EUROPEAN COMMISSION

Brussels, 27.10.2021

SWD(2021) 320 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the documents

Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Regulation (EU) No 575/2013 on prudential requirements for credit institutions as regards requirements for credit risk, credit valuation adjustment risk, operational risk, market risk and the output floor

Proposal for a DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Directive 2013/36/EU as regards supervisory powers, sanctions, third-country branches, and environmental, social and governance risks, and amending Directive 2014/59/EU

{COM(2021) 663 final} - {SEC(2021) 380 final} - {SWD(2021) 321 final}

Table of contents

1.    Introduction: Political and legal context    5

2.    Problem definition    12

2.1.    What are the problems and problem drivers?    12

2.1.1.    Deficiencies in the current framework for calculating risk-based capital requirements    12

2.1.2.    No dedicated capture of ESG risks in the prudential framework    18

2.1.3.    Inconsistency of powers and tools made available to supervisors across the Union    19

2.1.4.    Fragmentation and inefficiency in the disclosure of banks’ prudential information    21

2.2.    How will the problems evolve?    23

3.    Why should the EU act?    26

3.1.    Legal basis    26

3.2.    Subsidiarity: Necessity of EU action    26

3.3.    Subsidiarity: Added value of EU action    27

4.    Objectives: What is to be achieved?    27

4.1.    General objectives    27

4.2.    Specific objectives    28

5.    What are the available policy options?    29

5.1.    What is the baseline from which options are assessed?    29

5.2.    Description of policy options    30

5.2.1.    Improve the current framework for calculating risk-based capital requirements    30

5.2.2.    Dedicated capture of ESG risks in the prudential framework    40

5.2.3.    Improve the consistency in the application of supervisory powers    42

5.2.4.    Reduce disclosure costs and improve market access to bank prudential information    43

6.    What are the impacts of the policy options and how would they compare?    44

6.1.    Improve the current framework for calculating risk-based capital requirements    45

6.2.    Dedicated capture of ESG risks in the prudential framework    53

6.3.    Improve the consistency in the application of supervisory powers    55

6.4.    Reduce disclosure costs and improve market access to bank prudential information    58

7.    Preferred policy options    59

7.1.    Effectiveness    59

7.2.    Efficiency    60

7.3.    Coherence    61

7.4.    REFIT (simplification and improved efficiency)    62

8.    How will actual impacts be monitored and evaluated    63

Annex 1: Procedural information    65

1.    Lead DG, Decide Planning/CWP references    65

2.    Organisation and timing    65

3.    Consultation of the RSB    66

4.    Evidence, sources and quality    71

Annex 2: Stakeholder consultation    72

Annex 3: Who is affected and how?    77

Annex 4: Evaluation    89

Annex 5: Additional information about policy options    90

1.    Improve the current framework for calculating risk-based capital requirements    90

1.1. Credit risk framework – Standardised approach    90

1.2. Credit risk framework – Internal model approach    107

1.3. Market risk framework    117

1.4. Operational risk framework    124

1.5. Credit valuation adjustment risk framework    130

1.6. Minimum haircut floor framework for SFTs    135

1.7. Standardised approach for counterparty credit risk    143

1.8. Output Floor    147

Annex 6: Specific impacts of preferred policy options    156

1. Impact on administrative and operational costs    156

2. Impact on competitiveness    158

3. Impact on SMEs    169

4. Macro-economic costs and benefits analysis    172

Annex 7: Analytical methods    178

1. EBA impact analysis    178

2. ECB macroeconomic costs and benefits analysis    183

List of abbreviations

Term or acronym	Meaning or definition
AIRB	Advanced IRB
BA-CVA	Basic Approached for Credit Valuation Adjustment
BCBS	Basel Committee on Banking Supervision
CCB	Capital Conservation Buffer
CCyB	Countercyclical Capital Buffer
CfA	Call for Advice
CIU	Collective Investment Undertaking
CMU	Capital Markets Union
CRR	Capital Requirements Regulation
CRD	Capital Requirements Directive
CVA	Credit Valuation Adjustment
EAD	Exposure At Default
EBA	European Banking Authority
ECB	European Central Bank
EGD	European Green Deal
ESMA	European Securities Markets Authority
EU	European Union
EUCLID	European Centralised Infrastructure for Supervisory Data
FRTB	Fundamental Review of the Trading Book
GaR	Growth-at-Risk
GFC	Great Financial Crisis
GDP	Gross Domestic Product
GI	Gross Income
ILM	Internal Loss Multiplier
IRB	Internal Ratings Based
LDP	Low Default Portfolio
MRC	Minimum Required Capital
MS	Member State
OF	Output Floor
PD	Probability of Default
QIS	Quantitative Impact Study
RW	Risk Weight
RWA	Risk-Weighted Asset
SA-CCR	Standardised Approach for Counterparty Credit Risk
SA-CR	Standardised Approach for Credit Risk
SA-CVA	Standardised Approach for Credit Valuation Adjustment
SFT	Security Financing Transaction
SFTR	Securities Financing Transactions Regulation
SME	Small and medium-sized enterprise
SSM	Single Supervisory Mechanism
SREP	Supervisory Review and Evaluation Process
TLAC	Total Loss-Absorbing Capacity
TRIM	Targeted Review of Internal Models

1.Introduction: Political and legal context

The Great Financial Crisis of 2008-09 (GFC) had its origins in various deficiencies of the financial sector, in particular in the banking sector. According to the ‘Report of the de Larosière Group’ 1 , the origins of the GFC emanated from the combination of several factors, mainly the low cost of borrowing prevailing before the GFC, the financial market search for ever–higher returns and the emergence, and widespread use, of complex financial products created by bundling up new tradable securities from existing underlying risky loans. The deterioration of the credit quality of a large number of these loans triggered a rapid contagion of financial difficulties across the banking sector, highlighting its various deficiencies, including the failure of credit agencies and financial institutions to appropriately assess the risks of these new securities, the excessive interconnectedness of financial institutions worldwide, the inadequacy of banks’ prudential framework to impose sufficient loss-absorbing own funds and liquid assets requirements and the insufficiency of supervisory oversight.

The consequences of the GFC on the financial sector resulted in major costs 2 for governments as they had to support the financial sector, and also in a massive contraction 3 of economic activity in the Union and across the world. In response to the GFC, the Union implemented substantial reforms of the prudential framework applicable to banks in order to enhance their resilience and thus help prevent the recurrence of a similar crisis. Those reforms were largely based on international standards adopted since 2010 by the Basel Committee on Banking Supervision (BCBS) 4 . The standards are collectively known as the Basel III standards, the Basel III reforms or the Basel III framework 5 . A summary of the content and timelines of those reforms, as well as their implementation in the Union, is provided at the end of this Section (see Figure 3 ).

The global standards developed by the BCBS have become increasingly important due to the ever more global and interconnected nature of the banking sector. While a globalised banking sector facilitates international trade and investment, it also generates more complex financial risks. Without uniform global standards, banks could choose to establish their activities in the jurisdiction with the most lenient regulatory and supervisory regimes. This might lead to a regulatory race to the bottom to attract bank businesses, increasing at the same time the risk of global financial instability. International coordination on global standards limits this type of risky competition to a large extent and is key for maintaining financial stability in a globalised world. Global standards also simplify the life of internationally active banks – among which are a good number of EU banks – as they guarantee that broadly similar rules are applied in the most important financial hubs worldwide. The EU has therefore been a key proponent of international cooperation in the area of banking regulation

In the Union, the first set of post-crisis reforms that are part of the Basel III framework have been implemented in two steps:

·in June 2013 with the adoption of Regulation (EU) No 575/2013, also known as the Capital Requirements Regulation (CRR) 6 , and Directive 2013/36/EU, also known as the fourth Capital Requirements Directive (CRD IV) 7 ;

·in May 2019 with the adoption of Regulation (EU) 2019/876 8 , also known as the second Capital Requirements Regulation (CRR II), and Directive (EU) 2019/878, also known as the fifth Capital Requirements Directive (CRD V) 9 .

The reforms implemented so far focused on increasing the quality and quantity of regulatory capital that banks have to hold to cover potential losses. Furthermore, they aimed at reducing banks’ excessive leverage, increasing banks’ resilience to short-term liquidity shocks, reducing their reliance on short-term funding, reducing banks’ concentration risk, and addressing too-big-to-fail problems 10 .

As a result, the new rules strengthened the criteria for eligible regulatory capital, increased minimum capital requirements, and introduced new requirements for credit valuation adjustment 11 (CVA) risk and for exposures to central counterparties 12 . Furthermore, several new prudential measures were introduced: a minimum leverage ratio requirement, a short-term liquidity ratio (known as the liquidity coverage ratio), a longer-term stable funding ratio (known as the net stable funding ratio), large exposure limits 13 and macro-prudential capital buffers 14 .

Thanks to this first set of reforms implemented in the Union 15 , the EU banking sector has become significantly more resilient to financial shocks. One key indication of this increased resilience is the overall increase in regulatory capital EU banks have: between the end of 2014 and mid-2020 16 , the average Common Equity Tier 1 (CET1) 17 capital ratio 18 of EU banks improved by 2.5 percentage points (pp) to 15% 19 , as shown in Figure   1 20 .

Figure 1: Weighted average capital and leverage ratios for EU banks over time.

Source: Risk Assessment of the European banking system, EBA, December 2020.

Note: the above ratios are based on the prevailing rules of the prudential framework of the Union at the time, including under the prevailing transitional arrangements.

As a result, the EU banking sector entered the COVID-19 crisis on a significantly more resilient footing when compared to its condition at the onset of the GFC. In addition, temporary relief measures were taken by supervisors and legislators at the outset of the COVID-19 crisis. In its Interpretative Communication on the application of the accounting and prudential frameworks to facilitate EU bank lending - Supporting businesses and households amid COVID-19 of 28 April 2020 21 , the Commission confirmed the flexibility embedded in the prudential and accounting rules as highlighted by the European Supervisory Authorities and international bodies. In June 2020, co-legislators adopted targeted temporary amendments to specific aspects of the prudential framework – the so-called CRR “quick fix” package 22 . Together with resolute monetary and fiscal policy measures 23 , this helped banks to keep on lending to households and companies during the pandemic (as can be observed from Figure 2  below, which also shows some of the impacts of the relief measures for the Euro area). This, in turn, helped mitigate the economic shock 24 resulting from the pandemic.

Figure 2: Annual growth rate of loans to Euro area households (left hand-side) and some aggregate impacts of the temporary relief measures provided by supervisors and legislators in Q2 2020 on banks’ CET1 ratio (right hand-side).

Source: European Central Bank (ECB) Financial stability review, November 2020.

Note: In the two right-hand side charts, the y-axis represents the aggregate impacts of temporary relief measures expressed in percentage points changes between the banks’ CET1 ratio between end-Q1 2020 (i.e. before the measures applied) and end-Q2 2020 (i.e. after the measures applied). In the left-hand side chart, the x-axis represents the distribution of the aggregate impacts of temporary relief measures across individual banks supervised by the ECB. In the second chart, the x-axis represents the inter-quantile distribution of the individual banks’ impacts of two specific temporary relief measures, specifically the amendments to the CRR related to transitional arrangement for the application of IFRS 9 provisions and the date of application of the SME factor.

However, while the overall level of capital in the EU banking system is now considered satisfactory on average, some of the problems that were identified in the wake of the GFC have not yet been addressed. Analyses performed by the EBA and the ECB (see Section 2.1.1) have shown that the capital requirements calculated by EU banks using internal models demonstrated a significant level of variability that was not justified by differences in the underlying risks, ultimately undermining the reliability and comparability of their capital ratios. In addition, the lack of risk-sensitivity in the capital requirements calculated using standardised approaches results in insufficient or unduly high capital requirements for some financial products or activities (and hence specific business models primarily based on them). In December 2017, the BCBS agreed on a final set of reforms 25 to the international standards to address these problems. In March 2018, the G20 Finance Minister and Central Bank Governors welcomed these reforms 26 . In 2019, the Commission announced its intention to table a legislative proposal to implement these reforms in the EU prudential framework. 27

In light of the COVID-19 pandemic, the preparatory work has been delayed, reflecting the BCBS’s decision of 26 March 2020 to postpone the previously agreed implementation deadlines for the final elements of the Basel III reform by one year 28 . Beyond the temporary measures adopted to facilitate bank lending in the context of COVID-19 referred above, this delay has allowed the Commission services to reassess the impact of the planned reform in light of the potential consequences of the COVID-19 pandemic. The temporarily stressed economic conditions have not altered the Commission services’ views on the need to deliver on this structural reform. Completing the reform will address the outstanding issues highlighted above and will thus further strengthen EU banks’ financial soundness, putting them in a better position to support economic growth and withstand potential future crises. It will also give banks the necessary regulatory certainty, completing a decade-long reform of the prudential framework for banks. The Commission services consider that the reform can be carried out in a manner that will not disrupt the recovery from the COVID-19 crisis.

This would also be in line with the actions of other members of the BCBS that are committed to implementing the reform timely and faithfully. Indeed, major jurisdictions, (US, UK, JP, HK, CA, AU and SG), have publically committed to adopting rules implementing the reform by 1 January 2023. Some of them, namely JP, CA, HK and SG, have already published draft rules. Those jurisdictions expect the EU to stick to its commitment to implement the reform on time.

Figure 3 below provides an overview of the first and final set of Basel III reforms, as well as the timelines of their adoption in the prudential framework applicable to EU banks. 

Figure 3: Summary of the main revisions to the Basel framework adopted by the BCBS post-GFC and their implementation timelines in the Union.

Source: European Commission.

Note: The dates coloured in yellow at the top are the dates of adoption of the various Basel III standards by the BCBS. The dates coloured in blue at the bottom are the dates of adoption of standards in Union law.

The completion of the reform of the prudential framework for banks following the GFC is not the only important initiative related to the banking sector.

Another initiative is linked to the Commission’s ongoing work on the transition to a sustainable economy. The Commission Communication on the European Green Deal (EGD) 29 clearly set out the Commission’s commitment to transform the EU economy into a sustainable economy while also dealing with the inevitable consequences of climate change. It also announced a Sustainable Finance Strategy 30 that will build on previous initiatives and reports, such as the action plan on financing sustainable growth 31 and the reports of the Technical Expert Group on Sustainable Finance 32 , but will reinforce the Commission’s efforts in this area to bring them in line with the ambitious goals of the EGD. The Taxonomy Regulation 33 will play an important enabling role in this context, by establishing a list of environmentally sustainable economic activities.

Bank-based intermediation will play a crucial role in financing the transition to a more sustainable economy. At the same time, the transition to a more sustainable economy is likely to entail risks for banks that they will need to properly manage to ensure that risks to financial stability are minimised. This is where prudential regulation can play an important role. The abovementioned Strategy acknowledged this and highlighted the need to include a better integration of climate and environmental risks into the EU prudential framework.

Work in that direction has already started. In the CRR II, the co-legislators introduced a requirement for large, listed banks to disclose environmental, social and governance (ESG) risks and mandated the EBA to prepare two reports on how the treatment of ESG risks could be incorporated in the prudential framework. In October 2019, the EBA published its Action Plan on sustainable finance 34 , outlining how it intends to achieve the three actions above. In this report, among other things, the EBA encouraged banks to integrate ESG risks, identify simple climate-risk metrics, adopt climate change related scenarios and use scenario analysis.

Work is also ongoing within the international supervisory community. In 2017, central banks and supervisors launched the Network for Greening the Financial System (NGFS) 35 . The aim of the network is to enhance the role of the financial system to manage risks and to mobilise funding for green and low-carbon investments in the broader context of environmentally sustainable development. In May 2020, it has published a handbook for supervisors on how to incorporate climate-related risks into supervision. 36

The final area of focus is the proper enforcement of prudential rules. Both supervisors and markets play a crucial role in this respect. In order for rules to achieve their intended effect, they need to be properly enforced. For this to happen, supervisors need to have at their disposal the necessary tools and powers (e.g. powers to authorise banks and their activities, require information from them, or sanction them in case they break the rules). The Commission keeps monitoring the functioning of the supervisory framework laid down in the CRD, including through close dialogue with national supervisors, the ECB and the EBA, in order to ascertain whether the powers and tools made available to supervisors are adequate, complete and used appropriately.

Market discipline is another important tool. In order to for investors to properly exercise they role of monitoring the behaviour of banks, they need to access the necessary information. This is why the CRR requires banks to disclose certain information to the markets. As in the case of the supervisory framework, the Commission keeps monitoring disclosure rules, including through dialogue with market participants, to gauge whether the information disclosed by banks is sufficient and easy to obtain.

The above monitoring activity allows the Commission to identify areas where rules need to be adjusted in order to address identified issues.

2.Problem definition

2.1.What are the problems and problem drivers?

2.1.1.Deficiencies in the current framework for calculating risk-based capital requirements

Banks are exposed to different types of risk as part of their day-to-day business. While the specific types of risks (as well as the intensity of those risks) a bank will be exposed to will depend on the business model of that bank, the four main types of risk that might result in financial losses for banks are credit risk, operational risk, market risk and counterparty risk 37 . In order to ensure that banks have sufficient amounts of regulatory capital to cover unexpected financial losses caused by those risks, banks are subject to binding risk-based capital requirements under the prudential framework ( Figure 4 below shows the aggregate level of risk weighted assets 38 (RWA) for those risks for EU banks over the last few years).

Despite the wide-ranging first set of reforms implemented in the Union after the GFC, increased levels of capital (as shown in  Figure 1 ) have not yet allowed to fully restore the confidence in the EU banking sector and some problems identified during the GFC remain. One important reason for this is the lack of trust in the risk-based capital requirements calculated by using internal models 39 .

Figure 4: Amount in EUR trillions of risk weighted assets by type of risk for EU banks over time 40 .

Source: Risk Assessment of the European banking system, EBA, December 2020.

Indeed, banks can use two types of approaches to calculate their risk-based capital requirements: the standardised approaches, which banks have to use by default, or the internal model approaches (based on banks’ own modelling assumptions), which banks may use upon the permission from their supervisors.

Standardised approaches are benchmark risk measurement techniques which banks have to us by default unless they have been granted permission to use the internal model approaches. Under these approaches, banks have to calculate their risk-based capital requirements according to standard formulas and pre-defined parameters (e.g. regulatory risk weights, loss-given default parameters, market volatilities, etc.) specified in the legislation. This ensures that banks apply those approaches in a uniform manner which makes the calculation of capital requirements under those approaches largely comparable across banks for similar risks. The standardised approaches’ parameters are intended to capture a conservative estimate of the average risk of an exposure in a way that is sufficiently simple for a widespread use. The majority of EU banks relies on standardised approaches to calculate their capital requirements.

Standardised approaches have shown a number of weaknesses during the GFC which prevent them from acting as solid benchmark. First, some observations of actual losses incurred by banks during the GFC showed that the existing standardised approaches do sometimes underestimate the risks of certain types of exposures leading to insufficient amounts of capital required. The opposite was also found to be true: sometimes standardised approaches overestimated risks, leading to excessive amount of capital required. This can be explained, to an important extent, by the fact that they are designed to be simple. They hence do not always properly reflect the various characteristics of financial products, especially the most complex ones. This may, in turn, have an impact on banks’ activities. For example, if the capital requirement for a certain type of loan is too low compared to the riskiness of that loan, then the bank may grant too many of those loans while having insufficient capital if those loans start defaulting (and vice versa).

The lack of risk-sensitivity of standardised approaches has been observed for all types of risks, although to different extents. For example, in its 2019 benchmarking exercise of internal models for credit risk 41 , the EBA highlighted the high variability of the ratio between the risk weights generated by banks’ internal models for credit risk and the corresponding risk weights under the standardised approaches, for different types of credit exposures. As shown in Figure 5  below, the high discrepancies of those ratios across institutions cannot be explained solely by the high variability in banks’ internal model approaches but also by the lack of risk-sensitivity in the standardised approaches.

Figure 5: Comparison between risk weights implied by individual banks’ internal models for credit risk or ‘Internal Rating Based’ (RW IRB) and risk weight of the credit risk standardised approach (RW SA) for mortgage exposures.

Source: EBA.

Notes: Each point represent one EU bank participating to the exercise.

Unlike the standardised approaches, internal model approaches allow banks to estimate most or all the parameters required to compute capital requirements on their own. Since putting in place and maintaining such internal models requires significant resources, the cost of operating internal model approaches are significantly higher than the costs of using standardised approaches. This is why the number of banks that use internal models is much smaller than the number of banks using standardised approaches: according to the EBA’s 2019 study of the impact of the final elements of the Basel III reform, , only 79 banks out of 189 participating to the EBA data collection were using internal models 42 . However, those banks tend to be the largest ones in the EU, accounting for a large proportion of the total EU assets. Hence, the capital requirements calculated under the internal model approaches by those banks represent more than half of the overall capital requirements of EU banks as shown in Figure   6 .

Figure 6: Breakdown of EU banks’ RWAs calculated under the standardised and internal model approaches of the current prudential framework for credit, market and CVA risks.

Source: Basel III reforms: Impact study and key recommendations, August 2019, EBA.

When using internal models, banks can capture risks more accurately by taking into account their own assessment of the characteristics of exposures, such as loans (e.g. the likelihood that the borrower would default and the size of the loss the bank would incur in case there is a default). Since the use of internal models is predicated on close monitoring and assessment of the risks banks are exposed to, banks have a better understanding of how to manage and mitigate those risks.

However, the freedom that internal model approaches give to banks has potential downsides. Depending on how these models are built and on the modelling assumptions underpinning them, internal models of different banks can produce different estimates of risks and hence different levels of capital requirements. Given that each bank originates loans to different clients, invests in different assets and trades with different counterparties, a certain degree of variation is to be expected. However, a range of studies conducted at both international 43 and EU 44 levels found a level of variation in capital requirements across banks using internal models that cannot be explained solely by differences in the riskiness of banks’ exposures. In fact, those studies have shown that internal models can produce very different capital requirements for very similar or even identical exposures. In some cases, capital requirements for the exact same portfolios of exposures have shown a variation of more than 600%, as illustrated by  Figure 7 below in the area of market risk 45 . This variation makes it difficult to compare capital ratios across banks, puts in question their calculation and undermines confidence in capital ratios and distorts competition across banks.

Figure 7: Variability of capital requirements produced by internal models for market risk for different types of trading activities. Ratio between the individual bank “Value-at-Risk” risk measures for a given portfolio with the median for that portfolio.

Source: EBA report - results from the 2019 market risk benchmarking exercise.

Notes: The category ‘FX’ refers to trading portfolios with foreign-exchange rate risk, ‘Comm’ to trading portfolios with commodity risk (e.g. energy or agricultural goods) and ‘CTP’ to trading portfolios with subject to the specific Correlation Trading Portfolio capital requirement (e.g. collateralised debt obligations).

Moreover, a recent study 46 published by the Bank for International Settlements (BIS) illustrates that, besides a wide degree of variability in capital requirements among banks using internal models, the market perception of the financial risks that banks face is persistently higher than banks’ own assessment of risk when calculating capital requirements with internal models. As shown by  Figure 8  below, this problem of perceived underestimation of risks by internal models is significant for certain banks.

The reliability and robustness of the capital requirements produced by internal models is predicated on the condition that banks should not use internal models to reduce their capital requirements to levels which jeopardise the very objectives of safeguarding financial soundness and covering potential losses. For this purpose, the current prudential framework already provides a number of safeguards, most notably the need for supervisory approval to use an internal model as well as its regular monitoring and review. However, given the growing number and sophistication of models used by banks and the updates made to those models 47 , their supervision is becoming more complex and more resource intensive. 48 Furthermore, a reliable internal model requires sufficient data of sufficient quality to be available. This condition is not always met: in some cases the amount of available data is insufficient to allow for reliable and robust modelling of losses. This in turn produces unreliable estimates of the size of the risks a bank is exposed to and hence leads to inaccurate capital requirements.

Figure 8: Comparison of the Variability Ratio* across countries.

Source: “Variability in risk-weighted assets: what does the market think?”, BIS, 2020.

Notes: * The variability ratio is defined as the risk perception by the market compared to the risks measured in the regulatory framework across banks in a given country. The higher the ratio, the larger the perceived underestimation of risk.

