10 years on – What have we learned? - Deputy Governor Sharon Donnery

It is a great pleasure to be with you today, and I would like to thank the organisers for arranging the event and inviting me to participate on this panel.¹

The Dublin Economics Workshop has always been a key policy forum and September 2018 – ten years on – gives us a timely opportunity to look back and consider not only what we have learned - the Central Bank included – but also, what has been achieved, and critically, what must yet be addressed.

The crisis affected every aspect of Irish society, of the Irish economy, of the Irish banking system. A decade later, significant numbers of people across the country are still dealing with its legacies. Whilst it may be tempting to look at the uniqueness of the Irish experience, taking a wider perspective shows that “the last 400 years has been replete with financial crises, which often followed increases in supplies of credit, greater investor optimism, and more rapid economic growth.”² This is not to minimise the Irish crisis and its effects, merely to stress that history shows us that we must always be on guard against the next potential crisis.

The story of the causes of the Irish crisis has been well told, so I will not repeat it here. ³ The herding behaviour, groupthink, speculative mania, and information gaps, have all been well documented.⁴ Today, I would like to highlight what I see as five key lessons. Whilst not unique to the Irish experience, some are lessons which we in Ireland learned, to devastating effect, the ‘hard way’.

What I will say is that the overall governance, regulatory and supervisory system is designed with lines of responsibility to manage and mitigate risk effectively. In the banking and financial sector, the primary responsibility for risk management in individual financial institutions lies with their executive management and boards of directors. This is supported by internal audit and compliance functions within the banks, and external auditors. Legislators set rules, regulators define and implement the regulatory framework, and supervisors oversee firms’ behaviour, enforce this framework and enforce sanctions in the case of non-compliance.⁵  Each plays a role in the system with its own responsibilities.  

“Systemic financial crises – both in Ireland and elsewhere – are uncommon because they require such a large number of simultaneous institutional and judgmental failures.” ⁶

For central banks and regulatory authorities, and particularly the Central Bank of Ireland, the crisis transformed everything we do: our mandate, our organisation, our frameworks for assessment, our policy instruments, our culture.

So whilst I will not try and be exhaustive today, as I said, I will focus on five key lessons from both a domestic and European perspective.  

Lesson 1: we must act to mitigate the build-up of systemic risk

Before the crisis, Irish banks built up a heavy reliance on funding from wholesale markets and hence were more vulnerable to a sudden stop. The crisis also laid bare the lack of analytical frameworks to assess financial imbalances and the lack of policy tools to mitigate them.⁷ Therefore, following the crisis, the argument for new policy instruments to mitigate risks to financial stability became compelling, and the development of macroprudential policies and instruments followed. ⁸

The ultimate objective of macroprudential policy is to mitigate systemic risk, as major disturbances to the financial system can disrupt the provision of financial services, which can have serious negative consequences for the real economy.⁹

In Ireland, the Central Bank was designated the National Macroprudential Authority in 2014 and the introduction of macroprudential policies by the Central Bank has been a key step for promoting financial stability.¹⁰ These tools aid the Central Bank in building resilience for households and the banking system, such that they can better withstand economic shocks or adverse movements in credit or property prices.

The mortgage rules are now an established element of the macro-financial landscape and have an important structural dimension, with the aim of permanently dampening the pro-cyclicality of credit and house prices (relative to the experience prior to the rules being in place).

The widespread public understanding for the Central Bank’s mortgage measures is positive. We are very conscious that the deposit requirements are a heavy burden on many people trying to build for their future. However, there is I think, a public understanding that as hard as that is, these measures are necessary and better than the alternative. Ultimately, they are just as much about protecting households, as they are about promoting stability.

From a broad cyclical perspective, ensuring resilience is built early in the cycle is most effective – and this has informed the timing and calibration of the mortgage measures and the counter cyclical capital buffer (CCyB). This so-called CCyB is another example of a tool that can build resilience in the banking system to withstand future shocks.  We recently announced an increase in the CCyB to 1 per cent, which becomes effective in July 2019.¹¹ Structural systemic risk, related to the “too-big” or “too complex” to fail phenomenon, is mitigated by the Other Systemically Important Institutions (O-SII) buffer, which has been set at rates up to 1.5 per cent for the relevant institutions, with a phase-in period from 1 July 2019 to 1 July 2021.¹²

One key point to note when looking back at the crisis is that while the use of macroprudential tools may not have averted the crisis a decade ago, the scale of the downturn in Ireland could have been limited and the high cost to citizens reduced. We must therefore continue to guard against complacency and build up resilience across the system.

