, as a result of its conduct leading up to the financial crisis, settled with the U.S. Department of Justice (DOJ) for US$1.4 billion and with California Public Employees’ Retirement System for US$125 million. In 2017, Moody’s suffered the same fate and settled for US$864 million with the DOJ, the District of Columbia, and 21 states. Caption: Village in Niger | Source: Climate Investment Funds CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 19 The Principles for Responsible Investment (PRI) ESG in questioned whether CRAs are suitable in the new financial Credit Risk and Ratings Initiative, launched in 2016, has sector architecture that is centered on sustainability and been important in promoting discussion on the relevance forward-thinking finance. The recent J.P. Morgan survey of ESG factors for credit risk assessments.20 The initiative highlights that the investor community clearly believes that has been important in increasing engagement of the CRAs improving ESG fundamentals leads to lower sovereign credit with the industry and investor community, as well as igniting risk, which highlights the potential relevance of these issues CRA communication and research on ESG issues (box 1.2). for CRAs. Indeed, in another recent joint J.P Morgan-World For example, the initiative produced four reports, based on Bank survey of EM asset managers ( “the climate strategy investor and CRA roundtables, helping to foster much industry survey”) on EM sovereign debt climate strategies, over 40 per debate.21 The reports highlighted the complexity of ESG factors cent of respondents agreed that climate risk is not adequately in credit risk analysis, the diversity of perspectives prevailing reflected in sovereign credit ratings. Of particular relevance on this issue, and the challenges pertaining to the topic. for CRAs, respondents also emphasized that their focus is on issuer-level ESG metrics as opposed to specific instrument- The initiative continues to publish quarterly progress level assessments. Among the ESG issues most relevant reports on how the CRA industry is integrating ESG factors for EM sovereign debt investors, institutional strength, social into its credit risk analysis.22 The updates are designed to cohesion and climate change ranked highest. inform investors on relevant CRA tools that may be useful for their ESG integration process. It also allows CRAs to highlight Empirical evidence that demonstrates causation their transparency on this topic and showcase any new between ESG fundamentals and credit risk is mixed. innovations as well as continue the investor-CRA dialogue. The role of governance factors to complement financial and Twenty-six CRAs have now signed the UN PRI’s ESG in credit macroeconomic metrics in sovereign credit analysis is better risk and ratings statement.23 This participation shows the understood. The materiality of environmental and social industry’s commitment to incorporate ESG into credit ratings factors is not as well documented, and the studies are often and analysis in a systematic and transparent way. affected by an ingrained income bias (Gratcheva, Emery, and Wang 2021). Yet events such as climate disasters or Despite these developments, many financial market the COVID-19 pandemic have underscored that ESG can participants continue to see CRAs as not adapting fast materially affect the long-term profile of an issuer. or deeply enough and have been calling on the CRA industry to do more. Further, sustainability advocates have > > > BOX 1.2 - Participant Views in United Nations PRI Study These key challenges were identified in the Principles for Responsible Investment (PRI) dialogues and publications: The time horizon that CRAs use for their credit assessments. CRAs argue that the ESG-related issues are more difficult to capture over longer time horizons.a Nevertheless, the dominant view among investors is that ESG- related risks are not adequately captured within CRA ratings and that CRAs should explicitly weight ESG factors in their methodologies. At the same time, a typical horizon for CRAs’ credit assessments is about 3–5 years, with 10 years for long-term ratings, while the nature of the investors’ portfolios determines the investors’ horizon. Continue on next page 20 The UN’s Principles for Responsible Investment was launched in 2005. It is supported by, but not a part of, the United Nations. The aim of the initiative is to enhance the transparent and systematic integration of ESG factors into credit assessments. 21 For the June 11, 2018, report, see https://www.unpri.org/credit-risk-and-ratings/esg-credit-risk-and-ratings-part-2-exploring-the-disconnects/3250.article. 22 For an example, see https://www.unpri.org/credit-risk-and-ratings/credit-rating-agency-cra-quarterly-updates/5631.article. 23 See the statement on ESG in credit risk and ratings: https://www.unpri.org/credit-risk-and-ratings/statement-on-esg-in-credit-risk-and-ratings-available-in-different- languages/77.article. 20 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Continued from previous page For long-term investors, their respective horizons are longer than CRAs’ horizons, and those investors view themselves as more exposed to ESG risks than is measured in credit assessments. The position of CRAs is that their appropriate time horizon is shorter than the materialization horizon of many ESG factors. They also argue that the incorporation of ESG factors is already implicit within other included credit factors and that they do not ignore ESG issues altogether as some market participants perceive.b The visibility of material ESG risks (figure B1.2.1). Participants mentioned the lower visibility of material ESG risks as a major challenge with various views expressed on this matter. Some investors expressed the opinion that CRAs should leverage the issuer-pays CRA model to increase the amount of ESG-related information that is available to the financial system. Several investors also argued that CRAs should highlight the ESG issues that affect an issuer irrespective of whether the materialization of those issues would be later than the time horizon of the credit rating assessment. This suggestion could prove difficult from a regulatory viewpoint. Conflicting objectives for incorporating ESG into investment decisions—credit risk or sustainability. In the CRAs’ view, investors need to better understand the objectives behind their ESG considerations and not to confuse risk management with impact investing.c In addition, CRAs suggest that investors need to engage more with the rating industry on how ESG factors are integrated into credit assessments. FIGURE B1.2.1 - CRA assessments are generally shorter term, but many longer-term issues are material to the assessment of sovereign credit risks CRA assessment horizon Materiality Forward-looking horizon GDP growth Human capital Inflation Trade balance Forest capital Current account Fossil fuels Protected areas I international investment position Mineral wealth Short-term growth depends on long-term growth potential Source: World Bank staff illustration. Note: CRA = credit rating agency; GDP = gross domestic product. a. This point of view continues to be the key issue that CRAs articulated during the authors’ discussions with them for this paper in the latter part of 2020. b. In the authors’ discussion with CRAs, this point of view was still held by the three leading CRAs, while the smaller CRAs acknowledged the possibility that, going forward, ESG risks are likely to materialize sooner and have a greater effect than in the past. Although the Big Three did mention that climate, environmental, and social risks will become more material in the coming years, they noted that the data and methodologies to capture these effects are not readily available. c. The authors address investors’ conflation of the use of ESG factors for risk management versus sustainability in “A New Dawn: Rethinking Sovereign ESG” (Gratcheva et al. 2021). CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 21 2. Caption: King penguins in South Georgia and the South Sandwich Islands | Source: Unsplash.com 22 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS >>> ESG Factors and Sovereign Credit Ratings The CRA industry emphasizes that only ESG credit factors (that is, ESG factors that affect creditworthiness of an issuer) should be included in a sovereign credit assessment and that these factors should not be conflated with other ESG factors that do not affect the ability of a sovereign issuer to repay its debt obligations (see figure 2.1). CRAs argue that sovereign ESG credit factors that can be modeled on the basis of current data and existing analytical methods already in sovereign credit assessments. Although governance factors have historically been a key part of the sovereign credit risk assessment, environmental and social factors have not. Furthermore, on the environmental side, the inclusion of low-probability, high- impact events is analytically challenging. Nevertheless, contingent liabilities are often a driver of sovereign credit ratings, and most CRAs attempt to capture such risks in their sovereign credit assessments. In a similar way, many environmental risks, when considered via the spectra of climate tipping points,24 could be thought of as credit-relevant contingent liabilities. FIGURE 2.1 - The intersection of sovereign ESG factors and sovereign credit factors ESG ESG factors credit Credit factors factors Source: Adapted from S&P 2021. 24 A climate tipping point is a threshold that, when exceeded, leads to large and often irreversible changes in the state of the system (Hoegh-Guldberg et al. 2018). CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 23 Figure 2.2 is based on a keyword search of CRA methodological and can affect asset classes and countries in different ways, documents25 and highlights the limited focus on E and S all of which make it a difficult concept to formalize. Although factors in sovereign credit rating assessments’ methodologies. much of the discussion around materiality focuses on the Although some CRAs may argue that these factors are corporate sector, we argue that it is particularly relevant in included outside their quantitative assessment, the lack of the context of sovereign ESG. Certain E or S factors may not clarity for investors on their inclusion makes the investment be material from a credit factor perspective (for example, they process more cumbersome and less transparent. would not result in a change to a sovereign credit assessment), but they could still be material from a financial viewpoint29 The concept of materiality is fundamental to the discussion (for example, they would result in an asset repricing30). The around ESG credit factors. Materiality26 is a principle that persistence of this asset repricing would depend on the defines which information is useful to an investor given materiality of the credit factors. Moreover, sustainability topics specific investment objectives, such as using ESG factors as that are considered immaterial to financial value creation today an input or output in the investment process.27 Materiality can may become material over time, either gradually or suddenly be transitory in nature, can be important from both a financial because of catalyst events, policy, or regulatory reaction as and impact perspective (for example, double materiality28), well as innovation.31 Caption: Building climate resilience of watersheds | Source: Climate Investment Funds 25 Based on the authors’ judgment. 26 See https://www.sasb.org/blog/double-and-dynamic-understanding-the-changing-perspectives-on-materiality/. 