The identified problems have raised doubts on the adequacy of internal models. Supervisory approval is required for banks to use internal models and supervisors can naturally decide not to allow a bank to use inappropriate internal models. However, the supervisors’ approval process of internal models cannot prevent in itself the variability observed across the outcomes of banks’ internal models. The supervisory authorities have already undertaken dedicated initiatives to reduce to some extent such variability. In 2016, the EBA produced a roadmap to comprehensively review the current rules for credit risk internal models and in order to issue guidelines, opinions or develop technical standards 49 . Also since 2016, the ECB/SSM has carried out a large-scale targeted review of internal models (TRIM) for the banks under its direct supervision in cooperation with the national supervisory authorities 50 . However, there is a limit to the number of interventions supervisors can make to address issues with individual models used by individual banks, because supervisors have limited resources at their disposal. Once that limit is reached, more structural solutions, like changes to the rules governing models, may need to be contemplated.

2.1.2.No dedicated capture of ESG risks in the prudential framework

Climate change and the profound economic transformations that are needed to contain it pose significant risks to banks, primarily in the form of transition risk (whereby the transition to a sustainable economy can result in big shifts in asset values) and physical risks (whereby more frequent or more severe weather events impact banks and their customers). Climate (and, more broadly, environmental) risks are often considered together with social risks 51 and governance risks 52 under the heading of ESG risks, as these risks share a number of characteristics and are often intertwined. 53  ESG risks, in turn, are closely linked with the concept of sustainability, as ESG factors represent the main three pillars of sustainability.

ESG risks affect different types of banks’ exposures differently: over the longer term, exposures related to the financing of sustainable activities are most likely less risky for banks than exposures financing unsustainable activities. If these risk differentials are not adequately reflected in banks’ decision-making, banks may underestimate the overall level of risk that they face, which raises financial stability concerns. For example, Alessi, L., Di Girolamo, F., Petracco-Giudici, M. and Pagano, A. (2021) argue that transition risks might result in an increase of bank losses by 4% in a crisis. Also, banks may also underestimate the risks of unsustainable activities compared to sustainable activities and as a result may overinvest in unsustainable activities while underinvesting in sustainable activities. An adequate reflection of ESG risks in banks’ decision-making in turn would help addressing this misallocation of resources and hence make it more likely that banks finance sustainable activities, enabling the Union to reach the EGD’s goals.

Against this background, it is essential that banks are able to measure and monitor their exposure to ESG risks, also to enable supervisors and market participants to appropriately assess the ESG risks faced by each bank in order for supervision and market discipline to function effectively.

The current legal framework does not prevent banks from considering ESG risks in their decision-making nor from disclosing information on their exposure to such risks. While availability of relevant data for banks has been an obstacle in this context, steps have been taken to facilitate banks’ access to such data, for example by means of requirements under the Corporate Sustainability Reporting Directive (CSRD) 54 and the Taxonomy Regulation. These measures should put banks in a better position to manage ESG risks.

Also, EU co-legislators have deemed that a dedicated approach to capture ESG risks in banks’ financial activities could help address the aforementioned challenges and introduced in CRR II provisions aimed at improving the capture of ESG risks. First, large banks with publicly listed issuances will start disclosing information on ESG risks from 2022 onwards. Second, the EBA has been mandated to assess by June 2021 the potential inclusion of ESG risks in the supervisory review and evaluation process (SREP) performed by supervisors. Third, the EBA has been mandated to assess by 2025 whether a dedicated prudential treatment of exposures related to assets or activities associated substantially with environmental and/or social objectives would be justified. However, the immediate effectiveness of these provisions is limited, as a large number of banks are outside of the scope of the CRR disclosure rules, and any advice from the EBA on the other two areas under investigation would require subsequent changes to the CRR.

As a result, the present legal requirements alone are insufficient to provide incentives for a systematic and consistent management of ESG risks by banks. This has also been recognised by the EBA in its aforementioned discussion paper, which states that it “sees the need for enhancing the incorporation of ESG risks into institutions’ business strategies, business processes and proportionately incorporate ESG risks in their internal governance arrangements”. The EBA considers the current legal requirements insufficient for this purpose and therefore “recommends to incorporate ESG risk-related considerations in directives and regulations applicable to the banking sector (e.g. CRD and CRR)”.

2.1.3.Inconsistency of powers and tools made available to supervisors across the Union

In order to perform their duties, national and European 55 competent authorities in charge of banking supervision have to use their powers under national laws transposing the CRD. In this regard, the CRD requires Member States (MS) to provide competent authorities with a minimum set of powers to exercise their supervisory functions 56 (thereafter “supervisory powers“) and to impose sanctions through administrative measures 57 and administrative penalties 58 (thereafter “sanctioning powers”) for banks breaching regulatory requirements (as set out in the CRR rules or national laws transposing the CRD). While the CRD ensures a minimum level of harmonisation across the Union, some MS have identified 59 a number of areas for which they considered it necessary to further elaborate the rules and/or to introduce additional powers for supervisory authorities. Consequently, these MS introduced additional provisions in their national laws, making use of the discretion allowed under the CRD. This has led to a situation where supervisors in different MS have different powers.

In the area of sanctioning powers, some MS included additional provisions in order to, inter alia, sanction banks for breaches 60 other than those contained in the minimum list provided in the CRD 61 , and determine administrative penalties 62 incurred by banks in case of breaches of CRD/CRR, including the maximum amount 63 of administrative penalties.

In the area of supervisory powers, some MS further specified the assessment of the prudential soundness of banks in case of acquisitions of material holdings in entities other than banks 64 , material transfers of assets and liabilities between a bank and a third party, and mergers or de-mergers with other banks.

As regards the supervision of members of a bank’s management body and of key function holders 65 , the CRD sets a number of principles to assess their suitability (‘fit-and-proper assessment’). However, the CRD lacks details on how and when supervisors should conduct fit-and-proper assessments of board members and how to identify the key function holders and assess their suitability. While the publication of joint guidelines 66 by the EBA and the European Securities Markets Authority (ESMA) published in 2017 and the guide to fit and proper assessments published by the SSM 67 in 2018 improved the harmonisation 68 of practices across MS, material divergences in national laws remain. For instance, the supervisors in some MS assess the suitability of board members only a significant period of time after their appointment 69 while in the majority of MS supervisors perform this assessment prior to their appointment. In the case of key function holders, some supervisors do not properly identify them and therefore do not carry out an assessment of their suitability to perform their duties, while other do it in a variety of ways 70 .

This fragmented regulatory landscape in the definition of certain powers and tools available to supervisors and their application across MS undermines the level playing field in the Single Market and raises doubts about the sound and prudent management of EU banks and their supervision. This problem is particularly acute in the context of the Banking Union, as already highlighted in the Commission’s report on the SSM 71 . Differences across 19 different legal systems prevents the SSM from performing its supervisory functions effectively and efficiently 72 . Moreover, as a result of the diverse transposition of the CRD at national level, cross-border banking groups have to deal with a number of different procedures for the same prudential issue, unduly increasing their administrative costs. Banks as well as supervisors, in particular in the SSM, and MS have therefore acknowledged the problem (for more details see Section 6.3. and Annex 2).

2.1.4.Fragmentation and inefficiency in the disclosure of banks’ prudential information

Prudential regulation requires banks to publicly disclose financial and other quantitative and qualitative information 73 so that investors, clients, depositors and other interested stakeholders can gauge their level of risk. Banks’ disclosure of financial information also contributes to enhanced transparency and market discipline, thereby promoting sound risk management.

In the Union, the CRR II implemented the revised BCBS framework 74 on public disclosure (also known as the ‘Pillar 3’ framework), and adjusted the content 75 and scope of bank disclosures to make them applicable to all EU banks in a proportionate manner. Under the CRR II, the amount of information that banks need to disclose depends on the size and complexity of their activities (the larger and more complex a bank is, the more information it is required to disclose).

Banks are currently required to disclose all relevant prudential information in one single document or a separate section of their financial report prepared under the applicable accounting standards. Information on banks are therefore scattered on their individual websites and other media platforms. However, stakeholders (e.g. investors making investment decisions or analysts making recommendations to their clients) are interested in gathering, analysing and comparing information across several banks.

At present, it is considered difficult and burdensome for these stakeholders to access and aggregate this information. For example, available information on smaller, non-publicly listed banks tends to be harder to find and is usually only available in the language of the Member State in which the bank is established. The current difficulties related to the access to prudential information deprive market participants from the information they need about banks’ prudential situations. This ultimately reduces the effectiveness of the prudential framework for banks and potentially raises doubt about the resilience of the banking sector, especially in periods of stress.

In addition to having to disclose certain information, banks must also report certain information 76 to their supervisors as part of a separate process. Two separate processes to share similar financial information imply undue administrative costs for banks 77 ; this also makes these processes more vulnerable to a certain type of operational risk (i.e. the risk of misalignment in the same type of information communicated through two different channels).

Since 2018, the EBA, in cooperation with the ECB and national competent authorities, has been working on the creation of the European Centralised Infrastructure for Supervisory Data (EUCLID) to aggregate in a centralised integrated system the reporting information shared by supervisors on the largest EU banks 78 . This system will be particularly useful to feed public reports and analysis with aggregated data and risk indicators on the overall EU banking sector. However, the prudential framework does not yet grant powers to the EBA to disclose individual bank data that is reported to supervisors. Introducing those powers would allow banks to only report information to their supervisors and the EBA which would then proceed to disclose the required parts of that information on behalf of banks. Banks, in particular small and non-complex ones, have repeatedly called to further reduce the administrative burden stemming from reporting 79 and disclosure requirements whereas other market participants (investors, analysts) highlighted the need for a centralised and easy access to banks’ prudential information.

2.2.How will the problems evolve?

As far as the deficiencies of internal models in the current framework for calculating risk-based capital requirements are concerned, confidence in the risk based capital requirements could be partially restored through the supervisory exercises run by the EBA and the ECB 80 . Those exercises aim at ‘model repair’ by means of harmonising certain modelling assumptions across EU banks. However, the completion of those exercises could not fully resolve the issues, given the amount of flexibility the current rules still provide banks to design their individual internal models. In addition, enhanced supervisory review would not address the deficiencies identified in the standardised approaches for calculating risk-based capital requirements. In this context it has to be recalled that capital requirements calculated under the standardised approach represent a significant share of total capital requirements across the Union.

In the absence of changes to address the identified deficiencies related to internal model approaches, the risk-based capital requirements calculated by banks using those models would remain incomparable across banks and in some cases may be too low in relation to the risks of certain exposures. If the deficiencies related to the standardised approaches would remain unaddressed, some financial products or activities (and hence specific business models primarily based on them) would attract either insufficient or unduly high capital requirements. In both cases the potential mispricing of risks by individual banks and consequently the inadequate capitalisation of those risks by those banks would persist.

As a consequence, some mistrust in the EU banking sector would persist with negative consequences for banks’ market valuations and funding costs which could in turn undermine their ability to finance the EU economy. The lack of confidence in the EU banking sector could ultimately lead to a higher probability that future periods of stress, whatever their origins, could turn into more severe financial crises.

As regards the capture of ESG risks in the prudential framework, the problems that such capture could potentially address (i.e. inadequate management of ESG risks) would become ever more pressing. Both transition risks and physical risks to banks would inevitably increase as the economic restructuring gains speed and as environmental events increase in magnitude and frequency. In the absence of timely legislative action to address these problems (in addition to the limited measures included in CRR II and elsewhere, see Section 2.1.2), banks might continue to misprice ESG risks, which would in turn lead to inadequate financing of the transition to a more sustainable economy risks. It could also lead to undercapitalisation of banks which could lead to financial stability issues.

On the supervisory and sanctioning powers, in absence of a legal initiative the discrepancies observed across national laws transposing the CRD would continue to exist, maintaining the current fragmentation and the un-level playing field with regard to the application of supervisory powers and the imposition of sanctions. This would have negative consequences in two respects. On the one hand, some MS would maintain a relative low degree of supervision, allowing some of their banks to perform certain operations that could be risky, thus fuelling mistrust in the soundness of banks. Furthermore, in some MS supervisors would still not be empowered to impose sanctions for breaches of certain prudential requirements and would thus continue to lack an important supervisory tool. On the other hand, other MS would continue to exercise the legal flexibility available in the CRD in a more conservative manner. Additional supervisory and sanctioning powers in these MS could incentivise some banks to carry out certain activities in other MS without such powers. Finally, without a change, the ECB/SSM would not be able to apply the same supervisory tools and exercise the same powers to all banks under its supervision in a consistent way which would impede the effectiveness of its supervision.

The existing inconsistencies and identified deficiencies of the application of fit-and-proper requirements lead to a less effective and efficient supervision regime. In a number of MS, members of the management body would continue to take up their position without having been vetted upfront by supervisors. This involves the risk that unqualified managers could contribute to key decisions for the banks’ businesses and its risk management. In the absence of proper definitions of key function holders and rules for their assessment the current un-level playing field across institutions within the Union would continue to exist, creating reputational risk for EU banks and their supervisors, undermining trust in the banking sector.

Finally, in the absence of actions to address the fragmentation and the inefficiency in the disclosure of banks’ prudential information to the public, analysing and comparing information on individual banks would remain burdensome and costly for the relevant stakeholders and would continue to undermine the effectiveness of market discipline. In addition, banks’ processes to report information to supervisors and to disclose information to the public would continue to run in parallel, unduly maintaining maintaining unnecessary administrative costs.

Figure 9: Mapping of problems, problem drivers and their consequences if not addressed

3.Why should the EU act?

3.1. Legal basis

All actions considered frame the taking up, pursuit and supervision of the business of banks within the Union, with the objective of ensuring the stability of the internal market. One of the fundamental components of the Union’s financial system, banking is currently providing the largest part of financing within the internal market. The Union has a clear mandate to act in the area of the internal market and the appropriate legal basis consists of the relevant Treaty Articles 81 underpinning Union competences in such area.

The legal basis falls within the internal market area, which is considered a shared competence, as defined by Article 4 TFEU. Most of the actions considered represent updates and amendments to Union law, and as such, they concern areas where the Union has already exercised its competence and does not intend to cease exercising such competence. A few actions (particularly those amending the CRD) aim to introduce an additional degree of harmonisation in order to achieve consistently the objectives defined by that Directive.

3.2.Subsidiarity: Necessity of EU action

In the context of global cooperation on financial stability, supervisors and regulators meeting within the BCBS, including from several EU Member States, the ECB, the EBA and the Commission, have developed common international standards that members jurisdictions should apply to their internationally active banks. Following the GFC, the BCBS launched a fundamental review of the international standards to strengthen the resilience of the global banking system and improve comparability across banks worldwide. A number of the revised standards have already been incorporated into Union law by means of the CRR and the CRD IV, as subsequently amended by the CRR II and CRD V. However, a number of additional revisions adopted by the BCBS in December 2017 in relation to credit risk, operational risk, CVA risk, market risks and the replacement of the Basel I floor by an aggregate output floor have still to be transposed in Union law. These proposed revisions address remaining shortcomings in the international prudential framework as identified during and after the GFC.

The objectives pursued by those revisions of international standards can be better achieved at Union level rather than by different national initiatives as they represent adjustments to the EU prudential framework. The identified problems (see section 2.1.1) and the underlying causes are similar across Member States and potential differences pertain to the behaviour and business model of individual institutions, not their location within the Union. No action by the Union would render the existing prudential framework outdated in relation to evolving market challenges and would create major misalignments with standards applied by other jurisdictions. This would have reputational, financial stability and market impacts.

The ability of Member States to adopt national measures to address the identified issues is limited, given that the CRR and the CRD already regulate those aspects, and changes at national level would not be able to derogate from Union law currently in force. If the Union were to cease regulating those aspects, the internal market for banking services would become subject to different sets of rules, leading to fragmentation and undermining the recently built single rulebook in this area. National measures would affect the degree of cross-border service provision, capital flows and market integration across Member States. This would be detrimental to effectively ensuring financial stability in the internal market.

With regard to the current rules on supervision (e.g. fit and proper rules, supervisory powers and sanctions), the current national laws supplementing the prudential framework have displayed some substantial differences. The absence of sufficient common rules does not allow for a level-playing field, potentially fuelling regulatory or supervisory competition. Also, in the context of the Banking Union, where the ECB exercises direct supervisory powers set out in national laws transposing the CRD, the ECB does not have the same range of powers with regard to all banks under its supervision. The objective of efficient and harmonised supervision throughout the Union cannot be achieved by individual MS actions.

3.3.Subsidiarity: Added value of EU action

There are clear benefits from action at Union level. The Union’s prudential framework for banks would be aligned with the latest international standards, thus becoming more fit for purpose and ensuring a greater resilience of the Union’s banking sector. It is more efficient to change the current Union rules than repealing them and replacing them with national rules. EU action also allows for a more homogenous approach, taking into account all EU specificities in a comprehensive way.

As regards those aspects that would supplement existing rules (e.g. fit and proper rules, supervisory powers, sanctions) additional harmonisation will contribute to a more homogenous approach and reduce the fragmentation of the internal market. At the same time EU action contributes to establishing a level playing field and a higher quality of supervision across the Union.

4.Objectives: What is to be achieved?

4.1.General objectives

There are two general objectives pursued by this initiative:

1.Contributing to financial stability. Only a stable and financially sound banking system, which is well capitalised and where risks are adequately managed, will lead to a reduced probability of banking crises, and reduce the impact of such crises should they occur. This will in turn allow to maintain investors’ and depositors’ confidence in the banking system, especially in periods of stress.

2.Contributing to steady financing of the economy in the context of the recovery post-COVID-19 crisis. The EU economy is heavily reliant on financing provided by the banking sector. While initiatives are underway to develop a Capital Markets Union (CMU) that might over time reduce this reliance, the banking sector will in particular play a key role in financing the medium-to-long term recovery from the COVID-19 crisis.

4.2.Specific objectives

The two general objectives pursued by this initiative can be broken down into the following four specific objectives:

1.Strengthen the risk-based capital framework, without significant increases in capital requirements. To ensure a resilient and stable banking system, it is essential to have a solid prudential framework in place which ensures that risks are accurately measured and adequately covered by capital. In particular, trust in the banking system is conditional upon trust in the reliability of the risk-based capital framework. Internally modelled approaches that EU banks use for calculating risk-based capital requirements should result in adequate capital levels and produce comparable outcomes. At the same time the prudential framework should provide robust and sufficiently risk-sensitive standardised approaches for banks using them. However, strengthening the risk-based capital framework should not come at the cost of significantly increasing capital requirements as requested by the EU Parliament and the Council 82 .

2.Enhance the focus on ESG risks in the prudential framework. Including a sustainability dimension in the prudential framework would ensure a better management of ESG risks and a better allocation of bank funding across projects, thus helping the transition to a more sustainable economy.

3.Further harmonise supervisory powers and tools. Supervisory tools such as fit-and-proper assessments, supervisory powers and sanctioning powers play a key role in ensuring the safety and soundness of individual banks and the stability of the EU banking system as a whole. Certain tools should therefore be made available to supervisors in all Member States and applied consistently. This would also help reduce administrative costs resulting from the current fragmentation.

4.Reduce banks’ administrative costs related to public disclosures and improve access to banks’ prudential data. Public information on individual banks that is easily accessible and comparable should enhance the ability of bank clients, investors and other market participants to monitor and exert market discipline on banks’ behaviour. Banks in turn would benefit from a more efficient system that integrates supervisory reporting and disclosure, and thereby reduces their administrative burden.

5.What are the available policy options?

5.1.What is the baseline from which options are assessed?

The baseline option from which the considered policy options are assessed assumes no legislative change to the prudential framework applicable at Union level.

Under this option, the structural shortcomings of internal models of the current prudential framework as described in Section 2.1.1 would remain unaddressed, and the Commission would solely rely on the supervisory initiatives taken by the EBA and the ECB under their current remits to improve the situation. In addition, the EBA will continue performing its annual supervisory benchmarking exercises on credit and market risk internal models to highlight the degree of variability of EU banks’ internal models. The EBA and ECB initiatives could achieve to some extent a reduction of the variability in capital requirements calculated based on internal models for credit, counterparty and market risks 83 . While acting as important complements, these supervisory initiatives cannot substitute for the necessary reforms of the binding requirements included in the current prudential framework. Without those reforms, the problems described in Section 2.1.1 would continue to exist.

In the same vein, without changes to the standardised approaches, they would remain insufficiently risk-sensitive. Since those standardised approaches are legally binding, they are virtually identical for all banks that apply them. Therefore initiatives employed by supervisory authorities cannot be used to tackle the deficiencies identified in Section 2.1.1.

Moreover, under this scenario, the only binding requirement related to ESG risks that would apply would be the requirement for large banks to disclose information on those risks from end-June 2022 onwards, as highlighted in Section 2.1.2. While competent authorities will continue to be able to exercise their Pillar 2 powers to require banks to address ESG risks, due to the lack of an explicit reference to those risks in the CRD, those powers will not be applied systematically and consistently. Any additional requirements that the EBA may propose in the reports due in 2021 and 2025 (see Section 2.1.2) would require legislative action.

Not changing legislation would also mean that divergences among MS in the area of supervisory powers made available to competent authorities under the CRD, identified in Section 2.1.3, would persist. For example, the prudential framework would still not require MS to give competent authorities powers to sanction certain types of regulatory breaches. Furthermore, there would be no legal requirement for an intervention in the case of specific situations that could raise prudential concerns, such as the acquisition of material holdings in entities other than banks, material transfers of assets and liabilities, and (de-)mergers. MS would still have ample leeway to grant the competent authorities these supervisory powers at their own discretion. The CRD would also still not require competent authorities to assess the suitability of key function holders. At the same time, the prudential framework would continue to lack specifications on how competent authorities should conduct fit-and-proper assessments. As a result, the rules would remain subject to broad margins of interpretation by MS, leaving clearly identified prudential concerns unaddressed. The initiatives conducted by the EBA and the ECB/SSM (see Section 2.1.3) could help increase the harmonisation of the fit-and-proper assessment across Member States but, only if MS would empower supervisors by amending their national frameworks. In absence of such national empowerments, inadequate supervision of key function holders would persist in some MS, as associated with the risks of employing unsuitable individuals as members of the management body.

Finally, under this scenario of no change to the current prudential framework, banks’ investors and other stakeholders interested in banks’ prudential information would still find it burdensome to collect such information on individual banks’ platforms. In addition, banks would continue to have to follow two separate processes to report and disclose the required information, which would continue to generate an undue administrative burden for them, as described in Section 2.1.4.

5.2.Description of policy options

5.2.1.Improve the current framework for calculating risk-based capital requirements

Option 1 -     Implement the Basel III reforms in full alignment with the BCBS standards and implementation timelines

This option would implement in Union law the final elements of the Basel III reform (adopted by the BCBS between December 2017 and July 2020) in full alignment with the standards and the timelines agreed by the BCBS. The agreement was the result of a strategic review of the international prudential standards for banks, which was conducted by the BCBS in the wake of the GFC, with a view to improving the balance between simplicity, comparability (mainly of the internal model approaches) and risk-sensitivity (mainly of the standardised approaches) of those standards.

This option would entail implementing the final elements of the Basel III reform agreed in December 2017, namely:

·the revisions to the standardised approach for credit risk (SA-CR) to improve the robustness and risk sensitivity of the existing approach;

·the revisions to the IRB approaches for credit risk to reduce unwarranted variability in banks’ calculations of RWAs;

·the minimum haircut floors for non-centrally cleared securities financing transactions 84 (SFTs) to limit the pro-cyclicality of these transactions and the build-up of excessive leverage in the financial system;

·the revisions to the CVA risk framework consisting of the removal of the use of an internally modelled approach and the introduction of a new basic approach (BA-CVA) as well as revisions to the standardised approach for CVA (SA-CVA) to enhance the risk sensitivity, strengthen the robustness and improve the consistency of the framework;

·the new standardised approach for operational risk (SA-OR), replacing all the existing standardised and internal model approaches for this risk to simplify the framework and increase comparability; and

·the aggregate output floor (OF) to limit the unwarranted variability in the regulatory capital requirements produced by internal models and the excessive reduction in capital that a bank using internal models can derive relative to a bank using the revised standardised approaches.

In addition, this option would implement the revised 85 (i.e. January 2019) version of the original (i.e. January 2016) market risk standards, known as the ‘fundamental review of the trading book’ (FRTB), for the purpose of calculating capital requirements. Most of the revised FRTB standards have already been implemented in EU law as part of the CRR II. Following an agreement between the European Parliament and the Council, they were implemented for reporting purposes only (i.e. banks are not required to use them to determine their capital requirements). Under this option they would be converted into a capital requirement, fully aligned with the final 2019 standards.

Table 1 below presents a more detailed list of the key elements of the reforms included in the December 2017 agreement, as well as the revised market risk framework adopted in 2019.

