INTRODUCTION

The global financial crisis has woken up the European authorities prompting them to review the existing frameworks to manage the crisis both domestically and on a cross-border basis. In view of the series of events in the autumn of 2008, the established public safety net arrangements, designed to create safeguards for depositors, investors and policyholders, failed to ensure market confidence in the midst of a long-lasting systemic crisis. This article explores the limitations of deposit guarantee schemes (DGS) in the context of the overall safety net arrangements in Europe and proposes concrete avenues for reforms.

Initiated in the United States in the summer of 2007, the financial crisis worsened and became global in 2008. The collapse of Lehman Brothers, the rescue of AIG and the bail out of Fannie Mae and Freddie Mac in September 2008 were seismic events in financial history. The crisis of Fortis, Dexia and the crisis in Iceland, to cite a few relevant European developments, evidenced the international nature of the crisis, which has been rightly characterized as the most severe one since the Great Depression. The period of September/October 2008 will be remembered as the time in which the financial system was reshaped and when previous assumptions about the nature of markets and the benefits of government intervention were questioned again, given the unprecedented scale of the actions undertaken by the authorities to combat the crisis. Many national responses ranging from liquidity assistance to DGSs’ activation and other state aid measures (guarantees, recapitalization plans and others) were poorly coordinated showing the limits of the current arrangements for European supervision and crisis management.

After the run on Northern Rock in September 2007 (see Figure 1) and the likely spread of such a threat (a looming banking crisis) to other European Union (EU) Member States, the Irish government, acting alone, was leading1 the scene by guaranteeing 100 per cent of the deposits of the six major Irish banks and increasing the statutory limit for the scheme for banks and building societies from €20 000 to €100 000 per depositor per institution. This triggered a series of similar actions by other governments. The UK authorities increased deposit insurance coverage to £50 000 (from £35 000). Other European governments followed suit (in Greece, Spain, Germany … ) (see Figure 2).

Figure 1
figure 1

The run on Northern Rock in September 2008.

Figure 2
figure 2

Deposit insurance in 2008 from limited to blanket coverage.

The protection of private savings generated a debate in the EU about the effectiveness and the appropriate design of deposit insurance (see Table 1). Triggered by these individual Member States’ actions, European finance ministers decided during a meeting on 7 October 2008 to increase Europe-wide deposit insurance, for an initial period of at least 1 year to a minimum of 50 000 euros, which could be raised to 100 000 euros to contain the competitive distortions that such actions could generate in the future.

Table 1 DGSs in some European countries: EFDI DEPOSIT GUARANTEE SYSTEMS, FIRST REPORT OCTOBER 2006

However, these actions did not succeed in calming the disquiet of the market nor, in some cases, did they halt the depositors from queuing in front of their banks. Indeed, governments needed to resort to many other measures, including massive injections of liquidity, fiscal stimulus, and quantitative easing, to restore confidence.

It is in this context of safeguarding and restoring confidence where our analysis comes into place. Notwithstanding the importance of other components of the safety net ranging from lender of last resort (LOLR), emergency liquidity assistance, bank insolvency proceedings to early intervention measures (such as the new Special Resolution Regime (SSR) introduced in the United Kingdom by the Banking Act 2009),2 this article focuses on explicit deposit protection, which is one key element that – if properly designed and supported by effective bank supervision – greatly contributes to maintaining or restoring the confidence of individual depositors in banks and the confidence of the general public in the banking system at large.

THEORETICAL CONSIDERATIONS ON DGS IN A BROADER CONTEXT OF SAFETY NET ARRANGEMENTS3

Since the banking industry is inherently unstable, the authorities always need to be prepared to confront the possibility of crises or problems. Over the years, a number of preventive and remedial instruments have been devised to strengthen the banking system and defend it against any negative contingencies. Ex ante measures comprise better banking regulation and supervision, transparency and disclosure. Ex post mechanisms, designed to ‘steer the boat through a rowing sea’,4 include the suspension of convertibility of deposits into cash, the LOLR role of the central bank, deposit insurance schemes, bank insolvency proceedings and government policies of implicit protection of depositors (both insured or uninsured) or banks (the ‘too-big-to-fail doctrine’ and the multiple variants of this doctrine during the crisis: too interconnected to fail, too many to fail, too big to save and so on).