27 “ESG as input” and “ESG as output” are two mutually nonexclusive approaches to ESG investing: (a) ESG integration or a purpose-neutral approach using ESG factors as an input into the investment process to manage ESG-related risks that affect the financial risk of the investment portfolio and (b) a purposeful approach using ESG factors as an output of the investment process to achieve measurable, sustainable results. 28 As stated in European Commission, “Guidelines on Non-financial Reporting: Supplement on Reporting Climate-Related Information,” June 20, 2019. 29 Certain E or S factors could be more relevant from an economic viewpoint and be reflected in various economic variables. These factors may or may not implicitly affect financial market price action. 30 A natural disaster is unlikely to result in most sovereigns defaulting on their debt obligations, but it could be reflected in financial market price action and result in market volatility. Nevertheless, such natural disasters may be more material for lower-income countries that rely heavily on a certain industry. 31 See http://eifrs.ifrs.org/eifrs/comment_letters/570/570_27198_SonalDalalSustainabilityAccountingStandardsBoard_0_SASBResponsetoIFRSConsultation11Dec2020.pdf. 24 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS FIGURE 2.2 - ESG factors and sovereign credit rating methodologies The figure shows the 100 most-mentioned terms in the methodology documents of the seven CRAs. As in FIgure ES1, environmental and social terms applear to play a very limited role. P GD Source: World Bank staff illustration. Data from sovereign credit rating methodology by Fitch Ratings, Moody’s Investor Service, S&P Global Ratings, Kroll Bond Rating Agency KBRA, DBRS Morningstar, Scope Ratings GmbH, and HR Ratings CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 25 Caption: Aerial view | Source: Unsplash.com Sovereign credit ratings directly influence a country’s scorecard in a qualitative way, if viewed as relevant for cost of capital (that is, a lower rating implies a higher sovereign creditworthiness.32 Currently, only one CRA— cost of capital), while there is no clear link between Scope Ratings GmbH—explicitly includes ESG credit factors sovereign ESG scores influencing sovereign borrowing in its sovereign credit rating methodology (see appendix A).33 costs. Lower-rated sovereigns generally pay a higher interest Figure 2.3 shows examples from S&P and Moody’s research rate because of perceived higher credit risk. The higher publications that illustrate the materiality of ESG factors in cost of capital can also affect market access and therefore credit ratings. For sovereigns, governance factors dominated significantly limit sovereign financing options. A sovereign’s while social and environmental factors were less material. credit rating generally represents the highest rating attainable CRAs have recently been launching additional products by most issuers domiciled within the country. on the other designed to show how ESG factors are relevant in their hand, the effect of a change in a sovereign’s ESG score on credit assessments and products similar to scores by the sovereign’s cost of capital is much more ambiguous. ESG providers that assess countries on a broad set of Although a less positive ESG score could affect some investor ESG factors. For example, in 2020 Moody’s introduced E, allocations over time, such effects are currently much more S, and G issuer profile scores and credit impact scores34 difficult to pinpoint. As a result, most countries are less designed to provide the market with more granularity on how incentivized to improve their sustainability performance. ESG risks influence the rating process while Fitch launched CRAs have devised other ways to illustrate the extent to its ESG relevance scores35 for sovereigns to show how which ESG factors are material in their sovereign credit ESG risks affect sovereign rating decisions. In 2021, S&P rating decisions. Some CRAs have published research that Global has also begun issuing ESG credit-related indicators highlights the materiality of ESG factors in rating assessments, reflecting a qualitative assessment—determined typically by and republished their rating methodologies to clarify more a rating committee—of whether ESG factors have a neutral, explicitly the role that ESG factors may play in a credit positive, or negative influence on the key components assessment. CRAs have also started to provide clarifications of a credit rating assessment.36 Other CRAs indicated in in their formal rating assessments on the credit relevancy discussions with the team that they are also working on of ESG factors in the rating decision. In many instances, introducing similar new products. CRAs assess E and S risks outside of a rating methodology 32 Qualitative analysis requires the ratings analyst to exercise judgment. Rating assessments are also based on scorecards, a component of which is qualitative, and these scorecards serve as guides for discussion in rating committees. 33 See the Scope Ratings publication “Sovereign Rating Methodology: Sovereign and Public Sector,” October 8, 2019, https://www.scoperatings.com/ScopeRatingsApi/api/ downloadmethodology?id=01508950-119c-4ab5-9182-54fffdc1003f. 34 See https://www.moodys.com/research/Moodys-updates-its-methodology-for-assessing-environmental-social-and-governance--PBC_1254678?showPdf=true. 35 The scores are subjective judgments on the extent to which E, S, and G risks elements influence credit rating decisions. 36 See https://sustainableinvest.com/sp-global-ratings-to-issue-esg-credit-related-indicators/. 26 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS FIGURE 2.3 - Materiality of ESG Factors on Credit Ratings a. Breakdown of S&P credit rating actions (2021) affected by ESG factors ESG-Related Credit Rating Actions Including Structured Finance (January–October 2021) INFRASTRUCTURE PUBLIC FINANCE INTERNATIONAL CORPORATES STRUCTURED SOVEREIGNS U.S. PUBLIC INSURANCE FINANCE FINANCE FI AND TOTAL AND Downgrade 10 10 67 52 122 4 265 CreditWatch negative 0 0 44 16 15 0 75 Downward outlook revision 6 1 40 19 0 66 Upgrade/Upward outlook revision 5 2 9 53 11 2 82 Total ESG-related rating actions 21 13 160 140 148 6 488 Of which social 13 11 33 97 131 2 287 Of which governance 12 2 75 17 8 6 120 Of which environmental 0 0 70 30 9 109 a. Breakdown of S&P credit rating actions (2021) affected by ESG factors Economimc Systemic Currency ESG Risks Recession Banking Crisis Crisis (15 cases) (37 cases) (14 cases) (18 cases) Sovereign Debt Crisis Source: Panel a, Copyright 2021, Standard & Poor’s Financial Services LLC. All rights reserved. Panel b, Moody’s Investor Service. Sample includes he 42 sovereign bond defaults from 1997 to July 2020. Note: Panel a, FI = financial institutions; SovlPF = sovereigns and international public finance. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 27 3. Caption: Pollinating bee | Source: Pexels.com 28 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS >>> Sovereign Credit Ratings and Sovereign ESG Scores In this section we empirically analyze how sovereign ESG scores are manifested in sovereign credit ratings. For the subsequent analysis, we use the average credit ratings from the three largest CRAs—Fitch, Moody’s, and S&P—and relate them to the average sovereign ESG scores from leading ESG providers (Gratcheva, Emery and Wang 2021). We use the average of the three CRAs to proxy a country’s creditworthiness, given the highly concentrated nature of the credit rating industry and because they rarely assign fundamentally different ratings. For sovereign ESG scores, however, it is not always clear what constitutes good ESG performance (Gratcheva, Emery and Wang 2021). This disagreement mostly affects the environmental pillar and less the social or governance pillars. Nonetheless, taking the average over six leading ESG providers should yield a useful proxy of the countries’ ESG profiles. Thus, the term “sovereign ESG scores” used in this paper represents a useful proxy for the sovereign ESG sector as a whole.37 We find that on average, sovereign credit ratings and sovereign ESG scores are highly correlated—especially for governance and social scores—which is likely explained by the predominant role of the ingrained income bias (Gratcheva, Emery and Wang, 2021). Figure 3.1 shows that the aggregated S and G scores are highly correlated with credit ratings, 83.1 percent and 81.6 percent, respectively, while the E score is correlated to a lesser degree, with 66.5 percent. The correlation of credit ratings and ESG scores is consistent with our previous findings about the predominant role of the country’s level of income—or development, in general—in current industry measures of sustainability (Gratcheva, Emery, and Wang 2021; Gratcheva et al. 2021). These sustainability metrics include sovereign ESG scores and other measures such as the SDG index and environmental indices such as the Notre Dame Global Adaptation Initiative (ND-GAIN) Country Index, ND-GAIN adjusted for gross domestic product (GDP), and Environmental Performance Index (EPI). As we will see next, in order to understand the materiality of ESG factors for sovereign credit ratings, it is important to account for the predominant role of income. 37 In the following, we only compare the average credit rating and the average ESG scores with each other. Thus, we only mention the term “average” when necessary. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 29 FIGURE 3.1 - Do better ESG scores equal better credit ratings? Correlating scores of each ESG pillar with the credit ratings of 115 countries reveals a strong positive correlation. Countries with better ratings also tend to have better ESG scores. This correlation is particularly true for social and governance scores and less so for environmental scores. Environmental pillar Social pillar Governance pillar r=66.5% r=83.1% r=81.6% Credit rating Score value Score value Score value Source: World Bank staff calculations. Note: The vertical axes depict the rating scale, where higher is better, and the horizontal axes measure the aggregated E, S, and G scores, averaged over six leading ESG providers. The dashed line distinguishes between investment-grade ratings (above) and non-investment-grade ratings (below). Analyzed for each World Bank income group, the compared to their credit ratings? and (b) why are ESG materiality of ESG scores for the credit rating gradually scores almost entirely uninformative about low-income breaks down as we move lower down the country income country’s credit ratings? It is worth recalling that different specturm. Figure 3.2 repeats the correlation analysis and income levels also reflect different stages of development. accounts for the different income levels by grouping countries Thus, how informative a sovereign ESG score is depends according to their income classification. We see that, for high- on how developed a nation is. For instance, research and income countries only, ESG factors are strongly aligned with development (R&D) expenditure in the governance pillar sovereign credit ratings. As we consider upper-middle income is a more useful indicator for high-income countries, whose and lower-middle income groups, we find that the alignment economies strongly depend on human capital. For low-income between sovereign credit ratings and ESG factors slowly countries—assuming the values are not missing—the same breaks down until we reach low-income countries, where indicator is less relevant because human capital makes up the correlation vanishes almost entirely. This trend is clearly a comparatively smaller share of a country’s growth. At the visible in figure 3.2, where the black regression line starts off same time, social indicators such as infant mortality or with a clear upward slope but flattens as we traverse along undernourishment metrics are arguably more informative for the income categories. In other words, ESG factors and credit lower-income countries and are important drivers of these ratings are strongly correlated with each other for high-income countries’ development in the long term. However, from the countries, but for low-income countries, the exact opposite perspective of CRAs, the importance of social indicators for holds. No matter whether low-income countries score high on creditworthiness over a shorter time horizon likely pales in E, S, or G, they all share a similar credit rating.38 For example, comparison with traditional macroeconomic, geopolitical, Burkina Faso, Mozambique, Rwanda and Tajikistan are rated or fiscal variables. Box 3.1 takes a more robust statistical B or lower by the main CRAs but display high ESG scores.39 approach to understanding the relationship between sovereign credit ratings and sovereign ESG scores, and this approach Our findings raise two questions: (a) what additional also confirms our initial analysis. information do ESG scores as currently produced by the ESG providers convey about high-income countries 38 Low-income countries are clustered between B and CC ratings, which impedes their analysis. The 10 low-income countries studied here are all non-investment-grade level and already in an elevated default risk state. This situation presents a challenge for our analysis of materiality of ESG to credit ratings. Furthermore, because our sample comprises only 10 low-income countries, results may be driven by a very few outlier countries. 