This option would address the main deficiencies, identified in Section 2.1.1, with the current prudential framework for calculating risk-based capital requirements applicable to EU banks 86 . At the same time, this option would entail no adjustments to the final Basel III standards to cater for specificities of the EU economy and would also remove some of the EU-specific rules currently in place to fully align the EU prudential framework with those standards (e.g. it would remove the exemptions from the CVA risk charge and the small and medium-sized enterprise (SME) supporting factor).

Table 1: Overview of the main revisions to the Basel framework introduced by the final elements of the Basel III reform.

Risk area	Main revisions
Credit risk – standardised approach	·For rated exposures to banks, some of the RWs have been recalibrated. In addition, the RW treatment for unrated exposures to banks is more granular than the existing flat RW, which depends on the RW applicable to the central government of the Member State in which the bank is established. ·For exposures to corporates, a more granular RW treatment has been developed. In addition, a specific treatment for exposures to project finance, object finance and commodities finance was developed. ·For real estate exposures (both residential and commercial), more risk-sensitive approaches have been developed to better reflect different funding models and stages in the construction process. ·For retail exposures, a more granular treatment was developed, which distinguishes between different types of retail exposures and reflects FX risk. ·For subordinated debt and equity exposures, a more granular and generally stringent RW treatment was developed. ·For off-balance sheet items, a more risk-sensitive treatment was developed, which is more stringent for unconditionally cancellable commitments.
Credit risk – internal models approach	·The possibility to use of internal models was either limited or altogether removed for portfolios and risk parameters where the BCBS had concluded that the available data was insufficient to ensure reliable modelling (i.e. exposures to financial institutions and large corporates, equity exposures). ·New minimum values (‘input floors’) were introduced for banks’ estimates of the probability of default, loss-given default and exposure at default (EAD). ·The option for banks to pick and choose between the use of internal models and the standardised approach per asset class was introduced.
Market risk (FRTB)	·More objective rules were introduced to allocate transactions either to the trading book, or to the banking book. ·Both the standardised approach and the internal model approach were be completely overhauled to better capture market risk. ·New tests were introduced to ensure the robustness of the internal models and leave less flexibility for banks to use their own modelling assumptions.
CVA risk	·Internal models were replaced by standardised approaches, leaving banks with a choice between a more sophisticated and a simpler approach. ·For banks with non-centrally cleared derivatives contracts with a combined notional value of less than EUR 100 bn, a simplified approach was made available.
Securities Financing Transactions	·For certain non-centrally cleared SFTs with certain counterparties, minimum haircut floors were introduced; as a result SFTs which do not meet the haircut floors must be treated as unsecured loans.
Operational risk	·The current internal model approach and the three existing standardised approaches were replaced with a single risk-sensitive standardised approach to be used by all banks. ·The capital requirement under the new standardised approach depends on the size of the bank (expressed in terms of a refined measure of gross income) and the bank’s operational risk-related loss history.
Output floor	·A revised output floor was introduced. The floor sets a lower limit to the capital requirements that are produced by a bank’s internal model at 72.5% of the capital requirements that would apply if the bank would calculate its capital requirements using standardised approaches 87 .

Option 2 - Implement the final Basel III reforms with EU-specific adjustments and in alignment with the BCBS implementation timelines

Similar to option 1, option 2 would also implement all the final elements of the Basel III reform. Compared to option 1, this option would include a number of targeted adjustments to cater for the specificities of the EU banking sector and the funding structure of the EU economy, and taking into account the context of the recovery, with the objective to avoid disproportionate impacts or unintended consequences of the reforms on essential activities and financial services provided by EU banks. In fact, the Basel standards are designed to capture common financial risks that can be observed across the world. They are calibrated based on averages of data collected across banks and financial markets located in different regions of the world. As a result, the capital requirements under some Basel standards may not capture more specific financial risks appropriately which could lead to disproportionate impacts on the corresponding financial activities. In addition, a few other targeted adjustments under this option would help to ensure an international level playing field for trading activities. Finally, this option would also make use of the flexibility the international standards provide in order to harmonise their application across the Union, in line with the objectives set out for this legislative proposal.

Specificities of the EU banking sector and EU economy

EU businesses (and in particular SMEs, see also Section 3 of Annex 6) rely heavily on bank lending to finance their investment and working capital needs, much more than in other major jurisdictions. However, the international standards do not always sufficiently take into account the specific financing structures and risk management strategies used in the EU. To address this issue, option 2 includes several targeted adjustments:

·Treatment of unrated corporates: corporate lending in the EU is predominantly provided by banks using IRB models. With the implementation of the OF those banks would also need to apply the SA-CR which relies on external ratings to determine the credit quality of the corporate borrower. Most EU corporates, however, do not typically seek external credit ratings, due to the cost of establishing a rating 88 and other factors 89 . Given that capital requirements calculated under the SA-CR are, on average, more conservative for unrated corporates than for corporates that have a rating, 90 the implementation of the OF could cause substantial increases in capital requirements for banks using internal models (because the standardised approach would be used to calculate the OF). To avoid disruptive impacts on bank lending to unrated corporates and provide enough time to establish public and/or private initiatives aimed at increasing the coverage of credit ratings, option 2 would introduce a transitional period. During this period banks using internal models could apply a favourable treatment when calculating their OF for exposures to unrated corporates. This transitional arrangement would be coupled with an empowerment for the Commission to further extend the length of the period, based on a report by the EBA. This would ensure sufficient access to bank funding by unrated corporates during the transition period, as banks could continue to apply lower capital requirements whilst initiatives to foster widespread use of credit ratings would be established. After the transition period banks would refer to credit ratings to calculate capital requirements for most of their exposures to corporates in accordance with the Basel III standards. More details about this specific issue and the dedicated adjustment to address it are provided in subsection 1.1.1 of Annex 5.

·SME supporting factor: Given their fundamental role in creating jobs and economic growth in the EU, EU co-legislators decided that capital requirements for SME exposures should be lower than those for large corporates to ensure appropriate bank financing of SMEs. As a result, an SME supporting factor was introduced in the CRR and its scope extended in CRR II 91 . Option 2 would maintain this SME supporting factor in the prudential framework, which would result in lower capital requirements for SMEs than the specific treatment provided by the Basel III standards.

·Infrastructure supporting factor: bank financing of infrastructure and other specialised projects is also a defining characteristic of the EU economy, as compared with other jurisdictions where such projects are predominantly financed by capital markets. Large EU banks are major providers of funding for specialised projects 92 , objects finance 93 and commodities finance 94 , in the EU and globally. They have developed a high level of expertise in those areas. A preferential treatment has been introduced in CRR II to foster bank finance and private investment in high quality infrastructure projects (‘infrastructure supporting factor’). Option 2 would maintain this treatment, which would result in lower capital requirements for infrastructure projects than the specific treatment provided by the Basel III standards. Furthermore, a new preferential treatment under the standardised approach for “high quality” object finance would be introduced. It would apply where financial risks are specifically managed. Conflicting signals to banks active in this market segment should be avoided. 95  This option would also empower the Commission to adopt delegated acts that would adjust the internal models approach, if such adjustment would be found appropriate based on a comprehensive assessment by the EBA during the implementation period.

·Equity exposures: Many EU banks hold long-standing, strategic 96  equity participations in financial and non-financial corporates. The final Basel III standards increase the RWs for all kinds of equity exposures over a 5-year transition period without providing a specific treatment for strategic equity investments. Applying the more conservative approach embedded in the Basel III standards to the whole stock of existing equity holdings could jeopardise the economic viability of existing strategic relationships. Option 2 would exclude equity holdings in entities within the same banking group or covered by the same institutional protection schemes (IPS) 97 from the application of the more conservative treatment. In addition, it would grandfather existing strategic participations where banks exercise influence, including via holdings in insurance undertakings, whilst applying the new, more conservative treatment to new equity exposures. More details about this specific issue and the dedicated adjustment to address it are provided in subsection 1.1.3 of Annex 5.

·Treatment of Collective Investment Undertakings 98 (CIUs) used for trading purposes: CIUs play a crucial role in facilitating the accumulation of personal savings, whether for investments or for retirement. The seamless provision of CIUs as investment product hinges on banks’ ability to continuously offer to their clients the possibility to buy or sell back those instruments. For that purpose, banks must keep inventories of CIUs in their trading books. The revised market risk standards adopted in 2019 rely on a number of conservative assumptions and complex operational requirements 99 that could increase significantly the capital requirements for those instruments, therefore restricting their supply. To avoid this unintended effect on those trading activities, Option 2 would provide a number of adjustments to the treatment of CIUs under the market risk rules. More details about this specific issue and the dedicated adjustment to address it are provided in section 1.3 of Annex 5.

·Treatment of financial products based on the EU emission trading scheme (ETS): banks play an important role in providing liquidity to the EU market for carbon emissions allowances. Banks typically fill their clients’ estimated demand for allowances at a future date via derivatives (‘forward’) transactions. Under the revised market risk rules, the exposures to carbon emission allowances are assimilated to electricity contracts, and therefore get allocated to the same risk weight (60%). In light of the historical price volatility of EU ETS, this risk weight appears excessively high. Furthermore, in its Communication 100 on fostering openness, strength and resilience of the European economic and financial system, the Commission announced its intention to support further ETS trading activity in the EU, notably as part of the ‘green’ transition 101 . Option 2 would therefore introduce a specific category for ETS allowances, distinct from electricity, for which RWs would be calibrated to reflect the actual price volatility of this commodity in the EU ETS market. More details about this specific issue and the dedicated adjustment to address it are provided in section 1.3 of Annex 5.

·Exemptions from capital requirements for CVA risk: When implementing the initial Basel III reforms in Union law through the CRR, the EU co-legislators exempted certain transactions 102 from the calculation of capital requirements for CVA risk. These exemptions were agreed to prevent a potential excessive increase in the cost of some derivative transactions triggered by the introduction of the capital requirement for CVA risk, particularly when banks could not mitigate the CVA risks of certain clients 103 . While the introduction of the revised capital requirement for CVA risks adopted as part of the final elements of the Basel III reform would improve to some extent the risk-sensitivity of the standardised approaches for CVA risks, the level of capital requirements would still remain very high for the exempted transactions, notably due to the removal of the internal model approach for CVA risk. In this context, Option 2 would maintain the existing CRR exemptions. To help supervisors monitoring the CVA risks arising from the exempted transactions when they are excessive, option 2 would require institutions to report to their supervisors the calculation 104 of capital requirements for CVA risks of the exempted transactions. In addition, option 2 would introduce EBA guidelines to help supervisors to identify excessive CVA risk. More details about this specific issue and the dedicated adjustment to address it are provided in section 1.5 of Annex 5.

·Introduction of the minimum haircut floor framework for non-centrally cleared SFTs: SFT markets play an essential role 105 in the EU financial system by allowing financial institutions to manage their own liquidity position and support their securities market-making activities, as well as central banks to transmit, via financial institutions, their monetary policy plans to the real economy. However, SFTs can also enable market participants to recursively leverage their positions by reinvesting cash collateral and re-using non-cash collateral, respectively. To address some concern with respect to the risk of build-up of excessive leverage outside the banking sector, the Financial Stability Board (FSB) published 106 in 2013 a recommendation to its member jurisdictions to introduce minimum collateral haircuts for some non-centrally cleared SFTs traded between banks and non-banks, either, at the discretion of each jurisdiction, directly via a market regulation or indirectly via a more punitive capital requirement that was later developed by the BCBS 107 . Presently there is still a lack of certainty about the impact of implementing this FSB recommendation on the EU SFTs market and whether its prudential objective could be attained without creating undesirable consequences. In this context, option 2 would propose to delay its implementation in the EU until the EBA and ESMA provide in the coming years a joint report to the Commission assessing its impact and recommending the most appropriate approach. More details about this specific issue and the dedicated adjustment to address it are provided in section 1.6 of Annex 5.

International level playing field for trading activities. 

Banks trading activities in wholesale markets can easily be carried out across borders (for certain activities, even between the EU and non-EU countries). The capital requirements applied to these activities should therefore converge as much as possible across jurisdictions to avoid a potential competitive advantage for those banks for which the domestic rules are more lenient.

In this context, option 2 would turn the existing reporting requirement for market risk based on the BCBS FRTB framework into a capital requirement, as proposed under option 1. However, option 2 would contain a safeguard that would allow addressing disruptions to the playing field for EU banks’ trading activities that could materialise if other major jurisdictions would delay the implementation of the FRTB framework or adjust its calibration 108 . Specifically, option 2 would introduce an empowerment for the Commission to delay, if necessary, the application of the capital requirement based on the FRTB framework and/or to adjust its calibration considering international developments. More details about this specific issue and the dedicated adjustment to address it are provided in section 1.3 of Annex 5.

A second adjustment under option 2 would address the conservative calibration of the standardised approach for counterparty credit risk (SA-CCR) currently applied to derivative transactions and already transposed 109 in Union law by means of the CRR II. At the time of adoption, the EU co-legislators requested the EBA to report by June 2023 on the current SA-CCR calibration. Subsequently, in the Capital Market Recovery Package 110  they requested the Commission to review the appropriateness of the SA-CCR calibration by June 2021. Considering that at least one major jurisdiction 111  has already lowered the calibration of SA-CCR for certain types of derivatives transactions, option 2 would temporarily lower the existing calibration of SA-CCR for all derivatives transactions when calculating the OF. This would provide sufficient time to further discuss this issue at international level. More details about this specific issue and the dedicated adjustment to address it are provided in section 1.7 of Annex 5.

Flexibility in the Basel III standards

The Basel standards usually leave some flexibility for their local implementation, either by explicitly allowing jurisdictions to choose between different implementation options at their own discretion or by implicitly not providing all the technical details for the operationalisation of the standards. Option 2 proposes to make use of this flexibility and would implement those discretions in a harmonised manner.

First, the introduction of the OF in the prudential framework represents one of the key elements of the final Basel III reform, aimed at reducing the excessive variability of banks’ capital requirements calculated with internal models and thereby enhancing the comparability of capital ratios. The Basel III standards arguably leave some room for interpretation as regards the requirements 112 to which the OF should apply and, in case of banking groups, the level 113 at which it should apply. Under option 2, the OF would include all the existing capital requirements in scope of Union law, including the specific SRB and P2R capital requirements imposed to EU banks by supervisors and competent authorities. This is coherent with the current prudential framework and consistent with the objective of the OF. In order to avoid disproportionate capital impacts, while still strengthening financial stability and complying with the Basel III standards, option 2 would also prescribe that any overlap between the risks captured by the OF and the risks captured by any of the these specific capital requirements imposed by supervisors and competent authorities must be avoided 114 .

Moreover, the Basel standards, including the OF, apply foremost at the level of a banking group (consolidated level). In the EU prudential framework, capital requirements usually apply both at the level of individual banks/subsidiaries (individual/solo level) and at consolidated level. Some concerns have been raised that certain EU banking groups would incur an excessive increase in capital requirement due to the introduction of the OF if it were applied at solo level, in addition to its application at consolidated level. In this context, option 2 would introduce the OF at consolidated level only. However, to adequately capture the risks of both parent entities and their subsidiaries, whilst remaining consistent with the logic of the Single Market, option 2 would require any additional capital resulting from the application of the OF at consolidated level to be distributed fairly across the various entities of the group according to their risk profile as if the OF was applied at individual level 115 .

More details about the impacts of the output floor and the different implementation options are provided in section 1.8 of Annex 5.

In addition, the revised standardised approach for operational risk (SA-OR) introduced a discretion to allow jurisdictions to disregard the inclusion of banks’ own historical losses related to operational risk through the Internal Loss Multiplier (ILM) indicator in their calculations of capital requirements for operational risk under this approach 116 . While there is empirical evidence showing that banks experiencing greater operational risk losses historically are more likely to experience operational risk losses in the future, the events that have led to the largest operational losses are less amenable to prediction based on historical loss data than for other types of risks. For this reason, option 2 would exercise the discretion provided by the Basel III standards and set the ILM to 1. More details about this specific issue and the dedicated adjustment to address it are provided in section 1.4 of Annex 5.

Finally, option 2 would also exercise the flexibility provided in the Basel III standards to clarify that holdings of unlisted equities with a holding period of at least 3 years would not be considered as speculative holdings and would therefore not be subject to the most conservative treatment.

Option 3 - Implement the Basel III reforms with EU-specific adjustments and transitional arrangement adapted to the COVID-19 crisis

This option would mirror option 2, but with a later date of application of the reform. Instead from 1 January 2023, the reform would apply from 1 January 2025, followed by a 5-year transitional period as proposed by the BCBS. This would lead to the full application of the reform by 1 January 2030.

Option 3 would fulfil EU’s commitment to implement the international standards, and provide certainty for banks’ capital planning and lending decisions, whilst at the same time leaving them more time to comply with the revised capital requirements. The extended implementation period would allow banks to focus on managing their financial risks stemming from the COVID-19 crisis and financing the recovery and give them enough time to adjust before the reform would reach its full effect. In addition, to reinforce the signal that banks would have ample time throughout the recovery phase to adjust to the new rules and thereby help avoid that other market participants would put pressure on banks (in particular the few most impacted banks) to frontload 117 the new requirements, this option would entail dedicated communication efforts to explain the additional implementation period. 

5.2.2.Dedicated capture of ESG risks in the prudential framework

Option 1 - Introduce measures for a better management of ESG risks by banks

Under this option, several measures would be introduced to improve the management of ESG risks by banks and to reinforce the specific review by bank supervisors as well as the degree of market discipline, without directly targeting banks’ minimum capital requirements. These measures would aim to improve the resilience of banks to ESG risks and increase bank funding of sustainable activities by means of an improved understanding of the risks involved and an anticipation of expected market pressures.

As regards risk management, there is currently no explicit requirement for banks to have in place internal processes to manage ESG risks, nor an explicit requirement for ESG risks to be part of the supervisory review and evaluation process (SREP). Under this option, the current CRD V mandate for the EBA to issue a report to assess the inclusion of ESG risks in the SREP would be reinforced by explicitly requiring supervisors to ensure that banks manage ESG risks adequately. To the same end, the EBA's mandate would be clarified to require an assessment of how such risks should be included in the SREP via guidelines. Finally, the EBA would be mandated to specify further how ESG risks should be identified, measured, managed and monitored. This would include internal stress tests on banks’ resilience to climate change risks and long-term negative impacts.

The annual stress tests performed by supervisors are an important tool to gauge the viability of banks under adverse conditions. At present, the CRD requires supervisors to perform annual stress tests on the banks they supervise, while the EBA is mandated to define a common stress test methodology via guidelines. Under this option, this requirement would be extended to include regular climate change stress tests and to mandate the EBA together with the other ESAs to develop a methodology for that purpose.

In the area of disclosure, the disclosure requirements related to the disclosure of ESG risks would be extended to a larger universe of banks (i.e. beyond large, listed banks to whom the existing requirement will apply from 2022) while respecting the proportionality principle.

As the EU economy is transitioning towards a sustainable economic model, while at the same time being exposed to sustainability risks, exposures due to the financing of sustainable activities might be expected to be less risky for banks than exposures financing unsustainable activities. While this option would not include comprehensive adjustments to capital requirements for these exposures, it would contain a specific review clause that would allow such adjustments to be made in the near future, once sufficient evidence would be available 118 . In addition, the deadline for the EBA to deliver its report on the prudential treatment of these exposures would be advanced from 2025 to 2023 in order to ensure greater timeliness of any changes to the prudential rules that may be needed.

Option 2 - Adapt minimum capital requirements to reflect ESG risks

Option 2 would adapt the minimum capital requirements under so-called “Pillar 1” of the prudential framework for the financing of certain activities and/or products where specific evidence exists that they are more or less risky than comparable exposures as a result of ESG factors. For the exact delimitation of environmentally sustainable activities deemed to be of lower risk based on such evidence, the Taxonomy Regulation and its delegated acts would be used. 119  Under this option, capital requirements for sustainable exposures would be lowered, while capital requirements for unsustainable activities would be increased, to reflect the differences in ESG risks inherent in the two types of exposures. As under option 1, the deadline for the EBA to deliver its report would be advanced from 2025 to 2023.

5.2.3.Improve the consistency in the application of supervisory powers

 Option 1 - Clarify and complement certain provisions on supervisory and sanctioning powers while leaving ample flexibility to Member States

Under option 1, several elements of the prudential framework related to supervisory and sanctioning powers would be clarified compared to the status quo in order to ensure a more consistent application of the supervisory toolkit. However, this option would still grant flexibility to MS to detail certain supervisory and sanctioning powers in national laws.

As regards supervisory powers, this option would introduce harmonised obligations for banks to notify competent authorities ahead of specific events with prudential relevance (acquisitions of holdings, transfers of assets and liabilities, mergers and demergers). To avoid placing an unnecessary burden on competent authorities and banks alike, the notification obligation would be subject to a materiality threshold. This option would abstain from specifying relevant related powers or obligations for the competent authorities. This means that MS would be left with the discretion to grant competent authorities ex ante supervisory powers to oppose or to approve these events.

As regards sanctioning powers, this option would introduce a generic sanctioning provision, giving competent authorities a general power to sanction all possible breaches of the regulatory framework 120 . This option would also include a clarification of the distinction between the enforcement dimension (e.g. compelling entities to comply with the rules) and the punitive dimension (sanctioning a breach or a misconduct) of sanctioning powers. Lastly, this option would mandate the EBA to harmonise the basis for the calculation of pecuniary sanctions applicable to EU credit institutions (e.g. the total annual net turnover) by means of regulatory technical standards.

Concerning the fit-and-proper framework, this option would introduce harmonised rules for banks by requiring them to carry out an assessment of members of the management body and of key function holders before they take up their positions. This would be mandatory for all the banks in the group in which those persons are supposed to hold their functions. In addition, the option would address the assessment procedure in order to make it more reliable. This would include a specification of competent authorities’ powers and processes for the assessment of members of the management body 121  but not for key function holders as this category comprises very different roles 122 .

Option 2:    Ensure a greater level of harmonisation of the provisions on supervisory and sanctioning powers by narrowing down the flexibility of Member States

This option would go a step further compared to option 1. It would frame the current flexibility of Member States in determining which powers and tools to grant supervisors in national laws.

As regards supervisory powers, based on the notification requirement already foreseen in option 1, option 2 would introduce in the CRD explicit powers for competent authorities to ex ante oppose or approve events with prudential relevance (e.g. acquisitions of material holdings, material transfers of assets and liabilities, mergers and demergers).

As regards sanctioning powers, compared to option 1, the list of key CRD/CRR breaches subject to administrative sanctions would be completed based on a risk-based approach 123 . The list of sanctionable breaches would be expanded (e.g. breaches of additional reporting requirements and capital requirements 124 would be included). In this regard, this option would ensure that the breaches to these important regulatory requirements would fall under the scope of the sanctioning regimes of all Member States. Like option 1, this option would introduce a clarification between enforcement and punitive dimension of sanctioning powers. But in addition to that it would grant an additional enforcement tool (periodic penalty payments 125 ) to competent authorities in order to reinforce their ability to ensure a strict application of prudential requirements. Furthermore, this option would include a harmonisation of the definition of total annual net turnover by using the indicator currently used in the calculation of capital requirements for operational risk 126 in order to foster a comprehensive and consistent application of administrative penalties across the EU and ensure comparability of sanctions. 

Finally, under this option, the obligation for competent authorities to assess members of the management body and key function holders before they would occupy their positions would be limited to positions in the parent entity and material subsidiaries of large banks only (i.e. those powers would not be extended to small banks) following a risk-based approach. This would allow both supervisors and banks to focus their resources on the most important appointments. As with option 1, EU law would specify the procedural requirements for the fit-and-proper assessment, irrespective of whether the assessment is done ex ante or ex post.

5.2.4.Reduce disclosure costs and improve market access to bank prudential information

Option 1 -    EBA provides a single electronic access to EU banks’ quantitative disclosures

Under this option, the EBA would provide investors and other stakeholders with a single electronic web-based access to quantitative information that EU banks are required to disclose. Specifically, the EBA would publish this information on its website. The data would be sourced directly from the supervisory data that banks are required to report to their competent authorities, with the frequency determined by the disclosure rules. Specifically, they would be sourced from the EUCLID platform which is expected to be launched in 2021. This would be similar to what the EBA does as part of its EU-wide Transparency Exercises. It is important to highlight that the EBA would only provide the platform for the centralised disclosure; the ownership of the data and the responsibility for its accuracy would remain with the banks that produce it. Under this option, any qualitative information that banks have to disclose would not have to be published on the centralised platform; the disclosure of that information would be left to banks to manage.

