Interview at the 27th European Financials Conference organised by Goldman Sachs

Transcript of a conversation between Andrea Enria, Chair of the Supervisory Board of the ECB, and Chris Hallam, at the Goldman Sachs European Financials Conference on 13 June 2023

14 June 2023

So let’s start with some of this recent turbulence. How was this crisis, or the recent turbulence, different from the issues we saw around COVID-19 and the Russian invasion of Ukraine? Does the fact that this crisis originated within the banking system itself change the key lessons to be learned?

Yes. For the banking sector, COVID-19 and the Russian aggression against Ukraine were purely exogenous shocks that tested its regained resilience and its ability to support the economy in an environment of heightened risk. By contrast, the issues of the regional banks in the United States and of Credit Suisse have been more rooted in weaknesses in the banks themselves. So for banks, and especially for supervisors, this triggers questions over what we can do better. Are there lessons for us? I think that the US authorities have also been pretty candid when identifying the main shortcomings on the supervisory side, in terms of the focus of their supervision, the ability to shift quickly to the risks posed by a rapidly changing interest rate environment, and especially the effectiveness of supervision. So the ability to remediate issues once they were identified in the banks, which is a particularly delicate issue that we are also focusing a lot of our attention on. There are lessons in terms of governance, risk management and business models that we need to reflect on.

I’ve seen that there is a tendency for the banking sector to say that the cause was the rapid shift in the interest rate environment, but that was actually the trigger. The real underlying problems were weaknesses in governance, poor asset-liability management practices, poor management of interest rate risk in the banking book and a lack of attention on the liability structure and to the deposit structure. So all these issues are now capturing more attention – and rightly so – and I think that we should focus the attention of the banks on them.

In the space of just several weeks, we’ve had several bank failures in the United States. We had a failure of a global systemically important bank here in Europe. Does it surprise you at all that, effectively, the ripple effects on the rest of the European banking system have been relatively minimal?

We’ve been saying from day one that the read-across from these cases to European banks was not there. In the banking union we don’t have any banks with as extreme a business model as Silicon Valley Bank or the other regional banks, in terms of high dependence on a concentrated and connected network of uninsured depositors like Silicon Valley Bank, or such an extreme exposure to interest rate risk in the banking book, or such a huge amount of unrealised losses. So that was simply not present in our banking sector.

Credit Suisse had weaknesses in the business model, difficulties in remediation or in terms of profitability and depressed valuations. This is something that we know well in the European banking sector and that we have also seen at European banks. But, there has also been significant change here in the last few years. We have seen profitability improving at European banks as well. It has been more an issue of who has been best at increasing dividend payments and buybacks rather than still being in the process of remediating those difficulties. We have made a lot of improvements in internal risk management in recent years. So, from that point of view, it was not a surprise.

We have seen a very fragile market sentiment, a lot of volatility, a lot of searching for the next weak link, and this concerns us. Neither we nor the banks should be excessively relaxed. We should still be focused on risk management. The impact on asset quality of the change in the interest rate environment has not yet materialised, but it will be coming, so I think that we need to keep our eyes on the ball and stay focused on risk management.

You talked a bit about the transformation of the banks post-financial crisis, both in terms of the actions taken by the supervisors, but also the banks themselves, if we think about higher profitability, lower non-performing loans, better capital ratios, etc. Two questions to follow up on that. Number one: how important has that transformation been in actually adding resilience at system level? Number two: if we can take it as read that the banks are in a much better position now, is one of the bigger risks from here actually complacency?

First of all, let me say that it has been a long, difficult journey. European banks have to some extent been laggards in remediation post-great financial crisis. They also had the double whammy of the great financial crisis and the sovereign debt crisis. But the speed of reorganisation, of cleaning banks’ balance sheets, of strengthening the capital position, of restoring profitability has been much slower than for banks in other jurisdictions. So we have been struggling with a banking sector that was very slow in coming back from that blow.

But I would say there has been a significant improvement in the last few years. In terms of asset quality, even the banks that were the extreme outliers in terms of legacy assets have now made major progress. Capital ratios are in the same ballpark as banks in the United States or the United Kingdom. Profitability is not yet completely where we and the managers and particularly the shareholders would like it to be, but there have been major improvements. In the first quarter of 2023 we saw listed banks having a double-digit return on equity and now approaching the cost of equity, and most banks, especially the largest ones, are now credibly projecting earning their cost of equity in the near future. So I think that we have really turned the page for the focused efforts of the management of the banks. Our supervisory pressure as well. And we should, of course, be happy about that.

