Alix Auzepy Christina E. Bannier Integrating Climate Risks in Bank Risk Management and Capital Requirements Integrating Climate Risks in Bank Risk Management and Capital Requirements Alix Auzepy · Christina E. Bannier Integrating Climate Risks in Bank Risk Management and Capital Requirements Alix Auzepy FB 02 – Wirtschaftswissenschaften Justus-Liebig-Universität Gießen Gießen, Germany Prof. Dr. Christina E. Bannier FB 02 – Wirtschaftswissenschaften Justus-Liebig-Universität Gießen Gießen, Germany ISBN 978-3-658-47060-9 ISBN 978-3-658-47061-6 (eBook) https://doi.org/10.1007/978-3-658-47061-6 © The Editor(s) (if applicable) and The Author(s) 2025. This book is an open access publication. Open Access This book is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this book are included in the book’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the book’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer Gabler imprint is published by the registered company Springer Fachmedien Wiesbaden GmbH, part of Springer Nature. The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany If disposing of this product, please recycle the paper. Foreword As climate-related events become ever more frequent and policy landscapes evolve, credit institutions are under increasing pressure to consider a broader spectrum of risks, including physical risks from extreme weather events and transition risks associated with regulatory changes. This study examines how credit institutions are adapting their risk management frameworks and capital adequacy strategies to better account for such emerging risks. Via semi-structured interviews with 30 risk experts and a review of institutions’ disclosure re- ports, we consider two key questions: (i) how are institutions integrating climate risks into their risk management frameworks, and what challenges are they encountering?; (ii) to what extent are climate risks being incorporated into capital considerations, and through which mechanisms are they influencing the determination of capital requirements? By highlighting both the approaches taken and the challenges faced by institutions, this work sheds light on the different ways in which climate risks are reshaping risk management practices. A simple but telling example in this context is that of stress tests, where climate risks are prompting credit institutions to extend the time horizons they typically consider in such exercises — from the traditional 3-5 years to up to 30 years — in order to better account for potential long-term impacts. This work aims to provide a detailed analysis of institutional practices and regulatory devel- opments that we hope will be of value to academics, industry practitioners and policymakers engaged in this field. By presenting concrete examples of approaches used by large credit institutions in the European Union, enriched with insights from risk experts, our study also provides guidance for the climate risk integration efforts of smaller credit institutions that may be in the early stages of their “journey”. Importantly, this work aims to paint an ob- jective picture of current developments and does not advocate for a particular policy stance. As this book is the result of an extensive research project, its structure is similar to that of a research paper, including a comprehensive discussion of the theoretical background, cov- ering key regulatory and supervisory developments, and a short literature review describing the empirical evidence on the growing relevance of climate risks for financial institutions and supervisors. Before presenting the results, a methodology section outlines the empirical approach and the conceptual framework that guides our analysis. We conclude with a dis- cussion of limitations and recommendations for further research and policy initiatives. This project was made possible by the support of a number of contributors, to whom we express our sincere gratitude. First and foremost, we would like to thank the risk experts who participated in our interviews and provided very valuable insights into current prac- V tices and challenges. Their contributions add a practical, hands-on dimension to our study that enhances the relevance of the findings. We would also like to thank the Frankfurt Institute for Risk Management and Regulation (FIRM) for facilitating the dialogue with key stakeholders whose expertise was instrumental to this research. We are thankful for financial support from the German Federal Ministry of Education and Research (BMBF) through the funding measure “Climate Protection and Finance” (KlimFi). While the BMBF had no direct involvement in the design and implementation of this research, the funding (grant no. 01LA2210C) greatly contributed to enhancing our research in the area of climate- related scenario analysis, stress testing and transition planning. This funding also facilitated collaborative research initiatives and exchanges with the Frankfurt School of Finance and Management and the German Institute for Economic Research (DIW). We are also grateful for financial support from CRIF GmbH, which made it possible to publish this research as open access. We firmly believe that an effective and scientifically- grounded integration of climate risks into risk management depends on the availability of research findings to a wide audience, including practitioners, researchers, and policymak- ers. CRIF’s commitment to fostering best practices in climate-related risk management aligns with the goals of this work. Lastly, we thank the University Library and the Giessen Graduate Centre for Social Sciences, Business, Economics and Law of the Justus-Liebig- University Giessen for providing additional funding for the publication of this work via open access. Sincerely, Alix Auzepy Christina E. Bannier VI Table of contents 1 Introduction 1 2 Theoretical background 9 2.1 Regulatory and supervisory developments . . . . . . . . . . . . . . . . . . . 9 2.2 Literature review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3 Research method and sample 17 3.1 Semi-structured interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 3.2 Archival data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.3 Conceptual framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4 Results 27 4.1 Pillar I: Integrating climate risks into the standardized approach (SA) . . . . 27 4.1.1 Background and rationale . . . . . . . . . . . . . . . . . . . . . . . . 27 4.1.2 Credit ratings and dialogue with rating agencies . . . . . . . . . . . . 29 4.2 Pillar I: Integrating climate risks into internal models (IRB) . . . . . . . . . 30 4.2.1 Background and rationale . . . . . . . . . . . . . . . . . . . . . . . . 30 4.2.2 Materiality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 4.2.3 Rating criteria and climate-related scores . . . . . . . . . . . . . . . . 36 4.2.4 Override framework and margin of conservatism (MoC) . . . . . . . . 42 4.2.5 Model calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.3 Pillar II: Integrating climate risks into the ICAAP and broader risk management 48 4.3.1 Background and rationale . . . . . . . . . . . . . . . . . . . . . . . . 48 4.3.2 Risk identification, materiality assessment & risk inventory . . . . . . 52 4.3.2.1 Identification of risk drivers and relevance assessment . . . . 52 4.3.2.2 Materiality assessment . . . . . . . . . . . . . . . . . . . . . 56 4.3.2.3 Risk inventory . . . . . . . . . . . . . . . . . . . . . . . . . 63 4.3.3 Risk quantification approaches . . . . . . . . . . . . . . . . . . . . . . 64 4.3.3.1 Credit rating simulations and credit portfolio models . . . . 66 4.3.3.2 Climate-related scoring methodologies . . . . . . . . . . . . 68 4.3.3.3 Climate-related data and data collection . . . . . . . . . . . 75 4.3.4 Risk quantification with climate stress testing . . . . . . . . . . . . . 80 4.3.4.1 Integration of climate stress testing into stress testing frame- work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 4.3.4.2 Climate scenario choices, assumptions and related challenges 88 VII 4.3.4.3 Choice of baseline scenario . . . . . . . . . . . . . . . . . . . 102 4.3.4.4 Top-down modeling approaches . . . . . . . . . . . . . . . . 107 4.3.4.5 Bottom-up modeling approaches . . . . . . . . . . . . . . . 113 4.3.4.6 Balance sheet assumptions . . . . . . . . . . . . . . . . . . . 116 4.3.4.7 Physical risk scenario analysis . . . . . . . . . . . . . . . . . 120 4.3.4.8 Data for climate stress testing . . . . . . . . . . . . . . . . . 124 4.3.4.9 Integration of climate stress test results into business and risk strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 4.3.5 Internal capital adequacy assessments and capital considerations . . . 129 4.3.5.1 Integration of climate risks into scenarios for capital ade- quacy and capital planning . . . . . . . . . . . . . . . . . . 129 4.3.5.2 Climate-related capital adequacy assessments . . . . . . . . 132 4.3.5.3 Climate-related capital buffers . . . . . . . . . . . . . . . . . 137 4.3.5.4 Economic and normative perspectives . . . . . . . . . . . . 143 4.4 Pillar II: Integrating climate risks into risk appetite, risk metrics and limit system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 4.4.1 Linking risk appetite, risk strategy & business strategy . . . . . . . . 145 4.4.2 Key risk indicators (KRIs) . . . . . . . . . . . . . . . . . . . . . . . . 149 4.4.3 Limits and exclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 4.5 Pillar I & II: Integrating climate risks into loan pricing . . . . . . . . . . . . 159 4.5.1 Background and rationale . . . . . . . . . . . . . . . . . . . . . . . . 159 4.5.2 Pricing of climate risks through capital considerations . . . . . . . . . 160 4.5.3 Pricing of climate risks through credit risk and funding considerations 163 4.6 Pillar II: Integrating climate risks into the Supervisory Review and Evaluation Process (SREP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 4.6.1 Background and rationale . . . . . . . . . . . . . . . . . . . . . . . . 165 4.6.2 Business model, governance and risk management assessments . . . . 168 4.6.3 Pillar 2 Requirement (P2R) and Pillar 2 Guidance (P2G) . . . . . . . 