Option 2 -    EBA provides a single electronic access to EU banks’ quantitative and qualitative disclosures.

Under this option, the EBA would provide investors and other stakeholders with a single electronic web-based access to both quantitative and qualitative information that EU banks are required to disclose. As under option 1, the quantitative information would be sourced from the EUCLID platform. Conversely, the qualitative information would be sent to the EBA by banks. The qualitative information could be simply sent in the format that institutions currently use (e.g. in the form of a pdf document; that document would then be published on the centralised platform). Subsequently, the format could be amended 127 in line with developments related to other initiatives, e.g. the European Single Access Point (ESAP). As under option 1, the ownership of the information and the responsibility for its accuracy would remain with the banks that produce it.

6.What are the impacts of the policy options and how would they compare?

In this section, the policy options identified to address each problem in Section 5 are assessed against three criteria:

·effectiveness: the extent to which the different policy options would achieve the objectives;

·efficiency: the analysis of the costs versus the benefits of the different policy options; and

·coherence: the coherence of the different policy options with the overarching objectives of EU policies

Based on the analysis provided, a score is assigned to each policy option for each criterion. This helps to understand the selection of the preferred policy option, i.e. the policy option with the highest overall score.

6.1.Improve the current framework for calculating risk-based capital requirements

As explained in Section 5.2.1, option 1 would address the main outstanding deficiencies of the prudential framework identified after the GFC, by strengthening the calculation of risk-based capital requirements and ensuring more comparability in this calculation across banks. Despite the ongoing COVID-19 crisis, these reforms are still needed since they address structural shortcomings of the prudential framework. Those shortcomings undermine the reliability of banks’ risk measurement and calculation of capital requirements. This can, in turn, have negative consequences on financial stability in situations of future financial crises or wider economic downturns. In its response to the CfA published in December 2020, the EBA showed that the reforms would meet their purposes to reduce the variability across banks’ internal models (due to the introduction of new constraints in using internal models) and to provide banks with more risk-sensitive standardised approaches.

However, the implementation of the final elements of the Basel III reform without adjustments would affect the overall level of EU banks’ minimum capital requirements. The EBA’s updated impact analysis, which uses the latest available estimates (based on Q4 2019 data), confirmed that implementing the final Basel III reforms under option 1 (i.e. in full alignment) would significantly increase the overall minimum capital requirements for EU banks when the reforms apply in full in January 2028 (i.e. once the transitional period would end). By this date, EU banks included in the EBA sample 128  would face an average 129 increase of 18.5% in total minimum capital requirements 130 , with 13 out of the 99 banks in the EBA sample that would have a combined capital shortfall 131 of EUR 52.2bn, all else equal (see Table 2 ). As shown in Table 4 , this impact would be driven by the introduction of the output floor (contributing to an average increase of 6.7% in total minimum capital requirements), the revisions to the operational risk framework (+3.8%), the revisions to the credit risk framework (+2.6% for the IRB and +2.4% for the SA) and the revisions to the CVA risk framework (+2.1%).

In the updated impact analysis published in December 2020, the EBA also estimated, based on a smaller sample of 45 banks 132 out of the 99banks of the EBA sample, the total shortfall in eligible liabilities (TLAC and MREL requirements) resulting from the introduction of the final Basel III standard under option 1: between EUR 7bn to EUR 8.6bn (for the purposes of comparison, the combined capital shortfall for these 45 banks represents EUR 41bn out of the EUR 52.2bn combined capital shortfall for the 99 banks of the EBA sample). Naturally, institutions could not only use own funds, but also MREL-eligible instruments to cover this shortfall.  It has to be noted, however, that the shortfalls calculated are an approximation 133  and these estimates should be considered with great caution 134 . 

The impacts of option 1 would be less significant in January 2023, when the revised rules would start to be phased in under the transitional arrangements, mainly due to the lower value of the output floor (i.e. 50%). However, they would remain relatively important: there would be an average increase of 11.8% in total minimum capital requirements, with 10 out of the 99 banks in the EBA sample having a combined capital shortfall of EUR 27.6bn.

Table 3  below highlights the profile of the EU banks that would incur a capital shortfall under option 1. The vast majority of the combined capital shortfall in 2028 is due to those EU banks that would continue to use internal models under the Basel III standards and for which the OF would become the binding capital requirement 135 . It is important to note that the capital shortfall of those banks would build up progressively during the transitional period. In addition, the EBA updated impact analysis published in December 2020 indicates that the combined capital shortfall under option 1 would be concentrated within the largest banks in the EU, while small and medium-sized banks would incur negligible capital shortfalls or no shortfalls at all (as highlighted in Section 2 of Annex 6, capital requirements of small banks would even decrease on average as a result of the reform). Due to their simpler business models and usually very limited use of internal model approaches, small and medium-sized banks would be mainly affected by the revisions to the standardised approach for credit risk. Indeed, for these banks, the impacts of the revision would remain relatively contained, as the increases for some types of credit risk exposures would be offset by capital requirements decreases for other types of credit risk exposures.

Although the updated analysis shows that the overall impacts would decrease compared to the previous EBA impact analysis 136 , option 1 would still not meet the specific objective of “no significant increase” in the minimum capital requirements of EU banks. As a consequence, EU banks that would have a substantial capital shortfall under this option might find it difficult over the next few years to build up 137 the amount of capital required under the revised rules, especially in the current low profitability environment. Consequently, EU banks might be obliged to significantly reduce their activities in certain segments, or sell existing assets (“deleverage”), which may be detrimental to the real economy. Therefore, option 1 would not meet one of the general objectives of this legislative initiative, i.e. to contribute to the steady financing of the EU economy in the context of the recovery post COVID 19 crisis.

In the two public consultations launched by the Commission services in 2018 and 2019 on the implementation of the final elements of the Basel III reform, most respondents from the banking sector and some banks’ clients raised concerns regarding the increase in capital requirements resulting from the implementation of the reform without any adjustments (see Annex 2 for a full summary of the responses). In its responses to the Commission’s CfA on the impact of the reform, the EBA supported the overall implementation of the final elements of the Basel III reform under option 1. While Member States share the overall view that those reforms are necessary to address the outstanding deficiencies of the prudential framework (see Annex 2), some of them also expressed concerns about the impacts if the international standards would be implemented without adjustments.

The introduction of EU specific adjustments in the implementation of the final Basel III reforms as proposed under option 2 would more than halve the estimated impact of option 1, as shown in Table 2 . In fact, when the reforms would apply in full in 2028, the estimated average increase in total minimum capital requirements under this option would be between 6.4% and 8.4%, with 10 banks out of 99 banks that would have a combined capital shortfall of less than 138 EUR 27bn. The lower impact of option 2 would be even more pronounced at the beginning of the transitional period (i.e. in 2023): the estimated average increase in total minimum capital requirements would range between 0.7% and 2.7%, with only 7 out of 99 banks that would have a combined capital shortfall of less than EUR 7.5bn. Table 4  below provides more details on the mitigating impacts of the various EU specific adjustments under option 2 as compared to option 1.

Under option 2, the estimated shortfall in eligible liabilities would also reduce: between less than EUR 2bn and less than EUR 2.01bn for the 45banks analysed.

The introduction of EU specific adjustments would not jeopardise the overall objective of the Basel III reform, i.e. to address the outstanding deficiencies of the prudential framework. At the same time, option 2 would implement the reform in a way that would meet the specific objective of “no significant” increase in capital requirements across EU banks, as highlighted by the above analysis.

To put the capital shortfalls under the two options into perspective, the 99 banks included in the EBA sample held a total amount of regulatory capital worth EUR 1414bn at the end of 2019 and had combined profits of EUR 99.8bn in 2019. Knowing that option 2 would halve the capital shortfalls of EU banks as compared to option 1, these banks would be in a better position 139 to build up the amount of capital required under the new rules over the next few years under option 2, without the need to abruptly reduce their exposures, particularly in certain financing activities that are key to the EU economy.

Option 2 is broadly aligned with the views expressed by the bank respondents to the two public consultations: they also proposed some specific adjustments when implementing the final elements of the Basel III reform in the EU. After a careful assessment, the adjustments that were deemed justified and appropriate have been included under option 2 (e.g. specific treatment for certain equity and specialised lending exposures, postponement of the implementation of the haircut floor framework, etc.). Other EU specific adjustments proposed under option 2 would simply maintain previously agreed rules already catering for EU specificities (e.g. SME supporting factor and the CVA exemptions). Mixed views have been expressed by the EBA, the ECB, Member States and national supervisors during expert group meeting organised by the Commission on the specific adjustments proposed under option 2. The views of Member States generally depended on the extent to which the EU specificities identified in Section 5.2.1 are present in their banking sector.

Table 2 : Summary of overall key impact estimates of the final Basel III reforms on EU banks, under both the implementation policy options 1 and 2.

Sources: Basel III reforms impact study, EBA, December 2020; European Commission estimates.

Table 3: Number and profile of EU banks incurring a capital shortfall with the full application of the final Basel III reforms in 2028, under the implementation policy options 1 and 2.

Sources: Basel III reforms impact study, EBA, December 2020.

Table 4: Breakdown of the impacts of the individual elements of final Basel III reforms under options 1 and 2 in 2028 141 and of the individual EU specific adjustments considered under option 2.

Sources: Basel III reforms impact study, EBA, December 2020; European Commission estimates.

The above impact estimates provided by the EBA are based on Q4 2019 data. They implicitly 143 assume that the economic activity and EU banks’ balance sheets would recover to pre-crisis levels by the time the final elements of the Basel III reform would start applying, i.e. by January 2023. However, it cannot be ignored that over this horizon, banks’ balance sheets might still be affected by the consequences of the COVID-19 crisis.

To this end, the EBA’s updated impact analysis also provides the Commission with a qualitative and a quantitative analysis of the combined effect of the reform and the potential consequences of the COVID-19 crisis over the short-term, i.e. at the start of the application of the reforms in 2023.

First, one key conclusion of the qualitative analysis performed by the EBA is that the effects of the COVID-19 crisis would not necessarily lead to higher capital impacts under the final Basel III standards than they would do under the current prudential framework. Indeed, the EBA expects an increase in banks’ capital requirements as a result of the expected deterioration in ratings, higher probabilities of default and higher expected losses caused by the COVID-19 crisis. In the short-term, this increase would occur under the current prudential framework irrespective of whether the final Basel III standards would be implemented and would, to a certain extent, mitigate the capital impact due to the reform. For instance, for banks using internal models under the Basel III standards, the impact of introducing the output floor might be mitigated by an increase in the capital requirements generated by internal models, which might happen in the coming years due to the higher credit risk of banks’ borrowers.

Second, the quantitative analysis performed by the EBA defines two hypothetic, adverse economic scenarios which would imply a significant deterioration of the financial situation of EU banks’ borrowers over the short-term horizon as a potential negative consequence of the COVID-19 crisis (the two scenarios differ in terms of how severe 144 this deterioration and the resulting increase of their non-performing loans would be, all other risks of EU banks being equal). Under each scenario, the EBA recalculated the average change in the minimum capital requirements and the resulting capital shortfall of EU banks only under option 2 and only in the short-term, i.e.at the start of application of the reforms in 2023.

As compared to the impact of option 2 (i.e. average increase in banks’ minimum capital requirements between +0.7% and +2.7% and a combined capital shortfall of less than EUR 7.5bn as shown in Table 2 ), the average increase in banks’ minimum capital requirements under the less (resp. more) severe if the adverse economic scenarios in the short-term would go up to between +4% and +6% (resp. +5% and +7%) leading to a capital shortfall of EUR 30.4bn (resp. EUR 59.8bn) .

As recognised by the EBA in its report, the impact estimates under these two adverse scenarios are more significant than using the Q4 2019 data but the assumptions and the methodology used by the EBA in these scenarios were very conservative and need to be interpreted with caution. In fact, the EBA methodology uses a stress-testing approach which assumes a simultaneous deterioration in the financial situation of all borrowers of EU banks, i.e. even for those borrowers that do not currently experience difficulties. Second, the higher combined capital shortfall under the adverse scenarios would not be only due to higher capital shortfalls incurred by banks under option 2 based on Q4 2019 data, but also due to more banks that would incur a shortfall under that scenario. Some of the banks that would incur a capital shortfall under the adverse scenarios would also incur a shortfall if the current prudential framework was used instead of the final Basel III standards (the EBA analysis does not disentangle the two effects – the COVID-19 effect and effect of the final Basel III reforms).

Taking into account these caveats, the EBA analysis still provides a useful indication of what the upper bound of the impacts of the final elements of the Basel III reform could be if the financial situation of EU banks were to significantly deteriorate over the next few years. The continuing EBA monitoring of the impacts of the reform on EU banks will be particularly useful to inform EU co-legislators during the first stage of negotiations of this legislative initiative about the evolution of the impacts of those reforms with more concrete first signs of the consequences of the COVID-19 crisis on EU banks’ balance sheets.

Option 3 would lead to the same overall impacts on total capital requirements as option 2 but would give EU banks two more years to comply with the new capital requirements. Due to the uncertainty with the evolution of the financial situation of EU banks’ clients, as shown in the above analysis, this delay would give banks time to absorb potential losses coming from the most fragile borrowers that are (or will be) affected by the COVID-19 pandemic, while still leaving them enough room to support the financing of the EU economy during the recovery phase.

Option 3 would provide much needed certainty to the banks that the impact of the EU implementation of Basel III is manageable. Banks usually refer to the EBA’s estimates on how much capital requirements would increase if the reforms were implemented in full alignment with the Basel III standards. Given the lack of clarity about the EU’s implementation approach and missing certainty on the timeline, banks expect a significant increase in capital requirements as a result of the implementation and the most impacted or weakest amongst them could see this as another reason 145 to hold back from using their capital to lend. Coming forward with option 3 for the implementation would signal that the impact would be quite limited and provide certainty for banks’ capital planning and lending decisions throughout the recovery phase while reaffirming the EU’s international commitment to the Basel III implementation. In addition, clear communication on the difference between entry into force and effective application of the new requirements would help reinforce the signalling effect. Postponing further, by contrast, would nurture speculations about the EU’s commitment to implement the final elements of the Basel III reform, damage the EU’s reputation as a reliable partner in international fora/negotiations, and be likely interpreted as a sign that the EU banking sector is too weak to accommodate the Basel III reforms (with potential negative consequences for EU banks’ market valuations and funding costs).

In their communications with the Commission, a good part of the banking sector and some of their clients have been asking for a postponement of this legislative proposal implementing the final elements of the Basel III reforms. Option 3, while not entailing further postponement of the legislative proposal, but given the extended implementation timeline, would largely meet this demand. Member States and supervisors were more supportive of the idea of extending the implementation timelines as an alternative to structural deviations from the final Basel III standards. Most Member States and key MEPs have recently restated their support for a timely and faithful implementation of the final Basel III reforms – taking account of EU specificities – notwithstanding the COVID-19 crisis. They expect the Commission to table a legislative proposal by mid-2021.

In light of the above analysis, which led to the overall score of each policy option in Table 5  below, option 3 is deemed the preferred policy option to improve the current framework for calculating risk-based capital requirements.

Table 5 : Comparison of policy options against effectiveness, efficiency and coherence criteria

Magnitude of impact as compared with the baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

6.2.Dedicated capture of ESG risks in the prudential framework

Option 1 would provide incentives for banks to improve the management of ESG risks by reinforcing banks’ obligations as regards the management of ESG risks as well as the supervisory review thereof. This would help ensure that banks would be able to handle their exposures to ESG risks more effectively and aligning their investment strategies accordingly, in line with the stated objective pursued by this initiative. In addition, the suggested reinforcement of disclosure requirements would allow market participants to scrutinise the degree of exposure of banks to ESG risks as well as how close those banks are to delivering on any sustainability commitments they had already made (or would make in the future).

By introducing a dedicated review clause to adapt capital requirements based on concrete evidence, option 1 would pave the way for better aligning capital requirements with the riskiness of ESG assets.

Through the envisaged change to bring forward the delivery date for the EBA report, option 1 would allow to minimise delays in effectively addressing any additional underlying problems that the report may identify. At the same time, it would ensure that any changes in capital requirements would be based on adequate evidence, which would be coherent with the principles of risk-based capital requirements and thus the objective of financial stability.

Option 1 would result in some costs for banks as they would have to adapt their risk management systems and processes and collect the necessary data. To the extent that the data must be obtained from customers, the latter would also incur additional costs as a result. Supervisors would incur costs as their supervisory processes would become more complex.

In contrast to option 1, option 2 would adapt capital requirements based on currently available evidence without awaiting the EBA report on the relative riskiness of exposures. It would thus have the advantage of effectively responding to calls for an early intervention in view of the urgency of environmental (and more broadly sustainability) challenges. However, research on how the riskiness of bank exposures differs based on sustainability criteria is in its early stages and empirical evidence in this area is still limited. The available research is focused on a very narrow subset of bank exposures, namely lending for the financing of the construction/purchase of energy-efficient buildings or for the “upgrading” of energy inefficient buildings. That research provides tentative evidence that such lending may be slightly less risky than “traditional” mortgage lending 146 . At the same time, other research concludes that such differences either do not exist or can be explained by other factors than environmental ones that are correlated with “green-ness”, such as income differences between borrowers.

An attempt to introduce changes to capital requirements for ESG risks based on this limited and inconsistent evidence would likely result in an inadequate calibration of such capital requirements. This would be incoherent with the principles of risk-based capital requirements and would negatively impact financial stability in particular if capital requirements for certain exposures were to be too low to cover the real risks. Also, a premature change in capital requirements would entail the risk that new evidence may subsequently emerge that contradicts the assumptions on the basis of which the changes had been made in the first place, thereby creating the need to undo these changes. At the same time, the available evidence does not allow a detailed assessment of the impact that option 2 might have.

Many stakeholders recognise that prudential requirements for banks must reflect ESG risks, with the EBA recommending “to incorporate ESG risk-related considerations in directives and regulations applicable to the banking sector” 147 . This holds true in particular with respect to proposals to strengthen risk management requirements, supervisory review and stress testing as well as disclosure requirements, even though with respect to the latter some stakeholders have cautioned that these must not result in disproportionate costs.

However, views vary as regards changes to capital requirements based on ESG factors. Banks tend to be open to the idea of lower capital requirements for exposures to sustainable activities and/or products but strongly oppose higher capital requirements for exposures to activities and/or products deemed to be unsustainable. Civil society stakeholders tend to view increased capital requirements for unsustainable activities and/or products more favourably, while some consider that the potential benefit of a decrease in capital requirements for sustainable activities and/or products would be outweighed by a potential negative impact on financial stability. Supervisors consider that any change in capital requirements must be based on solid evidence of risk differentials based on ESG factors, which they do not consider to be available at present.

In light of the above analysis, which led to the overall score of each policy option in Table 6  below, option 1 is deemed the preferred policy option to introduce a dedicated capture ESG risks in the prudential framework.

Table 6 : Comparison of policy options against effectiveness, efficiency and coherence criteria

Magnitude of impact as compared with the baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

6.3.Improve the consistency in the application of supervisory powers

On supervisory powers

In relation to supervisory powers, option 1 would draw awareness across the Union to the importance of assessing prudentially relevant events in a timely manner. The suggested mandatory ex-ante notification, e.g. of a merger, would allow supervisors to be informed at an early stage and to react promptly, if deemed necessary from a prudential viewpoint. Yet, MS would remain free to implement corresponding supervisory powers. The likely effect would be a perpetuation of the absence of equal rules and limitations for several authorities, including the ECB, to intervene ex ante. Option 1 would imply some costs for banks in the form of administrative burden (due to the notification obligation) and would for individual banks depend on whether ex ante notification obligations already exist in the relevant MS. However, the materiality thresholds for notifications included in this option would limit those costs, as only event exceeding the thresholds would need to be notified.

On the supervisory powers and procedures concerning the suitability assessment, option 1 would mean assessing ex ante all members of the management body by the competent authorities, without considering the characteristics of banks or of the different entities within the group in which those persons hold their functions. While this would ensure a high degree of harmonisation, it would require significant additional efforts from authorities that currently perform ex-post assessments, notably to the numerous small banks in their supervisory remit.

By contrast to option 1, the harmonisation of supervisory powers in EU law as suggested in option 2 would achieve a level playing field, endowing all competent authorities with sufficient capacity to react to prudentially relevant events. It would impact to a different extent banks and supervisory authorities, depending on whether supervisory powers 148 and ex ante notification obligations already exist and how they are designed. The incurred costs would in most cases remain ultimately less important than those for ex-post supervisory assessments. As in the case of option 1, the costs would be contained due to the materiality thresholds. 

As regards the fit-and-proper assessment, under option 2, the ex-ante assessment would be introduced only for members of the management body in the parent entity and in material subsidiaries of large banks. Competent authorities would therefore be able to continue ex-post assessments for smaller banks which would entail no additional costs. Option 2 would extend the scope to key function holders and require competent authorities to assess them ex ante for the same type of entities as for the members of the management body. The proposed framework would also provide further specifications concerning the supervisory procedures for the assessment of both, board members and key function holders. Furthermore, it would set out criteria for carrying out the assessment of key function holders 149 .

Overall, option 2 would achieve a more balanced reform of supervisory powers than option 1. It would ensure a sufficient level of supervisory convergence and keep the administrative burden to a reasonable level.

On sanctioning powers

Option 1 would clarify the distinction between enforcement measures and sanctioning tools for supervisors. This clarification would leave nonetheless an important discretion to Member States to introduce or not additional enforcement 150 and sanctioning powers in their national laws.

Furthermore, option 1 would achieve a significant increase of the harmonisation of sanctioning powers by introducing a general power for supervisors to sanction all potential breaches of the regulatory framework. However, this option might create legal uncertainty as regards its application. A generic clause might not be effective when breaches of a bank are not identified by national and European supervisors under a common standard. In addition, some Member States impose the obligation to state the breach for which a sanction is applicable 151 . Therefore, introducing a general sanctioning power could potentially raise constitutional issues in some Member States 152 .

Finally, under option 1 the basis for the calculation of pecuniary measures imposed on EU banks to sanction breaches would be specified by MS, agreeing on a common definition in a Level 2 text (e.g. Regulatory Technical Standards) that could leave some additional flexibility.

None of the above elements is expected to lead to undue increases in costs for either banks or their supervisors 153 .

Option 2 would provide more legal certainty as regards breaches that would be sanctioned than option 1. MS would implement an identical list of key CRD/CRR requirements, which would eliminate inconsistencies as regards the scope of sanctioning powers of the competent authorities. This option considers the proportionality principle. It increases the level of harmonisation as regards sanctioning powers without leaving supervisors and credit institutions with legal uncertainty which is the case for solely generic sanctioning powers under option 1.

Like under option 1, all the important breaches would be subject to administrative sanctions by supervisors and clarifications would be provided on the articulation between enforcement and punitive measures taken by supervisors. However, option 2 would go beyond option 1 by providing an additional enforcement tool to supervisors in the form of periodic penalty payments).

Finally, option 2 suggests a harmonisation of the calculation of pecuniary sanctions based on the notion of total annual net turnover. The use of total annual net turnover as a criterion would benefit from a clarification on the highest level of the EU legislation (e.g. CRD), compared to option 1 (Regulatory Technical Standards).

Similarly to option 1, the changes contemplated under option 2 would not lead to undue increases in costs for banks or for their supervisors (for the latter, the sanctioning procedures could stay unchanged, it is only the scope of breaches to which those procedures would apply that would be expanded).

* * *

Most stakeholders acknowledge that differences in the powers available to supervisors and their application across MS are contrary to the level playing field principle and effective supervision. While supervisors (in particular the ECB) and civil society stakeholders highlight the prudential risks and the uncertainty resulting from the lack of certain powers, definitions and common procedures, the banking industry is concerned about possible distortions to the competition across MS and notes that the status quo would prevent groups from reaping the synergies expected from cross-border acquisitions.

As regards concrete policy options to address the problems, the views of MS and supervisors are largely correlated with their current practices. In particular, those being home and/or predominantly in charge of smaller banks are concerned about potentially increased administrative burden. Their support is hence conditional on the introduction of materiality thresholds which would exempt a significant number of events and the provision of targeted flexibility (particularly regarding the scope of ex-ante fit-and-proper assessments) allowing for risk-based adjustments as envisaged under option 2. Some MS and the majority of banks, by contrast, prefer limiting the flexibility left to MS and supervisors to the extent possible.