Having said that, you’re right. We should be careful not to be complacent. What concerns me the most is that, as I mentioned before, we still see in this market environment that in moments of turmoil, there is sometimes this sudden shift by investors and markets from a balance sheet view of the banks to a mark-to-market view in which a bank’s capital ratio or liquidity ratio does not matter anymore. You look at the mark-to-market view, at how much the bank is valued on a mark-to-market basis. This shift can be very destructive, so we need to place a lot of attention on that. Banks need to be careful when they manage interest rate risk, not only from the earnings perspective but also in terms of the economic value of equity. We need to make sure that we also understand how the markets look at banks. Disclosures need to be very granular to explain where the banks stand. So we still need to be very focused on the risk environment, which is still shifting. The hiking cycle is not over yet, so we need to avoid complacency and remain very focused on risk management.

The Fed put out a pretty detailed review of the failure of Silicon Valley Bank, and I think reading through that, if you think about the three pillars of regulation, governance and supervision, its conclusion was that this was more of an issue around governance and supervision than regulation per se. Is that a conclusion that you would agree with?

I wholeheartedly support that view. I think that in the industry and in the regulatory community we are all exhausted from the long reform cycle we have had. I think that we are bringing this to an end with the implementation of the final Basel package. I don’t see any reason for any wholesale change in the regulatory framework. It would be a mistake to calibrate international liquidity standards based on an extreme business model of a regional bank in the United States. If you find these types of issues, these fragilities in the funding structure, you should address them with supervisory tools, not by changing the international standards.

The Credit Suisse case will still require reflection in the sense that there were some pretty upsetting conclusions reached by some Swiss authorities, particularly the Finance Ministry, on the effectiveness of resolution. I sometimes hear quite dismissive comments from bankers saying “We spent a lot of money on recovery and resolution planning and this doesn’t seem to be useful.” I don’t see this. I still think that all the work that we are doing collectively – supervisors, resolution authorities and banks – to prepare for a crisis is very useful. I’m still convinced that the framework for resolution works, and if there are issues that we need to adjust to make sure that it does work, then we should focus on that.

Before I turn to the audience Q&A, I just wanted to touch on this point around Credit Suisse. So one of the notable consequences of the failure of Credit Suisse was the impact on the broader Additional Tier 1 market. Do you think that post that event, there is now sufficient clarity on where Additional Tier 1 fits within the broader hierarchy/capital structure of European banks?

It should not be a surprise that if a bank goes into a crisis, holders of Additional Tier 1 take the losses. These instruments are counted as regulatory capital so they are supposed to absorb losses in a situation like this. Of course, as you said, there was this point that really upset the market, which was the fact that while equity holders were maintaining value in the bank, Additional Tier 1 holders lost all their investment. So there was a perception that there was an inversion in the hierarchy of claims and that Additional Tier 1 holders are basically junior to equity. Together with the with the Single Resolution Board and the European Banking Authority (EBA) we came out pretty strongly to say that this cannot happen in the European Union.

First of all, we don’t have instruments with supervisory triggers, so the write-down cannot simply be triggered by public intervention or a decision of the supervisor.

Second, we have a majority of instruments which have a temporary write-down rather than a permanent write-down. And most importantly, our resolution framework would push us to respect the hierarchy of claims. And legally, one of the things that is clear is that the European Union is a community based on the rule of law. So we couldn’t do anything but follow the hierarchy. And we would do it even in the case, which I personally deem very unlikely, where a rescue operation were to be framed outside of resolution. So I hope this clarifies that.

So far, we have not seen many AT1 issuances. We see some banks starting to consider issuances. I hope that investors understand that, although there is sometimes the tendency to group European banks and European instruments together as if they were the same, there are differences across Switzerland, the European Union, and the United Kingdom, and sometimes these have to be appreciated by investors, I think.

My question would be on asset quality, essentially the commercial real estate market. Given the higher rate environment, is it a segment which concerns you? Because it is a real topic for the investor community.

If you go back to the supervisory priorities we identified for 2022 at the end of 2021, at that time we saw the change in interest rate environment coming, so the exit from the negative interest rate policy coming to the European banking market, and we very much focused our supervisory programme for last year on interest rate and credit spread risk and on sectors which are particularly sensitive to interest rates. So, commercial real estate first. We did a campaign of on-site inspections and a targeted review last year that identified weaknesses in the risk management of commercial real estate exposures. We sent banks a number of recommendations to remediate these weaknesses: weaknesses in origination, weaknesses in customer monitoring and in the valuation of collateral. So all these topics have been brought to the attention of banks’ management and since last year we have been exerting a lot of supervisory pressure in relation to these issues.

We are now rolling out this type of exercise to residential real estate, which is also showing signs of deterioration in some countries. Although we haven’t yet seen an impact on asset quality, we should be prepared. So the supervisory work is being done. There are weaknesses, but banks are very alert to the topic.

My question is on Additional Tier 1 treatment of coupon payments versus dividend payments. Should we expect an enhancement of the framework that will clearly prioritise bondholders versus shareholders in this respect going forward?

