Address by Governor Patrick Honohan to the SUERF Conference

20 September 2010 Speech

Banks and the Budget: Lessons from Europe

My topic today goes to the heart of the financial crisis from which we are emerging. For it is abundantly clear that a breathtaking series of banking failures have, directly or indirectly, been at the root of the budgetary problems now being faced by so many countries, and the impact of the banking crisis on national budgets is set to colour the pace, character and geographical differentiation of the economic recovery generally over the coming years.

But there are several important channels of influence between banks and the budget, some more obvious than others. Because of the scale and pervasiveness of banks, and the speed with which liquidity conditions can change, the impact of banks on the budget can be crucial. At the same time, budgetary decisions including tax and quasi-taxes crucially impact on the banks themselves as do indirect influences through the direction of credit; when banks get into trouble, they are more likely than almost any other sector to be rescued by a government whose interventions are backed up by budgetary resources.

All of this has been dramatically illustrated in recent years, months and indeed weeks. And although the epicentre of the global financial crisis was the US investment banking market, the links between banks and the budget have arguably been recently illustrated in no region more than Europe and in few European countries more than in Ireland.

Two-way channels of influence between banks and the budget – the Irish example

The channels of influence are present on the way up and the way down, and in both directions between banks and the budget.

Let me make this concrete by giving you some flavour of the Irish situation. From 2003 to 2007 the Irish banking system imported funds equivalent to over 50 per cent of GDP to fund a runaway property and construction bubble. The tax revenue generated by the boom came in many forms: capital gains on property, stamp duty on property transactions, value added tax on construction materials and income tax from the extra workers – immigrants from the rest of Europe, from Africa, from China, flooded in as the construction sector alone swelled up to account for about 13 per cent of the numbers at work (about twice the current level, which is closer to what would be normal).

For over a decade the budget was in surplus (averaging over 1½ per cent of GDP) almost every year. No need, it seemed, for restraint in spending, and so, after years of relatively disciplined government budgeting, there was a relaxation of spending controls—one I will say which was broadly welcomed across most of the political spectrum. Alas, that the apparent solidity of the public finances was all a mirage was brutally exposed when global financial confidence collapsed. Already the Irish boom was well over: house prices were falling and the government deficit widening from early 2007. The global recession did its damage to the budget also through the direct effect on world trade and accordingly on economic activity, given the very open nature of the Irish economy. But a more direct effect on the budget was yet to become manifest. The progressive tightening of short-term financial markets during the second half of 2007 and through 2008, peaking in those dramatic weeks after the bankruptcy of Lehman Brothers, eventually exposed the fact that, not only had the banks fuelled an unsustainable property bubble, but they had not safeguarded their own solvency through adequate collateral and guarantees. The Government had little alternative to announcing an extensive guarantee of bank liabilities. This has proved not to be costless, and is imposing a net cost which will place a heavy burden on taxpayers and the users of public services in Ireland for several years -- though, as I have repeatedly stated since the broad magnitudes became clearer in March of this year, it is manageable from the point of view of the national public finances and a good deal less than some alarmist commentary had suggested.

To be specific, every one of the top eight retail lenders – three of them subsidiaries or branches of well-known international banking groups – have been recording loan losses which would eventually eat substantially into their capital were it not for the fact that this is in the process of being replenished. Thus the entire banking sector got caught up in an unreal scenario. Each firm was looking over its shoulder at the others and reluctant both to miss out on a seemingly profitable growth business and fearful of the shareholders’ reaction to declining market share. In an ideal world, the prudential regulator would have called a halt, but – as I have recently documented in my detailed report to the Minister for Finance on the matter – despite some misgivings about the growing exposure, the regulatory apparatus did not fully appreciate the scale and imminence of the risks whether at the systemic or individual bank level and acted only timidly and ineffectually to restrain the lending boom.

Rather than see the banks limp on, crippled by non-performing loans that would both inhibit their ability to retain depositor funding and discourage any new lending, the government already stepped in at the beginning of 2009 to begin the portfolio restructuring of the system.

