Interview with AFME-OMFIF

Transcript of a conversation between Andrea Enria, Chair of the Supervisory Board of the ECB, and Adam Farkas, at an AFME-OMFIF event on 23 May 2023

24 May 2023

I think one undisputed success of creating the Single Supervisory Mechanism (SSM) is the increased consistency of the Supervisory Review and Evaluation Process (SREP), the Pillar 2 risk assessment process that banks are subject to. This started with set guidelines, and was then put into practice with the creation of the SSM. Of course, there were two particular challenges in this process. On the creation of the SSM, one challenge was to try to ensure consistency among 19 national practices in supervising banks and carrying out the Pillar 2 assessment. More recently, another challenge has been the significant inflow, or emergence, of international banks in the eurozone, which have grown in size and complexity, but whose home supervisor is often in a third country. In such cases, you have to integrate their risk assessment into the European framework.

My first question is where you think this process stands. What are the major achievements? What are the outstanding challenges? We know that an expert group, as requested by you and the Supervisory Board, recently published an external review. So you also benefited from some outside observers’ views on these questions. Where does this integration of supervisory practices stand today, particularly in the area of Pillar 2, and what are the necessary next steps?

The SREP has been one of the most challenging areas for integration, because we know from the European Banking Authority (EBA) times that there were a wide variety of different practices across the union. So the banking union was a once in a lifetime opportunity to bring this together and achieve single, fully unified supervisory practices, and I think this has been achieved. I think the SREP process is now fully integrated; it is single, and it is run in a very structured way. I’d say that we are now in a transition, to which the report by these experts, who we mandated to review our processes independently, gave further input. I think that at the beginning – I was not there at the time, but I would have probably made the same choices – when you have to find integration and harmonisation at the European level, what we do is we sit around the table and write a paper. We codified the practices in very structured, process-oriented methodologies; we ticked the boxes, you could say.

When I joined the SSM, I received a clear cry for help from the teams who said we could not really supervise banks. We could not devote enough time to monitoring risk because we are always conducting this monster process that more or less takes up the full year and absorbs a lot of energy. So I think that we are already moving on from the start-up phase − in which the SREP process was very codified and process-oriented and was geared to achieving consistency in a rules-based manner − to a process which is much more risk-focused, leaving room for judgement to the line supervisors, and then having internal processes, second lines of defence, for instance, that challenge the line supervisors and give peer comparisons to ensure consistency across the board. The report that was published pushes us to go further in that direction. For instance, we are now starting to apply a risk tolerance framework: we define the priorities and then strongly empower the supervisory teams to decide where to put their chips in terms of resources when carrying out the SREP for their own bank. So we leave a lot of leeway to the supervisory teams, while also having mechanisms in place to ensure greater consistency.

The other point is that we are trying to streamline the process somewhat, to shorten the time to make it more compact, more focused. Another idea is to undertake a multi-year process. Rather than ticking all the boxes every year, we would try to structure it within three years, focusing more on certain aspects in one year than another. So, a lot of, maybe boring, but still very important improvement to make it more risk-focused, lighter and geared in terms of processes, but still very demanding towards the banks. The full point of the report was not to water down the degree of ambition, but rather to change the way in which the process is structured. So that’s where we are right now.

One small additional question around the Pillar 2 process, the SREP process. There has always been a discussion, globally as well as in Europe, on the transparency of the process, of the findings, the capital add-ons, and the reasons behind these add-ons vis-à-vis the supervised entity and also vis-à-vis the public. Where does this debate now stand within the eurozone?

The first point is a complaint that I received from banks. Banks told me that sometimes they receive SREP letters of 60 or 70 pages with a lot of recommendations, and that it was then difficult for them to see what was really relevant in order to change the scores, and effectively change the capital add-ons, and receive a slightly better assessment from the supervisors. So we have now introduced an executive letter, which identifies the main areas we ask the bank to focus on in order to improve its scores and its assessment by the supervisors. We are trying to prioritise what we want the management of the bank to do. I think this is appreciated. What the banks sometimes ask for, which I don’t think we could or should deliver, is to have full disclosure of the black box: how the scores come out, what is the algorithm that produces the score. The reality is that these algorithms are not such a surprise. They are really very basic, but the process is very much based on judgement and if we were to publish the full details of the methodology, it would probably end up in a game of arbitrage or in trying to window-dress the indicator that is known to be used for the score rather than really having substantive engagement with the underlying issue.