So in short, a key lesson from the crisis therefore was the importance of developing credible counter-cyclical macroprudential policy and strengthening micro-prudential policy to mitigate systemic risk. This is what the Central Bank has been doing for the last decade.

Lesson 2: all actors in the economy need to build resilience

As a small, particularly open economy, Ireland can be significantly influenced by external events. Ireland has benefitted greatly from our openness. However, the other side of this coin is our vulnerability to external shocks, over which we may have little control. We do not have to look far for evidence of this. Over the past twenty-five years or so growth in employment, incomes, consumption, domestic investment, real estate prices and credit have been more volatile in Ireland than in the euro area, the United Kingdom, Sweden or Denmark.¹³ The key to dealing with this vulnerability is not to retreat from the global marketplace, but to recognise that we need to be prepared for inevitable downturns. In this context, building resilience, whether in banks, at the individual, household, corporate, or national level is key.

All stakeholders have their part to play in this process.

We have seen progress, for example, in household debt statistics – Irish household debt as a proportion of disposable income fell by 49.8 percentage points between Q1 2014 and Q1 2018. However, despite this progress the legacy of the crisis is still apparent, as Irish households continue to be the fourth most indebted in the EU.¹⁴

From a national perspective, fiscal policy is one of the most important policy levers available. One of the trade-offs of joining a monetary union is the loss of an independent monetary policy. This trade-off increases the relative importance of fiscal policy, in addition to macroprudential policy, as a lever that can be used to mitigate country specific shocks in the way that a broad, union wide, monetary policy cannot.

The fiscal position in Ireland in the years after joining the economic and monetary union (EMU) appeared favourable. The debt to GDP ratio continued to trend downward and the budget showed a surplus. However, these headline figures obscured a significant shift in the structure of tax revenues. The share of revenue collected by income taxes fell, while the share for property related taxes such as stamp duty and capital gains tax increased.¹⁵

Debt to GNI* - a measure which avoids the distortions encountered when using GDP - stood at 28 per cent in 2007 and peaked at 158 per cent in 2012.¹⁶ In 2017, debt to GNI* stood at 111 per cent.

The adverse effects this tax structure had in the crisis years, highlights the need for a broad-based and stable tax base, rather than one dependent on windfall revenues. It also exposed that using potentially transitory revenue items to fund regular expenditure can exacerbate deteriorations in the fiscal position during a downturn.

The recovery in the public finances has been dramatic relative to the situation faced during the early crisis years. Significant fiscal consolidation played a large part, as did some favourable external factors, including unexpectedly positive growth and low interest costs. ¹⁷

These factors illustrate the need to strike a cyclically-appropriate fiscal balance to strengthen resilience and build buffers to counteract future shocks.

The characteristics of Ireland, as an extremely open economy with significant trade and financial linkages to the broader international economy, as well as downside risks to the economy, such as Brexit, add to this need.

So, the second key lesson is all actors in the economy need to build resilience, be they individuals, households, corporates, banks or indeed governments.

Lesson 3: in addition to the level of regulatory capital, the type and quality matter

Turning specifically to the banking sector, whilst in Ireland, aggregate capital resources of banks grew from 2000-2007, and banks maintained their regulatory requirements, the composition of capital changed materially.  The proportion of shareholder equity in the covered banks’ capital decreased significantly, with the balance being made up by subordinated loan capital.¹⁸

Ordinary shares, however, were found to be the main effective loss-absorbing instrument on a going concern basis.

In response to the crisis, European legislation was agreed to create a safer financial sector. A key aspect is the establishment of a more robust framework for setting minimum capital requirements. The Capital Requirements Regulation (CRR) and Capital Requirements Directive (CRD IV) were agreed in 2013 and revised the capital requirements for banks.