39 For example, Burkina Faso and Rwanda have standardized governance scores of 40.6 and 39.6, respectively (the country with the highest governance score in the sample receives a score of 100). As a reference, the average governance score of lower and upper middle income countries are 20.0 and 37.7, respectively. Similarly, Tajikistan has a social score of 40.8 while the average social score of lower and upper middle income countries are 36.7 and 50.4, respectively. 30 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS FIGURE 3.2 - The lower the income group, the less the ESG scores and ratings are correlated Rather than taking a global view of all 115 countries together (as in figure 3.1), we examine the relationship between ESG scores and credit ratings for each income group. Interestingly, the relationship is strongest for high-income countries as reflected by the steep regression line. As we consider upper-middle, lower-middle, and low-income groups, the line flattens more and more. For low-income countries, ESG scores appear to be almost immaterial. All countries High income Upper-middle Lower-middle Low income (n=115) (n=44) income (n=34) income (n=27) (n=10) Environmental pillar Social pillar Governance pillar Investment grade Non-investment grade Source: World Bank staff calculations. Note: The vertical axes depict the rating scale, where higher is better, and the horizontal axes measure the aggregated E, S, and G scores, averaged over six leading ESG providers. The dashed line distinguishes between investment-grade ratings (above) and non-investment-grade ratings (below). CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 31 > > > BOX 3.1 - Another Approach to Assessing the Materiality of Sovereign ESG Credit Factors A more rigorous statistical model reveals that only S and G scores are material to a country’s sovereign credit ratings, while E scores are largely irrelevant. Figure B3.1.1 displays the results of a probability regression model for each individual ESG pillar. The results show how the likelihood of a country’s credit rating’s belonging in one of the six credit quality scores (CQS) would change if that country were to improve its E, S, or G score. Countries with higher social and governance scores are more likely to have a higher credit rating, while countries’ E scores are not statistically material to sovereign credit ratings. The probability or likelihood of the ESG factors being credit material is depicted by the length and color of the arrow. These findings corroborate our previous findings that sovereign credit ratings incorporate S and G factors but not necessarily E factors. FIGURE B3.1.1 - Are countries with better ESG scores more likely to belong to a better rating group? A probabilistic approach If we consider a hypothetical “average” country from the sample of 115 countries (such as a middle-income country with a BBB– credit rating), what would happen if the country then had a higher E, S, or G score? Would it remain in the BBB+ to BBB– rating group, or would it be more likely to be part of the AAA to AA– group? The arrows show how this change would play out for each ESG pillar. Better S or G scores are more likely associated with ratings of BB– or better and less likely with ratings below B+. For the E scores, no such statements can be made. Environmental Social Governance less likely more likely less likely more likely less likely more likely Statistically Statistically significant changes insignificant changes Source: World Bank staff calculations Note: We use an ordered probit model for panel data with income-group fixed effects. It estimates the probability of belonging into one of the six rating groups (credit quality steps; see table B.1) on the vertical axes, based on higher E, S, and G scores. The effect sizes reported are the marginal effects for the average observation. The green and red arrows show how such a change would increase or decrease the probability of being in the respective CQS group. Transparent arrows indicate that the effects are not significant on the 5 percent significance level. The sum of all arrows, or the net change of probability, is zero. For regression details, see table B.2,in annex. 32 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Caption: Schoolboys in Mogadishu, Somalia | Source: Unsplash.com CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 33 4. Caption: Bird migration | Source: Unsplash.com 34 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS >>> Sovereign Creditworthiness and a Country’s Wealth This section introduces the World Bank country wealth data set (World Bank 2021) and empirically examines its usefulness to CRA and ESG providers in their respective sovereign assessments. A country’s wealth is a record of its natural, produced, and human capital, and a fall in wealth indicates that a country is depleting its assets and may not be able to sustain its future GDP growth and vice versa. The data set expresses a country’s wealth in terms of economic or monetary value, which allows for better comparability across scores and countries. The data are relevant for CRAs and investors alike, both of whom acknowledge the need to also take a longer-term perspective on a country’s sustainability to inform a more holistic credit assessment, beyond current credit rating horizons. This data set is also relevant for ESG providers who want to measure a country’s sustainability metrics in an economically meaningful way. We find that sovereign credit ratings are strongly correlated with human and produced capital while natural capital exhibits a much less clear-cut relationship (figure 4.1). Human and produced capital40 are measurements of a country’s prosperity and therefore relevant to its ability to service its debt obligations. Natural capital, in contrast, exhibits a much less clear- cut relationship with of sovereign credit ratings as currently measured. Although figure 4.1 illustrates an overall positive correlation, the wide confidence intervals prevent us from drawing any definite conclusions. As in section 3, examining the relationship from high-income groups to low-income groups reveals a similar flattening effect. Figure 4.2 illustrates that the strength of this relationship drops as we move from higher-income to lower-income countries. 40 Human capital is computed as the present value of future earnings for the working population over their lifetimes. Produced capital includes the value of machinery, buildings, equipment, and residential and nonresidential urban land. See appendix A, Wealth Data section. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 35 FIGURE 4.1 - More wealth = better credit ratings? Correlating the wealth accounts for human, produced, and natural capital with the credit ratings of 115 countries reveals a positive correlation. Countries with more human and produced capital also tend to have better credit ratings. However, this tendency is not the case for natural capital. Human capital Produced capital Natural capital correlation = 84.5% correlation = 80.3% correlation = 29.0% Investment grade Non-investment grade credit rating Wealth indicator Wealth indicator Wealth indicator Source: World Bank staff calculations. Note: The vertical axes depict the rating scale, where higher is better, and the horizontal axes measure the corresponding wealth indicators (World Bank 2021). The dashed line distinguishes between investment-grade ratings (above) and non-investment-grade ratings (below). Caption: Lao girls | Source: Climate Investment Funds 36 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS FIGURE 4.2 - The lower the income group, the less the correlation between wealth and ratings Rather than taking a global view of all 115 countries together (figure 4.1), we examine the relationship between wealth and ratings for each income group. The relationship is strongest for human and produced capital in high- income countries, as reflected by the steep regression line. As we consider upper-middle-, lower-middle, and low-income groups, the line flattens more and more. For natural capital and its components, the line is flat for all income categories, indicating that natural capital does not manifest itself in credit ratings. In low-income countries, the line is even downward sloping. All countries High income Upper-middle Lower-middle Low income (n=115) (n=44) income (n=34) income (n=27) (n=10) Human capital corr=84.5% corr=80.0% corr=18.1% corr=10.0% corr=44.4% Production capital corr=80.3% corr=69.1% corr=10.4% corr=–15.3% corr=13.0% Natural capital corr=29.0% corr=12.7% corr=–15.0% corr=–31.5% corr=–29.6% Renewables (forests, agriculture, protected areas) corr=27.5% corr=12.3% corr=4.3% corr=–25.2% corr=–16.3% Subsoil assets (fossil fuel, minerals) corr=19.1% corr=26.5% corr=–13.8% corr=–11.0% corr=–54.1% Investment grade Non-investment grade Source: WB staff calculations Note: The vertical axes depict the rating scale, where higher is better, and the horizontal axes measure the corresponding wealth indicators (World Bank 2021). The dashed line distinguishes between investment-grade ratings (above) and non-investment-grade ratings (below). CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 37 We find that a country’s natural capital is not a strong consistently negative but weak relationship between natural credit factor in sovereign credit assessments for all capital components and sovereign credit ratings. This finding income levels. This finding is not surprising for higher- might be supported by the “natural resource curse,” whereby income countries, which enjoy better credit ratings and whose countries rich in natural capital but weak in governance are economies are less dependent on natural wealth. However, prone to corruption or rent-seeking behavior. Those harmful the low relevance of natural capital for lower-income countries types of activities jeopardize a country’s economic outlook