Interview with Reuters

Interview with Pedro Machado, Member of the Supervisory Board of the ECB, conducted by Francesco Canepa on 3 March 2026

5 March 2026

What’s the exposure of European banks to the Iran war?

We have been monitoring the direct exposures on both the assets side and the liabilities side very closely and we will continue to do so. The aggregate exposure of significant institutions (i.e. institutions under the direct supervision of the ECB) to Israel and Iran is very modest in terms of both lending and deposit-taking. Exposures to Israel and Iran, to give you a perspective, are overall around 0.7% of Common Equity Tier 1 (CET1) capital on the assets side and 0.6% of CET1 capital on the liabilities side. Even if you include neighbouring countries, the exposures are pretty contained, representing slightly less than 1% of supervised entities’ total assets. There is just one country in the region where there are a few subsidiaries of European banks, and that is Turkey. But we don’t see any sort of stress there for the moment – although the situation may change rapidly.

If energy costs go up in Europe, that has an impact on companies and households. What are your thoughts on that?

There I can only lean on what has already been said by the ECB’s chief economist, Philip Lane. I think he was pretty clear on the fact that this all depends on the breadth and extension of the conflict and the impact that it can have on energy prices. In terms of supply, Europe is not that dependent on that region when it comes to oil and gas. Of course, in the long term, if we have energy prices heating up, we might have an inflation spike with potential recessionary impacts in terms of economic activity. And this translates into a potential impact on unemployment, which is a variable that is quite important for banks.

Do you have any numbers on that?

We don’t. It’s too soon and, by the way, this is not something that is done by supervisors. This would be done by our colleagues on the central bank side. But I think this confirms that banks should assess and embed the possible impact of geopolitical events in their capital planning as well as in their internal stress-testing frameworks. The 2026 geopolitical reverse stress test becomes a case in point.

Will there be a liquidity element too in the stress test?

No, it’s a capital-driven exercise. We are going to ask for complementary liquidity and funding information from the banks, but the exercise remains purely capital-driven.

Dollar liquidity hasn’t been a problem for many years, thanks to the Federal Reserve’s swap lines. But if that liquidity evaporates, European banks that have a large exposure to the dollar may wind up in trouble.

We monitor the foreign exchange exposures of banks in quite some detail. We pay close attention to funding and lending activities in foreign currency, to any concentration of counterparties and instruments, to internal risk limits by currency and tenor, to structural hedging strategies and to collateral management. Then we assess the inclusion of these financial foreign currency shocks in the internal capital adequacy assessment process (ICAAP), and in terms of the overall liquidity metrics and the recovery options. US funding comes very much into this framework of analysis. Of course, if we see that there are significant maturity mismatches in foreign currency activities or even deficiencies in the stress testing framework, then we will, of course, strengthen limits, require the refinement of models and ask for the improvement of contingency planning, which could be a case in point here. These may then be subject to certain remediation decisions if deficiencies are detected. If remediation measures do not follow or prove insufficient, we stand ready to escalate them in terms of intrusiveness and ultimately enforceability.

Do you have any indication that the banks are overexposed to the dollar on any of the dimensions that you mentioned?

We don’t see, at the moment, any stress in terms of US funding activity or in accessing US capital markets.

The European Commission has launched a consultation on bank competitiveness. They ask what’s stopping the banking union from coming to life. What are the main hurdles to a banking union and could “branchification” be a solution?

The lack of a European deposit insurance scheme (EDIS) continues to exacerbate the sovereign-bank nexus, in particular for banks that operate across borders through branches, and to a certain extent promoting ring-fencing practices by national authorities, in particular in the countries where subsidiaries are established.

Branchification could in theory overcome somehow this fragmentation but switching from one national deposit guarantee scheme to another may result in significant costs. For example, the fact that you might not be able to take advantage of the past contributions to the deposit guarantee scheme when you convert subsidiaries to branches might be one reason why we don’t see many of the banks engaging in this so-called branchification. There may also be business model reasons, even historically driven, for banks keeping subsidiary structures in place.

