“What got you here won’t get you there”
Fernando Restoy, Chair of the Financial Stability Institute at the Bank for International Settlements and member of the expert group that reviewed the ECB’s SREP, explains the group’s recommendations to streamline processes, redefine the scoring, focus on qualitative measures and modify the approach to calculating P2R.
17 May 2023
Over seven months, you reviewed the effectiveness and efficiency of the ECB’s Supervisory Review and Evaluation Process (SREP) and how it relates to other supervisory processes. What is your view of European banking supervision?
The exercise confirmed that the Single Supervisory Mechanism (SSM) is fulfilling its mandate. In record time, the ECB and participating competent authorities managed to put in place a fully effective supervisory regime for all significant banks in the euro area. Moreover – and this was a big challenge from the very beginning – our review confirmed that a well-defined and articulated supervisory methodology has been established, which is consistently applied to all banks thanks to the ECB’s internal controls, including a strong and rather unique second line of defence. Finally, a common supervisory culture is certainly being progressively developed by taking the most positive elements of national supervisors’ varied practices.
But, of course, the success achieved so far does not guarantee future success. As it is often said: “what got you here is not necessarily going to get you there”. ECB supervision needs to continually evolve to continue being effective. For that purpose, the ECB could leverage the maturity of the common supervisory procedures reached to date and, in particular, the strong commitment to ensuring a level playing field across institutions. The credibility gained so far allows the ECB to take ambitious steps to further improve. That could be accomplished by increasing the risk sensitivity of supervisory actions, streamlining processes, better establishing priorities, applying more judgement to complement the information provided by quantitative methodologies for risk assessment and strengthening the role of qualitative measures aimed at promoting adequate risk management.
You flagged the need to streamline the supervisory processes. How could the SREP become leaner? After all, it is a process spanning 21 countries, more than 100 banks and thousands of staff members.
Indeed, by definition the SREP needs to provide on an annual basis an overview of the risk profile of the institutions and the measures required to preserve their safety and soundness. This entails, and will always entail, a fair amount of complexity. That said, there is clear dissatisfaction on the part of all relevant stakeholders regarding the length and operational costs of current procedures. The expert review group believes that there are ways to make the process leaner and shorter. Our main recommendation in that regard is to reduce and simplify SREP-specific analysis and to rely much more on ongoing supervisory actions. Moreover, supervisory decisions taken outside the SREP could well feed into SREP decisions and, possibly, be escalated if so warranted.
We also believe that by separating the risk assessment component of the SREP from the decision on capital add-ons (Pillar 2 requirements (P2R)), running some of the process in parallel and reducing the touch points with the Supervisory Board, the length of the SREP process could be significantly shortened, possibly by up to three months.
Another interesting recommendation was to score a bank’s ability to act, instead of scoring the viability risk for banks. To what extent would this improve the SREP?
The current scoring system aims to provide an overall assessment of a bank’s viability risk and does a good job in that regard. Yet, by definition, that risk does not vary much over time, nor do the scores. That stickiness over time may dampen the ability of the current scoring system to reflect banks’ improvements and the system may therefore fail to provide incentives for banks to address their weaknesses.
Following the example of other important jurisdictions, we believe that rather than reflecting banks’ absolute viability risks, scores could be redefined to reflect supervisors’ assessment of banks’ ability to address their most important weaknesses. As progress made by banks to meet supervisory expectations varies in both directions over time, so will the scores they receive. Importantly, that redefinition would clarify the rationale behind score assignments and strengthen banks’ incentives to implement the actions requested by the supervisor.
This links back to your recommendation to make more use of qualitative measures to move away from the ECB’s capital-centric approach. What is the added value, bearing in mind that more qualitative measures would not result in lower capital requirements?
Yes, we believe that the SREP methodology is quite capital-centric at present. This has nothing to do with the current level of capital requirements, which we feel are broadly appropriate. Our concern is that a specific element of the SREP, namely P2R, is seen as the key outcome of the exercise, if not the only important one. This is unfortunate as the main outcome of supervision should not be the determination of banks’ required loss-absorbing capacity but the identification and monitoring of the actions they should undertake to mitigate those losses. Moreover, if you think of risks related to governance or business model sustainability, and also others like operational resilience, it is hard to argue that there is a level of capital that could effectively cover those risks by itself. Promoting sound governance, and appropriate management strategies and actions is essential to ensure a satisfactory supervisory outcome.
The prominent role currently played by P2R is the consequence of several features of the current methodology. First, contrary to what happens in other jurisdictions, P2R is, by and large, directly derived from overall risk scores. In other words, there is no specific analysis of whether the risks identified for a particular institution would be best addressed by additional capital or by qualitative measures and, in case capital should be the right instrument, whether or not the required capital coverage is already provided by Pillar 1. Second, the qualitative measures are not always formulated in the most effective way and are often insufficiently prioritised.
