Response to the European Commission's consultation on further possible changes to the capital requirements directive ('CRD IV')
Finanstilsynet og Norges Banks brev av 16. april 2010 til DG Internal Market and Services, Banking and Financial Conglomerates Unit, European Commission
1. On 26 February 2010, the European Commission launched a public consultation on further possible changes to the capital requirements directive (CRD IV). This consultative statement was prepared jointly by Norges Bank and Finanstilsynet. Norway is taking part in the quantitative impact study coordinated by the Committee of the European Banking Supervisors (CEBS), and will consider additional comments on the background of results from this exercise.
2. Norges Bank and Finanstilsynet broadly endorse the Commission’s proposals. The proposals address a wide range of problems and deficiencies that have been identified in the current regulation of bank capital and liquidity management. However, we note that the existing and proposed regulations result in a comprehensive and complex regulatory system that entails significant challenges for both banks and authorities. The international financial crisis has clearly again demonstrated that the costs to society of financial instability can be very high. The proposed increased capital and liquidity requirements will contribute to reducing systemic risk, but will at the same time increase the costs of financial services for households and non-financial companies. These costs may delay the recovery and economic growth after the crisis. The final calibration and timing of implementation must carefully address the balance between these considerations. The cost of compliance is likely to vary between jurisdictions, and is at risk of becoming particularly high in jurisdictions with limited availability of certain assets and funding sources.
3. Norges Bank and Finanstilsynet generally support the Commission’s efforts to improve transparency and market discipline by requiring banks to disclose metrics showing how they comply with the standards.
4. Norwegian banks experienced funding problems during the recent financial crisis. The problems were alleviated by extraordinary liquidity measures on the part of the Norwegian Government and Norges Bank. Norwegian banks had limited exposure to structured credit or the type of securitisation and derivative transactions that caused solvency problems internationally. Norwegian authorities have applied rather strict rules and supervision after the Norwegian banking crisis in the early 1990s. Together with relatively prudent credit practice and emphasis on risk control on the part of the banks, this may have contributed to mitigating the impact of the international financial crisis on the Norwegian financial sector.
Strengthening the resilience of the banking sector
5. Norges Bank and Finanstilsynet generally support the proposals to strengthen the resilience of the banking sector. The effect of many of the proposals, and the joint effect, will depend largely on calibration issues that are not settled at this point. Specifically, the proposals aimed at reducing procyclicality and excessive credit growth, require further investigation.
Quality, consistency and transparency of the capital base
6. Norges Bank and Finanstilsynet welcome measures to enhance the quality, consistency and transparency of the capital base. In particular, we support setting a separate quantitative requirement for common equity. Ideally, we would like the Commission to consider removing hybrid capital altogether from tier 1 capital. Alternatively, we recommend that the quantitative requirements for common equity, tier 1 capital and total capital are calibrated rather strictly and more strictly than existing regulation. Sufficient weight should be put on the common equity component (including retained earnings). As long as there are hybrid instruments in going-concern capital, we strongly urge that the minimum standard includes announced trigger points for write-downs and conversion of hybrid instruments into equity. The trigger points should be set with sufficient distance to the minimum regulatory capital adequacy requirement. A bank will need to be recapitalised long before capital reaches the minimum regulatory requirement. If hybrid instruments do not contain sufficient loss absorbency mechanisms, it will be more difficult to recapitalise the bank since the shareholders will bear all the losses. We strongly suggest that the requirements for loss absorption on a going-concern basis explicitly apply for all tier 1 capital (not being common equity), not only for instruments that under the relevant accounting rules are recognised as liabilities.
7. We note that the Commission is particularly aware of the special adjustments of the proposed measures that need to be taken in order to encompass co-operatives and savings banks without common stock. Norwegian banks are not likely to be significantly affected by the proposals, as Norwegian legislation applies rather strict rules for both the composition and amount of tier 1 capital.
8. We strongly support that prudential filters and deductions are generally made at the level of common equity including retained earnings.
Risk coverage and counterparty credit risk
9. Norges Bank and Finanstilsynet support the decision to increase the risk weight for exposures to financial institutions. Furthermore, we agree that the issues of wrong-way risk and credit valuation adjustments are lessons at the heart of the propagation mechanisms that were at work during the financial turmoil in 2008/2009.
