Basel 4 and the reduction in RWA variability: capital buffers and capital erosion

9 gennaio 2020

Massimiliano Coluccia, Marco Franchini, Elisa Quinto

A reduction in the variability of RWAs is an objective of all regulators, but a lot still needs to be done

White paper

In recent years, a key objective of the main regulatory revisions to capital ratios has been reducing the excessive variability of risk-weighted assets (RWAs) among different banks and jurisdictions. After the introduction of Basel 3, which included a narrower definition of eligible capital instruments, the European Banking Authority (EBA) and other regulators focused on the RWAs of banks authorized to use internal approaches for the calculation of capital requirements. A number of studies, done by the regulators themselves as well as academics and the private sector, pointed to a worryingly large variation in banks’ estimated RWAs; benchmark reports showed how differences in the RWAs were due to fundamental factors (i.e. risk, collateralization, portfolio mix, etc.) as well as to inconsistencies and discrepancies in the estimates, thus paving the way to new regulation to address the issue and restore trust in internal models.


Among the main initiatives (Fig.1), it is worth mentioning the EBA considerations about the regulatory review of the IRB approach in its Discussion Paper “The Future of the IRB Approach”, published in March 2015: this was the starting point of the roadmap to repair internal models, set out in 2016 and that should be completed by mid-2020 [1].  

From a supervisory perspective, and in conjunction with the regulatory initiatives, in December 2015 the Single Supervisory Mechanism (SSM) launched the Targeted Review of Internal Models, or TRIM: a multi-year project to reduce unwarranted variability of capital requirements that began in 2016 and that will be formally concluded in the first half of the current year. It envisages on-site investigations at more than 60 banks directly supervised by the SSM, with inspections covering the most material and critical models for credit risk and all internal models for market and counterparty credit risk [2]. The issues identified have already been communicated to the banks and have material impacts on capital ratios.

Basel 4: a top down approach with relevant effects

All these initiatives interact with the final Basel 3 framework, published by the Basel Committee on Banking Supervision (BCBS) in December 2017 (which hereafter will be called Basel 4 for the sake of simplicity) [3]. 

This regulation is a central element of the Basel Committee’s response to the global financial crisis to address the major fault lines left in the regulatory framework, i.e. the way in which RWAs are calculated. 

The aim of this reform is to restore public confidence in the banking sector by making their capital ratios sounder and more comparable focusing on three main issues: improving the robustness and risk sensitivity of the standardized approaches for credit and operational risk; constraining the use of internally modelled approaches, also by removing the use of the most advanced approaches for certain credit risk asset classes (namely, low default portfolios and equity) and for operational risk calculation; complementing the risk-weighted capital ratio with a finalized leverage ratio and a revised and robust output floor.

As far as RWAs variability is concerned, Basel 4 differs from the other initiatives (by the EBA and the ECB) as it is a top-down approach. Figure 2 describes the main elements of the revision [4]: 

  • amendment to the standardized approach (SA) for credit risk, which becomes more granular and more risk sensitive;
  • in-depth review of internal models for credit risk that introduces constraints on their use;
  • revision of the output floor that limits the amount of capital relief a bank can obtain from the use of its internal models: banks’ calculations of RWAs generated by internal models cannot, in aggregate, fall below 72.5% of the risk-weighted assets computed with the standardized approach on a fully-phased basis (2027);
  • setting a new standardized approach for credit valuation adjustment (CVA) risk and operational risk, replacing the current methods;
  • adding a leverage ratio surcharge for the largest banks (G-SIBs), equal to 50% of their risk-based capital buffer.

The agreement also includes implementation dates and long phase-in arrangements (Fig.3): most changes should be operational from 2022, while only for the output floor has a phase-in period that is set to end in 2027.


However, it is worth noting that Basel 4 is a supranational regulation that must be transposed into national laws to be effective (Fig.4). In the European Union, after the recent entry into force of the CRR 2 and CRD V, a new set of rules is expected to transpose Basel 4 regulation shortly: by summer 2020 the drafts of the CRR 3 and CRD VI should be ready, in order to allow its application in a timely manner. 

Regarding implementation, the EBA made a series of recommendations to the European Commission, the most important being an accurate implementation of Basel 4 standards into European legislation to ensure a globally levelled playing-field (as recently advocated also by EBA president José Manuel Campa). 


By reading the rules published by the Basel Committee, Prometeia expects that their introduction will entail an increase in RWAs. To quantify it, we performed an impact study of the implementation of Basel 4 rules on credit risk. The sample of the analysis includes 28 significant banks: 7 Italian, 10 German, 5 French, 6 Spanish [5].  

Due to the complexity of the exercise, several simplifying assumptions had to be adopted, the most relevant being static balance sheets: we assessed the impact of full Basel 4 implementation on the balance sheet at June 2018 and we did not consider any managerial action or change in balance sheet composition.  The estimates focus only on the changes in credit risk, the introduction of the output floor fully phased (72.5%) and the removal of the scaling factor, which is the 1.06 multiplier that is currently applied to RWAs determined by the IRB approach to credit risk. 

In summary, the assumptions are rather conservative and the output is a comparison with the actual capital ratios that are compliant with CRR/CRD IV rules.

