Address by Governor Patrick Honohan to the IMF-CEPR-Central Bank of Ireland Conference
19 January 2015
Speech
‘Some Lessons Learnt from the EU-IMF Programme’
This occasion is of course an opportunity to mark the successful conclusion of the EU-IMF Programme of assistance to Ireland. I share the assessment that the Programme has been a significant success, without which the wellbeing of the people of Ireland would be very much lower than it is today.
Reading social media and other populist commentary might suggest otherwise, but (without in any way suggesting that economic policy could not have been improved over the past four years) I want to begin by insisting that such commentary is poorly informed. The Programme helped Ireland a lot.
Indeed I would go further to say that, partly due to some good fortune as well as to the sustained adherence to an effective fiscal adjustment programme, Ireland has done much better than it might have – and much better than I and others expected – in turning around a situation, the gravity of which became increasingly evident during 2009 and 2010.
Still, it would be incorrect to insist on unqualified positivity about the Programme, in the context of an overall economic situation in Ireland which has been fraught, looked unpromising, and still leaves a long-lasting residue of high unemployment, and financial distress related to the over-indebtedness of households and firms.
So I would like us to take a few minutes to recall that this started out as a high-risk Programme, why that was so, and how things evolved so that it ended up being a success.
A high-risk Programme
As designed in November 2010, the EU-IMF Programme of assistance was acknowledged to be one that lacked a strong probability of success (in the limited sense of the country’s return to a sustainable fiscal position, recognised as such by private investors). This was the view of the IMF staff,1 and it was also the view of the Irish staff negotiating the programme.
Three elements in particular were lacking. First, the interest rate was too high, considering the starting debt-to-GNP ratio. Second, the high tail-risks associated with the loan portfolio of the banking system were not being addressed with an efficient instrument: an additional debt burden was placed on the Irish State, whereas a tail-risk insurance or direct capital injection from external sources would have been appropriate. Third, the speed of the required bank deleveraging risked imposing substantial fire sales of assets, which would add to the losses of the Irish Government.
Accordingly the risk was high that, at the end of the three years of the programme, the indebtedness of the Government, taking account also of the servicing costs of that indebtedness, would be too high for the government to have access to the market at any reasonable terms, and would give rise to the need for a second programme and a prolonged period of uncertainty.
So why then did Ireland agree to enter the Programme? First, because despite these shortcomings, it nevertheless offered the least painful path for fiscal adjustment: that is, the most gradual path relative to anything else that could reasonably have been expected at that time.2
Essentially, no-one was willing to lend to Ireland any more money to allow a more gradual adjustment. Indeed, a comparison of the Irish Programme with what was called for in other Programme countries reveals that the Irish fiscal adjustment, though effective, was more gradual than most others. That the Troika were prepared to go along with a more gradual path may be attributable to the fact that Ireland entered the Programme with still sizable cash balances, which could be employed alongside the available official funds to smooth the adjustment path. In this way Ireland’s prompt entry into the Programme (not waiting for months until all cash had been exhausted) provided valuable room for manoeuvre.
From the lenders’ point of view, it was thought appropriate to proceed with the loan despite the unpromising calculations, in the light of the likely spillover effects on other member states of an Irish sovereign default.
In short, from the Irish point of view, proceeding with the official borrowing was the best thing to do, even though what was on offer was disappointing. Given the large primary deficit that then prevailed, any alternative would have been much more painful for the Irish people in terms of drastic cuts in services and increases in taxation. What this involved was not merely substituting official for privately-held debt, but also a substantial part of what was being provided was going to pay for further primary (i.e. non-interest) deficits for the full period of the Programme.
If the Programme did fail to restore market confidence, that would have to be dealt with when the time came: holding out for better terms at the outset would have been self-defeating.