170 5 Discussion and conclusion 173 References 175 Appendix 185 VIII List of abbreviations A-IRB Advanced Internal Ratings-Based Approach BCBS Basel Committee on Banking Supervision CCF Credit Conversion Factor CET1 Common Equity Tier 1 CQS Credit Quality Step CRD Capital Requirements Directive CRR Capital Requirements Regulation CSRD Corporate Sustainability Reporting Directive EAD Exposure at Default EBA European Banking Authority ECB European Central Bank ESG environmental, social and governance ESMA European Securities and Markets Authority EU European Union F-IRB Foundation Internal Ratings-Based Approach GDP Gross Domestic Product GVA gross value added ICAAP Internal Capital Adequacy Assessment Process IEA International Energy Agency IFRS 9 International Financial Reporting Standard 9 ILAAP Internal Liquidity Adequacy Assessment Process IPCC Intergovernmental Panel on Climate Change IRB Internal Ratings-Based Approach ITS Implementing Technical Standard KRI Key Risk Indicator LCR Liquidity Coverage Ratio LGD Loss Given Default LSI Less Significant Institution M Maturity MoC Margin of Conservatism NGFS Network for Greening the Financial System P2G Pillar 2 Guidance P2R Pillar 2 Requirement PD Probability of Default IX RCP Representative Concentration Pathway RSU Rating Service Unit GmbH & Co. KG RWA risk-weighted assets SR Sparkassen Rating und Risikosysteme GmbH SREP Supervisory Review and Evaluation Process SSM Single Supervisory Mechanism X The authors Alix Auzepy is a research assistant and doctoral candidate at the Chair of Banking and Finance at the Justus-Liebig- University Giessen, Germany. Prior to her PhD, she worked at KfW IPEX-Bank GmbH as a portfolio manager in the renewable energy sector and later as a project manager in the strategy and sustainability department. She is a certified Sustainability and Climate Risk Analyst (GARP SCR®) through the Global Asso- ciation of Risk Professionals. Her research focuses on sustainable finance, climate risks and investor preferences. Christina E. Bannier is a Professor of Banking and Finance at the Justus-Liebig-University Giessen, Germany. Previously, she held professorships at the Johannes Gutenberg University Mainz, the Frankfurt School of Finance & Management, and the Leibniz University Hannover. In addition to her academic roles, she is a member of the supervisory boards of DWS Group GmbH & Co. KGaA, Clearstream Banking AG, and Eurex Clearing AG. Her expertise lies in the areas of sustainable finance, corporate gov- ernance, (climate-related) risk management, and data analytics. XI 1 Introduction If left unaddressed, today’s risks can easily turn into tomorrow’s losses. It is therefore vital for credit institutions1 to have in place a robust and comprehensive risk management framework, including models and processes that ensure a clear understanding of risks and effective mitigation strategies. Nevertheless, even the most flawless risk management cannot fully eliminate the possibility of losses. Thus, a key imperative for these institutions is also to hold sufficient capital to be able to absorb losses exceeding expected levels (Holscher et al. 2022). This can be achieved by ensuring that capital requirements are met and that capital adequacy is maintained on an ongoing basis.2 In addition, maintaining sufficient capital is essential for institutions to support their business models and pursue their strategic objectives (Gruenewald et al. 2023).3 The risks posed by climate change, including transition and physical risks, have in recent years contributed to reshaping the risk landscape known to credit institutions. Physical risks, such as acute and chronic weather events, and transition risks, including the economic impacts associated with the process of adjustment towards a lower-carbon economy, are receiving increasing attention from civil society, regulators and investors alike (Demekas & Grippa 2022).4 There are also concerns that such risks may change the distribution of losses that banks face (Holscher et al. 2022). In an evolving risk landscape, institutions need to adapt their risk management practices and continuously assess the robustness of their overall risk strategies (Buch 2024). Climate risks, therefore, compel credit institutions to rethink their risk monitoring processes and to reassess their approach to capital adequacy from a new angle. However, as new (or “novel”) risks, for which historical data provides limited insight into future developments, climate risks are, by definition, inherently uncertain (Buch 2024, ECB 2024c). Accounting for such risks therefore also poses significant conceptual, analytical and data challenges (Demekas & Grippa 2024). This study seeks to contribute to the discussion on climate risk integration by addressing two key broad questions. First, how are credit institutions integrating climate risks into their 1In ECB (2020b), supervised credit institutions are commonly referred to as “institutions”. Throughout this study, we use the terms “credit institutions”, “banks” and “institutions” interchangeably. 2Capital adequacy refers to the degree to which risks are covered by capital (ECB 2018). 3The weaker an institution’s capital base, the more difficult and costly it may be to pursue its strategy. For example, lower capital levels may lead investors and counterparties to demand higher risk premiums, thus reducing profitability (Gruenewald et al. 2023). 4Acute physical risks stem from extreme events such as droughts, floods and storms, while chronic physical risks result from gradual changes, such as rising temperatures and sea levels. Transition risks stem from shifts in climate-related policies, technological advancements or changes in market sentiment and preferences (ECB 2020b). © The Author(s) 2025 A. Auzepy and C. E. Bannier, Integrating Climate Risks in Bank Risk Management and Capital Requirements, https://doi.org/10.1007/978-3-658-47061-6_1 1 https://crossmark.crossref.org/dialog/?doi=10.1007/978-3-658-47061-6_1&domain=pdf risk management frameworks, and what challenges are they encountering? (“Question 1”). Second, to what extent are climate risks being incorporated into capital considerations, and through which mechanisms are they influencing the determination of capital requirements? (“Question 2”).5 To address these questions, we examine institutions with total assets of more than EUR 30 billion, focusing particularly on 15 credit institutions directly supervised by the European Central Bank (ECB) and subject to the Capital Requirements Regulation (CRR) and the Capital Requirements Directive (CRD). In recent years, the ECB has been at the forefront of defining strategies to address climate risks in the banking sector (Gruenewald et al. 2023). In November 2020, it published 13 broad supervisory expectations regarding the management of climate-related and environmental risks (ECB 2020b). These expectations are to be fully implemented by the end of 2024, making ECB-supervised institutions an ideal context for examining how banks are incorporating climate risks into their overall risk management practices. In 2022, the ECB conducted both a climate stress test and a thematic review of climate-related risk management practices. The ECB found that most institutions had implemented at least “basic practices” across most areas of the expectations, but also voiced “significant supervisory concern” about the institutions’ ability to achieve a higher level of implementation (p.2 & p.5, ECB 2022e). Among these institutions, we conduct a series of semi-structured interviews with 30 respon- dents. Our interviewees have expertise in risk management, risk controlling, risk modeling, stress testing and related fields. Although the ECB has set out clear supervisory expec- tations, it does not prescribe specific approaches for their implementation. Our interviews therefore allow us to gain in-depth insights into to the various — and often heterogeneous — approaches for integrating climate risks, reflecting both current practices and ideas for future developments. In addition, in light of the concerns voiced by the ECB, it appears crucial to shed more light on the specific challenges at hand. This study therefore presents a structured discussion not only of the climate risk integration approaches and their underlying rationale, but also of the barriers identified by respondents. Finally, given that these respondents play a crucial role in implementing the ECB’s supervisory expectations, understanding their perspectives is particularly valuable. By better identifying and describing these challenges, we aim to contribute to a smoother, more effective integration of climate risks into banking practices. 5Capital requirements refers, in this context, to the determination of capital needs in a broader sense, i.e., amount needed to cover risks, rather than compliance with minimum capital requirements in a regulatory sense. 2 We complement the semi-structured interviews by a review of the institutions’ disclosure re- ports (“Pillar 3” reports). An important development is that these reports are now required to include information on environmental, social and governance (ESG) risks,6 with a partic- ular focus on physical risks and transition risks (see Art. 449a of the CRR).7 Specifically, qualitative disclosures are required to include information on risk management related to en- vironmental risks, while quantitative disclosures cover exposures to transition risks, physical risks and mitigating actions. Even though these reports, particularly the qualitative sections, vary in their level of detail and do not directly highlight specific challenges, they provide a valuable overview of the climate-related risk management practices in place. The reports typically cover topics such as risk identification, materiality assessments, scenario analysis, stress testing, risk metrics, and integration into risk appetite. We therefore use the infor- mation from these reports to both inform our semi-structured interviews and complement interview findings. Examining the integration of climate risks into risk management and capital considerations requires an understanding of the regulatory and supervisory frameworks governing financial institutions. These frameworks are shaped by a range of laws and regulations, largely derived from the Basel Accords (“Basel framework”) (EBA 2022b).8 In the European Union (EU), the Basel framework is implemented through the CRR and the CRD, and is further supported by technical standards and guidelines issued by EU bodies, such as the European Banking Authority (EBA) and national supervisors, such as the BaFin and Bundesbank in Germany.9 It is widely accepted that climate risks are not a stand-alone risk category but act as drivers of “conventional” (i.e., prudential) risk types, such as credit, market and operational risks (BIS 2022b). As a result, it is useful to contextualize the integration of climate risks within the broader framework designed to manage these prudential risks. To guide and structure our discussion, we therefore propose a conceptual framework aligned with core elements of the “Pillar 1” and “Pillar 2” of the Basel framework. Pillar 1, on the one hand, specifies standard rules for calculating minimum capital requirements relative to risk-weighted assets (RWA). In essence, riskier assets receive a higher risk weight, so that capital increases with risk (Holscher et al. 2022). These rules mainly apply to credit risk, 6Climate risks are typically treated as a subset of ESG risks, which falls under the environmental pillar. 7For the technical standards, see Implementing Regulation (EU) 2022/2453 of 30 November 2022 amend- ing the implementing technical standards laid down in Implementing Regulation (EU) 2021/637 as regards the disclosure of environmental, social and governance risks. 