Address by Jonathan McMahon, Director of Credit Institutions, to the Financial Services Innovation Centre, University College Cork

Thank you to UCC and the Financial Services Innovation Centre for organising today's event. The Central Bank has participated in a number of events at the Centre in recent months, and we are grateful for its role in providing a platform for informed discussion of financial sector issues.

I want to speak today about the Irish banking system. To be specific, I want to discuss what the Central Bank’s recent announcements might mean for individual banks and the system as whole.

My starting place is the 31 March announcements.

In the past fortnight a lot has been written about the stress test, the Minister’s announcement on restructuring, and the prospects for the Irish economy. Given this attention, I think it is important to explain the genesis of the measures announced on 31 March, and also to try and identify what further changes are likely for the banking and credit union sectors in Ireland.

The Prudential Capital Assessment Review (“PCAR”)

The result of the stress test was the core of the 31 March announcement. The anticipation surrounding this exercise grew in the preceding weeks, to a point where it was clearly going to be evaluated as much for what it would reveal about the Irish authorities resolve to tackle the banking crisis, and thus their credibility, as it would the cost of ‘fixing’ the Irish banks. This is hardly surprising. It became clear during 2010 that despite previous attempts to size the losses embedded on Irish banks’ balance sheets, market confidence in banks’ own estimates of capital needs, and indeed official estimates, deteriorated. This in turn impacted the Sovereign’s standing in markets. Restoring some measure of confidence in the likely worst outcome therefore became essential.

Restoration of confidence itself depended on understanding the drivers of the loss of confidence. The central issue appears to have been a marked uncertainty about the direction of asset prices in Ireland. Absent confidence that prices would stop falling, observers expected – not unreasonably - that asset quality would decline. An offsetting factor was the low, by historic standards, carry cost of assets. Yet it was equally clear that the stock of debt was high and that interest rates could not stay low forever. It was further understood that the value of collateral supporting real estate exposures has in most cases fallen, and in some instances markedly. This suggested a problematic path for banks’ exposures: first, that debt servicing problems could increase; and second, albeit more controversially, that a combination of debt service challenges and negative equity might impact borrowers’ appetite to service that debt.¹ As these risks would ultimately manifest themselves in the banking system, confidence in the resilience of Irish banks declined in 2010.

Developments in the third and fourth quarters reinforced the impression that material additional credit risks existed on the balance sheets of Irish banks. Additional NAMA transfers, based on steeper discounts, required more capital to be injected into the banking system and, critically, above that earmarked in the March 2010 stress test. The estimates of loss for Anglo Irish Bank increased. The non-domestic banks active in the retail market did the same. An uptick in Eurosystem borrowing, particularly (the sometimes misunderstood) Emergency Liquidity Assistance, fed a perception of a banking system in deep trouble. Finally, the weakened position of the Sovereign, itself under extreme pressure in bond markets, suggested that the State might not have the resources to continue supporting the banking system.

The need to expose the scale and nature of the possible losses in the banking system therefore became imperative. It also became necessary to do so in a way that would allow the market to reach its own view as to the adequacy of the Central Bank’s assessment of loss. In other words, rather than place an interpretation on the numbers, itself inherently problematic, we concluded that a restoration in confidence would be best served through a policy of transparency. Furthermore, and here the counsel of the IMF was important, the credibility of the exercise could be further enhanced by the pursuit of an independent loan loss assessment. This, of course, led to the appointment of BlackRock Solutions.

Although very obviously the right approach, this did create an immediate practical problem. The PCAR is, of course, a three year stress test. Its purpose is to make sure that banks have adequate capital at the end of three years even after a severe shock. It is a very different thing to try and capture losses beyond the lifetime of a stress test. Yet this is what needed to be done. (And as we now know, sticking to a three year stress ‘horizon’ would have provided a somewhat truncated view of losses.)

The application of existing accounting standards also had to be questioned.

The existing accounting standard, IAS 39, requires that banks recognise the entire anticipated loss associated with an impaired loan in the accounting period in which the loan becomes impaired. Applied literally, this would have left the banks short of required provisions. Fortunately, pending changes to accounting standards, under which banks will have to provision for those losses expected to arise over the life of a portfolio (the expected loss model), provided a rationale for a more realistic approach to provisioning. The addition of the capital buffers allowed us to overcome the limitations the existing standards impose.

Our report on the Financial Measures Programme describes how we married these two exercises: one a three year stress test and the other an assessment of lifetime loan losses.²  This was a complex and involved process, but is illustrated in a very simple way by the fact that a high proportion of the losses BlackRock identified have been brought forward into the period of the stress. This is important for two reasons.