Then if you look at cross-border mortgage lending or lending to small and medium-sized enterprises, it’s practically non-existent. And this has to do with the significant differences that you still have between insolvency laws, collateral enforcement, property laws and property registries. This prevents us from having true cross-border retail lending activity. Cross-border deposit-taking also remains quite residual.

Why has Europe not gone back to the same level of cross-border mergers and acquisitions that we saw before the financial crisis?

The lack of EDIS, of course, plays a significant role. There are also some differences in tax treatments and accounting treatments between the different jurisdictions that might sometimes hamper the ability to carry out mergers on economic terms which are deemed more favourable by the parties.

One of the obstacles to cross-border mergers has come from politics, for example in the case of UniCredit’s investment in Commerzbank. How do you keep politics outside the banking sector?

From our perspective as supervisors, we don’t take into consideration any political dimension of the mergers. We approach any merger based on whether there is a sound case in terms of viability of the business model, whether the risk management is going to remain prudent, and whether the governance is appropriate or needs some enhancement for the combined entity. And of course, we assess the different capital and liquidity positions that have to be in place for the combined entity. In a nutshell, we assess compliance with prudential requirements and the safeguard of effective and prudent risk management.

We know that there are sometime political discussions around these cases. We would expect that any sort of interference is shielded by appropriate governance of banks. That’s why the ECB has been so active in assessing and revising the governance frameworks that are in place for banks since the very inception of European banking supervision.

The ECB has been accused by part of the Italian press of being asleep at the wheel when approving the Mediobanca-Monte dei Paschi operation, because there had been publicly expressed concerns about two actors acting in concert. How does the ECB assess the question of acting in concert? Does it need to take the market authority’s word for it, essentially?

We do not comment on specific cases, and this is no exception. What I can say in general terms is that the joint European Securities and Markets Authority (ESMA) and European Banking Authority (EBA) guidelines provide us with some factors to assess whether certain shareholders are potentially acting in concert. We perform a case-by-case assessment on the basis of that and of the applicable national law, but we do so based on the information that is available to us. Any relevant information that comes from a national authority might eventually be taken into account if it’s deemed relevant.

We assess the information that is publicly available or that comes to us via official channels. We cannot rely on suspicions or speculation regarding potential acting in concert. That said, we would expect that the ownership structures are transparent and the potential relationships between relevant shareholders are duly assessed. But we cannot go beyond what is in our regulatory framework.

Does that mean that you cannot carry out investigations, for example?

We are not prosecutors and the instruments we can use are those specified in the law. We work with the relevant information that we can either gather within our powers or that is communicated to us via official channels. That also applies to qualifying holdings and fit and proper assessments.

You mention fit and proper. The CEO of Monte dei Paschi is under investigation along with some shareholders for an alleged hidden accord to gain control of Mediobanca and Generali. When is a new fit and proper assessment triggered for an existing board member or an existing shareholder?

Again, let me repeat that I will not comment on individual cases. It is however important that there is a clear understanding of how we operate. Let’s start with what the law says: as soon as new facts come to the knowledge of the supervisor that might affect the suitability of a member of the management board, the supervisor is duty bound to reassess their suitability in the light of these facts. You don’t need a judicial decision to trigger that reassessment and eventually to come to the conclusion that these new facts may impair the suitability of the board member. But you do need new facts. These can come from, for instance, a sanction that is applied by the ECB at the end of a sanctioning procedure, where we become aware of behavioural issues. The information can also come, for example, from the findings of on-site inspections. It can come from information that we collect from other authorities, like market and anti-money laundering authorities. It can also come from criminal investigation authorities. Sometimes it comes from the banks themselves.

What ECB sanctions are you thinking about here?

To give you an example, if a board member has actively instructed their teams not to report certain flaws in the reporting system, we then communicate that for a potential suitability reassessment.