8The Basel Committee on Banking Supervision sets standards for prudential regulation and serves as a forum for cooperation on banking supervision. It is composed of central banks and supervisory authorities. These standards are regularly updated, with the most recent series of reforms known as Basel III. 9For example, in Germany, even the MaRisk, referring to the minimum requirements for risk management, implement qualitative requirements derived from the Basel framework into German law. 3 market risk and operational risk. Examining the integration of climate risks within Pillar 1 is especially relevant for addressing Question 2 on capital considerations. Nonetheless, since Pillar 1 defines minimum capital requirements, its outcomes also feed into broader capital considerations and risk management strategies under Pillar 2. Given the importance of climate risks primarily for credit risk, we focus on the standardized and internal ratings- based approaches — two credit risk-related methods for determining risk weights under Pillar 1 and capital requirements for credit risk (EBA 2023). Pillar 2, on the other hand, consists mainly of the Internal Capital Adequacy Assessment Pro- cess (ICAAP) and the Supervisory Review and Evaluation Process (SREP).10 The ICAAP is the process by which institutions assess the adequacy of their capital using their own internal methods and risk management processes (ECB 2018). It considers not only the risk types covered under Pillar 1, but also all other material risks to which the institutions are exposed. In this context, the capital considerations under the ICAAP go beyond the minimum capital requirements set under Pillar 1. The ECB expects institutions to achieve full integration of climate risks into the ICAAP by the end of 2024 (ECB 2022e). Examining such integration in the context of this study is therefore relevant for addressing both Question 1 on risk man- agement and Question 2 on capital considerations. As the ICAAP is a multi-step process, we break down each of the key steps (e.g., risk identification and inventory, risk quantification, stress testing and capital allocation) and discuss the approaches and challenges involved in incorporating climate risks at each stage. The ICAAP is reviewed by supervisors as part of the SREP, which is an annual process where supervisors evaluate, on a case-by-case basis, the risks institutions face and their ability to manage these risks in terms of risk management, capital, and liquidity (ECB 2023g). One key outcome of the SREP is the Pillar 2 Requirement (P2R), a binding capital requirement set by supervisors (ECB 2023f). This requirement can be applied, for example, if the ECB determines that an institution underestimates or inadequately manages certain material risks. As a result, the SREP process can shape institutions’ risk management practices regarding climate risks and affect their capital requirements. Therefore, examining the role of climate risks within the SREP framework is also relevant for both Question 1 and Question 2. To structure our analysis, we categorize our results based on their main relevance to either Pillar 1 and/or Pillar 2. For example, the ICAAP is generally regarded as a process associated 10The Internal Liquidity Adequacy Assessment Process (ILAAP) is also part of Pillar 2, but is not covered in this study. 4 with Pillar 2 and, therefore, all results related to the ICAAP are classified under this pillar.11 Within this conceptual framework of Pillars 1 and 2, our interview-based analysis delivers the following key takeaways and observations, which will be elaborated in more detail in the following sections: • Capital for credit risk - Internal ratings-based approach (Pillar 1): A key de- velopment is that the ECB expects that where climate-related risks drivers are “found to be material, institutions should include such risk drivers in their internal models” ap- proved for use for the calculation of capital requirements for credit risk (ECB 2024a). The integration of climate risks into internal ratings-based models is still in its early stages, with a key challenge being the definition and assessment of materiality. While materiality assessments performed under the ICAAP are generally forward-looking, the quantitative integration of risk drivers into rating systems under Pillar 1 relies on statis- tical assessments and calibration requiring historical data. Short-term climate risks with financial impacts are considered by most respondents to be implicitly captured in the ratings, whereas medium- and long-term risks rely more heavily on ad hoc rating over- rides and/or additional Margin of Conservatism (MoC). One institution quantitatively integrates climate risks into its ratings and Probability of Default (PD) estimates. Most institutions rely mainly on qualitative considerations, client engagement and apply rating overrides where necessary. The overrides are based on expert judgement derived from in- ternal and highly heterogeneous climate-related scorecards, introduced by the institutions for risk management purposes. While overrides offer some flexibility and are based on a granular assessment of climate risks in the short term, they bear risks from a model val- idation perspective. They may also hinder a more systematic and consistent integration of climate risks into PD estimates.12 • Capital for credit risk - Standardized approach (Pillar 1): At present, the main method to integrate climate risks into this approach is through external credit ratings. Institutions primarily rely on the three major rating agencies — Moody’s, Standard & Poor’s and Fitch. The broad coverage provided by these agencies is a key reason for their selection. Consequently, the extent to which climate risks influence risk weights and capital requirements largely depends on how these agencies incorporate climate factors into their assessments, rather than on an active decision by the institutions themselves. 11We note, however, that certain aspects of Pillar 1 are also relevant to the ICAAP and, by extension, to Pillar 2. For example, models developed for Pillar 1 may also find application in the context of Pillar 2. 12This study primarily focuses on PD considerations under Pillar 1. However, it is important to acknowl- edge that other credit risk parameters, including loss given default and the treatment of collateral values, also play a role in determining minimum capital requirements for credit risk. 5 With the “output floor” coming into effect in January 2025, the standardized approach is expected to gain importance in the coming years and see even wider adoption. This shift is likely to let the role of external rating agencies and their methodologies become even more relevant in shaping minimum capital requirements in relation to climate risks. • ICAAP - Risk identification, materiality assessment and risk inventory (Pillar 2): Institutions are increasingly incorporating climate risks into their internal risk identi- fication process and annual risk inventory as part of the ICAAP. This involves identifying relevant climate-related risk drivers and determining their materiality. A key challenge in this process is the materiality assessment, which is multifaceted and requires detailed, granular data from both internal and external sources to evaluate how different climate- related risk drivers will affect various portfolios and risk types over short-, medium- and long-term horizons. Most institutions rely on qualitative approaches, such as expert judgement, while some further substantiate their materiality assessments with insights from quantitative methods, such as risk-type-specific scenario analyses and exploratory stress tests. Approaches to defining materiality as well as the types of materiality thresh- olds applied vary widely, highlighting the complexity of integrating climate risks based on materiality considerations. • ICAAP - Risk quantification methodologies and data collection (Pillar 2): In- stitutions are increasingly using forward-looking simulation techniques to project changes in credit ratings based on climate scenarios. Some institutions are working to improve credit portfolio models by integrating climate risks, including through correlation factors. In addition, institutions are developing climate-related scorecards to quantify transition and physical risks at the individual client level. The scores resulting from these scorecards are also used in the context of rating overrides (see internal ratings-based approach under Pillar 1). The design of climate-related scorecards (e.g., criteria, weights) varies across institutions. Transition risks are assessed using criteria such as greenhouse gas emis- sions, technology risks and transition plans, while physical risks rely on criteria related to the regional or international nature of companies, geolocation and hazard-related data. Important data gaps remain, including on geolocation of corporate assets, production facilities and insurance coverage. • ICAAP - Risk quantification with climate stress tests (Pillar 2): Institutions are gradually integrating climate stress tests into their broader stress testing frameworks. While many have introduced climate stress tests as a separate scenario class, others have also incorporated climate-related elements directly into their quarterly, institution-wide stress assessments. Most institutions rely on publicly available external climate scenarios, 6 employing a diverse mix of short- and long-term scenarios, with considerable variation in scenario selection. Institutions make various adjustments to the scenarios, not only in terms of scenario expansion, but also by adding new assumptions, such as CO2 cost pass- through rates. A key challenge is the selection of an appropriate baseline scenario, with institutions adopting different approaches. Some institutions incorporate climate risks into the baseline macroeconomic forecast, while others use a dedicated climate scenario as the baseline. Modeling dynamic balance sheet assumptions is another major challenge. Finally, to achieve a comprehensive assessment and granular quantification of climate risks, institutions tend to combine top-down and bottom-up modeling approaches. • ICAAP - Internal capital adequacy and climate-related capital buffers (Pil- lar 2): Several institutions use a combination of exploratory scenarios, which focus on specific risks such as credit or operational risk, and comprehensive scenarios, which en- compass multiple risk types while considering macroeconomic impacts. Climate-related aspects, including CO2 prices, are increasingly incorporated into these comprehensive sce- narios, supporting both capital adequacy assessments and capital planning efforts. Due to modeling uncertainties — particularly the limitations of traditional risk models in capturing medium- to long-term climate risks — some institutions have started to intro- duce climate-specific capital buffers. These instruments, often precautionary and based on expert judgement, vary in form, including general buffers, economic capital add-ons and adjustments to management buffers. One institution combines an economic capital add-on for climate risks under the ICAAP with a management overlay for expected loss models. • Risk appetite, risk metrics and limit systems (Pillar 2): Institutions are increas- ingly integrating climate risks into their risk appetite frameworks, including their credit risk appetite. Some institutions have formalized their climate objectives within their risk appetite statements, linking these objectives directly to risk metrics and limits. Institu- tions are also introducing a wide range of Key Risk Indicators (KRIs) specifically designed to track climate risk exposures and setting specific limits within their risk appetite frame- works. These limits vary in scope and focus, but typically target sectors and activities most exposed to transition and physical climate risks. • Integration of climate risks into loan pricing (Pillar 1 & 2): Institutions are increasingly incorporating climate risks into loan pricing by factoring in credit risk costs, funding costs and capital costs. While climate risk-based pricing is still in its infancy, some institutions use innovative tools, such as green refinancing curves and margin premi- ums linked to internal ESG scores and loan maturities. Scenario analysis is also becoming 7 increasingly used for loan pricing purposes, allowing institutions to develop default prob- abilities that are conditional on climate scenarios and determine credit provisions based on such scenarios. One institution includes climate risks into macroeconomic scenarios used for expected credit loss. • SREP, P2R and P2G (Pillar 2): Climate risks are now integral to three critical SREP components: business model assessment, governance and risk management assess- ment and risks to capital assessment. The business model assessment holds particular importance, as supervisors increasingly evaluate institutions’ strategic and operational plans, as well as their ability