10. We support the efforts to strengthen incentives for using central counterparties. The strengthened incentives should be accompanied by internationally harmonised regulation of central counterparties.
11. Norges Bank and Finanstilsynet agree that the proposed measures for risk coverage and counterparty credit risk may alleviate certain systemic risk issues. We endorse that the Commission work on possible policy approaches to ensure that prudential requirements for financial institutions are commensurate with the risks they pose to the financial system.
Norges Bank and Finanstilsynet strongly support that improved resolution tools are key requisites for the treatment of systemically important financial institutions. Moreover, we believe all cross-border banks should be required to have a ”living will” in place, not only the largest ones. That way, one avoids, at least for this purpose, having to decide which cross-border banks are systemic and which are not. For small banks with uncomplicated structures, setting up a living will involves minor costs. 
12. We believe systemic risk and macroprudential issues should be given a more pronounced and broader treatment in the final reform package. Pillar 1 of Basel II framework is based on loss experiences at the level of individual banks. Experience, including the recent financial imbalances, suggests the existence of externalities in the financial system that magnify the interface between the financial system and the real economy. Leverage in the household sector is one example. While a single bank may be faced with limited credit risk on its mortgage portfolio, high leverage in the household sector makes the economy vulnerable. Economic disturbances will more likely lead to losses on loans to corporates that are dependent on domestic demand. We urge the Commission to investigate whether such externalities can be better accounted for within the regulatory framework. In the case of housing loans, risk weights differ significantly between IRB banks and banks applying the standardised approach. We doubt the underlying risks are accordingly different, especially for loans with moderate loan-to-value ratio. A housing loan is a fairly homogenous financial product that should be subject to relatively homogenous capital requirements. We therefore urge the Commission to investigate whether there is a basis for recalibration of the IRB risk weight curve for housing loans, with the aim of aligning the risk weights under the IRB approach with the standardised approach. We see the introduction of a leverage ratio as an important measure to contain the macroprudential risks for high levels of housing loans.
13. Norges Bank and Finanstilsynet support the introduction of specific requirements when loans to real estate can receive preferential risk weight. Finanstilsynet has recently issued similar, written recommendations to banks operating in Norway.
There seems to be an inconsistency between the definition applied for loan to value and loan to income on page 60 in the consultation. While loan to value presumably is computed with loan as the nominator, loan to income seems to be computed with loan as the denominator. The range that is indicated, 33.3%-50%, is extremely low for the loan to income ratio, if loan is the nominator. We assume this range is meant for a ratio with loan as the denominator and income as the nominator. If so, we find the range sensible.
14. Norges Bank and Finanstilsynet support to impose a leverage ratio as a means to set up a safety net against model risk and measurement error. We are generally supportive of the proposed design and the efforts made for ensuring comparable calculations across institutions and jurisdictions.
Basel II has introduced new challenges to comparing banks, the main reasons being differences in risk assessment between standard and IRB banks and differences in how banks report the effect of the contemporary floors. The leverage ratio may have a complementary role in comparisons between banks.
Calibration of the leverage ratio is crucial. Norges Bank and Finanstilsynet stress that the leverage ratio should be a supplement to the risk-weighted capital requirement. The risk-weighted requirement should be binding for most institutions without the leverage ratio inflicting. In general, we support that leverage ratios be applied both at the individual entity and consolidated level. However, the calibration and exact implementation at the individual level should be carefully based on the results of the impact studies.
15. While Norges Bank and Finanstilsynet find merit in each of the proposals to mitigate procyclicality, we recognise the need for further investigation of the effect of each proposal and in particular the joint effect. A key premise is the planned assessment of the degree of procyclicality of Basel II, and to what extent IRB models should be adjusted for the cycle.
Procyclicality and cycle adjustments
16. Norges Bank and Finanstilsynet agree that a certain degree of cyclicality in minimum capital requirements is acceptable in order to preserve IRB risk sensitivity across institutions at each point in time. We support to continue the ongoing research to assess the degree of Basel II cyclicality. We believe supervisors should increase their effort to require IRB banks to rate through the cycle in a forward looking manner. A harmonisation of international practice may be warranted. We recognise that the use of cycle adjustment methods to assess downturn probability of default for different exposure classes could provide useful benchmarks for national supervisors under pillar 2 of Basel II.
