Speech by Ignazio Angeloni, Member of the Supervisory Board of the ECB,
Fact-finding inquiry into the conditions of the Italian banking and financial system and the safeguarding of savings, also with reference to supervision, crisis management and European deposit insurance,
Rome, 3 May 2016
Mr Chairman, Honourable Senators,
It is a pleasure to be here today. I thank you for your interest in the ECB´s perspective on banking supervision. In my previous hearing with you, last June, we discussed the set-up of the new European supervision in general terms. Today, I can be more specific about the banking developments and the activities that the ECB has conducted in its first year of operation as supervisor.
I will first provide some background on how banking supervision functions in the euro area since November 2014, when the Single Supervisory Mechanism (SSM) started to operate. I will then turn to the recent developments in the banking sector, focusing on how Italian banks compare with their European peers, especially regarding non performing loans. This will also give me the opportunity to clarify certain misconceptions regarding the “ECB´s approach” to supervision. I will next comment on the recent initiatives undertaken in Italy in the field of banking. Finally, I will discuss two pending regulatory issues on which concern has recently been expressed: the new crisis management framework and the prudential treatment of sovereign bonds. After that I will be available to answer questions.
1. Single Supervision in the EU
The Single Supervisory Mechanism – which comprises the ECB and the national supervisory authorities of the countries that participate in the banking union – assumed its full duties and responsibilities on 4 November 2014. On that date, the responsibility to supervise “significant” institutions (banks of special relevance because of their size/- and or systemic characteristics, a group that includes the 14 major Italian banks) was conferred on the ECB. Like other national supervisory authorities, the Banca d’Italia is involved in the working teams supervising the banks, the so called Joint Supervisory Teams, but supervisory decisions are taken by the Supervisory Board of the ECB, a body of 25 members in which the Banca d’Italia is represented. In line with the SSM legal framework, the Banca d’Italia is directly responsible for the supervision of the less significant Italian credit institutions (at the end of June 2015, there were 481 of these banks, representing about 20% of the Italian banking sector by balance sheet size). Finally, the ECB is responsible for certain types of decision regarding all SSM banks; for example, authorisation of new banks, or withdrawals of authorisations, and the acquisitions of qualifying holdings.
In accordance with this division of responsibilities, a new framework for accountability has been put in place (Articles 20 and 21 of the SSM Regulation). The ECB is accountable to the European Parliament and to the EU Council of Ministers for its supervisory actions. The ECB also reports to national parliaments, through its Annual Report (which you received at the end of March) and, on request, also through exchanges of views like the one we are holding today. The Banca d’Italia remains accountable, in line with national law, for the supervision of less significant banks and for the other tasks not conferred on the ECB (for example, anti-money laundering and the supervision of certain non-bank entities).
The ECB is also responsible for the proper functioning of the SSM as a whole. This includes ensuring that supervisory activities carried out by the national supervisors within the system are of a consistently high standard, and in particular are in line with regulatory requirements and common supervisory policies. To this end, the ECB develops joint standards and methodologies, of a general nature, also for the supervision of less significant institutions. This power does not extend, however – and let me make this very clear – to individual decisions on less-significant institutions, which remain the competence of the respective national authority
It is important to realise that the new supervisory framework did not automatically produce common supervisory practices; it was rather the premise and the starting point for the ECB to build its own, harmonised supervisory model. This is an area where the SSM has put a lot of effort last year. Let me mention just two major achievements. First, we built a harmonised Supervisory Review and Evaluation Process (SREP) for all banks under direct ECB supervision. The SREP is our methodology for assessing bank risks and for setting the prudential requirements (solvency, liquidity, etc.), applied once every year to each bank. This methodology, whose main points are now published, was used last year for the first time and is being used, with improvements, again in 2016. Second, the ECB has conducted, and is about to conclude, a major project aimed at harmonising the application of European banking rules within the SSM. Within European law, the national supervisors were using many of these rules differently before the establishment of the SSM; in the last year, the ECB has worked closely with the national authorities to eliminate these regulatory asymmetries. A harmonised policy for the exercise of the supervisory “options and discretions” (this is the technical name for the legal provisions that allow for different modes of application) for banks under direct ECB supervision has been approved and is now coming into force.
