Address by Governor Patrick Honohan to the David Hume Institute and the Scottish Institute for Research in Economics, Edinburgh 13 November 2012 Speech Insulating the euro from both scepticism and complacency How is it that the euro, having confounded its critics by displaying resilience and tranquillity for a decade, is suddenly plunged by the global financial crisis into a severe structural crisis? The view from Ireland, one of the economies most damaged in the crisis, allows us to see just how release 1.0 of the euro was under-designed, and robust only to moderate shocks. Also evident from Ireland is how the plausibility and credibility gained from the outset by the system resulted in market and official complacency that allowed imbalances of indebtedness to grow too large. The debts reached such a point that, when a crisis of confidence hit, it was likely to have consequences too large to be coped with by conventional tools. Now non-standard measures of monetary policy and intergovernmental lending have been put in place, while the institutional architecture of the euro area is being rapidly overhauled towards a fairly early release date for euro 2.0. Although it is said that Celts are prone to analysing and re-analysing the historical antecedents of present woes; and although I cannot claim to be free of this vice, if vice it be, I will try not to rake over at any length the sorry story of how Ireland’s macroeconomic performance became bogged-down in the first years of the new millennium. Instead I want to speak more about how the euro is involved in this story, and how a way forward is being mapped. Scotland and Ireland in the crisis I know I will be understood in using a bog as a metaphor when speaking to a Scottish audience. Indeed, I am aware that, although Ireland prides itself on the extent and richness of its bogs, raised and blanket, Scottish naturalists claim that you have to go to Caithness and Sutherland to find “arguably the finest blanket bogs on the planet”. In the 1990s, Ireland had stuck to the arduous but dry uplands of growth through competitiveness and fiscal discipline that alone lead safely to sustainably higher income, full employment and steady prosperity. But sometime around the turn of the century, lured by a Will O’ the wisp, or tine ghealáin, or teine biorach, in the form of a property and construction bubble financed by easy access to foreign borrowing, Ireland stumbled into a morass of over-indebtedness from which it is now extricating itself. It is the international context in which the extrication is being engineered on which I want to focus today and in particular to the European context. We, Irish and Scots, live in Europe and depend on our European connections for much of our prosperity. If I may be permitted to consider the economy of Scotland in isolation, I might say that in some respects it depends on European connections even more so than Ireland. Even excluding intra-UK trade, it seems that European countries represent nine of Scotland’s top ten export customers. (For Ireland, the same applies with one difference: China is a bigger customer of Ireland than Norway). Of course the rest of the world matters, and will do increasingly, but proximity counts too, and Europe is the context. This is even more so given the recent stresses which have been felt by the euro area, and which have had their effect worldwide, but more particularly throughout Europe. Scotland is not in the euro, of course, but like Ireland, the economy of Scotland too uses a common currency managed in the context of a larger economic entity. Actually, were it not overshadowed by that of Ireland, the recent economic performance of Scotland, might be characterised in much the same language, if I am to look at the latest GDP figures which show Scotland’s economy to be still 3 per cent below its 2007 peak, with construction – 15 per cent below peak – the weakest sector. The difference is in amplitude: Irish GNP is about 10 per cent below peak and building and construction is down almost two-thirds from the peak. The legacy in terms of over-indebtedness is likewise much more acute in Ireland. I do not ignore the fact that Edinburgh, long a world centre of banking, has experienced some hard knocks in that sphere, but the extraordinary loan-losses experienced by the Irish banking system, whose shareholders lost essentially all of their investment and still left a heavy burden, are in a class of their own, exceeded perhaps only by Iceland. Weighed-down by what may (relatively speaking) be the largest ever banking tab picked-up by a Government, Ireland, after years of apparent fiscal discipline, has recently been somewhat precariously, but nevertheless rather effectively, fighting to restore international confidence in its creditworthiness. Why has the poison of the crisis become concentrated in the euro area? Latvia and Iceland, non-euro countries both, had equally severe economic collapses resulting from credit-fuelled bubbles in the 2000s. Some aspects of euro area membership are helpful in the recovery phase. Thus, while the siren seduction of a devaluation fix for restoring competitiveness is tied down, inflation remains under control and extensive liquidity provision by the central bank met the outflows of Government-guaranteed funds. Yet the crisis has seemed more deep-rooted and intractable in the euro area than elsewhere. Far-reaching corrective action needs to take account of the underlying reasons. I was beaten to the label euro 2.0 by the Vice President of the European Commission, and I think it is undeniable that the transformation of the institutional, legal and policy