In light of the above analysis, which led to the overall score of each policy option in Table 7  below, option 2 is deemed the preferred policy option to improve the consistency in the application of supervisory powers

Table 7 : Comparison of policy options against effectiveness, efficiency and coherence criteria

Magnitude of impact as compared with the Baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

6.4.Reduce disclosure costs and improve market access to bank prudential information 

Both options would increase the efficiency and effectiveness of disclosures, particularly for small and non-complex banks. These banks would be relieved from the burden to prepare disclosures of prudential information 154 , which will be disclosed centrally based on prudential information provided to supervisors. For other banks, the benefits from option 1 may be limited: given that they would still need to manage their disclosures of qualitative information, they may simply decide to continue disclosing the required qualitative and quantitative information as they currently do. At the same time, option 1 would not entail any additional costs for those banks, nor it would entail any additional costs for banks or supervisors more in general as changes to their existing systems would not be required.

For other stakeholders (e.g. investors, academics) option 1 would achieve all benefits associated with a single, free access point to prudential data (i.e. quantitative information) on all EU banks. It would allow them to undertake a meaningful analysis across EU banks. The disadvantage (but not associated with additional cost) of this option for these stakeholders would be that in order to access banks’ qualitative information, they would still need to gather it from the websites of individual banks like they currently do.

Option 2 would entail the same benefits as option 1. In addition, it would eliminate all its drawbacks: all banks would be able to use the centralised platform for all their disclosures (and hence avoid duplication), while the other stakeholders would have a single point of access to those disclosures.

In light of the above analysis, which led to the overall score of each policy option in Table 8  below, option 2 is deemed the preferred policy option to reduce disclosure costs and improve market access to bank prudential information.

Table 8 : Comparison of policy options against effectiveness, efficiency and coherence criteria

Magnitude of impact as compared with the Baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

7.Preferred policy options

7.1.Effectiveness

The implementation of the final elements of the Basel III reform under the preferred policy option will address the shortcomings of risk-based capital requirements under the prudential framework that have been identified during the GFC. This will ensure an accurate measurement of risk and an adequate capitalisation of institutions which will in turn ensure financial stability. A more stable and resilient EU banking sector will, in turn, effectively strengthen the trust of global financial markets and international partners in the EU financial system.

The preferred option will also take due account of the specificities of the EU banking sector through a number of specific adjustments to the Basel III standards as well as a longer implementation timeline of the reforms. In this way the preferred option will most effectively help to mitigate the risk associated with a sharp increase in capital requirements for institutions, especially in view of the COVID-19 –crisis. This will ensure a smooth provision of essential financial services to the EU economy during the recovery phase and in the long term.

The preferred policy option will enhance the focus on ESG risks in the prudential framework with a dedicated approach to capture those risks. Improved management of ESG risks will ensure that institutions will be better prepared to withstand those risks, in particular risks due to climate change. Moreover, the explicit reference to ESG risks in the supervisory framework will increase supervisors’ focus of those risks. Finally, the enhanced transparency about institutions’ exposures to those risks will give markets the necessary tools for an effective monitoring of the sustainability of institutions’ activities. The combination of these measures will create the necessary incentives for institutions to allocate more financing to more sustainable investments.

The preferred policy option regarding supervisory and sanctioning powers will further harmonise and strengthen the toolkit available to supervisors across the Union, improving the robustness, application and enforcement of the prudential framework applicable to institutions. By strengthening supervisors’ powers to ensure institutions’ compliance with the prudential framework across the Union, and by giving supervisors the necessary powers to intervene in transactions that can have a significant prudential impact on institutions, the preferred option will ensure a more effective supervision of institutions and therefore a safer banking sector.

Next to supervision, scrutiny by market participants and other stakeholders constitutes another important control mechanism in the prudential framework. The effective functioning of this mechanism is conditional upon the transparency of banks’ activities, their financial position and the risks they face. The preferred policy option would ensure that information (i.e. qualitative and quantitative) that institutions have to disclose would be easily accessible in one place and would be easily comparable, thus enhancing the ability of clients, investors and other market participants to monitor and exert market discipline on institutions.

7.2.Efficiency

The preferred policy option will achieve the desired objectives with enhanced efficiency. By adapting the final Basel III framework to several specificities of the EU banking sector, the preferred option would avoid disproportionate capital requirements for certain financial products or activities provided by banks and are essential to the EU economy. Without adapting the Basel III standards to EU specificities, the resulting increase in capital requirements would likely be significant, resulting in higher costs for institutions’ clients, including SMEs. This could ultimately undermine the clients’ economic activities or capacity to hedge their financial risks. The proposed two-year postponement of the date of application of the final elements of the Basel III reform in the preferred policy option will enable banks to support the recovery post COVID-19 crisis over the short- to medium-term whilst providing EU banks with regulatory certainty early on, thereby allowing for proper planning and a smooth implementation of the revised prudential framework.

The preferred policy option to address ESG risks will provide the necessary incentives for banks and, indirectly, their clients to take due account of the sustainability of their economic activities and will therefore ensure a more efficient allocation of economic resources. This will, in turn, support the effort to transform the EU economy into a resource-efficient, sustainable and competitive economy.

As regards supervisory and sanctioning powers, the preferred option addresses the current fragmentation of supervisory powers by ensuring that scarce supervisory resources are used in the most efficient way. This is done by allowing supervisors to focus only on events that can materially affect the prudential standing of banks. While the preferred policy option involves an increase in costs for at least some supervisors (especially for those that currently do not have the powers that would be introduced under the preferred option) and the banks they supervise, the increase in costs is limited (due to the in-built proportionality of the rules that focuses only on material events). The costs would be outweighed by the benefits of more harmonisation which will contribute to levelling the playing field in the single market and render supervision in the Banking Union more efficient.

Finally, under the preferred policy option for the disclosure of prudential information, banks will benefit from a more efficient system that integrates supervisory reporting and disclosure, and thereby reduces their administrative burden. At the same time, access to bank data for stakeholders will become more efficient as all the relevant information will be available in one place.

7.3.Coherence

In addition to the effectiveness and efficiency of the preferred policy options chosen to address the problems identified, the preferred options shall be coherent with each other as well as the whole package with other policy initiatives at EU level.

The prudential framework for banks in the Union consists of three main pillars, each of which plays a distinct, key role in ensuring the stability of individual institutions and the banking sector as a whole. The preferred policy options propose changes to each of these pillars, which will increase the coherence of the overall framework.

The first pillar consists of minimum capital requirements, in the form of quantitative and qualitative rules. In 2013, in the wake of the GFC, these rules were moved from a Directive to a Regulation to form a “Single Rule Book”, and a large number of national options and discretions was removed. This change improved the uniformity of application of minimum capital requirements in all MS, closed regulatory loopholes and thus contributed to a more effective functioning of the single market for banking services. The current proposal further improves the consistent application of these rules by limiting banks’ freedom in calculating their capital requirements; this will make capital requirements and reported capital ratios more comparable across the Union.

The second pillar consists of the supervisory review of banks’ activities and risks. This review is crucial to ensure a consistent application of the prudential framework, in particular in the Banking Union. However, where powers and tools made available to supervisors in conducting this review differ across the Union, prudential rules are often applied inconsistently. The preferred options will address shortcomings in respect of strengthening supervisory powers and the sanctioning of breaches.

The third pillar consists of market scrutiny. For banks to be subject to comparable levels of scrutiny, stakeholders must have access to comparable information. The preferred policy options will improve such access and will also increase the comparability of the disclosed information.

Furthermore, the preferred policy options are coherent with other policy initiatives at EU level, in particular:

·The Banking Union aims to increase financial integration and stability in the Economic and Monetary Union. Common supervision is a central element of the Banking Union. However, where the ECB exercises direct supervisory powers set out in national law transposing the CRD, the ECB does not have the same range of powers with regard to all banks under its supervision because of differences in the transposition. This impinges on attaining the objective of efficient and harmonised supervision within the Banking Union. The present initiative aims at addressing some of these obstacles.

·The CMU aims to improve the access to financing for companies and projects across the Union. This overall aim is coherent with the general objectives of the present initiative, whereas the specific measures are complementary: banking regulation mostly relates to bank financing, while CMU mostly concerns non-bank financing. This initiative takes into account the fact that EU banks also play a role in facilitating companies’ access to financial markets, and ensures, by considering EU specificities, that the proposed measures do not unduly constrain this important role of banks.

·The EGD Communication announced that environmental risks would be better integrated into the EU prudential framework, and that the suitability of the existing capital requirements for green assets would be assessed. The present initiative puts this announcement into practice. It will help ensure that the banking sector can play an appropriate role in achieving the ambitious aims of the EGD.

·The ESAP aims at providing investors with easy access to regulated financial information of companies listed on the EU‘s regulated markets. Although the scope of this initiative is different from the scope of the ESAP, the aim of the two is fully compatible. Depending on the final design of the ESAP, the EBA centralised disclosure platform could either feed information into the ESAP or the ESAP could provide a gateway to the information stored on the EBA platform.

7.4.REFIT (simplification and improved efficiency)

This initiative is aimed at completing the EU implementation of the international prudential standards for banks agreed by the BCBS between 2017 and 2020. It would complete the EU implementation of the Basel III reform that was launched by the Basel Committee in the wake of the GFC. That reform was in itself a comprehensive review of the prudential framework that was in place before and during the GFC, namely the Basel II framework (in the EU that framework was implemented through Directive 2006/48/EC, i.e. the original CRD). The Commission used the results of that review, together with input provided by the EBA and other stakeholders, to inform its implementation work. A fitness check or refit exercise of the EU implementation of the Basel III reform has not been carried out yet because all the elements of the reform need to be put in place before one can be carried out.

Implementing the final elements of the Basel III reform with the EU specific adjustments envisaged under the preferred option would simplify the risk-based capital framework, notably by removing more complex internal models approaches to calculate capital requirements in for operational and CVA risks and by limiting the scope of internal models for credit risk. This would positively impact the recurring administrative and operational costs of EU banks (see Section 1 of Annex 6) and facilitate their supervision.

In relation to disclosure, this legislative initiative would introduce measures to reduce redundancies in respect of information reported to supervisors and disclosures to markets by centralising disclosures via a European data infrastructure based on supervisory reporting data. This would contribute to a reduction in the administrative burden of banks, in particular small ones.

8.How will actual impacts be monitored and evaluated

The changes contained in this legislative initiative would start applying in 2025 and become fully applicable in 2030. After that date, an evaluation of the reform will be carried out in principle three years after the latter date.

The below Table 9 presents some indicators that would help the Commission to monitor and evaluate the effectiveness and the appropriateness of the proposed preferred policy options to achieve the specific objectives mentioned in Section 4, based on the data/information available. The Commission will mostly use data/information from the European Supervisory Authorities (the EBA, the ESMA and the ESRB), the national supervisory authorities and the ECB/SSM, the BCBS, and other market data indicators to monitor and evaluate the impacts of the proposed preferred policy options.

Table 9: Summary of indicators to monitor and evaluate the impacts of the preferred policy options

(1)

See “ Opinion of the European Economic and Social Committee on the Report of the de Larosière Group (Own-initiative opinion)”.

(2)

The total amounts of state aid used by Member States to shore up the banking sector during and after the GFC (i.e. over the 2008-2017 period) is estimated to €2 trillion.

(3)

While the annual EU GDP growth remained slightly positive in 2008 (+0.8%), it dropped in 2009 (-4.2%) (See https://ec.europa.eu/economy_finance/publications/european_economy/2010/interim_forecast_2010-09_en.htm ).

(4)

Members of the BCBS comprise central banks and bank supervisors from 28 jurisdictions worldwide. Among the EU Member States, Belgium, France, Germany, Italy, Luxembourg, the Netherlands, and Spain, as well as the European Central Bank are members of the BCBS. The European Commission and the EBA participate in BCBS meetings as observers.

(5)

The consolidated Basel III framework is available at https://www.bis.org/bcbs/publ/d462.htm .

(6)

Regulation (EU) No 575/2013 of the European Parliament and of the Council of 26 June 2013 on prudential requirements for credit institutions and investment firms and amending Regulation (EU) No 648/2012 (OJ L 321, 26.6.2013, p. 6).

(7)

Directive 2013/36/EU of the European Parliament and of the Council of 26 June 2013 on access to the activity of credit institutions and the prudential supervision of credit institutions and investment firms, amending Directive 2002/87/EC and repealing Directives 2006/48/EC and 2006/49/EC (OJ L 176, 27.6.2013, p. 338).

(8)

Regulation (EU) 2019/876 of the European Parliament and of the Council of 20 May 2019 amending Regulation (EU) No 575/2013 as regards the leverage ratio, the net stable funding ratio, requirements for own funds and eligible liabilities, counterparty credit risk, market risk, exposures to central counterparties, exposures to collective investment undertakings (CIU), large exposures, reporting and disclosure requirements, and Regulation (EU) No 648/2012.

(9)

Directive (EU) 2019/878 of the European Parliament and of the Council of 20 May 2019 amending Directive 2013/36/EU as regards exempted entities, financial holding companies, mixed financial holding companies, remuneration, supervisory measures and powers and capital conservation measures.

(10)

See https://www.bis.org/publ/bcbs189.htm .

(11)

CVA is an accounting adjustment to the price of a derivative to account for counterparty credit risk. For more details, see Section 1.5 in Annex 5.

(12)

These were the only significant changes to the part of the standards that deal with risk-based capital requirements that were introduced as part of the first stage of the Basel III reform.

(13)

A minimum requirement on large exposure limits was already a feature of Union legislation, but was a novelty for the Basel standards.

(14)

More specifically the capital conservation buffer (CCB), the countercyclical capital buffer (CCyB), the systemic risk buffer (SRB), and capital buffers for global and other systemically important banks (respectively, G-SII and O-SII).

(15)

Those first set of reforms have also been implemented in most jurisdictions worldwide as can be observed in the eighteenth progress report on adoption of the Basel regulatory framework published in July 2020 (see https://www.bis.org/bcbs/publ/d506.htm ).

(16)

Before the publication of this impact assessment, these data will be updated with the latest figures as of Q4 2020 which will be published by the EBA in Q2 2021 it their next Risk Assessment Report.

(17)

CET 1 capital is the form of banks’ capital recognised by the prudential framework for having the highest capacity to absorb unexpected losses that arise during the normal course of banks’ businesses. It is mainly composed of banks’ common shares and retained earnings.

(18)

The CET 1 ratio is calculated by dividing a bank’s CET 1 capital by its total assets weighted by their relative riskiness (“risk-weighted assets” or RWA). The CET 1 ratio is a key indicator of banks’ resilience to idiosyncratic risks.

(19)

EBA Risk Assessment of the European banking system, December 2020 (see https://www.eba.europa.eu/risk-analysis-and-data/risk-assessment-reports ).

(20)

Since 2016 the weighted average leverage ratio of EU banks, which will become a binding requirement in June 2021, remained relatively stable and well above the minimum requirement that would be applicable in the EU (3% for all EU banks from June 2021, at least 3.5% for G-SIIs from January 2023). The fact that the leverage ratio did not increase similar to the capital ratios could be explained by the fact that EU banks reduced their exposures to risky assets over the last few years which has no effect on the leverage ratio.

(21)

See https://ec.europa.eu/info/publications/200428-banking-package-communication_en .

(22)

See https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32020R0873&from=EN .

(23)

A comprehensive list of such measures has been collected by the ESBR, see “ Policy measures in response to the COVID-19 pandemic”.  

(24)

In its COVID-19 vulnerability analysis published in July 2020, the ECB showed that the largest euro area banks would be sufficiently capitalised to withstand a short-lived deep recession and that the number of those banks with insufficient capital resources in case of a more severe recession would be limited (see https://www.bankingsupervision.europa.eu/press/pr/date/2020/html/ssm.pr200728_annex~d36d893ca2.en.pdf ).

(25)

See https://www.bis.org/bcbs/publ/d424.htm

(26)

See http://www.g20.utoronto.ca/2018/2018-03-30-g20_finance_communique-en.pdf . The relevant passage of the statement - the latter was agreed by the European Union as a member of the G20 - reads: “We welcome the finalisation of Basel III, which completes main elements of the post crisis reforms. We remain committed to the full, timely and consistent implementation and finalisation of the reforms and their evaluation to help identify and address any material unintended consequences and ensure that the reforms accomplish their objectives.” The message has regularly been repeated in subsequent G20 press statements. 

(27)

See https://ec.europa.eu/commission/presscorner/detail/en/SPEECH_19_6269 .

(28)

More specifically to 1 January 2023 for the starting date of application and to 1 January 2028 for the full application of the final elements of the reform.

(29)

See https://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1588580774040&uri=CELEX:52019DC0640 .

(30)

See https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12635-Renewed-sustainable-finance-strategy

(31)

See https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:52018DC0097 .

(32)

https://ec.europa.eu/info/publications/sustainable-finance-high-level-expert-group_en

(33)

https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32020R0852

(34)

  https://www.eba.europa.eu/sites/default/documents/files/document_library/EBA%20Action%20plan%20 on%20sustainable%20finance.pdf .

(35)

See https://www.ngfs.net/en .

(36)

https://www.ngfs.net/sites/default/files/medias/documents/ngfs_guide_for_supervisors.pdf

(37)

Counterparty risk relates to bilateral transactions (e.g. derivatives or securities financing transactions) and include two types of risks: the risk of losses upon the default of the counterparty (default risk) and the risk of market value losses on bilateral transactions due to the decrease in the creditworthiness of the counterparty (CVA risk).

(38)

In the prudential framework, banks have first to calculate the corresponding RWAs for those risks (the RWAs are calculated by multiplying the size of a bank’s exposure (e.g. a loan) with the appropriate risk weight, which captures the degree of riskiness of the exposure) and then determine the capital requirements as a small portion of those RWAs.

(39)

See for instance https://www.wsj.com/articles/basel-committee-to-stop-banks-gaming-risk-models-1446472711 ; https://blogs.cfainstitute.org/marketintegrity/2015/03/17/bank-risk-weighted-assets-how-to-restore-investor-trust/ ; or p. 53 of https://voxeu.org/sites/default/files/file/Post_Crisis_Banking_Regulation_VoxEU.pdf .  

(40)

 Before the publication of this impact assessment, these data will be updated with the latest figures as of Q4 2020 which will be published by the EBA in Q2 2021 it their next Risk Assessment Report

(41)

See https://eba.europa.eu/regulation-and-policy/supervisory-benchmarking-exercises .

(42)

In reality, the proportion of EU banks using internal model approaches would be much smaller since the EBA data collection does not include the vast majority of the thousands of small and medium-sized banks established in the EU, most of which do not use internal model approaches.

(43)

For the BCBS Regulatory Consistency Assessment Program reports on the variability of risk-weighted assets, see https://www.bis.org/bcbs/publ/d363.htm , https://www.bis.org/publ/bcbs267.htm , and https://www.bis.org/bcbs/publ/d337.htm.

(44)

For EBA benchmarking exercises see https://eba.europa.eu/regulation-and-policy/supervisory-benchmarking-exercises .

(45)

Risk of losses due to adverse price movements in trading activities.

(46)

See https://www.bis.org/publ/work844.htm .

(47)

Banks change the design and calibration of (parts of) models to keep pace with changing portfolios, new data and modelling techniques.

(48)

In principle, supervisors have the ability to withdraw the permission to use internal models from a bank. However, such measure can have a significant impact on the bank that loses the permission, which is why it is usually used as a very last resort.

(49)

For the current status of this exercise, EBA report on progress made on its roadmap to repair IRB models, July 2019 (see https://eba.europa.eu/eba-publishes-report-on-progress-made-on-its-roadmap-to-repair-irb-models ).

(50)

In April 2021, the ECB published the outcomes of the TRIM exercise which resulted in Euros 275 billion increase in RWAs over the last three years and more than 5,000 findings for banks to remediate (see https://www.bankingsupervision.europa.eu/press/pr/date/2021/html/ssm.pr210419~94c010eb9d.en.html ).

(51)

According to the EBA Discussion Paper on management and supervision of ESG risks for credit institutions and investment firms, “[s]ocial risks are the risks posed by the exposure of institutions to counterparties that may potentially be negatively affected by social factors”, with social factors in turn being “related to the rights, well-being and interests of people and communities, which may have an impact on the activities of the institutions’ counterparties”.

(52)

According to the EBA Discussion Paper on management and supervision of ESG risks for credit institutions and investment firms, “[g]overnance risks are the risks posed by the exposure of institutions to counterparties that may potentially be negatively affected by governance factors”, with governance factors in turn covering “governance practices of the institutions’ counterparties, including the inclusion of ESG factors in policies and procedures under the governance of the counterparties”.

(53)

According to the EBA Action Plan on Sustainable Finance, “[e]nvironmental and social considerations are often intertwined, as especially climate change can exacerbate existing systems of inequality. The governance of public and private institutions, including management structures, employee relations and executive remuneration, plays a fundamental role in ensuring the inclusion of social and environmental considerations in the decision-making process.

(54)

See https://ec.europa.eu/info/business-economy-euro/company-reporting-and-auditing/company-reporting/non-financial-reporting_en .

(55)

Since its entry into force in November 2014, the Single Supervisory Mechanism (SSM), established within the ECB, is in charge of the direct supervision of 115 significant banks of the Members of the Eurozone (82% of banking assets in those countries) and of non-Eurozone Member States which have entered, on voluntary basis, into close cooperation with the ECB. The other banks of those Member States (the less significant banks) continue to be supervised by their national supervisors, in close cooperation with the SSM. The action of the SSM is framed by the SSM Regulation and the SSM Framework Regulation, which specifies the functioning and powers of the SSM, within the remit of the broader supervisory framework set at European level by the CRD and transposed in national laws.

(56)

For instance, the power to require institutions to have additional own funds in excess to those required pursuant to Pillar 1 requirements; or the power to restrict or limit the business, operations or network of institutions.

(57)

For instance, the withdrawal of a banking license.

(58)

For instance, fines paid to the supervisory authority.

(59)

The difference of approaches from MS in this area may come from, differences in their legal system prior to the application of Union law, the structure of the national banking sector or the supervisory culture of the MS.

(60)

For instance, breaches of capital requirements, internal models approval and remuneration requirements.

(61)

The SSM estimates that for several significant institutions under its direct supervision breaches of CRD/CRR requirements with material impacts cannot be sanctioned due to the lack of powers in the relevant national laws.

(62)

Some MS introduced additional powers to impose periodic penalty payments, for instance daily payments until the breach justifying this penalty has ended.

(63)

Some MS further specified the definition of ‘total annual net turnover’ (used in the determination of the maximum amount of administrative penalties) since the CRD lacks details on the inclusion of important elements reflecting the ordinary activities of institutions, for instance interest payables and similar charges, commissions and fees, net profit on financial operations.

(64)

Some MS introduced an ex ante notification requirement for banks that allows supervisory authorities to oppose the operation in case of prudential concerns.

(65)

According to the applicable EBA/ESMA guidelines key function holders means persons who have significant influence over the direction of an institution, but who are neither members of the management body and are not the CEO. They include the heads of internal control functions and the CFO, where they are not members of the management body.

(66)

See https://www.eba.europa.eu/regulation-and-policy/internal-governance/joint-esma-and-eba-guidelines-on-the-assessment-of-the-suitability-of-members-of-the-management-body .

(67)

See https://www.bankingsupervision.europa.eu/ecb/pub/pdf/ssm.fap_guide_201705_rev_201805.en.pdf .

(68)

For example, the Guidelines identified more specifically key functions holders as the heads of the internal control functions and the CFO, where they are not members of the management body, and provide provisions identify other key functions holders based on an assessment of their materiality by institution.

(69)

Assessment is carried out ex post in eight MS (DK, DE, EE, EL, FR, FI, IT, AT), and partially ex post in four MS (CZ, PL, SI, SE).

(70)

For instance, some MS will only identify key function holders as those performing anti-money laundering or audit duties within an institution while others MS will rely on the definition provided in the EBA/ESMA guidelines, or even extend the scope of this definition.

(71)

Cf. SWD(2017) 336 final.

(72)

On the issue of fit-and-proper assessments see the in-depth analysis requested by the ECON committee: Is the current “fit and proper” regime appropriate for the Banking Union? , March 2020.

(73)

The institution’s capital and liquidity ratios are examples of the former, while a description of the institution’s processes for managing credit risk is an example of the latter.

(74)

Pillar 3 requirements have been developed in stages and finalised in December 2018. For more information, please be referred to https://www.bis.org/bcbs/publ/d455.htm .