to assess climate risks from a business model perspective and to integrate them into broader business strategies. This qualitative integration of climate risks into the SREP framework influences SREP scores and is expected to more directly shape P2R levels over time. The Pillar 2 Guidance (P2G) is expected to play a greater role in the future as well and to further integrate climate risks into capital recommendations. Our study contributes to the growing body of literature on the role of climate risks in the banking sector, particularly in the context of banking supervision. To date, the ECB has published good practice reports on climate-related risk management and stress testing (ECB 2022c,d). The EBA has also issued several reports on the integration of climate risks into Pillar 1 (EBA 2022b, 2023). We add to this literature by providing a comprehensive analysis that highlights the interplay between Pillar 1 and Pillar 2. Furthermore, our study contributes to the growing empirical literature on both the relevance and the implications of climate risks for financial institutions. We discuss the existing empirical literature in more detail in Section 2.2. The remainder of this study is organized as follows. Section 2 offers an overview of the latest regulatory and supervisory developments related to climate risks. Furthermore, it reviews the literature and examines the relevance and implications of climate risks for financial institutions and supervisors. In Section 3, we describe our data and sample, and outline the conceptual framework used to structure our analysis. Section 4 presents our findings. Note that within the results section, each sub-section is designed to be largely self-explanatory. Section 5 discusses the limitations of our study and concludes. Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. 2 Theoretical background 2.1 Regulatory and supervisory developments The management of climate risks has received considerable attention from legislators and supervisors, especially in Europe. In particular, there has been growing interest and work by these actors on the extent to which climate risks are captured and addressed (or not) within the prudential framework. The European legislator has issued several mandates related to the integration of climate risks into the three pillars of the Basel III framework, making these risks a key supervisory concern (Gruenewald et al. 2023). For example, the EBA has a mandate under the CRD to issue guidelines to credit institutions in all EU member states on minimum standards for the identification, measurement and management of ESG risks.13 Between 2021 and 2023, the EBA published, by exercising its mandate, various discussion papers and reports exploring the justification for a prudential, risk-based approach to ESG risks (EBA 2021, 2022b, 2023). These reports address the integration of environmental risks, including transition and phys- ical risks, into Pillar 1 (EBA 2023) and into risk management practices under Pillar 2 (EBA 2021). Similarly, the Basel Committee on Banking Supervision (BCBS) released a report in December 2022, outlining how climate risks may be more consistently incorporated into the existing Basel framework and reflected within existing Pillar 1 standards (BIS 2022a). In the same year, the European Commission adopted new rules requiring additional disclo- sures under Pillar 3, including both qualitative information on ESG risk management and quantitative information on transition and physical risks.14 In line with these developments, the ECB has increasingly incorporated climate and environ- mental considerations into its supervisory practice. The ECB has made the management of climate and environmental risks one of its supervisory priorities for 2024-2026 (ECB 2024f). Through its role in the Single Supervisory Mechanism (SSM), the ECB monitors how signif- icant institutions in the euro area address climate risks (ECB 2020b). To support this, the ECB published a guide setting out 13 supervisory expectations for the integration of climate and environmental risks (the “ECB Guide”). It also set staggered deadlines for supervised institutions to comply with these expectations by the end of 2024. In terms of content, these 13See Directive 2013/36/EU on access to the activity of credit institutions and the prudential supervision of credit institutions and investment firms. 14See Implementing Regulation (EU) 2022/2453 of 30 November 2022 amending the implementing tech- nical standards laid down in Implementing Regulation (EU) 2021/637 as regards the disclosure of environ- mental, social and governance risks. © The Author(s) 2025 A. Auzepy and C. E. Bannier, Integrating Climate Risks in Bank Risk Management and Capital Requirements, https://doi.org/10.1007/978-3-658-47061-6_2 9 https://crossmark.crossref.org/dialog/?doi=10.1007/978-3-658-47061-6_2&domain=pdf expectations focus primarily on Pillar 2 aspects, including risk management and internal capital adequacy assessment process. Specifically, institutions are expected to consider cli- mate risks in their business strategies, risk appetite frameworks and credit processes, while assigning clear responsibilities for managing these risks. They are also required to collect and report relevant data internally, incorporate climate risks into capital adequacy consider- ations and conduct stress tests that include these risks. Finally, institutions are expected to consider the potential impact of climate risks on their liquidity position, business continuity and reputation. Another important supervisory development is the growing number of climate stress tests initiated by central banks, supervisors and regulators. In Europe, the ECB conducted one of the most comprehensive tests in 2022, building on earlier national-level efforts. De Neder- landsche Bank, the Dutch central bank, was among the first to introduce a mandatory stress test focused on transition risks in 2018, while the Banque de France launched a voluntary ex- ercise in 2020 (Alogoskoufis et al. 2021). In 2021, the Bank of England followed with a stress test for major UK institutions, covering both transition and physical risks, and the EBA initiated an EU-wide pilot climate exercise that same year. Outside Europe, central banks in Canada (2020), Australia (2021), Singapore (2022) and the United States (2023) also con- ducted climate stress tests (AFME 2023). Moreover, over the past few years, more than 100 central banks and financial supervisors have joined the Network for Greening the Financial System (NGFS), a forum aimed at enhancing supervisory practices for managing climate risks and developing climate scenarios (NGFS 2024a).15 This trend once again highlights the growing importance of climate risks for regulatory authorities and supervisors. However, integrating climate risks into the broader risk management framework is particu- larly challenging because climate risks act as risk drivers for existing risks, rather than being a separate, stand-alone type of risk. For example, both physical and transition risks can impact institutions by reducing borrowers’ ability to repay loans or by affecting the pricing of different asset classes (e.g., equities, bonds and commodities), thereby affecting loan books and trading books (Acharya et al. 2023). As illustrated in Figure 1, climate risks therefore materialize through “conventional” risk types such as credit, market and operational risks, and can affect both their likelihood and severity in different ways. As a result, incorporating climate considerations into risk management requires a nuanced, yet broad and comprehen- sive approach. This also requires a deep understanding of how climate risks interact with and amplify conventional risks, as well as evaluating where climate-related factors may be inadequately addressed in current risk models and processes. 15We discuss some of the NGFS scenarios in Section 4.3.4.2. 10 In this context, regulators and supervisors have emphasized that climate risk integration should focus on identifying and monitoring those risks that have, or may have, a material impact on institutions’ financial health — i.e., their financial position (balance sheet and capital resources), financial performance (profits and losses) and liquidity position (BaFin 2020, BIS 2022b). In essence, the primary objective is to raise awareness among institutions about climate risks and their transmission channels, particularly those that manifest through conventional risk types and ultimately affect balance sheets, profitability and liquidity. In terms of conventional risks to focus on, the EBA has particularly stressed the relevance of credit risk in the context of climate risk integration for EU institutions (EBA 2023). This emphasis stems from the fact that their activities and capital requirements are primarily driven by credit risk, as opposed to market or operational risks (EBA 2023). As a result, this study pays particular attention to the relationship between credit and climate risks, especially when discussing Pillar 1 considerations. 11 F ig ur e 1: C lim at e ris k tr an sm iss io n ch an ne ls. T he fi gu re il lu st ra te s th e ch an ne ls th ro ug h w hi ch c lim at e ris ks m ay a ffe ct p ru de nt ia l r isk t yp es a nd , in t ur n, t he fi na nc ia l p os iti on , fi na nc ia l p er fo rm an ce a nd li qu id ity p er fo rm an ce o f i ns tit ut io ns . B as ed o n B aF in ( 20 20 ), EB A ( 20 22 b) , B IS ( 20 22 b) , A ch ar ya e t al . ( 20 22 ). N ot e th at t he r isk s in cl ud ed in t he t ax on om y ar e no t ex ha us tiv e. 12 2.2 Literature review What are the motivations for supervisors and credit institutions to integrate climate risks into their prudential and risk management frameworks? In this section, we discuss the growing empirical literature on both the relevance and the implications of climate risks for these actors.16 While the scope of their involvement remains a subject of ongoing debate (see e.g., Hellwig 2022, Demekas & Grippa 2022 and Demekas & Grippa 2024), central banks are key actors in the discussion on climate-related risk management. Climate risks are becoming increasingly important to these institutions on two fronts: price stability and financial stability. In the euro area, the ECB has a primary mandate to pursue price stability by ensuring that inflation remains low, stable and predictable.17 An emerging strand of literature shows that climate risks can contribute to inflationary pressures, making these developments relevant for the ECB to monitor and incorporate into its inflation forecasts. Thus, the ECB’s price stability objective is seen as part of the justification for the climate-related policies it has introduced in the last years (Faccia et al. 2021, van ‘t Klooster & de Boer 2022).18 On the one hand, climate change, in particular its physical manifestations, may disrupt supply chains and production processes and reduce agricultural yields, all of which have important implications for inflation dynamics and economic output (Acharya et al. 2023). For example, Pankratz & Schiller (2022) show that weather shocks (extreme heat and floods) negatively affect the financial performance of suppliers and that the financial impact of these shocks is transmitted to customers through existing supply chain links. Faccia et al. (2021) provide evidence for the impact of extreme temperatures on price development as captured by different measures of prices, including consumer and producer prices. Ciccarelli et al. (2024) assess the four largest euro area economies (Germany, France, Italy and Spain) and find that hotter summers lead to upward pressures in terms of food inflation and service inflation. De Winne & Peersman (2021) report that extreme weather events, such as droughts and heatwaves, influence fluctuations in global agricultural commodity prices and impact economic activity of both advanced and developing countries. Finally, Bilal & Känzig (2024) find that every additional 1°C rise in the average temperature on earth means a 12 percent 16For extensive literature overviews on the effects of climate risks on financial market participants, see de Bandt et al. (2023), Eren et al. (2022) and Giglio et al. (2021). 