is unexpected because natural resources are the primary and, therefore, its ability to service debt obligations. We add driver for early economic development, fueling the transition a caveat for these results because the number of low-income into an economy that is less reliant on rents from natural countries included in our study is low. In section 5, we find resources. Moreover, natural capital, such as sustainably an additional explanation for the irrelevance of natural capital: managed protected areas, is important for industries such as nonrenewable resources also constitute part of natural wealth tourism. For countries with a significant tourism industry, it is and, in the context of the low-carbon transition, subsoil assets particularly important that CRAs account for the potential of can have both positive (resource blessing) and negative natural capital in their credit rating assessment. As economies (resource curse) effects on sovereign credit risk. grow in prosperity, the share of natural wealth of their total We extend our investigation of the relationship between wealth continuously drops while produced capital and human wealth variables and sovereign credit ratings by using capital gain in importance (figure 4.3). This effect is also reflected in the drop of natural resource rents as part of GDP a more rigorous statistical approach. By employing an as countries grow richer.41 ordered probit model with income-group fixed effects, we can estimate how a hypothetical change in a country’s wealth We dig deeper into the natural capital pillar and find that account affects its likelihood to belong to a specific rating both renewable (ecosystems) and nonrenewable (fossil category.42 More concretely, consider a middle-income country fuels) natural capital appear to have only a weak, negative with a credit rating of BBB– and a wealth profile as depicted in association with credit ratings, if at all. If we focus on figure 4.3.43 Given what we know about the wealth profiles of lower-middle-income and low-income countries, we do find a all other countries and their corresponding ratings, what would FIGURE 4.3 - Share of natural capital for different income groups This figure shows how natural capital’s share of total wealth drops as countries climb the development ladder. It is replaced by the growing importance of produced capital, which is highly correlated with GDP. High-income 3.3 34.4 62.3 Middle-income 8.8 26.0 65.2 Low-income 24.8 26.8 48.4 0 25 50 75 100 Share of total wealth (%) Natural capital Produced capital Human capital Source: World Bank (2021). 41 For high-income countries, the total natural resource rents as share of GDP in the observation period (1996–2018) was 4.00 percent, for upper- and lower-middle-income countries—7.02 percent and 7.44 percent, respectively, and 12.62 percent for low-income countries. 42 The ordered probit model is a static model. Thus, changes in probabilities of belonging to one of the six credit quality steps (CQS) groups are calculated on the basis of cross-sectional variation among the 115 countries in 2017. 43 The results of the ordered probit model hold for a hypothetical average representative country of the sample that contains 133 countries. In our case, such a country would be a middle-income country with a BBB– rating. 38 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS happen if the BBB-rated country had more human capital metrics are likely to be the key driver of credit assessments, while everything else stays equal? Figure 4.4 shows us that the inclusion of data on natural capital could be an important countries with above-average human capital are more likely insight on a country’s longer-term growth model and thus be to have an AA– rating or better (11.9 percent and less likely to relevant for a sovereign credit assessment. Natural capital is be rated between B+ and B– (11.3 percent A similar result is also relevant for the ESG provider industry. Box 4.1 provides found if the middle-income country with BBB– rating has more an overview of the relationship between these two measures. produced capital, but natural capital remains the exception. Given the strong conceptual foundation of natural capital wealth data, the data set could be highly relevant for assessing Our findings suggest that natural capital could be an a country’s sustainability profile. important input into credit rating assessments of lower- income countries. Although macroeconomic and debt credit FIGURE 4.4 - Are countries with higher wealth accounts more likely to belong to a better rating group? A probabilistic approach If we consider a hypothetical “average” country from the sample of 115 countries (such as a middle-income country with a BBB– credit rating), what would happen if such a country had more human, produced, or natural capital? Would it remain in the BBB+ to BBB– rating group, or would it be more likely to be part of the AAA to AA– group? The arrows show how this scenario would play out for each ESG pillar. Countries richer in human or produced capital are more likely associated with ratings of BB– or better and less likely with ratings below B+. For natural capital, no such statements can be made. Source: World Bank staff calculations. Note: We use an ordered probit model for panel data with income group fixed effects. It estimates the probability of belonging into one of the six rating groups (credit quality steps; see table B.1) on the vertical axes, based on higher E, S, and G scores. The effect sizes reported are the marginal effects for the average observation. The green and red arrows show how such a change would increase or decrease the probability of being in the respective CQS group. Transparent arrows indicate that the effects are not significant on the 5 percent significance level. The sum of all arrows, or the net change of probability, is zero. For regression details, see table B.2. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 39 > > > BOX 4.1 - Natural Capital Components and Environmental Indicators: Two Measures of the Same? At first glance, natural capital components, which are wealth accounting variables (World Bank 2021), and environmental indicators, as defined on the World Bank Sovereign ESG Data Portal,a appear to measure very similar things. However, as figure B4.1.1, panel a, demonstrates, the relationship is not a simple one-to-one mapping. It is sometimes tempting to see a natural capital indicator (left) as the monetary equivalent of an environmental indicator (right), such as forest capital and forest land. These two factors – while having similar names – represent different metrics: , forest area is not featured in the forest capital calculation, which instead depends on the lifetime of the asset, the rents it produces, and the production (or extraction) quantity (World Bank 2021). Figure B4.1.1, panel a, also presents another type of relationship: growth in a natural capital account can lead to measurable changes in an environmental indicator. For instance, growth in oil, gas, and coal capital (left), which involves extraction, production, and transportation, can lead to measurable CO2 emissions (right). Another type of link is established when an environmental indicator is relevant for growth in a natural capital account. An example is the case of cooling degree days (right) and agricultural capital (left), because crop health and plant growth require sufficient days above crop-specific threshold temperatures over extended periods of time. Thus, natural capitalb and environmental scores are related in multiple ways. However, figure B4.1.1, panel b, shows that these conceptual relationships do not necessarily translate into quantifiable correlations. This translation difference can be explained in various ways. Natural capital focuses on the long-term economic value of forests, agricultural land, protected areas, and fossil fuel and mineral resources. In contrast, the E pillar aims at characterizing a country’s current state in terms of emissions, energy use, climate risks, and food security. This distinction is also reflected in the units of measurement: E indicators are calculated as (shares of) metric tons, square kilometers, temperature degrees, and so on. Natural capital is always presented in dollar values. As Gratcheva and Wang (2021) discuss, this low correlation is the result of two complementary approaches to measuring sustainability. Neither approach alone is sufficient to tell a complete story. Instead, a combination of both is necessary for the data to become relevant for economic and financial decision-making. As those authors mention, wealth accounting variables and environmental indicators are natural allies. a. For the portal, see http://esgdata.worldbank.org/.  atural capital is computed as the present value of future economic rents of various renewable (such as forests, b. N agriculture, protected areas) and nonrenewable (such as oil, gas, minerals) resources. The lifetime of the resource depends on its extraction rate and, in the case of renewables, its regeneration rate. See appendix A, Wealth Data. Continue on next page 40 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Continued from previous page FIGURE B4.1.1 - Natural capital components and environmental indicators (selected from World Bank data portal) a. Natural capital components (left) and environmental indicators (right) are conceptually related b. However, the conceptual relationship does not translate into a strong quantitative relationship High income Investment grade r=10.0% Upper- middle Environmental pillar income r=46.6% Non-investment grade Lower- middle income r=76.1% Low income Natural capital r=45.1% r=22.3% Source: World Bank staff calculations. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 41 5. Caption: Wheat field | Source: Pexels.com 42 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS >>> The Low-Carbon Transition and Stranded Assets in Sovereign Credit Ratings The incorporation of climate change risks, both physical and transition risks, into sovereign credit assessments poses a particular challenge for CRAs.44 This section focuses on the so-called stranded assets risks45 that many countries face in transitioning to a low-carbon economy. These risks relate to the possibility that certain industries or natural resources that may now be regarded as a credit-positive sovereign asset would depreciate significantly in value as society moves toward cleaner energy sources and therefore negatively affect sovereign creditworthiness. Recent global policy actions46 promoting the green agenda could also speed up such a transition and pose a significant challenge to economies that are less diversified. Indeed, the investment community is increasingly aware of these risks as evidenced by a recent J.P. Morgan survey (“the climate strategy survey”) of EM investors that indicate that stranded asset risks are now considered on a par with climate change as an investment risk. Sovereigns with economies highly dependent on fossil fuels are most exposed to stranded-assets risks (figures 5.1 and 5.2). As demand for fossil fuels wanes, unless the sovereigns are successful in diversifying their economies away from such fuels, the sovereigns will face a drop in fiscal revenue and export receipts, leading to economic decline, higher government debt, political instability, and rising financing costs. Sovereigns with strong balance sheets and potential to diversify their economies are better placed for the carbon transition. Because the path of the energy transition is highly uncertain, CRAs maintain that there is low visibility to such factors. It is expected that stranded-asset risks will build up over time and trigger more rating changes as the effects become clearer, closer, and more material (Fitch 2021). 