Translating Politics into Technocracy in the European Banking Union

William O’Connell

University of Toronto

William O’Connell is a PhD candidate in the Department of Political Science at the University of Toronto. His research focuses on the politics of international finance, global governance, and the political economy of multinational corporations. Prior to pursuing his PhD, William worked as a policy analyst for the Government of Canada.

William’s dissertation research focuses on the history of cross-border bank failures and the development of the post-2008 regime for resolving too-big-to-fail banks. His work examines the extent to which cooperative relationships among individual technocrats, or between regulatory agencies, can compensate for or overcome more politicized barriers to international cooperation. More broadly, his dissertation explores the limits of the ‘soft law’ process typical of modern global governance in responding to international crises. Some of William’s other research has focused on the implications of the rise of China’s dominant consumer market on foreign censorship (published in Review of International Political Economy) and the regulation of cryptocurrency markets, for which he has made several media appearances.

During his time at AGI, William intends to synthesize findings from his dissertation fieldwork, including data from interviews and archival research conducted in the United States and Europe. A core element of this work will involve gaining a deeper understanding of the role of Germany, both historically through the Basel process and, more recently, in the Banking Union, in shaping international standards for financial regulation as a result of its unique market structure.

The failures of two Belgian-based banks, Fortis and Dexia, have cast a long shadow over European financial regulation. Once viewed as a model of regulatory cooperation, the Fortis collapse in 2008 quickly descended into an ugly, chaotic dispute between Belgian and Dutch authorities. Dexia’s 2008 and 2011 bailouts by Belgium and France raised thorny questions about burden sharing and liquidity provision in banks which are European in life, but national in death.[1] These failures, and the other banking and sovereign debt crises that have plagued Europe since 2008, spawned reforms embodied in the Banking Union, which began implementation in 2014. Yet, after nearly a decade, it is unclear whether this new, pan-European approach to crisis management can overcome the political and technical challenges associated with a major cross-border bank failure.

The Banking Union was intended to be built on three pillars. The first pillar is the Single Supervisory Mechanism (SSM), now in place, which entails a common rulebook for bank regulation and designates the European Central Bank (ECB) as the supervisor of the eurozone’s largest banks. The second pillar is the Single Resolution Mechanism (SRM), which establishes the Single Resolution Board (SRB) as the resolution authority for ECB-supervised banks and harmonizes national resolution procedures through the Bank Resolution and Recovery Directive. The SRM is also now in place, though with a number of gaps which are discussed below. The final pillar is a European Deposit Insurance Scheme (EDIS), which would see the creation of a jointly-funded and managed deposit insurance system. EDIS was dead on arrival, and member states have struggled even to harmonize their national insurance schemes.

The shortcomings of the SRM and the failure of EDIS stem from barriers to cooperation at both the global and EU levels. At the global level, the successful resolution of a cross-border bank requires overcoming several technical and political challenges. On the technical side, insolvency laws need to be harmonized between countries, information sharing and planning procedures need to be created among authorities, the complexity of bank structures needs to be drastically reduced, and a mechanism for private loss absorption must be created. On the political side, officials must resist the temptation to ringfence assets in a crisis and must be willing to share the financial burden of winding down a cross-border entity. They must also legally recognize resolution actions taken abroad, accept the possible termination of their national champion banks, and be willing to impose losses on shareholders and at least some portion of creditors in a consistent manner.

However, the soft law standard-setting process typical of global financial regulation is much better suited to addressing those technical barriers than the political ones. Since 2008, cooperation among regulatory agencies has dramatically improved: most G20 countries have broadly similar resolution regimes, cross-border contingency planning and information sharing have become routine, global banks have made significant progress in becoming more “resolvable,” and technical regulations and guidance on executing a “bail-in”—a method for haircutting bondholders in a crisis to recapitalize a failing bank—have been established. These provisions are embodied in the Financial Stability Board’s Key Attributes of Effective Resolution Regimes for Financial Institutions, a G20-endorsed international standard which was first published in 2011.