The framework provides for a Pillar I Requirement, which is the regulatory minimum amount of capital that banks must hold. This is a minimum total Capital Ratio of 8% of their Risk Weighted Assets (RWA). As part of those revisions, the highest quality capital, Common Equity Tier 1 (CET1), typical items of which include ordinary shares and reserves, must now, following various deductions and adjustments, correspond to 4.5% of RWA. Pillar II is an additional capital requirement that may be applied to institutions by supervisors and is based on the Supervisory Review and Evaluation Process (SREP). The institution specific nature of Pillar II ensures that a bank’s capital is tailored to their individual business models and risk profiles. In addition to the Pillar II capital requirement, competent authorities may also set non-binding Pillar II capital guidance that relevant banks will be expected to hold.

The Combined Buffer Requirement is the collective term for the four capital buffers provided for in CRDIV. This includes the aforementioned CCyB and O-SII buffer, as well as the Capital Conservation Buffer and Systemic Risk buffer.

The Combined Buffers act firstly as a mechanism for pro-cyclical accumulation of capital, so that institutions can build a store of supplementary own funds in ‘good times’ in order to protect their regulatory minimum during periods of adverse conditions. If a bank breaches the buffer, automatic safeguards apply to limit the amount of dividends and bonus payments it can make.

The key point here, however, is buffers must be made of CET1, the highest quality capital.

And much progress has been made, since the end of 2014, the fully loaded CET1 ratio of Significant Institutions in Ireland has increased from 9.4 per cent to 17.7 per cent at end-2017.

So in short, the third key lesson from the crisis I want to mention today was that, in addition to the level, the type and quality of capital matter. Progress has been made in many other areas too. My colleague Ed Sibley, Deputy Governor, Prudential Regulation outlined in more detail the specific regulatory and supervisory response, earlier this week. ¹⁹

Lesson 4: some banks will inevitably fail - we just have to be better prepared

The level of public support to the banking system during the crisis in the EU was staggering. In 2009, Member States provided the equivalent of 7.3% of EU GDP, or EUR 906 billion in liquidity support and guarantees to the financial sector. Ireland stands out beyond any other Member State in terms of the relative support it provided following the far-reaching 2008 decision to guarantee all liabilities and deposits. Beyond liquidity and support, Member State governments provided capital support amounting to 4.5% of EU GDP between 2008 and 2012. Ireland, for example spent the equivalent of 37.3% of GDP recapitalising its banks over the period.²⁰

This motivated a fundamental change in how banks and authorities prepare for and manage bank failure. In Ireland, as a direct response to the crisis, we introduced the Credit Institutions Resolution Act in 2011. In a European context, the Bank Recovery and Resolution Directive (BRRD) now provides authorities with common European rules for the recovery and resolution of failing banks. When banks get into difficult financial situations, the supervisor now has a number of early intervention powers at its disposal. The BRRD also requires each bank to prepare a recovery plan which outlines what it will do if it gets into difficulty. The Central Bank of Ireland was designated as the Irish National Resolution Authority in 2015. In this context, the Bank is also part of the Single Resolution Mechanism and we draw up firm-specific resolution plans with the aim of enhancing the resolvability of the firms.

Central to the BRRD is that the authorities have a variety of powers to try and minimise the cost to taxpayers of a bank failure. Critically, the new framework introduces bail-in tools and also aims to ensure that banks have a sufficient stock of bail-inable liabilities. This is known as the minimum requirement for own funds and eligible liabilities (MREL). In short, the objective of this is to ensure that shareholders and creditors bear the brunt of losses in bank failures, thus moving from bail-out to bail-in. Ensuring banks have sufficient and high quality MREL goes a long way to increasing their gone-concern loss absorbing capacity.

So in short, the fourth key lesson is that some banks will inevitably fail, to protect taxpayers we must actively prepare for their failure and ensure that impediments to resolvability are removed. 

Major legislative and institutional changes took place to ensure authorities are equipped with the tools that enhance both the resilience and the resolvability of  institutions, which will be better prepared to deal with, and recover from, a crisis situation. But ensuring the effectiveness of the framework remains a challenge for the years to come.

Lesson 5: the need for cross-border authorities

The final lesson I wish to discuss today is the need for cross-border authorities.

