Final report on EU CLO credit ratings – risk of conflicts of interests relating to methodology changes

7 December 2023 ESMA80-189-33520 1 Final Report EU CLO credit ratings – risk of conflicts of interests relating to methodology changes

2 Table of Contents 1 Executive Summary ......................................................................................................3 2 Introduction ...................................................................................................................5 3 CLO market environment ..............................................................................................7 3.1 Overview of the CLO market ..................................................................................7 3.2 CLO market participants.........................................................................................9 4 CLO credit ratings business and the selection of a CRA..............................................11 5 CLO rating methodologies...........................................................................................12 5.1 The role of a CLO rating methodology..................................................................12 5.2 The changes to CLO rating methodologies ..........................................................12 5.3 Risks to the robustness of CLO rating methodologies ..........................................14 6 CLO rating practices....................................................................................................15 6.1 CLO rating analysts’ activities ..............................................................................15 6.2 Risks to the independence of CLO credit ratings..................................................17 7 Conclusion ..................................................................................................................18 7.1 Main supervisory concerns and risks....................................................................18 7.2 ESMA actions and next steps...............................................................................18 8 Annexes ......................................................................................................................20 A.Statistical information ..................................................................................................20 B. Overview of changes in regulatory framework ............................................................26 C. List of abbreviations ...................................................................................................27 D. List of references........................................................................................................28

3 1 Executive Summary Reasons for publication In 2020, ESMA published a Public Report on Credit Rating Agencies’ (CRAs) rating processes and methodology validation for collateralised loan obligations (CLOs). In this report ESMA noted, amongst other points, that CRAs mainly interact with CLO arrangers and managers in the rating process, whereas in other asset classes CRAs mostly interact with issuers. CRAs provide them with tools to simulate the expected ratings. As CLO arrangers and managers can identify which CRA may assign the best credit ratings for each CLO tranche, it is key that CRAs ensure the independence of their rating process from any commercial influence. After publication, ESMA continued to monitor CLO rating methodologies and changes to them. ESMA observed that in certain instances after the largest EU CRAs changed their CLO rating methodologies, there was significant market share fluctuation. At the end of 2021 ESMA opened a thematic investigation into the drivers of changes to CLO methodologies in recent years. All CRAs investigated have procedures in place to identify, manage and mitigate the risk and perception of conflict of interests, in particular from fee related discussions and negotiations. ESMA however observed the controls around other forms of commercial influence to be less developed. ESMA found that it is common practice for CRA analysts to participate in market outreach activities, including with arrangers (who have a key role in choosing a CRA for a transaction) and key investors. ESMA recognises the benefits of analytical outreach with industry. This outreach includes exchanges of views on CRAs’ rationale for their methodologies, and feedback on how well the methodology is understood by the market. ESMA found analysts involved in this outreach were frequently also involved in the development of changes to CLO methodologies. In practice, there is a risk that analytical outreach becomes commercial feedback depending on the context, and the type of information provided by market participants. Analytical outreach and the boundaries of that outreach therefore need to be carefully safeguarded and communicated to staff. In some instances, market outreach included the sharing of information that ESMA considers ‘non-analytical’. This included on occasion market participants sharing with CRAs feedback on the commercial perception of a methodology versus their peers’ methodologies. CRAs should have in place sufficient safeguards to ensure that changes to their methodologies are based on objective reasons. ESMA observed that CRA policies, procedures and their implementation had varying degrees of coverage, detail and guidance to staff to prevent non-analytical information from potentially influencing CLO methodology development for the period reviewed. In order to ensure that non-analytical information does

4 not influence the development of CLO methodologies and to address the risk of potential conflict of interests, ESMA has directed CRAs to make remediations where necessary. ESMA has given the CRAs under investigation the opportunity to comment on possible factual inaccuracies of this report ahead of its publication. Contents Section 2 of this report explains the reasons why ESMA opened a thematic investigation into CLO rating practices and its scope. Section 3 presents an overview of the CLO market during the investigated period. Section 4 shares ESMA’s findings with regard to the CLO ratings business and the potential drivers for the selection of a CRA to rate CLOs by market participants. Section 5 presents the role that the CLO methodology plays in the selection of a CRA and summarises the changes that CRAs made to their CLO methodologies in recent years. Section 6 presents ESMA’s findings on the organisational arrangements that CRAs have for analysts to participate in CLO market outreach. Section 7 provides a summary of ESMA’s main supervisory concerns and risks. Next Steps ESMA has communicated its concerns to all CRAs whose practices were reviewed and directed remediation where necessary. ESMA will continue to monitor the developments in CLO markets. It will also closely follow the changes and developments in CLO credit ratings, rating practices and rating methodologies.

5 2 Introduction

1. Securitisation plays a useful role in financial markets by converting non-tradable balance sheet exposures into tradable securities. The Capital Markets Union (CMU) Action Plan1 highlighted that securitisation could enhance the resilience of the financial system as securitisation instruments allow the transfer and dilution of risks across market participants and jurisdictions.
2. In recent years, there have been heightened regulatory concerns surrounding the CLO market. These concerns were driven by the complexity of these instruments, the increasing risk appetite in leveraged loan markets, as well as the covenant-lite trend which has eroded lenders’ protection and generated a concentration of leveraged loans at the lower end of the credit quality spectrum.2
3. Given these growing concerns, and CRAs’ central role of credit assessment in the CLO market, ESMA carried out a thematic review into CLOs and CRAs during 2019 and 2020. The review focused on the CLO credit rating process and methodologies in relation to the obligations set forth in the Credit Rating Agencies Regulation (CRAR). 3 The review included the largest EU credit rating agencies (CRAs) and concluded with a Public Report summarising ESMA’s findings on CRAs’ practices for the issuance and monitoring of CLO credit ratings (ESMA 2020 CLO Report). 4 A summary of these findings is listed in Box 1 below. Box 1: Thematic review of CRA CLO practices Main findings outlined in the ESMA 2020 CLO Report: ◊ The internal organisation of CRAs – the CLO rating process is segmented between the CLO analytical and the corporate analytical team. A smooth and ongoing exchange of information between internal teams is key to ensure a holistic assessment of CLO creditworthiness. CRAs should ensure the timely identification of all inherent risks to CLOs. ◊ The interactions with CLO issuers – CRAs mainly interact with CLO arrangers and managers in the rating process and CRAs provide them with tools to simulate the expected ratings. As CLO arrangers and managers can identify which CRA may assign the best credit ratings for each CLO tranche, it is key that CRAs ensure the independence of their rating process from any influence from their commercial teams and/or arrangers. ◊ Model/third party dependencies leading to potential operational risks – CLO credit ratings are mainly the outcome of models and other rating tools, formalised in applications developed by teams that are separated from the rating analysts. The dependency on rating models and data provided by third parties, and the high automation of processes, present operational risks that need to be monitored by CRAs to avoid potential errors in credit ratings. ◊ Rating methodologies, modelling risks and commercial influence – CLO rating methodologies are underpinned by assumptions and modelling approaches that can have an impact on credit ratings. It is key that CRAs provide transparency to market participants on the limitations of methodological approaches. In addition, CRAs should ensure that changes in CLO rating methodologies are not influenced by commercial interests. ◊ The thorough analysis of CLOs – it is key that CRAs continue to monitor market trends and perform a thorough analysis of all relevant developments in CLO contractual arrangements. 1 European Commission, A Capital Markets Union for people and businesses – New Action plan, 24 September 2020 2 ECB, Leveraged transactions – supervisory expectations regarding the design and functioning of risk appetite frameworks and high levels of risk taking, March 2022. 3 Regulation (EC) 1060/2009 of the European Parliament and of the Council of 16 September 2009 on credit rating agencies, Official Journal of the European Union (OJ L 302, 17.11.2009). 4 ESMA 2020 CLO Report of 3 May 2020 | ESMA80-189-6982