The approach being adopted is comprehensive and transparent. It is designed to remove the overhang of large hard-to-recover property loans from the books of the banks, freeing their management to concentrate on providing the credit and other services needed for the economic recovery. The banks are being recapitalized – by parents in the case of the foreign-owned banks, through market-sources, backed-up by injections of capital by government where necessary for the rest.

The asset management agency NAMA has been established to purchase the largest property related loans from the banks on an ambitious scale – indeed, relative to the economy NAMA will be among the largest agencies of its type – China providing possibly the closest parallel. The NAMA approach is to buy the loans at market-related prices, with a view to NAMA breaking even or better over time as it recovers on the loans. Despite the haircuts that this inevitably entails, all five of the main locally-controlled banks have agreed to participate in NAMA. The process of valuing and transferring the loans is a painstaking one which is taking longer than everyone hoped, but it is well under way, and the discounts on the initial tranches have been large – over 50 per cent in the first two tranches on average.

Of course this crystallization of losses erodes the capital of the banks, and has given rise to a very visible and transparent recapitalization need which has been addressed very energetically by the Central Bank. (I have heard some suggestion that such up-front transparency could have been avoided by brushing the problem under the carpet in any of a variety of ways. Bank resolution history does not provide much support for such a suggestion. Denial and obfuscation would, I believe, not have been credible anyway, and would have prolonged the crisis and increased its ultimate cost.)

It has been clear to the authorities that the banks would need to be recapitalized to a comfortable and credible level if they were to recover their financial independence and operate without placing additional burdens on the budget. Well ahead of the recent EU-wide stress tests, the Central Bank conducted its own stress tests, which we called the prudential capital adequacy review (PCAR), starting with the two main banks. This entailed a detailed assessment of loan-loss prospects across the non-NAMA part of the portfolio to the end of the current loan-loss cycle, assumed to be 2012, as well as a projection of operating income. As a result of this assessment, which challenged and went further than not only the banks’ existing provisions, but their management’s projections of loan losses, the Central Bank determined for each of the banks an additional capital amount that they would have to raise, one way or another. The additional amount was designed to cover not only the loan losses implied by reading across the first NAMA haircuts to the rest of the non-NAMA book, but also to ensure that the banks would have ample capital to meet the future losses. Thus, on the base case projection – which as mentioned is highly stressed relative to IFRS provisioning and relative to the banks’ own projections – the banks will come through the bottom of the cycle with at least 8 per cent core tier 1 capital (7 per cent equity).

These new capital requirements for Irish banks were announced at the end of March, to be met by the end of the year. Bank of Ireland already satisfies the requirement, having inter alia raised equity capital through a rights issue. AIB, the other large core bank in the system had more to do, and is still working on it, having announced the satisfactory sale of its Polish subsidiary within the last few days. To the extent that AIB may not raise the full amount on its own, the Government has undertaken to meet any deficiency, through the conversion of its preference share stake and if needed with an additional equity injection, presumably using the resources of the National Pension Reserve Fund.

Recapitalization and restructuring of the three smaller institutions is well under way also. Bids have been received from a number of interested parties in EBS, and I expect its future ownership to be resolved quite soon. The disposition of the retail business of Irish Nationwide will follow soon thereafter. The PCAR exercise for the bancassurer ILP has recently been completed and involves only a small additional sum to be raised.

The recapitalization process also benefits the banks’ liquidity position, both in terms of cash raised and by reducing the loan to deposit ratios.

The Government’s decision on Anglo – a bank whose astonishing level of loan losses has been widely aired, and which has damaged the domestic and international reputation of Irish banking – comes as a welcome relief. The proposed structure will provide a secure home for Anglo’s depositors, insulating them from the process of running-off the loan assets over time.