The crisis exposed the weaknesses and heterogeneity in banking supervision across Member States.  Of course, the failures in banking supervision in Ireland have been comprehensively examined and well documented, particularly in the (2010) Honohan Report. ²¹ Some of the conclusions of the report showed that the then ‘Financial Regulator’ relied “excessively on a regulatory philosophy emphasising process over outcomes,” that “a greater degree of intrusiveness and assertiveness” was required on the part of the regulator and there was a need for “additional staff resources and training.” ²²

The crisis also exposed the weaknesses in the framework for resolution – as I just discussed, but also the fundamental mismatch between cross-border banking and domestic crisis management, in addition to the lack of adequate, and harmonised powers authorities had to deal with crises.

Prior to the crisis, the authorities, legal powers, objectives and tools for crisis management and resolution varied considerably across the EU. For example, in some cases the relevant powers were derived from specific resolution or insolvency frameworks for banks, while in others only the general insolvency proceedings were available. In some cases, Member State authorities had the power to transfer assets to a third party purchaser by executive order, in others by judicial proceedings, for example.²³ These differences, coupled with the lack of a harmonised approach to burden-sharing and a weak framework for cooperation (in the form of voluntary cross-border colleges/stability groups) made effective, rapid, coordinated cross-border intervention very challenging.

In a monetary union, where subsidiaries were the main form of cross-border banking holding almost EUR 4 trillion of assets by 2006, the lack of incentives to cooperate and coordinate compounded the lack of trust across the system.²⁴ The well documented collapse in the inter-bank market led to the need for central banks to aggressively step in to provide liquidity in order to prevent the collapse of the financial system.

The response to the crisis was a wholesale change in the legislative framework, policy tools and the institutional architecture for banking supervision and resolution. This culminated most visibly with the establishment of the Single Supervisory Mechanism (SSM) with the ECB in Frankfurt at its centre, and the Single Resolution Mechanism (SRM) with the Brussels based Single Resolution Board (SRB) at its centre.

So in short, the final lesson that became abundantly clear was the need for cross-border authorities. A major institutional overhaul of the architecture of Economic and Monetary Union ensued to complement Banking Union. Key elements have been addressed but the work is not yet complete.

If “stability” is the short-hand we use to describe the objective of these changes, it is important to stress that we seek to safeguard stability precisely because of its importance for citizens. Building on this, the changes I outlined above have been accompanied by a parallel, and interrelated, series of changes to strengthen the consumer protection framework.

European legislation improved protection for consumers in areas such as mortgage credit, payment services, insurance distribution and investments. The Central Bank both contributed to this work and strengthened the domestic framework. This included the enhancement of the Consumer Protection Code, the introduction of the Code of Conduct on Mortgage Arrears, the SME Lending Code and subsequent SME Lending Regulations, and more.

Conclusion

So, is it all sorted then? Clearly no, much remains to be done.

On the institutional side, the review of the European Supervisory Authorities (ESAs) and ongoing debate on the need for a European Deposit Insurance Scheme and a pan-European Anti Money Laundering (AML) authority show the institutional architecture necessary to sustain Banking Union is not yet complete and further work must be undertaken.

Enhancing the MREL targets based on the outcome of resolvability assessments and developing a policy for transfer strategies all remain ongoing initiatives.²⁵ Moreover, the backstop to the Single Resolution Fund needs to be quickly operationalised. Now that important steps have been taken to address bank failure, the authorities will need to turn their attention to other significant players in the financial sector such as financial market infrastructure and insurance firms.

Turning to resilience, debt to GNI* in Ireland is still elevated and remains above 100 per cent. As outlined by the Governor of the Central Bank in a recent speech, given the particularly favourable economic growth in recent years, the projection of a general government deficit in 2019 does not appear to be a sufficiently ambitious fiscal policy. ²⁶

On the macroprudential side, building resilience most effectively requires balancing timing, policy-mix and policy calibration, depending on the nature of the systemic risk being addressed. Technical work is ongoing to assess whether we can strengthen the framework further. ²⁷

However, much progress has been made.

And ten years on, it is important to take stock and acknowledge this.