6 4. In the Public Report, ESMA stated that CRAs mainly interact with CLO arrangers and managers in the rating process, whereas in other asset classes CRAs mostly interact with issuers. CRAs also provide market participants with tools to forecast the expected credit ratings for different CLO tranches. This enables CLO arrangers and managers to identify which CRA may assign the best credit ratings for each CLO tranche. Given this, it is key that CRAs ensure the independence of their rating process from any influence from their commercial teams and/or arrangers. 5. The CRA Regulation requires that CRAs ensure that their rating methodologies are rigorous, systematic, continuous and subject to validation.5 In its Public Report, ESMA highlighted that commercial interests should not hamper: a) the analytical soundness of CLO rating methodologies, or b) the quality of credit ratings, via rating inflation to retain or attract clients. 6. ESMA committed in the Public Report to continue to monitor CLO rating practices. As part of this work, ESMA observed that in certain instances the largest EU CRAs experienced a significant change in their market coverage following changes to their CLO rating methodologies (Figure 1). 5 Article 8(3) of the CRA Regulation in corroboration with Commission Delegated Regulation (EU) No 447/2012 of 21 March 2012 laying down regulatory technical standards for the assessment of compliance of credit rating methodologies, Official Journal of the European Union (OJ L 140, 30.5.2012) and ESMA/2016/1575, Guidelines on the validation and review of Credit Rating Agencies’ methodologies, 23 March 2017. 14%21%27% 15%18% 14% 21% 20% 18% 18% 15% 21% 26% 25% 25% 22% 22% 23% 21% 18% 14% 6% 6% 7% 7% 10% 25% 20% 55% 67% 81% 83% 72% 70% 62% 65% 64% 65% 61% 66% 63% 60% 58% 51% 53% 56% 33% 28% 5% 8% 16% 40% 45% 53% 50% 48% 37% 25% 44% 23% 14% 34% 35% 46% 39% 39% 35% 43% 42% 41% 35% 34% 39% 38% 40% 69% 67% 88% 88% 85% 69% 59% 51% 48% 44% 51% 37% 0% 20% 40% 60% 80% 100% 120% 140% Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 2015 2016 2017 2018 2019 2020 2021 2022 Figure 1: CRA's market coverage based on new CLO issuance Fitch Moody's S&P Notes:

• The percentage reflects the number of ISINs rated by each CRA over the total pool of ISINs issued in the quarter;
• New CLOs include reset and refinancing transactions, hence in certain instances CRAs may have reported different ISINs for the same tranche;
• Data are updated to 2022 Q3 and include EU and endorsed credit ratings;
• The CRA flags refer to material methodological changes. Source: Refinitiv, ESMA. June 2019 Oct 2020 Sept 2021 Dec 2020 Dec 2021

7 7. In light of this observation, ESMA opened an investigation at the end of 2021 to assess changes to CLO rating methodologies and their rationale. Among other points, ESMA examined if the CRAs complied with the obligations to: a) ensure that the accuracy and independence of credit rating activities has not been impaired by business influences or constraints, or by any existing or potential conflicts of interests or business relationship, b) ensure that conflicts of interests are properly identified, managed and disclosed, c) establish a review function responsible for periodically reviewing its rating methodologies, models and key rating assumptions and any changes thereto as well as the appropriateness of those rating methodologies, models and key rating assumptions, and d) establish, maintain, enforce and document an effective internal control structure, preventing and mitigating possible conflicts of interests and ensuring the independence of credit ratings, rating analysts and rating teams. 8. ESMA’s investigation focused on: (i) EU and endorsed6 CLO credit ratings, and (ii) the three largest CRAs, accounting for the vast majority of CLO credit ratings, namely Fitch Ratings (Fitch), Moody’s Investors Service (Moody’s), and S&P Global Ratings (S&P). 9. As part of this investigation, ESMA reviewed extensive information from the three CRAs, including policies and procedures, reports and presentations to the entities’ boards, meeting logs, reports by control functions, emails between staff and other forms of communication exchanges. ESMA also held meetings with key staff from the three CRAs, as well as with other key participants in this market, including CLO arrangers, managers and investors. 3 CLO market environment 3.1 Overview of the CLO market 10. In the European Union (EU) securitisation has continued to play a limited role after the Global Financial Crisis (GFC) (see Figure A1 in Annex A). EU issuance peaked at over EUR 800 billion in 2008 but has averaged EUR 260 billion from 2009 to 2022. In 2021, only approximately EUR 230 billion was issued in the EU (see Figure 2 and Figure A5 in Annex A). 6 Endorsement is one of the options provided by the CRA Regulation for credit ratings issued by non-EU CRAs to be used for regulatory purposes in the EU. More information can be found on ESMA’s website: https://www.esma.europa.eu/supervision/non￾eu-credit-rating-agencies. Proportion of endorsed outstanding CLO ratings increased from 73% in 2017 to 87% in 2022