Prometeia’s assessment is that the reform has a material impact on capital ratios and that the impact is rather different across banks. At country-level, the reduction in capital ratios ranges from 144 basis points for Spanish banks on aggregate to 300 basis points for German institutions (for which RWAs increase by almost 30%) (Fig. 5). These differences are also the result of the heterogeneous use of internal models by European banks: as shown in Figure 6, internal models for credit risk are not equally widespread among Significant European banks, with a range going from over 90% of the credit portfolio for Nordic intermediaries to less than 50% in some Mediterranean countries.


The greater impact on banks that make a more intensive use of IRB models does not come as a surprise, as the main target of Basel 4 rules is to reduce capital savings stemming from internal modelling.


However, the solvency position of European banks seems solid, as the banks most impacted by Basel 4 reform are those with higher “management buffers”, that can therefore afford a reduction in their capital ratios without incurring in any sanction. Moreover, the recently approved CRD V states in Article 104 that banks can comply with Pillar 2 requirements not only with Common Equity Tier1 capital (CET1), as it is the case now, but also with AT1 and Tier 2 instruments. This implies that banks will need less CET1, all things equal. This element of the new directive could therefore somehow mitigate the impact of Basel 4 on banks’ capital requirements.

The latest EBA quantitative analysis on the impact of Basel 4 implementation on a sample of 189 European banks (December 2019) shows that the minimum capital requirement will increase by 23.6% on average (+14% the impact of credit risk and output floor). This increase in capital requirements implies an aggregate shortfall in total capital of about 125 billion euros.

A relevant feature of Basel 4 is its impact on capital ratios through the increase in RWAs. This has important second-round effects on other regulatory requirements, the most important being the minimum requirement for own funds and eligible liabilities (MREL). In fact, higher RWAs entail higher MREL requirements, since with the second Bank Recovery and Resolution Directive (BRRD 2) they are based on RWAs and not on Total Liabilities and Own Funds (TLOF), as it was in the previous BRRD. Thus, all else equal, an increase in RWA implies higher funding costs for banks. That is why, in the process of implementing Basel 4 rules in EU legislation, the EU Commission required the EBA to conduct more detailed analysis on the impact on MREL requirements. The response will be delivered in the first quarter of 2020. 

In principle, there could be effects also on Pillar 2 requirements and other regulatory requirements that are expressed as a percentage of RWAs. However, on this last point, we think that capital requirements will be reviewed in light of the recalibration of Pillar 1 requirements, as the EBA recommended to the European Commission.


Basel 4 represents the final milestone in the process of reform of the regulatory setting after the Great Financial Crisis. Prometeia’s analysis focused on the changes to credit risk and the introduction of the output floor, and showed that the impact on the capitalization of European banks will be substantial, although dependent on the use of internal models. We estimate that CET1 capital ratios will decrease by 140 basis points for Spanish banks up to more than 300 pips for German ones.

Several other studies also assess the macroeconomic effects of the Basel 4 reform. The evidence so far is that long-run benefits outweigh short-term costs. However, as advocated by many stakeholders, we think that an overall assessment of the post-crisis regulation should be performed, in order to obtain a holistic view of the impact on banks’ strategies and performances on one side and on the real economy, on the other.

Could Basel 4 be the regulatory ‘end of the game’, as stated by former ECB President Mario Draghi in 2017? It could be not: the regulatory treatment of sovereign exposures is still on top of the euro area political agenda and it has resurfaced the public debate recently in relation to the completion of Banking Union. In 2017 the Basel Committee released a discussion paper declaring that the Committee had not reached a consensus on any changes to the treatment of sovereign exposures, therefore deciding not to consult on the ideas presented in the paper.

However, there are many signs pointing to a change in attitude. One of these is the recent BCBS discussion paper fostering the adoption of a disclosure template on sovereign exposures: the 'end of the game' might still be some way away.

[1] It consists of four phases: (1) IRB assessment methodology; (2) definition of default; (3) risk parameter estimation and treatment of defaulted assets; (4) credit risk mitigation. Supervisory benchmark exercises support these regulatory enhancements in order to support authorities' assessments of internal models and produce comparisons with EU peers. The implementation date for institutions of all regulatory products developed based on the IRB roadmap has been aligned and specified as 31 December 2020 in the original plan, so that all requirements can be fully applied starting from 1 January 2021, in accordance with the application date specified in the RTS and GLs.
[2] TRIM covers internal models for credit, market and counterparty credit risk: operational risk has been excluded, given the Basel Committee on Banking Supervision’s stance against using internal models for such risk. The main findings of the TRIM are that some banks lacked a policy for model changes, did not back-test their internal models annually, used the standardized approach to calculate own funds requirements without formal approval by the supervisor, and/or failed to ensure that staff validating the models were not involved in developing those same models.
[3] Taking Basel 4 into account, the EBA  for example  allowed for the changes in the loss given default (LGD) and conversion factors models for low default portfolios to be implemented by the end of 2023.
[4] For a detailed analysis of Basel 4 rules refer to our 2018 white paper: “Basel 3 or Basel 4? Where do we stand?
[5] The banks are: Unicredit, Intesa Sanpaolo, Banco BPM, Monte dei Paschi di Siena, Ubi, Bper, Credem; Deutsche Bank, DZ Bank, Commerzbank, LB Baden-Württemberg, Bayern LB, Deutsche Apotheker und Ärztebank, Erwerbsgesellschaft der S-Finanzgrupp, Hamburg Commercial Bank, Helaba, Aareal Bank AG; BNP, Credit Agricole, Société Générale, BPCE, Credit Mutuel; Santander, BBVA, Banco Sabadell, Bankinter, Caixa, BFA Holding (Bankia).
Leggi tutti gli insight