In the interest of time, I will not discuss these initiatives further. More detail can be obtained on request or by consulting the ECB Banking Supervision internet website.
2. Recent developments in the banking sector and credit markets
In recent years, following policies enacted as a result of the crisis and, in Europe, especially after the launch of the banking union, banks have become more resilient. This is best illustrated by the increase in capital ratios. Common Equity Tier 1 (or CET1) ratios of the significant institutions have increased on average from approximately 9% in early 2012 to above 13% now. Italian significant banks have lagged behind somewhat, but their CET1 ratio has nonetheless increased markedly, from just above 8% to 11.5% on average, over the same period.
As in the euro area as a whole the growth of bank lending in Italy remains modest, but it is gradually recovering. For the euro area, the data from February 2016 show loans to households growing at 2.3% y/y, while bank loans to non-financial corporation are still falling slightly, at -0.2%. In Italy, loans to households grew by 1.4% y/y at the end of February 2016 (against -0.1% a year earlier), whereas for loans to non-financial corporations the figure is -0.2% y/y (-2.6% a year earlier). This subdued credit growth reflects both supply and demand factors. On the positive side, the ECB bank lending survey shows that banks are easing their lending policies, both in the euro area and in Italy.
It is important to note that the strengthening of banks´ balance sheets has not prevented a recovery in credit supply, and indeed of economic growth more broadly. This is true in the euro area and also, to a slightly lesser extent, in Italy. The strengthening of the capital position of banks, therefore, has not resulted in a “credit crunch”, as some had feared. The credit supply mechanism is progressing, though at a slow pace. The expansionary stance of the ECB´s monetary policy, far from being inconsistent with the strategy pursued by the ECB banking supervision, has facilitated the joint achievement of the two results.
The total size of the balance sheets of Italy´s significant credit institutions remained broadly stable in 2015. In Italy, bank assets consist mainly of direct loans and advances; at the end of 2015, loans to non-financial corporations constituted 46% of the total loan portfolio, and loans to households 28%. Compared with other significant institutions in the euro area, Italian banks have a comparatively large share of loans to non-financial corporations in their loan portfolio.
Some tentative progress can also be seen in bank profitability, which has remained for several years at historically low levels. Italian significant institutions, on aggregate, returned to profitability in 2015, after the losses recorded in 2014. However, their return on equity remains low, at 2.7% in 2015, also relative to the euro area average of 4.4%. Again, these are average data: some Italian banks still remained unprofitable. Weak profitability in Italy mainly reflects the severe negative impact of loan-loss provisions for the large stocks of non-performing loans or NPLs; even though impairments have decreased. Relatively high operating costs are also a factor weighing on bank profits.
3. Non-performing loans
Let me now look more closely at the specific issue of non-performing loans (NPLs). In the euro area, progress was made in 2015: the overall NPL ratio decreased by 0.7 percentage points, to 7%, while the coverage ratio, i.e. the ratio of loan loss reserves to NPLs, has increased by 1.3 percentage points to 45.4%. This average, while somewhat comforting, hides very large differences across countries and individual banks.
For the Italian economy, bank asset quality is one of the most acute problems at the moment. For Italian significant institutions, the stock of NPLs reached €274 billion at the end of last year. Including smaller banks, the total stock of NPLs stands at about €360 billion, representing about 18% of total loans in June 2015 (compared to 16.8% a year earlier). In nominal amounts, the volume of NPLs in Italy is by far the highest among the euro area countries. At the end of last year, the level of provisions stood at 45.6% of NPLs, up 0.2 percentage point from a year earlier. The so-called Texas ratio, a measure monitored by analysts that combines NPLs with loss absorbing buffers, is almost double, on average, in Italy relative to the euro area. A mitigating factor consists in the fact that loans are partly covered by collateral, which has the potential to provide an additional buffer. The word “potential” is needed, though, because collateral only provides a buffer if it can be speedily repossessed and liquidated. This is far less the case in Italy than in other European countries.
The incidence of NPLs in Italy is larger for banks characterised by a cooperative structure (NPL ratio of 20.3%) than in other banks (15.5%; figures refer to significant institutions). In the rest of the euro area the opposite is true: 5.2% vs 7.7%. This difference may be due to many factors and one should not generalise. However, it raises awareness on the fact that the regional and local rooting of cooperative banks can, in certain situations, not fulfil its promises and turn instead into a factor of risk.