8 11. While EU securitisation issuance has broadly remained flat, collateralised loan obligations (CLOs) 7 have grown in importance in both the securitisation and leveraged finance (leveraged loans and high yield bond) markets. CLOs accounted for 19% of the 2021 European securitisation issuance compared to 9% in 2016 (see Figure A3 in Annex A). In the global leveraged loan market, CLOs have risen to approximately half of the global leveraged loan market. In the EU, leveraged finance issuance reached EUR 415 billion in 2021, with an increase of approximately 35% compared to 2020 (see Figure A4 in Annex A).8This growth came in the context of persistently low interest rates (see Figure A2 in Annex A), with a limited availability of investments with higher yields, investors’ preference for riskier exposures(due to higher returns), and the lower perceived risks and high credit ratings in the leveraged loan market (as is the case for many CLO tranches). 12. The majority of securitisation issuance has been in the US. Securitisation issuance in the US significantly decreased between 2007 (over EUR 2 trillion) and 2008 (below EUR 1 trillion) following the Global Financial Crisis (GFC). Since 2008, securitisation issuance has steadily grown; between 2008 and 2021, securitisation market issuance volume increased by 85%, peaking in 2021 with an issuance volume of over EUR 3.8 trillion and an average annual volume of EUR 1.9 trillion from 2009 to 2022. 13. In both EU and US markets, CLO market activity has undergone significant change, in particular the growing trend and market share of refinanced and reset legacy CLO transactions (see Figures A6 and A7 in Annex A).9 7 CLOs are securities backed by portfolios of loans to highly leveraged business that are typically rated in the non-investment grade category (leveraged loans). For further details on the CLO features, please see ESMA80-189-6982, Thematic Report – EU CLO credit ratings, an overview of Credit Rating Agencies practices and challenges, 13 May 2020 (hereinafter – ESMA 2020 CLO Report) as well as FSB, Vulnerabilities associated with leveraged loans and collateralised loan obligations, December 2019. 8 AFME, European High Yield and Leveraged Loan Report - Q2 2022, 27 September 2022. 9 Refinancing permits the holders of some CLO tranches to refinance them after the expiration of the ‘non-call’ period. This could be done with or without the consent of the collateral manager. Resetting implies a redemption and replacement of all outstanding debt tranches with debt tranches issued at current market spreads. Further, the resetting also implies an extension of the initial maturity profile, reinvestment period and weighted average life (WAL) test. -50 50 150 250 350 450 550 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 (YTD) Figure 2: US CLO vs. European CLO Total Issuance ($/€ bn) US Issuance European Issuance Note: All issuance types, per year. 2023 data up to 17 November. Source: J.P. Morgan Securities LLC, ESMA.

9 14. Despite its benefits, securitisation can also pose risks to financial stability. As seen during the GFC, securitisation can amplify the impact of a crisis through various channels: (i) loose underwriting standards (covenant-lite approach), which results in a deterioration in the credit quality of underlying assets and less signalling to lenders of this deterioration in credit quality 10 , (ii) complex and opaque products with associated model risk, which impairs investors’ ability to perform their own due diligence, and (iii) misaligned interests 11 between participants in a securitisation transaction, particularly originators, arrangers and investors.12 15. After the GFC, many regulatory reforms and initiatives were adopted, both at international and EU level. Their aim was to allow market participants to continue to benefit from securitisation while mitigating the risks that could arise from this market (see Annex B for an overview of reforms). 16. These regulatory reforms have shaped the decisions of CLO market participants in terms of portfolio composition, deal structure, and investment preferences, allowing for risk sharing across a more diverse set of market participants. 3.2 CLO market participants 17. Since 2016, the parties involved in a CLO deal have remained predominantly the same, that is originators, arrangers, issuers, managers and investors. 18. In recent years, there was higher competition for loans securitisation mandates across CLO arrangers. During the period 2016 – 2022, the concentration of CLO arrangers remained moderate in the US but higher in Europe; there were 16 European CLO arrangers (see Figure 3.1 and Table A2 in Annex A). During the same period, the level of concentration for CLO managers was lower than arrangers; however European CLO managers concentration was still higher than in the US (83 EU and 212 US CLO managers for the same period, see Figure 3.2 and Table A3 in Annex A). 10 ESMA 2020 CLO Report 11 There are many papers discussing the various interests of market participants in the securitisation industry, with a particular emphasis on the impact of misaligned interests and conflicts of interests during the GFC. For further reference see BIS, The Joint Forum – Report on asset securitisation incentives, July 2011. 12 ESRB, Monitoring systemic risks in the EU securitisation market, pages 16-17, July 2022. For example, the complexity and opaqueness of securitised products fuelled erroneous modelling assumptions, which implied the underestimating and under-pricing of risk among market participants. This weakness in the credit rating assessments coupled with inadequate transparency standards, contributed to dramatic imbalances in the financial markets, particularly in relation to the US mortgage-backed collateralised debt obligations (CDOs).

10 19. The CLO market remains a concentrated market from an investor perspective despite the additional involvement of financial institutions in recent years. As shown in Table 1, a small number of large global banks, with substantial cross-border activities, hold the largest share of senior and highly rated CLO tranches. 13 At the same time, asset managers and hedge funds focus on lower rated tranches with higher risk and higher yields. 13 FSB, Vulnerabilities associated with leveraged loans and collateralised loan obligations, December 2019 and ECB, Leveraged transactions – supervisory expectations regarding the design and functioning of risk appetite frameworks and high levels of risk taking, March 2022. 0 300 600 900 1200 1500 2016 2017 2018 2019 2020 2021 2022 2023 YTD Figure 3.1: Herfindahl-Hirschman index (HHI) for Arrangers of US and European CLO Issuances US Deals - Arrangers HHI European Deals - Arrangers HHI Note: 2023 data up to 24 July. All CLO deal issuances per year. The HHI is a market concentration metric that takes into account the issuance size distribution of CLO arrangers. Generally, markets in which the HHI is between 1,500 and 2,500 points are considered to be moderately concentrated. Source: J.P. Morgan Securities LLC, ESMA Staff calculations 0 300 600 900 1200 1500 2016 2017 2018 2019 2020 2021 2022 2023 YTD Figure 3.2: Herfindahl-Hirschman index (HHI) for Asset Managers of US and European CLO Issuances US Deals - Asset Managers HHI European Deals - Asset Managers HHI Note: 2023 data up to 24 July. All CLO deal issuances per year. Source: J.P. Morgan Securities LLC, ESMA Staff calculations Table 1: CLO Investor base breakdown (2022) Investor type AAA AA A BBB BB B Equity Banks 54% 10% 9% 3% 4% 4% 0% Pension funds 5% 5% 7% 5% 7% 9% 11% Insurance 2% 12% 15% 10% 1% 0% 0% Asset managers 38% 67% 62% 70% 65% 41% 64% Hedge funds 1% 6% 7% 13% 22% 46% 24% Source: Bank of America Securities, European SF Markets 2022-2023, 9 January 2023.