(75)

With limited exceptions, the CRR currently gives institutions a certain degree of freedom in terms of the detail of the information to be disclosed and a significant degree of freedom in terms of the format of the disclosure. The entry into application of the amendments to the disclosure rules (introduced by the CRR II) will bring about a significant harmonisation of disclosures. In particular, this will be done through implementing technical standards to be developed by the EBA, which will contain detailed templates (and related instructions to fill those templates) that institutions will need to use for their disclosures.

(76)

There is an overlap between the information that must be disclosed and the information that must be reported, although the amount of information that needs to be reported is normally much larger (and more detailed).

(77)

The size of the undue costs cannot be estimated due to lack of data.

(78)

EUCLID is expected to be launched in 2021.

(79)

See, for example, https://op.europa.eu/en/publication-detail/-/publication/4b62e682-4e0f-11ea-aece-01aa75ed71a1 .

(80)

See  https://www.eba.europa.eu/regulation-and-policy/supervisory-benchmarking-exercises

and https://www.bankingsupervision.europa.eu/banking/tasks/internal_models/trim/html/index.en.html  

for EBA and ECB, respectively.

(81)

The relevant Treaty Articles conferring the Union the right to adopt measures are those concerning the freedom of establishment (in particular Article 53 TFEU), the freedom to provide services (Article 59 TFEU), and the approximation of rules which have as their object the establishment and functioning of the internal market (Article 114 TFEU).

(82)

See https://www.europarl.europa.eu/doceo/document/TA-8-2016-0439_EN.pdf for the EP and https://www.consilium.europa.eu/media/22659/st11052en16.pdf for the Council.

(83)

Internal models used to calculate capital requirements for operational risk are not covered by TRIM.

(84)

SFTs are secured funding or lending transactions that imply a temporary exchange of assets with one leg of the transaction serving as a guarantee (collateral). Repurchase agreements (repos) constitute the most important category of SFTs in terms of outstanding amounts and turnover. They are generally motivated by the need to borrow cash. From the borrower's point of view, the transaction consists of selling securities against cash, while agreeing in advance to buy back the securities at a predetermined price. The sold securities serve as collateral for the buyer (provider of cash) in the repo. Securities lending, the second largest category of SFTs, is primarily driven by market demand for specific securities, e.g. for short selling or settlement purposes. In this type of transaction, one counterparty lends securities for a fee against collateral in the form of cash or other securities given by another counterparty.

(85)

Apart from modifying the calibration of the original FRTB standards, the revised version also aligned the implementation date of the standards with the implementation date of the overall framework (the date was originally set to January 2019).

(86)

Note that the final elements of the Basel III reform were not specifically intended to address the problems identified with the risk-based capital requirements of EU banks; the same problems have been identified with their international peers with similar business models and activities. When developing the reform, the BCBS took a rather global approach in addressing those problems, therefore not necessarily taking into account the specificities of the banking sector of each of the BCBS members.

(87)

For more details on the level of application see section “Flexibility in the Basel III standards” below.

(88)

For example, according to one provider of credit ratings the average cost of obtaining a rating for an SME is approximately between EUR 40000 and 50000 for the initial evaluation, and then EUR 30000 to 35000 annually for keeping the rating updated. However, these figures are only indicative, as the cost depends on various factors, such as the complexity of the company, its financial structure, and the size of the debt issue.

(89)

For example, according to one ratings provider some of the other reasons why SMEs do not seek to obtain ratings are the availability of financing from banks (ratings are usually sought when debt is sold in the markets), unmet expectations about the quality of the assigned ratings (ratings turn out worse than the company issuing the debt expects), and reluctance to make financial data available.

(90)

A bank’s exposure to an unrated company is assigned a 100% risk weight under the SA-CR, whereas an exposure to a company with a good credit rating can attract a lower risk weight (20% in case of the highest rating).

(91)

As a result of CRR and CRRII, banks can now apply a 23.81% discount when computing their capital requirements on SME exposures of up to EUR 2.5mn EUR whereby the part of an SME exposure exceeding EUR 2.5mn EUR are subject to a 15 % reduction in capital requirements (so-called SME supporting factor).

(92)

Namely, loans funding long-term important infrastructure or industrial projects.

(93)

Namely, loans funding the acquisition of physical assets such airplanes, ships, satellites, railcars, fleets, etc.

(94)

Namely, financing exchange-traded commodities like crude oil, metals or crops.

(95)

This would avoid undermining incentives for proper risk-mitigation of those transactions, for instance through close monitoring and various forms of collateralisation. The EBA would be mandated to develop criteria to determine what constitutes a high-quality project and to determine the calibration of the applicable RW.

(96)

E.g. investments in equities of corporates with which the bank has a long-term business relationship.

(97)

An institutional protection scheme (IPS) is defined in the CRR as a contractual or statutory liability arrangement which protects its member institutions and in particular ensures that they have liquidity and solvency needed to avoid bankruptcy where necessary.

(98)

Under CRR, a “CIU” means an Undertakings for Collective Investments in Transferable Securities (UCITS) as defined in Article 1(2) of Directive 2009/65/EC of the European Parliament and of the Council or an alternative investment fund (AIF) as defined in point (a) of Article 4(1) of Directive 2011/61/EU of the European Parliament and of the Council.

(99)

As described in Annex XX, these requirements include the restricted permission to capitalise exposures to CIUs under the internal model approaches only if the bank can look through the CIUs’ composition and the conservative calibration of some of the standardised approaches available for those exposures.

(100)

COM(2021) 32 final.

(101)

For specific policy options to deliver on the EGD by addressing ESG risks see next section.

(102)

The exemptions cover derivative transactions with certain non-financial companies, governments, pension funds, but also intragroup transactions within the same banking group.

(103)

These situations arise where banks could not put exchange collateral on derivative transactions with their clients or where they cannot find guarantee provided by third-parties (e.g. via credit derivatives) to protect them from the non-payment of the clients obligations under the derivative transactions.

(104)

That is the calculation of capital required for CVA risks if the transactions were not exempted under CRR.

(105)

As an example of the importance of the EU SFTs market, the total value of the EU repo market (the most widely traded form of SFTs in the EU) amounted to around EUR 8.3 trillion at the end of 2019, according to the ESRB.

(106)

FSB: Strengthening Oversight and Regulation of Shadow Banking , 29 August 2013.

(107)

The implementation of this FSB recommendation has been aligned with the implementation of the Basel III reforms, including the one-year postponement to 1 January 2023.

(108)

Hong Kong and Singapore publically announced the application of the final FRTB standards as a reporting requirement from 1 January 2023 and committed to implement the standards as a capital requirements at a later stage. Other jurisdictions have already publically indicated a delay of the application of the final FRTB standards as capital requirement, as compared to the BCBS recommended implementation date: Q3 2023 for Japan; 1 January 2014 for Canada and Australia. Finally other major jurisdictions (e.g. US, CH) have publically announced their commitment to implement the FRTB framework as capital requirements without more details at this stage.

(109)

The Basel standards for SA-CCR were adopted in 2014.

(110)

See EUR-Lex - 32021R0337 - EN - EUR-Lex (europa.eu)  

(111)

In their implementation of SA-CCR adopted in November 2019, the US authorities introduced a number of adjustments to the Basel standard which would reduce the capital requirements of some derivate transactions with corporates by about 30%.

(112)

More specifically the Basel III standards refer to the Pillar 1 requirements, the capital conservation buffer requirement, the countercyclical capital buffer requirement, as well as the buffer requirements for global systemically-important and, respectively, other systemically-important institutions (G-/O-SIIs) and the total loss-absorbing capacity (TLAC) requirements. However, they do not include a reference to the systemic risk buffer requirement (SRB) and the Pillar 2 requirement (P2R), as these are not implemented, or not implemented in comparable ways, in all member jurisdictions of the Basel Committee.

(113)

Capital requirements can be calculated at the level of individual banks (“individual level”) or at the level of a banking group (“consolidated level”). The Basel standards, including the OF, apply foremost on a consolidated level. The requirements in the CRR are principally applied at both individual and consolidated level.

(114)

Supervisors would hence need to reduce some of the requirements as certain risks would be captured by the OF.

(115)

This means that the distribution key for any additional capital required by the OF would depend on the contribution of each entity to the consolidated floor requirement.

(116)

The inclusion of banks’ own historical losses through the ILM indicator would either increase their capital requirement for operational risk in case the banks suffered large operational risk losses in the past (in this case ILM would be higher than 1) or decrease it if banks did not suffer such losses (in this case ILM would be lower than 1). A supervisory discretion introduced in the Basel III standards allows supervisors to set ILM to 1 for all banks in their jurisdictions, in order to disregard banks’ own historical losses as a driver of the level of their capital requirement for operational risk.

(117)

NB: Large banks and their supervisors have started to prepare for the implementation of the final elements of the Basel III reform soon after their adoption by the Basel Committee in 2017. This frontloading behaviour is also confirmed by the EBA’s Basel III monitoring exercises, which show a steady decline in capital shortfalls at EU banks with respect to the requirements implied by the Basel III reform (see for instance also https://www.spglobal.com/marketintelligence/en/news-insights/latest-news-headlines/dutch-banks-unflustered-by-front-loading-of-capital-rules-54955966 ) . The presentation of the Commission proposal is hence not going to cause any additional or accelerated frontloading but rather slow it down.

(118)

Such evidence would primarily come from the EBA report mandated under Article 501c of the CRR.

(119)

 At present, the work on the taxonomy related to environmental risks is the most advanced one: a first delegated act on sustainable activities for climate change adaptation and mitigation objectives (see https://ec.europa.eu/info/publications/210421-sustainable-finance-communication_en#taxonomy ) was formally adopted on 4 June 2021 for scrutiny by the co-legislators. A second delegated act for the remaining objectives will be published in 2022. On 12 July 2021, the Platform on Sustainable Finance has published two draft reports on the potential extensions of the taxonomy framework to cover i) social objectives and ii) activities that are significantly harmful to environmental sustainability, and those that have no significant impact on it (see https://ec.europa.eu/info/publications/210712-sustainable-finance-platform-draft-reports_en ).

(120)

This option would also include a general obligation for Member States to ensure that breaches of all relevant CRD and CRR provisions are sanctioned

(121)

For instance, competent authorities should assess them prior to their appointment (and not ex post).

(122)

Ranging from e.g. chief financial officers to persons in any other control function.

(123)

The additional breaches would concern only elements of the regulatory regime applicable to supervised banks which are, according to national and European competent authorities, as the most sensitive from a prudential perspective.

(124)

Financial sanctions would be applicable exclusively to the members of the management bodies of institutions and not to the institutions themselves.

(125)

The institution concerned has to pay a daily amount – up to 5% of its average daily turnover – for every day the infringement continues during a maximum period of six months.

(126)

as defined in Article 316 of the CRR

(127)

Please note that this would not require any changes to the CRR.

(128)

99 EU banks representing 75% of all EU bank assets according to the EBA.

(129)

The EBA impact analysis also shows the distribution of banks’ individual total impact in minimum capital requirements as a result of implementing the final Basel III reforms under option 1. It can be noted that a quarter of EU banks in the EBA sample will incur a total impact in minimum capital requirements lower than +2.2%, while another quarter of these EU banks will incur a total impact in minimum capital requirements higher than +20.5%.

(130)

The EBA‘s methodology estimates the impact of the Basel III reforms on EU banks‘ minimum capital requirements, i.e. taking into account the Pillar 1 minimum requirement, the Pillar 2 requirements and the fully loaded combined buffers requirement.

(131)

Banks incurring a capital shortfall as the result of implementing the final Basel III reforms would have to procure collectively that capital amount by the date of application of the reforms to meet the revised minimum requirements introduced by the reforms. The calculation of capital shortfalls in the EBA’s methodology therefore does not take into potential capital management buffers that banks often hold in addition to the minimum required capital amount. A potential management buffer could in theory be used to compensate the capital shortfall incurred by the implementation of the final Basel III reforms.

(132)

The reduced sample included 45 of the 99 banks of the EBA sample used to calculate the estimates of the final Basel III reforms impacts on own fund requirement, accounting for roughly 85% of the total RWAs of these 99 banks (see Annex 5 of the CfA December 2020 report)

(133)

 

(134)

For instance, the calculations do not take into account the significant discretion of a resolution authority to adjust MREL decisions upwards or downwards, which is meant to ensure that MREL remains a bank specific requirement. In addition, any changes introduced from BRRD2 besides the subordination requirements have been excluded from the analysis.

(135)

A bank will be bound by the OF where its total minimum capital requirements calculated by its internal models would be lower than 72.5% of its total minimum capital requirements calculated under the standardised approaches, for the same exposures.

(136)

In its report published in December 2020, the EBA also provided the impacts on the same sample of 99 banks but based on Q2 2018 data which was used in their previous impact analysis. From Q2 2018 to Q4 2019, the total increase in minimum capital requirements decreased by over 5 percentage points (i.e. from +24.1% to +18.5%), while the capital shortfall across these banks has more than halved (from EUR 109.5 bn to EUR 52.2 bn). According to the EBA, this reduction can be mostly explained by the strengthening in the total capital positions of EU banks between the two dates, but also by the reduction in the overall impact of the output floor for few large banks and the revision of the calibration of the CVA risk framework adopted by the BCBS in July 2020.

(137)

Banks can usually build up additional amount of regulatory capital by either retaining more earnings or by issuing new common shares or other forms of regulatory capital. While the former strategy is inherently difficult in a low profitability environment, the later strategy can also be challenging because the future expected profitability of a bank is reflected in its share price. The lower the profitability outlook, the lower the share price and the larger the number of new shares that have to be issued to raise a given amount of capital, making it more difficult to reach a given new issuance target.

(138)

 The impact of option 2 in terms of % change to the total MRC includes the Commission estimates for some of EU specific adjustments proposed in Section 5.2.1 that the EBA has not been able to quantify in its impact study (see Table 4 for the breakdown of these estimates). However, the Commission services did not have the ability to reflect these estimates in the TC shortfall under this option which would require to have access to individual banks data. Therefore, the TC shortfall amounts shown in this table, which only reflect the EU specific adjustments quantified by the EBA, should be interpreted as an upper bound of the actual TC shortfall incurred by banks in the EBA sample.

(139)

Banks without capital shortfalls would also be in a better position to maintain their current level of capital ratio over the next few years under option by building the required amount of capital to do so.

(140)

Minimum Required Capital.

(141)

The breakdown of these individual impacts in 2023 is basically the same, expect the impact of the output floor which reduces to 0% since its lower value of 50% does not bind EU banks.

(142)

In the “EU single stack” approach proposed in this note, supervisors will be able to further adjust the overall than was estimated by the EBA in its impact assessment.

(143)

These impacts estimates are calculated by the EBA under a ‘static balance sheet’ assumption, meaning that they only capture the impacts of a change in the prudential framework at the date they have been calculated. Therefore, the impact estimated presented in this impact assessment implicitly assumes banks‘ balance sheet

(144)

 The first adverse scenario applies a stress effect on the EU banks’ credit risk provisions and their credit risk capital requirements to all their borrowers, based on the hypothetical shock arising from the 2018 stress test; the second adverse scenario cumulative the stress effect of the first scenario with more specific stress effects applied to bank’s exposures from economic sectors that are the most affected by the COVID-19 crisis. This methodology, which has previously been used by the EBA thematic note on the impact of COVID-19 in the EU banking sector published in May 2020 (see https://www.eba.europa.eu/risk-analysis-and-data/risk-assessment-reports ), is described in Annex 6.

(145)

Recent market observations (e.g. ECB’s lending survey or EBA’s risk and vulnerabilities assessment) indicate that lending conditions remained broadly favourable in 2020 but started to tighten. This is mainly driven by banks’ assessment of the risk related to the deteriorating creditworthiness of borrowers affected by the pandemic, whereas banks’ capital position remains strong and did not contribute to the tightening. In the Euro area, banks do not expect that regulatory or supervisory action will constrain their capital positions or lead to a decrease in their total assets.

(146)

Does energy efficiency predict mortgage performance?, Bank of England, 2020 (see https://www.bankofengland.co.uk/working-paper/2020/does-energy-efficiency-predict-mortgage-performance ).

(147)

See the EBA Discussion paper on management and supervision of ESG risks for credit institutions and investment firms.

(148)

Most MS have already provided their supervisors with at least some approval powers.

(149)

Article 91 of Directive 2013/36/EU already provides criteria for assessing the members of the management body.

(150)

The possibility for supervisors to impose periodic penalty payments applicable to credit institutions breaching their regulatory requirements

(151)

It is the case especially if the breach would lead to financial penalties

(152)

In some Member States, it would not be possible to introduce such general administrative sanctioning power because national laws already foresee criminal sanctions for such CRD/CRR breaches

(153)

Under this new regime banks could be hit by sanctions for certain breaches, which would of course represent a cost for those banks. However, those costs are not considered as undue.

(154)

Following the amendments to the disclosure rules introduced by the CRR II, small and non-complex banks are required to disclose quantitative information only.

EUROPEAN COMMISSION

Brussels, 27.10.2021

SWD(2021) 320 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the documents

Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Regulation (EU) No 575/2013 on prudential requirements for credit institutions as regards requirements for credit risk, credit valuation adjustment risk, operational risk, market risk and the output floor

Proposal for a DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Directive 2013/36/EU as regards supervisory powers, sanctions, third-country branches, environmental, social and governance risks, and amending Directive 2014/59/EU

{COM(2021) 663 final} - {SEC(2021) 380 final} - {SWD(2021) 321 final}

Annex 1: Procedural information

1.Lead DG, Decide Planning/CWP references

This legislative proposal (CRR III/CRD VI) was prepared under the lead of DG FISMA in association with DG JUST. Within the Agenda Planning of the European Commission, the project is referred to under item 21. In the Adjusted Commission Work Programme for 2020, the Commission committed under the header “An Economy that Works for People” to review of the CRR and the CRD and adopt a legislative proposal by Q2 2020.

The Decide Planning references are:

·REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Regulation (EU) No 575/2013 (Capital Requirements Regulation - CRR) as regards risk-based own funds requirements

PLAN/2019/5320

·DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Directive 2013/36/EU (Capital Requirements Directive - CRD) as regards risk management and review processes

PLAN/2019/5321

In light of the COVID-19 pandemic, the preparatory work has been delayed, reflecting the BCBS’ decision of 26 March 2020 to postpone the previously agreed implementation deadlines for the final set of Basel III reforms by one year 1 . This delay has allowed the Commission services to focus their attention on the response to the COVID-19 crisis. It has also allowed them to reassess the impact of the planned reforms in light of the potential consequences of the crisis. 

2.Organisation and timing

DG FISMA work on this legislative proposal started in spring 2018 with the publication of the first public consultation requesting inputs from external stakeholders on the implementation of the final elements of the Basel III reform (see Annex 3).

An Inter-services Steering Group assisted DG FISMA in the preparation of this Impact Assessment report. The Steering Group was set up on 11 September 2019 and included colleagues from DG ECFIN, DG GROW, DG JUST, DG COMP, and DG TRADE.

Two additional Steering Group meetings were organised on 2 December 2019 (with colleagues from the same DGs as in the previous meeting) and on 23 January 2020 (with colleagues from the same DGs as in the previous meeting, as well as colleagues from the Legal Service). At each occasion, the members of the Steering Group were given the opportunity to provide comments in writing on the draft versions of the documents presented.

A final Steering Group meeting took place on 12 February 2021 (with colleagues from the same DGs as in the previous meeting, as well as colleagues from DG CLIMA and DG ENV) to discuss the revised text before its submission.

All the meetings were chaired by the Secretariat General.

The contributions of the members of the Steering Group have been taken into account in the content and shape of this impact assessment.

3.Consultation of the RSB 

A first version of the Impact Assessment report was submitted to the Regulatory Scrutiny Board (RSB) on 12 February 2020 and discussed during a physical meeting with the RSB on 4 March 2020. The RSB gave a negative opinion on 6 March 2020, pointing to several shortcomings in the report (see left column of Table 10 for the list).

The Impact Assessment was modified to address the identified shortcomings (see right column of Table 10 how they were addressed) and was re-submitted via written procedure to the RSB on 22 June 2021.

Table 10: Summary of RSB comments in its 4 March 2020 opinion

4.Evidence, sources and quality

A number of inputs and sources of data were used in the preparation of this impact assessment, including the following:

·advices from the EBA, delivered to the Commission in August and December 2019 and December 2020, and other reports of the EBA referred to in footnotes to this impact assessment;

·information supplied in the context of the public consultation described in Annex 2;

·publications of the ECB, the ESRB, the FSB and the BCBS referred to in footnotes to this impact assessment;

·newspaper articles, scientific journal articles, and other sources referred to in footnotes to this impact assessment.

The vast majority of the data underpinning the quantitative analysis contained in this impact assessment was provided by banks. Given that the data used in the EBA, ECB and BCBS analyses underwent quality checks by those organisations, the data quality can be considered reasonably good. Nevertheless, the figures provided by banks are (more or less accurate) estimates based on assumptions made by banks. As such, they need to be interpreted with caution.

Annex 2: Stakeholder consultation

1. Public consultation

As part of the implementation process of the final elements of the Basel III reform in the EU, the Commission services gathered stakeholders’ views on specific topics in the areas of credit risk, operational risk, market risk, CVA risk, securities financing transactions as well as in relation to the output floor.

Beyond these topics related to the Basel III implementation, the Commission services have also consulted on certain other subjects with a view to ensuring convergent and consistent supervisory practices across the Union and alleviating the administrative burden.

The public consultation carried out between October 2019 and early January 2020 2 had been preceded by a first exploratory consultation conducted in spring 2018 3 , seeking first views of a targeted group of stakeholders on the international agreement. The results of the two consultations have fed into the preparation of the legislative initiative accompanying this impact assessment.

Stakeholder groups 

There were 119 responses to the public consultation. As illustrated by Figure 10  and  Figure 11 , most responses came from the financial industry (i.e. individual banks, banking associations and) and half of them came from respondents established in three Member States (Belgium, Germany, France and the United Kingdom).

Results

Stakeholders overall agreed on the necessity to complete the implementation of the Basel III framework in the EU. While the financial industry called for several adjustments and additional transition periods, supervisors and public authorities took a more conservative approach and preferred an implementation of the remaining reforms closer to the international standards.

Figure 10: Nature of respondents to the public consultation

Source: Commission, DG FISMA

Figure 11: Country of origin of respondents to the public consultation

Source: Commission, DG FISMA

The feedback received on the individual elements of the reforms can be summarised as follows:

·Regarding the output floor the financial industry asked to limit the scope to the requirements explicitly listed in the corresponding Basel standard by applying it as a “parallel stack” and at consolidated level only. In contrast, a majority of supervisors were in favour of a “single stack” implementation of the output floor at all levels of consolidation.

·As regards the treatment of unrated corporates in the credit risk framework, supervisors and the financial industry have expressed different views. Industry advocated for a so-called “hybrid approach” while supervisors prefer to implement the Basel standard. Views were divided among supervisors on the option to remove the SME supporting factor from the risk framework, while the financial industry was in favour to keep it.

·On specialised lending, views were mixed between Member States and supervisors. Some showed openness to introduce more granularity in particular with regards to object finance than it is envisaged under the final Basel III standards. This approach was supported by the financial industry.

·Supervisors and the industry also expressed different views on the treatment of equity exposures. Supervisors preferred to apply the Basel treatment without modifications while the industry and some Member States argued for a more granular treatment for long-term equity holdings in unlisted entities, and notably a more favourable treatment for intragroup and IPS exposures.

·There was consensus among Member States and supervisors in favour of the continuation of the “loan splitting approach” (instead of implementing the “whole loan approach”) for real estate exposures. The financial industry preferred to provide flexibility for banks to choose its approach. Industry’s plea to continue allowing for the upward revaluation of property values after origination found some (conditional) support with Member States and supervisors.

·Several stakeholders from industry were in favour of not fully applying the new constraints for the use of internal models for credit risk. This option found only little support from Member States and supervisors.

·The industry favoured a delay in the implementation of minimum haircut floors for securities financing transactions, or, as a second best, an implementation via market regulation. Member States and supervisors were mostly silent on the issue.

·Member States and supervisors were largely in favour of taking historical losses into account to compute large banks’ capital requirements for operational risk, while the majority of stakeholders in the financial industry favour neutralising the impact of past losses or taking them into account only when they have a beneficial impact.

·The consultation showed no opposition against the implementation of the simplified standardised approach for market risk in line with the Basel calibration. Also, the financial industry unanimously supported the need for more flexibility in the treatment of CIUs while Member States and supervisors were mostly silent on the issue.