Deposit insurance has been the subject of much debate in the literature. In 1959, Milton Friedman expressed a widely held view when he asserted that the introduction of US federal deposit insurance after the bank crisis of 1929–1933, as part of the New Deal legislation under President Franklin D. Roosevelt, was ‘the most important structural change in our monetary system in the direction of greater stability since the post-Civil War tax on state bank notes’.5

Although deposit insurance has been adopted by law in many jurisdictions around the World, particularly in the last two decades, its very existence has been much debated in the literature,6 though those who opposed deposit insurance in the past may have changed their views on this subject in light of the crisis.

STRUCTURE OF DEPOSIT INSURANCE

The structure of DGSs varies greatly from country to country, with differences with regard to their funding, coverage and administration.7 In this article, we do not analyze the specific policy and structural features of deposit insurance, but rather some of the issues that are most relevant from the point of institutional design. The first issue is the difference between explicit and implicit deposit insurance, the second issue is the status of ‘preferred creditors’, which insured depositors have under an explicit DGS. The third issue is the mandatory nature of deposit insurance, as opposed to the contingent nature of the LOLR role of the central bank.

EXPLICIT VERSUS IMPLICIT DEPOSIT INSURANCE

Explicit deposit insurance is the formal creation of a DGS by law, with specific rules concerning the extent of the ‘insurance’ or protection, the operation of the scheme and the type of deposits/depositors protected. Explicit deposit insurance can be useful as an instrument of protective bank regulation. Explicit deposit insurance has traditionally served two purposes: consumer protection and the prevention of bank runs, as well as the broader objective of contributing to financial stability. The rationale behind depositor protection is the presumed inability of ordinary depositors to monitor the riskiness of banks in which they place their funds as well as the potentially severe cost of deposit losses to individual savers. The rationale behind the prevention of bank runs is the inherent fragility of the banking system. Owing to the first come, first served nature of bank liabilities, and because loans (unless securitized) are highly illiquid, and worth less at liquidation than on a going concern basis, depositors have a rational propensity ‘to run’ at the first sign of trouble. Bank failures become highly contagious, thereby exposing the financial system to the risk of depositor panics. We argue that a third rationale of explicit deposit insurance (in addition to consumer protection and prevention of bank runs) is that it allows the public authorities to close banks more easily, as it becomes politically acceptable to liquidate insolvent institutions, in the knowledge that unsophisticated depositors are protected.

Under an explicit DGS, depositors are only paid once the bank is closed. Thus, there can be no deposit insurance if the bank remains open. Therefore, explicit deposit insurance presupposes that a bank has failed and, hence, it is not compatible with the ‘too big to fail’ doctrine.

The European Shadow Financial Regulatory Committee (ESFRC)8 took the view in 1999 that explicit deposit insurance can and should play a key role primarily in facilitating the liquidation of insolvent banks without the need for implicit deposit protection. The ESFRC argued that the practice of bailing out insolvent institutions (implicit protection) creates expectations of official support beyond deposit insurance limits, thereby distorting market incentives and undermining financial discipline (the so-called moral hazard problem). It is the strongly held view of the ESFRC that deposit insurance should be designed and operated in a way that allows, and indeed requires, national authorities to liquidate insolvent banks, thereby exposing uninsured depositors and other creditors to default risk. Such an approach ensures that high-risk institutions pay a market penalty in terms of higher funding costs. In this way, excessive risk taking can be discouraged.

Implicit deposit insurance, as opposed to explicit deposit insurance, is a ‘blanket guarantee’ for all sorts of depositors (insured and uninsured), other creditors, shareholders and even managers. Implicit deposit insurance often presupposes that the bank remains in business (either because it is ‘too big to fail’ or because it is politically difficult to close the bank) thus creating pervasive moral hazard incentives. While explicit deposit insurance is applied ex post (following the closure of a bank), implicit deposit insurance is often applied while a bank is still in operation.

Explicit deposit insurance inflicts only very limited damage upon taxpayers, and, depending on the funding of the scheme, there may no damage at all. However, implicit deposit insurance has the potential of shifting the burden onto taxpayers, because rescue packages tend to be financed by the government. The use of rescue packages results not only in moral hazard considerations but may also affect competition, especially if a too-big-to-fail doctrine is applied.

‘PREFERRED CREDITORS’

Explicit deposit insurance is a guarantee limited to one type of ‘preferred creditors’, that is insured depositors. Under explicit deposit insurance, uninsured depositors, other creditors, shareholders and managers are not protected. Therefore, explicit deposit insurance is more compatible with market discipline, as uninsured depositors and other creditors have an interest in monitoring the solvency of the bank while still in operation. Explicit deposit insurance can co-exist with insolvency laws that give preference to depositors and also with insolvency laws that do not establish such preference in their bankruptcy procedures. (As acknowledged, in some jurisdictions, insolvency laws have a system of depositor preference, while in other jurisdictions, depositors’ claims run pari passu with other creditors’ claims).