• First, it recognises the reality that many losses will only arise after 2013, reflecting the size of the residential mortgage portfolios.
• Second, it evidences a willingness to work against the grain of some of the assumptions held locally about the path of losses in Ireland.

One challenge to BlackRock’s assessment is that there is little evidence, contemporary or historic, to suggest that Irish banks will foreclose on borrowers in the aggressive way that BlackRock’s model might suggest. There is also little evidence to suggest borrowers will, or indeed can because of personal insolvency laws, simply stop repaying loans. These points should not be dismissed lightly, and for some observers BlackRock’s view of Irish loan losses therefore jars with actual experience. Yet I am certain that a stress test predicated on a well meaning but questionable narrative of local exceptionalism would not have withstood external scrutiny. More importantly, it is simply not possible to dismiss BlackRock’s central conclusion: material loan losses exist even if they are yet to be realised. It is unlikely that this will lead Irish banks to adopt restructuring practices inimical to desirable social and economic outcomes. But neither is the strategy of booting the can down the road any longer an option: there are large credit problems and they have to be addressed.

The Prudential Liquidity Assessment Review (“PLAR”)

The PCAR has re-emphasised the appalling consequences of the weak credit standards which existed during the property boom. These weaknesses would have mattered less if banks had not borrowed expansively to fund the creation of new assets. Similarly, the PLAR emphasises the consequences of weak Asset and Liability Management standards during the same period. We have therefore spent a lot of time thinking about how we strengthen the liquidity regime for Irish banks. We have also devoted considerable resources to analysing how the banks will transition to the Basel III liquidity regime. Together this work comprises our new Prudential Liquidity Assessment Regime.

Where the PCAR is principally concerned with the asset side of the balance sheet, the PLAR is its analogue for the corresponding liabilities. Its creation fills a marked gap in the domestic regulatory regime, and aligns us with the post-crisis international regulatory landscape. Like the PCAR, the PLAR combines quantitative standards with qualitative interventions. For example, the PLAR is the tool we will use to validate the data that banks report to evidence compliance with these ratios. It is the tool we will employ to assess how well banks manage liquidity risks. The PLAR enables us to determine whether the structure of a bank’s liabilities allows it address those risks which arise from the duration of an asset. It is a major addition to our supervisory armoury.

Banking sector reorganisation

The PCAR and PLAR are powerful analytical tools; and the Central Bank’s access to market and non-market information means that we can use them to conduct very fundamental analyse the banks we supervise in some depth. In recent months, the iterative outputs from the PCAR and PLAR have informed work on the reorganisation of individual banks and the system as a whole. Put another way, the PCAR and PLAR have allowed us to assess which structures work and which do not using detailed, current data and with a high measure of analytical rigour.

Our approach to restructuring was initially informed by two related assumptions. First, that the banking system as then structured could not be considered sustainable from a commercial or financial stability perspective: banks’ business models were malfunctioning, there was an excessive reliance on Eurosystem support, and the future path of loan losses looked likely to overrun capital cushions. Second, and for this reason, the credibility of the announcement on 31 March could be enhanced if the public authorities were to address, and be seen to address, these structural issues in one swoop. Achieving some measure of credible reorganisation was therefore always a key objective for the Central Bank.

The main features of the reorganisation are now clear. It is worth reflecting on the extent of change this will bring about, and that which has already occurred. In 2007, there were six domestic banks. Once EBS has merged with AIB, there will be three. Of the non-domestic banks, Ulster, KBC, ACC and National Irish Bank remain important participants, but one major player, Bank of Scotland (Ireland), has exited the market. This is a significant contraction by any measure. Yet it is one that provides a basis for competition and creates a more stable platform for new lending.

The necessity for viable business models

The ultimate test of these reforms will be the banking sector’s ability to generate sustainable profits. Only then can we consider the banking system stable on a key measure: its ability to operate through successive economic cycles without State and monetary authority support. If banks can make profits they will be able to generate the capital they require to invest and create a buffer for future losses. If they can, then private markets are more likely to provide capital or debt. The alternative is that the taxpayer remains the main source of capital for the banks. This would not be a good outcome.

But what might characterise sustainable profitability in this market? It is, after all, only 36 months since Irish banks booked large profits. We now know that these were illusory. In contrast, sustainable future profits will be based on the maintenance of sound business models. But what could this mean in practice?

It is clear that Irish banks will be smaller, have fewer business lines, and will operate predominantly in the Republic of Ireland. (I will return to the remaining UK activities in a moment.) Their main activity will be the origination of loan assets, with the provision of residential mortgages and loans to small and medium sized enterprises likely to form the majority of this lending. Clearly this will not be on the scale of the past; nor can it account for such a high proportion of new lending.