The Government has asked the Central Bank to determine the appropriate levels of capital needed in the two institutions into which Anglo is being split. There are some technicalities involved, and, while I don’t want to anticipate the exact numbers to be published, it may be worth explaining that the type of capital assessment we are doing for Anglo is different from those we have already conducted for the other Irish banks, due to the nature of the split bank structure, the goal of working out assets and of derisking the balance sheet and the prospect of a change in the regulatory status of the asset recovery entity over time. Our analysis of base capital calculations will therefore focus on the amount required to ensure the new structure is capitalized in accordance with current Basel capital requirements in light of existing provisions and immediate expected losses. We will also be publishing a stress assessment of the potential future losses taking account of stress funding costs as well. Thus, recognizing that losses may exceed the base assessment over time, by setting out a stress case we aim to provide as much certainty as is reasonably possible as to the potential exposure presented by Anglo under a severe hypothetical stress scenario in the run off of its book.

I’m not planning to revisit today details of the overall net fiscal cost of the Irish banking crisis. Everyone has known for months that the exposure is large, with a wide, though narrowing, range of uncertainty. Current work seeks to narrow the range as far as possible, though historical experience cautions against expecting great precision. The Anglo base capital and stress calculations will definitely help, and the completion of the NAMA process will also narrow the range – for example with regard to INBS.

Much of the exceptional government financial support for bank restructuring arises this year, and followers of budgetary statistics will see a sizable step jump in Irish government debt, and a sharp transient spike in government borrowing in 2010. Interested market participants are being kept well aware of this, and publication of the exact numbers should not come as a surprise. Actually, though, these exceptional financial injections into banks are smaller than the rest of the government deficit in 2009-10. Just as expansive credit conditions filled the coffers in the good years, so the unwinding of an unsustainable domestic boom is draining the coffers in much the same way and to a similar magnitude.

As many of you will know, the Government has already taken prompt and painful steps to readjust its spending and tax profile, most conspicuously by effectively cutting public sector pay rates. It has kept to the targets it set for net budgetary savings in 2009 and 2010.

I have recently been looking more closely though at the multi-year prospects for the budget. Of course it can be said, if the economy stays close to the track originally envisaged, the deficit would come close to 3 per cent by 2014. But as the IMF and others have noted, the real economy, the price level and also interest rates on Government borrowing, have evolved in a less favorable way. Servicing of the additional debt related to bank restructuring is also a negative factor.

Some explicit reprogramming of the budgetary profile for the coming years is clearly necessary soon if debt dynamics are to be convincingly convergent. Recent movements in the yield spread on Government debt – both for Ireland and for some other countries – readily demonstrate the costs that can result unless international lenders remain convinced that the budget is going to be kept on a convergent path, as indeed the Government is committed to ensuring.

During the 1980s Ireland paid a high price in terms of borrowing costs because the markets feared much steeper exchange rate depreciation than actually occurred. An equilibrium of pessimism, with the economy struggling, and investors requiring a risk premium that imposed additional costs on the taxpayer, displaced what could have been an equilibrium of self-fulfilling optimism. It is important now to re-set the fiscal path to ensure a virtuous cycle of lower borrowing rates contributing to even faster fiscal adjustment and a lower overall cost of the adjustment to society at large.

While there has been an international debate on this matter for larger countries, there seems to me to be no question, for Ireland and for other small financially-stressed sovereigns, but that national growth is best served by ensuring that the public finances are convincingly on a convergent path: the impact on funding costs and confidence surely more than offsets any short-term adverse impact on domestic demand from lower net public spending.

Other recent European examples of banks imposing costs on the budget

There have been several other high profile cases in Europe – Iceland and Latvia most conspicuously – where we have seen the public finances (and the economy generally) crippled by local banking excesses that coincided with, but were not in any close way related to, the contemporary excesses related to the US mortgage market.

These two cases provide a contrast to the Irish one in regard to the impact of bank recapitalisation on the national budget. As Mervyn King has remarked, the current crisis has shown the extent to which, no matter how global banks are in life, in death – or near death experiences – they are local; in this context imposing on the national budget.

