The Danish Bankers Association, the Danish Mortgage Banks' Federation and the Association of Danish Mortgage Banks would like to thank the Commission for the opportunity to comment on the possible impact of the CRR and CRD IV (the new regulation) on credit institutions' lending.
In our opinion, the CRR and CRD IV regulatory framework is generally reasonable. Many of the hurdles in the original regulation and directive proposals were solved in the legislative process. Credit institutions focus today on the final wording of pending regulation under the CRR and CRD IV framework as well as on potential future requirements. In this connection we find it worrying that the proposals generally tend to not take account of a risk-based approach.
We share the view of the European Commission that an adjustment of any unintended consequences of the introduction of the CRR and CRD IV is a precondition of building a Capital Markets Union (CMU) and the successful revitalisation of the European economy.
Therefore, we find it essential that the evaluation of the impact of the new rules should be as comprehensive as possible and also take into consideration the new capital requirements and other central elements of the new regulation. This also includes the new short-term and long-term liquidity requirements (LCR and NSFR) and potentially binding leverage ratio requirements. The mere scope and complexity of the new regulation implies that the consequences often need to be considered jointly and jointly with other regulation, to reach a fair evaluation of these interaction effects.
In a Danish context, particular focus is on the derived effects observed in the bond market. An efficient bond market and an efficient repo market are essential to the Danish market given the significance of mortgage financing to total lending in Denmark, including home financing and financing of SMEs.
These elements will also appear from our answers to the individual questions in the consultation paper.
1. What role has been played by the CRR and CRD IV requirements in the recapitalisation process, in terms of the timing and overall effect on the levels and quality of capital held by banks? How have market, supervisory and regulatory capitalisation demands interacted to make banks adjust the level of capital they hold to the current level? Whilst these three factors may be interlinked, is it possible to identify which has/have played the most important role?
The increased capital requirements laid down in the CRR have been the driving force of the recapitalisation of the financial sector. The magnitude of the new regulatory capital requirements as well as the new requirements regarding the quality and composition of capital has overshadowed the market and supervisory driven demands for capital in comparison to the situation before the financial crisis.
On the other hand, the significantly higher regulatory capital requirements have raised market driven capital requirements to higher levels than before the adoption of the CRR and CRD IV. This effect is amplified by the increased qualitative requirements on capital including the introduction of new triggers regarding conversion or write downs of capital instruments causing increased investor demands for an additional buffer on top of the regulatory capital requirements. This additional buffer has increased the total capital level further.
2. If you consider that capital levels go significantly beyond what is necessary in light of the level of risk incurred and posed by banking activities in certain areas, please specify those areas and back up your view with specific evidence.
The CRR/CRD IV is an implementation of Basel III into EU law. The Basel III recommendations were aimed explicitly at large international banks.
Unlike its predecessors, Basel III introduced a number of capital buffers, a leverage ratio requirement and requirements for the composition of banks' short-term and long-term liquidity. On top of that, the quality requirements were increased. Apart from raising the total capital requirement significantly, this "multi-layered" approach may to some extent have caused overlapping of regulatory requirements. An overall assessment by legislators of these "overlappings" caused by this multi-layered approach to regulation has not yet been made.
In addition, the consequences of this multi-layered approach are amplified by other new regulatory initiatives with a direct impact on the capital requirements for credit institution, for example the recovery plan requirement and MREL requirement of the BRRD.
In the short term increased capital requirements will lead directly to higher loan costs for borrowers. Therefore the capitalisation level of the individual institution should be considered carefully, partly on the basis of an assessment of the institution's risk profile and business model.
Not all the capital requirements in the new regulation are based on the credit institutions risk-weighted assets. For example, a potential decision to introduce a hard requirement for a leverage ratio or new capital floors for internal rating based approach institutions will not take into consideration the actual risks of institutions. This would seriously impair the competitive condition of low-risk business models relative to business models with medium to high risk. If the leverage ratio requirement is set so high that the implied risk weight exceeds the average risk weight of a low-risk credit institution, the institution would end up having to hold more capital than needed when taking into account the real risk related to its assets.
A high leverage ratio requirement or capital floor requirement would introduce a biased incentive structure where high-risk lending will be more capital efficient than low risk lending. Institution with business models constrained to low risk lending according to their articles of association or by law will be particularly hard hit. Such institutions, for example Danish mortgage banks, will be forced to increase their earnings on loans through higher spreads or fees in order to fulfil the new capital floor requirements or leverage ratio.