What is missing from the Key Attributes are credible commitments towards liquidity provision, burden sharing, and mutual recognition, all of which are contentious issues of fiscal responsibility which are beyond the scope of technocratic agencies. Without ex-ante agreement on the terms of cross-border intervention, the temptation to ringfence may mean fifteen years of careful planning will quickly fall apart. Governments may find themselves back in 2008, facing a choice between state-sponsored bailouts and chaotic, value-destroying breakups along national lines.

In the context of managing a banking crisis, the need for clear, quick, and decisive intervention calls into question the viability of a rigid, multilayered resolution process.

This is a challenge facing Europe and the rest of the world alike. What is unique to the EU is the mismatch between financial and fiscal authority. The other home authorities of large banks have the spending power and capacity to provide a safety net. They offer nationwide deposit insurance, liquidity in resolution, and last resort lending backed by the fiscal authority of the state. In some cases, including in the United States and the United Kingdom, there are statutory mechanisms for authorities to draw directly from the Treasury to finance a resolution. While these policies have clear rules stipulating their use, they are sufficiently flexible to allow regulators to respond to a variety of circumstances. The willingness of national authorities to cooperate across borders may be unclear, but at a national level, these countries’ capacity to intervene is not in doubt.

This is not the case in the eurozone. Without a fiscal backstop, EU institutions need to compensate with additional layers of bureaucracy, including stringent rules, quantitative thresholds, and member state vetoes. This is a familiar story in European integration, but in the context of managing a banking crisis, the need for clear, quick, and decisive intervention calls into question the viability of a rigid, multilayered resolution process.

Specifically, executing a resolution under the SRM entails a six-step process with multiple veto points. First, the ECB or SRB declares a bank “failing or likely to fail,” with the other confirming this assessment. Then the SRB undertakes an additional assessment to determine if a resolution is in the public interest (compared to a liquidation). The SRB then draws up a resolution plan in consultation with the relevant national authority, which must then be approved by the European Commission (and, presumably, the home authority government), approved again by the European Council, and, finally, implemented by the national authorities. This process must be done under a considerable time crunch, as a typical resolution is done over a weekend.[2]

Complicating matters further is a mandatory minimum threshold of eight percent of liabilities being “bailed-in” (wiped out) before the Single Resolution Fund (SRF) or any other public money can be put up to facilitate a resolution. This is sensible when faced with a single, idiosyncratic failure. However, in a systemic crisis, where potentially multiple large banks are under pressure, a mandatory bail-in may trigger firesales of eligible bonds, making it difficult for banks to meet the required threshold or leading to further resolutions. Moreover, the current maximum amount available to the SRF is €160 billion. Again, this is likely sufficient in an idiosyncratic failure but unlikely to be in a systemic crisis.

Further, these bail-in-eligible bonds are often prepositioned between a banking group’s various national subsidiaries. This creates an additional complication if problems in a failing bank are not distributed proportionally among its subsidiaries. Member states may be incentivized to ringfence assets or may otherwise refuse to cooperate if doing so means having losses imposed on their jurisdictions for activity undertaken abroad.

Finally, without a common deposit insurance scheme, the SRB’s public interest assessment must necessarily be sensitive to the various designs throughout the eurozone, many of which have different creditor hierarchies with different rules for repayment of the deposit insurer when the failed bank is finally liquidated. These arcane legal differences can alter who gets what during the resolution process in ways that may vary across member states, a problem which is antithetical to the stated purpose of the Banking Union, which is to finalize a single market for financial services.

The necessity of increasing European resources for combatting a crisis is universally recognized, though member states may be reluctant to contribute in the near term.

These shortcomings are rooted in familiar tensions which have characterized, to varying degrees, most aspects of European integration: conflict between Germany and France, between Northern Europe and Southern Europe, and between large, “home” states and small “host” states.

