Emmanuel Macron’s recent proposals for European reform have concentrated on fiscal issues but also include the demand that the European banking union should be completed, since its third pillar (a pan-European deposit guarantee scheme) is not yet implemented. The formation of the European banking union, initiated in 2012, is the last major reform in Europe and the most significant one since the introduction of the euro as a common currency. Prior to this reform, countries on the euro (now nineteen out of twenty-eight member states) already shared the same currency, the same central bank, and the same monetary policy. But supervision of banks remained a national responsibility, and there was no mechanism for financial solidarity between the various countries and governments.
The decision to create the banking union was made very quickly in June 2012, in the midst of the euro crisis—or more precisely, the crisis of indebted governments in the eurozone, mainly in southern Europe and in Ireland. One of the key aspects of that crisis, particularly in Spain and Ireland, was the problem created by the inherent link between the national government and banks in the same country. The governments of these two countries were not highly indebted by European standards, but they felt it necessary to rescue their banks with taxpayers’ money, which massively increased their own indebtedness. In all these countries, especially in Italy, banks held (and still hold) very significant amounts of bonds issued by their governments, which meant that a default of those governments would cause the banks to default, as well. The fatal link between them went both ways. In addition, many people suspected that national governments were not supervising the banks closely, and would not make sufficiently dramatic interventions when necessary. Critics denounced a form of cozy complicity between banks and supervisors, implying that problems were too often underestimated or hidden, and solutions dangerously delayed.
The idea of the banking union was then to cut those links in two ways: first by transferring the supervision of most banks in the eurozone to a common supervisor under the umbrella of the European Central Bank; second, by devising a system of bail-in for all European countries and by creating a common resolution authority for eurozone countries (complete with a dedicated fund to help the rescue process). A third pillar was proposed, to create a unified deposit guarantee scheme or at least a form of solidarity between existing schemes. The banking union proposal also included a mechanism for some sort of solidarity between states generally.1 At least for significant financial institutions, the banking union is supposed to transfer all aspects of banking supervision and crisis prevention and resolution to the European level.
In practice the union is limited to the euro area, since from the beginning it excluded Britain, Scandinavia, and the eastern European countries. Moreover, its whole purpose was to respond to the European debt crisis, which began when heavily indebted governments and banks in Greece, Portugal, Ireland, Spain, and Cyprus could no longer refinance their debt and suddenly required emergency assistance. As it happens, the eurozone crisis was “solved,” at least in its acute form. But the solution came about without any contribution from the banking union, which occurred much later and initially offered only a type of psychological impact. Rather it was the European Central Bank which, beginning in May 2010, began intervening to relieve the indebted governments and banks. This decisive step came when Mario Draghi explained that they would do “whatever it takes” to solve the crisis: they would buy huge quantities of bonds for years and years (as they are still doing now). Indeed, they would take a completely different approach from the banking union.
Solving National Economic Crises in Europe
Any evaluation of the banking union is based, then, not on its role in solving the crisis, but on the examination of what could happen in the future. Let’s first approach the issue of how to solve a banking crisis such as those mentioned above. Consider the Irish or Spanish situation, in which the banks were in bad shape and the government was unable to save them without sinking itself. In both Ireland and Spain, this situation followed a real estate bubble—an insane collective speculation that inflated the balance sheets of plain vanilla commercial banks with loans based on inflated property values. That arrangement quickly proves devastating for banks when prices plummet and borrowers collapse. In Spain, some of those banks were the “Cajas”—savings banks that were (and in many cases still are) closely linked to local authorities, most of them small- or medium-sized. More than the big Spanish banks, the small savings banks fueled the bubble and were threatening to explode. In Ireland the banks at risk were more typical commercial banks.