We are planning to disclose much more anyway. We have already started being more transparent by publishing more of our methodologies, and we will publish more before I leave at the end of this year. We already started to immediately publish the Pillar 2 requirements in 2019 and in 2020. These are published every year, and I would personally be in favour of also publishing the Pillar 2 guidance. On that I would say I have a little bit of resistance in my board, but also the feedback from analysts and investors that if the Pillar 2 guidance levels were published, they would be perceived as a hard-wired minimum, whereas we want them to be a real buffer in case of need. So there are pros and cons to the issue of disclosure. It’s debatable whether it would be a good thing or not.

Let me move to another topic, which I know is very close to your heart, namely the obstacles to integrating the banking system or creating a genuine single market for banks within the eurozone. In other words, the completion of the banking union. We all know what the stumbling blocks are at the highest political levels, i.e. the key areas where agreement has not been reached, meaning that the banking union cannot be considered complete. But I would also be interested from a supervisory point of view, because you actually see the banks in action, the banks which are exploring different options to benefit from the single market. What obstacles do you still see blocking the optimal functioning and integration of the eurozone banking system?

This could keep me going for much longer than we have. The completion of the banking union is the final goal, because the main argument that is sometimes used against integration is that the final bill would eventually be footed by the national deposit guarantee schemes (DGSs), and that they therefore need to be protected by some sort of guarantee provided locally in terms of capital and liquidity. That’s important, and until we nail down this point, this argument will always be there. But to be honest, in my view, that argument is preposterous. There is a long way we could go within the current framework. We have all the tools to do much more than we are doing right now, and I must say, I’m disappointed because I have made a huge effort. When I started my mandate, I identified this idea of integration, of generating the banking union as a domestic market for our banks as one of the priorities for my mandate, and I must say that’s an area where I didn’t achieve a lot.

I think we have taken many good steps in this direction. When I started, I was told that the SSM policies were preventing consolidation and would therefore be an obstacle to reorganising the sector at the European level. So we tried to clearly state our approach to consolidation, showing that we would not have charged additional capital if two banks were merging and that we would recognise badwill where applicable rather than taking more restrictive approaches. We also provided some leeway in terms of internal models, namely the ability to gradually phase in the integration of the internal models’ frameworks. So we gave a lot of support. A little consolidation took place here and there in Spain and Italy, but nothing really happened cross-border. We made it clear that we were pretty neutral, also because of that dimension. We tried to make it clear that also on the waiver side, on the small waivers that we have on liquidity in the legislation, we were open to considering specific arrangements for banks, also maybe providing some guarantees to address the legitimate concern of some host countries − that if the liquidity flies away, or goes to the parent in good times and then when there is a crisis, the local establishment is left alone − to have a sort of guarantee built into a recovery plan. So we provided some options there as well.

Finally, having seen, as you mentioned in your introduction, what the banks relocating after Brexit have done here in the banking union, with most of them basically establishing the top parent company as a European company, and then using the cross-border merger directive to merge all their subsidiaries into the parent, branchifying basically, and exploiting the branch as a way to provide services across the union without any trapped capital or liquidity. So we said we are also open to considering banks that want to move in that direction. Unfortunately, there have been few initiatives in those areas. There are a number of (possibly legal) impediments here and there and sometimes some headwinds, or concerns expressed by national authorities in a more or less polite way.

The point I want to make is that I don’t see sufficient ownership of this issue of integration on the side of the industry. I remember when I was at the Committee of European Banking Supervisors in the early 2000s, the industry clearly had a strong agenda to push for integration, more harmonisation, integrated reporting, a single supervisor. This was also having a strong impact on the policy tables in Brussels and at the national level. I understand that there is frustration and disappointment about the slow pace of progress in this area – but now that we have the banking union, I think that would be a real opportunity and it’s a pity that the industry is not pushing more for these objectives.

Again, on these areas, branchification, waivers; I get the feeling that they are not really taking the risk of reinterpreting the current framework as it is, i.e. bringing the issues to the attention of the European authorities and triggering the European process instead of just probing the ground in the national environment.