11 20. Both the EU and US CLO market shared the same top investors for the period reviewed (see Table 2). Regulatory changes and stricter risk retention rules have encouraged investors from other regions, such as the US and Japan, to become more active in terms of European CLO investments. 4 CLO credit ratings business and the selection of a CRA 21. Securitisation rating revenue in the CRAs investigated varied between (on average) 10% and 20% of CRA total rating revenues (see Table A1 in Annex A). In recent years, CLO revenues represent, on average, around 30% of each of the three CRAs revenues from rating structured finance instruments (see Figures A8.1 and A8.2 from Annex A). This revenue share has increased over time in Europe for all three CRAs, at times coinciding with changes in a CRA’s CLO rating methodology, or a renewed focus in market outreach by a CRA. 22. Securitisation products are different from traditional securities, as issuers are special purpose vehicles (SPV) and so related third parties to the issuers, namely arrangers and managers, play a significant role in the selection of a CRA. The role of arrangers and managers is to attract investors to the deal across all rated classes. This is particularly relevant in cases where these parties play a dual or triple role as manager, arranger and/or investor. Unlike in the issuance of other instruments, where the issuer typically has the most important role in the selection of a CRA, it is the CLO related third parties that choose the CRA that best meets their interests. 23. During the investigation, ESMA found in internal reporting of one CRA that related third parties select a CRA based on whether the CLO methodology is favourable to investors (in terms of the quality of the credit rating methodology, the CRA’s longstanding position in the specific market etc.), on whether the methodology allows for sufficient flexibility in the structuring and managing of a deal (as the CLO managers trade on the underlying loans during the lifetime of a deal), and offers rating stability. 24. Changes in the CLO market observed in recent years, in combination with the revisions to the regulatory framework (see in Annex B), have further driven market participants’ choice Table 2: Top CLO investors by CLO holdings (2019, $bn) Name Type Country CLO Holdings ($bn) Norinchukin Bank Japan 68 Wells Fargo Bank US 34.6 JP Morgan Bank US 20.5 Citibank Bank UK 18.1 Japan Post Bank Japan 10.6 Mitsubishi UFJ Trust (MUFG) Bank Japan 1.6 Total 153.4 Note: The information is reported as of 31 March 2019. S&P reports 18 US banks investing in CLOs with collective holdings of around 90 $bn (the top two US banks are reported in the above table). Source: S&P, The $700B in US CLOs: Who holds them, what risk they pose, 21 June 2019.

12 of CRAs to rate CLO deals. Specifically, in the European area, the CRA choice has been influenced by the 2016 entrance of two large non-EU investors, which pursued certain investment mandates and had specific CRA preferences. 5 CLO rating methodologies 25. The three CRAs employ global CLO methodologies. In the 2020 CLO Report, ESMA flagged that this approach allows CLO credit ratings to be compared irrespective of their geographical location, addressing the fact that some assets are originated in different jurisdictions. The main local differences are captured through the credit ratings of the underlying corporate loans, and not by the CLO rating methodologies. 5.1 The role of a CLO rating methodology 26. CRAs secure securitisation mandates on an issuance-by-issuance basis. This makes it a more competitive market, and more so in the EU because there are a limited number of arrangers and managers structuring securitisations and sourcing loans. A CRA’s relationship with these participants is therefore key for a CRA being consistently given new CLO mandates. 27. CLO credit ratings rely heavily on quantitative factors (such as statistical models), and less on qualitative factors derived from expert judgment and specific to each transaction.14 A CRA’s rating methodology will outline and explain the use of these quantitative factors in the determination of a credit rating. CRAs may also provide market participants with tools to forecast the expected credit ratings for different CLO tranches. Given that the rating methodology plays a fundamental role in a CRA’s selection, changes to rating methodologies can have a substantial impact on a CRA’s market share. 28. ESMA observed that changes to CLO rating methodologies in recent years have been generally followed by fluctuations in the respective CRAs’ market share (see Figure 1). 5.2 The changes to CLO rating methodologies Description of the changes to the CLO rating methodologies of CRAs 29. During its investigation, ESMA assessed why and how the three CRAs changed their methodologies during the sample period. 30. S&P recalibrated the CLO scenario and breakeven default rates in 2019, which was the first recalibration since 2009. This change involved analysing the relationship between corporate defaults and macroeconomic variables by adding ten years of additional data from between 2009 and 2019. S&P also incorporated an additional ten years of performance data for CLOs, which covered the performance of nearly all CLOs issued in or before 2009 (“generation 1.0 CLOs”) and allowed for the analysis of the entire life cycle of these 14 For more details see ESMA 2020 CLO Report.

13 securitisations. The additional data also provided insights into the performance of CLOs issued in 2010 or later (“generation 2.0 CLOs”). S&P further updated its CLO methodology to revise the “archetypal” pool of assets used in modelling to more closely approximate the makeup of actual pools of assets securitized into CLOs. 31. One impact of these revisions was to reduce the CLO par-subordination level, and in turn allow a cushion for CLOs to raise the embedded credit leverage at the same rating levels. An additional cushion can: (i) be used to create more rating cushion in the CLO structure to protect against future downgrades, (ii) allow a CLO to include riskier underlying assets, or (iii) decrease CLO par-subordination and increase leverage. The new historical data included in the recalibration was from 2009 to 2019, a relatively benign environment in the leveraged loan market, as well as a period with low interest rates. The inclusion of this period in the data set offered no additional insights into the likely performance of a CLO in a stress scenario. 32. Fitch made several changes to its CLO rating methodology during 2017-2021. The changes considered material by Fitch were published in 2020 and 2021. The changes in 2020 were mostly related to recovery assumptions. In 2020, during the initial stages of the Covid-19 pandemic, Fitch lowered its applicable recovery assumption for calculating its Weighted Average Recovery Rate by 5% if no recovery rating or recovery estimate was available from the Fitch corporate rating team. The Fitch methodology change in 2020 was considered to be more conservative following the initial stress added due to the Covid-19 pandemic. In 2021, similar to S&P, Fitch included an update to its base-case probability of default assumptions, default timing update and portfolio risk horizon. This resulted in Fitch lowering the rating default rate scenarios for its base calibration in the CLO model but maintaining the rating default rate scenario for high investment grade stresses. This change had a positive rating impact for sub investment grade and some investment grade rating levels. 33. Moody’s made a change to its CLO rating methodology in 2020 following a study to assess the impact of credit watches and outlooks on the underlying leveraged loans ratings. This study considered the impact of rating changes in the underlying leveraged loans, and not to CLO ratings directly. The review resulted in Moody’s changing its approach when assessing an obligor’s default probability rating for CLOs. Before the 2020 change, Moody’s would (i) adjust down the obligor rating by one notch if the obligor had been assigned a negative outlook, (ii) adjust down the obligor rating by two notches, if the obligor was on review for a potential downgrade. This approach changed in 2020 so that where an obligor was on review for a potential downgrade, the obligor rating was reduced by one notch rather than two. Moody’s kept the practice of adjusting up by one notch if the obligor was on review for possible upgrade. 34. Moody’s also made a change in 2021 to its approach to modelling a reinvesting CLO’s weighted average life (WAL). Moody’s previously modelled the CLO WAL as the covenant maximum of the transaction with amortisation distributed evenly over 2.5 and 5 years. This was changed to using the longer of the WAL covenant minus one year or the portfolio WAL plus one year capped by the covenant, with amortisation that is evenly distributed over a 2.5-year period. This change to the WAL by Moody’s in 2021 effectively simplified the