The high level of bank NPLs in Italy impacts not only on banks but on the whole economy. Banks incur sizeable costs and losses in the process of recovering debts, so borrowing costs for firms and households rises and credit supply is curtailed. High NPL levels hamper bank profitability, discourage investors in bank capital and reduce credit available to new creditworthy firms and households. Ultimately, they reduce economic growth and employment.
Not all NPLs that we observe today are the result of the recent financial and economic crisis. An NPL problem, albeit of a smaller scale, was already present in Italy before 2007. Even then, Italian NPL ratios stood at more than double the euro area average, and oscillaed between 6% and 7% between 2000 and 2006. Part of the NPL problem is linked to corporate sector profitability. According to international comparisons by the IMF, even before 2007 the Italian corporate sector appeared vulnerable due to relatively high leverage and low profitability, and this translated into a persistently high share of bank NPLs. At the same time, relatively weak loan origination standards have contributed to the increase in NPLs, as has the long duration of insolvency proceedings. The ECB itself, in its supervisory dialogues, has identified several shortcomings in the recognition, monitoring and management of non-performing exposures by some Italian banks as well as in the credit origination process. Some of these issues are being tackled by the ECB in the context of a recently launched targeted project on NPLs; I will return to this project in a moment.
It should be clear that no single or rapid solution to this problem exists. Bringing NPLs down to sustainable and internationally comparable levels will require years of consistent effort by the banks and authorities. The initiatives that are needed differ across banks depending on the nature and the seriousness of the problem. Legislation to accelerate insolvency procedures and the liquidation of collateral should be an important part of the solution. The ECB, in its role as banking supervisor, is fully committed to pressing ahead for progress, working with the Banca d’Italia. Banks must, to varying degrees, adopt better internal procedures to recognise NPLs early on and put in place appropriate provisioning, as well as write down or off, without undue delay. Adequate resources should be devoted by banks to this function. Origination standards should be improved. Bank lending decisions should become more objective and merit-based, less reliant on collateral and more focused on the concrete prospects of borrowers’ returns and cash flow.
4. Recent initiatives in Italy
Italian authorities have recently undertaken a number of initiatives to strengthen the governance and financial position of the banking sector. Let me share a few thoughts on these, starting, since we are currently on this topic, with those addressing NPLs.
The ECB welcomes the guarantee scheme for securitisations of NPLs, announced in January 2016. This scheme may lead to a reduction in funding costs for such securitisations. If used efficiently to transfer NPL risks off credit institutions’ balance sheets, it may have a positive, albeit incremental, effect on financial stability. More recently, the newly established Atlante fund is a further step in the right direction. Its mission to intervene both on capital and NPLs rightly recognises the fact that the disposal of bad exposures and the strengthening of the capital position are, in many concrete cases, two sides of the same coin.
We should not, however, expect these two initiatives to deliver a breakthrough on their own. Investor interest in the securitisation scheme as proposed remains to be tested, especially since the gaps between market-based valuation of NPL portfolios and prices expected by banks remain wide. Given its present limited size, the Atlante fund will be able to intervene in only a limited number of medium or small-sized banks. The leadership of the company managing the fund will have to play, as shareholder, a crucial role in strengthening the governance and internal controls of the banks in questions; within the limits of our supervisory mandate, we will follow and support their action. The first operation conducted by the fund, regarding Banca Popolare di Vicenza, will be an important test; if successful it may trigger a virtuous circle and also attract other investors. A further inflow of fresh capital, notably from outside of Italy, would be highly desirable.
These schemes need to be seen in conjunction with other measures that go beyond the sphere of banking. Further steps should be taken to streamline the lengthy court procedures for the enforcement of collateral. Out-of-court procedures, where creditors and debtors reach a solution without resorting to formal legal processes, can be strengthened and their use should be promoted; the measures announced last week by the government go in this direction. This would increase the value of NPLs, allow banks to fully benefit from a large stock of received collateral, and stimulate investors’ interest. A fast resolution of debt is also beneficial for troubled borrowers, who would be able to restructure their businesses and start afresh.