The last question I would ask before opening the floor to a few questions from the audience is as follows. Given what has happened recently, we cannot avoid having a question on recent global developments – in the United States, Switzerland and, to a much lesser extent, Europe. In terms of what we have learnt so far, or what you have learnt so far within the SSM – especially with respect to supervision or supervisory priorities, and supervisory processes or approaches – what are the main lessons from the recent events?

I appreciate your question, because we have this gut reaction when there are crises to immediately go back to financial reforms, to ask: what are the rules we need to change? I think that actually, especially in this case, it’s more an issue for supervision than for regulation. There might be some tweaks in the rules that we might consider, but I don’t see a need for a wholesale review of the regulatory framework or the components of regulatory frameworks. From the supervisory point of view, I think the first point is governance, which is a difficult challenge. All the cases that we have seen are cases in which there were major failures in the governance and internal controls of the banks, and that’s a difficult field for a supervisor. It’s difficult to step in, but I think that we also need to have the strength and courage, as supervisors, when we see that the internal controls or governance of the bank are not working, to ask for change and to ask for change fast.

The second point that struck me in the Federal Reserve’s report, which is something that is also connected to the expert group review that we were discussing before, is the issue of the effectiveness of supervision. Sometimes supervisors are good at identifying issues, but then there is not a sufficiently structured escalation process for having these issues remediated. That was a point that came out clearly in the Federal Reserve’s report on Silicon Valley Bank, and it’s something, honestly, that applies to us as well. Sometimes we have a number of findings that remain unremediated for a long period of time; a large stockpile. So we need to be more effective in identifying what the findings are and which measures are really important – the ones we want remediated within a certain timeframe – and to be able to escalate faster to enforcement or capital; whatever measures we deem appropriate. While for the others, maybe we should sometimes deprioritise them and pass them over to the internal audit of the bank and let the bank deal with them. But supervisory effectiveness, in my view, is one of the most important lessons.

The last lesson, maybe, is that there are moments in which markets switch from taking a going-concern earnings view of banks to a mark-to-market view of banks. This happened when I was at the EBA during the sovereign debt crisis. You’ll remember that there was immediately a lot of focus on the mark-to-market of sovereign portfolios, and now again, with the issue of unrealised losses, with an increase in interest rates, the sudden change in the interest rate environment. So I think we also need to focus more on the market valuations in the supervisory process, much more than we are doing currently.

Thank you Andrea, as always it is a real pleasure to listen to you. We have some time for a few questions from the audience now.

Do you think some European banks need bigger liquidity buffers? And is liquidity a special focus of the current SREP process after the recent events in the United States and Switzerland?

Liquidity has been identified as a priority. Liquidity and funding were already identified as a priority for ECB Banking Supervision in December last year. So this will indeed be a focus for our SREP process this year. We also have a number of targeted reviews and onsite inspections. We are entering into a dialogue on banks’ funding plans, also with the exit from the targeted longer-term refinancing operations. So there is a lot of work under way on liquidity, and this will definitely be factored into this year’s SREP process.

I’m also seeing some inaccurate reports based on the interpretation that the SREP process equals capital requirements. The SREP process is much more than that. The core of the SREP process is actually qualitative recommendations to the banks where we identify weaknesses, to ask for remediation. So that’s the bulk of what I would expect banks to do. If we identify weaknesses in their liquidity management, then indeed we will ask them to remediate, and I cannot rule out that, in some cases, we might also ask them to beef up their liquidity buffers a bit. But that’s not something that is an objective or something that I have instructed my teams to do.

Going back to the situation in the United States, the Barr Report. From the perspective of a supervised bank, I’ve come to the conclusion that it will not happen to us. I see the flaws on interest rate risk, the flaws on liquidity, and I go back to the review that we go through with the Joint Supervisory Teams every year. Again, it is not the case for us, although for sure there is room for improvement, as you say. But this is the impression I have, and I would like to also know your views on that.

There is a bit of creeping complacency in the European banking sector that we are out. It’s true that there is less exposure to interest rate risk. For instance, the IMF published a graph that shows the different impact of unrealised losses. So, losses on sovereign portfolios, which are at amortised cost, if you were to mark them to market, what would the capital impact be? Their data showed more than 250 basis points for the US banks, and less than 50 basis points for the European banks. So there is less exposure.