The principal source of funding for Irish banks will be the domestic retail deposit market, with domestic corporate deposits an important additional source. Wholesale funding will be small relative to this domestically sourced pool, and where it exists, it should be longer term.

An emerging feature of the current banking system is the increasing application of fees to customer activity. I would expect this to continue as banks seek to reduce their reliance on net interest margins.

In time, banks will also develop particular business lending niches that we cannot foresee today.

The presence of Irish banks in the UK will be much reduced, but the residual activities still important. The main benefit of access to this market is funding: both AIB and Bank of Ireland will be able to raise deposits through their UK subsidiaries which, in time, will help them to meet funding targets. We expect the amount of new lending in the UK to be limited as the supply of credit to the Irish economy has been identified as the principal priority. Both banks will need to maintain a level of lending in the UK consistent with preserving value in their operations.

Our work during the PCAR suggests that there is a basis for the emergence of stable, profitable institutions, but this will require hard work, and will occur in the context of a system in ongoing receipt of money authority support.

Management of non-core assets

A pre-condition for viability is balance sheet contraction: it is well understood that the banking system outgrew its ‘natural’ funding capacity. There is now no alternative for the banks but to shrink: the funding constraints they face are an irresistible force. The Central Bank’ s loan-to-deposit target of 122.5% for the end of 2013 is the mechanism we have chosen to compel asset disposals, with the consequence that €73bn across the system has been earmarked for disposal.

The restructured banking system has been deigned to incentivise de-leveraging. It has also been set-up to allow it to occur as smoothly as possible. Yet it will be a major operational challenge for the banks to do so as they transition to functioning business models at the same time.

A further challenge is the banks’ need to work-out non-performing loans. Under the split balance sheet structures, non-performing loans will sit in both core and non-core portfolios. History suggests that it can be challenging for banks to resolve a back book while also trying to re-start normal patters on lending. This was one attraction of a clear legal split rather than the virtual model finally adopted. There is, though, abundant international evidence that the virtual split can be made to work; and there are compelling reasons why it is not the right solution for the Irish banking system at present.

The form of the split is, however, a secondary issue. Given the scale of the task, and given also the public resources being committed to the banking system, the extraction of value from non-core assets, and the expeditious clean-up of balance sheets, is now a critical process for Ireland. In other words, now that the future structure of the system is clear, it is time for the banks, under the supervision of the Irish public authorities, to extract as much value as possible from existing loans as balance sheets are cleansed. This will help reduce the ultimate burden on taxpayers. It will allow repayment of creditors. It will support economic activity when a performing loan is refinanced. How Ireland decides to run its non-core banking system is thus a major public policy issue.

Basel III

The repair of the Irish banking system is occurring, of course, while regulatory standards for banks internationally are on the rise. This has been an important consideration during the work on reorganisation: not only does the banking system need to function under today’s rules, it will also have to be able to function under new ones.

The catalyst for this process has been the development of a new Basel Accord. This has prompted banking supervisors and governments to increase banks' levels of equity capitalisation. It is also leading them to implement solutions to meet Basel III's more demanding funding standards. As relevant, though, is how banks are responding to the new requirements with own initiative capital and funding actions. This can be seen in the issuances of capital by banks globally. And although the Basel Committee’s position on capital levels is largely settled, the actions of certain jurisdictions, including Ireland, suggests this process has further to run.

Indeed the Central Bank’s decision to raise the minimum capital target for Irish banks to 10.5% Core Tier One reflects this global drift upwards in the amount of loss absorbing capital on bank balance sheets. Since 2008, and following the March exercise, Irish banks will have received capital injections, predominantly but not exclusively from the State, to the value of 71.7bn.³ And not only has the amount of new capital injected been extraordinary, the high quality composition of that capital is notable. This can be seen most clearly in changes to the capital structures of both AIB and Bank of Ireland. The re-capitalisation process was therefore increasing the quality of capital in the banking system, and thus its capacity to absorb future shocks. It is an oblique achievement, but oblique of course to a calamitously expensive exercise in public sector support for private sector institutions.

Capital is only part of the Basel III story, though.

In 1988, the original Basel Accord set a capital floor for international banks of 8%. This was amended in 2004 as part of the Second Basel Accord. What did not happen during this period was a companion exercise in raising and harmonising liquidity requirements for banks. The Basel Committee’s decision to close this gap has been widely welcomed for this reason.

The two key liquidity ratios under Basel III are the Net Stable Funding Ratio (“NSFR”) and the Liquidity Coverage Ratio (“LCR”). They serve related but different purposes. The NSFR will compel banks to source long term funding equivalent to the duration of longer term assets. This ratio has been designed to increase the stability of funding models by compelling banks to move away from reliance on short term sources of funding. The LCR requires banks to retain highly liquid assets sufficient to cover a liquidity shock lasting up to 30 days. Taken together, the NSFR and LCR will strengthen the prudential regime for banks internationally.