But the reassessment is not a sanctioning procedure, it’s a prudential procedure. So it doesn’t have to follow the same due process as a sanctioning procedure notwithstanding the observance of certain procedural guarantees.

Has this happened on your watch?

I cannot give you concrete examples, but it happens.

J.P. Morgan CEO Jamie Dimon recently said blowups in private credit are like cockroaches. If you see one, there are more. Are you seeing many cockroaches from your vantage point?

At the moment, we don’t have any particular evidence of that in the euro area banking landscape. We think that banks overall have evolved pretty well in terms of their risk management across the board. That being said we don’t think we can be complacent, and neither can the banks, regarding potential excessive exposures to non-bank financial institutions. Maybe this statement by Mr Dimon was very much related to the private credit in the United States. We see clearly that this is a less material sector here in the euro area. But there are also other parts of exposures that we should be attentive to.

Are you referring to risk transfers? They lend themselves to becoming a circular game, where the bank lends to the fund that buys the risk.

Precisely. We cannot lose sight of a much more broader macro perspective, which ultimately might endanger financial stability. Euro area banks are becoming very active in synthetic securitisations. We have to remain attentive to whether there is a significant risk transfer with these transactions, so that the risk does not come back to the banking sector – including through the financing of those exposures. This is something that we will also seek to monitor through the applications for synthetic risk transactions. We intend to collect individual information on those transactions to then try to have a much more aggregate view on those, both in terms of volume, but also in terms of potential exposure through the backdoor.

How concerned are you about the state of play in the European securitisation market?

We see a very significant rise in terms of the synthetic securitisation structures. If you compare the first half of 2024 with the first half of 2025, there was an 85% rise. I think this is basically related to the fact that synthetic risk securitisations can be labelled as standardised simple transactions under a change that was introduced to the regulation in 2021. We will have to carefully monitor this in liaison with our colleagues from the financial stability department.

The problem with synthetic risk securitisation rules lies very much in the maturity mismatch of the portfolios that you are insuring, particularly on the mortgage portfolios. This rollover risk, of course, might amplify financial stability concerns in case of unwillingness of investors to renew the credit protection, or if they translate into an exposure of the banks to those insurers in the face of stressful events.

There are several proposals for reforming the Additional Tier 1 bond market. What changes would you advocate?

I think there are some weaknesses that have to do with backward-looking CET1 triggers that may not fire before a non-viability event. There are issues regarding the reliance on discretionary point of non-viability activation. There is always a strong market expectation that calls are redeemed at the first opportunity. And then coupons today tend to be treated de facto as a fixed obligation.

So these are areas where I think it would be worthwhile for the lawmaker to revisit: more forward-looking triggers or more equity-like conversion mechanisms, tighter call conditions, or remuneration structures which are more linked to distributive items.

How far are we from an AI powered SREP?

We are far from it in the sense that I don’t think we will ever have it. SREP is conducted by supervisors, by our teams, and with the full involvement of the middle and senior management. AI tools will be a very supportive element to improve, for example, the analytical capabilities of our teams in different dimensions. They might also be powerful tools to free our teams to concentrate on more material risks and process substantial volumes of findings, and to structure those findings. But they are not a replacement for supervision. This is not going to be conducted by algorithms. It’s going to be conducted by people.

Even from a decision-making point of view, I would very much doubt that it would even be lawful to delegate, by whatever means, supervisory decisions to an algorithm.

Have you seen anything that the banks are doing which goes a little bit too far in that direction?

Banks have to be able to understand how the models work and the outcomes that the models generate. Robust model risk management and governance frameworks need to be in place, together with strong data quality assurance processes. And here is where banks really have to make an investment. We have been consistently pushing banks on data quality issues, yet we still observe a very uneven landscape across the institutions.

I think that these dimensions are critical for AI models to be adopted in a sound way by banks and not to generate prudential concerns, in particular due to improper risk assessment – notably on mispricing of risk or concentration effects – leading to flawed capital calculations for the risks that they are taking