17See Art. 127 (1) of the Treaty on the Functioning of the European Union (TFEU). Price stability is currently defined by a 2% inflation target over the medium term. 18The ECB’s second mandate to support broader economic policies by and in the EU is also partly seen as an element of this justification, although we do not discuss this mandate in detail (van ‘t Klooster & de Boer 2022). 13 hit to global Gross Domestic Product (GDP), arguing that the macroeconomic effects of climate change have so far been largely underestimated. On the other hand, climate change may also pose a threat to the overall stability of the financial sector. As noted earlier, climate risks represent a new risk driver that can affect institutions’ financial position (balance sheet), financial performance (income statement - P&L) and liquidity position.19 In this context, these risks are not only of concern to central banks in their role as supervisors, but also a factor that institutions are increasingly required to take into account. In terms of mechanisms, there are three key channels through which climate risks are expected to affect institutions: the credit, market and liquidity risk channels (Acharya et al. 2023, de Bandt et al. 2023). To illustrate the credit risk channel, climate risks may affect loan books, with both transition and physical risks reducing borrowers’ ability to repay their outstanding loans and increasing their default rates (Acharya et al. 2023, Cepni et al. 2024). For example, this may hold for loan exposures to borrowers in sectors with high carbon emissions, facing higher operating costs as a result of regulatory interventions. Similarly, extreme weather events may lead to the destruction or depreciation of physical assets, resulting in higher default probabilities for corporate loans and mortgage loans (Cepni et al. 2024). Ultimately, this may potentially translate into increased loan losses for the banks (Brei et al. 2019). Nevertheless, the empirical evidence on the credit risk channel is mixed. In particular, while transition and physical risks are a source of financial risk for borrowers (see e.g., Ehlers et al. 2022, Seltzer et al. 2020), the effects on institutions’ financial position and financial performance are contrasted. This may be due to factors such as composition and maturities of loan portfolios, but also to differences in the types of institutions (large, well- diversified institutions versus smaller and more local ones) and the wide range of climate risks considered (carbon taxes, heat, flood, etc.). So far, a large strand of the literature has focused on the link between physical risks and institutions’ profitability. For example, Blickle et al. (2021) report that weather disasters in the last quarter century had limited effects on U.S. banks’ performance. The authors find that loan losses tend to be offset by an increase in loan demand after the disasters, boosting profits at larger banks. On the contrary, Schubert (2021) reports that flood shocks negatively affect U.S. banks’ performance, as 19To understand the potential effects of climate risks on institutions’ financial position and financial performance, it is useful to recall the structure of banks’ balance sheets and the link with their P&L. On the asset side, institutions typically hold cash, loans (loan book) and trading assets (trading book). On the liabilities side, institutions typically have deposits, short-term and long-term debt (e.g., obligations and commitments related to derivatives contracts) and equity. Net income (from the P&L) increases equity, while a net loss decreases it. 14 measured by the return on assets (ROA). Ge et al. (2022) and Sastry et al. (2024) observe that the underlying assets of U.S. institutions with large real estate portfolios in climate- vulnerable regions (e.g., coastal areas prone to sea-level rise) are at risk of increased losses due to rising trends in insurance premiums, insurance company withdrawals and a decline in the quality of insurance quality. Vollmar & Wening (2023) analyze the effects of heat on regional institutions in Germany and find a negative impact in terms of profitability (ROA) for institutions servicing industries in heat-affected regions compared to institutions in less heat-affected regions. On the transition risks side, while there is evidence that institutions are managing these risks, including by adapting their lending, re-balancing their portfolios within borrower industries and adjusting their pricing (see e.g., Chava 2014, Delis et al. 2019, Ivanov et al. 2023, Martini et al. 2023), their impact on bank performance is less evident. For example, Jung et al. (2023) find that the credit exposures of U.S. institutions to transition risks are modest and, on average, do not exceed 14% of their loan portfolios under several scenarios. Referring to an ECB study, the authors also underscore that “expected losses of European banks’ credit portfolios are minimal, around 0.7% of the total loan exposure under both the accelerated and the delayed transition scenarios” (Jung et al. 2023, Emambakhsh et al. 2023). While the credit risk channel is widely regarded as the most relevant one for banks (EBA 2023), market risk is considered the second most critical. Market risk refers to the potential for losses arising from fluctuations in market prices, including the risk of losses from trading assets held in the trading book and the potential for losses due to foreign-exchange risk or commodity risk from assets in the banking book (EBA 2023). Bonato et al. (2023a) analyze the role of climate risks in forecasting realized volatility of exchange rate returns of major fossil fuel exporters and find that climate risks (temperature, wind, precipitation) have incremental out-of-sample predictive value for exchange rate volatility. In a follow- up study, Bonato et al. (2023b) report that climate risks have predictive value for realized stock market volatility at the U.S. state-level. Bertolotti et al. (2019) analyze the impact of physical events on the stocks of U.S. electric utilities and find significant market price reactions after hurricanes. Furthermore, since climate risks are anticipated to affect future cash flows, market risk also relates to the effects of such risks on the present-day values of financial assets (Acharya et al. 2023). For example, if climate risks are considerably under- priced, changes in investor perception may lead to re-pricing of assets and, potentially, to losses for asset holders such as banks (Demekas & Grippa 2024). A large strand of the literature therefore focuses on the extent to which climate risks explain stock market returns (see e.g., Bolton & Kacperczyk 2021, Ardia et al. 2023, Aswani et al. 2023), are priced on stock and bond markets (e.g., Huynh & Xia 2020, Painter 2020, Acharya et al. 2022) and 15 affect investor beliefs (e.g., Bakkensen & Barrage 2021, Ceccarelli & Ramelli 2024). Finally, in addition to credit and market risks, there is also a potential liquidity risk channel to consider (Acharya et al. 2023). However, this channel appears not very well understood to date and research remains limited (de Bandt et al. 2023). So far, there is empirical evidence that climate risks may lead to both decreases and increases in deposit withdrawals after major disasters. For example, Choi et al. (2022) find that institutions with a poor environmental reputation are more likely to face a decline in their branch deposits in regions exposed to severe physical risks. In some cases, climate risks, particularly physical risks, are shown to increase loan demand, which offsets losses but puts pressure on liquidity buffers. Brei et al. (2019) report that following natural catastrophes (hurricanes) institutions face large deposit withdrawals and experience a negative funding shock to which they respond by reducing their loan supply and drawing on their liquid assets. Koetter et al. (2020) study the role of German local banks in the context of natural disasters (flooding) and find that local banks that are domiciled in unaffected counties but are exposed to disaster-ridden firms in affected counties lend 3% more in the post-flood period compared with unexposed local banks. Cortés & Strahan (2017) find that institutions are incentivized to cut lending from markets unaffected by natural disasters in order to have enough liquidity to support disaster-affected markets facing increased credit demand. Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. 3 Research method and sample 3.1 Semi-structured interviews We adopt an empirical approach based on semi-structured interviews with risk experts, combined with insights from archival data, primarily from Pillar 3 reports. This approach allows us to gather information on both the climate-related risk management practices of institutions and the experiences and views of practitioners. We also use data from the Pillar 3 reports to guide and refine our interviews. This data proved particularly helpful in gaining an overview of the climate-related risk management processes in place. Finally, we analyze supervisory and regulatory guidelines to gain a deeper understanding of the expectations and recommendations set by supervisors and regulators. We conducted semi-structured interviews with risk experts mainly from significant institu- tions that are directly supervised by the ECB and subject to the CRR. The ECB maintains and regularly updates a public list of these institutions (ECB 2024d). For our study, we used the list as of January 1, 2023 and selected institutions classified as “credit institutions”. We particularly focused on institutions established in Germany due to our geographical prox- imity and personal contacts, which contributed to a higher response rate. As of January 1, 2023, the ECB’s list of supervised entities included 18 SIs established in Germany. Our sam- ple covers experts from 12 of these institutions (approximately 67%). Furthermore, we also targeted experts of large credit institutions established in France and the Netherlands. Our sample covers three of these institutions. In addition, our sample includes one expert from a Less Significant Institution (LSI) in Germany that is preparing for ECB supervision and thus actively integrating climate-related risks.20 Finally, we conducted one interview with an expert in PD modeling and stress testing from a large German institution that is exempt from the scope of the CRR, but is subject to key banking regulations and supervised by the German banking supervisory authorities. To sum up, our sample includes 17 institutions, of which 15 are credit institutions currently directly supervised by the ECB. We identified respondents through LinkedIn and via personal recommendations. Employees of relevant institutions were screened on LinkedIn using keywords such as “climate risk”, “ESG risk”, “credit risk”, “credit risk modeling”, “risk controlling”, “risk management”, “stress testing”, “scenario analysis”, “IRB”, “rating”, “PD model”, “environmental risk”. This process led to 62 potential respondents being contacted. Overall, 30 risk experts par- 20LSIs have been explicitly invited by the ECB to apply its climate-related expectations in a manner that is proportionate to the nature, scale and complexity of their activities (ECB 2020b). Furthermore, 79 LSI participated in the 2022 thematic review on climate-related and environmental risks (ECB 2022e). 