Following on what you said about qualitative measures, they seem to be a hard nut to crack for supervisors and banks alike: they must be measurable, achievable and provide a clear sense of direction and prioritisation. What should the ECB change here?
You are right. Effective qualitative measures are hard to define as this requires thorough knowledge of the strategy, business model and internal organisation of the institution. Moreover, banks often find them excessively intrusive. Yet supervisors find them essential, particularly in the current context in which the banking sector increasingly faces risks outside the traditional credit and market categories.
Qualitative measures are part of the current methodology and are frequently used within and, mostly, outside the SREP process. However, the high number of qualitative measures currently conveyed by the ECB to the banks are not sufficiently prioritised. Sometimes they are not drafted precisely enough, and often they do not contain a sufficiently clear timeline for actions that could facilitate their monitoring. It is very important for banks to know what priority actions are required and by when. The SREP communication is the most powerful instrument to convey this information to the bank, which should be adequately framed in a sufficiently rich description of the bank’s risk profile and the challenges faced in the environment in which it operates. The ECB is already moving in that direction, but more needs to be done. Finally, for the sake of clarity and prioritisation, we propose to the ECB to channel all relevant communication to the bank through Joint Supervisory Teams (JSTs). The direct transmission to banks of the often detailed findings of on-site inspections or horizontal reviews without JSTs’ participation does not always help to convey the key supervisory expectations clearly and consistently to the banks.
The report implies that the ECB overengineered the process for setting P2R and you suggest rethinking it. What led you to this conclusion?
As I have said before, since the beginning the ECB’s determination of P2R has been mostly based on the overall risk scores. Recently, the ECB made the process more sophisticated by adding a decomposition of score-based P2Rs in different risk components. That decomposition is based on banks’ internal methodologies to determine capital adequacy (ICAAP) which must be assessed by the supervisor. The experience so far is that the outcome of that risk-by-risk decomposition is of little practical use although it is operationally complex and consumes a relatively large amount of resources. Moreover, this mixed approach produces conceptual inconsistencies as it tries to combine two mutually incompatible methodologies to determine capital needs. Importantly, it relies on internal methodologies which are not generally considered to be sufficiently reliable.
Against that background, we feel that this rather artificial ICAAP-based risk-by-risk decomposition adds little value. In fact, we believe that a non-ICAAP-based risk-by-risk approach could be a more rigorous one for calculating P2R. In particular, we propose that the ECB develop its own methodologies to identify the risks banks face that, while requiring capital coverage, this is not sufficiently provided for under Pillar1. That approach would allow the determination of P2R to be separate from risk scores, allow qualitative and quantitative measures to address each specific risk to be more effectively combined, and minimise the reliance on insufficiently robust ICAAPs.
Something that could further help to identify adequate capital requirements for each bank would be to anchor those requirements to the outcome of a previous discussion on the average capital needs for the system from a microprudential perspective. That discussion, which could take place when the Supervisory Board discusses supervisory priorities every year, could lead to both the formulation of an expected average P2R and the desired stringency of the stress test exercise used to determine the other capital component of Pillar 2, namely Pillar 2 guidance (P2G).
Finally, you also highlight supervisory culture as one of the key avenues for evolution. Can you explain this aspect a little more?
Any relevant modification to supervisory approaches and methodologies can only become effective if it is in line with the culture of supervisory organisations. The ECB has developed that culture over the years, leveraging a strong “tone from the top” approach. Supervisors do have a clear view of the main strategic priorities and all seem attuned to key objectives that include not only the preservation of the safety and soundness of supervised institutions but also the development of a level playing field across them. Moreover, the ECB has taken steps to formalise that culture by publishing a mission statement and developing a high-level risk tolerance framework.
In the report, we propose some changes to the supervisory approaches and methods that, while not affecting the main values and the mission of ECB banking supervision, may require some adjustments to that supervisory culture. In particular, the suggestions to further increase the risk sensitivity of supervisory actions, to enlarge the space for judgement and to strengthen the role of qualitative measures may imply a reordering of priorities and a slightly higher appetite to take on legal risks. In that context, we propose that the ECB take further steps to develop and formalise the current supervisory culture, expected behaviours of supervisors and the risk tolerance framework.
- The expert group also included Sarah Dahlgren, former Executive Vice President, Federal Reserve Bank of New York, Partner, McKinsey & Company; Matthew Elderfield, former Deputy Governor of the Central Bank of Ireland and former Chief Risk Officer of Nordea; Ryozo Himino, former Commissioner of the Financial Services Agency of Japan (JFSA) and former Secretary General of the Basel Committee on Banking Supervision; and Carolyn Rogers, Senior Deputy Governor of the Bank of Canada and former Secretary General of the Basel Committee on Banking Supervision.
European Central Bank
Directorate General Communications
- Sonnemannstrasse 20
- 60314 Frankfurt am Main, Germany
- +49 69 1344 7455
Reproduction is permitted provided that the source is acknowledged.Media contacts