architecture of the euro area that is under way – and needs to be under way – definitely amounts to a new release, and could not be merely labelled – shall we say – euro 1.2. After all, current data for existing members does not shape up to the prerequisites for admission to the euro in 1998. It is not only Ireland, Greece and Portugal that need to tackle public debt ratios in excess of 60 per cent of GDP and rising; fully two-thirds of the membership, including Germany, has debt levels in excess of that famous, albeit somewhat arbitrary, threshold [Chart 1]. The same with the deficit [Chart 2] and, of course, and most shockingly, long-term interest rates [Chart 3]. Only the spread of inflation rates remains close to being within the Maastricht range [Chart 4] – a tribute of course to it being the target of ECB policy. In a way, that last slide is quite instructive in showing the capacity of the monetary union to achieve what was, after all, set by Treaty as its over-arching goal, namely price stability. As of 2012, then, what were thought of as prerequisites for a successful operation of the euro in its initial guise are hardly present. But why has it come to this? We can answer this from a close-in, blow-by-blow perspective, or, standing back from consideration of the system in the round. Adopting the first perspective, it is by now well understood, at least by close observers, this situation has arisen because of the interaction between financial market sentiment and the behaviour of public and private sector borrowers. Complacent regulatory and fiscal policy design meant that the entire euro area became increasingly vulnerable to a turn in sentiment that would make the servicing of vastly increased debt ratios problematic. Some countries with high public debt at the start of EMU did not bring their debt ratios down: with the newly lowered interest rates, the situation looked manageable. Public debt even in countries with relatively low credit ratings traded during the early 2000s at tight spreads allowing these governments to roll over their debt easily, and to rely on insecure revenue sources. Then, when the global financial crisis hit in 2008, all governments naturally and rightly allowed automatic fiscal stabilisers full rein and many added discretionary stimulus in accordance with most textbooks to help offset the collapse of global demand. Accumulation of private debt was the other dimension of creeping vulnerability which was exposed when the global crisis hit. Liberalised finance generated balance of payments flows on current and private capital account resulting in large cross-border claims and liabilities to an extent which was scarcely monitored, especially within the euro area: indeed it was, for a time, considered politically incorrect in some quarters to think of national balance of payments and other macroeconomic data for individual euro area member states as worthy of analysis. After all, how much attention does Scotland pay to its balance of regional payments within the United Kingdom? Largely unnoticed against a background of huge international flows in the rest of the world, the intra-euro area flows were large. Some of the private international flows within the euro area were, no doubt, eased by the removal of exchange rate risk; but it was a time of vast flows anyway, and not all of these flows were well-used by the under-regulated intermediaries and, in turn, by their borrowers. Banking losses in turn were socialised, adding to the pressure on the public finances in some countries, but especially in Ireland. Perhaps it is too much to ask of my audience tonight to picture themselves in an economy where all of the banks had failed and required recapitalisation in the tens of billions; but then again, perhaps it’s not so hard for you to imagine this. In effect, public and private sectors both largely discounted the risk of a crisis of creditworthiness in the euro area until mid-2010. Since then, though, the reaction of financial markets has been to tighten a vicious feedback loop adding to the pressures both on those countries that had become over-indebted even at the old interest spreads, and on those whose debt had seemed manageable at lower interest rates but now looked less so at the higher rates demanded by the market. [Chart 1 again] Far from trading at spreads that were insensitive to ratings (as had been the case just a few years before), euro area government bond yields became super-sensitive to ratings and other risk factors. Nothing alerted governments to the sharp widening of yield spreads for highly indebted countries from 2010 on: suddenly levels of debt and of public spending that seemed perfectly sustainable at the old levels of risk appetite and the old assessments of credit risk in the euro area, were no longer secure [Chart 5]. The risk of debt restructuring, reinforced by the case of Greece, where sizable losses have been incurred by investors in government bonds, has caused yields on government debt in many parts of the euro area – and as a result most other forms of debt in those countries – to soar. Debt levels that seemed quite manageable at lower yields now look potentially unsustainable to some market participants. In addition, the volatility of market perceptions of credit risks has been contributing to the level of yields. Short-term investors in what are now relatively high-yield bonds face considerable uncertainty about the price of those bonds in the market if they want to sell before maturity. It may very well be that they are caught needing to sell just when the market happens to be in a pessimistic frame: through this mechanism yield