As I mentioned, the ECB has launched a dedicated taskforce on NPLs, with two objectives. The first is to improve the common understanding of the size and composition of NPLs as well as the best practices pursued by banks and supervisors. The second is to develop and implement a consistent supervisory approach to NPL management. In its work, the task force draws on experience and knowledge gained through the 2014 comprehensive assessment, as well as on expertise from other international sources, such as the IMF. We expect to see initial results in the course of this year.
I finally come to the reforms of the popolari and banche di credito cooperativo. As the ECB has already stated in public legal opinions, we consider these reforms to be a very positive step to address the governance shortcomings in the banks concerned, and we support the Italian authorities in this reform effort. In both groups of banks, these reforms should enhance the monitoring powers of shareholders, increase banks’ access to capital, reduce the risk of concentration of power in a minority group of shareholders and provide opportunities for cost synergies. Consolidation of both sectors would help them enhance efficiency and increase profitability, putting them in a better position to support the overall economy.
5. On the ECB’s approach to banking supervision
Let me take this opportunity to address some misconceptions sometimes voiced regarding the ECB approach to banking supervision.
I sometimes hear that we are biased in favour or against certain business models. This is not correct. The SSM approach is neutral with regard to different business models, as long as they are sustainable and do not risk compromising the viability of the institution. The supervisory practices of the SSM, and the SREP methodology in particular, are tailored to the importance and the risk profile of each credit institutions; they are sensitive to the business orientation, but not biased in relation to it. The intensity of the SSM’s supervision varies across credit institutions, being highest for the largest and most systemically important banks, the so-called G-SIBs. These banks often have a higher incidence of market exposure, and correspondingly lower credit exposure, and tend to make wider use of internal models for determining their capital charges. Internal models have merits if they enable a more accurate and risk-sensitive determination of capital requirements, but also entail dangers, such as a potential for bias or even manipulation when calculating risk weights. For this reason, a non-risk-weighted leverage ratio has been introduced internationally, including in European legislation, as a supplementary measure to contain risks stemming from an excessive size of the balance sheet in relation to capital. In Europe, the leverage ratio will become mandatory in 2018; before that date, it is disclosed, in order to enhance market discipline, and monitored by supervisors. Within the limits of the law, the ECB has made a considerable effort, notably in designing its policy on options and discretions, to ensure that the leverage ratio is calculated correctly and without improper exemptions. Based on this metric, Italian banks fare relatively well in the European comparison.
In order to restore the credibility and appropriateness of the Pillar 1 internal models used by significant institutions in the SSM, the Supervisory Board has launched a targeted review of internal models (TRIM) to assess the reliability and comparability of internal rating systems and models. Analyses by the Basel Committee on Banking Supervision and the European Banking Authority have shown that part of the observed variability of risk weights across banks is driven not by differences in the risk of the underlying exposures, but by banks’ modelling choices. Typically, large banks with large market exposures tend to have a lower ratio of risk-weighted assets to total assets. The review started by the ECB aims at identifying, and ultimately at removing, these sources of non-risk-based variability and possible bias in the risk weights. Also the Basel Committee has launched, this year, a review of the market risk framework, to which the ECB is actively contributing.
Furthermore, some observers and bank representatives expressed surprise, last year, in seeing that as a result of our SREP analysis the Pillar 2 capital requirements had been increased, even for banks that had successfully passed the comprehensive assessment a year earlier or that had enacted a capital increase as a result of it. This surprise was unjustified for two reasons. First, the risk assessment by the supervisor is by nature evolutionary, and needs to be constantly updated. Second, the 2014 comprehensive assessment, consisting in an asset quality review and a stress test, was targeted to specific risks. As such, it did not take into account other potential risk factors, such as those deriving from the quality of the governance and risk controls of each bank. The assessment conducted by the ECB in 2015 was broadened to take on board these additional elements. That said, it should also be noted that the increase in the Pillar 2 requirements decided in 2015 by the ECB was very limited, amounting to 30 basis points on aggregate across all significant institutions. After that increase we now consider that, given the overall configuration of the risks affecting the banks under our direct supervision, an appropriate degree of capitalisation has for the moment been reached. This means that, all other things being equal, we do not expect the average capital requirement for the system as a whole to increase further.
