Response to consultation on “Reforming the structure of the EU banking sector”

European Commission





The Nordic/Baltic economies are small, open economies with a well-functioning social safety net for citizens allowing households to take a longer term view on investments and place their financial assets in e.g. asset funds or securities. Institutional investors tend to place a large part of their portfolios abroad, and therefore need to be able to hedge themselves against various risks. These economies also have a number of large and middle-sized export companies and several of the countries have not yet introduced the euro. The markets for government bonds and corporate bonds, covered bonds, interest rate derivatives and foreign exchange derivatives, are, to a large extent, provided by a few banks acting as market makers. The Nordic/Baltic economies are consequently highly dependent on the trading activities of a few universal banks. They provide services that are essential for the real economy in these countries. Without those services, institutional investors such as pension funds and other asset management funds and large export corporations could not handle their risks effectively. Furthermore, in the Nordic/Baltic countries, as well as in many other countries, financial crises have mainly been caused by lending to the real estate sector – not by trading activities.

We have two key messages on the issue of structural reform:

First, mandatory separation is not a good solution to stem disproportionate risk taking or speculative activities. Instead, the focus should be on good risk management and improved supervision. It is the elements of excessive risk-taking in trading, that policy makers fear could contribute to aggregated systemic financial risk in the EU economy – not trading in general. Higher capital requirements, or in extreme cases, separation within the supervisory authorities’ mandate given by the BRRD, would be better tools to handle that problem.

Second, on the issue of thresholds / criteria for capturing banks, we believe that to target trading activities of universal banks based on accounting based parameters (market valued assets in relation to total assets) would be problematic if the applied method did not adequately reflect the levels of risk in the trading business. We note that the Commission has given some thought to how one should avoid the inclusion of assets held in order to secure compliance with new liquidity rules (AFS), but the problem remains. If a proposal on structural reform is presented, it should catch the right banks – no matter what the proposed measures are. We strongly advocate that selection criteria reflect risks in the trading activities, such as potential losses and other types of trading related risks.

1. Can structural reform of the largest and most complex banking groups address and alleviate these problems? Please substantiate your answer.

According to the Commission, the intention of a separation proposal would be to prohibit that trading activities within a bank can rely on guaranteed deposits to secure funding at a lower cost than it would otherwise have been able to on a stand-alone basis. The assumption is that the cross-subsidization between the two businesses would give a bank the opportunity to take on higher risks at a larger scale in its trading activities.

We argue that separation does not solve the alleged problem of intra-group subsidization of trading activities. It is the level of risk in the trading activities that should be targeted, rather than the issue of cross subsidization. One method to come to terms with that is through strengthened supervision and better monitoring of risks in the trading branch of a universal bank.

2. Do you consider that an EU proposal in the field of structural reform is needed? What are the possible advantages or drawbacks associated with such reforms?

We do not believe that an EU proposal in the field of structural reform will solve the problems described in the HLEG report. Many of the problems that a structural reform would try to address are already being addressed through other legislation under way. Increased supervisory powers and possibilities to set higher capital requirements for certain types of activities are provided for in CRR/CRD4. Within the new EU framework for crisis management (BRRD) supervisors have the authority to demand a bank to separate its activities. A common EU proposal on structural reform is therefore not needed even against the background of national proposals in some member states.

In the Nordic/ Baltic countries the larger universal banks are market makers in e.g. government bonds, covered bonds and in the foreign exchange market. Large issuers of high quality bonds (governments and mortgage institutions) could suffer severely if our banks had to reduce or give up their market making activities and it could have serious consequences for the local mortgage systems. This is also true for our export industry regarding foreign exchange related financial products. For our economies, a mandatory separation would have severe negative effects on the activity in the capital markets and liquidity would suffer.

3. Which of the four definitions is the best indicator to identify systemically risky trading activities? If none of the above, please propose an alternative indicator.

We do not support any of the indicators presented in the Commission consultation paper for the following reasons:

1. Assets held for trading and available for sale in relation to total assets.
This measure creates several problems. Both assets held for trading (HFT) and available for sale (AFS) are accounting based and on their way to be altered in a future accounting context. New liquidity requirements require banks to hold larger liquidity reserves, which can affect both assets available for sale on a bank’s balance sheet, and assets held for trading, depending on the strategy of the individual bank. As a consequence, a bank could exceed the first threshold for a mandatory separation as a consequence of being compliant with the new liquidity requirements. In the current market environment several banks hold a significant part of their liquidity reserves as deposits in central banks. In a more normal market situation these balances are likely to be shifted into highly liquid low risk securities (classified in the accounting as held for trading or available for sale) unless there are regulations punishing such a shift. A threshold based on market valued assets in relation to total gives a bank incentive to keep activities below a certain level, but does not give incentives to keep down risk levels in the trading business. The threshold would potentially catch a large number of European universal banks, but not necessarily the ones that it is intended to catch.