Explicit deposit insurance, by limiting the protection of ‘insured depositors’, exposes uninsured depositors, general creditors, subordinated debt-holders, shareholders and management to increased risk exposure, thereby encouraging them to monitor and limit the riskiness of the bank. These incentives are very important, particularly in the case of shareholders, whose limited liability renders them more prone to lend on a high risk/high return basis, while restricting their own exposure through high leverage.9 In the absence of open bank assistance, management will also be inclined to run the institution in a prudent manner, or risk being removed from office.

In its statement of October 1999, the ESFRC recommended that uninsured deposits and other liabilities should be ‘credibly uninsured’, meaning that holders of such claims have no expectation of official support in the event of a bank insolvency. Explicit deposit insurance must be set at a level that enables national authorities to accept the political consequences of bank liquidations.

MANDATORY VERSUS CONTINGENT GUARANTEE

Deposit insurance provides a guarantee on certain deposits that is non-contingent. It provides legal certainty as to the way the depositors will be protected and the amount covered, should a bank be closed. LOLR, on the other hand, is contingent. The injection of liquidity in times of crises is not mandatory, but discretionary, that is subject to the discretion of the central bank authority. There is always a degree of uncertainty regarding the provision of emergency liquidity assistance by the central bank.

To minimize the risk of moral hazard, it is important to demarcate what each institutional arrangement can do and what it cannot do or should not do. Explicit deposit insurance can protect insured depositors, but it cannot – or should not – protect other depositors or creditors, shareholders, or managers. Explicit deposit insurance cannot protect the banks, because it can only be activated once a bank is closed.

In our opinion, DGSs should be mandatory and explicit in nature, credible and limited in the amount covered.10

DGSs IN EUROPE

In Europe, the 1994 Directive (94/19/EC) provided a minimum harmonization background for establishing DGSs. It resulted in a decentralized approach to deposit insurance. Although the criteria used in the Directive were generally harmonized in terms of the scope (the exclusion of the interbank and corporate deposits) and the minimum of coverage (fixed at 20.000 euros per person per bank), they were not sufficient to ensure a sound European deposit guarantee system. Indeed, the directive was implemented unevenly in the Member States as a result of divergent interpretations of its provisions. Table 1 shows differences between selected Member States in the legal framework, in the administration of the schemes, the extent of coverage, the co-insurance practices and the sources of funding. In the light of continuing trends of cross-border banking, these divergences in implementation address major challenges. Although foreign branches of EU banks are covered by the home DGS, foreign subsidiaries of EU banks are covered by host DGS. Both schemes can be intrinsically different and depositors are not necessarily aware of such differences, and in the majority of the cases they are not protected evenly. The situation gets further complicated in the case of branches of non-EU banks and in countries where no DGSs exist.11

This situation raises competitive considerations and it can be the source of potential conflicts of interest between the host and the home countries’ authorities. In addition it adds further confusion and complications for depositors, particularly if a bank failure occurs.

These limitations were acknowledged by the European Commission in a communication12 in November 2006, less than a year before the eruption of the crisis. The conclusions concurred not to make changes to the directive, but to work on some interpretative guidance and recommendations on the main aspects of the directive.13

The financial crisis has not only put into doubt the adequacy of the existing national schemes and the European Commission's decision not to amend the directive earlier, but has also pressed governments in the Member States of the EU and European Economic Area (EEA) to take individual un-concerted actions to help restore confidence in their domestic markets. These individual actions have then prompted EU policymakers to revise the original directive to prevent competitive distortions. Amendments were introduced to reassure depositors rather than to promote the convergence of DGSs. These amendments revised three key areas: (a) the increase of the coverage level (from a minimum of 20.000 euros to first 50.000 euros and within a further year to at least 100.000 euros14); (b) the reduction of the payout delay to a maximum of 3 days15; and (c) the termination of co-insurance.