Thus, a possible impact of the new regulation is that the most secure types of lending will become more expensive – both in absolute terms and relative to riskier loans.
3. What role have the additional capital requirements and buffers exceeding the harmonised requirements set out in the CRR played in the capitalisation process? Are such additional micro- and macroprudential capital requirements and buffers commensurate to the level of risk incurred and posed by banks? Please back up your view with specific evidence.
The introduction of the countercyclical buffer is a new and additional element of uncertainty in the financial sector. The countercyclical capital buffer is de facto a permanent element of the total capital requirement for commercial and mortgage banks. All in all, the current micro- and macroprudential buffer requirements account for a substantial part of the de facto increase in the total capital requirements.
Seen in relation to loss data for recent years, it raises the question whether the total capital requirement derived from micro- and macroprudential capital requirements is higher than necessary based on the actual risk in the European credit institution industry at present.
4. Have increased capital requirements influenced the overall capacity of banks to lend? Which factors, including demand-side factors, regulatory changes and other supply-side factors (such as the volatility of interbank and capital markets), contributed most significantly to the change in the volume of loans? How do you think bank lending would have developed had regulatory changes to capital requirements not been introduced?
In Denmark total bank and mortgage lending as a percentage of GDP fell marginally at the onset of the financial crisis and then rose slightly again.
The lending costs of Danish commercial and mortgage banks rose as a result of the new regulation. The financial sector in Denmark has generally had to increase margins to build and hold sufficient capital to comply with the increased capital requirements.
On top of that, price differences have emerged in the market - lately as a result of differences in the sizes of covered bond series - which directly results in differences in yield levels and thus in the loan rates of comparable mortgage loans funded by covered bonds (the Danish mortgage model). This is a problem for mortgage banks with fairly small bond series in particular. As described in our comments to Question 6, this is a consequence of the provisions regarding series size under the LCR.
From a Danish perspective, the increased margins resulting from the new regulation do not seem to have had any decisive impact on the demand for loans and thus lending volumes, as interest rates have been historically low.
5. Are the effects of increased capital requirements, such as they are, generally temporary and transitional or have structural changes been seen? Has the requirement to hold higher levels of capital increased the cost of funding banks? Has the per-unit cost of bank capital decreased as banks have become less risky?
The adaptation to the increased capital requirements has put the financial sector under massive pressure. There is thus no doubt that the price of borrowing (the margin to the credit institution) has generally risen due to the sector's efforts to raise the required capital by 2019.
At the same time, the funding costs for all funding sources, measured by spreads against swaps, have risen markedly. Increased requirements for both issuers and investors have led to a demand for higher returns.
Even after 2019, Danish mortgage banks, which provide nearly three quarters of real estate financing in Denmark, will – all other things being equal – have to charge higher administration margins because of the higher capital requirements. This is because the Danish mortgage model is a pass-through model according to which changes in loan rates (the yield-to-maturity of the issued covered bonds) are passed directly through to borrowers. Thus, mortgage banks do not obtain a reduction of their funding costs, but still have to fulfil a higher total return requirement on significantly higher equity as a result of the new regulation.
6. Have increased capital requirements affected the market for some categories of assets more than others? If so, which ones and how? Which of the provisions contained in the CRR, apart from those establishing capital ratios, are likely to have created the effects experienced by specific markets and/or exposures?
Please refer to our comments to Question 2, stating that a possible impact of introducing a minimum leverage ratio requirement is that the competitive conditions of low-risk business models will be seriously impaired relative to those of business models with medium to high risk.
Generally, the introduction of new regulatory measures have reduced liquidity and increased volatility in the global financial markets – even in formerly ultra-liquid markets such as the markets for US and German government bonds. The same trends have been observed in the Danish markets for government and mortgage covered bonds. One reason is that, because of the LCR requirement, investors tend to keep the most liquid bonds, which in reality makes them less liquid. A lack of risk capacity and repo capacity among market participants has also contributed to reducing market liquidity. In general, the sales opportunities for bonds have deteriorated.
In a Danish context, the new liquidity requirements have had consequences in terms of costs both between asset classes and between the individual bond series/securities.
