Address by Governor Patrick Honohan to the European Commission DG-ECFIN Conference

Ireland’s EU-IMF Programme: Delivering what it said on the tin

Introduction

The EU-IMF programme of financial support for Ireland, negotiated in November 2010 and with the final tranches of lending being completed about now, delivered what it said on the tin. Amid turbulent market conditions, it provided a safe harbour into which Ireland was able to retreat, in order to clarify its ability and determination to deal with the severe financial problems that had so destructively erupted during the global financial crisis in September 2008. Those problems had their origin in the property bubble that had already begun to deflate a little earlier, and which had not only generated the huge latent banking losses that have been so much discussed, but also incubated severe fiscal and macroeconomic imbalances.

The key to the return of market confidence to the extent that now exists has undoubtedly been rigorous adherence to fiscal goals. Over the three years, a continuation of the momentum of fiscal adjustments already initiated in 2008 has brought the public finances back within striking distance of EU norms. The debt-to-GDP ratio has reached a peak and is on target to fall in the coming year. Economic growth, albeit modest, has returned on a broad front; both full time and part-time employment have been growing for many months now. Residential property prices in the capital have bounced back a little from their lows of two years ago, and have on average been broadly stable in the rest of the country also for some months. Later I will speak a bit about how far the economy is nevertheless away from where we need it to be. But it cannot be denied that, reflecting both policy and general economic conditions, market confidence in Irish creditworthiness is higher than at any time since well before the Greek crisis developments of May 2010.

It was not always obvious that this restoration of market confidence was going to work out. The IMF staff appraisal of the initial programme proposal in December 2010 emphasised that the risks were high. And, after the programme began, the euro area slipped into a second dip recession which had its effect in slowing the Irish recovery. The cumulative change in GDP, consumption and employment over the three years may have been as much as 2 percentage points lower than projected (though GNP did not undershoot by much), and we end with an unemployment rate at around 12½ per cent instead of coming in below 12 per cent as was expected.

Still, compared with the experience of other countries, the macroeconomic and especially the fiscal outturn have been notably close to projection, and the macroeconomic shortfalls seem attributable to the disappointing external factors and not to any miscalculation about the inevitable extent to which the fiscal contraction would dampen the recovery (relative to the infeasible alternative of unchanged fiscal stance).

In addition to fiscal discipline, improved financing terms that emerged in various manners in the course of the programme represented a major contributing factor to the improvement in debt sustainability and in market confidence, enabling the Irish state to fund itself in the coming years.

In my remarks this morning, I will concentrate on the matters where the Central Bank was most closely concerned, namely the broad liquidity, fiscal and debt issues and repair of the banking system. A large number of other policy areas have seen action, consistent with what was set out at the start. Many of these would no doubt have happened in time, given that they had been long on the national policy agenda; some were no doubt accelerated. My overall impression is that most of the specific measures urged on the Government by Troika staff as the programme unfolded were sensible or inevitable; few were really bad ideas. And, as had been foreseen, the Government had a considerable leeway in choosing specific measures to meet the quantitative budgetary targets.

Going into the programme

The contributing factors to Ireland becoming the second euro area country to seek the protection of an international loan from the IMF and European partners included fiscal and banking factors, and a market reappraisal of Europe’s attitude to sovereign bondholder bail-ins.

On the fiscal side the market began -- by the second half of 2010 -- to realise that, despite significant fiscal adjustment since late 2008, the Government’s budget remained widely unbalanced since tax receipts had collapsed in the immediate aftermath of the property bust, and with the additional spending costs related to the associated surge in unemployment. Announced budgetary plans were not going to close this gap.

The scale of banking losses, already acknowledged by April 2010 to imply a net budgetary cost in the tens of billions, also continued to creep up during the summer, especially noticed after the Government finally decided to wind-down Anglo Irish Bank. The degree to which property-backed lending had distorted the banks’ balance sheets meant that placing a credible bound on potential future losses was hard: the potential for tail risks to generate losses that might be unaffordable for the Exchequer to cover could not be convincingly ruled out.