14 process for calculating the WAL and affected a smaller number of ratings than the change in 2020. Implications of the changes to the CLO rating methodology 35. Most of the changes to the CLO rating methodologies of the three CRAs reviewed as part of ESMA’s investigation, resulted in increases in rating cushion that gave more flexibility to CLO managers in the transaction structure to satisfy all types of investors: (i) to decrease CLO par-subordination and increase leverage for risk-seeking investors, and (ii) to increase cushion in CLO structures for certain investors, at a time when corporate leverage and loan covenants had deteriorated. 5.3 Risks to the robustness of CLO rating methodologies 36. The changes to the CLO methodologies that gave more flexibility to CLO managers coincided with several changes in the CLO market, such as those related to CLO structures and the loosening of the underwriting standards related to the underlying loans. 37. ESMA also notes that CRAs have included additional data in their models since 2009 that predominantly includes benign economic and financial conditions. This additional data does not represent a full credit cycle and includes observations from a period when interest rates were historically low and corporate defaults were limited.15 This gives rise to the risk that the additional data may artificially reduce the perceived underlying risk for CLOs. Since the credit ratings of CLO tranches are a key parameter for CLO investors, the potential for sudden CLO rating downgrades represents a risk for markets and investors. 38. ESMA also notes that there are inherent differences between CLOs issued before and after the Global Financial Crisis of 2008. Under current CLO structures (CLO 2.0), equity tranches are thicker which implies a higher required level of default for AAA tranches to be affected. CLOs 2.0 have tighter collateral eligibility requirements (including on the place of issue of leveraged loans) and shorter reinvestment periods, which reduce interest rate risk for investors. Finally, the risk retention rule set out in EU regulations after the crisis requires that the originator of the CLO retains at least 5% of the risk of the exposure on its balance sheet, in contrast to the originate-and-distribute model used before.17 CRAs have adjusted the archetypal pools used for modern CLO ratings to reflect these changes. 39. The current environment of higher interest rates has significantly changed from the historically low interest rates observed from 2009 to 2021. ESMA has raised with CRAs that, where additional data from after the Global Financial Crisis of 2008 is used as part of methodology development, CRAs should demonstrate that the total data used sufficiently reflects market stresses, so that a shift in the credit cycle and market conditions (e.g., higher interest rates) does not present a challenge to the continuous application of the current CLO methodologies. 15 ESMA50-165-883, ESMA Reports on Trends, Risks and Vulnerabilities, No.2, 10 September 2019 and ESMA50-165-1287, ESMA Reports on Trends, Risks and Vulnerabilities, No.2, 2 September 2020.

15 6 CLO rating practices 6.1 CLO rating analysts’ activities 40. ESMA has observed that rating analysts can typically be involved in three main activities: (i) credit rating activities, (ii) market outreach, and (iii) the initiation and development of the rating methodology. Rating analysts and credit rating activities 41. Rating analysts’ primary activity involves the issuance, approval and monitoring of credit ratings (see ESMA 2020 Public Report for more details). During its investigation, ESMA observed that in some CRAs, during the preparatory phase of a credit rating, rating analysts exchange with: (i) the CLO arrangers and managers in relation to the proposed structuring of a CLO deal and the issuance of the expected credit rating16 and the pre-sale report, and (ii) the CLO investors in relation to the issuance of the preliminary credit ratings. A first rating committee17 is typically convened to assign the expected credit rating. 42. After receiving all relevant and updated information, rating analysts perform analysis to assign the proposed final credit ratings on different tranches and call a rating committee (where they participate and sometimes chair). When the CLO portfolio is fully constituted, the rating analysts check its composition. Finally, the rating analysts monitor the performance of the CLOs and the compliance with the covenants as part of their surveillance activities. Rating analysts and market outreach activities 43. ESMA noted that CRAs have developed market outreach activities with the purpose of ensuring analysts have relevant and up-to-date knowledge of the products they rate, and so that CRAs can educate market participants on their methodologies and processes. The CRAs have established specific teams or programs for: (i) providing research and educational support to CLO market participants, (ii) receiving feedback about the quality of the CRA’s services and methodologies, (iii) monitoring market trends, (iv) ensuring transparency of CRA's rating practices and methodology, and (v) building investor relationships and stimulating demand for CRAs ratings. 44. From the information received in its investigation, ESMA noted that, at certain CRAs, rating analysts, typically senior lead analysts, participated in CLO market outreach activities on occasion with commercial staff and engaged with key CLO arrangers, managers and investors. 45. ESMA has further observed that CRAs have set up frameworks for market outreach, with varying degrees of formalisation around the ways analytical and commercial staff may 16 Different CRAs have different nomenclature for this point in a public rating phase. This point in the rating phase should not be confused with a ‘preliminary rating’, which is a distinct and separate product sold by some CRAs. 17 A Rating Committee is a committee constituted by the CRA to formally assign a credit rating. Generally, it is formed of senior analytical staff members of the CRA.

16 cooperate and interact with market participants. The leading practice from one CRA in this area was policies and procedures that outlined the limitations to analysts’ involvement in market outreach, including scenarios where analysts would have to recuse themselves from a meeting. ESMA observed that CRA policies, procedures and their implementation had varying degrees of coverage, detail and guidance to staff. 46. ESMA also observed that CRAs have set up different channels to record and exchange information between analytical and commercial staff, where commercial staff and, in at least one CRA, also analysts, keep records of the feedback received from various market participants. 47. In certain instances, CLO investors, arrangers and managers also shared their views on: (i) how they select CRAs and their needs, (ii) the commercial perception of a methodology versus other CRAs’ methodologies, and (iii) the relative strengths and weaknesses of CRAs’ rating practices compared to competitors. In certain instances, ESMA observed that at CLO outreach meetings where rating analysts participated, market participants explained to the CRA the reason for lost mandates. In certain cases, information on rating mandates and competitive disadvantages was also distributed to rating analysts via internal communication exchanges. 48. ESMA observed that the participation of rating analysts in market outreach activities related to CLOs increased, to varying degrees, during the preparation of draft changes to CLO methodologies, the consultation period, and the publication of new CLO methodologies. Rating analysts and rating methodology activities 49. In addition to their primary credit rating activities, ESMA observed that rating analysts have on occasion been part of the teams involved in initiating and driving the development of the CLO rating methodology changes. 50. These teams were predominantly formed by analytical staff (rating analysts and methodology development experts). In one CRA the review function staff performed their role in parallel to the CLO rating methodology development team as part of a wider project team. Among other controls and procedures in place for the development of rating methodologies, these methodology development teams submitted their proposal for material changes to the CLO rating methodology for: (i) validation and approval by the review function, (ii) pre-approval by a specific committee, and (iii) approval of the material changes by the relevant Board of Directors. The role of rating analysts and performance objectives 51. During its investigation, ESMA observed that, to varying degrees, the rating analysts of the three CRAs were involved in all three activities outlined above. In one CRA there was an increase in the responsibilities of analysts in knowledge sharing and outreach to CLO market participants. Given the breadth of analysts’ role, it is important that CRAs ensure that there are appropriate safeguards in place to prevent analysts receiving non-analytical information. Without appropriate safeguards, non-analytical information can influence both the rating process and methodology development.