The short-term liquidity requirement of the CRR (LCR) has caused an increase in the costs of bond-funded lending, such as mortgage lending funded by covered bonds. This is due to the fact that most assets, except government bonds, are subject to a "liquidity premium" (in the form of a haircut).
In addition, the requirements for series sizes in some asset classes have contributed to differentiating the costs within the individual asset class.
For example, the requirements for the series sizes of covered bonds used to fund loans secured by mortgages on real estate have led to price differences between Danish covered bonds that are exclusively due to differences in series sizes. At the latest large Danish refinancing auctions in August/September 2015, price differences of 0.6-0.9 point were observed for 3-5-year bonds between the largest series of bullet covered bonds and series that are not expected to reach the volumes required to fulfil the series size criteria for Level 1b assets. Similarly, price differences above 1 point were observed for longer-dated bonds up to 30 years.
Price differences resulting from series sizes are problematic from a mortgage finance perspective, as the mortgage security and the statutory framework are basically the same for large and small series. Thus, the series size requirement under the LCR has a direct and undesirable impact on the price structure in the Danish market for mortgage lending.
The costs related to holding repos have also risen considerably because a leverage ratio (LR) requirement is already factored in. The LCR also has an impact on mortgage banks' repo transactions running for more than 30 days. This is reflected in the reduction in repo market volumes observed recently. At the same time, there has been a distinct shift towards short-term repo transactions, or up to 30 days at the most.
The long-term liquidity requirement (NSFR) may also impact some asset classes more than others to such an extent that it could be anti-competitive. If the NSFR is introduced in accordance with the recommendations of the Basel Committee, one of the consequences will be that loans secured by mortgages on real estate and funded by covered bonds, all other things being equal, will become significantly more expensive than, say, loans secured by mortgages on real estate that are not funded by covered bonds.
The difference in funding costs arises because of a provision intended to prevent excessive asset encumbrance at commercial banks, as this could lead to structural subordination in respect of other creditors. Creditors of specialised institutions which are only permitted to issue loans based on covered bonds are, in the nature of things, not at risk of structural subordination. Nonetheless, the said provision also applies to these specialised institutions, resulting in higher funding costs. Lastly, a requirement for series sizes under the NSFR, relative to the requirement of stable funding on the part of investors, will also result in price differences on covered bonds with different series sizes – like the price differentiation resulting from the LCR requirement already observed.
All in all, a number of the new rules are deemed to have, or potentially have, an adverse impact on the availability of cheap funding in the form of covered bonds. This means that the relative competitive advantage of business models based on covered bonds funding is affected in an unfortunate way by the new requirements. This is not compatible with the aim of creating a Capital Markets Union that can foster growth and employment.
7. Do you think the phase-out of the transitional provisions under CRR could have an incremental impact on future lending decisions? If so, please explain how.
Lending to SMEs
8. To what extent has this provision been effective in supporting lending to SMEs? Could you provide any evidence, preferably quantitative, of the change in lending to SMEs due to the introduction of the supporting factor as from 2014?
In our opinion, the reduced capital requirements on lending to SMEs (SME supporting factor) have the potential to change the limits for SMEs' access to funding. Similarly, the exemption from the CVA charge for non-financial counterparties is important in order to be able to offer risk hedging to small corporates. It is therefore important that credit institutions can continue to apply the SME supporting factor.
However, the impact of the SME supporting factor has been limited by the uncertainty as to whether it will become permanent. This means that several institutions do not factor in the discount in their capital planning and loan decisions. In order for the SME supporting factor to have the desired impact, it has to be a permanent measure.
It should also be noted that the extensive requirements for documentation of SMEs in the form of customer data may in practice significantly put a damper on the use – and thus the impact – of the reduced capital requirements on lending to SMEs.
9. What specific difficulties do banks face when lending to SMEs, compared to when lending to larger corporates? Are these related to the CRR? How could the CRR and other prudential regulations contribute to addressing some of these difficulties in other ways than by adjusting rules for SMEs, or do they need to be resolved by some other means? If so, what other means would be adequate?
Lending to infrastructure
10. Has the CRR influenced the capacity of banks to provide loans to infrastructure projects? Which provisions are most relevant?
11. What are the specific difficulties that banks face when lending to infrastructure projects? Are they related to the CRR? How could the CRR and other prudential regulations contribute to addressing some of these difficulties or do they need to be resolved by some other means? If so, what other means would be adequate?