When a huge block of Government-guaranteed banking debt matured in September, the banks required much more central bank refinancing; not surprisingly, the ECB also began to focus on the Irish outlook with increased concern.

Talk of default in many quarters added to market anxiety and an outflow of deposits resumed, with about €100 billion (almost three-quarters of that year’s GNP) leaving in the course of the year, the bulk of it in the last five months, and a good segment financed by emergency liquidity assistance.

With the Deauville agreement (on future EU policy with regard to sovereign creditor haircuts) casting further doubt on the sovereign’s ability to continue to underpin both the continuing part of the banking guarantee and its bond issuance, Irish credit spreads moved out to unsustainable-looking levels.¹

From the combination of all these factors, by early November, it was clear that the protection of an official programme would be needed to enable the Government’s spending programme (which by then had been revised to deliver a convergent path for the public finances) to be maintained.

Changing terms of Government debt in the programme

As initially agreed, the programme disappointed the Irish negotiators in a number of dimensions, especially the rate of interest and the other side’s inability to factor in the banking risks in a way that would break the pernicious link between the sovereign and the banks, a link which continued to inhibit the funding of both. Had the stress test of 2011 obliged the Government to inject as much as €35 billion into the banks (as was pencilled-in by the Troika staff) – more than twice the figure finally struck in March 2011 – the sustainability of the Government’s debt profile would have been even weaker.

As we said publicly and privately at the time, alternative financing approaches, such as an insurance scheme against extreme loan losses, or a direct recapitalisation by a European entity, would have allocated risk more efficiently. But they were ruled-out by the other side, who argued that no mechanism was available to accomplish this. That was certainly the case for the IMF. Arguably, though, it would have been an appropriate time for further European institutional innovation. Actually, had a European entity invested an equity stake, it could have also used its own strong balance sheet to engineer² much lower funding costs of the banks; and it would have had an incentive to do so as it would thereby have generated a sizable additional upside potential to the profits from its equity investment. This opportunity was not taken.

More generally, forgoing -- or at least lacking -- the enhanced risk-sharing that such mechanism would have afforded, the lenders proceeded with a programme which, at the outset, had less favourable debt sustainability than was possible to achieve. Accordingly, the lenders entered into what was in fact a riskier situation for them than necessary, although we can now say that this risk has not materialised.

The interest rate initially charged on the European funds was in part modelled on the IMF lending rate conventions, which envisage a sizable spread over the cost of funds. That is what had been set for Greece in May 2010 and it was presented to the Irish negotiators as non-negotiable. Whereas for developing countries such rates are typically attractive and sustainable given the modest debt ratios that generally prevail, applying them to the levels of indebtedness involved in the European loans was always going to be problematic. All calculations (including those published by the IMF at the programme’s outset), indicated serious sustainability concerns at the terms offered.

I will not review here again the vexed question of guaranteed and unguaranteed bank debt. Suffice it (in the present context) to say that the relevant unguaranteed Irish bank debt that was still outstanding in November 2010 was left to mature, and largely did so before official Europe had finally accepted a more incentive-compatible understanding of how the cost of bank failures should be allocated. The bulk of the Government indebtedness attributable to the bond-holder bailouts has, following the liquidation of IBRC (successor to Anglo Irish Bank), now been folded into a portfolio of very long-term floating-rate notes (issued in place of the non-transferable promissory notes, which would have been unsuitable instruments for the Central Bank to hold).