As the Programme proceeded, the severity of the three risk factors that I mentioned dissipated. The interest rate was sharply reduced; the tail risk on bank losses did not materialise as much as many feared, and the additional cost of rapid deleveraging turned out to be less severe than it might have been. The two successive Irish Governments’ firm adherence to the Programmed fiscal adjustment was a key element: without it, external official and market confidence in the policy path would not have been restored. (It is worth recalling that fiscal adjustment was already well under way for more than a couple of years when the Programme began.) Close attention to the details of implementation and a solid record of working closely with Troika officials to tweak Programme conditionality in order to avoid self-defeating actions also helped in delivering this outcome. (Examples relating to the banking sector included small but important operational adjustments to the deleveraging path and the timing of stress tests).
While the Programme’s aggregate story thus ends well, it is instructive to consider in some detail how it might have been better designed. I will focus on some of the financial aspects, which, after the fiscal component, were indeed the Programme’s centre-piece.
The bank bond issue
Much has been made of the issue of “burning bondholders”, and rightly so. Just as the scope of the two-year blanket guarantee of September 2008 had been criticised, the idea that the newly-unguaranteed remaining bondholders in the failed banks Anglo and INBS might now receive Government largesse was rightly repugnant to many observers at home and abroad. The Irish government was very interested in seeking some form of bail-in to undo some of the damage caused by the original guarantee, especially if done as part of an international assistance Programme. It was unclear how much burden-sharing could in practice be achieved for the going-concern banks (discussions in the corridors around during the negotiations included suggestions that some form of debt-equity swap might have been achievable). But, for the gone-concern banks being wound down, the question of whether repayment at maturity of these bonds could be avoided, while legally complex,3 was already being closely examined by the Irish authorities quite independently of the Troika.
Pre-emptively, though, the Troika made it clear that no bail-in of senior bondholders could form part of the Programme.4 This must be considered a significant flaw in the Programme design. If the reason was potential spillover effects onto international banking markets from a bail-in, then it would only have been fair, and very much feasible, for the Troika to arrange the difference being made available from other sources. (True, there would probably have been some adverse reputation effect for Ireland from a bail-in, but very limited if it had been part of the Programme design as was envisaged by some Troika officials.)
However, this matter must be kept in perspective. Most of the bonds outstanding from the Irish banks at that time were not in practice bailable because they were either Government guaranteed or asset-covered bonds. The corridor discussions related to a figure of about €16 billion of unguaranteed bonds, of which the two gone-concern banks accounted at stage for less than a third. This is still a lot of money of course, but would not have resulted in any relatively large or noticeable alleviation of the belt-tightening, bearing in mind that each year from 2008 saw an additional fiscal consolidation effort to reach an annual figure of around €30 billion by 2014. I leave it to others to choose how exactly to add up the cumulative total of “austerity” measures since 2009, but it’s a large multiple of what might have been saved by any technically feasible bank creditor bail-in in 2010 or 2011.5
It is worth noting that the new European BRRD legislation which was negotiated in 2012 completely reverses the earlier presumption and explicitly calls for bail-in of unsecured wholesale funding of this type in banks with losses of the scale of Anglo and INBS, in order to reduce the subsequent burden put on taxpayers. This indeed has been a remarkable and welcome turnaround in European policy attitudes, but it came too late for Ireland.6
In the end, of course, the sums injected by the Government into these two banks have been financed at very low interest cost following the liquidation of IBRC and the exchange of the Promissory Notes for long-term bonds. Nevertheless, the pressure that was brought to bear on the Government in this episode continues to rankle in Ireland.
Capital injection into the banks
Not helped by the decision not to include a bondholder bail-in in the Programme, the amount of new government borrowing that the Troika team pencilled-in as being required to put the banks on a more solid capital basis was high.
This was one of the major reasons for doubts as to the likely success of the Programme at the outset. The new and more onerous capital requirements that were set for the banks gave rise to a government cash injection as high as the €35 billion projected by the Troika staff. It was striking how little some of the Troika staff seemed to appreciate that over-indebting the Government in order to recapitalise the banks could be counterproductive. A national banking system will not long be healthy if the Sovereign’s finances are shaky: this has been known for centuries!