Procyclicality and forward looking provisioning
17. Norges Bank and Finanstilsynet firmly support the Commission to continue its work to promote a change in provisioning standards that improves the usefulness and relevance of financial reporting for stakeholders, including prudential regulators. An accounting model for provisioning should allow for earlier identification and recognition of losses compared with the prevailing incurred loss model. Effort should be taken to limit the complexity of the accounting model to ensure transparency and to avoid obstacles to implementation. For IRB banks, imposing a separate accounting based model to assess expected losses will result in multiple definitions and a duplication of tasks. Therefore, Norges Bank and Finanstilsynet support that the Commission investigates a model based on expected loss estimates from banks’ IRB models.
Reduced procyclicality through capital conservation
18. Norges Bank and Finanstilsynet support the proposed framework of establishing a buffer range above the regulatory minimum, in which banks’ discretionary payments and share buy-backs are restricted. As proposed by the Commission, restrictions on discretionary payments should explicitly include payments to hybrid instruments in tier 1 capital.
Varying the buffer with macro indicators to mitigate excessive credit growth
19. Norges Bank and Finanstilsynet support the effort to find applicable measures to mitigate excess credit growth. We believe varying the capital buffer with macro indicators is a viable option. Macro credit growth relative to trend could be a suitable indicator for this purpose. However, there are challenges to tying the buffer range to macro variables.
First, there are challenges to applying varying capital buffer on cross-border banks. The macroeconomic environment, including credit growth, typically varies across countries. The Commission proposes that internationally active banks should have buffer capital commensurate with a weighted average of the buffer requirements in the jurisdictions the bank operates. To ensure level playing field, the weighted average buffer should be designed such that increased capital required by issuing a new loan in a certain jurisdiction is the same whether the loan is provided by domestic banks or foreign branches. Furthermore, the weighted average buffer requires cooperation between home and host authorities, as the home authorities are responsible for capital requirements.
Second, a varying buffer requirement may become complex and difficult to predict. We believe a capital buffer above the already advanced and complex risk-weighted minimum requirement should be kept simple. For example, with a varying buffer that restricts dividends, it will be difficult for investors to form expectations about future dividends.
Third, in periods of high credit growth, banks with excess capital may choose to increase lending further, due to the increased buffer range restricting dividends. When the variable buffer is reduced in periods of low credit growth, restrictions on dividends will ease and banks still running a surplus may pay higher than prudent dividends. Thus the effect may turn out to be the exact opposite of the intentions.
A virtue of the capital buffer is its simplicity relative to through-the-cycle provisioning. However, banks will nevertheless have to report loss provisioning in the accounting statements. Even with a capital buffer in place, one might argue that loss provisioning should be sufficiently through-the-cycle to reduce the procyclicality of bank earnings. Capital buffers will not to the same extent reduce volatility of earnings, at least during periods when the buffer requirement is fulfilled. However, we are concerned that if the proposal is implemented above the line, it will interfere with the ongoing work by IASB.
Based on the above considerations we encourage the Commission to further investigate the likely consequences of tying the buffer range to macro variables. Furthermore, we suggest considering a wider range of measures to mitigate excessive credit growth. National supervisors can already use pillar 2 of Basel II to address macroprudential concerns. Further work by the Commission on mitigation of excessive credit growth could be used as a benchmark for applying pillar 2 requirements for macroprudential purposes.
20. The Commission proposes to supplement the newly adopted qualitative liquidity standards with quantitative standards specifying minimum levels of liquid assets and longer-term structural funding. Norges Bank and Finanstilsynet support the introduction of such requirements.
Scope of application
21. We agree that the liquidity standards shall apply at both consolidated and individual entity level. The proposal does not encompass a mandatory sub-consolidation at the level of a parent in a Member State where that parent is itself a subsidiary of a bank in another Member State. In our view the scope should be equivalent to that of the solvency requirements. Such sub-consolidation should therefore be mandatory. We agree that there should be a waiver to derogate from the individual entity requirement if strict conditions are met (i.a. no constraints on transferability of assets, binding mutual support commitments and central liquidity risk management).