volatility translates into high average yields. All in all, the market’s reassessment of the degree to which things can go wrong has generated a potentially lethal combination of high yields and high debt, to the point where the concept of redenomination risk, previously all but unimaginable to most analysts, became factored into market considerations. The commitment device bites back This blow-by-blow account fails, however, to explain why these mechanisms have been so uniquely pernicious in the euro area: after all, all of the world’s economies have been hit by the crisis to some extent or another. In order to understand why the euro has been at the epicentre of the global crisis the key concept to keep in mind is the concept of a commitment device. The euro is a commitment device whereby member countries increased the cost of policy laxity with the intention of obtaining the benefits of enhanced credibility that policy would be disciplined. Unfortunately, the discipline was nevertheless not maintained, and the commitment penalty has been correspondingly severe. Let us recall that the euro was set up after 30 years of debate on how to reduce the disruptive impact of divergent monetary and exchange rate developments, including numerous devaluations, on real trade and investment integration in Europe. It also reflected the chequered experience of many European countries in the 1970s and 1980s with surges of inflation in response to monetary and fiscal indiscipline. The lack of fiscal and monetary credibility resulted in high nominal and real interest rates and recurrent relapses. Was there a way for these countries to acquire the credibility of the Bundesbank? Both of these problems could seem to be solved if a single currency was adopted and put beyond the reach of politicians with their tendency to succumb to a short-term policy perspective. In effect, a non-state currency. Granting full independence from government in a democracy to a central bank meant that its mandate would have to be a focused one. Central banks through history have occasionally acted beyond a narrow inflation targeting remit to address acute financial stability crises in the past (as when the Bank of England famously secured permission in a moment of panic to exceed the limitations of its charter – first in 1847 in the crisis that coincided with the Great Irish Famine). Indeed, the exception that proves the rule was the US Federal Reserve’s passivity which, it is generally accepted, lengthened and deepened the Great Depression of the 1930s. But that experience was not really taken into account in Maastricht Treaty planning. By the time of that planning, deep systemic crises were probably seen as a thing of the past for the advanced economy in part because of improved legal and financial infrastructures and also because, since the time of Keynes, fiscal policy was seen as having the tools to stem and reverse any aggregate demand declines that might occur. Against this background, designing a mechanism which could cope with the moderate fluctuations of the normal business cycle and sporadic fiscal pressures and was sure to eliminate inflationary surges and exchange rate changes seemed safe and even prudent. For the mechanism to work, however, sovereign states had to go beyond declarations of intent. They had to design and adopt a commitment device. As a commitment device (think of Ulysses and the sirens, Alexander or Cortez burning their boats, or medieval rulers exchanging hostages as an earnest of peace), the EMU should have functioned very well indeed. Debt, competitiveness and balance of payments problems were always going to be harder to fix if they emerged in a euro area country. Perhaps that was not fully and constantly brought home to decision-makers. Instead, they seem to have assumed that all problems of macroeconomic imbalance would automatically be self-correcting in the new environment. For example, by removing the possibility of devaluation, membership of the euro ensures that any country which allows its competitiveness to get out of line will ultimately suffer a costly period of adjustment likely associated with unemployment. It was always going to be much more difficult and more painful to readjust aggregate competitiveness. Did that stop countries from losing competitiveness: no! [Chart 6]. Indeed, wide discrepancies in competitiveness quickly opened-up and became progressively worse until 2007. Where one might have expected the financial markets to enforce discipline, they did not always do so. In particular, a market understanding that there really was going to be no bailing-out from the rest of the union of national governments that got into trouble, no transfer union, should have implied that market investors would penalise countries heavily for accumulating debt at an excessive rate. Some market participants did recognise the existence of differential credit risks. Let me return to the chart which I showed on ratings and yields [Chart 5]. The much maligned credit rating agencies did distinguish between euro area sovereigns – albeit only in a relatively narrow range that extended down to A (for Greece). But the range of market yields on government debt did not reflect even this range of credit assessments: in effect real money must have assumed that governments would be bailed out, despite the “no transfer union” dogma. Accordingly some governments were able to spend and borrow without seeing the cost of funds move up against them (though admittedly, the accumulation of excessive debt before the crisis was not as large as is sometimes carelessly asserted.) (Note that the markets were wrong: they