In the event, taking account of the March 2011 stress test, and the burden-sharing with – or bail-in of – holders of bank subordinated debt, the ultimate cash injection requirement came out at about a half of the €35 billion. This was the first piece of good news in the working through of the Programme and indications of a lower-than-feared tail-risk.7
Interest rate and maturity
The IMF interest rates are set in relation to the scale of borrowing (by reference to each country’s quota). Where very high borrowing is employed in IMF programmes, its early repayment is encouraged by means of interest rate surcharges, and similar pricing was at first built into this wave of official European lending to distressed euro area member states when the Greek Programme was designed in May 2010. But these “disciplining” surcharges result in interest rates which are dangerously destabilising when applied to high percentages of a country’s GDP. In practice, the least problematic way of dealing with government over-indebtedness on the scale that has been observed in the present crisis is to pass on funds borrowed by official agencies at close to cost, still, of course, as part of a strict fiscal adjustment Programme. This was eventually recognised and implemented from the middle of 2011, with Ireland benefitting significantly from this and from successive extensions of the maturities involved. Examination of the market yields on Irish Government bonds shows clearly that this alleviation of the financial terms of the assistance was a turning point in achieving the restoration of market confidence.
Combined with the fact that tail-risk on banking losses did not materialise on the feared scale, and the success with the liquidation of IBRC, this meant that, provided the Government stuck to its promised fiscal adjustment path8 – which it did – what had been a high risk Programme turned out a success in the end. Thus, the fear that the Programme would have to be extended or renegotiated receded sharply in the course of 2013, so that plans for a possible follow-on precautionary programme could be shelved; a decision welcomed in many quarters as showing that adjustment was possible – at least in the conditions faced by Ireland.
Private sector indebtedness
One aspect which has not been fixed as quickly as had been envisaged and indeed worsened during the course of the Programme is the extent of non-performing bank loans, reflecting the over-indebtedness of many households and firms. Troika and Irish officials shared the view that this debt overhang was holding back the recovery and needed to be dealt with quickly. Why did that not happen?
There is room for different views as to why this situation has persisted. I think that both the Troika and the Irish officials began with the presumption that, once sufficiently capitalised, and when appropriate amendments had been made to the legislation on insolvency and on repossession of the security on defaulted mortgages, the lenders would proceed speedily to deal with the arrears, efficiently making the triage between those loans that should be restructured (potentially with a sizable NPV concession) and those that needed to be foreclosed on. The Programme thus relied mainly on capitalisation and these pieces of legislation. The fact that there were delays in finalising the legislative changes, allowed this belief to persist.
By late 2011 however, the Central Bank realised that most of the banks had not developed sufficiently effective policies or operational capacity to deal with the arrears situation given the still growing number of cases involved. It took more than a year for this challenge to be adequately met. Even then, and with the legislation fixed by early 2013, progress has remained slow – and not just because court proceedings entail substantial delays.
To be sure, although on the official side there has been consistency in wanting the problem to be addressed quickly, there has also been a wide divergence between those who see the problem as chiefly one to be resolved through repossessions and improved payments discipline, and those who argue for greater use of concessionary restructuring of loans.
In my own view, a major factor behind the slow progress has been the banks’ persistent optimism about their ability to recover on the loans (especially the mortgage loans) if they wait long enough. While they have been restructuring loans in a way designed to ensure sustainability as we have defined it, as well as proceeding to legal recovery action as part of their response to the central bank’s Mortgage Arrears Resolution Target process, the banks have certainly not been inclined to err on the side of pessimism in calibrating restructures. Matters are indeed improving for the mortgage arrears situation, with house prices recuperating from the undershooting that followed the collapse of the bubble, but it is arguable that future redefaults on mortgages would be lower with a more liberal approach to restructures. The vigour of the economic recovery would be enhanced by such an approach. However, given the large sums and the large numbers of individual loans involved, it is not clear what additional policy tools could cost-effectively be brought to bear on this problem by the authorities.
Could a programme have been avoided?