This does not, however, exclude the possibility that specific requirements may be imposed on banks in particular situations. One such situation may arise, for example, in case of mergers. Mergers are complex and risky operations, more so in the current, challenging economic and banking environment. It is the ECB’s view that the conditions required for their authorisation should guarantee the best chance of success of the operation, by making the new entity strong from the very beginning. The merger should be supported by an ambitious but realistic business plan, sustainable over the medium term. The governance structure should provide for clear decision-making, empowering the CEO, and for a high degree of professional qualification and independence in the governing bodies. The risk characteristics of the combined entity, and in particular the standards regarding NPLs and their coverage, should be benchmarked with the peer group relevant for the new entity. This may imply, in some situations, the need to raise capital above the amount deriving from the simple sum of the two originating entities.
6. Current regulatory issues
At the beginning of this year, the Bank Resolution and Recovery Directive (BRRD) and the Single Resolution Mechanism (SRM) entered fully into force. This was certainly not unexpected: the BRRD was approved in its definitive form by all EU Member States in May 2014 and was subsequently transposed into national legislation, with a deadline of 1 January 2015, and with full application of the bail-in rules from the beginning of this year. In Italy it was transposed in November 2015.
By implementing in Europe principles established by the Financial Stability Board, the BRRD and the SRM provide a coherent bank crisis management framework including several elements: adequate instruments and powers for supervisory and resolution authorities, transparent burden-sharing arrangements, a single resolution authority in the banking union responsible for significant institutions and cross-border groups (while less significant institutions remain the competence of national resolution authorities – in Italy, the Banca d’Italia) and, finally, a single resolution fund. For each bank, a minimum endowment of capital resources (the minimum requirement for own funds and eligible liabilities, or MREL), set by the resolution authority in line with the BRRD and the European Banking Authority’s regulatory technical standards, will need to be available at all times to ensure that banks have in their own balance sheet sufficient loss-absorbing capacity when resolved. Resolution authorities are required to adopt resolution plans for banks, usable in case of need, and to remove ex ante any legal and operational impediments to resolvability. Since it typically involves the use of collective resources, resolution can only be enacted if a public interest exists, in the sense of the public interest being better protected than in the case of normal insolvency. A condition in the resolution process is that no creditor should lose more than it would in case of liquidation (the “no creditor worse off” principle).
The new framework implies a major shift of the potential burden of bank crises away from the taxpayer and on to the direct stakeholders of each bank — shareholders and creditors. In the loss-making hierarchy envisaged by the BRRD, shareholders and subordinate creditors are first in line, followed by senior bond holders and depositors. Smaller-sized deposits (below €100.000) are protected by the deposit guarantee schemes. Large corporate deposits and senior unsecured bonds are treated “pari passu” by the BRRD.
There is no doubt that such a shift – politically driven by the perception that the crisis incurred major costs for ordinary citizens – will contribute, after proper adjustment of the financial system, to more balanced-burden sharing and to sounder incentives for investors and bank managers. But the transition may be complex and challenging. The new regime requires full awareness by all investors of the characteristics of the instruments and their status under the BRRD. Adequate transparency and financial education are essential. Both the BRRD and the pre-existing State-aid framework provide for safeguards and flexibility to cater for the risks of financial instability and disproportionate effects. The authorities involved (ECB, Single Resolution Board, as well as the competition authority in the European Commission) will need to actively cooperate to prevent and diffuse potential crises. Also for that purpose, ECB Banking Supervision and the Single Resolution Board have recently signed a detailed Memorandum of Understanding. The communication and cooperation between these two authorities is active and well-functioning.
Looking ahead, the experience in applying the new framework in the coming years may provide useful lessons. The periodic reviews of the BRRD will offer opportunities to adjust the legislation, if needed, in the light of those lessons.
A second central issue on the regulatory agenda at present is the completion of the banking union. The ECB firmly believes that the European Deposit Insurance Scheme (EDIS) is a necessary pillar of a well-functioning banking union. In our view, the Commission´s proposal, involving a gradual set-up of the scheme in three phases to be concluded by 2024, is sound because it properly balances clarity of purpose and gradualism. The build-up of a common safety net for deposits using mutualised resources is the logical complement of elevating the responsibility for bank supervision and resolution to the European level and should proceed in parallel with the gradual reduction of banking risks in the system, a process to which ECB Banking Supervision is actively contributing . A common, credible system for deposit protection will underpin confidence, help stabilise the system and contribute to a level playing field.