It is clearly important to prevent this kind of situation and be able to cure it—but does that imply a common resolution process at European level? First of all, a bail-in system is now in place everywhere in Europe which is considered to be the main available instrument to help solve such situations. But it’s entirely possible to have a bail-in system without a unified supervision and resolution system: non-eurozone countries such as the UK and Sweden testify to that. The objective of the bail-in system is to make sure that no state intervention will happen in the future. But if that works, there is no need for a common resolution process, except perhaps for very large pan-European institutions (a specific issue I will examine later).2
At its most general level, the stated objective of the banking union has been to reduce or eliminate the mutual dependency of banks and states. But that dependency is largely natural: banks are sensitive to the economic health of their country. Another issue is that banks have bought huge quantities of government bonds from their country of origin (sometimes under pressure). But that situation largely results from misguided regulation, notably the fact that government bonds issued by European countries are, even today, all weighted 0. This zero weighting is in tension with the idea that there is no significant fiscal solidarity between those countries, and in this respect a European banking union does not change anything. A large part of this situation could be reduced by adapting the prudential rules, but the problem here is that those weightings are decided at European level. Although weighting the bonds at the European level makes sense in theory, unfortunately it appears to be politically impossible to differentiate the weightings between the various borrower states.
Finally, it is said that Europe can help to secure a country’s banks if solidarity between the deposit guarantee schemes is introduced. But what banks’ customers in the countries under attack generally fear is not primarily the health of their banks, but their country’s exit from the eurozone—and no one can ever guarantee that their deposits will maintain their value in euros if a country leaves the eurozone. Besides, for the time being solidarity between deposit guarantee schemes is missing from the banking union proposals. Many countries, Germany above all, are extremely reluctant to contemplate solidarity at that level.
National Banks, European Technocracy
A similar set of problems emerges in considering the role a banking union would play in prevention and supervision. A banking union means that the European supervisor is responsible for actively preventing banking crises that can occur anywhere in the union (more precisely, in the eurozone). In the case of banks with predominantly national importance (as is most often the case), the supervisory role of the banking union can only be justified if local supervisors are less capable than a European supervisor of carrying out their tasks—presumably because local supervisors would be too “sensitive” to the local situation. But this has not been shown. In fact, there are good reasons to doubt Frankfurt’s ability to monitor banks in nineteen countries with different laws and languages—and, in addition, to impose measures that locals will not perceive as a violent aggression by foreign technocrats. The same is of course true in case of resolution: in 2017 the Italian government intervened with taxpayer money, though the banks involved were not major European institutions (Monte dei Paschi di Siena was the largest), because the implementation of the mechanical rules would have ruined too many small investors and, in the view of public authorities, would have hurt the whole Italian banking system.
Moreover, the de facto result of entrusting supervision to the banking union is to put banking policy itself in the union’s hands, and that represents a major transfer of sovereignty. The result, however, is that the banking dimension is transferred to the union while the management of the national economy is left in the hands of each state. The supervision of banks becomes federalized in a context that is not otherwise federal in the economic and political senses. On the other hand, we do have experience of European supervision in the case of states—the euro convergence criteria or the Maastricht criteria—and it is not good. There, too, only some of the tools are transferred to the European level, and the others remain at the level of national responsibility. The process leaves the political responsibility at national level, while transferring significant powers to a technocracy and relieving local authorities of that responsibility. In Europe the only democratically responsible political powers and the only living political debate are at national level. Elections at the European Parliament are made on the basis of national political considerations and political parties.
In sum, given the importance of the local economic situation for banks and vice versa, it is certainly feasible but not coherent to separate the supervision of banks from the management of the local economy. In both the Irish and Spanish cases, the past mistake was to allow a huge real estate bubble to grow to uncontrollable dimensions. With the benefit of experience, it would have been necessary first to diagnose this bubble, then to act to slow down the speculative movement, and finally to intervene early by restructuring the relevant banks (along with other tools such as budgetary measures). But to do that from Frankfurt would require a violent interference with the local process. It would also have been necessary to raise the central bank’s interest rates. But instead the eurozone has unified these rates despite the fact that the problems are very different in different countries.
In short, the right solution should not have been a wholesale transfer of responsibility to the distant and disconnected European level, but rather a clear local responsibility, complemented by measures such as the introduction of the bail-in device and changes in prudential rules. Then one could have added a form of supplementary, second-level European supervision, including the monitoring of systemic risks given the possible impact on neighbors. But except in the case of extremely large and clearly pan-European institutions, there was no need for a banking union.