The reality for Irish banks is that convergence with these ratios will take time given the poverty of their starting place. It will also take time because of the global banking systems’ re-financing needs. But convergence will have to occur. These ratios are very likely to enter EU legislation in due course.⁴  Arguably it is important that compliance be achieved sooner rather than later as it is clear that private markets are increasingly using the NSFR, and to a lesser extent the LCR, when judging counterparty risk: in other words, those institutions which will, in time, start to lend again to our banks will use these new regulatory ratios as measures of risk. There is, of course, a potential circularity here: Irish banks will need to fund in international capital markets, particularly long term debt markets, to meet the new ratios, especially the NSFR; but to do so, counterparties will expect to see how well Irish banks are doing against these ratios. For this reason, the Central Bank will track closely Irish banks to make sure that they are doing all that they can to improve the quality of funding sources. The cost of this funding is likely to be prohibitive for the foreseeable future, but it is essential that banks take even the smallest steps to re-order their funding models.

Fixing the infrastructure which supports the banking system

The stress test has reminded us that fixing the banks will take time. There are very basic but important gaps in data to be resolved. For example, many mortgage files do not include details of a borrower’s employment status. There is also a very significant effort required to update and streamline IT systems. All of this deserves management attention, and it something on which the Central Bank will focus its supervisory activities.

An area where significant change is required, and required soon, is within the infrastructure supporting the aggregation and supply of credit information. I have spoken about this subject in the past,⁵ and the Central Bank has done a lot of work to investigate how better credit decisions could follow from a reformed system of information sharing between lenders. The central issue is a misalignment between lending decisions and borrower information. For example, the absence of complete information on a borrower’s existing credit exposures, and therefore that borrower’s debt service capability, is a foundational obstacle to credit risk management.

The Department of Finance led initiative to modernise the credit information regime is therefore very welcome. Other countries have shown what can be done; we must emulate those examples.

Credit Unions

The decisions made towards the resolution of banking system issues in recent months may provide something of a model for changes likely to be required in the credit union sector.

In the past year we have deepened our understanding of the financial position of individual credit unions. Unsurprisingly this has revealed that the tough macroeconomic conditions are impacting credit unions. To date, the Irish League of Credit Unions has played a central role in supporting credit unions facing financial difficulties, therefore offsetting downside risks which would otherwise arise from the insolvency of one or more credit unions.

While ILCU support has been important to the maintenance of the sector’s stability, the Central Bank considers that somewhat more definitive actions will also be required if the sector is to be placed on a stable long term footing.

Earlier this week we met with credit unions and their representatives to discuss the future direction of the sector. This was a highly constructive exchange.

Our starting position is that strong credit unions, operating close to the population they serve, should remain a central part of the Irish financial services landscape. The sector’s hard working volunteers will also need to remain an essential feature of the sector. We therefore see no tension between the preservation of the sector’s character and its reform. In fact, the process of consolidation already well under way tends to support this contention.

Put differently, fewer balance sheets in the sector is not the same thing as fewer credit unions. Indeed, there is an opportunity now for a strong mutual sector, providing financial services to all sections of society, to emerge and replace that extinguished by the banking crisis. This will require leadership and imagination, but there are good international precedents for what can be achieved. The Central Bank is certainly of the firm belief that a strengthened credit union sector, albeit in a different form, can be made a reality, and we will engage constructively with those who share this objective.

Closing remarks

In recent months the Central Bank has, in cooperation with the Department of Finance and the NTMA, sought to put individual banks and the system as a whole onto a more stable footing. The emergence of a self-sufficient banking system will, however, take time. The measures announced in recent weeks will aid this process, but sustained economic growth is the sufficient condition for a full recovery of the banking system. In the meantime, we will continue to address the residual effects of the crisis. Looking to the future, we will further strengthen our supervisory practices. There is much to be done, but we continue the process with some confidence that major steps towards the recovery of the banking system have been taken.

Thank you.

______________________________________________________________________________________________________

[1] http://www.irishtimes.com/newspaper/opinion/2010/1108/1224282865400.html

[2] See a response to this at: http://www.irishtimes.com/newspaper/opinion/2010/1112/1224283148351.html

[3] This is before any offsetting actions, for example liability management exercises.

[4] Banks need to be Basel III compliant by Jan 2018 in the case of the NSFR and Jan 2015 for the LCR.

[5] Address by ADG Financial Institutions Supervision to IBF National Conference