44 Physical risks include the frequency and severity of natural disasters and extreme weather events, rising sea levels, and other consequences of climate warming. Transition risks are risks related to the process of transitioning from traditional economic models that heavily rely on fossil fuels to a low-carbon economy that is based on low-carbon power sources. 45 “Stranded assets” often refers to those assets that depreciate because of declining demand mainly caused by their negative effects on climate. Oil, coal, and gas are considered typical examples of stranded assets, but such assets might also include power plants that retire early because new emission regulations are imposed or fossil-fuel fields that are discovered but are not developed because of new regulations or low demand (IEA 2013). 46 These actions include the European Union’s green recovery plan and the United States’ new targets for the reduction in greenhouse gas emissions by 2030. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 43 FIGURE 5.1 - Selected countries’ rents Fossil Fuel Rents, 2018 Oil Natural gas Coal (% GDP) 60 40 20 0 Congo, Rep. Iraq Kuwait Saudi Arabia Oman Azerbaijan Angola Qatar Gabon Kazakhstan Russian Federation Usbekistan Norway United Arab Emirates Mongolia Nigeria Mozambique Source: Fitch Ratings, World Bank. FIGURE 5.2 - Selected countries’ receipts Fossil Fuel Receipts, Percentage of Total, 2019 (%) 100 80 60 40 20 0 Iraq Oman Kuwait Bahrain Congo, Rep. Saudi Arabia Angola United Arab Emirates Azerbaijan Nigeria Gabon Kazakhstan Malaysia Usbekistan Russian Federation Bolivia Cameroon Mongolia Indonesia Ghana Qatar Ecuador Source: Fitch Ratings, World Bank. The assessment of a sovereign’s low-carbon transition economy or to move to more sustainable energy sources that is challenging because of uncertainty around its can positively affect credit ratings. For example, the “economic trajectory and how sovereign’s economic activity and strength” factor by Moody’s includes “other considerations,” creditworthiness will be affected. Most CRAs include a which are mainly qualitative and can include adjustments small weight in their methodologies to capture a country’s for sovereigns that are undergoing structural changes that exposure to fossil fuels. There are different types of factors may not be appropriately captured by other indicators. S&P that could partly capture a country’s exposure such as (a) mentions that a country’s shift to a renewable energy source commodity dependence, economic diversity, or economic as a substitute for an imported energy source could lead to concentration indicators; (b) fiscal flexibility indicators and improving external metrics, for example, related to the country’s fiscal revenue structure (such as capturing the concentration of imports or volatility of the country’s terms of trade (S&P 2018). fiscal inflows on the fossil fuel industry); (c) overall robustness Fitch includes “commodity or sector dependence” as an of government policy framework and governance strength; (d) indicator in the external finance component of its sovereign volatility of GDP growth; (e) scenario analysis; and (f) additional rating methodology. Box 5.1 provides an example of recent qualitative considerations. Some CRAs’ methodologies include CRA treatment of sovereign stranded-asset risks. qualitative assessments of a sovereign’s ability to diversify its 44 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS > > > BOX 5.1 - How Do Stranded-Asset Risks Affect Credit Ratings of Persian Gulf Countries? Persian Gulf countries are among the most exposed to stranded-asset risks globally, although they display moderate resilience to these risks. Figure B5.1.1 illustrates that Persian Gulf countries display a relatively small advantage on resilience scores relative to peer countries while being significantly more exposed to stranded-asset risks. Stranded-asset risks are not seen by CRAs as a driver of current credit assessments for many of these countries. For example, Fitch, which affirmed Qatar at AA– in June 2021, did not mention stranded-asset risks in its rating action commentary.a Although the assessment did mention the country’s hydrocarbon dependency, the overall focus of the commentary is from the fiscal and debt perspective. On ESG credit factors, only governance credit factors were mentioned, and other environmental or social credit factors were deemed either credit neutral or having only a minimal credit impact. Moreover, in the rating sensitivities section—where the CRA outlines factors that could lead to a negative rating action or downgrade—any mention of stranded-asset risks or indeed other E issues were absent. FIGURE B5.1.1 - Low-carbon transition (LCT) resilience and exposure scores of Persian Gulf countries, compared to the average scores of rating peers a. LCT resilience scores b. LCT exposure scores Resilience to LCT Exposure to LCT United Arab Emirates (AA/Aa2/AA) United Arab Emirates (AA/Aa2/AA) Qatar(AA−/Aa3/AA−) Qatar(AA−/Aa3/AA−) Saudi Arabia (A/A2/A) Saudi Arabia (A/A2/A) Oman(BB+/Ba1/BB+) Oman(BB+/Ba1/BB+) Kuwait(AA/Aa2/AA) Kuwait(AA/Aa2/AA) Iraq (B−/B3/B−) Iraq (B−/B3/B−) 0.00 0.20 0.40 0.60 0.80 0.00 0.20 0.40 0.60 0.80 1.00 Peer average Country score Peer average Country score Source: World Bank calculations using data from Peszko et al. (2020) and Fitch, Moody’s, and S&P. Notes: Sovereign credit ratings as available at the end of October 2021. a. See https://www.fitchratings.com/research/sovereigns/fitch-affirms-qatar-at-aa-outlook-stable-18-06-2021. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 45 THE R E LAT IO N S H I P B E T WE E N RIS KS/O PPO RTU NITIES REL ATED TO STR A N D ED AS S E TS A N D S OV EREIG N CRED ITWO RTHINES S We assess how the risks and opportunities related industries, (b) the carbon intensity of manufacturing, and to sovereign stranded assets are reflected in current (c) the size of accessible fossil fuel reserves. Resilience sovereign credit ratings. For this assessment, we use usually involves assessing institutional strength and different sovereign exposure and resilience scores calculated by perspectives of the viability of the country’s economic Peszko et al. (2020) for more than 90 countries (box 5.2). ecosystem, such as technological absorption, human capital, A country’s exposure to the low-carbon transition is typically wealth, and other factors. Box 5.2 provides more information measured by a combination of indicators such as (a) how on the construction of the sovereign exposures and resilience much the country’s exports are concentrated in fossil fuel scores used for our analysis. > > > BOX 5.2 - Sovereign Exposure and Resilience Scores Constructed by Peszko et al. (2020) The exposure score consists of four factors that measure how heavily a country is exposed to low- carbon transition risks. The first dimension of this measure is the timing of an exposure: (a) current exposure: carbon intensity of manufacturing exports and fossil fuel exports as a proportion of gross domestic product (GDP); (b) expected exposure: committed power emissions as a proportion of current annual power generation and expected resource rents as a proportion of GDP exposure. The second dimension is the type of exposed assets: (a) reliance on underground asset reserves: expected resource rents as a proportion of GDP and fossil fuel export as a proportion of GDP; (b) reliance on carbon-intensive built capital: carbon intensity of manufacturing exports and committed power emissions as a proportion of current annual power generation. The resilience score is a combination of 11 factors that measures a country’s ability and flexibility in adapting to structural changes caused by the low-carbon transition. Accordingly, the set of subfactors of this index is quite broad and includes five categories: (a) built, human, and institutional assets; (b) macroeconomic and financial flexibility; (c) economic performance and complexity; (d) business environment;a and (e) position on the global supply curve. Figure B5.2.1 contains the list of factors that are combined into one composite resilience score. FIGURE B5.2.1 - Overview of measurement factors of countries’ exposure and resilience to low-carbon transition Low-carbon transition Exposure Resilience Built, human, and Quality of infrastructure institutional assets Forward-looking Fossil fule rents Education Committed power emissions Institutional quality and governance Macroeconomic and Macroeconomic stability financial stability Financial market development Economic performance Adjusted net savings and complexity GDP per capita Economic complexity Current Fossil fule exports Technology absorption Business environment CO2 Intensity Ease of dong business Oil extraction costs Position on global supply curve Source: Peszko et al. 2020; World Bank’ infographics. Continue on next page 46 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Continued from previous page Figure B5.2.2 provides an overview of sovereign exposure and resilience scores. Countries such as Iraq and Libya are both highly exposed and least resilient to stranded-asset risks. Other Persian Gulf countries like Kuwait and Oman are also highly exposed; however, their resilience is assessed at moderate levels. On the other end, wealthier countries such as Singapore and Switzerland have high resilience and low exposure to stranded-asset risks. FIGURE B5.2.2 - Overview of low-carbon transition exposure and resilience of the countries most exposed 1.0 Iraq Libya Kazakhstan High Exposure Qatar Brunei Darussalam Kuwait Saudi Arabia Oman Nigeria Vietnam Azerbaijan Iran Equatorial Guinea Guyana Algeria Venezuela, RB Botswana Russian Federation Congo, Rep. South Africa Cambodia 0.5 United Arab Emirates China Ukraine Ghana Indonesia Norway Malaysia Bolivia India Angola Egypt Low Exposure Australia Argentina Bangladesh Mozambique Thailand Chile Colombia Korea, Rep. Canada Poland Namibia Malawi United States Turkey Cote d'Ivoire Germany Italy Brazil Pakistan Japan Mexico Mongolia Philippines Tanzania Sweden United France Uganda Kingdom Kenya 0.0 0.0 0.5 1.0 High Resilience Low Resilience Source: Peszko et al. 2020. Caption: Engineer at NOOR solar plant in Ouarzazate, Morocco | Source: Climate Investment Funds CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 47 Caption: Panels at NOOR solar plant, Ouarzazate, Morocco | Source: Climate Investment Funds We find that there is no empirically discernible relationship the communication by authorities and policy makers on how between current sovereign credit ratings and the level of countries are adapting and managing the transition to a low- a country’s exposure to a low-carbon transition (figure carbon economy will become increasingly important. 