22. We agree that the liquidity standards shall apply to all credit institutions and certain investment firms.
Level of detail
23. The proposed liquidity standards are specified with a very high level of detail. It is our concern that this raises a range of issues. Some of these are outlined in 23a-23c.
Level of detail and the relevance for future liquidity crisis
23a. Complex and highly specific regulatory systems are known to be most appropriate when there is very specific knowledge about the type of problematic situations that may arise. Liquidity risk seems to materialise in a wide range of ways, and often with an element of surprise to regulators and banks. When imposing a very detail-oriented regulatory system there may be a greater risk of over-emphasising the most recent experiences.
Implementation may be difficult due to differences between jurisdictions
23b. In order to preserve a level playing field Norges Bank and Finanstilsynet support the need for regulatory harmonisation of both capital and liquidity standards. However, the availability of specific assets and funding sources may vary extensively over jurisdictions. Thus, harmonisation of liquidity standards may be difficult when the regulatory framework is specified at a very detailed level.
Very detailed standards may reduce diversity among institutions
23c.When there is genuine uncertainty concerning the exact nature of future shocks that may hit the financial sector, having a diverse bank sector may be an important factor for robustness. We are somewhat concerned that the ability of institutions to comply with the proposed regulatory framework will depend largely on size and type of business activity. Small and domestically oriented institutions, and in particular those operating in jurisdictions with limited availability of specific assets and funding sources, may find it particularly difficult or costly to comply. As a result the banking sector may evolve over time in the direction of less diversity.
Calibration of stress tests
24. The recent financial crisis left little doubt that liquidity management and resilience to liquidity shocks must be improved substantially in financial institutions. The severity of the stress test scenarios applied should be carefully balanced against a consideration of what type of shocks the sector should be able to withstand without extraordinary policy actions taking place.
Minimum standards and the need to consider flexibility
25. Norges Bank and Finanstilsynet support that the minimum standards relating to the Liquidity Coverage Ratio (LCR) and Net Stable Funding Ratio (NSFR) should be defined with a view to ensuring flexibility in times of stress. In particular it is of importance that LCR is a liquidity buffer that can actually be drawn on. If standards are formulated as an absolute minimum, this will not be the case as the standards must be upheld at all times. We therefore agree with the proposed text stating that under stress credit institutions could fail to meet the requirements. In such circumstances, credit institutions would be required to restore compliance over a short timeframe. We agree that competent authorities in such cases should be required to define a restoration plan and to follow up its implementation. Such a flexible regime could also include a “duty to tell” the supervisor when departure from the standards occurs, or is likely to occur in the near future.
26. The Commission outlines different alternatives for the treatment of intra-group transactions at individual entity level, and favours a symmetrical treatment under the LCR (i.e. a consistent treatment for the two entities involved). We support this approach. Symmetrical treatment in a prudent way should assume that intra-group lines of credit are not drawn upon under stressed conditions. This principle should apply to both LCR and NSFR.
Partly owned firms and calculation on a consolidated basis
27. Under the proposed standards, it could be the case that a bank has on its balance sheet a partly owned firm that falls outside the consolidation standards, but still represents a major liquidity risk to the bank. It is important to ensure that the regulatory framework is such that the share of the liquidity risk that can be related to the bank’s ownership is included in the proposed indicators (LCR and NSFR). An interesting example taken from the Norwegian banking sector regards part ownership in a credit institution whose purpose is to issue covered bonds. In this particular case, the ownership is divided among an alliance of legally independent savings banks that can transfer loans to the credit institution as a means to obtain funding.
Liquidity coverage ratio
28. Norges Bank and Finanstilsynet are supportive of the proposed ratio. We believe it could be successful in ensuring that banks hold a sufficient amount of liquid assets within a time horizon of 30 days.
