Interview with Financial Times

Interview with Claudia Buch, Chair of the Supervisory Board of the ECB, conducted by Martin Arnold

18 March 2024

Let’s start with your assessment of the main risks facing euro area banks. How do you view these?

Something one needs to say upfront is that actually European banks are in a good position in terms of capital levels. Of course, the main reason why we now have good levels of resilience in the system is because of the post global financial crisis reforms. Looking forward, however, we have a few challenges that the banks have to deal with.

There has been a series of shocks hitting European economies, with the pandemic, Russia’s war in Ukraine, and the increase in energy prices. We have also had rather unprecedented policy support for the corporate sector and for households, which has indirectly benefited the financial sector because losses were contained during this period. This brings me to the forward-looking part. Certainly we’re not out of the woods yet, because the significant increase in interest rates still has to filter its way through the financial system. We’ve seen the relatively short-term impact on banks in terms of higher profitability, but it takes a while until all assets have been revalued and until we’ve seen how higher interest rates are affecting creditworthiness and levels of distress in the corporate sector.

Then we have new risks that the banks have to deal with. We’re constantly reading in the media about geopolitical risks, and the banks generally have to deal with a new macro-financial environment. There is a need for structural change in our economies, which will inevitably show up on banks’ balance sheets because they are a mirror image of the underlying real economy. We’re already seeing increases in defaults, underperforming loans and arrears. We are paying very close attention to this. It is the reason why we have to be very vigilant and also make sure that the system is sufficiently resilient.

Some European bank bosses complain they are held back by regulation? Does the relatively low market valuation of European banks worry you?

This is what we also sometimes hear from industry – that we are too strict. I actually think good capitalisation makes a bank stronger and more competitive. So we’ve looked into this argument; we’ve looked at what it would mean if European banks, particularly the larger globally systemic ones, were under the same rules that apply in the United States as of now. If anything, we don’t find evidence that our rules are stricter for these largest banks. For Europe’s smaller and mid-sized banks, US regulation would result in slightly lower capital requirements. But I’m quite glad about our stricter approach, given what happened recently at several mid-sized US banks. In any case, the banks’ argument, if you look at the facts, doesn’t really play out.

If you look at the fact that valuations of European banks are lower than those of US banks and that this is a relatively persistent pattern over time, it’s interesting to note that the same actually holds for other industrial sectors. So that brings us to a broader question. What is driving these differences in valuations? Is it market depth? Is it market liquidity? And then there are structural differences and differences in growth potential that come into play.

How much of a concern is commercial real estate?

Typically, real estate, and in particular commercial real estate, is an area which is vulnerable. So that’s why we’ve looked at this very carefully over the past years. We’ve been working very closely with the banks and done some deep dives to understand how banks are exposed to risk in commercial real estate. It is typically a very cyclical business. And now we have this additional layer that, since the pandemic, many people are working from home. This has an impact on the demand for office space in the inner cities and this is having an impact on valuations. Many of the loans given out are bullet loans, so they’re very sensitive to refinancing risk and to changes in interest rates.

Some euro area banks still have large operations in Russia more than two years after the invasion of Ukraine. Is this a source of frustration?

We look at it as prudential supervisors, and what we’ve told the banks early on is that divesting activities in Russia would be very prudent. First, there’s a reputational risk to being active in a country under international sanctions. And there are also issues related to the ability to control and manage risk in such a difficult environment. But it’s not entirely true to say there’s been no change. Actually, there’s been a decline of around 50 per cent in the activities of euro area banks in Russia since the invasion of Ukraine. For the banks that are still there, and I don’t comment on any specific cases, we’ve also given them clear expectations on how we expect a downsizing of activities and exit strategies.

Climate change risks in banks is another focus for the ECB. How close are you to imposing fines on banks that don’t meet your expectations on this?

This area is important for us because we know as societies that climate and environmental risks are very real. When we started working with the banks on these issues in 2019, we noticed that the majority of the banks were not really prepared to assess properly the materiality of credit risk or liquidity risk resulting from transition and physical risk. So this is why we urgently had to do something to bring the banks up to speed. We set various interim deadlines, to give banks time to meet our expectations, but by the end of this year all banks are expected to be in full compliance with all our expectations. And we need to see what comes out of this. Banks that meet our expectations would not face periodic penalty payments, but those that don’t meet them sufficiently would. However, I can’t tell you anything because the final bank-specific deadlines have not expired yet.

Banks say they can’t get enough information from companies to meet the ECB’s expectations. Is that fair?

I think our expectations are realistic. We’re looking at what information the banks could have available. Let’s say the climate certificate of a particular building being used as collateral for a mortgage loan. That is something one can, in most of the countries, easily get at a certain price. It’s not like this information is not available now. I would say that, even in that space, we’re seeing deficiencies, so the banks are not getting the information that they should get in order to assess these risks. You sometimes hear that they can no longer lend to a company that has highly CO2-intensive production. But that is not the case. They could still lend to that company, but they have to account for the risk that is embedded in that contract.