In its design, the 1994 Directive fails to tackle some key aspects. It leaves to the discretion of the Member States a number of issues that are not harmonized, such the funding of the schemes, their risk sensitivity and the consideration of home and host country conflicts, which have been manifested in the cases of Ireland and Iceland (though for different reasons and with different considerations as we further discuss below). Although the line between what should be harmonized and what should be left to the discretion of national authorities is a fluid and dynamic one, the experience of the current crisis suggests that a greater degree of harmonization and, possibly, the introduction of a Community arrangement16 is needed with regard to DGSs, given that with the freedom to move capital and provide financial services and products across the EU (which corresponds to the Single Market philosophy), regulatory measures in one country (for example, the guarantee of private savings in the six largest Irish banks by the Irish government) have competitive implications in other jurisdictions.

The crisis also raises important issues about the ability of some governments to underwrite their DGSs and fulfill their obligations under the Deposit Guarantee Schemes Directive (as in the case of Iceland). The ‘natural’ tendency of countries to protect their own nationals in a crisis needs to be reconciled with the obligations of EU membership or participation in the EEA.

The responsibility of the home country supervisor for branches – in particular with regard to the obligations of the DGS – has been put under the spotlight following the collapse of the Icelandic banks, where the Icelandic authorities were not in a position to fulfill such obligations. Notwithstanding the subsequent handling by the UK authorities, the truth of the matter remains that reliance on home country control proved very problematic, thus leading some commentators to push for host country control, perhaps by making branches of countries whose financial soundness is in doubt, be converted into subsidiaries. This is, of course, a departure from one of the principles of the single market, the home country control and single passport, which has been rather successful until relatively recently. If a retrenchment of the single market is to be avoided, perhaps we need a move to a pan-European supervisory arrangement (though that would mean the need to tackle the fiscal issue), as Howard Davies proposed in a contribution to the Financial Times (Europe's Banks need a Federal Fix, Financial Times, 14 January 2009).

The collapse of the Icelandic Landsbanki in October 2008, along with several other difficulties affecting the country's banking institutions and the alleged unwillingness or inability of the Icelandic Government to accept responsibility to save UK deposits has soured relationships between the United Kingdom and Iceland, with the Treasury threatening legal proceedings to enforce compliance with the EU's Directive on DGSs. The UK Treasury also took the extreme step of invoking the Anti-Terrorism, Crime and Security Act as a legal basis to block Landsbanki assets in the United Kingdom, effectively as a means of securing assets for the benefit of depositors. As is invariably the case, blocking orders of this kind impose a significant burden of compliance on financial institutions.17

The 2008 amendments of the directive 94/19/EC provide useful lessons for a fundamental review of the adequacy of the existing schemes and for the discussion of possible improvements to enhance the level playing field in Europe.

First, the coverage limit that had been fixed by the directive to a minimum of 20.000 euros was not effective in preventing bank runs and re-establishing the general public confidence. The differences in coverage and scope in an internal market lead to competitive distortions and market confusion. The amount covered must be credible and harmonized. A credible deposit insurance system requires inter alia prompt payment of depositors (next business day as in the United States is considered ideal by many) and a reasonable amount of coverage (neither too meager to be non-credible nor too generous to incur moral hazard incentives).18 A harmonized coverage will avoid competitive distortions and other burdens related to topping up and information exchange between schemes.

Second, at a national level when more than one competent authority is entrusted with responsibilities in a bank failure situation, it is important to establish a mechanism to set the scope and the hierarchy of various powers ex ante in order to insure the timely payouts of depositors. The failure of such arrangement provoked a bank run on Northern Rock in the United Kingdom. A hypothetical bank run on a cross-border institution in Europe would raise several questions on such arrangements between the different deposit insurers of the host and home countries. The new banking legislation in the United Kingdom has introduced early intervention mechanisms in the pre-insolvency phase, with new powers vested in the Bank of England in the exercise of that SRR, which comprises three stabilization options: transfer to a private sector purchaser, bridge bank and temporary public ownership. Although there was some debate in the consultation, that led to the Banking Act, as to who the competent authority for running the SRR should be, in the end it was decided that the Bank of England (which also receives a statutory mandate with regard to financial stability) should be such authority, though the Financial Services Authority (FSA) remains in charge of bank supervision and will be the institution that will pull the trigger to assess when an institution should be subject to SRR.

Third, the topping-up arrangements19 must be revisited in line with further harmonization of DGSs. The branches that opted for a topping-up solution are insured by two deposits insurers (home and host) and pay premiums for both. The 2008 amendment of the 94/19/EC directive explicitly requires schemes to cooperate with each other to avoid complications related to delays of payments, breakdown between what should be paid under which scheme and exchange of information between schemes. Notwithstanding such an improvement, depositors may not be necessarily informed about such arrangements, which may create further disarray in a situation of crisis. In addition, such arrangements only deal with the provision of services via branches, and do not consider direct provision of financial services.