When Germany and France agree, integration typically moves forward. On banking resolution, however, major structural differences in French and German banking have produced seemingly intractable preferences on a pan-EU resolution and deposit insurance system. French banking is highly concentrated, highly internationalized, and has a single, publicly-funded deposit insurer. In Germany, banking is diffused between a few large, universal banks (Deutsche Bank and Commerzbank) and many smaller, localized Sparkassen savings banks with fewer international exposures, significant political influence, and a variety of semi-public and private deposit insurance and institutional protection schemes. Accordingly, German officials have successfully exempted the Sparkassen from EU oversight and resolution and have drawn a red line over mutualizing deposit insurance.

Beyond differences in the structure of deposit insurance systems, a key source of German reluctance stems from conflict between Northern Europe and Southern Europe in the wake of the sovereign debt crisis. German banks and officials—along with those of the Netherlands, Austria, and other Northern and Central European member states—have vehemently opposed any system where their contributions might be used to compensate for perceived weakness in Southern European banks, particularly in Italy and Greece. The same essential problem has stymied the development of a robust SRF and the rules around mandatory bail-in thresholds, which are intended to ensure EU resources are not used to bail out governments and banking systems perceived as behaving irresponsibly.

These two axes of contention have led to a watering down of European resolution procedures. The complexity of what has been put in place is, however, instead a function of conflict between large member states and small ones—dubbed the “home-host problem” in the context of banking. Here, large member states where a systemically important bank is headquartered (Germany, France, Spain, Italy, the Netherlands) are contrasted with smaller member states whose banking systems are dominated by subsidiaries of large foreign banks (e.g. Portugal, Belgium, etc.). For host states, this creates a twofold problem of having far lower fiscal capacity to respond to major crises, while also housing subsidiaries which are much more important to their financial systems than they are to the banking conglomerate as a whole (and, by extension, the home authority).

Consider BNP Paribas Fortis,[3] currently Belgium’s largest bank which accounts for roughly 22 percent of assets in the Belgian banking system. By contrast, it accounts for just 10 percent of BNP’s assets. This asymmetry means Belgian authorities are, to an extent, reliant on French authorities honoring soft commitments not to ringfence in a crisis as the lynchpin of their banking system relies on the availability of liquidity from a foreign parent company. Given the trauma of 2008, where many host states experienced the brunt of the losses from this type of ringfencing, there is considerable distrust among smaller states that large ones will not abandon them in another systemic crisis. The result is regulations around “pre-positioned” bail-in capital which, while insulating host states, further complicates a resolution, creates additional incentives to ringfence, and indicates a lack of trust which is, again, antithetical to the principle of the single market.

The result is that, in the absence of a supranational fiscal authority, EU rules on bank resolutions must necessarily be even more complex, rigid, and bureaucratic than their peers, whose rules are themselves plagued by difficulty in establishing credible political commitments. Pillar I of the Banking Union—the Single Supervisory Mechanism—is up and running. Pillar II—the Single Resolution Mechanism—is functional but marred by political barriers which jeopardize its ability to execute its mandate. Pillar III—a European Deposit Insurance Scheme—has yet to materialize, with little indication of possible future progress.

While incomplete, the Banking Union represents an obvious improvement over the pre-2008 status quo. An orderly resolution of a major, cross-border European bank is now possible, legally, if not politically. The necessity of increasing European resources for combatting a crisis—be it deposit insurance, resolution funding, or last-resort lending—is universally recognized, though member states may be reluctant to contribute in the near term. The seeds of a deeper, more complete Banking Union are present, but it may take another crisis before they are realized.

[1] The notion of banks being “global in life, but national in death” is attributed to former Bank of England governor Mervyn King in a 2010 speech at the Bank for International Settlements available here: https://www.bis.org/review/r101028a.pdf

[2] See David G. Mayes, “Banking Union: The problem of untried systems,” Journal of Economic Policy Reform 21, no. 3 (2018): 178-189. https://doi.org/10.1080/17487870.2017.1396901

[3] Following its failure in 2008, Fortis was acquired by BNP Paribas and retained the Fortis branding in Belgium.

Supported by the DAAD with funds from the Federal Foreign Office (FF).