As has been thoroughly documented by the report of the inquiry there, Iceland’s banks had made a leap for scale which was not simply based on importing funds that pumped up the local economy. Indeed they greatly outgrew the local economy (which after all is very small, with a national population of less than half a million persons) and were heavily involved in the financing of international investment ventures of some of their Iceland-based customers. When these banks – middle sized in an international comparison – failed, the national authorities regarded the burden of meeting all of their liabilities as being too great to be assumed, and their rescue and restructuring of these entities was partial only, with much of the foreign liabilities not covered, a matter which led to a fraught international negotiation, especially with the authorities in the UK and Netherlands.

The Iceland case probes the limits of the impact that banking system rescues can have on the budget. Indeed the general issue of limiting the scope of bailouts has come under increasing international scrutiny with academics and policymakers trying to design viable ways of reducing the number of banks that can be considered too big, too complex, or too interrelated with the rest of the system, to be allowed to fail with losses to uninsured creditors. Policies designed to limit the size, remove the complexity, and pre-design an orderly liquidation scheme (living will) are all being considered as solutions in this direction. There is much to be said for this line of policy exploration; especially when it seems that some of the institutions that got into trouble globally have become too complex to be reliably managed. But I have to say that I am not altogether optimistic that there will quickly be a full solution that both preserves the effectiveness and cost efficiencies of modern finance and genuinely removes the need for rescues in all conditions.

Latvia’s reliance on foreign-owned banks points to the opposite end of the budgetary impact of financial globalization. Although there was a significant local bank failure, much of the actual and prospective loan losses in the Latvian downturn are being borne by the foreign-owned banks that made the loans (the bulk of them denominated in foreign currency, a factor which certainly complicates the assessment of risks and narrows the range of policy options).

These contrasts show the uneven degree of financial globalisation in the European Union (and the European Economic Area, which embraces Iceland). The banking system in some countries, mostly smaller countries in Central and Eastern Europe, is dominated by foreign-owned banks with headquarters elsewhere in the EEA. The direct recapitalisation costs of banking failures in those countries are exported back to the home countries, providing a kind of automatic insurance to the host country. (The primary supervisory responsibility also generally lies with the home country, though the geographical match of risk and regulation is certainly not perfect.) Most EEA countries rely principally, though, on locally-owned and controlled banks. Indeed, banks that have got into trouble and required rescuing have been shrinking back into their home countries, partly in order to help them meet regulatory capital requirements and partly because the European Commission has been taking a jaundiced view of state-aid being used to bolster geographically over-extended business models. The old vision of a seamless European banking market where, regardless of their home country within the EEA, the most cost-effective providers would deliver the financial services needed by firms and households may never have been wholly realistic, at least when it comes to retail services. But this vision has certainly suffered a setback from the Iceland case, to the extent that the regional ambitions of any bank may in practice be limited by the perceived fiscal capacity of its home government.

I have so far dwelt on the problems posed for government budgets by failing banks. The country examples I have chosen are not, of course, the only ones affected. Belgium, Germany, the Netherlands, Spain, Switzerland and the United Kingdom are other European countries which have seen the need for public injections of funds in one form or another. In some cases (such as Switzerland) the main source of the problem was the participation of banks in the tainted securitisations coming from the US; in other cases national bubbles were also present.

From budget to banks

But there are also pressures in the opposite direction, i.e. from budget to banks, which have only now come to the fore for the first time in advanced economies (though this was long an important dimension of the banking crises suffered by many developing countries in the 70s, 80s and 90s).

Two aspects have so far been central here, one of which relates to the pressure that a fiscally weak sovereign places on the funding costs of its home banks; the other to pressures that can be placed on banks in other jurisdictions who have or are thought to have, exposures to those sovereigns.