5.3).47 This finding is robust for all income levels. However, Currently, many CRAs appear to treat the stranded-asset any assessment of sovereign exposures also needs to risk inconsistently and their approach is focused on the consider institutional and economic resilience, which may typical time horizon of a sovereign credit assessment. For mitigate risks linked to higher exposure. We find that resilience countries with significant stranded-asset risk exposure, CRAs scores do display a high and statistically significant correlation frequently mention rating weaknesses around a dependency with sovereign credit ratings. However, this relationship is on fossil fuel. However, in most cases these weaknesses less robust for lower-income countries (0.84, 0.71, and 0.63 are outweighed by the short-term benefits of fossil fuels, for high-income, upper-middle income, and lower-middle particularly on the fiscal side. For example, the precedence of income countries, respectively). Moreover, resilience and rating actions taken by S&P when it downgraded Kuwait’s and exposure scores, if combined into a single joint preparedness Oman’s sovereign credit ratings and revised the rating outlook index, show a statistically significant correlation with credit for Bahrain because of the significant decline in oil prices ratings in the high-income group only. This finding indicates illustrates how credit factors that are treated as a strength in that sovereign credit ratings may not fully capture risks or the short term can quickly become the main cause for a rating opportunities related to stranded-asset risk for lower-income downgrade.48 sovereigns and that other factors, such as macroeconomic and debt metrics, which are perceived as having more materiality Finally, sovereign debt managers and other policy makers to credit, drive the credit rating assessment. should understand that the investment community, ESG Our findings highlight that risks and opportunities related providers, and CRAs are already forming opinions on a to stranded-asset risks may not be adequately reflected sovereign’s sustainability performance. As a result, issuers in current sovereign credit rating assessments. This is need to be proactive about providing information to these key especially the case for lower-income countries. The approach stakeholders on how the country is planning to transition to a to sovereign exposure and resilience that we present could low-carbon economy. Failure to do so will likely result in higher be relevant for CRAs in their sovereign assessments. We credit spreads and a higher cost of borrowing over time. expect that stranded-asset risks will become increasingly Recent World Bank publications provide advice to sovereign priced into financial markets over the coming years and that debt managers on this key area.49 47 Assessments have been conducted using the data for 23 lower-middle, 27 upper-middle, and 40 high-income countries. 48 https://www.spglobal.com/marketintelligence/en/news-insights/latest-news-headlines/s-p-downgrades-kuwait-oman-revises-outlook-on-bahrain-on-plummeting-oil- prices-57780806 49 See, for example, Boitreaud et al. 2020; Boitreaud et al. 2021; and Hussain 2020. 48 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS FIGURE 5.3 - The relationship between current sovereign credit ratings and the level of a country’s resilience and exposure to a low-carbon transition Resilience score Exposure score Resilience score Exposure score (higher score - higher resilience) (higher score - higher exposure) (higher score - higher resilience) (higher score - higher exposure) Upper middle income Low income Upper middle income Low income Upper middle income Low income Upper middle income Low income Lower middle income High income Lower middle income High income Lower middle income High income Lower middle income High income 1.0 1.0 1.0 1.0 0.9 0.9 0.9 0.9 0.8 0.8 0.8 0.8 0.7 0.7 0.7 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.5 0.5 0.4 0.4 0.4 0.4 0.3 0.3 0.3 0.3 0.2 0.2 Upper- 0.2 0.2 Upper- middle 0.1 0.1 middle 0.1 0.1 Income Income 0.0 0.0 0.0 0 5 10 15 20 0.0 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 Sovereign Rating Sovereign Rating Sovereign Rating Sovereign Rating (numerical scale; 1 - lowest, 21 - highest rating) (numerical scale; 1 - lowest, 21 - highest rating) (numerical scale; 1 - lowest, 21 - highest rating) (numerical scale; 1 - lowest, 21 - highest rating) Resilience score Exposure score Resilience score Exposure score (higher score - higher resilience) (higher score - higher exposure) (higher score - higher resilience) (higher score - higher exposure) Sub-Saharan Africa Latin America & Caribbean Sub-Saharan Africa Latin America & Caribbean Sub-Saharan Africa Latin America & Caribbean Sub-Saharan Africa Latin America & Caribbean South Asia Europe & Central Asia South Asia Europe & Central Asia South Asia North America Europe East Asia& Central Asia & Pacific South Asia North America Europe & Central East Asia Asia & Pacific North MiddleAmerica East & North Africa East Asia & Pacific North MiddleAmerica East & North Africa East Asia & Pacific Middle East & North Africa Middle East & North Africa 1 1 1 1 0.9 0.9 0.8 0.8 0.8 0.8 0.7 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.4 0.4 0.4 0.4 0.3 0.3 0.2 0.2 0.2 0.2 0.1 0.1 0 0 00 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 Sovereign Rating Sovereign Rating Sovereign Rating Sovereign Rating (numerical scale; 1 - lowest, 21 - highest rating) (numerical scale; 1 - lowest, 21 - highest rating) (numerical scale; 1 - lowest, 21 - highest rating) (numerical scale; 1 - lowest, 21 - highest rating) Source: World Bank calculations; Peszko et al. 2020; and CRAs. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 49 6. Caption: Marine line near Caye Caulker, Belize | Source: Unsplash.com 50 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS >>> Conclusions This paper sheds light on select issues related to how sovereign ESG factors affect sovereign credit ratings. We look at deeply embedded practices and methodologies that affect CRAs’ response to the market demand for mainstream sustainability in their services and products and provides new evidence to inform the industry’s evolution. Given the regulations and mandate for CRAs and prevailing constraints, both methodological and data related, on forward- looking analysis of ESG factors, the systematic integration of those factors into sovereign credit assessments is challenging. In recent years, CRAs have provided more clarity on their treatment of E, S, and G factors, but the many different CRA methodologies and new products are seen, by many, as not living up to the current sustainability challenge. Our analysis provides mixed evidence on the extent to which ESG factors are included in sovereign credit rating assessments. The CRA industry emphasizes that only factors that influence a sovereign’s creditworthiness over a particular time horizon (typically less than 10 years) are included in a credit rating assessment—and that this assessment should not be conflated with investment strategies that target social or environmental returns, in addition to a financial return. As a result, risks that are easier to understand and model are more explicitly included in credit assessments. Thus, G factors are included more often, but the explicit inclusion of E and S factors remains more limited. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 51 K EY AC T ION P O I N TS We suggest that there is scope to improve the current CRA approach by better reflecting ESG factors in sovereign credit rating methodologies. Materiality is a key concept in the discussion of the inclusion of credit- relevant ESG factors. Although on a conceptual level many ESG factors can affect a country’s sustainability profile, their financial materiality may be hard to measure for now and, therefore, they are currently not included in the overall credit assessment. Nevertheless, given that the ESG “lens” is now a structural tenet of the investment process, we suggest that credit rating agencies should also adapt their methodologies to align with this structural shift. Increasing transparency here will benefit the whole financial system. The CRAs and ESG provider mandates need to become clearer and distinguishable. Both industries have an important role to play. CRAs are focused on their medium-term credit assessments from a financial risk perspective. ESG providers need to focus on longer-term sustainability measurements from a sovereign ESG impact perspective—an area of increasing interest from investors. Furthermore, CRAs need to begin to plan for a longer-term credit measure that can complement the traditional CRA rating. CRAs could also consider adjusting their current rating scales toward greater granularity so that the financial materiality of ESG factors, as well as the financial materiality of other factors, could be better reflected on the credit rating scale. Currently, as the impact of the ingrained income bias dominates ESG factors, their signal is not “strong” enough to be regarded as being material from a credit perspective. Bringing more granularity to the rating scale, as well as clarity on what factors are driving differences, even if small, could provide more clarity to investors. As already mentioned, given the dynamic nature of materiality, some of these financially material factors may soon become credit relevant. World Bank research shows that wealth accounting data related to a country’s natural capital and information on a sovereign’s stranded-asset risks and opportunities represents additional ESG relevant information that could be better reflected in sovereign credit assessments. The wealth accounting approach is inherently forward-looking and captures determinants for long-term sustainable growth. Additionally transition risks and opportunities related to stranded assets are often not clearly taken into account in current sovereign CRA assessments- and CRAs could improve their analysis of these potentially relevant credit factors, going forward. Policy makers as well as sovereign debt managers have an important role to play in this changing ecosphere (Boitreaud et al. 2020). A sovereign’s debt management office should play a key role in engaging with investors on the sovereign’s ESG credentials and providing material information to market participants. This transparency will ultimately be valued by participants and allow a sovereign to explain how it is adapting to a more sustainable future. This approach is important for sovereigns that may be regarded by the market as lagging peers. 52 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS References Almeida, H., Cunha, I., Ferreira, M.A., Restrepo, F. (2017). The Real Hoegh-Guldberg, O., D. Jacob, M. Taylor, M. Bindi, S. Brown, Effects of Credit Ratings: The Sovereign Ceiling Channel. Journal of I. Camilloni, A. Diedhiou, R. Djalante, K. L. Ebi, F. Engelbrecht, Finance, vol. LXXII (1). J. Guiot, Y. Hijioka, S. Mehrotra, A. Payne, S. I. Seneviratne, A. Thomas, R. F. Warren, G. Zhou, and P. Tschakert. 2018. “Impacts of Bank of America. 2021. “ESG Matters Global. Sovereign ESG 1.5ºC Global Warming on Natural and Human Systems.” In Global Investing: You Can Do Good and Do Well.” BoA Securities Warming of 1.5°C: An IPCC Special Report on the Impacts Investment Strategy. December 1, 2021. of Global Warming of 1.5°C above Pre-industrial Levels and Related Global Greenhouse Gas Emission Pathways, in the Boitreaud, S., T. Emery, L. Gonzales, B. Gurhy, F. Larrain, and Context of Strengthening the Global Response to the Threat C. Paladines. 2021. “Paving the Path: Lessons from Chile’s of Climate Change, Sustainable Development, and Efforts to Experiences as a Sovereign Issuer for Sustainable Finance Action.” Eradicate Poverty. Geneva: The Intergovernmental Panel on EFI-Insight Finance. World Bank, Washington, DC. Climate Change. Retrieved August 6, 2021, https://www.ipcc.ch/site/ assets/uploads/sites/2/2019/02/SR15_Chapter3_Low_Res.pdf. Boitreaud, S., E. Gratcheva, B. Gurhy, C. Paladines, and A. Skarnulis. 2020. “Riding the Wave: Navigating the ESG Landscape Hussain, F. 2020. “Engaging with Investors on Environmental, for Sovereign Debt Managers.” EFI Insight-Finance. World Bank Social, and Governance (ESG) Issues: A World Bank Guide for Group, Washington, DC. Sovereign Debt Managers.” World Bank, Washington, DC. ESMA (European Securities and Markets Authority). 