In the event, the €67.5 billion borrowed from the European and IMF sources almost covered the Government deficit from December 2010 to the end of 2013, of which about one-quarter represented cash bank recapitalisation. There were long-term Government debt repayments also in that period, but these were roughly balanced by new issues. This pattern is seen from the “sources and needs” table summarised in Figure 1.³

Indeed many countries experienced banking failures in 2008 of comparable absolute magnitude to that of the Irish banks. Like Ireland, each of Britain, Germany, the Netherlands, Spain, France and Belgium saw banking failures that required their governments to step in for €50 billion or even more. Given its smaller overall economy, however, such a sum, when combined with the sudden erosion of the tax base, proved to be more than Ireland’s public finances could easily absorb. (Banking losses in Cyprus and Iceland were even larger than those of Ireland in relative terms).

Gradually, Europe began to realise the broad interdependence of member states, especially among euro area members in the banking sphere: poorly performing member economies contributed to heightened systemic risk and slower growth across the entire euro area. The single banking market and the single currency implied such an interdependency and had encouraged policy thinking that focused on the euro area as a whole, and not on individual countries. Awareness of the interdependency led, fairly early on, to a lowering of the interest rate on the official borrowings from Europe and an extension of the maturities. When combined with the lengthy maturity of the floating rate notes issued by the Government in respect of the liquidation of IBRC, these new terms for a large fraction of official indebtedness (amounting to over 50 per cent of GDP) have made all the difference to debt sustainability calculations, both in terms of net present value, and also in terms of refinancing risk.

The Banks: Liquidity

What happened to all the money that flowed in during the 2000s? The answer can be expressed in different ways, and I will be selective here. From one point of view, the money flowed out again: the pension funds, insurance companies, banks, sovereign wealth funds and others who had invested in Irish bank bonds and wholesale deposits were repaid, at first out of borrowings made from the eurosystem, and then increasingly out of the realisation and sale of assets and the repayment of loans made by the banks.

Given all of the emphasis that has been placed on the different elements here, it is perhaps worth looking at the magnitudes. Figures 2 and 3 look at the “Irish headquartered group” of credit institutions which is the most relevant for our purposes. There has been massive downsizing of this category of banks.⁴  (There has also been downsizing of the other banks active in Ireland, but these are less central to the fiscal-banking nexus that has been at the heart of the Irish crisis, so I will not dwell on that here).

A few points are worth noting from these figures. First, the relative importance of bonds and deposits: deposits very much larger at all time periods; bonds⁵ disproportionately invested by foreign concerns. Second, the changing relative importance of foreign business on both the asset and liability side – growing up to the beginning of 2009, shrinking thereafter. Third, the way in which central bank financing was used effectively in the classic lender of last resort function during the crisis.

The Banks: Troubled loans

As was already foreseen at the outset of the programme, repairing the banks is a lengthy process. At first, negotiators on the other side were inclined to wonder why more action had not already been taken. But already by the time the programme had been negotiated, they realised that this was going to take time. As IMF staff put it at that time: “The critically-weakened banking sector can be returned to health only at a calibrated pace.”

Indeed, the textbook first steps: triage the viable banks from the nonviable; recapitalise the former, and resolve the latter; were hampered both by the straitjacket created by the guarantee, and by the potential scale of needed recapitalisation, and its threat to the Sovereign.

This meant no asset fire-sales, and the target, ultimately achieved, of sharply reducing the loan-to-deposit ratio was kept under review, not least to try to prevent the outbreak (frequently threatened) of a destructive deposit price war. On the other hand, for example, the other side’s insistence that deposits should be promptly transferred out of the two fatally damaged banks, Anglo and INBS, actually suited the authorities’ intention to wind these entities down as soon as the guarantee (which had effectively precluded such action) had ended.

Had there been sufficient fiscal headroom, or if the damage had been limited to a segment of the banking system, instead of infecting it all, more drastic de novo approaches to establishing a well-functioning banking system might have been available options.

In the event, even injecting the proposed amount of capital in mid-2011 presented fiscal risks. Although seen as newly flush with capital, the banks still lacked the confidence of the market, which saw the fiscal situation as an over-arching threat to the banks. Paradoxically, the attempt to strengthen the banks by sharply recapitalising was sufficiently credit negative for the sovereign as to limit at first improvement in the banks’ access to the market.