Fourth, the scheme needs to be widely known by depositors. Ample publicity should be given to the scheme in order to make it credible; depositors should be in no doubt that their deposits will be covered, up to the amount specified in the law.

Fifth, risk-based contributions must be the way forward to ensure fairness among banking institutions with different risk profiles. High-risk institutions are expected to pay higher contributions and vice versa. Stressing on risk-based schemes will add further incentives to banks to improve risk management, which is a natural development in view of the risk-sensitive elements introduced by the capital requirement directive.

Sixth, for large cross-border banks, considerable resources are required if a DGS was to play a meaningful role in any crisis management. The combination of today's collection of DGS may prove to be inadequate; therefore, proposals are needed to ensure an adequate funding. Preference should be given to private pre-funding mechanisms, but if this is not sufficient then state topping up could be envisaged.

MOVING TO A PAN-EUROPEAN DEPOSIT GUARANTEE SYSTEM

We suggest that a single market in financial services requires a European solution with regard to deposit insurance. The possible establishment of a European deposit guarantee system could address the problems for large cross-border banks with major cross-border exposures through branches or subsidiaries.20 Such a scheme would, in principle, be a more efficient solution than the current fragmented framework of almost 40 DGS in the 27 Member States. It would also help remove competitive distortions, deal with administrative burdens, avoid branch/subsidiaries’ consumers confusion and most importantly preserve the internal market for retail banking. The financial crisis episodes of bank runs and tax payers’ money spending to save major reputable banks were vivid illustrations of the interconnectivity of banks, the scale of their operations in their local and regional economies and most importantly the speed of the contagion. The fundamental question is how to prevent the internal market for financial services from disintegration and at the same time protect EU depositors and savers. The establishment of the European supervisory authorities (for Banking, Securities and Insurance) and the European Systemic Risk Board21, 22 are steps into the direction of finding European solutions for European financial institutions. The next step is thus to complete the regulatory and crisis management architecture with a European safety net system. Home country safety net per se is not a sufficient option any longer. This reinforces the case for a European deposit protection solution, which appears to many as a logical development in today's financial market.23

There are, in principle, three alternatives for such a construction to happen:

  1. 1

    An optional DGS that is complimentary to the 27 existing DGSs – ‘a 28 regime’

  2. 2

    A single European DGS that replaces the existing 27 DGSs

  3. 3

    A European system of DGS – a sort of college of DGSs providing each other mutual assistance

For each alternative, the scope of coverage/protection (in our opinion, only insured deposits), the coverage level, the financing mechanisms, the payout delays and modalities and the interactions with the other mechanisms of the safety net need to be thoroughly examined.

The diversity of Europe's banking market should also be taken into consideration. Large cross-border, regional and local banks coexist, while serving different segments of the market. A European deposit insurance scheme covering cross-border banks and national schemes covering all other domestic banks can coexist if they are harmonized in terms of their operations, coverage and financing mechanisms and interaction with the other mechanism of safety net.

Deposit insurance, if properly designed and administrated, makes it politically feasible and practically possible to close a bank because the authorities know that depositors are protected, as we pointed above.

We argue that the establishment of a European system of DGS (or a network of DGSs) would better fit with the needs of the single market in financial services than the current decentralized structure.

CONCLUDING REMARKS

There is strong evidence that some ‘conventional tools’ failed to act as ‘lines of defense’ in the crisis and that other tools and institutional arrangements ought to be considered (from non-conventional monetary policies to new arrangements for a cross-border resolution). This is the context of the arguments underlying this article. Greater reliance on EU arrangements appears as the logical solution to the inconsistencies of the operation of the single market in financial services, in particular in the euro-zone. The difficulties of solving these inconsistencies are further compounded by the domain of fiscal policy in the EU.

Going forward we need better regulation, better supervision and better crisis management (including credible explicit deposit insurance) on a cross-border basis (balancing the need to preserve fair competition and the need to achieve financial stability). However, actions by the public authorities (whether supervision, regulation or resolution) cannot succeed unless they are accompanied by better risk management and corporate governance by the financial institutions that are supervised and regulated. Furthermore, European initiatives need to be aligned with international efforts, as most pan-European banks are also global banks.

Our proposal is a modest step in the direction of better cross-border crisis management arrangements to prevent and contain the effects of possible future financial crises.