did eventually lose significantly on Greek debt). The rapid expansion in cross-border borrowing and lending in the euro area unrelated to Governments was a phenomenon driven as much by global financial liberalisation as by the existence of the euro. It led to a huge accumulation of banking claims. The threat this accumulation presented to the public finances (given the likelihood that Governments would step in to cover at least a part of the losses generated by bank failures, were they to occur), which should have acted as a reason for national Governments to exercise some restraining hand on the exposure, was greatly underestimated and Governments did little or nothing about it. Indeed, by failing to include any significant system-wide banking surveillance in the euro area, the governance structures of the euro area left open a wide gap presenting huge risks. In addition the euro area authorities largely neglected to analyse national balance of payments trends on the mistaken assumption that only euro-area wide aggregates needed to be considered for central banking policy, and indeed the view was that focusing on national macroeconomic aggregates could be a damaging distraction to the formulation of the correct policies for achieving the price stability objectives for the euro area as a whole. So although competitiveness, indebtedness and banking failures were going to be much more difficult to resolve for a country in the euro area, this commitment device did not in practice discipline countries. They continued to operate as if occasional devaluations could be adopted to restore competitiveness, rebalance external payments and reduce the real value of debt. With the low interest rate environment that prevailed due to the removal of market discipline, the imbalances had the potential to grow much larger than they had when markets had feared devaluation. Only when they got into trouble did countries realise how costly fixing the imbalances was now going to be: they had adopted a commitment device and were paying the penalty for ignoring it in their behaviour. The ECB response The deep crisis of the euro area (at first just an aspect of the global financial crisis) clearly required exceptional measures to bring things back onto a stable path and restore efficient market functioning with high levels of employment. The ECB was well-equipped to take some of the necessary steps, even though some of them pushed the frontier of conventional monetary policy. Thus, having already been the first of the major central banks to provide unlimited financing to banks (initially on a one-off basis in August 2007) it made this open-ended provision continuous from October 2008, liberalised its collateral rules and then provided opportunities for extended duration lending at first of 12 months duration (July 2009) and then of 36 months (December 2011). However, as the fiscal response to the initial stages of the crisis expanded deficits and debts, and as the market’s heightened sensitivity to credit risk started to impose high spreads on borrowers across a wide swathe of the euro area the needed policy response became more difficult to implement, especially given the implicit and explicit constraints of the ECB’s mandate. In addition, the distributional impact (for example between countries) of monetary policy actions was beginning to become evident. A regime which had been set up to deal with problems of moderate scale was now being required to deal with a financial and balance sheet crisis of severity unmatched since the 1930s. For some analysts, no attempt should have been made to use central banking tools to address problems not conceived of in the Maastricht Treaty. But for most it was clear that central banking tools did have to be used even if this was not explicitly envisaged in the Treaty. Although the Treaty did wish to preclude certain types of destructive action that some misguided central banks had fallen prey to historically, its drafters did not envisage that the euro area would have less of a capacity to resolve emergent problems of macroeconomic stability. The ECB had to act, but without trespassing on matters that are properly the domain of national governments. This was especially challenging given the potential, as Greece faced the need for a debt restructuring, of central banks’ actions resulting in very evident and indisputable redistributions as between member states. Getting the design of those actions right now presents an acute problem for the central bank. Entirely new issues have presented themselves. Previously there was no great difficulty in ensuring that the desired interest rate levels would be transmitted by market forces more or less uniformly throughout the euro area: that can no longer be taken for granted. The emergence of redenomination risk in market perceptions and its disruptive influence have generated undesired and damaging yield volatility. Additional corrective action, as recently announced, is clearly necessary. In considering and taking such corrective action, the ECB has not only to ensure no compromise with its Treaty mandate and obligation to maintain price stability, but must also avoid the appearance of partiality, and retain the confidence and respect of the body politic. It has been argued by some that the ECB could have moved further sooner; history is of course littered with at least as many unhappy episodes consequential on hyperactive central banks and those which were destructively compliant with the desires of spendthrift governments and imprudent bankers as with those resulting from excessive central bank caution, and it is history that will be