Banks are now jointly supervised and resolved by cross-border institutions. Banks and authorities are better prepared for bank failures, which should reduce the cost to taxpayers in the future. The level, type and quality of regulatory capital has increased and the loss-absorbing capacity of banks has improved. We now actively intervene to address systemic risk and pro-cyclicality in the system. Also, there is a cognisance to Ireland’s vulnerability as a small open economy and the need for all sectors to build resilience.

We should keep the lessons from the crisis to the forefront of our mind, in our work, in our thinking, and in our communications. Then, our combined efforts to highlight and address risks and build resilience will make the financial system stronger and safer, thereby safeguarding stability and protecting consumers.

Thank you

¹ I would like to thank Mícheál O’Keeffe and Paul Reddan for their contributions to my remarks.

² See Aliber, R.Z. and Kindleberger, C.P., 2015. Manias, panics and crashes. Hoboken, pp. 340.

³ See Lane, P.R., 2011. The Irish Crisis. CEPR Discussion Paper No. 8287.

Also see Fitzgerald, J., 2014. Ireland’s recovery from crisis. CESifo Forum (Vol. 15, No. 2, pp. 8-13).

Also see Whelan, K., 2014. Ireland’s economic crisis: The good, the bad and the ugly. Journal of Macroeconomics, 39, pp.424-440.

⁴ See Nyberg, P., 2011. Misjudging risk: Causes of the systemic banking crisis in Ireland. Ministry of Finance, Dublin, March.

⁵ For an overview, see Lastra, R., 2003. The Governance Structure for Financial Supervision and Regulation in Europe. Columbia Journal of European Law, 10(1), pp.49-68.

⁶ See footnote (iv), pp 5.

⁷ See ‘Lessons from the past, safeguarding stability for the future,’ address by Sharon Donnery, at the CEPR Economic History Symposium 9 June 2016. Available here.

⁸ See ‘Macro-prudential policy: action in the face of uncertainty,’ address by Sharon Donnery, at the Dublin Economic Workshop Annual Economic Policy Conference 24 Sep 2016. Available here.

⁹ See footnote (viii).

¹⁰ See A Macro-Prudential Framework for Ireland 2014.

¹¹ See Central Bank of Ireland Countercyclical capital buffer rate announcement, 5 July 2018.

¹² See Central Bank of Ireland Identification of Systemically Important Irish Institutions and

Announcement of Buffer Rates, 8 December 2015.

¹³ See  Central Bank of Ireland table for Standard Deviation of Annual Growth Rates (%): 1991-2017.

¹⁴ See Central Bank of Ireland Quarterly Financial Accounts – Q1 2018.

¹⁵ See Lane, P., 2010. A new fiscal framework for Ireland. Journal of the Statistical and Social Inquiry Society of Ireland, 39, pp.144-165.

¹⁶ Modified GNI is an indicator that was recommended by the Economic Statistics Review Group and is designed to exclude globalisation effects that are disproportionally impacting the measurement of the size of the Irish economy.

¹⁷ See Smyth, D., 2017. The Recovery in the Public Finances in Ireland following the Financial Crisis. Journal of the Statistical and Social Inquiry Society of Ireland, 46, pp.149-174.

¹⁸ See footnote (iv), pp 41.

¹⁹ See ‘The Banking Crisis – A Decade On,’ address by Ed Sibley, at the Trinity College Dublin “Behind the Headline” Series 12 Sep 2018. Available here.

²⁰European Commission State Aid Scorecard

²¹ See Honohan, P., Donovan, D., Gorecki, P. and Mottiar, R., 2010. The Irish banking crisis: Regulatory and financial stability policy 2003-2008, Report to the Minister for Finance.

²² In 2003, the Irish Financial Services Regulatory Authority (the ‘Financial Regulator’) was established within the overall new structure of the Central Bank and Financial Services Authority of Ireland.

²³ See European Commission 2009.

²⁴ See EU Banking Structures October 2007. The subsidiary number reflects the EU rather than the euro area specifically.

²⁵ See Minimum Requirements for Own Funds and Eligible Liabilities (MREL).

²⁶ See ‘Macro-Financial Policies for the Short Term and the Long Term,’ address by Governor Philip R. Lane, at the MacGill Summer School 27 July 2018.

²⁷ A Systemic Risk Buffer is one potential instrument allowed under CRR/CRD to mitigate non-cyclical systemic risks, such as those related to exposure concentration.