Let me conclude with a few words on the interesting question of whether, given the fall-out from the boom-and-bust, and the guarantee of September 2008, a programme could have been avoided. The drip-feed of bad bank-loss news in August and September of 2010 was the main trigger of the loss of financial market confidence. But if the full extent of the losses had been known and made public all in one go (say in March of 2010 when the first NAMA purchases happened) it is likely that this would have immediately triggered the same loss of confidence. In theory the ECB might have done more to reassure bank creditors that, come what may, Irish banks would continue to be provided with sufficient liquidity. But that would have generated a sizable moral hazard. Indeed, the ECB can hardly be faulted as far as actual provision of liquidity is concerned. By the time the Programme began, Eurosystem lending to Irish banks far exceeded Irish annual GDP – an astonishingly large figure: no wonder there was concern in Frankfurt. No, by August 2010, with the markets already jumpy because of Greece, and the fiscal prospects deteriorating, the mounting loan losses removed the small but distinct prospect of avoiding a bail-out that had still remained in the spring.
Thanks to robust interaction and good cooperation, despite the various missteps which inevitably accompanied the initial design of the Programme of assistance, it worked out to deliver, as I have said on a previous occasion, what it said on the tin: restoring market confidence and stabilising not only the public finances but also the economy.
The rest is up to us, the Irish policy makers and the Irish people, to complete the recovery of the economy – which has been on a persistent growth path both in terms of output and employment for the past three years – and to rebuild the economic aspects of Irish society in a way which fully resonates with our shared vision and goals.
It is often observed these days that key among the characteristics needed to prosper in what seems to be a rapidly changing and increasingly unpredictable global economic environment are resilience and the adaptability to take advantage of new opportunities and trends. I believe that Irish resilience has been fully demonstrated in the way in which the economy has been recovering from the disaster that struck in 2008. Ireland's adaptability and openness to the opportunities that emerge in the world have been well demonstrated in the rapid and firmly founded growth of the mid- to late-1990s – the true "Celtic Tiger" period. That is why I believe there are solid reasons for optimism about the future, building on the platform of restored financial stability which the EU-IMF Programme helped us rebuild.
--------------------------------------------------------------------------------
1 Who placed on record that it was “difficult to state categorically” that government debt would, “with a high probability”, “be reduced to sustainable levels in the medium-run”.
2 This is especially true, given the prevailing reluctance in official Europe to rely to a greater extent on Keynesian reasoning which, in the interests of maintaining higher levels of economic activity and employment, might have tolerated a more gradual fiscal correction path in the post-crisis adjustment phase.
3 This is because of the pari passu nature of bond and deposit liabilities of Irish banks at the time (unlike in the US), and because the two banks involved had had to be recapitalised to maintain their access to central bank refinancing.
4 This despite much discussion in the corridors of the negotiations – and, I understand, pre-negotiation discussion of this point between the members of the Troika team of experts.
5 I am speaking here of the situation at November 2010. The overall fiscal and economic costs imposed by the bank failures, by the September 2008 guarantee, and above all by the collapse of the distorted productive and fiscal structures which had been generated by the bank credit excesses, are a different matter.
6 Although in some important respects not well implemented, a version of this principle was applied in Cyprus in 2013.
7 It may be asked why it was necessary to get the consultants BlackRock Solutions to design the models used in the 2011 stress test and why the needed capital had not already been identified. This story has been dealt with in detail in my 2012 lecture on bank recapitalization (reprinted as Honohan, Patrick, “Recapitalization of Failed Banks: Some Lessons from the Irish Experience” Manchester School Vol. 81, No. S1, 2013, pp. 1-15). One aspect not addressed in that paper is why external consultants such as BlackRock had not been used in the 2010 capital adequacy exercise. The answer to that is that several high profile international firms had indeed been engaged by the Irish authorities during 2008 and 2009 with a view to finding out how big the capital hole was. Each of these firms had proposed what soon proved to be absurdly low numbers, an experience which encouraged the more home-grown approach of March 2010. Actually, as noted in Honohan (2012) much of the additional capital called-for in the March 2011 exercise relates to the higher capital ratios and projected added costs of the faster deleveraging required by the Programme. In the event, the deleveraging was achieved with lower costs than projected at March 2011.
8 As it had already been doing for over two years before the Programme.