The first problem has been affecting bank lending rates across much of Europe, and is most evident in the euro area, where these effects are unrelated to currency differentials. The sharp jump in sovereign spreads in the first half of May, 2010 is only the most dramatic illustration of the degree to which these pressures matter. The tensions were transmitted immediately and fully to bank funding conditions affecting both cost and availability of wholesale funds in the relevant countries. With the European Council acting to put in place a huge commitment of public funds – some €750 billion, which is huge in relation, for example, to the outstanding stock of debt in the countries affected – as a financial back-stop mechanism to ensure access of governments to funding needs (conditional on compliance with adequate budgetary discipline – requiring quite sharp fiscal adjustment measures), the ECB stepped in promptly to purchase securities outright, with the aim of unblocking money market conditions, thereby restoring a more normal functioning of debt markets and ensuring a more effective transmission of the monetary policy stance, which has been geared for many months now to providing adequate liquidity at a low interest rate consistent with the overriding goal of low inflation.

Although market functioning improved following these actions, spreads on the debt of several sovereigns have widened again to levels that are exceptionally high by historical standards and this is still passing through to bank interest rates in the relevant countries, and adds to the cost and difficulty of rolling over wholesale deposits and debt issuance by banks in the relevant countries.

As I mentioned at the outset, Irish banks became heavily dependent on foreign funding sources at the height of the boom. It will clearly take some time for the Irish banks, like those in many other jurisdictions, to readjust their funding profile.

For example, as is well know, the Irish banks have been facing an increase in liability refinancing this month. Several policy tools have been available to assist with this process, both from the ELG guarantee scheme -- which I understand will very shortly be formally confirmed as applying to 0-3 month maturities -- and from central banking liquidity resources to the extent required.

The second problem has been the emergence of a market perception that significant European banks faced a serious problem with regard to cross-border holdings of sovereign debt. I believe that this issue was greatly overplayed by market participants and the recent EU-wide stress test exercise has, I think, gone a long way to putting it into a more realistic perspective, especially through the extensive disclosure of the relevant exposures.

Heightened awareness of the interaction between the budget and banks is stiffening a third form of pressure from budgetary policy to the banks. This relates to the new wave of proposals to tax banks one way or another. Given what has happened, it’s a natural impulse to see if some revenue can be clawed back from the banking system, especially in countries where most or all of the banks have now weathered the storm and are reporting healthy results. For my part, having seen and analysed in the past some of the damaging economic distortions around the world caused by poorly designed bank taxation, explicit and implicit, I am relieved that an international consensus seems to be building around forms of taxation that are less likely to add to damaging economic distortions. Not only can many types of bank tax fall disproportionately in practice on the smaller customers of the banks rather than on the shareholders or senior managers of the banks (who are sometimes the political target of such taxes), but the potential for disintermediation to undermine the revenue can mean both a disappointing revenue and sometimes an increase in prudential risks. Fortunately, these dangers are being recognised, thus, for example, imposing a levy on uninsured liabilities of banks, or on a surrogate for the value-added of the banking system, seems more likely to have an acceptable ratio of deadweight cost to revenue, than other forms such as a tax on interest rates, or than a general tax on bank transactions as has been proposed by some, but which would in my view not generate as much revenue as is often supposed, and would not be effective in discouraging the forms of financial excess that caused the crisis. It may be that well-targeted corrective Pigouvian taxes can do even better, for example if they are targeted at excessive reliance by banks and other financial firms on short-term borrowing.

Taxation is of course only one aspect of the heightened policy attention now being given to banks given their role in the crisis generally and not least in the damage they have, directly and indirectly, done to national budgets. Never more, I think, will banks be given the leeway to expand their intermediation activities with only a light regulatory touch to verify soundness. Banks must reconcile themselves to a heightened and more skeptical regulatory scrutiny, such as that which is being put in place in Ireland as in several other countries. The new capital and liquidity requirements recently agreed at Basel, and the heightened international surveillance of banking systems, exemplified by the proposed European Systemic Risk Board, represent an inevitable recognition of the interdependence of public and private finance, and the proposition that a banking license is a privilege, not a right, and one which needs to be exercised with greater regard to the public good.