Should Denmark and Sweden Join the Banking Union?

There is an important policy discussion in Denmark and Sweden on joining the banking union. This article reviews the pros and cons of joining. The main rationale for joining the banking union is cross-border banking in the EU internal market. Reviewing the banking systems, we find that Denmark and Sweden have the same cross-border characteristics as the euro area countries, which suggests that the rationale for joining is similar. Moreover, both countries have large banks which may be too big to save at country level, but not at the bank union level. Nevertheless, there are some governance concerns. While euro area countries have an automatic and full say in all banking union arrangements, the out-countries lack certain formal powers in ultimate decision-making. We find that this may in practice be less of a problem. Finally, the out-countries have the nuclear option to leave the banking union.


Introduction
The euro crisis was the immediate cause for establishing the banking union in Europe in 2012. The more structural reason for the banking union is the internal market in banking, which facilitates cross-border banking. That raises the question whether out-countries, that participate in the EU internal market, should also join the banking union. With Brexit, this question has gained renewed interest in the remaining out-countries with large banks, such as Denmark and Sweden.
The out-countries have the option to join the banking union. This paper considers whether Denmark and Sweden should join the banking union. The assessment is made from the perspective of a country that is a member of the European Union (EU) but which has not yet adopted the euro as its currency (the so-called out-countries). Essentially, the focus of the paper is on whether it would pay for Denmark and Sweden to join the banking union. It fits in an emerging literature on the pros and cons of joining banking union (Darvas and Wolff, 2013;Hüttl and Schoenmaker, 2016;Beck, 2019;Ekholm, 2020;Jensen, 2020).
We consider the degree of cross-border banking in Denmark and Sweden. It appears that both countries have the same cross-border characteristics as the euro area countries, which suggests that the rationale for joining is similar. But there are some governance concerns. While euro area countries have an automatic and full say in all banking union arrangements, the out-countries lack certain formal powers in ultimate decision-making. This may in practice be less of a problem than in theory, as the governing council of the European Central Bank (ECB) tend to rubberstamp decisions from the supervisory board, of which the out-countries are full members. Moreover, the out-countries have the nuclear option to leave the banking union.

Cross-border banking
The perceived need for a banking union in Europe has been widely motivated by the negative spiral that can result when banks hold sovereign bonds and governments bail out banks. This close link between banks and government solvency has since the euro sovereign debt crisis been seen as one of the biggest threats to financial stability in Europe. Therefore, creating a supranational supervisor and bank resolution regime seemed a logical response to this threat.
However, the academic literature has for a long time, and long before the euro sovereign debt crisis, pointed to the need for a banking union when there is cross-border banking. Early papers include Folkerts-Landau and Garber (1992), Schoenmaker (1997) and Vives (2001). With crossborder banking, there is a "financial trilemma" stating that the three objectives of financial stability, cross-border banking and national financial policy cannot be achieved at the same time (Schoenmaker, 2011). More generally, the interests of home and host countries of cross-border banks are likely to deviate in distressed situations.
The key point can be illustrated as follows: Suppose country A is not only the home country of banks from country A but also host country of banks from country B. In order to provide financial stability in country A, the authorities in country A would need information (about capital and liquidity positions of distressed banks) from the supervisory authorities of country B. However, country B may have reasons to hold back such information. Failure to get this information from the foreign supervisor, fully and on time, might (seriously) jeopardise the possibility for country A to deliver financial stability in country A. So, there is a need for a supranational authority.
Against this background, the pros and cons of taking part in the banking union should be assessed in terms of the magnitude of cross-border banking. While it is widely agreed that a currency union does not work without a banking union, the question is if EU Member States outside of the euro area have a large foreign component of their total banking assets from other EU countries. Table 1 seeks to illuminate this. Apparently, the foreign component is significantly larger in the EU countries outside the euro area (non-banking union) than in the countries within the euro area. From this perspective, the case for joining the banking union is strong. The question is whether the numbers for Denmark and Sweden are well represented by the average numbers for the nonbanking union.  Table 2 offers a closer look at individual EU countries outside of the euro area. Two observations leap to the eye. First, the magnitude of cross-border banking penetration in Denmark and Sweden is much lower than in non-banking union EU countries in Eastern Europe. Specifically, the home share of total banking assets in Denmark and Sweden is 86% and 82%, respectively. Second, the pattern found for Denmark and Sweden does not deviate much from what is found for members of the euro area (see Table 1). Therefore, the degree of cross-border penetration of Scandinavian banking is likely to be high enough to constitute a (strong) case for joining the banking union. Yet, in future work, it would be interesting to study more closely if there is a critical level of cross-border banking such that if a country's cross-border banking exceeds a certain 'threshold' (e.g., 15%), the country would benefit from joining the banking union.