17 © The Author(s) 2025 A. Auzepy and C. E. Bannier, Integrating Climate Risks in Bank Risk Management and Capital Requirements, https://doi.org/10.1007/978-3-658-47061-6_3 https://crossmark.crossref.org/dialog/?doi=10.1007/978-3-658-47061-6_3&domain=pdf ticipated, which corresponds to a high response rate of 48%. The breakdown of respondents is presented in Table 2. To maintain full anonymity, the names of the respondents and their institutions are not disclosed. Furthermore, we deliberately use the pronoun “they” instead of “she” or “he”. Each respondent is assigned a “respondent ID”, and each institution receives an “institution ID”, which are used to present our findings. As can be seen, our sample includes experts from various risk-related functions, offering different perspectives on climate risk integration, covering both Pillar 1 and Pillar 2 aspects. Furthermore, our sam- ple includes a wide range of large private commercial institutions (53%), public institutions (35%) and cooperative institutions (12%), with total assets ranging from EUR 30 billion to over EUR 1,000 billion. The decision to use semi-structured interviews, rather than a traditional survey with a set of pre-defined answers, provided several important advantages for our research. First, this approach allowed us to ask detailed questions about climate risk integration and thus better capture the complexity of the topic at hand. In addition, by allowing respondents to freely express their thoughts and views, we encouraged a more open and nuanced discussion, shed- ding light on the challenges currently faced. Second, the flexibility of the interviews enabled us to tailor questions to each respondent’s specific background, responsibilities and exper- tise. While we adhered to a core set of questions, some respondents had specific expertise in areas such as stress testing, while others had broader risk management roles, allowing us to explore certain topics in more depth in individual interviews. Importantly, we carefully se- lected respondents to ensure they had substantial expertise in their respective fields. Third, the interviews allowed respondents to introduce additional topics or concerns. This approach enabled us to incorporate insights from earlier interviews into subsequent ones, integrating emerging themes as they arose (Hummel & Bauernhofer 2024). The interviews were conducted in German or French and structured around open-ended questions. Each interview lasted approximately 60 minutes. The interviews were conducted between February and November 2024. Prior to the interviews, respondents were asked for their consent to be recorded using an audio recorder. We transcribed the interviews vebatim, resulting in about 500 pages of text. Key excerpts from the interviews were later translated into English. Thematic analysis was undertaken based on the interview transcripts (Braun & Clarke 2006). Using the software program MAXQDA Plus 2020, we carefully examined each interview transcript to perform qualitative data coding. This process allowed us to identify significant themes and refine our conceptual framework, presented in Section 3.3. This framework serves as the foundation for the structure of this study. 18 R es po nd en t ID R es po nd en t ex pe rt is e an d/ or ro le In st it ut io n ID T ot al as se ts C ou nt ry A 1 D ire ct or R eg ul at or y Po lic y, C ap ita l & L iq ui di ty A EU R 3 0- 50 b n G er m an y A 2 Se ni or S tr es s Te st in g Ex pe rt A EU R 3 0- 50 b n G er m an y A 3 R isk E xp er t St re ss T es tin g Ex pe rt A EU R 3 0- 50 b n G er m an y B1 D ire ct or R isk M et ho ds , R isk C on tr ol lin g B EU R 3 0- 50 b n G er m an y B2 M an ag er R isk M an ag em en t B EU R 3 0- 50 b n G er m an y C 1 Ex ec ut iv e D ire ct or , C lim at e R isk C EU R 5 0- 75 b n G er m an y C 2 Ex pe rt C re di t R isk a nd M ar ke t C ap ita l C EU R 5 0- 75 b n G er m an y D 1 Se ni or P ro je ct M an ag er , G ro up C re di t R isk M an ag em en t D EU R 3 00 -5 00 b n G er m an y D 2 Ex pe rt R isk M an ag em en t D EU R 3 00 -5 00 b n G er m an y E1 H ea d of D ep ar tm en t, O ve ra ll Ba nk R isk E EU R 5 0- 75 b n G er m an y F1 D ire ct or S tr at eg ic A na ly tic s F > EU R 1 ,0 00 b n G er m an y G 1 H ea d of S tr es s Te st in g M et ho do lo gi es a nd M od el s G > EU R 1 ,0 00 b n Fr an ce G 2 Se ni or R isk M an ag er , P hy sic al R isk L ea d G > EU R 1 ,0 00 b n Fr an ce H 1 Ex pe rt E SG R isk s H EU R 5 00 -1 ,0 00 b n G er m an y H 2 R isk M an ag er G ro up R isk C on tr ol lin g H EU R 5 00 -1 ,0 00 b n G er m an y H 3 H ea d of D ep ar tm en t St re ss T es ts a nd S ce na rio A na ly sis H EU R 5 00 -1 ,0 00 b n G er m an y H 4 St re ss T es tin g Ex pe rt H EU R 5 00 -1 ,0 00 b n G er m an y H 5 Ex pe rt R isk C on tr ol lin g, IC A A P, E SG R isk M an ag em en t H EU R 5 00 -1 ,0 00 b n G er m an y I1 Ex pe rt R at in g M et ho do lo gy , R isk C on tr ol lin g & E SG R isk I EU R 1 50 -3 00 b n G er m an y I2 G ro up M an ag er R isk C on tr ol lin g I EU R 1 50 -3 00 b n G er m an y J1 H ea d of S tr at eg ic R isk C on tr ol J EU R 3 0- 50 b n G er m an y J2 D ire ct or E SG R isk J EU R 3 0- 50 b n G er m an y K 1 H ea d of M od el R isk M an ag em en t an d Va lid at io n K EU R 7 5- 10 0 bn G er m an y K 2 Ex pe rt R isk C on tr ol lin g, N on -F in an ci al R isk a nd S us ta in ab ili ty R isk K EU R 7 5- 10 0 bn G er m an y L1 ES G S tr es s Te st in g Ex pe rt L EU R 3 0- 50 b n G er m an y M 1 Ex pe rt R isk C on tr ol lin g, C re di t R isk S tr es s Te st a nd P D M od el in g M EU R 5 00 -1 ,0 00 b n G er m an y N 1 M an ag in g D ire ct or R isk C on tr ol lin g N EU R 1 50 -3 00 b n G er m an y O 1 G lo ba l H ea d R isk C on tr ol O EU R 1 00 -1 50 b n G er m an y P1 Te am L ea d ES G R isk M an ag em en t P EU R 5 00 -1 ,0 00 b n N et he rla nd s Q 1 Ex pe rt E SG R isk a nd C lim at e R isk Q EU R 1 50 -3 00 b n N et he rla nd s 19 T ab le 2: Sa m pl e co m po sit io n. 3.2 Archival data We reviewed the 2023 Pillar 3 reports from the institutions in our sample. Although Pillar 3 reports (also known as regulatory disclosure reports)21 were not the focus of specific in- terview questions (e.g., we did not ask the respondents to discuss the preparation of these reports or the challenges their institutions face in disclosing climate-related information), we used the content of these disclosures to gain insights into the institutions’ climate-related risk management practices and to guide our interview questions. In addition, we used the qualitative information from the disclosures to complement our interview findings and of- fer a more comprehensive perspective of the risk management approaches employed by the institutions in our sample. Under the CRR, Pillar 3 disclosures are mandatory disclosures for significant institutions (see Art. 431 of the CRR) and must be published at least annually (see Art. 433 of the CRR). The content of the disclosures is also largely determined by the CRR, which specifies technical criteria for the disclosure of risk management objectives and policies (see Art. 435 of the CRR), own funds (see Art. 437 of the CRR), capital requirements (see Art. 438 of the CRR), capital buffers (see Art. 440 of the CRR) and leverage (see Art. 451 of the CRR), among others. Thus, Pillar 3 reports contain important information on both Pillar 1 and Pillar 2 aspects. In addition to the above, these reports are typically organized by risk type (e.g., credit risk, market risk, operational risk, liquidity risk, reputational risk, model risk, etc.) and provide specific information on the institutions’ exposure to and management of each risk type. An important development is the requirement for Pillar 3 reports to include detailed in- formation on ESG risks,22 with a particular focus on physical risks and transition risks. Institutions whose securities are admitted to trading on a regulated market in the EU have to comply with the disclosure requirements pursuant to Art. 449a of the CRR.23 In Jan- uary 2022, the EBA published the final Implementing Technical Standard (ITS) (EBA ITS 2022/01) on prudential disclosures on ESG risks, which defines the nature of these disclosure requirements and forms the basis for Implementing Regulation (EU) 2022/2453. In particular, this ITS specifies the set of qualitative and quantitative information to be disclosed by institutions regarding ESG risks. Qualitative disclosures include information 21Pillar 3 of the Basel framework focuses on market discipline through prescribed public disclosures. 22Climate risks are typically treated as a subset of ESG risks and fall under the environmental pillar. 23While Art. 449a of the CRR itself does not specify the content of the ESG risk disclosures, it refers to Art. 98 (8) of the CRD, which, in conjunction with Art. 434a of the CRR, gives the EBA a mandate to develop an Implementing Technical Standard (ITS). 20 on business strategy, governance and risk management related to environmental risks, while quantitative disclosures cover exposures to transition risks (templates 1 to 4), exposures to physical risks (template 5) and mitigating actions (templates 6 to 10).24 3.3 Conceptual framework Analyzing the interaction between climate risks and conventional risk types, as well as the integration of climate-related considerations, requires an understanding of the theoretical foundations of institutions’ capital requirements and risk management. This brings the banking prudential framework (the “Basel framework”) into focus (Redondo & Aracil 2024, EBA 2021, Gruenewald et al. 2023).25 The Basel framework, developed by the BCBS, is an internationally agreed set of measures and standards for prudential regulation. The EBA participates as an observer in the Basel Committee and plays a crucial role in implementing the Basel standards within the EU (EBA 2024). Far from being a set of loose guidelines, the Basel framework is directly applied in the EU through the CRR and the CRD.26 The institutions in our sample operate under both the CRR and the CRD, making the Basel framework applicable to them. The framework itself is structured around three fundamental pillars, which we explore further in this study. Specifically, we use key elements of Pillar 1 and Pillar 2 to guide our discussion of climate risk integration, as illustrated in Figure 2. 24See Implementing Regulation (EU) 2022/2453 of 30 November 2022 amending the implementing tech- nical standards laid down in Implementing Regulation (EU) 2021/637 as regards the disclosure of environ- mental, social and governance risks. 25The Basel I Accord, introduced in 1988, established a framework for minimum capital requirements based on risk-weighted assets to ensure institutions held adequate capital to cover credit and market risks. In 2004, the Basel II Accord expanded this framework by adding operational risk alongside credit and market risks, introducing three pillars: minimum capital requirements (Pillar 1), the supervisory review process (Pillar 2), and market discipline through transparency (Pillar 3). The Basel III reforms, developed in response to the 2007-2008 financial crisis, aim to enhance institutions’ resilience to financial and economic shocks by further strengthening risk management and addressing liquidity and funding risks. 