·Industry and supervisors were in favour of delaying the implementation of the CVA risk rules until the Basel standard is finalised. The industry unanimously favours keeping the existing exemptions. MS and supervisors are split on this issue.

Apart from questions on the implementation of the final elements of the Basel III reform in the EU, the consultation included some questions regarding supervisory reporting and public disclosure as well as the fit-and-proper assessment.

Views are all in all cautiously positive on the initiative to centralise supervisory reporting and public disclosure. A majority of industry players supported the approach but raised some doubts as to the size of the potential reduction in their administrative costs. Supervisors pointed to the need to address concerns about wrong expectations that the EBA would be responsible for the quality of the information disclosed by banks.

The views of Member States and supervisors on potential changes to the fit-and-proper assessment framework were largely correlated with their current practices. Doubts regarding the need for changes were mainly expressed by those Member States that currently do not assess key function holders and/or assess members of the management body after their appointment, while others were more supportive. Similarly, industry representatives in Member States applying only ex-post assessments were particularly concerned about potential difficulties in terms of administrative procedure and burden that could arise under an-ex ante system. Those industry representatives that already had experience with ex-ante assessments of members of the management body and of key function holders and/or with accountability regimes, generally reported a positive impact in terms of reducing risks for the sector and creating a level playing field.

2. Public conference

On 12 November 2019, DG FISMA held a public conference to discuss the impact and challenges of implementing the finalised Basel III standards in the EU.

More than 500 representatives of public authorities (Members of the European Parliament, Member State governments, bank supervisors and international organisations), the financial industry, non-financial companies, think tanks and non-governmental organisations physically attended the conference and 618 additional representatives watched the discussions online on that day.

The conference was comprised of three keynote speeches, delivered by staff of the European Commission, the chair of the EBA and the chair of the Single Supervisory Mechanism, as well as four panel discussions 4 : “Basel III in a global context”, “The impacts of Basel III on the EU economy”, “A proportionate implementation of Basel III” and “Basel III – are we done now?”.

Panel discussions took place on Basel III implementation in a global context, on its impact on the European economy as well as in view of proportionality. Panel discussions included also a regulatory outlook beyond the Basel reforms, chaired by officials of the European Commission, members of the European Parliament (ECON chair) and representatives of the Council (Financial Services Committee chair).

3. Ongoing exchanges with stakeholders

Since 2018 the Commission services have repeatedly consulted Member States on the EU implementation of the final elements of the Basel III reform and other possible revisions of the CRR and the CRD in the context of the Commission Expert Group for Banking, Payment and Insurance (EGBPI).

During the preparatory phase of the legislation, the Commission services have also held hundreds of meetings (physical and virtual) with representatives of the banking industry as well as other stakeholders.

Annex 3: Who is affected and how?

The purpose of this Annex is to set out the practical implications for stakeholders affected by this initiative, mainly banks, governments at the national and European level and the general public (namely companies and consumers acting as borrowers). The initiative aims to achieve the following objectives:

·Objective 1: strengthen the risk-based capital framework, without incurring in a significant increase of capital requirements;

·Objective 2: enhance the focus on Environmental, Social and Governance (“ESG”) risks in the prudential framework;

·Objective 3: further harmonise supervisory powers and tools;

·Objective 4: reduce administrative costs related to public disclosures and improve access to banking prudential data.

In order to “strengthen the risk-based capital framework, without incurring a significant increase of capital requirements” (objective 1), the preferred option would implement the final elements of the Basel III reform in Union law, subject to a set of adjustments. The proposed adjustments are intended to prevent an undue disruption of banks’ lending capacity during the (expected) post-COVID 19 pandemic phase.

The aim of the Basel III reform is to make banks more resilient and restore confidence in the banking system in response to the GCF. A more robust banking system has significant long-term benefits for the economy of the Union as a whole. Better capitalised banks will be more capable of withstanding future financial shocks and continue lending through economic downturns, which is likely to make those future downturns shorter in length and less severe. Hence, achieving the purpose of better capturing risks sought by the Basel III reform will directly benefit banks and, indirectly, all other stakeholders concerned. For instance, bank bail-outs and the recourse to governments to fund them in the event of a crisis can be expected to be less likely. At the same time, a steadier flow of credit may reduce the likelihood of failure for borrowers (namely businesses and households) that rely on bank lending as an essential source of funding.

Objective 1 will be achieved by various amendments to the current prudential framework contained in the CRR):

a)standardised approach for credit risk: changes to the regulatory capital treatment of rated exposures to banks, exposures to corporates, real estate exposures, retail exposures, subordinated debt and equity and off-balance sheet items;

b)internal ratings-based approach (IRBA) for credit risk: the use of internal models is either limited or precluded for certain portfolios that cannot be reliably modelled. Where models may be used, their parameters become subject to certain minimum values (“input floors”);

c)market risk (FRTB): the current reporting requirements based on the FRTB are transformed into substantive capital requirements;

d)credit valuation adjustment (CVA): the use of internal models for CVA is precluded and replaced by a revised standardised approach;

e)operational risk: the use of internal models is precluded and replaced by a single new standardised approach;

f)output floor: capital requirements that result from a bank’s internal models are floored at 72.5% of the requirements that would result from applying the corresponding standardised approaches.

The net effect of these changes in the CRR is an overall increase of regulatory capital requirements on banks. The effect mainly depends on the magnitude of the use of internal models, insofar as banks may be impacted by the output floor and/or become subject to the corresponding (normally more conservative) standardised approach or relevant regulatory input. As estimates show (see Section 6.1), the unfettered implementation of the final elements of the Basel III reform would lead to an average increase of 18.5% in banks’ capital requirements by the end implementation date (January 2028). The implementation would be phased in during a five-year period and the increase of capital requirements at the start of that period (January 2023) would be 11.8% on average. The impact on regulatory capital, at both the end and start dates of the standards’ implementation, would be significant and have considerable potential to reduce bank lending in the short term 5 . This would be particularly undesirable in the context of a post-Covid scenario when lending will be needed to fund the economic recovery.

In order to address this concern, it is suggested to mitigate the proposal’s impact through a set of substantive adjustments to the standards, namely:

a)provisions to cater for the “specificities” of the EU banking sector and its economy: these provisions would adjust the Basel III standards to the  distinctive features of the EU banking sector. In particular they aim at maintaining the flow of lending to EU businesses in general, and SMEs in particular. These include: 

·maintaining certain existing preferential treatments and exemptions for key exposures - the “SME supporting factor” and the “infrastructure supporting factor”, which lower capital requirements for these exposures compared to the corresponding Basel III standards, or the exemption for derivative transactions with certain qualifying parties from the CVA requirements;

·providing a transitional period for the implementation of the OF in relation to lending to unrated corporates – Basel III standards increase capital requirements on lending to unrated borrowers. However, most corporates in the EU (namely SMEs) are currently unrated. During the proposed transitional period and while solutions aimed at increasing the coverage of external ratings are rolled out, lending to unrated corporates by banks using the IRBA would be subject to a more favourable treatment than the one provided for in the standards (i.e. a lower risk weight under the OF);

·maintaining the existing treatment for certain types of equity exposures –the treatment of banks’ strategic holdings of equity issued by entities within the same banking group or covered by the same IPS would be left unchanged, and existing strategic participations in non-financial companies where banks exercise influence would be grandfathered. Hence, these particular equity exposures would remain subject to the capital requirements currently applicable under the CRR and, thus, exempted from the higher capital requirements on equity exposures that will result from implementing the Basel III standards;

·providing an ad hoc preferential treatment for certain exposure types under the new market risk rules - these comprise exposures to collective investment undertakings and financial products based on EU emission trading schemes; and

·postponing the implementation of the FSB’s recommendation of a minimum haircut floor for non-centrally cleared SFTs, awaiting a joint report on the matter by the EBA and ESMA;

b)the application of certain discretions contained in the Basel III standards:

·apply the OF on the consolidated level of a banking group taking into account all the risk-based capital requirements contained in EU law and adjusting the individual bank’s Pillar 2 requirement (“P2R”) and the Systemic Risk Buffer (“SRB”) to avoid double counting of risks. As P2R and SRB may currently be used to address, inter alia, shortcomings of internal models that result in underestimation of risks (“model risk”) and the OF is designed to address this same risk, supervisors would be required to adjust P2R and SRB in as much as necessary to avoid double counting of risks where the OF applies in relation to a bank:

·set the Internal Loss Multiplier (“ILM”) at 1 as part of implementing the new standardised approach for operational risk. The calculation of capital requirements for operational risks of EU banks would, thus, be based on their Business Indicator Component (“BIC”), which takes into account the main elements of a bank’s income and expenses. However, the banks’ historical operational losses would be disregarded for these purposes, which would significantly mitigate the impact of the new approach for calculating the capital requirements for operational risk.

c)safeguards related to banks’ trading activities: in order to preserve the international playing field for EU banks, the Commission would be empowered to delay the entry into force of the capital requirements based on the FRTB and/or to make certain adjustments to the framework. The Commission’s decision should, for instance, consider whether major jurisdictions failed to implement the FRTB. Similarly, it is suggested to temporarily lower the calibration of the current standardised approach for counterparty credit risk (“SA-CCR”) for all derivatives for the purpose of calculating the OF, taking into account international developments in this field (for instance, one major jurisdiction has lowered the calibration of the SA-CCR for certain derivative exposures). This would afford extra time to discuss the calibration of the SA-CCR at international level;

d)a delayed phase-in period for the new rules, relative to the phase-in period envisaged in the Basel III standards (from January 2023 to January 2028), as further explained below.

The adjustments to the Basel III standards referred to in points (a) to (c) would significantly lower the expected increase in banks’ capital requirements that results from implementing the Basel III standards. Capital requirements would go up on average between +0.7% to +2.7% at the start of the phase-in period (in contrast to the 11.8% increase without the adjustments), and between +6.4% and +8.4% at the end of the phase-in period (in contrast to the 18.5% increase without the adjustments). In this modified scenario, bank lending would not be impeded and the prudential benefit of the reform would be preserved.

While the quantitative impact at the beginning of the phase-in period may be moderate per se, the market may exert pressure on EU banks to start building up capital from early on to anticipate future capital requirements. This may coincide with the recovery phase from the COVID-19 crisis and have short term negative effects on bank lending and the wider economy. Accordingly, it is suggested as part of the preferred policy option for Objective 1, that the phase-in period of the new standards’ implementation in the EU would be delayed by two years. The starting date of the phase-in period would be January 2025 (instead of January 2023 under the Basel III scenario) and the new standards would only be fully effective from January 2030 (instead of January 2028). This additional two-year period would give banks enough time to start building up capital without compromising their short-term lending ability. At the same time, the completion of the bank reforms, albeit delayed, would give all market players certainty about the final shape of the regulatory landscape.

Objective 2 of the proposal is to “enhance the focus on ESG risks in the prudential framework”. This objective seeks to address, among others, the emerging risks that climate change and the resulting economic transformations pose to banks, primarily in the form of transition risk. The transition to a sustainable economy may lead to substantial shifts in the value of assets. At the same time, more frequent and /or more severe weather events will present new physical risks.

As explained in Section 6.2, the preferred policy option would introduce in the CRD a general requirement for banks to manage their ESG risks. At the same time, competent authorities would have to supervise compliance with that requirement as part of the supervisory and review assessment process. In addition, the CRR’s requirement to disclose ESG risks related information would be extended beyond large banks. This means that, at this stage, banks would be subject to behavioural obligations in relation to ESG risks (that is, management, governance and disclosure) as opposed to a pre-determined minimum amount of capital to cover unexpected losses arising from those risks within the framework (i.e. no Pillar 1 capital requirements). The introduction of ESG-targeted Pillar 1 capital requirement could be decided at a later stage, following the publication of an EBA Report providing quantitative evidence on the appropriate treatment of ESG risks under Pillar 1.

The above-described policy option presents a number of benefits for all stakeholders concerned without imposing significant costs on banks in the short term. Under the preferred approach, banks would be obliged to adapt their risk management systems and incorporate the ESG dimension before Pillar 1 requirements are introduced. As a consequence, they would be better prepared once Pillar 1 requirements are introduced The prudential framework’s better capturing of ESG risks would also have positive effects for governments and the general public as banks would be less exposed to shocks that may result from transition and physical risks and be more likely to continue lending to the economy under these circumstances. The proposal would also contribute to the general public policies by facilitating a smooth transition towards a more sustainable economy.

By contrast, the costs associated with this policy option would be relatively contained and limited to administrative costs that banks would incur to adapt their risk management, governance and disclosure policies. Such costs should not materially affect banks’ lending capacity.

Objective 3 of the proposal is to “further harmonise supervisory powers and tools of banking competent authorities”. In order to achieve this objective, it is suggested to make various amendments to the CRD to harmonise the supervisory and sanctioning powers of those authorities as follows:

a)introduce ex-ante notification requirements for banks on “material” events with prudential relevance, namely acquisitions of holdings, transfers of assets and liabilities and mergers and demergers;

b)enhance the disciplinary framework whereby the list of breaches for which competent authorities would have explicit powers to impose sanctions would be expanded;

c)introduce harmonised requirements on the assessment of the fitness and propriety of members of banks’ management boards and key function holders before taking up their positions. 

The preferred policy option would ensure a more consistent application of the prudential framework across the EU in general, and within the Banking Union in particular, than currently the case. Enhanced consistency in regulatory and supervisory processes and outcomes would be beneficial for all stakeholders concerned, namely banks that would be able to operate across much more harmonised legal frameworks within the EU. This would reduce the compliance costs that currently arise from the need to deal with diverse and potentially inconsistent requirements. While some banks may initially incur some costs to meet the new requirements (in particular if the national rules they are currently applicable are less stringent than those under the preferred policy option), it is not expected that those costs would be significant enough to offset the preferred policy option’s benefit. It should be noted, in particular, that the most impactful set of amendments, i.e. those referred to in point (a), would be subject to a materiality threshold.

The preferred policy option would also assist competent authorities to discharge their legal duties in a more effective manner than they currently do, as they all would have access to the same full set of supervisory powers (which some competent authorities currently lack) and a more harmonised supervisory framework would mitigate the risk of regulatory arbitrage or the existence of loopholes that banks could potentially exploit. This would, in turn, contribute to fostering the general public’s confidence in the EU system of banking supervision.

Lastly, Objective 4 aims at “reducing administrative costs related to public disclosures and improve access to banking prudential information”. It would be achieved through the creation of a single electronic access to EU banks’ quantitative information that the EBA would source from its EUCLID platform (which will be fed through the existing periodic reporting made by banks), and qualitative information that the EBA would source from banks (see Section 5.2.4).

The preferred option would eliminate disclosure costs for small and non-complex banks, which are only required to disclose quantitative data. For all the other banks, the preferred option would entail neither additional costs nor cost savings.

For other stakeholders, namely market participants who are users of information disclosed by banks, this proposal would bring about material benefits in the form of greater market transparency and lower costs to search for and to access prudential data.

Objective 1 – Strengthen the risk-based capital framework for credit institutions

Preferred Option – Implement Basel III reforms with EU-specific adjustments and transitional arrangements adapted to the COVID-19 crisis

II. Overview of costs – Preferred option
	Citizens/Consumers/non-financial corporates		Banks		Administrations (including competent authorities)
	One-off	Recurrent	One-off	Recurrent	One-off		Recurrent
Direct costs			Costs to adapt banks’ systems to incorporate the changes made to the prudential framework	Increased cost of capital for exposures that would be subject to higher capital requirements compared to the current rules	Costs to adapt current supervisory practices and processes to the new standards	Costs for running the new procedures (depending on magnitude of change compared to current procedures)
Indirect costs		Increase in the costs for bank loans/financial products which are subject to higher capital requirements compared to the current rules (depending on the size of the increase in the capital requirements for the bank loan and the level of competition in the market)

Objective 2 – Incorporate sustainability risks in the prudential framework.

Preferred Option – Introduce measures for a better management of environmental risks by banks

II. Overview of costs – Preferred option
		Citizens/Consumers/non-financial corporates		Banks		Administrations (including competent authorities)
		One-off	Recurrent	One-off	Recurrent	One-off	Recurrent
Reinforced requirements for banks to manage ESG risks	Direct costs			Cost of adjusting risk management systems and processes to the new requirement.
	Indirect costs		Cost of providing additional information to banks.
Reinforced supervisory powers for ESG risks	Direct costs					Cost of setting up new supervisory processes and systems.	Costs of running the new processes and systems.
	Indirect costs
Reinforced disclosure of ESG risks by banks	Direct costs			Changes to systems to accommodate new disclosure templates.	Costs of preparing the new information for disclosure.
	Indirect costs		Cost of providing additional information to banks.

Objective 3 – further harmonise supervisory powers and tools

Preferred Option – harmonise the supervisory powers and tools of banking competent authorities to the greatest possible degree between two available options

II. Overview of costs – Preferred option
	Citizens/Consumers/non-financial corporates			Banks		Administrations (including competent authorities)
	One-off		Recurrent	One-off	Recurrent	One-off	Recurrent
Direct costs				Administrative costs to adjust internal processes to meet new requirements.	Administrative costs to comply with new ex ante notification and assessment requirements. Scope limited to “material” events for ex ante notifications.	Costs to change current supervisory procedures or to set up new procedures to meet the new requirements.	Costs to deal on an on-going basis with new ex ante notification and assessment requirements.
Indirect costs

Objective 4 – Reduce administrative costs related to public disclosures and improve access to banking prudential data

Preferred Option – centralise the disclosure of both quantitative and qualitative prudential banking disclosures

II. Overview of costs – Preferred option
		Citizens/Consumers/non-financial corporates		Credit institutions		Administrations (including competent authorities)
		One-off	Recurrent	One-off	Recurrent	One-off	Recurrent
	Direct costs					EBA to incur costs to build up the systems, processes and get the necessary resources to provide the centralised disclosures.	There will be increased on-going costs for the EBA to maintain and operate the disclosure platform.

Annex 4: Evaluation

As described in Section 1, this prudential framework applicable to banks in the EU has been significantly revised through two waves of reforms to address a number of issues observed following the GFC: the first wave of reforms, introduced by the CRR and the CRD IV, was adopted in June 2013 and the second wave, introduced by CRR II and CRD V, was adopted in May 2019.

This new legislative initiative aims to complement and finalise the above reforms with the implementation of the final elements of the Basel III reform adopted by the Basel Committee in December 2017. The initiative addresses the outstanding deficiencies in the prudential framework that have not been addressed by the previous rounds of reforms.

The vast majority of the changes proposed in this legislative initiative stem from changes to the international standards developed by the BCBS. The latter changes were adopted to address the deficiencies that the BCBS identified when it carried out an evaluation of the existing international standards. A number of supervisory initiatives performed at EU level, notably the EBA benchmarking exercises (see Section 5.1), the EBA work on IRB repair and the ECB TRIM exercise (see Section 5.1), confirmed the findings of the BCBS.

Given the reliability of the abovementioned evaluations, the Commission decided to rely on their findings instead of carrying out its own evaluation.

(1)

See https://www.bis.org/press/p200327.htm

(2)

See https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12015-Alignment-EU-rules-on-capital-requirements-to-international-standards-prudential-requirements-and-market-discipline-/public-consultation_en .

(3)

See https://ec.europa.eu/info/consultations/finance-2018-basel-3-finalisation_en

(4)

Faced with higher capital requirements, banks may choose to raise new capital to increase their ratios and/or reduce their exposures (i.e. reduce lending) to meet the new requirements.

EUROPEAN COMMISSION

Brussels, 27.10.2021

SWD(2021) 320 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the documents

Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Regulation (EU) No 575/2013 on prudential requirements for credit institutions as regards requirements for credit risk, credit valuation adjustment risk, operational risk, market risk and the output floor

Proposal for a DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL amending Directive 2013/36/EU as regards supervisory powers, sanctions, third-country branches, environmental, social and governance risks, and amending Directive 2014/59/EU

{COM(2021) 663 final} - {SEC(2021) 380 final} - {SWD(2021) 321 final}

Annex 5: Additional information about policy options

1.Improve the current framework for calculating risk-based capital requirements

1.1.Credit risk framework – Standardised approach

General background

Credit risk is the risk of loss resulting from a borrower's failure to repay a loan or meet contractual obligations on a debt (“default”). It accounts for the bulk of most banks’ risk-taking activities and hence the bulk of their capital requirements.

The standardised approach for credit risk (SA-CR) is used by the majority of banks across the EU to calculate the capital requirements for the majority of their credit risk exposures, even though the majority of total EU banks’ credit risk exposures – which are held by a relatively low number of large institutions – are treated under internal model approaches. The SA-CR is thus highly important in its own right. In addition, the SA-CR must serve as a credible alternative to internal model approaches and as effective backstop to them. The SA-CR’s importance is further increased under the final elements of Basel III reform, due to (i) the reduction of the scope of application of internal models approaches and (ii) the introduction of the OF.

During its review of the functioning of standardised approaches, the BCBS found the current SA-CR to be insufficiently risk-sensitive in a number of areas, leading sometimes to inaccurate or inappropriate measurement of credit risk (either too high or too low) and hence of capital requirements. 1 These problems put into question the SA-CR’s role as approach for measuring credit risk, including as an alternative for and a backstop to internal model approaches. Existing supervisory or macro-prudential tools could be used to partly address these problems, but neither are well-suited for the purpose.

One of the final elements of the Basel III reform therefore aimed to increase the risk sensitivity of this approach. To achieve this, the Basel Committee agreed to change the SA-CR in relation to several key aspects. These changes, if implemented in the EU, could be particularly impactful in the following key areas (due to e.g. the amount of exposures potentially affected, the nature of the changes or EU specificities):

·exposures to unrated corporates, including SMEs;

·exposures to project finance, object finance and commodities finance (specialised lending);

·equity exposures; and

·exposures secured by real estate (both residential and commercial).

As outlined in Section 6.1., the preferred option is a revision of the SA-CR in line with the Basel III standard, subject to some adjustments. The subsections below discuss each of the aforementioned key areas and assess which adjustments are necessary in each case.

1.1.1.Unrated Corporates

Problem definition

Given the aforementioned shortcomings of the existing SA-CR, the new Basel III standard sets out a modified treatment of corporate exposures under the SA-CR. These modifications make the framework more risk-sensitive and take better account of the specificities of exposures to SMEs and investment grade corporates.

The Basel III standards contain two different approaches to calculate RWs of corporate exposures under the SA-CR. The first can be used in jurisdictions that allow the use of external ratings (the external credit risk assessment approach, or “ECRA”), while the alternative one can be used in jurisdictions that do not allow the use of external ratings (the standardised credit risk assessment approach, or “SCRA”). The ECRA reflects credit risk in banks’ capital requirements better than the SCRA as it allows for a more granular set of RWs.

Under the ECRA, corporate exposures are assigned a RW between 20% and 150% depending on their external rating 2 . However, exposures to corporates that do not have such external ratings (“unrated corporates”) receive a RW of 100%, with the exception of corporate SMEs (RW of 85%) and retail SMEs (RW of 75%). 3 The ECRA is consistent with the way in which standardised RWs are currently assigned in the EU; its improved risk-sensitivity would lead to a slight decrease in RWAs for highly-rated corporates whilst the impact on the majority of SA banks’ exposures would be marginal.

However, an unintended indirect consequence of the application of the ECRA may arise in the EU as a result of the introduction of the OF combined with the fact that the vast majority of EU corporates is not externally rated. Those corporates are currently being predominantly financed by IRBA banks, and the internal models of those banks produce significantly lower RWs for exposures to many of those corporates compared to the 100% RW applicable under ECRA (and the current version of the SA-CR). Once those banks would be required to apply the ECRA for the purposes of calculating the OF, there would be a material increase in capital requirements for those exposures.

In this context, it should be noted that the CRR provides for a preferential treatment for exposures to SMEs: the RW of an exposure to a SME is multiplied by the so-called “SME supporting factor” of 0.7619 for exposures up to EUR 2.5 million. For the remaining part of an exposure exceeding that threshold, a RW of 85% applies, in accordance with the Basel III standard. The SME supporting factor is applicable to all exposures to SMEs (whether rated or unrated, and whether treated under the SA or the IRBA). Concerning retail SMEs, the RW of 75% is already implemented and applicable in the EU. Consequently, all policy considerations relating to the treatment of unrated corporates must account for the effect of the SME supporting factor.

Policy options

Baseline option - no change to the prudential framework

·The existing RWs for corporates remain unchanged, including for unrated corporates.

·The existing SME supporting factor is kept.