Consolidated banking supervision and burden-sharing
Another aspect of joining banking union is the effectiveness of national banking supervision. It is difficult for national home-based supervisors to monitor the foreign operations of large banks. National supervisors typically rely on supervisory colleges of home and host supervisors based on memoranda of understanding (MoUs). However, these MoUs are voluntary.
Moreover, supervisors lack incentives to cooperate and share information in times of crisis (Schoenmaker, 2011), as witnessed in the global financial crisis. A major achievement of the Single Supervisory Mechanism (SSM) is the consolidated supervision of cross-border banks at the euro area level (Schoenmaker and Véron, 2016). Consolidated supervision in banking union enables the ECB, as central supervisor, to have an overall view of a bank's euro area operations. Table 3 shows the cross-border operations of the major banks (with total assets of more than € 100 billion) in the banking union countries as well as Denmark and Sweden (Duijm and Schoenmaker, 2020a). For illustration purposes, we base the calculations on the assumption that Denmark and Sweden join the banking union, which we label the enlarged banking union. The geographic spread of the major banks' activities is divided between the home country, the rest of the enlarged banking union, the rest of Europe and the rest of the world.
For the major banks in the banking union, 62% is in the home country, 12% in the enlarged banking union, 11% in the rest of Europe and 15% in the rest of the world. The geographic spread of the major Danish and Swedish banks does not deviate much from that of the major European banks. The proportion in the enlarged banking union is even larger for both countries (17%) than for the banking union countries (12%). The case for joining banking union on this front is for the two Scandinavian countries even stronger than that of other banking union countries. Furthermore, the banking union allows countries to share the burden of resolving ailing banks (Goodhart and Schoenmaker, 2009). The Single Resolution Mechanism (SRM) resolves ailing banks at the banking union level with access to the Single Resolution Fund (SRF). The reformed European Stability Mechanism (ESM) Treaty provides a backstop facility to the SRF 1 , which constitutes burden sharing between ESM countries.
Denmark and Sweden will not join the ESM in case they join the banking union, unless they would at the same time become a member of the euro zone. Under the common backstop arrangement for the SRF, they will set up a "parallel credit line". They will participate in the common backstop on equivalent terms and thus join the burden sharing mechanism. 2 By contrast, Denmark and Sweden are, if outside the banking union, on their own if and when they have to resolve one of their major banks. The stability of a banking system ultimately depends on the strength and credibility of the fiscal backstop. While large countries can still afford to resolve large banks on their own, small and medium-sized countries have difficulties providing a credible fiscal backstop to any large cross-border banks they host. Table 4 shows that the potential fiscal costs of a severe systemic crisis could amount to 12.1% of Danish GDP and 10.5% of Swedish GDP, if the respective government needed to recapitalise the largest three banks. We calculated an indicative hurdle rate for fiscal costs of 8% of GDP (Schoenmaker, 2018). Below that rate, countries were able to resolve a financial crisis without external assistance during the global financial crisis. Above that hurdle rate, countries needed external support from the International Monetary Fund (IMF) or the ESM.
So, Denmark and Sweden cannot provide a credible fiscal backstop to their large banks. These countries have to manage this large and undiversified risk. In the case of an asymmetric shock to the Danish or Swedish economy (e.g. a national housing market shock), these economies are much exposed to their large banks. The Swedish government has introduced tax increases and extra regulation for its large banks. Moreover, its largest bank, Nordea, wanted for investors to be in a peer group of European banks instead of Swedish banks. Apparently, these pressures are behind the recent relocation of Nordea from Sweden to Finland (FT, 2017). Also the United Kingdom and Switzerland, with potential fiscal costs above the hurdle rate of 8% of GDP, have adopted policies to downsize their banking system. We suggest that Denmark should consider how to deal with a banking system that is too big to save. Summing up, the level of inward banking from other banking union countries and the level of outward banking to other banking union countries determine the benefits of joining banking union. The calculation of these benefits is very important for assessing what Ekholm (2020) reports as the clearest cost of joining banking union: the loss of regulatory and supervisory independence. However, as also noted by Ekholm (2020), the size of that cost may be small in a world where financial markets are highly integrated. With a sufficiently high degree of financial interdependence, the scope for regulatory and supervisory independence at the national level may cease to exist (Schoenmaker, 2011).