However, when I go back to the work we did on interest rate risk last year, there are a number of findings and a number of weaknesses that we have identified in a number of banks. Banks have been looking at the increased interest rates as a positive from the earnings perspective. But very few of them have focused on the economic value of equity, on the impact this would have on their net worth. Let’s say that their behavioural assumptions on depositors leaving the banks are sometimes very optimistic, and they are also sometimes very optimistic on the pass-through of rates to depositors.

I think we should also be very focused on the challenges that the increasing interest rate environment – which can be seen as a good thing – is posing to banks’ balance sheets and be very, very attentive in how we deal with that.

I wondered if you might share with us whether you have any reflections on the most recent set of proposals from the European Commission on crisis management and deposit insurance, and whether you maybe see any small shoots of growth and hope, green shoots for enabling the branchification processes that you talked about in that package.

First of all, it’s a good package. I’m concerned when I see these packages looked at with the usual political sensitivities, which make everything that has to do with the safety net so difficult at European tables. The issue here is making the exit of banks from the market smoother and easier. First of all, I don’t think we have a dysfunctional crisis management framework. So we can manage crises. The problem I see is that we have a huge pot of money, be it the Single Resolution Fund or the whole constellation of national DGSs. We are in the same ballpark as the Federal Deposit Insurance Corporation in terms of the overall amount of resources. But when we come closer to the wire and one bank is failing, it’s very difficult to activate this amount of money to finance a smooth exit from the crisis on a “least cost” basis – saving money for the taxpayer and for the fund itself. So it’s very difficult. We have very little time to do that, and the criteria are very difficult to manage.

So what I think the proposal does is to enable the authorities, on the basis of the experience of these first years of the banking union, to have a little bit more room for flexibility to activate the DGSs on a least cost basis, on an harmonised basis, to have the same type of approach to accompany a medium-sized or small bank out of the market without there being huge damage. The key point is that this package, the different components, all go together – the extension of resolution, the ability of the DGSs to support and bridge to the Single Resolution Fund threshold, the harmonisation of the least cost test. If you take one out, the whole construction crumbles. So that is a bit of a concern I have with the difficult political negotiations we have ahead of us, but I think it’s a good package and I hope that it is passed quickly.

On your second question, I was a bit disappointed because one point I brought to the attention of the Commission was the fact that the portability of the contributions to DGSs, or rather the lack of portability, is one of the constraints on branchification. So if you want to move a subsidiary into a branch, you will lose most of the contributions that you had made to the local DGS. Fixing this would have been an important improvement to support more branchification. This is not in the proposal now, but there is still room and I hope it could be considered in the debate going forward.

How would you rate the interplay in the Joint Supervisory Teams between the ECB and the national competent authorities? Just like you, I’m a practitioner myself. A practical example that leads to frustration, I should call it, for many of us, is the succession of board members, executive committee members and critical functions, which takes much longer than we would hope, contrary to market practice. I think it’s due to the fact that it’s a sequential process and it just takes much longer. That’s just one example. How would you rate the interplay?

We are doing a lot to improve the interplay. There is one important objective that is very strongly shared by the Supervisory Board – increasing integration within the SSM, so to work as one team. We have developed a number of initiatives to do exactly that. For instance, in the field you are mentioning, we are now relying much more on delegation. So in plain vanilla cases, we delegate fully to the national authority. They do basically all of the preparation, and then the decision is taken. There are no additional steps. So we are trying to simplify and streamline processes significantly.

I would also ask for your understanding. You know that we have a legislative package in the pipeline in which we repeatedly asked for the rules on fit and proper to be harmonised. You have to understand that we have to run different processes in each country, in all 21 members of the banking union, in terms of timeline, in terms of whether the assessment is ex ante or ex post, in terms of the documentation. It’s a mess, honestly. So it’s clear that this also has an impact on the process. There was a very good proposal by the Commission in this area to harmonise that. But I’m afraid that the Council has not been supportive of the proposals. We have to understand that, when we have to run a lot of different approaches, there is a cost to be paid in terms of procedural workload.