2019a. IEA (International Energy Agency). 2013. “Redrawing the Energy- “Guidelines on Disclosure Requirements Applicable to Climate Map.” World Energy Outlook Special Report, Paris, June Credit Ratings.” https://www.esma.europa.eu/file/51961/ 2013. https://www.iea.org/reports/redrawing-the-energy-climate- download?token=ePo3rnD0. map. ESMA (European Securities and Markets Authority). 2019b. “ESMA Lee, Allison Herren, commissioner, US Securities and Exchange Technical Advice to the European Commission on Sustainability Commission. 2021. “Living in a Material World: Myths and Considerations in the Credit Rating Market.” https://www.esma. Misconceptions about ‘Materiality.’” Keynote remarks at the 2021 europa.eu/file/51962/download?token=KLrr2xTA. ESG Disclosure Priorities Event Hosted by the American Institute of CPAs with the Chartered Institute of Management Accountants, Fitch Ratings. 2021. Climate Change ‘Stranded Assets’ Are a Sustainability Accounting Standards Board, and the Center for Audit Long-Term Risk for Some Sovereigns. https://www.fitchratings.com/ Quality, May 24. https://www.sec.gov/news/speech/lee-living- research/sovereigns/climate-change-stranded-assets-are-long-term- material-world-052421. risk-for-some-sovereigns-15-02-2021 Peszko, Grzegorz, Dominique van der Mensbrugghe, Alexander Gratcheva, E., T. Emery, and D. Wang. 2021. “Demystifying Golub, John Ward, Dimitri Zenghelis, Cor Marijs, Anne Schopp, Sovereign ESG.” EFI Insight-Finance. World Bank Group, John A. Rogers, and Amelia Midgley. 2019. “Beyond Stranded Washington, DC. Assets: Climate Strategies of Fossil Fuel–Dependent Countries.” Gratcheva, E., B. Gurhy, T. Emery, D. Wang, L. Oganes, J. K. Climate Change and Development. World Bank, Washington, DC. Linzie, L. Harvey, K. Marney, J. Murray, and R. Rink, R. 2021. “A S&P Global Ratings. 2018. “How Environmental, Social, and New Dawn—Rethinking Sovereign ESG. EFI Insight-Finance. World Governance Factors Help Shape the Ratings on Governments, Bank, Washington, DC, and J.P. Morgan, New York. Insurers, and Financial Institutions.” October 23. Gratcheva, E., and D. Wang. 2021. “Natural Allies: Wealth and S&P Global Ratings. 2021. “General Criteria: Environmental, Social, Sovereign ESG.” In The Changing Wealth of Nations 2021: and Governance Principles in Credit Ratings.” October 10. https:// Managing Assets for the Future. Washington, DC: World Bank www.spglobal.com/ratings/en/research/articles/211010-general- Group. criteria-environmental-social-and-governance-principles-in- Goldman Sachs. 2020. “Global Markets Analyst: An ESG Scorecard credit-ratings-12085396. for EM Sovereign Credit.” Goldman Sachs Global Investment World Bank. 2021. “The Changing Wealth of Nations 2021: Research. Managing Assets for the Future.” World Bank, Washington, DC. Heitor, Igor Cunha, Miguel A. Ferreira, and Felipe Restrepo. 2017. “The Real Effects of Credit Ratings: The Sovereign Ceiling Channel.” Journal of Finance 72 (1). CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 53 Additional Material Afonso, A., P. Gomes, and P. Rother. 2007. “What ‘Hides’ behind Erdem, O., and Y. Varli. 2014. “Understanding the Sovereign Credit Sovereign Debt Ratings?” ECB Working Paper Series, 711, Ratings Of Emerging Markets.” Emerging Markets Review 20 January. (September): 42–57. Afonso, A., D. Furceri, and P. Gomes. 2011. “Sovereign Credit ESRB. 2020. “Issues Note on Liquidity in the Corporate Bond and Ratings and Financial Markets Linkages Application to European Commercial Paper Markets, the Procyclical Impact of Downgrades Data.” ECB Working Papers Series 1347, June. and Implications for Asset Managers and Insurers.” May. Amstad, M., Packer, F. (2015). Sovereign ratings of advanced European Commission. 2015. “Study on the Feasibility of and emerging economies after the crisis. BIS Quarterly Review. Alternatives to Credit Ratings.” Final report. December 2015. Feyen, E., R. Utz, I-Z. Huertas, O. Bogdan, and J. Moon. 2019. Arslanalp, S., Drakopoulos, D., Goel, R., Koepke, R. (2020). “Macro-Financial Aspects of Climate Change.” World Bank Group Benchmark-Driven Investments in Emerging Market Bond Markets: Policy Research Paper 9109. Taking Stock. IMF Working Paper, WP/20/192. Fitch Ratings. 2017. “Fitch Ratings Approach to Capturing Baghai, Ramin and Becker, Bo and Pitschner, Stefan, The Private Environmental, Social and Governance Risk in Credit Ratings: Use of Credit Ratings: Evidence from Mutual Fund Investment Criteria Overview.” Special report. Mandates. (April 15, 2020). Swedish House of Finance Research Paper No. 18-13; European Corporate Governance Institute GSIA (Global Sustainable Investment Alliance). 2018. “Global - Finance Working Paper No. 612/2019. Available at SSRN: Sustainable Investment Review.” https://ssrn.com/abstract=3201006 or http://dx.doi.org/10.2139/ ssrn.3201006. FSB. 2020. “COVID-19 Pandemic: Financial Stability Implications and Policy Measures Taken.” Report submitted to the G20 Finance Bathia, A.V. 2002. “Sovereign Credit Ratings Methodology.” IMF Ministers and Governors, July 15. Working Paper, WP/02/170. IMF (International Monetary Fund). 2010. “The Uses and Abuses Berg, Florian, Julian Koelbel, and Roberto Rigobon. 2019. of Sovereign Credit Ratings.” Global Financial Stability Report. “Aggregate Confusion: The Divergence of ESG Ratings.” SSRN October. Electronic Journal. IMF (International Monetary Fund). 2016. “Analyzing and Managing BIS. 2017. “The Regulatory Treatment of Sovereign Exposures. Fiscal Risks--Best Practices.” June. https://www.imf.org/external/ Discussion Paper.” Issued for comment by March 9, 2018. np/pp/eng/2016/050416.pdf. Bolton, P., M. Despres, L. A. Pereira da Silva, F. Samama, and R. IOSCO (Board of the International Organization of Securities Svartzman. 2020. “The Green Swan: Central Banking and Financial Commissions). 2015. “Good Practices on Reducing Reliance on Stability in the Age of Climate Change.” Bank of International CRAs in Asset Management: Final Report.” FR08/2015. Settlements, Banque De France. Jaramillo, L., and M. M. Tejada. 2011. “Sovereign Credit Ratings Bouyé, Eric, and Diane Menville. 2020. “The Convergence of and Spreads in Emerging Markets.” Sovereign ESG Ratings.” February 28. Available at SSRN: https://ssrn.com/abstract=3568547 or http://dx.doi.org/10.2139/ Kiff, J., M. Kisser, and L. Schumacher. 2013. “Rating Through- ssrn.3568547 the-Cycle: What Does the Concept Imply for Rating Stability and Accuracy?” IMF Working Paper, WP/13/64. Brown, M., and A. Sienaert. 2019. “Governance Improvements and Sovereign Financing Costs in Developing Countries.” World Bank Kiff, J, S Nowak, and L Schumacher. 2012. “Are Rating Agencies Group Discussion Paper 14. Powerful? An Investigation into the Impact and Accuracy of Sovereign Ratings.” IMF Working Papers, 12/23. Cai, P., Q. Gan, and S-J. Kim. 2018. “Do Sovereign Credit Ratings Matter for Foreign Direct Investments?” Journal of International Luitel, P., and R. Vanpee. 2018. “How Do Sovereign Credit Ratings Financial Markets, Institutions and Money 55 (July): 50–64. Help to Financially Develop Low-Developed Countries?” European Capital Markets Institute Working Paper 8, November. 54 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Moody’s Investors Service. 2018. Environmental, Social and Pereira da Silva, L. A. 2020. “Green Swan 2—Climate Change and Governance Risks Influence Sovereign Ratings in Multiple Ways.” Covid-19: Reflections on Efficiency Versus Resilience.” Based on Sector In-Depth, June 27. remarks at the OECD Chief Economist Talk Series, Paris, April 23, and a Research Webinar at the BIS, May 13. Moody’s Investors Service. 2019. “General Principles for Assessing Environmental, Social and Governance Risks. Cross-Sector Rating Powell, A., and J. F. Martinez. 2008. “On Emerging Economy Methodology.” January 9. Sovereign Spreads and Ratings.” Research Department Publications 4565, Inter-American Development Bank. Moody’s Investors Service. 2020. “General Principles for Assessing Environmental, Social and Governance Risks: Proposed Raddatz, C., S. Schmukler, and T. Williams. 2017. “International Methodology Update. Request for Comment.” September 23. Asset Allocations and Capital Flows: The Benchmark Effect.” Journal of International Economics 108 (September): 413−30. Moor, L. de, P. Luitel, P. Sercu, and R. Vanpee. 2018. “Subjectivity in Sovereign Credit Ratings.” Journal of Banking and Finance 88: S&P Global Ratings. 2017. “How Does S&P Global Ratings 366–92. Incorporate Environmental, Social, and Governance Risks into Its Ratings Analysis.” November 21. Mrsnik, M., M. Kraemer, and A. Petrov. 2015. “The Heat Is On: How Climate Change Can Impact Sovereign Ratings.” S&P Global Ratings. 2019. “The Role of Environmental, Social, and Governance Credit Factors in Our Ratings Analysis.” September. Caption: CREDIT ESGview Aerial WORTHY: of UAE FACTORS | Source: CREDIT AND SOVEREIGN earth.google.com RATINGS 55 Appendix A Extended History of Credit Rating Agencies The Credit Rating Agency The Mercantile Agency perfected the approach of (CRA) industry has been a cen- Griffen, Cleaveland, and Campbell, with the agen- This account tral component of the financial cy consisting of lawyers from all over the coun- is based on architecture for more than 200 try who would report on local business entities.52 Daniel Cash, years and has continuously At this time, the main competitor was the Bradstreet Co., Sustainability adapted and evolved with the which pioneered the “Reference Book.” This book was sent Rating Agencies times. Although the first CRA was to subscribers and included “notification sheets” on changes vs Credit Rating founded in 1900 by John Moody, in credit risk assessments. These sheets brought their refer- Agencies: The credit reference agencies were ences up to date and were the original rating modifiers. This Battle to Serve a feature of the financial market- system moved the process toward more of a ranking. At the the Mainstream place even earlier, from the early same time, the two leading referencing entities faced a variety Investor (Palgrave 1800s.50 The earlier history of the of legal challenges in the United States against their right to Studies in Impact industry was sporadic, and the publish such information, with the common argument being Financing, industry proved agile, adapting that negative reports were libelous. Still, the courts predomi- London: Palgrave to societal, structural, and finan- nantly sided with the agencies and ruled that their reports con- Macmillan, 2021. cial system change. This adapt- stituted opinions protected under the US Constitution. ability has helped to sustain the Henry Varnum Poor published his History of Railroads industry. The first form of formally collating data on businesses and Canals in the United States in 1860 and in 1867, he to better understand their creditworthiness was in the United formed H.V. and H.W. Poor, which later evolved into Poor’s Kingdom, with Scottish and English mercantile societies form- Railroad Manual Company. In 1900, John Moody founded ing to provide members with protection against “swindlers and his first company, John Moody & Co, which published Moody’s sharpers.” However, as European nations began establishing Manual of Industrial and Miscellaneous Securities. As a result bankruptcy networks and associated protections, the nascent of the stock market downturn in 1903, he lost control of the American economy rejected such evolution. As global trade company and its name. grew, wealthy European banking entities faced the prospect of dealing with new customers. By the early 1800s, Baring Brothers surveyed the creditworthiness of American Houses FIGURE A.1 - Poor’s Manual and opportunities and ranked them on a scale of 1 to 11.51 In the United States, the first commercialized entity was the law firm of Griffen, Cleaveland, and Campbell in 1835. The firm developed a network of attorneys that produced reports on businesspeople. Subscribers could view the reports at the offices of the firm. However, appetite for credit risk analysis at this time was curtailed by a booming economy. By the late 1830s, sentiment toward credit risk had changed after a spike in bankruptcies. The spike led to the National Bankruptcy Act of 1841 to provide relief to the thousands of individuals who had gone bankrupt during the dramatic deflation that followed the crises of the late 1830s. Shortly after the enactment of this act, Lewis Tappan launched the Mercantile Agency, which became the first viable credit reporting service in the United Source: Amazon.com States. 