Although the European Union is supposed to be a unified market since most regulations are pan-European, EU economies are clearly not homogeneous. When an Italian bank fails, the clients are Italian and it is a problem for them in Italy, not elsewhere (except in large financial crises). The Italian real estate market, SME market, and lending markets are different and separate from the German ones, for example. The same is true for salaries, employment, retail trade—and, of course, political responsibility. Europe is quite different from the United States in this respect. Although the FDIC resolution process is at the federal level even in the case of community or regional banks, the American economy is broadly unified and movement from state to state does not imply subjection to a foreign government with new and distinct rules, governance, and priorities.
The Case of Large Banks
The case of very large banks, those spread across national borders, is different. A crisis in one part or another of the banking group, contaminating the group level, can in this case have consequences going far beyond the original cause and the country of origin. The case for a much more integrated supervision system, as well as a resolution process, is much more compelling for such institutions.
There are very few such banking groups in the eurozone, however. If we exclude the activity of the main European banks in London, which may be very significant but outside of the eurozone (and will be outside after Brexit), the banks whose activity is significant in the eurozone outside their country of origin are fewer than five: BNP Paribas (in Belgium and Italy, Société Générale (mainly in eastern Europe), Crédit Agricole (in Italy), UniCredit (in Germany), perhaps one or two more. All the other big banks (Deutsche, the Dutch, the Swiss, etc.) are not significant in the eurozone outside their country of origin, since the financial hub of London (and the UK in general) is outside of the eurozone. In all those cases, the activity abroad is often not strictly coordinated across countries even when under one bank brand, and bank branches in different countries often operate as semi-independent divisions (much like U.S. banks operating in Mexico, for example). Why? For at least two groups of reasons.
First, banking markets at the retail level are extremely different in Europe from country to country: prices, products, and dominant players are all very different, often there are greater differences from country to country than the between the U.S. market and the Canadian or Mexican markets. Wholesale and financial markets activities are of course much more unified, but in that case only due to concentration of the activity in London (outside of the Eurozone). Take the example of mortgage lending: in France it is fixed rate at extremely low spreads (banks use them to stabilize their retail clients). French mortgage lending is very safe (the property is in practice collateral for what is in fact a personal loan, with strict rules especially regarding the ratio of installments to income), and borrowers have by law the right to repay in anticipation at a low cost. In most other countries, by contrast, interest rates are variable; in Germany premature repayment is impossible; in Italy, repossession is very slow. Pooling or managing these loans together across country lines is thus impossible. Insolvency laws are also quite different, which means that lending to SMEs becomes a different exercise as soon as you cross a border. For a European bank, entering a new European market means entering a very different game with quite different effective rules. Of course, it can still be done. But economies of scale are not what you can expect between, say, two banks in two different American states.
In addition, the structure of the banking systems differs widely from country to country (even if the UK is excluded). At one end, in France, six banks control the bulk of local banking activity; four of them are widely diversified, with significant investment banking activity (notably for two of them), mainly in London. At the other end, in Germany banking activity is dominated by a host of publicly owned so-called savings banks, complemented by mutual banks. Commercial banks have a much more limited role there, and among them only Deutsche Bank is a major player abroad and in financial markets. In between, there are in general two or three dominant banks in each country, surrounded by a variable but generally non-negligible group of smaller institutions; some of these are commercial banks, others are mutual or publicly owned banks. Concentrating on really systemic institutions would then yield the four French banks plus Deutsche Bank. (Concentrating on smaller ones would yield no French banks at all.) In addition, supervision of that very heterogeneous set is all the more complex once differing national political sensitivities are considered.
A second major factor is that the fate of banks is of major importance to national, democratically elected political authorities. This fact shows in these authorities’ tendency to increase their capacity (or the capacity of their resolution and supervision authorities) to check what can happen to their local banks, even when they are the subsidiaries of a large bank. For example, political authorities can oblige banks to incorporate locally, can increase the rights of the “home” supervisor, or can in some cases block the capacity of banks to transfer cash from one country to another even within the eurozone. All these actions come about for very clear reasons of political accountability.