Recently, the establishment of an EDIS has become part of a broader discussion regarding the timing and sequencing of achieving risk reduction and risk-sharing within the banking union. It has also been linked to changing the prudential treatment of sovereign exposures. In present banking law, bank´s exposures to their domestic sovereigns are exempt both from risk weights and from large exposure limits. The argument behind changing the status-quo is, in essence, that as government bonds are far from being free of credit-risk, as the euro sovereign crisis as well as market valuations amply demonstrate treating them as such is not prudent and sets wrong incentives for banks´ investment decisions, unduly subsidising credit to governments at the expense of private borrowers and ultimately of economic growth.
The ECB recognises that there are indeed good reasons to consider in the future a modification of the prudential treatment of sovereign exposures. Any regulatory change, however, should be mindful of financial risks and should take into account the central role that sovereign debt instruments play in the financial sector and in the monetary policy process. Rigid limits should be avoided, whereas consideration could be given, for example, to more flexible approaches consisting in the application of gradually increasing risk weights to exposures concentrated on individual sovereigns, beyond certain thresholds. The result, ceteris paribus, need not be a decline in the total amount of sovereign bonds held by banks, but rather greater diversification by issuer. In any case, an appropriately long phasing-in period should be foreseen to avoid abrupt effects and be able to adjust in response to experience. The Basel Committee on Banking Supervision has initiated a reflection on the issue, which may lead to the establishment of an international standard. Any changes in Europe should, in our view, take place within that framework.
Let me conclude. The banking union is a paradigm shift for banking supervision and bank crisis management on our continent. Though still incomplete, it is already producing important positive effects, notable in the area of banking supervision, which was the first leg of the banking union to be implemented. From the start, the ECB has worked hard to improve supervisory standards, promote safer and sounder banks, and ensure a level playing field for all banks in all participating countries.
The interaction between the ECB and the national authorities can be a powerful driver of supervisory effectiveness, if, as I have seen happening in many cases already, the in-depth knowledge of individual banks possessed by national supervisors is combined with the independence, clarity of mandate and determination of the ECB.
Much work remains to be done. Our Annual Report, that you have seen, details a number of priorities for the coming years, on which I can provide more detail if needed. Regarding the situation in Italy, we are very encouraged by the awareness of the problems and the willingness to act shown recently by the authorities. A lot will now depend on how and how quickly the measures adopted and announced will be implemented. We are prepared to offer our full support to the Italian authorities in this process.
Thank you for your attention. I am now at your disposal for questions.
The average Texas ratio, defined as total gross NPLs over the sum of capital and provisions, for EA SIs is around 59% but for Italian SIs it amounts to 102%. If the Texas ratios are adjusted further to take into account collateral received from borrowers, they stand at about 44% in the euro area and about 69% in Italy. Higher Texas ratios indicate lower loss-absorbing buffers in relation to NPLs.
Based on IMF and World Bank data.
See IMF , “Technical note on the financial situation of Italian households and non-financial corporations, and risks to the banking system”, 2013.
See, for mortgages, Moccetti, S. and Viviano , E., “Looking behind mortgage delinquencies”, Working Papers, No 999, Banca d´Italia, January 2015..
For some historical data on the duration of bankruptcy proceedings, see Figure 7 in IMF, op. cit.
At international level, the Financial Stability Board has agreed on establishing, for globally systemic banks only, a new requirement for resolution capital called total loss absorbing capacity, or TLAC. TLAC is a similar concept to MREL, but there are some differences regarding scope, denominator, calibration, eligibility of instruments, relationship with capital requirements, and treatment of exposures to eligible instruments, among other things. 2016 will be a crucial year for ensuring consistency between TLAC and MREL, in an implementation process that falls under the responsibility of the SRM.
See for example the statement of G20 leaders in Seoul in November 2010 that “all jurisdictions should undertake the necessary legal reforms to ensure that they have in place a resolution regime which would make feasible the resolution of any financial institution without taxpayer exposure to loss from solvency support while protecting vital economic functions through mechanisms which make it possible for shareholders and unsecured and uninsured creditors to absorb losses in their order of seniority”. In fact, the burden varied greatly depending on country-specific circumstances, in some cases being negligible; see http://ec.europa.eu/competition/state_aid/scoreboard/financial_economic_crisis_aid_en.html.