2. A narrower definition that excludes available for sale as mostly held for liquidity purposes
The narrower definition is marginally better from a liquidity management perspective, in any case for those banks for which much of the liquidity reserve is reflected in AFS. However, there are several banks, among others some of the large Nordic / Baltic banks, where assets held in the liquidity reserve would be categorized as assets held for trading. So even with this narrower definition reflecting assets in the trading book in a more limited sense, the problem remains. The definition does not take into account any aspects of risk levels or the risk profile in the trading activities. A risk reflecting criteria would be essential to assess whether the trading activities of a bank should be subject to measures of some kind, be it supervision, capital requirements or separation as a stability measure within the common framework for crisis management.

3. Gross volume of trading activity (prop trading and market making)
Trading volumes are possibly a more appropriate indicator than market valued assets on the balance sheet if the aim is to capture banks with large trading branches in general. However, it still unclear, whether the distinction between gross volumes and net volumes would help in drawing accurately the line between proprietary trading and market making. If a bank’s trading business engages in proprietary trading it will seek to net positions and minimize asymmetries between positions if possible. On the other hand a market maker can temporarily have open positions that, at a certain point in time, do not reflect the fact that the market maker is acting on behalf of a client, seeking to close these positions at the nearest possible opportunity.

4. Net volumes (captures institutions with a higher share of unbalanced risk trading, supposedly prop trading)
For reasons indicated above, we do not believe that assessing gross versus net volumes will automatically reflect whether trading activities could be categorized as market making or proprietary trading in an accurate way. Using net volumes could be more relevant than using gross volumes, but it would be dependent on the principles applied in the netting. A narrow definition of netting, requiring that only exact matching positions or exact matching risks may be netted out against each other would cause difficulties when the measure is to be applied in practice.

4. Which of the approaches outlined above is the most appropriate? Are there any alternative approaches? Please substantiate your answer.

As stated above, we do not support mandatory separation as a tool to reduce systemic risk.

In several Nordic/Baltic countries, the secondary market for covered bonds, government bonds, interest rate and foreign exchange derivatives, is entirely provided through market making by a number of market participants. Should such activities become subject to mandatory separation, banks would either limit these activities, in order to keep them below the thresholds, or turn down the activities entirely, should they not be economically viable on a stand-alone basis. Both would have severe effects on liquidity in the local capital market. An alternative or second phase development could be that the trading activities in the Nordic/Baltic currencies are concentrated to one or a few big international mono line players. This would distort the competition in our markets.

Instead, we suggest that the need for some form of structural measure should be based on criteria reflecting the risk level in the trading activities. The Commission should consider criteria that reflect the different risks within the trading business. Furthermore, to avoid different regimes on banking structure within the single market, definitions should be clear and predictable.

5. What are the costs and benefits of separating market-making and/or underwriting activities? Could some of these activities be included in, or exempt from, a separation requirement? If so, which and on what basis?

The separated unit will suffer from higher administrative and organizational costs such as new management, new IT-systems, risk control etc. More importantly, due to the mono line nature of the separated businesses, the trading unit would receive a weaker credit rating, meaning higher funding costs. Higher funding costs would most likely lead to fewer market participants and a concentration of activities. Lower activity on the markets implies less liquidity. Splitting banks will create new systemic risks, as a few large players may emerge, which conduct only trading related banking and which other market participants will become dependent on – both other banks and companies from other sectors.

6. Should deposit banks be allowed to directly provide risk management services to clients? If so, should any (which) additional safeguards/limits be considered?

Yes, our opinion is that universal banks must be able to provide risk management services to clients without running the risk of being separated. If it is necessary, additional safeguards could be in the form of capital or corresponding buffers on the liquidity side. Safeguards should be based on an analysis by the supervisory authority, designed for this purpose.

7. As regards the legal dimension of functional separation, what are the costs and benefits of regulating intra-group ownership structures?


8. What are the relevant economic links and associated risks between intra-group entities?


9. As regards full ownership separation, what are the associated costs and benefits?

We believe full ownership separation will result in higher transaction costs in the capital market, for all market participants - banks and their customers -, and a higher level of risk in the financial system, while benefits remain hard to find. Higher costs of funding and capital for trading banks contribute to higher transactions costs. The loss of positive diversification effects turns both trading banks and deposit banks into mono line businesses. Separation is likely to lead to a concentration of market making activities to a few large streamlined investment banks. Banks will become less robust, especially in a crisis situation, which will increase the risk of a bank failure.

10. Does the above matrix capture a sufficiently broad range of structural reform options?

No, the matrix does not capture the relevant structural reform options as other alternatives – i.e. measures to stem systemic risk build-up in the trading branches of universal banks through additional capital or separation within the crisis management directive - have not been considered. Our Associations would welcome an in-depth analysis of how a combination of measures already accounted for in the CRR/CRD 4 and the BRRD could present an alternative to mandatory separation.

11. Which option best addresses the problems identified? Please substantiate your answer.



Thomas Östros - Swedish Banker's Association
Piia-Noora Kauppi - Federation of Finnish Financial Services
Jørgen A. Horwitz - The Danish Bankers Association
Katrin Talihärm - Estonian Banking Association
Stasys Kropas - Lithuanian Banks Association
Martins Bicevskis - Association of Commercial Banks of Latvia
Idar Kreutzer - Finance Norway