26See Regulation (EU) No 575/2013 on prudential requirements for credit institutions and investment firms and Directive 2013/36/EU on access to the activity of credit institutions and the prudential supervision of credit institutions and investment firms. 21 F ig ur e 2: C on ce pt ua l f ra m ew or k. O w n ill us tr at io n. 22 There are important differences between the two pillars, leading to different approaches, but also challenges, in integrating climate risks. Pillar 1 of the Basel framework focuses on minimum capital requirements for credit, market and operational risks. Capital requirements are determined as fixed percentages of RWA. The RWA is a measure of an institution’s assets adjusted for the level of risk inherent to those assets. This level of risk is partly determined by a risk weight. The key idea is therefore to capture the relative riskiness of various assets on an institution’s balance sheet (Holscher et al. 2022). The calculation of RWA is governed by a set of rules applied uniformly across institutions, aiming to establish consistent regulatory standards for risk-based capital requirements. An important metric under Pillar 1 is the Common Equity Tier 1 (CET1) ratio, which compares an institution’s CET1 capital to its total RWA (i.e., the sum of credit RWA, market RWA and operational RWA). The primary goal of minimum capital requirements, along with regulatory buffers, is to ensure that institutions can absorb losses beyond expected levels and thus maintain an amount of capital necessary to be regarded as a viable going concern by creditors and counterparties (Holscher et al. 2022). Expected losses are typically covered ex-ante by risk-based pricing and through loss provisioning under International Financial Reporting Standard 9 (IFRS 9). Expected credit loss provisions are deducted from capital (Krüger et al. 2018).27 In recent years, both the BCBS and the EBA have published discussion papers on incorpo- rating climate risks into the Pillar 1 framework (EBA 2022b, BIS 2022a). However, unlike credit, market and operational risks — where clear rules for calculating RWA are already well-established — neither the BCBS nor the EBA have proposed new, specific capital re- quirements, such as dedicated risk weightings, for assets exposed (or less exposed) to climate risks under this pillar.28 Instead, the emphasis has been on how these risks can be more consistently integrated into the existing frameworks and models, with a focus on credit risk and, to a lesser extent, market risk. For example, the ECB issued in 2024 an updated guid- ance on internal models, requiring institutions to incorporate material climate-related risk drivers into their existing Pillar 1 models for credit and market risks (ECB 2024a). Integrating climate risks into the existing Pillar 1 framework presents challenges due to several factors, including the framework’s rigidity, its short-term focus and its reliance on 27More specifically, expected credit losses provisions are deducted from accounting equity, which is the starting point for regulatory capital, e.g., CET1. Provisions are not covered in detail in this study, but we touch upon some aspects in Section 4.3.5.3 and Section 4.5.3. 28For example, the EBA has so far ruled out the introduction of a green supporting factor, which would lower capital requirements for environmentally sustainable exposures by reducing risk weights or adjusting the RWA calculation (EBA 2023). A similar instrument already exists under Pillar 1, however for small and medium-sized enterprises and infrastructure projects, allowing institutions to apply a discount factor, resulting in a downward adjustment in risk weights for SME and infrastructure-related exposures meeting certain eligibility criteria (EBA 2023). 23 historical data for calculations (Gruenewald et al. 2023). In addition, a less explored issue in the literature is the potential for double counting. For example, internal models under Pillar 1, which are designed to quantify the specific risk of an exposure, already incorporate various risk factors that may implicitly account for some climate-related risks (EBA 2022b). This broader context raises important questions about where climate risks may already be captured implicitly and where a more explicit approach is necessary. In contrast, Pillar 2 of the Basel framework encourages institutions to implement individual risk management strategies for identifying, measuring and monitoring their risks beyond minimum regulatory requirements. Key components of this pillar include the ICAAP and the SREP. The ICAAP is characterized by an internal capital adequacy concept comprising an internal capital adequacy calculation, stress tests and capital planning processes (BaFin 2018). It is closely linked to the development of both business and risk strategies, as well as risk management processes (BaFin 2018). Supervisors evaluate the soundness of an in- stitution’s ICAAP and overall risk management framework through the SREP. Based on their assessment, supervisors may prescribe or recommend the institution to hold additional capital (e.g., via the Pillar 2 Requirement or Pillar 2 Guidance) or take other corrective actions. Compared to Pillar 1, the treatment of climate risk integration is more concrete for Pillar 2 (Smoleńska & van ’t Klooster 2022, Gruenewald et al. 2023). The ECB’s 13 supervisory expectations regarding the integration of climate risks into risk management practices have placed significant emphasis on Pillar 2, particularly on the integration of these risks into the ICAAP. Notably, the ECB highlighted that the Guide on climate-related and environmental risks should be considered alongside the ECB guide to the ICAAP, indicating that these two guides are complementary (ECB 2018, 2020b). As the ICAAP is an internal and individual process, this also suggests that the integration of climate risks requires tailored approaches across institutions. Finally, the treatment of climate risk under Pillar 2 is also influenced by the integration of climate risks into the SREP, where the ECB now plays a critical role in reviewing how institutions manage these risks (Gruenewald et al. 2023). What does the combined consideration of Pillar 1 and Pillar 2 imply for the integration of cli- mate risks? Standardizing the treatment of such risks under Pillar 1 would likely contribute to establishment of consistent regulatory standards across all institutions, potentially result- ing in higher minimum capital requirements for institutions exposed to climate risks. As Pillar 1 capital calculations often serve as the basis for further risk assessments under Pillar 2 (BaFin 2018), a more systematic integration of climate risks into Pillar 1 may indirectly raise the baseline requirements and also impact capital considerations under Pillar 2. 24 The current guidelines and expectations for both Pillar 1 and Pillar 2 provide institutions with large flexibility in terms of implementation. This highlights the need to examine how individual institutions are integrating climate risks and the specific challenges they face in doing so. Furthermore, where climate risks act as drivers of other risk types, an important question emerges: to what extent, and how, are climate risks being factored into institutions’ risk management and capital considerations? If climate risks are (or are to be) incorporated, it is crucial to understand the mechanisms through which this integration occurs, as well as how institutions and supervisors determine the capital resources needed. To support this analysis, Figure 2 illustrates the conceptual framework of our study and outlines the key areas that were the focus of our interviews. Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. 4 Results 4.1 Pillar I: Integrating climate risks into the standardized ap- proach (SA) 4.1.1 Background and rationale We start by discussing important features of Pillar 1 and explore both potential and existing approaches for integrating climate risks. As noted in Section 3.3, Pillar 1 focuses on the establishment of minimum capital requirements for three primary types of risk: credit risk, market risk and operational risk. Under this framework, capital requirements are calculated as fixed percentages of RWA based on these risk types (Holscher et al. 2022).29. The EBA describes climate risks as a particularly relevant driver of credit risk, which in turn accounts for the largest share of institutions’ total capital requirements (compared to the capital requirements resulting from market and operational risks) (EBA 2022b).30 As a result, our analysis of Pillar 1 focuses primarily on the interaction between climate and credit risks. In this section, we delve into a key approach to credit risk: the standardized approach. Given its importance in determining capital requirements for credit risk under Pillar 1, we explore the main channel for integrating climate risks into this approach and address the challenges involved. As of June 2023, the standardized approach represented on average 54% of all credit risk-weighted exposure amounts in the EU (EBA 2023). Several participants (A1, B1) pointed out the increasing regulatory shift towards this approach over internal models (further discussed in Section 4.2). A notable example of this shift is the introduction of an “output floor”, applicable as of January 2025, designed to prevent institutions from using internal models to significantly lower their capital requirements compared to the standardized approach (Smoleńska & van ’t Klooster 2022). Specifically, with the introduction of the output floor, the RWA calculated using internal models will not be allowed to fall below 72.5% of those calculated by the standardized approach, limiting the potential capital benefit from internal models to 27.5%. The respondents observed that this restriction makes using internal models more challenging, which is likely to drive wider adoption of the standardized approach in the coming years. Under the standardized approach, the RWA resulting from credit risk are calculated as the 29For example, all institutions must hold capital equivalent to at least 8% of total RWA. 30The EBA notes that “credit risk is considered the most relevant part of the prudential framework, with RWAs attributable to credit risk accounting for over 80% of total RWAs” of credit institutions in the EU (see p.24, EBA 2022b). 27 © The Author(s) 2025 A. Auzepy and C. E. Bannier, Integrating Climate Risks in Bank Risk Management and Capital Requirements, https://doi.org/10.1007/978-3-658-47061-6_4 https://crossmark.crossref.org/dialog/?doi=10.1007/978-3-658-47061-6_4&domain=pdf product of the exposure value (net of credit risk mitigation instruments) and a specific risk weight, as illustrated in Figure 3. A key feature of this approach is the reliance on external credit assessments (i.e., external credit ratings) to determine the risk weight (see Art. 135 of the CRR). These external credit assessments reflect an agency’s opinion about the issuer’s (i.e., the borrower’s) ability to meet its financial obligations on time and in full. Each external credit assessment is assigned to a Credit Quality Step (CQS). The CQS, together with the exposure class, determine the risk weight. The exposure class depends on the type of issuer, such as corporate, retail or public sector entities (see Art. 112 of the CRR). Figure 3: Calculation of risk-weighted assets (RWA) under the standardized approach for credit risk. Based on Regulation (EU) No 575/2013 on prudential requirements for credit institutions and investment firms (CRR). Only credit assessments from external credit rating agencies explicitly approved by the reg- ulator are permitted.31 Under the standardized approach, institutions have to formally nominate a rating agency whose credit ratings they will use. Consequently, the degree to which climate risks are integrated into the standardized approach — and thus into the calcu- lation of the risk weight and, indirectly, the minimum capital requirements for credit risk — largely depends on the selected agency, its methodologies and its approach to incorporating climate-related factors into credit risk assessments. The EBA has already extensively discussed this issue, identifying two primary challenges in this context (EBA 2022b, 2023). First, the degree to which credit rating agencies incorporate climate risks into their credit assessments varies significantly from one to the other, leading to an inconsistent treatment of such risks across the industry. Second, the EBA also pointed to a lack of transparency in the methodologies used by credit rating agencies. In particular, the specific ways in which these agencies incorporate climate risks into their ratings are often unclear, making it difficult for institutions to fully understand whether and how climate- related factors influence their final rating decision. 