17 52. ESMA observed that the CRAs have staff performance assessment frameworks that link the objectives of analytical staff to all three activities performed by rating analysts, including market outreach and rating methodology development. 6.2 Risks to the independence of CLO credit ratings 53. Unlike the issuance of traditional corporate bonds where the issuer plays the most important role in the selection of a CRA, CLO related third parties (e.g., arrangers, managers) and investors, play a key role in the choice of CRAs. 54. Although CLO market participants (arrangers, managers, investors) may have divergent interests, their role in the CLO market makes them particularly relevant. Analyst interactions with CLO market participants without appropriate controls therefore exposes CRAs to a heightened risk of potential conflict of interests. This creates the risk that some market participants may try to influence CRAs’ analytical and methodological decisions. Analyst interactions with CLO market participants without appropriate controls therefore exposes CRAs to a heightened risk of potential conflict of interests. 55. ESMA notes that market outreach facilitates transparency towards market participants regarding CRAs’ rating process and methodology. It also allows CRAs to gather views from market participants on the quality of the services they provide, including the analytical quality of the credit ratings they issue and the rating methodologies they use. 56. However, in ESMA’s view, there is a risk that analytical outreach becomes commercial feedback depending on the context and type of information provided by market participants. Analytical outreach and the boundaries of that outreach therefore need to be carefully safeguarded and communicated to staff. The CRAs’ current market outreach arrangements may result in rating analysts receiving information that could pose a risk to their independence and objective judgement. This information could influence the core analytical activities that rating analysts perform, such as credit rating activities, and the initiation and development of rating methodologies. Moreover, this risk of non-independence may be heightened where performance assessment frameworks link the annual evaluation and subsequent remuneration of rating analysts to broad assessment metrics of market outreach activities (for example, the number of meetings held) that do not assess the content and substance of analytical engagement. 57. It is key that CRAs have strong internal control arrangements to ensure that the credit rating process and their rating methodologies are not unduly influenced by the interests of different CLO market participants. CRAs’ controls should reflect all potential sources of influence including from CLO arrangers, managers and investors, who are essential in the choice of a CRA to rate a securitisation instrument in concentrated markets such as CLOs. CRAs should also consider the potential perception of conflict of interest risk in certain arrangements and business practices for market outreach. 58. Potential conflicts of interests may lead to rating shopping, where the CLO arranger or manager may select the most favourable credit rating for each CLO tranche depending on their interests.

18 7 Conclusion 7.1 Main supervisory concerns and risks 59. Securitisation rating revenue in the CRAs investigated varied between on average 10% and 20% of CRA total rating revenues. CLOs make up on average one third of the CRAs’ securitisation rating revenue. CLOs are sophisticated securitisation products and given their complexity and opacity, CLO market participants, and in particular investors, continue to use credit ratings to support their decision making. 60. ESMA recognises market outreach is an important activity for CRAs and particularly important in an asset class such as CLOs. ESMA also recognises that there are benefits to the market from analytical market outreach. It gives an opportunity for market participants to better understand a CRA’s approach to ratings and methodologies, and for analysts to discuss market trends and share their thoughts on related research. 61. However, ESMA observed that CLO rating analysts who have participated in market outreach activities may have been exposed to non-analytical information, such as the preferences of CLO related third parties in relation to credit ratings and rating methodologies, and the commercial perception of their methodologies. 62. Non-analytical information could provide rating analysts, involved in the initiation and development of changes to a CLO rating methodology and rating process, with commercial information that results in a perception of or actual conflict of interests. This could impair the accuracy, objectivity and independence of CLO credit ratings. 63. Credit ratings continue to be essential for all CLO market participants. For this reason, it is crucial that CRAs have sound controls over their market outreach activity to ensure the accuracy, objectivity and independence of credit ratings, as well as the independent development of rating methodologies. ESMA found in its investigation that CRAs’ policies, procedures and their implementation had varying degrees of coverage, detail and guidance to prevent non-analytical information from potentially influencing the CLO methodology development and rating process. In order to ensure that non-analytical information does not influence the development of CLO methodologies and to address the risk of potential conflict of interests, ESMA has directed CRAs to make remediations where necessary. 7.2 ESMA actions and next steps 64. Following the investigation, ESMA reminded CRAs’ of the critical importance of ensuring the independence of rating activities. ESMA has communicated its observations to each

19 CRA and will direct individual remedial action plans to ensure appropriate safeguards and controls are in place. 18 65. ESMA will continue to monitor the developments in CLO markets. This will include closely following the changes in CLO credit ratings, rating practices and rating methodologies. 18 As of the date of this document ESMA has not determined whether any of the observations made in this report constitute serious indications of the possible existence of facts liable to constitute one or more infringements of the CRA Regulation.

20 8 Annexes A. Statistical information 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 YtD Figure A1: US vs. European Securitisation Issuance (€ bn) US issuance (€bn) European issuance (€bn) Note: All asset classes included (ABS, CDO/CLO, RMBS, CMBS, SME) 2023 data up to Q2. Source: AFME, ESMA.

21 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 Figure A2: ECB and US FED Policy Rates ECB US FED Note: ECB key policy rate (variable rate tender (minimum bid rate) 01/01/07 to 14/10/07, fixed rate tenders (fixed rate) 15/10/08 onwards) in percent. Data up to 1 August 2023. Source: ECB, Board of Governors of the Federal Reserve System (US), ESMA. % 0 50 100 150 200 250 300 2016 2017 2018 2019 2020 2021 2022 2023 ABS CDO/CLO CMBS RMBS SME Corporate Note: Data up to Q1 2023. Source: AFME, ESMA.

22 0 10 20 30 40 50 60 Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May Jul Sep Nov Jan Mar May 2020 2021 2022 2023 A4: Evolution of European leveraged finance issuance (€bn) Leveraged Loan (€bn) High Yield Bonds (€bn) Note: Data up to May 2023. Source: AFME, ESMA 0 50 100 150 200 250 300 350 400

20,000 40,000 60,000 80,000 100,000 120,000 140,000 160,000 180,000 200,000 2016 2017 2018 2019 2020 2021 2022 2023 (YTD) Figure A5: US vs. European New CLO Issuance Amount and Deal Count US Issuance Amount ($mm) Europe Issuance Amount (€mm) US Deal Count (right axis) Europe Deal Count (right axis) Note: New Issuance Amounts only (Refi/Repricing/Reset not included). 2023 data up to 20 November. Source: J.P. Morgan Securities LLC, ESMA.