50 Credit reference agencies are companies that collect and hold information that affects a subject’s creditworthiness, while not explicitly rating the subject. Credit rating agencies rate the issuer’s ability and willingness to repay debt. 51 Baring Brothers was a London merchant bank founded in 1763 by the Baring family to finance trade with the United States and India. 52 Reporters included Abraham Lincoln and Ulysses S. Grant before their presidencies. 56 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Moody reestablished an agency in 1907 under the Analyses In 1973, the Securities and Exchange Commission Publishing Co. The agency adopted a new approach by essentially began the era of forcing the rating agencies focusing on rating securities as well as companies. He upon the financial markets. This effort culminated with the launched Moody’s Analyses of Investments in 1910, and the registering of recognized rating agencies under the name of rating industry began to take shape as we know it. Following a Nationally Recognized Statistical Rating Organization. The the market crash of 1929, on May 27, 1933, legislators passed rating industry developed steadily from this point on. S&P the first major federal securities law—the 1933 Securities developed under the auspices of its parent company McGraw- Act—also referred to as the Truth in Securities Act, the Federal Hill. Warren Buffett’s Berkshire Hathaway bought Dun and Securities Act, or the 1933 Act. The act was seen as a reflection Bradstreet in 2000 and immediately spun the Moody’s brand of the existing market dynamic, because the reliance on the off. Following the failures of Enron in 2001 and WorldCom agencies, from the perspective of regulators and the judiciary, in 2002, regulatory scrutiny of the credit rating agencies began well before the 1933 Act. In 1941, Poor’s merged with intensified, and they became regulated for the first time in Standard Statistics, which had been established in 1906 as 2005, just before the Global Financial Crisis of 2007–08. the Standard Statistics Bureau. IN TEG RAT IO N O F ES G AS A S EPARATE CO M PO NENT INTO THE SOVERE IG N C R E DI T ME T H O D OLO GY Scope Ratings GmbH is currently the only CRA that environmental risks. The ESG pillar gets a 20 percent weight, includes ESG as a stand-alone factor in its sovereign with seven indicators representing this component (figure credit rating methodology. In October 2021, Scope published A.2). In addition to quantitative indicators, Scope’s sovereign a sovereign credit rating methodology.53 In general, Scope’s credit rating framework includes qualitative scorecards that sovereign credit ratings framework rests on five categories of span all five pillars. sovereign risk, one of which is directly tied to ESG: “Domestic Scope’s approach brings more transparency to the role that economic risk,” “Public finances risk,” “External economic risk,” ESG factors play in sovereign credit rating assessments. “Financial stability risk,” and “ESG risks.” The ESG credit risk Although many of the traditional indicators and qualitative component is established by expanding the previously used considerations in sovereign credit rating methodologies could “Institutional & political risk” pillar to account for social and already be labeled as ESG-related credit factors, the explicit FIGURE A.2 - Scope GmbH’s sovereign credit rating methodology Domestic Public Finance External Financial Stability ESG Risk Economic Risk Risk Economic Risk Risk 20% 35% 25% 10% 10% Environmental Social Governance Transition risks: Old-age depemdency ratio; CO2/GDP; WB Governance Indicators Income inequality; Natural disaster risks; (average of six indicators) Labor force participation. Resource risks. Source: Scope Sovereign Rating Methodology. 53 See https://www.scoperatings.com/ScopeRatingsApi/api/downloadmethodology?id=01508950-119c-4ab5-9182-54fffdc1003f. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 57 breakdown of these risks into an ESG sphere brings clarity to rating methodologies is not a goal in itself. However, there is the topic. Notwithstanding this improved clarity, the separation a need for a more robust understanding of which ESG factors of ESG as a discrete structural component of sovereign credit materially affect credit risk. WEALT H DATA The World Bank’s Wealth Accounting and the Valuation valuation of natural resources. A country’s fossil fuel wealth of Ecosystem Services (WAVES) partnership has created is calculated as the discounted value of future resource rents, a national wealth accounting framework that works to until this nonrenewable resource is depleted. Renewable incorporate the value of natural capital and human capital resources, such as forests or agricultural land, distinguish alongside traditional financial measures of national income themselves in that their discounting horizon depends on the to create a more holistic accounting of national wealth (figure rate of extraction versus replacement. For instance, forest A.3). The purpose is to move beyond traditional GDP and capital is a function of (inflation-adjusted) unit rents, production capture the interaction of economic activity, the environment, quantities, and the difference between deforestation and re-/ and the human capital of a country’s citizens. Human capital, afforestation rates. In principle, renewable resources can for instance, is calculated as the discounted expected lifetime produce rents in perpetuity. earnings of a population. A similar rationale applies to the FIGURE A.3 - Total wealth composition See the Changing Wealth of Nations, a World Bank report, for a database and an analysis of the world’s wealth accounts, covering 146 countries: https://www.worldbank.org/en/publication/changing-wealth-of-nations. Long-term Prosperity and Well-Being National Income/GDP Total Wealth Human Net Foreign Produced Capital Natural Capital Capital Assets Male/ Female Total Machinery Urban Energy/ Agricultural Protected and Assets Equipment Forests Land Minerals Land Areas Employed/ - Total Stryctures Self- Liabilities employed Source: World Bank staff illustration. Notes: Natural capital includes the value of natural resources and ecosystem services. Human capital is the present value of future earnings for the labor force. Produced capital is the value of machinery, equipment, and structures, as well as urban land. Total wealth captures the sum of the produced capital, natural capital, human capital, and net foreign assets. 58 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Appendix B Additional Tables TABLE B.1 - Sovereign credit ratings, ranks, and credit quality steps (CQS) CREDIT QUALITY MOODY’S S&P FITCH GRADE DESCRIPTION STEP RATING # RATING # RATING # Prime Aaa 1 AAA 1 AAA 1 Aa1 2 AA+ 2 AA+ 2 1 High grade Aa2 3 AA 3 AA 3 Aa3 4 AA- 4 AA- 4 A1 5 A+ 5 A+ 5 Investment 2 Upper medium grade A2 6 A 6 A 6 A3 7 A- 7 A- 7 Baa1 8 BBB+ 8 BBB+ 8 3 Lower medium grade Baa2 9 BBB 9 BBB 9 Baa3 10 BBB- 10 BBB- 10 Ba1 11 BB+ 11 BB+ 11 4 Speculative Ba2 12 BB 12 BB 12 Ba3 13 BB- 13 BB- 13 B1 14 B+ 14 B+ 14 5 Highly speculative B2 15 B 15 B 15 B3 16 B- 16 B- 16 Substantial risks Caa1 17 CCC+ 17 CCC+ 17 Non-investment Extremely speculative Caa2 18 CCC 18 CCC 18 Caa3 19 CCC- 19 CCC- 19 Default imminent Ca 20 CC 20 CC 20 6 Ca 20 C 21 C 21 C 21 D 22 DDD 22 In default C 21 D 22 DD 23 C 21 D 22 D 24 Source: World Bank staff calculations. CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS 59 TABLE B.2 - Ordered probit model results The subtables (a) and (b) report the underlying results for figures B3.1.1 and 4.4. Marginal effects are commuted for a linear-linear relationship for the hypothetical average country. Income FE stands for income group fixed effects (low, lower-middle, upper-middle, and high income). a. Sovereign ESG socres ENVIRONMENTAL SOCIAL GOVERNANCE MARGINAL EFF. COEF. STD.ERR. P-VALUE COEF. STD.ERR. P-VALUE COEF. STD.ERR. P-VALUE LiProb=cqs1 0.0017 (0.0011) 0.1320 0.0065 (0.0010) 0.0000 0.0053 (0.0008) 0.0000 LiProb=cqs2 0.0002 (0.0003) 0.4580 0.0008 (0.0009) 0.3260 0.0007 (0.0007) 0.3150 LiProb=cqs3 0.0002 (0.0002) 0.2830 0.0014 (0.0006) 0.0159 0.0012 (0.0006) 0.0262 LiProb=cqs4 0.0003 (0.0003) 0.3340 0.0010 (0.0008) 0.2020 0.0010 (0.0006) 0.1310 LiProb=cqs5 -0.0016 (0.0011) 0.1460 -0.0063 (0.0014) 0.0000 -0.0056 (0.0013) 0.0000 LiProb=cqs6 -0.0008 (0.0006) 0.1880 -0.0033 (0.0011) 0.0037 -0.0026 (0.0009) 0.0046 Observations 115 - - 115 - - 115 - - Income FE Yes - - Yes - - Yes - - b. Wealth accountig variables HUMAN CAPITAL PRODUCED CAPITAL NATURAL CAPITAL MARGINAL EFF. COEF. STD.ERR. P-VALUE COEF. STD.ERR. P-VALUE COEF. STD.ERR. P-VALUE LiProb=cqs1 0.0849 (0.0211) 0.0001 0.1190 (0.0168) 0.0000 0.0001 (0.0174) 0.9950 LiProb=cqs2 0.0096 (0.0121) 0.4290 0.0156 (0.0157) 0.3210 0.0000 (0.0020) 0.9950 LiProb=cqs3 0.0169 (0.0094) 0.0728 0.0253 (0.0122) 0.0381 0.0000 (0.0019) 0.9950 LiProb=cqs4 0.0134 (0.0108) 0.2160 0.0274 (0.0148) 0.0649 0.0000 (0.0027) 0.9950 LiProb=cqs5 -0.0814 (0.0243) 0.0008 -0.1130 (0.0268) 0.0000 -0.0001 (0.0163) 0.9950 LiProb=cqs6 -0.0433 (0.0177) 0.0145 -0.0740 (0.0257) 0.0040 -0.0000 (0.0077) 0.9950 Observations 115 - - 115 - - 115 - - Income FE Yes - - Yes - - Yes - - Source: World Bank staff calculations. 60 CREDIT WORTHY: ESG FACTORS AND SOVEREIGN CREDIT RATINGS Other insights into sustainable finance you may be interested in Riding the Wave: Navigating the ESG Landscape for Sovereign Debt Managers. by S. Boitreaud, E. Gratcheva, B. Gurhy, C. Paladines and A. Skarnulis Demystifying Sovereign ESG. by E. Gratcheva, T. Emery and D. Wang A New Dawn - Rethinking Sovereign ESG. by E. Gratcheva, B. Gurhy, T. Emery and D. Wang Credit Worthy: ESG Considerations in Sovereign Credit Ratings. by E. Gratcheva, B. Gurhy, F. Stewart, A. Skarnulis and D. Wang 1% Growth in Natural Capital: Why it Matters for Sovereign Bonds. by E. Gratcheva, B. Gurhy and D. Wang Natural Allies: Wealth and Sovereign ESG, in: The Changing Wealth of Nations 2021: Managing Assets for the Future. by E. Gratcheva and D. Wang Natural Capital and Sovereign Bonds. by D. Wang Paving the Path: Lessons from Chile’s Experiences as a Sovereign Issuer for Sustainable Finance Action. by S. Boitreaud; T. Emery; L. Gonzales; B. Gurhy; F. Larrain; C. Paladines Spatial Finance: Challenges and Opportunities in a Changing World by WWF and World Bank. Geospatial ESG – The emerging application of geospatial data for gaining 'environmental' insights on the asset, corporate and sovereign level. by WWF, World Bank and GC The Global Program on Sustainability (GPS) promotes the use of high-quality data and analysis on natural capital, ecosystem services, and sustainability to better inform decisions made by governments, the private sector, and financial institutions. Find out more on http://worldbank.org/gps