Again, these factors do not eliminate completely the advantages of integrated supervision at the European level in the case of the very big banks. Integration makes that supervision much more homogeneous; it creates a much better level playing field and may smooth the resolution process. In fact, the very big banks initially welcomed the banking union. They are less enthusiastic now, however, since they perceive some of the associated inconveniences—a much more abstract and quantified approach by the new supervisor, and a strong tendency to unify the approach even when the local situations are very different. In addition, a large part of their activity is still and will remain outside of the unified supervision, particularly investment banking and other wholesales activities in the UK or the United States. But on the whole, in the case of the supervision of those large banks, one can say that the jury is out or perhaps that the balance is positive.
Large Banks and Resolution at the European Level
Although one could argue that similar reasons (of speed and immediacy) in favor of EU-level supervision obtain in the case of resolution for large banks, resolution is in fact an entirely different matter. But then we have to take into account two major difficulties. The first is that there are strong reasons to doubt that the magic solution of the bail-in will work in the case of a major institution, even less so if the difficulties happen during a major financial crisis. First of all, those banks are extremely complex and difficult to assess (even if the contingency planning that is required of them is reasonably well done). But in addition, the basic issue is still with us: they are too big and too interconnected to take the risk of their resolution. This situation means that their eventual failure would send shockwaves through the whole system—and the fact is a bail-in is essentially a compact, quick, and controlled form of default. In any significant case, it would send the same kind of shockwaves as default, even more so in an environment of incipient crisis.
For large banks, then, the integration of the resolution process at the European level, although not completely useless, is not a substantial answer. In short, it may happen that good old methods will have to be used, such as some form or other of public intervention, especially if political sensitivity is too high. But if you have to accept the idea that at least in some cases you may need public intervention, you need a public authority with sufficient resources to do that. That’s where matters become complicated in the eurozone. If supervision and resolution are European responsibilities, you would expect the corresponding authority (and money) to come from European sources. But nobody has that kind of capacity there, except if they can draw on European funds supposed to help sovereign debtors, and even then it is limited. Furthermore a consensus would be required to use the money of some of the states (most of the time, the Germans) to pay for the mistakes of others (mostly in the South), implementing something that was supposed to be avoided in all cases.
In short, even in the case of resolution, pooling the mechanisms at a European level is not a surefire solution. This deficiency is not the result of limited technical flaws in the system that could be fixed in some way or another. Rather, it is based on very significant limits to the way those European common institutions work.
The Limits of European Financial Instruments
Finance is a risky business: it is about evaluating risks in the future, and the future is unknown. In addition, there is a built-in tendency for financial activities to be quite profitable and safe for a long time before problems appear. Finance is also an intrinsically social or even political activity: when a problem arises, it generally becomes a problem for many participants. Leverage also compounds problems to the level of systemic crises, since debt is the most efficient way to transmit insolvency and failures through an economic system. Regulations and supervisions are thus necessary to reduce those risks, but in general they are not sufficient. At the same time, because finance is essential to an economy and the consequences of a crisis too risky for the majority, allowing the link between finance and public policy to be completely cut, at least in the case of a significant crisis is a risk.
So, while supervision and resolution authorities and mechanisms are extremely important and should normally provide the right solution, they cannot be the final answer to all problems. Final responsibility has to remain, in some respect, linked to the rules common to a particular polity. In practical terms, this means that managing banks and their crises requires a strong link with the political (democratic) dimension of our life in common. In the case of Europe, a completely unified banking union would only make sense if it were to operate within a genuine federal framework, including an integrated fiscal and economic policy for Europe as a whole. But this in turn would imply a political Europe and, in an environment where the base of legitimacy is democratic, a European democracy. But that would suppose the existence of a European people or demos, and we know that we are extremely far from that. Unless we want to jump into the vacuum of bureaucratic federalism without democratic legitimacy, Europe should return to a pragmatic policy of progressive cooperation and solidarity where possible and, in the case of finance, should adjust its tools to that reality. The consequence is clear: banks should not be too big for the underlying countries.