Monetary union, monetary policy and decision-making
The banking union was introduced in 2012 to address the bank-sovereign 'doom loop'. The rationale for centralised supervision in this case arises partly because of cross-border externalities from sovereign default that are sufficiently large to justify cross-border transfers. Whereas the euro can be at stake for members of the euro area, for small stand-alone countries, such as Denmark and Sweden, the risks associated with a doom loop may be much smaller. In principle, their central banks are not constrained in the same way to act as lenders of last resort to the government as is ECB vis-à-vis the governments in the euro area.
This insight might weaken the case for Danish and Swedish membership of the banking union, as there is no currency union for them to defend and their access to lender-of-last-resort activities seems more straightforward. However, there is a substantive difference between the two countries' exchange rate policy, potentially implying a (big) difference in the pro-versus-con calculation. While Sweden pursues inflation targeting, Denmark pegs the krone to the euro, as part of ERM-2. The peg is a cornerstone of Danish economic policy and there is widespread support for the fixed exchange rate policy (Bergman, Hutchison and Jensen, 2015). It implies that monetary policy interest rates are solely used to keep the Danish krone stable against the euro, while other considerations are not taken into account. Therefore, stability of the euro area is more important for Denmark than for Sweden.
This takes us to discuss the role of decision-making in the banking union (see Véron, 2015, for an excellent overview). The point is that the location of the supervisory authority in the banking union is ultimately in the ECB. This may well present a challenge, as Denmark and Sweden have no representation on the ECB's governing council, which is the ultimate decision-making body on supervisory decisions. Nevertheless, the countries that join the banking union on a voluntary basis will be represented at the ECB's supervisory board, which prepares the supervisory decisions. 3 In that way, Denmark and Sweden would de facto be part of decision-making, as the governing council typically rubberstamps supervisory board decisions (Schoenmaker and Véron, 2016). Clearly, Denmark and Sweden would prefer a location of the Single Supervisory Mechanism (SSM) outside of the ECB, but this is hardly realistic at this stage of development of the banking union. Also, as monetary policy and macroprudential polices are intertwined, it can from a broader point of view be debated whether a separate location of the SSM is desirable.
Again, a difference between Denmark and Sweden can be pointed out (Bergman, Jensen and Thøgersen, 2018). Indeed, Denmark has two decades of experience with taking part in such an arrangement, by being de facto in the euro area when it comes to monetary policy but without having a representation in the governing council. Apparently, membership of the decision-making bodies has not proven to be decisive for reaping benefits in terms of macroeconomic stability. In the same vein, participation in the banking union, without joining the euro area, could be a combination likely to generate important benefits in terms of financial stability.
Another example where it could make a difference if a country is not only a member of the banking union but also of the euro area is resolution. Suppose Denmark joins the banking union and a Danish bank -e.g., Danske Bank -runs into big trouble. In the first round, decisions about resolution would be taken by the SRM, a body where Denmark has representation. 4 As for resolution, this might happen in a situation where resources would need to come from the ESM, the fiscal backstop for the SRF. The ESM is an intergovernmental institution of the euro area countries. Nevertheless, as discussed above, there are commitments in the draft revised ESM Treaty and the draft backstop guidelines, which ensure that countries joining banking union without joining ESM will be informed and involved equivalently in the decision-making for the backstop arrangements. 5 Thus, ultimately, the decision about resolution of Danske Bank might have been taken by a body without direct Danish representation. Admittedly, this might be regarded as a rather extreme case, but it illustrates the importance -for Denmark and Sweden -of knowing in advance whether their treatment would be the same as for members of the euro area.

Final considerations
The main rationale for joining the banking union is cross-border banking in the EU internal market.
Reviewing the banking systems, we find that Denmark and Sweden have the same cross-border characteristics as the euro area countries, which suggests that the rationale for joining is similar.
There is an (often neglected) area where we find the case for joining the banking union particularly strong. Indeed, the SSM already has significant resources and will over time gain extensive experience in supervising different types of institutions (Beck, 2019). The fact that it is located far from most of the institutions that it supervises may also reduce the risk of regulatory capture.
The SSM, based in the ECB, would not only be able to attract talent at the junior level but also to develop and maintain senior staff, thereby having a very experienced and highly competent staff. Financial Supervisory Authorities in smaller countries, such as Denmark and Sweden, typically have high turnover rates, with the best and most ambitious staff moving to the private financial sector. The point is that supervision is complex, and makes heavy demands on skills to match the expertise available in commercial banks. To us, this is a key benefit of joining the banking union.
Overall, we consider that the clearest economic benefit of enlarging the banking union is the prospect of more efficient resolution of cross-border banks at the banking union level (see also Ekholm, 2020). An equal distribution of the gains as well as full participation in decision-making are crucial for a lasting membership. Maybe a 'flexible' membership should be considered, by joining now and exiting later without big costs if membership does not live up to expectations. 6