For me, the framework is very clear. When you have Additional Tier 1 instruments, you have the maximum distributable amount trigger. So there is a level of the capital ratios that would trigger a suspension of coupons, and of course the same type of restriction on payments would apply to dividends. So there would not be any difference in that respect, and it would be the flexibility of payments that would represent the main element of loss absorbency, of the contribution of Additional Tier 1 to support the strength of the banks.

When you go down the ladder there are triggers that are defined in the contracts and by regulations that would then determine either the conversion or the write-down of the instruments, so their ability to absorb losses. In any case, they should be treated, in my view, as senior to equity. I mean, it’s surprising that you have to say it – that’s how it is. So I wouldn’t see any need for changes.

Again, it’s a moment in which if any clarification needs to be given, we stand ready to do so, but for us the European framework is pretty clear.

Any thoughts on Russian exposures?

Russian exposures have been coming down pretty steeply and I think the fourth quarter of last year was really a bit of a turning point. In 2022 we saw a 37% decline in exposures to Russian counterparties of European banks under our supervision, and a 25% decrease in the last quarter of 2022. We have been putting a lot of pressure on banks to reduce exposures. All banks are no longer originating credit vis-à-vis Russian counterparties and are trying to bring down their exposures. Most of those who have shops in Russia, let’s say subsidiaries or branches in Russia, are basically winding down operations, running down the books and significantly reducing. Some of them are also trying to sell. That’s difficult because the Russian authorities are putting a lot of pressure on banks and placing obstacles in their way to prevent them doing that. You need presidential approval and to take a significant hit on the investment to be able to do so, but I know that many banks are actively engaged in moving further.

That’s a process that we not only praise, but strongly encourage banks to perform, because there is a huge reputational risk in continuing to operate in Russia, in an economy that is shifting towards trying to limit the impact of sanctions and supporting the war effort. So I think that it is important that banks remain very focused on reducing their exposures further and ideally exiting the market as soon as they can.

Do you have any guidance on what the attitude of the ECB would be towards the results of stress tests? What should an outlier expect to have as a treatment for these results?

In the stress test, we do have a pretty harsh scenario that also has a significant change in the interest rate environment and a very negative macroeconomic situation, a strong impact on real estate and on stock prices. It’s the harshest on record in the European Union. Of course, there is the issue of unrealised losses that came to the fore after Silicon Valley Bank. Together with the EBA, we have requested that banks give us some additional information on unrealised losses, enabling us to conduct a sensitivity analysis of the unrealised losses.

Now, I must say that in general, this is important to give us a clear picture. With IFRS 9, banks are already under an obligation to disclose the unrealised losses that they have in their holdings at amortised cost. But we have no place where this is seen in a consistent way, also taking into account hedging practices. So that was, for us, a very important point.

I can say that for all the banks under the supervision of the ECB, the overall amount of unrealised losses is pretty contained. It is in the ballpark of €70 billion. Compared with what the US authorities have disclosed recently – which for the whole of held to collect and available for sale not factored into regulatory capital was north of 620 billion – it’s much more contained.

Therefore, for us, the point would be to provide input to the supervisors in their supervisory review process, to look at these figures and to look at unrealised losses also in combination with the overall exposure to interest rate risk, and maybe also the deposit base and the liability structure. So how volatile your liability structure is, how exposed you are to interest rate risk in the banking book, and how many unrealised losses you have, of course, would be indicators for some supervisory discussions.

Post-sovereign debt crisis, the mantra was that the regulator had to produce a clear firewall between sovereign risk and banking risk. And so far, if I look at the concentration of government portfolios in banks, especially the most exposed like the Italians, that hasn’t happened. Also, capital charges for a concentration of that nature is not there. Is that no longer an issue or no longer a worry for the ECB?

The picture is not as gloomy as you put it. There has been a significant decrease in the amount of concentration. I remember – I was at the EBA at the time – that there were cases of banks which had 12 times their capital in an individual sovereign. And most of it was at amortised cost and there was no hedging. So, you don’t see this situation anymore.

It’s true that this issue has not been addressed yet. That’s clear. There have been a number of initiatives that have been taken or proposed but there hasn’t been a lot of change there.

Anyway, for instance, when doing our supervisory work on interest rate and credit spread risk, we have looked at exactly that. We look at possible scenarios of credit spreads widening and how this would impact the banks and we ask banks to actively manage these risks – sometimes by diversifying their portfolios, hedging and things like that.

One point I’ve been making from time to time is that, ­as I personally don’t think that it will be easy to achieve any meaningful calibration of capital requirements for these types of risk – it’s very difficult, if you look where the internal models of banks are with reference to these types of risks, they are a bit all over the place – having a requirement for banks to hold at market value the assets that they use to fulfil the liquidity coverage ratio requirements would already provide a strong incentive for banks to more actively manage the risks of these types of exposure, including the concentration in the sovereign portfolios.

