Second ordinary hearing in 2019 at the European Parliament’s Economic and Monetary Affairs Committee

Introductory statement by Andrea Enria, Chair of the Supervisory Board of the ECB

Brussels, 12 December 2019

It is a pleasure to be back in front of your Committee to discuss the work of ECB Banking Supervision and thus fulfil our accountability and transparency obligations.

ECB Banking Supervision’s recent activities

We continue to make good progress in reducing risks in the banking sector. The capital position of banks under direct ECB supervision has further improved. Banks in the euro area have, on average, also reduced their leverage, and their liquidity positions are now well above minimum legal requirements.

The volume of non-performing loans, or NPLs, held by banks under our direct supervision stood at around €562 billion at the end of June 2019. When the ECB assumed its supervisory responsibilities five years ago, it stood at around €1 trillion. This means that the NPL ratio went down from 8% at the start of ECB Banking Supervision to 3.6% in the second quarter of 2019. When we look in more detail, we see that the NPL reduction has accelerated over the past two years and has been particularly rapid in countries with high levels of NPLs. The fact that this trend started with the publication of the ECB’s NPL Guidance in March 2017 illustrates the impact of supervisory action.

Despite good progress to date, the aggregate level of NPLs in the European banking sector remains elevated by international standards, putting a drag on bank profitability. We consider it essential that more progress is swiftly made on this front while economic conditions are benign. Banks have to maintain high focus in meeting the supervisory expectations about the coverage of the stock of non-performing exposures. We are also working with the national competent authorities on various initiatives aimed at further reducing NPLs held by less significant institutions.

Our targeted review of internal models, or TRIM for short, is nearing completion. Confidence in the robustness of internal models suffered in the wake of the crisis. They were perceived as increasingly complex and giving results that were significantly different across banks. TRIM was launched to check whether the internal models used by banks under the ECB’s direct supervision delivered reliable and comparable results. After four years of intense effort, we recently completed the final phase of the on-site investigations. For banks, TRIM has raised the bar. Institutions are now being asked to remediate the shortcomings identified by our inspectors. TRIM investigations of models for retail and small and medium-sized enterprise portfolios, for instance, concluded with an average of 20 findings on the estimation of the key parameters for credit risk modelling (i.e. probability of default and loss given default). Banks will need to improve how they implement and use their models to address these findings and also meet the new requirements set out by the European Banking Authority. Work in this direction is already under way. For supervisors, TRIM has confirmed that we need to continue investing resources in our ongoing supervision of internal models to maintain high quality and consistent standards going forward.

Looking to the future – the work ahead

Looking forward, fully implementing Basel III will bring the global reform process to a close and contribute significantly to risk reduction. Basel III reforms seek to close the gaps in the regulatory framework that were uncovered by the crisis. It is essential that Basel III is implemented faithfully, consistently and in good time. The EU should not take the lessons of the past lightly; it should not forget the huge impact of banking crises on our economies and on society at large; and it should not renege on its international commitments.

Banks are concerned about the impact of the finalised Basel III rules on their capital requirements. But the new standards do not aim at raising the bar for all banks. They target those banks that have unduly benefited from the use of internal models to calculate their capital requirements. The impact of the new international standards varies widely across banks. So the right question to ask is not whether the aggregate impact is high, but whether such an increase is justified from a prudential and level playing field perspective.

One important element of the upcoming reforms is the introduction of the output floor. The output floor acts as a backstop that ensures that banks using internal models cannot end up with a dramatically lower level of capital than banks that do not use such models. It will complement our supervisory work on TRIM which I mentioned earlier. Acting as a backstop to ensure a minimum level of risk weights across banking groups, the output floor complements other, more risk-sensitive initiatives such as our supervisory reviews. For our international partners, the output floor was essential to achieve an agreement.

When considering the impact of Basel III, we should also consider recent legislative changes that will lead to more lenient requirements. For instance, the Capital Requirements Directive V contains new rules on the quality of capital for Pillar 2 requirements, which will impose a change in the policy the ECB has pursued until now – that of focusing only on Common Equity Tier 1, or CET1 for short. When this change was being negotiated, ECB Banking Supervision and the European Parliament voiced their concern about this change, warning that high quality capital was essential. According to our calculations, the change will generate an average reduction in CET1 requirements of 90 basis points, as banks will be able to rely on lower quality additional Tier 1 and Tier 2 capital, which is now available at favourable conditions.

As supervisors, we can work to prevent any unwarranted consequences of Basel III for bank-specific Pillar 2 requirements. For example, model risks will in future be addressed under Pillar 1. We need to ensure there are no overlapping charges in Pillar 2. Also, we could sterilise the purely arithmetic effect of the increase in risk-weighted assets generated by the output floor in the computation of Pillar 2 charges: if the underlying risks now covered under Pillar 2 are not changed, an increase in capital requirements would not be justified.

But the effects of the new standards cannot, and should not, be fully compensated, as this would jeopardise the prudential objectives of the reform. We are aware of, and concerned about, the low profitability of European banks. But relaxing regulation and deviating from international standards is not the solution. Instead, Banks should focus on their business models, pursue enhanced cost efficiency and take advantage of the opportunities offered by new technologies.

Digitalisation poses significant challenges for banks and supervisors. Fintech firms wishing to engage in banking business must obtain a banking licence. We have published a guide explaining how we will assess banking licence applications from fintech firms, taking into account the specificities of their business models. In May this year, we hosted the first fintech industry dialogue, which focused on the risks and supervisory treatment of artificial intelligence-based credit scoring, robo-advice and cloud computing. It is very difficult to predict exactly how digitalisation will change the business of banking and the structure of the market. We will therefore remain vigilant and continue to engage with banks and industry players in order to adequately tailor our supervisory approach.

Completing banking union and improving crisis management

While progress is being made in reducing risks and finishing the regulatory reforms, we must not forget the need to complete the overall architecture of the banking union.

The crisis management framework still needs to be improved; it remains fragmented and some crisis management tools are missing, or cannot be employed effectively. Concretely, the early intervention framework should be revised in order to simplify the use of such measures and clarify their relationship with other supervisory measures.

Some challenges could be addressed by learning from other jurisdictions. Resolution is a single European process, whereas bank liquidation is governed by 19 different national processes. These national legal frameworks differ greatly, for example on the conditions for starting insolvency proceedings, the ranking of claims and the tools available for managing banks in a crisis. For a truly level playing field in the banking union, we need truly unified liquidation procedures for banks. We could learn from the US Federal Deposit Insurance Corporation when considering new administrative liquidation powers for the Single Resolution Board.

As you are well aware, the European deposit insurance scheme (EDIS), which represents the third pillar of banking union, is still missing. A fully fledged EDIS would ensure that people’s confidence in the safety of their deposits is equally high in all Member States. In addition, it would remove any justification for the rules and policies that are still preventing the full integration of banking business within a single, truly unified domestic market under the banking union. We trust that we can count on the support of the European Parliament in pursuing this crucial objective for our monetary and banking union.

Conclusion

As I conclude, I must of course also mention Brexit. We have continuously stressed that banks need to get operationally ready as soon as possible, and we expect banks to make further progress regarding the implementation of their target operating models in line with the previously agreed timelines.