Could the kind of escalation that you’ve talked about in climate change be applied to other areas?

We are using this now on the climate agenda, but it’s definitely something that we can look at in the overall space of becoming more effective and following up on the findings that we have. This is a lesson that we learned from the banking sector stress last year: very often the supervisors knew that there were certain issues, but they didn’t follow them up sufficiently fast. So, in that sense, it’s a general escalation tool that we would use for other issues.

You are in the middle of doing a stress test of cyber risks in banks. How is that going?

We have questionnaires out in the field where we ask about preparedness for an adverse cyber stress event. We’re working closely with the banks to see how prepared they are. We know that cyberattacks have gone up. Most of them are not that critical. But if one gets through, then it could be very bad for the bank in terms of operational resilience and reputational effects.

What about outsourcing of IT by banks? Does this create new risks?

First of all, outsourcing can be an efficient way to reduce costs, and therefore it can be a good business strategy. But of course, our thinking always has to be about what potential risk is associated with it. One is if you outsource to a service provider – and actually many banks do that for their critical services too – then you also need to be able to audit that service provider. Does that service provider have the same risk controls that we would require if the banks did the same service in-house? Then there’s the issue of what happens if there’s a cyberattack on one of those service providers. And what if that service provider is actually working with many banks? Could there be a systemic risk factor? With DORA, the Digital Operations Resilience Act, we have more tools to also conduct this oversight of the service providers.

How much are you using the latest artificial intelligence technology for banking supervision?

The more we can reduce routine tasks and the more we can relieve highly qualified colleagues from things that a machine could do better in terms of simple text analysis or text search, the better it is. So we’re definitely using AI tools. For instance, when doing the preparation for the fit and proper assessments we have to look at each application, and there’s a lot of very detailed information from prospective board members in the banks. So we use a tool which is helping to synthesise this information and sort it. It’s a big relief in terms of time for the colleagues working with this, because, as I said, it can be very labour intensive. But then there is still the human part, the judgement part, of course. These are very critical decisions. We’re using these AI tools to prepare them, but not to take them. That’s where you definitely need human judgement.

And are there concerns about banks increasingly using AI?

They’re using it, for instance, to process credit file applications. And I think it can be extremely efficient to do this. But at the same time, of course, the bank has to make sure that there are no biases in this data so there can be no discriminatory judgement. You have to factor this in when using AI tools. One really needs to understand what is happening there, what is the algorithm on the basis of which the decisions are being made or being prepared. And that’s also something where we’re working with the banks, of course, to understand what they’re doing. The banks have to show to us that they’re not just blindly using tools that may lead to decisions and to risks that we don’t want to see.

You seem to think that banks are not properly assessing these risks that are emerging from recent geopolitical developments. Why is that?

When you look at the expectations that are out there in the markets and the central forecasts from most of the economists, they expect a soft landing. That means the geopolitical risks, and also the changes that are coming up through structural change, the energy transition, demographics and digitalisation, and their effects on economies and on banks, are not being fully priced in by markets. But, of course, these risks will affect banks’ balance sheets and bank risk going forward. So this is why we really need to have a forward-looking approach. Most of the risk models that the banks are using don’t really give us a story about how risks will evolve in the future, because they are based on the past. And we are coming out of a period where we had relatively stable growth. Recent shocks have not really affected the banks’ balance sheets so much because of all the fiscal support. So it is going to be a big challenge for the banks to work with scenarios. What are the relevant scenarios for their credit books?

How will this work in practice?

For instance, when we form our expectations about dividend distributions and what we want to see in this medium-term capital planning, banks need to make realistic assumptions about adverse scenarios in the future. And I think it’s not enough just to look at the overall macro forecast and the GDP forecast; banks have to be more specific. Take, for example, the Red Sea scenario, or sources of fragmentation of global supply chains: how would that affect the specific corporate customers, the sectors to which the bank is exposed?

Is the risk of deindustrialisation in Europe one area that banks are not properly considering?

I don’t necessarily think we are going towards a period of deindustrialisation. But our industrialised regions will definitely look different in the future, depending on the availability of renewable energy in different countries. We will have more relocation of activities; we will have more sectoral relocation. And this is something that, in many of our economies, we haven’t had for a very long time. So the firms have to adjust. The corporate sector has to adjust. And this is something banks have to factor in. Coming out of a period where in many countries corporate insolvencies declined or were stable during the pandemic, banks need to consider how they model credit risk going forward.

Insolvencies and bad loans have been rising. Do you expect this to continue?

It’s hard to forecast, but I think it’s likely that we’re going to see more of this. We’re seeing an uptick in non-performing loans partly because we have higher interest rates. We have this pressure for structural change, and it’s just extremely unlikely that we would have a period of structural change where there’s no increase in defaults. This is why banks need to be resilient and sufficiently capitalised to be able to absorb potential losses.