Yet there are ways to combine the strength of a large financial group with the existence of separate, locally responsible entities, and of entities with different tasks such as retail vs. investment banking. The answer is probably to have mandatory holding companies, owning separate entities, with separate supervision and resolution processes, and of course without significant cross-exposures between countries (and so no significant pooling of monies). The holding company may then be in position to help an ailing subsidiary, using group-level equity or possibly selling some healthy assets if necessary, but without jeopardizing the various subsidiaries. This is the original spirit, if not the present reality, of American bank holding companies. The bank holding company remains a useful concept, and requires some separation of commercial banking from financial market activities and investment banking. Other directions could be explored as well. But the main point remains in maintaining the link between the necessary responsibility for banking activities and those communities and economies for which they are essential, while avoiding contagion between one part of an integrated group and others.
Banking Union and European Politics
Even though the banking union arose in response to problems in the eurozone, the designers of the union never questioned the euro itself or its role in the crisis. Most economists, however, have never considered a common currency to be a reasonable design for extremely different countries. The introduction of the euro thus carried obvious risks and was not simply justified by economic considerations. Instead it was seen as a partial step toward building a more integrated Europe. The euro was a technical attempt to bring about a political change, since achieving political union directly would have been inconceivable. The meaning of the euro was therefore purely political in an area that was not considered politically sensitive.
In practice the euro led to a significant reduction of market discipline and a corresponding accumulation of debt in certain countries during the period preceding the crisis. This debt increase was fueled by low interest rates and the quasi-disappearance of spreads between borrowers, since the story was that the eurozone would create convergence among countries and economies, and thus solidarity would arise among member states. The debt in question was government debt in most cases (Greece being an extreme example, plus Italy and Portugal) and bank borrowing in other cases, as in Ireland and Spain. In these last two countries, the enhanced capacity of banks to borrow, enjoying artificially low interest rates, fueled a huge real estate bubble. Shortly thereafter, the crisis of government and bank debt in those countries spread to the European level and led to a sequence of actions by the European Central Bank and other pan-European authorities that has still not reached its conclusion.
The banking union is eerily reminiscent of the usual process of European integration: as soon as a problem arises, no matter what the question is, the answer always is the same—more European integration. As with any secularized faith, no return is considered possible. Every part of the acquis communautaire, every past decision in the direction of more integration is permanent, and calling it into question would be considered a serious moral fault. The European faith is based on the conviction that history has a meaning, a single meaning of greater human unity, and that we know it: it is both a necessity and a duty. Even when a mistake has been committed, the answer is always the same: increase the doses. Indeed, crises are needed to accelerate integration, because normally governments, parliaments, and public opinion resist the further reduction of national sovereignty. Another feature of that complex process is that it always remains partial, lacking coherence.
Before the banking union, Europe had a common currency but no common budget and no common management of the financial system. That will remain the case: after the banking union, there will still be major differences between the various national banking systems. Once again, European leaders are attempting to use complex but apolitical and thus “uncontroversial” methods to drive further integration. Once again, their methods will partially resolve some problems. But the banking union, as with other reforms, will leave in place the fundamental political disjunction at the heart of the European project.
2 Of course, one may doubt that the bail-in system will succeed in avoiding state intervention in the future. And in fact, there are arguably cases in which state intervention can be better than a bail-in. Rare as that case should be according to the current approach, however, nothing prevents Europe from helping the state directly while keeping its basic national responsibility. If a government intervenes to save banks, then by definition it is because the stakes are important for the economy concerned. But, aside from the rare case of an international (pan-European) bank facing an international problem, this situation is found first in specific national economies. It is thus strange that, in the case of banks with essentially national utility and which have become precarious mainly for local reasons, it is the whole of Europe that has to manage the operation. Of course, a neighbor’s health is important to everyone, and it is normal to show solidarity in case of real need. But it would probably be more reasonable, and in line with the principle of subsidiarity (and responsibility), to leave the prevention and management responsibilities where they are and where they should be—at a national level, with additional assistance if really necessary.