In a recent speech, you discussed the impact of technology and the potential flight risk and speed of deposit outflows. What would be a good way to incorporate these lessons into the current environment? Is this a question of more capital in the system, or are there different levers that you, as a regulator, can pull to minimise these risks?

Sometimes it seems that capital is the one and only answer to all the questions ­– I don’t think that this should be the case. If there is an issue in terms of the volatility of the depositor base, this should be addressed by asking the banks to diversify their funding sources, to rely more on term deposits, or to possibly hold slightly larger liquidity buffers. So capital is not always the right answer. That’s my first point.

Second, I think we need to hold fire a bit and understand better. I also have the impression that there were very peculiar forces at play here, because there was a massive build-up of deposits during the pandemic, which was unprecedented. So, to some extent, having a reduction in deposits was logical after the pandemic. It was also reasonable to expect that once monetary policy started to normalise, it would also have an impact on deposits. These two things happened at the same time in a very fast and concentrated period, so it caused slightly accelerated outflows of deposits.

I still need to understand and quantify how much new technologies played a role here. It was mainly uninsured depositors or corporate institutional treasurers. I don’t know if they use their smartphone to move the deposits, I suppose they have other means as well. I think we need to go deeper to understand how much of this was triggered by the peculiar situation we are in right now and how much is indeed a structural change triggered by digitalisation and other changes, and then build up our supervisory thinking on that.

You highlighted how higher profitability has led to higher buybacks and dividends. Do you think there should be any constraint on that given the macro uncertainty you mentioned, with asset quality issues not yet having come through? Should banks be allowed to pay more than 100% of their earnings if they are in a good position?

I think we made our position clear on that. I mean, we are supervisors, so the only thing we look at when the banks make their distribution plans is their capital trajectory. We want a solid capital trajectory projected by the bank under a baseline and an adverse scenario, and if the bank is able to convince us that even under a decently conservative adverse scenario, they would be able to respect all the supervisory yardsticks, then there would be no constraints from our side in terms of their distribution. That’s the only thing we look at.

On liquidity requirements, it seemed like the failures of Silicon Valley Bank and Credit Suisse were significantly influenced by social media, so we saw panic spreading at a rate we hadn’t seen in the pre-social media era. Don’t you think that that in itself justifies some kind of raising of liquidity requirements despite business models? I only say that because Credit Suisse’s business model was pretty much unchanged for 20 years before it went bust.

We do have quite a significant amount of liquidity buffers at the banks. So the real question is, do we really need more? I’m yet to be convinced that we need more. Basically, the construction of the liquidity coverage ratio is: we as authorities want banks to have an amount of liquid resources that enable them to stand on their own feet without support from the central bank for a period of time that allows the authorities or the bank to identify a solution. This period of time is pretty large, so we target one month, and there were assumptions on outflow rates that were calibrated in order to make sure that these amounts of liquidity would be sufficient.

Do we need to review the calibration of these outflow rates? Maybe, but again, I still need to be convinced. For instance, my colleagues in the United States told me that some venture capital deposits were classified as operational deposits that had very low outflow rates and were not matched by the actual speed with which they flew out in the case of Silicon Valley Bank. That could be an issue that one could try to identify.

But again, for me, the real point is where you go with regulation and where you go with supervision. I think that we should really avoid trying to have the ambition to calibrate regulatory tools that really capture all the possible states of the world. This will only make regulation excessively complex and easily circumvented. In my view, the point is for the supervisor to identify situations in which the deposit structure of a bank is prone to panic runs, and in those cases, maybe, yes, ask that bank to have slightly higher liquidity buffers rather than go for a very granular recalibration of all the outflow rates. That would be my preference. The Basel Committee will look into that and come up with an answer to your question.

You will step down from your role at the end of this year. I’m sure everybody in the room wants a nice, quiet few months between now and then. What advice would you give to your successor?

My first piece of advice is that we made a significant effort to make the ECB’s supervision strong, intrusive and challenging, and I think that this should be maintained. The supervisory teams need to be sure that they have the backing of the Chair and the backing of the Supervisory Board when going to the banks and challenging them. If supervision becomes a sort of legal checklist, it becomes weak and we will have problems at banks. That is in my view the most important issue: strongly backing the supervisory teams.

The second is that I think it is important to understand, to listen to investors, to analysts, to understand how the market perceives the banks, to communicate, to try to make the markets understand what you are doing and what you are trying to achieve, and to try to be predictable. There is sometimes this perception I feel in the market that supervisors surprise, that banks surprise the markets, and I think that we should avoid that. I made a major effort to increase transparency, disclose methodologies, and explain what we do. I think that it is very important to continue this dialogue and make sure that markets understand what we do.

So be tough and be predictable would be my advice.