12. Should infrastructure projects continue to be treated as loans to corporate borrowers? If not, why? What common features of infrastructure projects or their subsets would justify a separate treatment from loans to corporate borrowers?
13. Should the provisions contained in the CRR allow for more differentiation in how they are applied to banks of different sizes or with different risk-profiles? How can they do this without compromising the objective of achieving financial stability and creating a level playing field within the single banking market? Are there any provisions that could potentially be applied with greater differentiation? If so, what are these provisions? Provided application on a differentiated basis is desirable, what considerations could be relevant to make such a differentiated application? Are any concrete changes desirable in this context? If so, what are these changes and the associated costs and benefits?
In our comments to the preceding questions, we have described and documented a number of undesirable consequences of the decision to make the Basel III recommendations binding on the entire European credit institution industry through the CRR and CRD IV. On a more general level, we fear that this "one-size-fits-all" approach will cause the diversity in the European financial industry to diminish over time.
In other words, a narrow range of business models will be those "best suited" for carrying on financial business. Accordingly, the business models that are fundamentally furthest away from the group of the "best suited" will gradually be outdone. The result will be a more homogeneous financial sector. In our view, a financial industry with low diversity is basically undesirable for financial stability purposes.
For example, due to its structure the Danish mortgage model has been difficult to fit into the CRR/CRD IV regulatory framework. The high level of security embedded in this model is actually an outright disadvantage in certain cases. An example is given in our comments on the consequences of implementing an NSFR according to the Basel recommendations.
So, we have the opinion
- that a "one-size-fits-all" approach will be detrimental to certain business models and thereby to diversity
- that a broad implementation of Basel III through the CRR/CRD IV does not necessarily promote the best alternatives, and
- that reduced diversity is not desirable – not least from a financial perspective of stability.
Scope for simplification
14. Which areas of the CRR could be simplified without compromising the Regulation’s objective of ensuring prudence, legal certainty and a level playing field? Are there areas that could be simplified, but only for specific types of bank or business models? Would it be useful to consider an approach where banks that are capitalised well above minimum requirements or that are less exposed to certain risks could be subject to simplified obligations? What would be the risks with such an approach?
Low-risk business models, such as Danish mortgage banks, should be subject to less regulation than high-risk models, or to regulation adapted to the lower risk. Generally, we support the view that the requirements should be risk-based.
The rules on prudent valuation are one example of very complex rules, and it is doubtful whether their actual effect is commensurate with their complexity.
The EBA's Guidelines proposing criteria for setting limits to and monitoring exposures to shadow banking entities are another example of complex and demanding regulation accompanied by reporting requirements which will likely be of limited value, at least as far as Danish conditions are concerned. Moreover, the LCR requirement has had a larger market impact than expected. As described in our comments to Question 6, the LCR requirement has led to decreasing liquidity in bond markets as well as differing costs within the individual types of bond. Against this backdrop, we find that an in-depth analysis should be made of the importance of the LCR requirement to liquidity and price formation in the markets and any negative impacts thereof.
Simplification of the entire reporting requirements section of the CRR should also be considered. The amount of detailed data is substantial and probably also larger than appropriate for some business models. Even on what was intended to be a simple metric such as the leverage ratio, the reporting requirement is substantial. The reporting scope has reached such proportions that transparency is lost. This also implies that it is virtually impossible to coordinate the reporting in different areas.
15. What additional measures could be taken in the area of prudential regulation to further promote integration and enhance a level playing field? Can you indicate specific examples and evidence of discretions that affect the cost and availability of bank lending?
Please refer to our comments about maintaining a certain level of diversity in business models, see Question 13.
We call for EU rules that can accommodate different business models. Otherwise, we consider it necessary that the national implementation of the EU rules allows for national business models. Greater clarity/harmonisation is also needed as regards the practical application of the Pillar II add-on. Binding technical standards should be an option in this area.
Finally, some of the answers in the EBA Q&As express a very narrow and not very feasibility-oriented interpretation of the CRR. One example is EBA Q&A 1263 about A-IRB institutions' use of their own conversion factors. The restriction on the use of own conversion factors for which the EBA argues in their answer does not seem to be in accordance with the intention of the CRR. The EBA's interpretation implies that the capital requirements for guarantees will be set at a level that will not reflect the actual risks. This will render bank guarantees unnecessarily expensive for ordinary businesses. In practice, it is not possible to appeal Q&A answers of the EBA.