the judge of all this. And now… Managing the transition from the emergent bad equilibrium of high debt, high perceived credit risk and risk aversion, and high and volatile yield spreads to a position where confidence has been sufficiently restored is the task of the policy and institutional innovations that form the “what’s new” of euro 2.0. To some extent the new institutions, policies and instruments are designed for a transitional period and to some extent as the framework in which the system is to settle down. Consistent with the analysis that I have proposed, it could be argued that only the transition matters, and that euro 1.0 had a sufficient structure, but was simply not well enough managed in terms of fiscal and especially financial regulation. That is a defensible position, but market and indeed general public confidence in the system has taken such a knock that to insist on it would surely be to tie one’s hands behind one’s back also during the transition. For this reason alone the task must be seen as the creation of a much more robust regime for the long haul. The four main dimensions of reform – relating to banking, budgetary framework, economic efficiency and competitiveness for growth and democratic legitimacy – have been conveniently set out in the report of the four presidents, (of the European Council, the Commission, the Eurogroup and the ECB) which was prepared in advance of the June 2012 Summit. Some of these are well understood, some well advanced, but there is a lot of work to be done and quite a bit of potential for misunderstandings as we go along. As far as the proposals now being fleshed out for a banking union are concerned, it is true that there are subtly different expectations concerning what can and will be achieved, and how. Three components are now envisaged in the banking union, namely a single supervisory mechanism, a common deposit guarantee regime and a common resolution agency. The first of these is in full active preparation; draft legislation is expected to be complete by the end of 2012 with an implementation phase during 2013. The other two would come later. Much is expected from these, and I believe that they are very important elements of euro 2.0. For the single supervisory mechanism, there is both the potential for an operational improvement and – even if there were to be no operational improvement – the potential for greater cross-border confidence in banks to an extent that could improve the access of banks in stressed countries to private and perhaps public debt and equity funding. My work on developing countries long ago convinced me that cross-border involvement of supervisors can help local supervisors achieve the distance and scepticism that can often be needed to detect under-assessed risks in a banking business, especially in the midst of an unsustainable boom. Moving to a single supervisory mechanism will tend to achieve this, the main challenges being how to do so without losing what is equally essential, the nose for local detail, rumour and feel of the market. Achieving the balance, which will require a subtle balance between national and central supervisors, will not be easy but can be developed over time. After all, the euro area is not the only entity covering thousands of banks in which centrally controlled supervision is in place. To what extent is this likely to represent an intrusion into what should properly belong to a national policy arena? I am doubtful that there is any intrusion at all. Safety and soundness of banking is a common goal of bank supervision: non-market goals of public banking policy can be effected through taxation or other means. Assessing creditworthiness and balance sheet risks is not really something that I believe has much room for national exceptionalism. Expectations with regard to the likely roles of a common deposit insurance scheme and of a common bank resolution agency are still diffuse, I feel. Deposit insurance of the first €100,000, currently provided at national level, offers peace of mind to retail users of banking services, and a strengthening of the perceived backstop by pooling the responsibility and resources for payouts could help this, especially at a time when the creditworthiness of some national governments has been called into question. But it is well to understand what a common deposit insurance scheme can and cannot deliver. Deposit insurance does not fully protect banks and banking systems against wholesale bank runs, which are typically much faster and larger than anything generated by retail holders. There is also the danger of divergent expectations with regard to the intended operation of a resolution agency. If greater private sector involvement in burden-sharing of future banking failures is envisaged, and if the single supervisory mechanism is sufficiently proactive and prompt in identifying the need for and taking corrective action, then it will not require or use substantial public funds. But that is not how the recent wave of failures has been dealt with. Can we be sure that enough will have changed in this respect for the future, or will the public purse be called upon again? Perhaps I have ventured enough into the uncertainties of the design and implementation of euro 2.0 to convince you that this makeover is indeed a large and ambitious venture. This is what we are working on. Its accomplishment over the coming years is the sine qua non to re-establishing smooth and effective operation of finance and banking in the euro area, and that in turn is of course a prerequisite of a return to sustained growth, employment and prosperity in all of the regions of Europe.