·The OF is not implemented for unrated corporates.

Option 1- full alignment with Basel III

·The ECRA approach applies to all corporate exposures (including unrated corporates), also for the purposes of the output floor.

·The existing SME supporting factor is removed.

Option 2 - implementation of ECRA in line with Basel III with a transitional treatment for IRBA banks in relation to unrated corporates

·For unrated corporates, a transitional period lasting until 2030 is provided. During this period, when calculating their OF, IRBA banks are allowed to apply the SCRA to exposures to unrated corporates which have a PD corresponding to an investment grade rating. This treatment applies to all unrated corporates, irrespective of whether they are listed or not.

·For rated corporates, when calculating the OF, IRBA banks are allowed to use the ECRA to calculate capital requirements.

·The Commission is empowered to extend the transitional treatment by up to three years based on a report by the EBA.

·At the end of in the transitional period, IRBA apply the ECRA to all corporate exposures when calculating the OF.

·The existing SME supporting factor is kept.

Impacts and comparison across options

Under the baseline option, the shortcomings in terms of risk-capture of the SA-CR for corporates would remain unaddressed and the framework would continue to lack a backstop (in the form of an output floor) limiting excessive variability in capital requirements. At the same time, the level of capital requirements for the exposures concerned would remain largely unchanged.

Option 1 would be fully compliant with the Basel III standards and consistent with the EBA's advice 4 . However, the removal of the beneficial treatment of SMEs that EU co-legislators had agreed upon (in the CRR) and recently extended (in CRR II) could lead to a tightening of financing conditions for SMEs. Indeed, this option would lead to a material increase in minimum required capital for both SA banks (due to the removal of the SME supporting factor) and also IRBA banks (due to the direct effect of the removal of the SME supporting factor and, indirectly, as a result of the application of the output floor). This would, in turn, lead to either a reduction in the amount of loans banks would be willing to provide to SMEs (to the extent banks could not pass the increased cost of capital on SMEs via higher interest rates) or to an increase in the interest rates banks would charge SMEs for the loans (to the extent banks could pass on the increased cost of capital) or a combination of both.

Option 2 would preserve the use of external ratings for corporates and increase the risk-sensitivity of their treatment under the SA-CR. This option would be in line with the final Basel III standard, except when it comes to the EU-specific treatment of SMEs. The transitional adjustment for IRBA banks’ exposures to unrated corporates would limit an increase in capital requirements as compared to option 1. It would avoid disruptive impacts on bank lending and leave sufficient time to establish public or private initiatives aimed at increasing the coverage of external ratings for corporates. Broadening the coverage of external ratings would also increase transparency in the EU corporate sector in the long run and thereby foster the Capital Markets Union (CMU).

Table 11. Comparison of policy options against effectiveness, efficiency and coherence criteria

Magnitude of impact as compared with the Baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

Preferred option

In light of the above analysis, which led to the overall score of each policy option in Table 11  , option 2 is the preferred policy option.

1.1.2. Specialised lending

Problem definition

There is no specific treatment for specialised lending (SL) exposures under the current SA-CR. Those exposures are treated as any corporate exposures even though they have a very specific risk profile. The latter depends more on the type of transaction and/or collateral provided to secure the transaction rather than on the creditworthiness of the borrower(s) since the borrower is typically a dedicated vehicle, which secures the debt with the cash flows generated by its assets. The lack of granularity in the current framework leads to under- or overestimation of risks under the SA-CR and does not allow to appropriately compare related capital requirements between SA and IRBA banks (the IRBA contains a specific treatment for SL exposures). The SA-CR can hence not act as a credible benchmark to internal models used for SL exposures.

To address these shortcomings of the SA-CR, the Basel III standard introduced a specific treatment for SL exposures, distinguishing between project finance, object finance and commodities finance based on the definitions of these three subcategories in the IRBA. Like for corporates exposures (see Section 1.1.1 of this Annex), two approaches are available in the Basel III SA-CR, one for jurisdictions allowing the use of external ratings for regulatory purposes and one for jurisdictions that do not allow it. The new treatment reflects more appropriately and accurately the risks associated with SL exposures and improves the consistency with the already existing treatment of those exposures under the IRBA. Under the new SA-CR, SL exposures for which no issue-specific external ratings are available would be assigned RWs ranging from 80% to 130%. The exact RW assigned to the exposure would depend on the relevant SL subcategory and, in the case of project finance, on the phase in which the project is (pre-operational/operational) as well as on whether certain quality criteria are met. 5  

However, similar to the case of exposures to unrated corporates, unintended consequences of the application of this new approach may arise in the EU, for essentially the same reasons: most SL exposures are not externally rated and are financed by IRBA banks that have in place internal models which produce materially lower RWs than those provided by the SA-CR. While the new standardised treatment for unrated SL exposures is more granular, it is not sufficiently risk-sensitive to reflect the effects of comprehensive security packages usually associated with these exposures (these packages comprise covenants and collateral subject to dedicated monitoring). The impact may be particularly felt in the case of object finance exposures in the context of the application of the OF. As a consequence, there could be a risk of discontinuation of these activities.

The impact on project finance is likely to be less significant because the EU has introduced a discount (= supporting) factor of 25% for exposures to high quality infrastructure projects that comply with a set of eligibility criteria capable to lower their risk profile and enhance the predictability of their cash flows. This treatment applies to both institutions using the SA and institutions using the IRBA.

Policy options

Baseline option: no change to the prudential framework

·No dedicated treatment for SL exposures would be available under the SA-CR.

·The OF is not implemented for SL exposures.

·The existing supporting factor for high quality infrastructure projects is kept.

Option 1 - full alignment with Basel III

·The new SL subcategories are introduced in the SA-CR framework while maintaining the use of issue-specific external credit ratings when available.

·The existing supporting factor for high quality infrastructure projects is removed.

·The new standardised RWs for SL exposures are used for the calculation of the OF.

Option 2 - alignment with Basel III with adjustments for project and object finance

·The new SL subcategories are introduced in the SA-CR framework while maintaining the use of issue-specific external credit ratings when available.

·The specific supporting factor for infrastructure projects remains applicable to complement the preferential treatment for high quality project finance which is limited to projects in the operational phasewhilst avoiding “double discounts” 6 . 

·A new subcategory for high quality object finance is introduced. The EBA is mandated to specify the relevant criteria via RTS.

·The new standardised RWs for SL exposures are used for the calculation of the output floor.

Impacts and comparison across options

Only a small portion of SL exposures in EU banks portfolios is rated (3.62%; see below Table 12 ). Among all SL exposures, non-rated project finance is the most important sub-class (81.5%), followed by non-rated object finance (9.3%) and non-rated commodities finance (5.6%).

Table 12. SL exposures under the SA-CR by sub-exposure class

Source: EBA, CfA response

Based on the EBA's analysis, the impact on exposures (classified as corporate exposures under the current SA-CR), which would be classified as SL exposures under the revised SA-CR, appears to be limited, due to the limited volume of specialised transactions under the SA-CR (14% of all transactions; see Figure 12 below).

As mentioned above, most SL exposures are in the portfolios of IRBA banks. Therefore, the new standardised RWs for SL would also indirectly impact those banks as a consequence of the introduction of the OF. 7  

Figure 12. Share of SL exposures RWA by regulatory approach in the current and revised framework

Sources: EBA 2018-Q2 QIS data and EBA calculations. Note: Based on a sample of 204 banks: IRB (78), SA (196).

Under the baseline option, SL would not be defined in the SA-CR. Consequently, a risk-sensitive and granular approach could not be used for those exposures. Instead, a flat RW of 100% would continue to apply to all forms of SL unless an issue-specific rating would be available.

Option 1 would allow to better reflect the specific risk-profile of SL under the SA-CR. However, the removal of the existing EU-supporting factor for high-quality infrastructure projects under this option would lower the incentives recently put in place in the legislative framework to foster private and public investments in high quality infrastructure projects with low risk profile 8 . In addition, some forms of SL (in particular object finance) would likely be materially impacted by the new RW calibrations, even though indirectly, i.e. through the application of the output floor.

Under option 2, the existing supporting factor for high quality infrastructure finance would be kept. At the same time, the lack of risk-sensitivity of the Basel treatment for unrated object finance exposures (RW of 100%) would be addressed by a specific treatment for “high-quality” object finance exposures and a revised calibration for those exposures, aligning the risk sensitivity for those categories with that of project finance. The indirect potential impact (as an effect of the output floor) for IRBA banks that have developed a SL activity would also be less material under this option than under option 1.

Table 13. Comparison of policy options against effectiveness, efficiency and coherence criteria

Magnitude of impact as compared with the Baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

Preferred option

In light of the above analysis, which led to the overall score of each policy option in  Table 13 , option 2 is deemed the preferred policy option

1.1.3.Equity exposures

Problem definition

The current (Basel II) treatment of equity exposures under the SA-CR is not risk-sensitive: In particular, it does not reflect the higher loss risk of equity compared to senior exposures nor does it differentiate between strategic and speculative (and hence riskier) investments. Furthermore, the different methods for calculating RWs for equities under the IRBA (simple risk weight method, internal model method, PD/LGD approach) have been found to be unduly complex, leading to different outcomes and to undue RWA variability. Finally, the current treatment allowed for regulatory arbitrage between the banking book and the trading book. 9

To address these shortcomings, the final elements of the Basel III reform amend the treatment of equity exposures in two ways. First, banks can no longer use the IRBA for those exposures. Second, the calibration of RWs under the SA-CR is more granular and more conservative: the default RW for equity exposures increases from 100% to 250%. A more conservative RW of 400% is assigned to “speculative unlisted equity exposures” (most of those exposures are currently labelled as “high-risk” equity exposures and hence subject to a 150% RW). Those revisions can be phased-in over a five-year period. The existing 100% RW remains available only for equity investments made pursuant to “national legislated programmes” which meet certain eligibility conditions and have been approved by the competent authority.

Table 14: Risk weights applicable to equity exposures during the phased-in implementation of the Basel standards

Source: EBA, CfA response (Table 11, p. 59)

The final Basel III standard does not differentiate between the riskiness of certain types of equity holdings existing in the EU, such as equity investments in entities that are included in the same banking group (intragroup) or the same institutional protection schemes (IPS), or long-term and strategic equity investments in businesses outside the banking group (including holdings in insurance undertakings).

Policy options

Baseline option - no change to the prudential framework

·The existing RWs (100% and 150%) for equity exposures under the SA-CR are kept.

·Banks may use the IRBA for equity exposures.

Option 1 - full alignment with Basel III

·The use of IRBA for equities is removed; in turn, the SA-CR becomes the only approach available for calculating capital requirements for the credit risk of equity exposures.

·For equity holdings made pursuant to national legislated programmes a preferential RW of 100% may be applied, subject to certain eligibility criteria and supervisory approval.

·The new RWs for speculative unlisted equity exposures (400%) and for other equity holdings (250%) are progressively phased-in during a 3-year period (see  Table 14 above).

Option 2 - implement Basel III with targeted clarifications and adjustments

·The current 100% RW remains applicable to intragroup equity exposures and equity holdings within institutional protection schemes (IPS).

·Existing long-term and strategic equity exposures to counterparties outside the banking sector (i.e. including in insurance undertakings) are subject to a grandfathering regime (i.e. banks are allowed to apply the existing RWs to those exposures).

·Only short-term equity investments with a holding period of less than 5 years are considered as speculative exposures and assigned a steady state RW of 400%.

Impacts and comparison across options

The baseline option would leave the identified problems largely unaddressed: Overly complex and discretionary methods under the IRBA would result in undue variability of RWAs whilst the SA-CR would provide insufficient risk coverage. As a consequence, banks would still have incentives to move equities from the trading to the banking book

The tightened treatment of equities under Option 1 has been identified by the EBA as a major impact driver for increased capital requirements in the area of credit risk (it represents 2.8 pp of the overall increase in RWAs; see Table 15 below) even without taking into account the increase in RWs for intra-group equity exposures 10 . This would be partly compensated by a decrease in RWAs for some of the equity exposures migrating from the IRBA to the SA-CR (as internal modelling of equities’ credit risk would not be allowed anymore; see Table 18 below).

Table 15: Percentage change in equity RWA (relative to total current SA RWA) for equity exposures currently under the SA-CR, by equity category

Source: EBA, CfA response

Table 16: Percentage change in equity SA RWA (relative to total current SA RWA) for equity exposures currently under the IRB, per equity category

Source: EBA, CfA response

The additional data collection at individual and sub-consolidated level performed by the EBA 11 shows the importance of the proportion of intragroup equity exposures at those levels and points to an even higher RWA impact on this subset of exposures.

Table 17: Percentage of exposures to equity and exposures to equity intragroup (over total exposure), by approach

Source: EBA response (5 March 2020) - Letter on additional analysis for the Call for Advice for the purposes of revising the own fund requirements for credit, operational, market and credit valuation adjustment risk: output floor and equity exposure class (Table 4, p. 14)

Table 18: Percentage change in equity RWA (relative to current equity RWA by approach), by equity sub-type

Source: EBA, Letter on additional analysis for the Call for Advice for the purposes of revising the own fund requirements for credit, operational, market and credit valuation adjustment risk: output floor and equity exposure class, March 2020 

Table 19: Percentage change in equity RWA (relative to total current RWA), by approach

Source: EBA, Letter on additional analysis for the Call for Advice for the purposes of revising the own fund requirements for credit, operational, market and credit valuation adjustment risk: output floor and equity exposure class, March 2020

Applying the increased RWs to all equity exposures at all levels (i.e. individual, sub-consolidated and consolidated level) would have a significantly higher impact than the application at consolidated application only. This would be mainly driven by intra-group equity exposures.

This may lead to unintended consequences (such as divestments) for existing structures and business models if such increased RWs would be applied to intragroup and IPS equity exposures, or to strategic investments outside the banking group (i.e. including holdings in insurance undertakings).

The targeted adjustments envisaged under Option 2 would prevent those unintended effects whilst addressing the shortcomings of the current treatment.

Table 20: Comparison of policy options against effectiveness, efficiency and coherence criteria

	EFFECTIVENESS	EFFICIENCY (cost-effectiveness)	Coherence	OVERALL SCORE
Baseline option	0	0	0	0
Option 1	+	-	+	+
Option 2	++	≈	++	++

Magnitude of impact as compared with the Baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

Preferred option

In light of the above analysis, which led to the overall score of each policy option in  Table 20 , option 2 is deemed the preferred policy option.

1.1.4.Real estate exposures

Problem definition

The GFC revealed a number of shortcomings of the current standardised treatment of real estate exposures which the Basel III SA-CR addresses.

Income-producing real estate exposures

Evidence showed that mortgage loans the repayment of which is materially dependent on the cash flows generated by the property securing those loans tend to be materially riskier than mortgage loans the repayment of which are materially dependent on the underlying capacity of the borrower to service the loan. However, under the current SA-CR no specific treatment is foreseen for such exposures, even though this dependence is an important risk driver. This may result in insufficient levels of capital to cover unexpected losses on this type of real estate exposures.

In order to address this shortcoming, the final Basel III standards introduced a new category of real estate exposures, namely income producing real estate (IPRE) 12 exposures, with a dedicated RW. This modification is intended not only to reflect more accurately the risk associated with those exposures, but also to improve consistency with the treatment of IPRE under the IRBA.

Land acquisition, development and construction exposures

Loans financing land acquisition, development or construction (ADC) of any properties incur a heightened risk where the source of repayment at origination of the loan is either a planned but uncertain sale of the property or substantially uncertain cash flows (e.g. this may happen if the property has not yet been leased to the occupancy rate prevailing in that geographic market for that type of property). The current SA-CR applicable in the EU provides for a flat RW of 150% for so-called “speculative immovable property financing”. The latter category is defined solely based on the borrower’s intention to resell the property for a profit, without taking into account to which extent the repayment is actually uncertain. As a result of this lack of clarity and risk-sensitivity, capital requirements for ADC exposures are currently often deemed to be too high or too low..

To better reflect the risk of ADC financing models, the final Basel III standards introduce a dedicated sub-exposure class, referring to loans to companies or SPVs financing any of the land acquisition for development and construction purposes, or development and construction of any residential or commercial property. ADC exposures are to be risk-weighted at 150%. However, provided that certain risk-mitigating conditions 13 are met, ADC exposures to residential real estate may be risk-weighted at 100%.

Whole-loan vs. loan-splitting approach

The final Basel III standards also provide two new, more granular and risk-sensitive approaches for general 14 residential and general commercial real estate exposures (GRRE and GCRE, respectively): (i) the loan splitting approach, which splits mortgage loans into a secured and an unsecured part and assigns a different risk weight to each of these two parts; and (ii) the whole loan approach, which considers mortgage loans as specific products and assigns a RW to the entire exposure based on its loan-to-value (LTV) ratio using different LTV buckets. According to the final Basel III standards, jurisdictions can apply either of these two approaches, but not both.

The rationale for using the LTV ratio as a risk driver for determining the applicable RWs is that the losses incurred in the event of a default and the likelihood of a borrower’s default are lower when the outstanding loan amount relative to the value of the real estate collateral (i.e. the LTV ratio) is lower. However, only the loan splitting approach is also sensitive to the type of borrower (as it applies the RW of the counterparty to the unsecured part) and reflects the risk mitigating effects of the real estate collateral in the applicable RWs even in case of high LTV ratios. While the loan splitting approach is currently in place in the EU, its RW calibration has been found too conservative for mortgages with very low LTV ratios (see  Figure 13 ).

Figure 13: Stylised illustration – RW function for residential real estate exposures under different approaches

Source: Commission

Prudent valuation of property

In the aftermath of the GFC, property prices dropped significantly and progressively in countries in- and outside of the EU. Those bank that had not revised downwards the value of the properties securing mortgage loans on a timely basis, to reflect their reduced market value, ended up underestimating the underlying credit risk of those loans and hence the corresponding capital requirement. Had they updated them, the LTV ratio of those loans would have increased thereby potentially increasing capital requirements).

To reduce the impact of cyclical effects on the valuation of property securing a loan and to keep capital requirements for mortgages more stable, the final Basel III standards cap the value of the property recognised for prudential purposes at the value measured at loan origination, unless modifications “unequivocally” increase the value of the property. At the same time, the standards do not oblige banks to monitor the development of property values. Instead, they only require adjustments in case of extraordinary events. By contrast, the current SA-CR applicable in the EU requires banks to regularly monitor the value of property pledged as collateral. Based on this monitoring, banks are require to make upwards or downwards adjustments to the property (irrespective of the property value at loan origination). The current SA-CR does not include a mechanism addressing cyclical effects in real estate, and hence it does not address the risk of overvaluation and volatile capital requirements for mortgages.

Policy options

Baseline option: no change to the prudential framework

·The current loan splitting approach is kept unchanged.

·No specific treatment for IPRE and ADC exposures is provided for.

·The current rules on property value monitoring and adjustment are kept.

Option 1: full alignment with Basel III

·The loan splitting approach is kept for GRRE and GCRE exposures, but with the revised calibration as set by the final Basel III standards.

·Specific treatments for IPRE and ADC exposures are introduced.

·Property values are capped at their value at loan origination and the current requirement for frequent monitoring is removed.

Option 2: alignment with Basel III with adjustment for property valuation

·Specific treatments for IPRE and ADC exposures are introduced.

·The loan splitting approach is kept for GRRE and GCRE exposures, but with the revised calibration as set by the final Basel III standards.

·The current requirement for frequent monitoring of property values is kept, allowing upwards adjustment beyond the value at loan origination but with a cap set at the average value over the last 5 years.

Impacts and comparison across options

GRRE and GCRE account for the vast majority (more than 90%) of total exposures secured by real estate, whereas all the remaining categories, i.e. IPRE and ADC exposures, together amount to less than 10% (see  Figure 14 ).

Figure 14: Exposure value breakdown as a percentage of total SA real estate exposure under the revised Basel III framework

Source: EBA – Basel III reforms: impact study and key recommendations (Figure 34, p. 88)

Under the baseline option, the lack of risk-sensitivity of the current framework in particular concerning IPRE and ADC exposures would remain, leading to inadequate levels of capital for covering unexpected losses on real estate exposures and potentially providing inappropriate incentives for banks’ lending decisions. Furthermore, the (pro-) cyclicality of property valuations and, by consequence, of capital requirements would remain under the baseline option.

As regards Option 1, while the share of exposures in each category would change slightly as a result of the implementation of the revised SA-CR (more precisely because of the new definitions under Option 1; see Figure 15 below), the RWs for exposures falling in the GRRE and GCRE categories would remain largely unchanged, as the revisions in these areas are marginal compared to the framework currently applicable in the EU.

For similar reasons, the impact of the new treatment for ADC exposures under Option 1 is expected to be limited. The overall impact is marginal as ADC exposures account for only 2.9% of total real estate exposures treated under the SA-CR.

By contrast, the application of the revised standardised treatment for IPRE would increase the RWAs for IPRRE by 47.6% under Option 1. This would be the most significant impact across all exposures secured by real estate. However, given the low share of IPRE in total exposures, the overall impact would be limited (IPRRE accounts for 2.5% of total real estate exposure, while IPCRE accounts for 1.1%).

Figure 15: Percentage change in exposures secured by real estate SA RWA, relative to current sub-exposure class SA RWA (left), and relative to total current SA RWA (right)

Source: EBA – Basel III reforms: impact study and key recommendations (Figure 35 and 36, p. 88)

The (re-)valuation approach provided by Basel III (option 1) is less pro-cyclical, but also less risk-sensitive than the current approach applicable in the EU. Under this option, banks would not be required to adjust property values downwards based on continuous monitoring. They would only need to make adjustments following a supervisor’s intervention or due to extraordinary events. This option would not fit European real estate markets, where mortgage loans usually have long(er) maturities (than in other Basel jurisdictions) 15 . Moreover, it would create perverse incentives for circumvention of the rules by repeated renewal of mortgage contracts, as highlighted in the EBA’s advice. It would also put the burden to require downwards adjustments of property values on supervisors and could also have a negative effect on banks’ risk management, as it would remove the requirement to frequently monitor property values.

Option 2 would have the same impacts as Option 1 on GRRE, GCRE, IPRRE and ADC exposures. By contrast, Option 2 would address potential pro-cyclical effects under option 1 by limiting any upward adjustments at the average value over a certain period thereby reducing the volatility in property values. This would address the disadvantages of the current approach while avoiding the problems inherent in the approach contained in Option 1.

Table 21: Comparison of policy options against effectiveness, efficiency and coherence criteria

Magnitude of impact as compared with the Baseline option (the baseline is indicated as 0): ++ strongly positive; + positive; – – strongly negative; – negative; ≈ marginal/neutral; ? uncertain; n.a. not applicable

Preferred option

In light of the above analysis, which led to the overall score of each policy option in  Table 21 , option 2 is deemed the preferred policy option

1.2.Credit risk framework – Internal model approach 

General background

As an alternative to using the SA-CR, banks may also use one of the two approaches based on internal models, the so-called internal ratings-based (IRB) approaches, to calculate capital requirements for credit risk (subject to supervisory approval). The advanced IRB (AIRB) approach allows modelling of all key risk parameters, whereas the foundation IRB (FIRB) approach is somewhat more restrictive as modelling is allowed for only one out of the three key risk parameters 16 . 

The IRB approaches are complex and are therefore used primarily by large, sophisticated institutions. However, given the large market shares of these institutions, the capital requirements for the majority of credit risk exposures in the EU banking sector are calculated using the IRB approaches (see Figure 16 below).

Figure 16: Exposure value: SA versus IRB composition (% of total credit risk exposure value)

Source: EBA 2018-Q2 QIS data and EBA calculations. Note: Based on a sample of 189 banks.

The GFC highlighted important deficiencies of the IRB approaches. A range of studies conducted at both international and EU level found an unacceptably wide variation in capital requirements across banks that cannot be explained solely by differences in the riskiness of banks’ portfolios. This is undesirable, as it hinders the comparability of capital ratios and impacts the level playing field among banks. Also, the crisis has revealed instances where the losses incurred by banks on some portfolios were significantly higher than the model predictions, which resulted in insufficient levels of capital held by individual banks.

The December 2017 Basel III agreement aimed to address these deficiencies primarily by limiting banks’ flexibility in calculating their capital requirements for credit risk 17 :

·the possibility to use internal models was either limited or altogether removed for portfolios and risk parameters where the BCBS had concluded that the available data was insufficient to ensure reliable modelling (i.e. exposures to financial institutions and exposures to large corporates fall in the first category, while equity exposures fall in the second one);