Only after sufficient further consolidation of the fiscal position (and a stabilisation of the wider situation in the euro area) did the market’s assessment of the creditworthiness of both sovereign and banks improve.

That said, other parts of the banking repair have taken much longer than expected. Even with Troika pressure the complex bankruptcy law reforms have come slowly; and on the ground, the mortgage arrears and wider impaired assets problems are only now showing clear signs of coming under control. These represent the major unfinished business as Ireland exits the programme. Progress is being made, and more is needed before the banks can be regarded as restored to fully effective and self-reliant operation. We will not relax in this area.

Conclusion

So where has the Irish economy arrived in macroeconomic terms after three years of the programme? Perhaps the best single picture for illustrating the pattern of macroeconomic stabilisation is Figure 4 which shows that aggregate employment started growing again in 2012 and suggests that this resumes a gradually slowing trend that was in place for more than a decade before interrupted by a construction related surge in the mid-2000s. To those who wish to get back to the favourable and soundly-based economic conditions of the late 1990s, this is probably the most encouraging indicator. There is plenty of scope for disagreement on the quantification, but the pattern is likely to be valid.

At the same time the figure also hints at how far below previous trend and medium term potential the economy remains at this point. Unemployment, albeit moderated by emigration and labour market exit, stands at 12.4 per cent.

In any assessment of the major macroeconomic adjustments that have occurred, the impact on income inequality needs to be taken into account. Actually, we don't yet have the comprehensive survey figures on this for 2012, but, while the available statistics indicate that the downturn has had a broadly similar average percentage effect on incomes across the distribution,⁶ an equal proportionate reduction in incomes of course hits lower income groups harder: there has been a substantial increase in the proportion of poor households suffering deprivation in the years since the crisis broke. A return to higher levels of employment will be a macroeconomic key to reversing this situation.

The crisis will have a lasting unfavourable legacy. The accumulation of debt, public and private, will continue to weigh on growth prospects in a variety of ways. And many households are being affected by long term unemployment. But the damage can be ameliorated by a variety of means, including work on labour market activation and continued improvement of fiscal policy and measures.

Limiting the legacy damage is also the rationale for the Central Bank’s persistence in pressing the banks, in accordance with our mortgage arrears resolution strategy and targets, to accelerate their work to ensure that non-performing loans are brought back into performing status, and dealing with over-indebtedness by moving to sustainable solutions. These are tasks which remain work in progress, though progress that is now accelerating.

In cushioning the impact of the loss of market confidence resulting from the crisis, the programme did no more and no less than was promised on the tin. The rest is up to us.

_____________________________________________________________________________________________________________________________________

¹ Deauville came just after the market had relaxed following the build-up of uncertainty with the big guaranteed bank debt repayments at end-September. To be sure, the post-Deauville impact on spreads was not as large as spread movements that were seen later in the euro crisis, but it was enough to move Ireland very much into the danger zone 

² For example by adding its guarantee to mortgage-backed securities. 

³ Which is based on the NTMA presentation at http://www.ntma.ie/business-areas/funding-and-debt-management/funding-needs-and-sources/ The pie chart excludes changes in cash balances and short-term borrowing; note also that “cash deficit” includes promissory note instalment payments. 

⁴ This downsizing has generally been labelled “deleveraging”, though I prefer not to use that term, as it could equally refer to a situation where total balance sheet size is maintained, but financed with a higher proportion of equity 

⁵ This term includes a wide range of different instruments, such as commercial paper, certificates of deposit, and notes as well as “own-use” bonds issued with a government guarantee and either held as an asset or employed in repo transactions. So there are a number of definitional complexities here. 

⁶ At least if we are to go by the CSO SILC survey results stating that, in 2010-2011, the Gini coefficient was the same or even a little better than the levels that prevailed in the years before the crisis.