23

50,000 100,000 150,000 200,000 250,000 300,000 350,000 400,000 450,000 2016 2017 2018 2019 2020 2021 2022 2023 (YTD) Figure A6: US CLO Market Activity ($mm) New Issuance ($mm) Refinance ($mm) Reset ($mm) Reissue ($mm) Note: Data up to 20 November 2023. Source: J.P. Morgan Securities LLC, ESMA.

10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000 2016 2017 2018 2019 2020 2021 2022 2023 (YTD) Figure A7: European CLO Market (€mm) New Issuance (€mm) Refinance (€mm) Reset (€mm) Reissue (€mm) Note: Data up to 20 November 2023. Source: J.P. Morgan Securities LLC, ESMA. 28% 33% 30% 38% 34% 20% 41% 28% 39% 32% 28% 29% 72% 67% 70% 62% 66% 80% 59% 72% 61% 68% 72% 71% 0% 20% 40% 60% 80% 100% 2017 2018 2019 2020 2019 2020 2021 2017 2018 2019 2020 2021 CRA1 CRA2 CRA3 Figure A.8.1: Global Structured Finance Revenues (%) Global CLO revenue (%) Global non-CLO revenue (%) Note: CRA3 does not include revenues from covered bonds. Source: CRAs, ESMA. 35% 26% 28% 35% 37% 20% 48% 24% 38% 38% 35% 45% 65% 74% 72% 65% 63% 80% 52% 76% 62% 62% 65% 55% 0% 20% 40% 60% 80% 100% 2017 2018 2019 2020 2019 2020 2021 2017 2018 2019 2020 2021 CRA1 CRA2 CRA3 Figure A.8.2: EMEA Structured Finance Revenues (%) EMEA CLO revenue (%) EMEA non-CLO revenue (%)

24 Table A1: CRA’s EU credit rating revenues (€mm or %) 2018 2019 2020 2021* Moody’s total EU credit rating revenues 443 465 492 413 Moody’s total EU SF revenues 114 102 95 77 Moody’s total EU SF revenues/total EU credit rating revenues 26% 22% 19% 19% Fitch total EU credit rating revenues 263 311 288 141 Fitch total EU SF revenues 54 55 46 17 Fitch total EU SF revenues/total EU credit rating revenues 21% 18% 16% 12% S&P total EU credit rating revenues 525 576 643 635 S&P total EU SF revenues 54 56 65 75 S&P total EU SF revenues/total EU credit rating revenues 10% 10% 10% 12%

• For 2021, Moody’s and Fitch include EU-only revenues, while S&P includes UK revenues since SPGUK was still a branch of SPGRE Table A2: Concentration of EU CLO arrangers and their respective CLO issuance volume (2016-2023) List of arrangers Total Issuance (€mm) % of Total 2016-2023 Issuances Citigroup 48,350 € 15% Barclays 43,875 € 14% Bank of America 33,140 € 11% Morgan Stanley 29,255 € 9% BNP Paribas 28,530 € 9% JPMorgan 25,080 € 8% Credit Suisse 24,650 € 8% Goldman Sachs 22,700 € 7% Deutsche Bank 21,730 € 7% Jefferies & Co. 21,450 € 7% Natixis 10,635 € 3% GreensLedge Capital Markets 2,335 € 1% Stifel Nicolaus 865 € 0% Royal Bank of Canada 415 € 0% Mitsubishi UFJ Securities 410 € 0% Royal Bank of Scotland 380 € 0% Total 313,800 € 100.0% Note: 2023 data up to 24 July. All CLO deal issuances per year. Source: J.P. Morgan Securities LLC, ESMA Staff calculations.

25 Table A3: Concentration of EU CLO managers and the respective CLO issuance volume (2016- 2023) Asset Managers Total Issuance (€mm) % of Total 2016-2023 Issuances GSO/Blackstone 19,840 € 6.3% CVC 16,275 € 5.2% Carlyle 14,670 € 4.7% PGIM 14,000 € 4.5% Investcorp 12,295 € 3.9% KKR 11,335 € 3.6% CSAM 10,065 € 3.2% ICG 9,855 € 3.1% Apollo / Redding Ridge 9,100 € 2.9% Alcentra 8,730 € 2.8% Blackrock 8,270 € 2.6% Barings 8,205 € 2.6% Ares 6,725 € 2.1% Cairn 6,430 € 2.0% Partners Group 6,395 € 2.0% Total top 15 Asset Managers 162,190 € 51.7% All Other 68 Asset Managers 151,615 € 48.3% Total 83 Asset Managers 313,800 € 100.0%

26 B. Overview of changes in regulatory framework In the post GFC period, there were several changes in the regulatory framework, including credit risk retention rules19 across various jurisdictions, most notably in the US, EU and Japan. These legislative initiatives apply to CLOs and other types of structured finance products and aim to align the incentives of the original portfolio underwriters with the end investors. ◊ US: The risk retention rules became effective for CLOs in December 2016 (among other asset classes), when the CLO managers of both “balance sheet” and “open-market” CLOs were required to continuously hold 5% of their CLO issue.20 These changes were subsequently challenged at the US Court of Appeal for the District of Columbia Circuit, which ruled, in February 2018, that CLO managers of “open-market” CLOs are not securitisation sponsors and therefore, not subject to these risk retention rules (however, the requirements still apply for “balance sheet” CLOs).21 The risk retention rules apply directly to CLO securitisation sponsors and not investors; however, many of the CLO investors in the US are regulated entities which are subject to various requirements. Further, US prudential requirements prevent banks from using external credit ratings, requiring the use of supervisory formula approaches that rely on standardised or modelled inputs (regulators apply a risk weight floor of 20% or higher). ◊ EU: The regulatory framework was strengthened to address the threats to financial stability that the GFC exposed, including the risks attached to the ‘originate to distribute’ model.22 First, the CRA Regulation introduced a common approach for regulating and supervising CRAs to ensure credit ratings quality and independence as well as investor protection. Further, it required structured finance transactions, including CLOs, to be rated by at least two CRAs. Second, the 2013 Capital Requirements Regulation (CRR) 23 imposes prudential requirements, including a capital retention requirement of 5% for the originating bank. This was amended by Regulation 2401/201724, with the introduction of Simple, Transparent and Standardised (STS) securitisations and the associated beneficial capital treatment. However, CLOs actively managed on a discretionary basis are ineligible for STS label. Further, the 2017 Securitisation Regulation 25 defined a common approach for supervising securitisation repositories, introducing a comprehensive set of rules for: (i) providing transparency and due diligence requirements, (ii) banning re-securitisations, (iii) regulating originate-to-distribute securitisations, (iv) avoiding securitisations without risk retention, (v) preventing investments without proper due diligence, (vi) prohibiting selling of securitisations to retail investors, and (vii) providing rules introducing geographical requirements for establishing securitisation special purpose entities. ◊ Japan: The risk retention rule was introduced in March 201926, and it applies to all securitisation instruments, requiring that Japanese banks hold excess capital against all securitization exposures unless the CLO originator retains, on a continuous basis and without hedging the credit risk, at least 5% of the exposure of the securitization or unless it is determined through due diligence that “no improper original assets are structured.” Since the exception provided is not equivalent to the US blank exception for “open-market CLOs”, Japanese investors are expected to conduct extensive due diligence. Specifically, within this exception, banks need to either: (i) demonstrate that the underlying assets of securitised instruments are properly formed, or (ii) apply risk weights that are three times higher than original risk weights. 19 The requirement for a securitizer to retain an economic interest in a portion of the credit risk for any asset that the securitizer, through the issuance of an asset-backed security, transfers, sells or conveys to a third party in a transaction. 20 The US regulatory framework adopted the Credit Risk Retention Rule in December 2014 and became effective in December 2016. US Regulatory Agencies, Credit Risk Retention, Federal Register, Volume 79, No. 247. 21 US Court of Appels, District of Columbia Circuit, Loan Syndications & Trading Associations v. Sec. & Exch. Commission, 9 February 2018, 882F.3d 220 (D.C.Cir.2018). It shall be noted that the bank “balance sheet” CLOs have generally disappeared after the GFC as they became unfeasible due to more stringent capital requirements regulation. 22 ECB, The Incentive Structure of the ‘Originate to Distribute’ model, December 2008. 23 Regulation (EU) No 575 of the European Parliament and of the Council of 26 June 2013 on prudential requirements for credit institutions and investment firms, Official Journal of the European Union (OJ L 176, 27.6.2013). 24 Regulation (EU) 2017/2401 of the European Parliament and of the Council of 12 December 2017 amending Regulation (EU) No 575/2013 on prudential requirements for credit institutions and investment firms, Official Journal of the European Union (OJ L 347, 28.12.2017) 25 Regulation (EU) 2017/2402 of the European Parliament and of the Council of 12 December 2017 laying down a general framework for securitisation and creating a specific framework for simple, transparent and standardised securitisation, Official Journal of the European Union (OJ L 347, 28.12.2017). Under this regulation, ESMA has direct responsibilities regarding the registration and supervision of securitisation repositories. 26 Banking Act of Japan, FSA Notice No.19, 2006, Chapter VIII, Section I, Article 248, paragraph 3.