31A list of such rating agencies is published by the EBA and regularly updated (see Art. 135 of the CRR). 28 4.1.2 Credit ratings and dialogue with rating agencies In our sample, all but one institution (A) rely on internal models to calculate capital re- quirements for credit risk. In this section, we build on the EBA’s earlier observations by adding two key insights shared by a respondent (A1) from this institution and exploring the challenges that were highlighted. First, the respondent stressed that large, globally operating institutions heavily rely on the three major credit rating agencies — Moody’s, Standard & Poor’s and Fitch. This expert noted that their institution uses ratings from all three, rather than opting for smaller, less established agencies that have more limited coverage. Currently, the broad coverage provided by these agencies is considered to outweigh concerns about their treatment of climate risks. Therefore, the extent to which climate risks are integrated largely depends on how these three leading agencies, in particular, incorporate such factors into their ratings: “If you are a globally operating bank, then there is a certain expectation in the market that you will use these rating agencies [Moody’s, Standard & Poor’s and Fitch]. But if you are a specialized bank, such as the Umweltbank, where sustainability is part of the core strategy, then I could imagine that you might try to take a closer look at the sustainability criteria applied by the rating agencies and their coverage of climate risks. In this case, you might also select a smaller agency with a stronger sustainability focus.” (A1) At this stage, it is also important to note that the European Securities and Markets Authority (ESMA), which oversees the rating approval process, conducted an assessment in 2019 of the extent to which credit ratings take into account ESG factors (ESMA 2019). The ESMA concluded that such factors, including climate-related factors, are taken into account in credit ratings, but that their importance and incorporation varies across asset classes. In addition, the ESMA advised against amending the regulation governing credit rating agencies to more explicitly account for ESG-related factors in credit risk assessments (ESMA 2019). Our respondent also pointed out that direct dialogue between institutions and rating agencies regarding their methodologies is rather rare. In practice, external ratings are imported into internal systems that assign the appropriate risk weights. As a result, this respondent believed it is unlikely that an institution could influence a rating agency’s methodology to better incorporate climate risks unless the agency is already doing so. An exception to this may arise with the implementation of the CRR III, which is due to take effect in January 2025. Specifically, the CRR III amends Article 138 of the CRR, 29 restricting the use of credit ratings that include assumptions of implicit government support. While this regulation applies to credit ratings for certain financial institutions rather than corporates, the respondent viewed it as an opportunity for institutions to re-examine rating methodologies more broadly: “New regulations could drive a broader discussion on rating methodologies. A clear example is the implementation of CRR III, where institutions, when assessing other institutions as counterparties, will only be allowed to use ratings that exclude embedded government support. This is likely to prompt institutions to engage with their credit rating agencies and take a closer look at the underlying methodologies, as such ratings are currently quite limited.” (A1) Pillar I: Integrating climate risks into internal models (IRB) 4.2.1 Background and rationale Most institutions in our sample use internal models under Pillar 1 to calculate RWA from credit exposures. This section outlines the key features of these internal models for credit risk, while the following sections consider approaches to integrating climate risks and discuss the main challenges highlighted by our respondents. With internal models and the Internal Ratings-Based Approach (IRB), institutions use their own modeling techniques and internal data to estimate key credit risk parameters. There are two types of internal models for credit risk: the Foundation Internal Ratings-Based Approach (F-IRB) and the Advanced Internal Ratings-Based Approach (A-IRB). Under the F-IRB, institutions model only the PD using their internal approach. In contrast, under the A-IRB, institutions can also estimate their own Loss Given Default (LGD) and Exposure at Default (EAD). Both approaches share the common feature that institutions develop their own PD models for credit risk exposures based on internal ratings, which help categorize borrowers by their risk of default. These models are subject to supervisory approval and have to fulfill a set of requirements set forth in the CRR. Figure 4 illustrates the calculation of RWA under the internal ratings-based approach. As shown, the credit risk parameters associated with the internal rating play a key role. The input parameters PD, LGD and Maturity (M) influence the risk weight and ultimately determine the RWA.32 32Note that several parameters, including PD, LGD, M and Credit Conversion Factors (CCFs), are involved in credit risk modeling (see EBA 2023). This study focuses on the integration of climate risks into internal ratings, particularly in relation to PD considerations. 30 4.2 Figure 4: Calculation of risk-weighted assets (RWA) under the internal ratings-based approach (IRB) for credit risk. Based on Regulation (EU) No 575/2013 on prudential requirements for credit institutions and investment firms (CRR). An important development introduced in the latest revision of the ECB guide to internal models is the expectation that institutions incorporate material climate risks in their internal models for credit risk (ECB 2024a). Specifically, the ECB stated that “where climate-related and environmental risks drivers are found to be relevant and material, institutions should include such risk drivers in their internal models approved for use for the calculation of own funds requirements for credit and market risk” (p.13, ECB 2024a). In addition, the EBA also commented on the integration of climate risks into internal models for credit risk and formulated specific recommendations (EBA 2023). In particular, the EBA noted that as the “impact of climate risks on defaults and loss rates becomes available, institutions should reflect these risks in their PD and LGD estimates through a re-development or recalibration of their rating systems in the long term” (p.58, EBA 2023). Overall, these expectations and recommendations have important implications, both in terms of Pillar 1 and Pillar 2 aspects. The direct implication is that institutions are required to test and evaluate whether climate-related risk drivers are a material predictor of credit risk within their portfolios. If they are, the ECB guide to internal models suggests that climate risks should be incorporated into internal ratings-based models and thus into PD modeling (and LGD modeling, depending on whether F-IRB or A-IRB is in place), requiring a com- prehensive review of the modeling process. In addition, it also raises important questions about how climate risks are to be incorporated, e.g., by adding risk factors quantitatively into the models or indirectly through the use of overrides and additional margins of conser- vatism. Institutions face further implications. For example, testing whether climate risks are a ma- terial predictor of credit risk requires access to comprehensive, high-quality data on relevant risk drivers integrated into a reference data set (EBA 2023). In cases where such quantitative 31 data is lacking, institutions have to establish processes and implement tools to collect the necessary climate-related information. These tools and processes are essential components of the risk identification, risk inventory and risk quantification processes under Pillar 2, as discussed in Section 4.3.2. While the ECB stresses the importance of integrating “material” climate risks into the models, it does not prescribe specific approaches for assessing mate- riality. As shown in Section 4.3, institutions have adopted different approaches to evaluate the materiality of climate risks. Finally, another important consideration, raised by several respondents, is that while in- stitutions may use internal models for credit risk, the models themselves may not always be developed in-house, but rather provided by external vendors. For example, they noted that many German “Landesbanken”, which are supervised by the ECB, rely on a rating tool developed by Rating Service Unit GmbH & Co. KG (RSU). Similarly, German savings institutions (“Sparkassen”) commonly rely on a tool developed by Sparkassen Rating und Risikosysteme GmbH (SR) for their rating systems. To ensure that these institutions inte- grate material climate risks into their capital considerations under Pillar 1, it is therefore essential to also assess the methodologies of these external model providers. In the following sections, we examine the approaches described by the respondents for inte- grating climate risks into internal ratings-based models and discuss the associated challenges and implications, with a primary focus on PD considerations. Figure 5 provides a simpli- fied representation of the rating process in the context of internal ratings-based approaches, illustrating the link between the rating and the final PD (see EBA (2017) for a detailed explanation of the PD estimation process). Ratings are typically based on a combination of quantitative and qualitative rating criteria or risk drivers, which we discuss in Section 4.2.3. One approach to integrating climate risks is to incorporate climate-related factors quantitatively, either as new risk drivers or through adjustments to existing rating criteria. Institutions may also apply overrides in the rating process when “individual circumstances, related to a given obligor or exposures, cannot be reasonably captured by the model” (p.12, EBA 2017). These overrides offer an indirect way of accounting for climate risks. In ad- dition, model recalibration may be necessary, meaning that the weights assigned to certain risk drivers may need to be adjusted to better account for climate risks as reflected in his- torical data. The final PD typically includes a MoC to account for uncertainties in the best estimate of the risk parameter (EBA 2017). This margin may also be applied to address data limit