27 C. List of abbreviations AFME Association for Financial Markets in Europe ABS Asset-Backed Securities BIS Bank for International Settlements CDOs Collateralised Debt Obligations CLO Collateralised Loan Obligation CMBS Commercial Mortgage-Backed Securities CRA Credit Rating Agency CRA Regulation Regulation (EC) 1060/2009 of the European Parliament and of the Council of 16 September 2009 on credit rating agencies EBA European Banking Authority ECB European Central Bank EMEA Europe, the Middle East and Africa ESMA European Securities and Markets Authority EU European Union FED US Federal Reserve Bank Fitch Fitch Ratings FSB Financial Stability Board GFC Global Financial Crisis Moody’s Moody’s Investors Service RMBS Residential Mortgage-Backed Securities SEC US Securities and Exchange Commission S&P S&P Global Ratings SME Small and Medium Enterprises SPV Special Purpose Vehicle STC Simple, transparent and comparable securitisations UCITS Undertakings for the Collective Investment in Transferable Securities UK United Kingdom US United States of America WAL Weighted Average Life

28 D. List of references AFME, European High Yield and Leveraged Loan Report - Q4 and full year 2021, 11 March 2022. AFME, European High Yield and Leveraged Loan Report - Q2 2022, 27 September 2022. BIS, The Join Forum – Report on asset securitisation incentives, July 2011. BIS, Report on asset securitisation incentives, 13 July 2011. BIS, The role of ratings in structured finance: issues and implications, January 2005. EBA, Report on qualifying securitisation, 7 July 2015. ECB, Leveraged transactions – supervisory expectations regarding the design and functioning of risk appetite frameworks and high levels of risk taking, March 2022. ECB, The Incentive Structure of the ‘Originate to Distribute’ model, December 2008. ESMA50-165-1842, ESMA Reports on Trends, Risks and Vulnerabilities, No.2, 1 September 2021. ESMA50-165-1287, ESMA Reports on Trends, Risks and Vulnerabilities, No.2, 2 September 2020. ESMA80-189-6982, Thematic Report – EU CLO credit ratings, an overview of Credit Rating Agencies practices and challenges, 13 May 2020. ESMA50-165-883, ESMA Reports on Trends, Risks and Vulnerabilities, No.2, 10 September 2019. ESMA/2016/1575, Guidelines on the validation and review of Credit Rating Agencies’ methodologies, 23 March 2017. ESRB, Monitoring systemic risks in the EU securitisation market, July 2022. European Commission, A Capital Markets Union for people and businesses – New Action plan, 24 September 2020. European Parliament, Understanding Securitisation, October 2015. FSB, Vulnerabilities associated with leveraged loans and collateralised loan obligations, December 2019. US Court of Appels, District of Columbia Circuit, Loan Syndications & Trading Associations v. Sec. & Exch. Commission, 9 February 2018, 882F.3d 220 (D.C.Cir.2018). S&P, Par Wars: The Investor Strikes Back, 2 May 2018. S&P, The $700B in US CLOs: Who holds them, what risk they pose, 21 June 2019. S&P, Par Wars: US CLO Document Provisions Evolve to Provide Managers More Flexibility, 23 September 2022. ***Banking Act of Japan, FSA Notice No.19, 2006. ***Regulation (EU) 2017/2401 of the European Parliament and of the Council of 12 December 2017 amending Regulation (EU) No 575/2013 on prudential requirements for credit institutions and investment firms, Official Journal of the European Union (OJ L 347, 28.12.2017). ***Regulation (EU) 2017/2402 of the European Parliament and of the Council of 12 December 2017 laying down a general framework for securitisation and creating a specific framework for simple, transparent and standardised securitisation, Official Journal of the European Union (OJ L 347, 28.12.2017). ***Regulation (EU) No 575 of the European Parliament and of the Council of 26 June 2013 on prudential requirements for credit institutions and investment firms, Official Journal of the European Union (OJ L 176, 27.6.2013). ***Commission Delegated Regulation (EU) No 447/2012 of 21 March 2012 supplementing Regulation (EC) No 1060/2009 of the European Parliament and of the Council on credit rating agencies by laying down regulatory technical standards for the assessment of compliance of credit rating methodologies, Official Journal of the European Union (OJ L 140, 30.5.2012). ***Regulation (EC) 1060/2009 of the European Parliament and of the Council of 16 September 2009 on credit rating agencies, Official Journal of the European Union (OJ L 302, 17.11.2009). ***US Regulatory Agencies, Credit Risk Retention, Federal Register, Volume 79, No. 247.