Stock of high quality assets
29. Norges Bank and Finanstilsynet underline the importance of formulating the listed criteria for “high quality assets” to enable banks to hold assets with reasonably stable access in stressed situations. The particular formulation that banks should hold liquid assets denominated in the same currency as the liquidity risk exposure, assumes reasonable availability of such assets. For Norwegian banks, a major part of such assets will have to be government bonds or central bank deposits. In Norway, the market for government bonds is particularly small and illiquid (which can be explained by the fact that the Norwegian government is a net positive asset holder). Norwegian government bonds are primarily held by foreign investors and life insurance companies. Demand for government bonds is likely to increase further, and significantly so, when Solvency 2 is implemented. Thus government bonds will only to a limited degree be available for inclusion in the proposed liquidity buffer.
30. For the extended definition of liquid assets, which includes a share of non-financial corporate bonds and covered bonds, the standards are so strict that in practice there will be no assets available to Norwegian banks denominated in the domestic currency. Thus in practical terms, the narrow and the extended definition coincide.
31. Norges Bank and Finanstilsynet suggest that the proposed standards be modified. We believe that the way forward is to put more emphasis on diversification, so that a wider definition can be applied. The share of government bonds which banks are required to hold must be consistent with the availability of these assets. For other assets, it is vital that the requirements are not so strict that too few available assets qualify. It is important that covered bonds with sufficiently low credit risk are included.
Net cash flows
32. By international standards, the share of deposits to total assets of Norwegian banks is very high. The assumptions regarding the run-off factor for deposits will therefore be crucial for the cost of compliance. The risk of deposit run-offs has been observed to vary over different jurisdictions. Possible factors like savings patterns, variation in deposit insurance agreements, as well as variations in public confidence in institutions and authorities may explain this observation. In Norway insured deposits were reasonably stable during both the Norwegian banking crisis in 1988-1992, the financial turmoil in 2002/2003, as well as under the recent financial crisis. The main pattern was then that depositors fractionalised over a larger number of banks to qualify for the deposit insurance, causing only limited net effects in each bank. Against this background, we believe that the run-off factors for deposits not made by financial institutions, insured deposits in particular, are set too high. If effectuated they are prone to initiate structural changes in the banking sector, which may be unwarranted. For instance they may induce smaller banks to replace deposits with less familiar funding sources. Furthermore, we believe that criteria for classifying deposits such as “banking relations” and “transactional accounts” should be dropped so that all types of insured deposits can receive similar treatment. “Banking relations” are subjective and not particularly transparent. “Transactional accounts” are more flexible, not less flexible than other accounts. The classification of different types of deposits may vary substantially between countries.
Net stable funding ratio
33. Norges Bank and Finanstilsynet support the structure of the proposed ratio. We believe that it can contribute to banks holding a sufficient amount of long-term funding. However, the specified time horizon of one year implies that debt maturing in just over one year is as stable as longer-term debt. The indicator measures the value of stable funding relative to an estimate of necessary stable funding. Again we wish to stress the importance of deposits as a source of financing for Norwegian banks. The available stable funding (ASF) factors, which capture the share of each financing source assumed available in a stressed situation, will have to be considered with care. We are concerned that the ASF factors for deposits are too conservative.
Supervisory responsibility for branch liquidity
34. Norges Bank and Finanstilsynet agree with the conclusion that for credit institutions with significant branches or cross-border services in another Member State, liquidity supervision should be the responsibility of the home Member State, in close collaboration with the host Member States.
35. Institutions should provide public disclosure of information on metrics and standards, similar to the disclosure of capital standards. More transparency means fewer banks will choose business models with high liquidity risk. A minimum international disclosure standard should be established.
Ensuring consistency with Basel
36. The Commission believes the lack of prudential policy coordination at EU level should be addressed in a way that is consistent with international developments, in particular the proposals tabled by the Basel Committee. We appreciate that the proposed EU liquidity standards are closely aligned with Basel. Some discrepancy is natural due to the proportionality principle. However, some deviations exist that do not seem to be intended. In Annex II of the consultation paper off balance sheet exposures (undrawn credit and liquidity facilities) are not mentioned in the list of items requiring stable funding. However, paragraph 12 states that contingent obligations have to be matched by stable sources of funding. Moreover, in Annex II the term “credit institution” replaces Basel’s term “bank”. A consequence of that is that covered bonds may require more stable funding in the Commission’s proposal than in Basel’s. We therefore recommend that the complete proposal is checked for unintended deviations.
Bjørn Skogstad Aamo (Director General, Finanstilsynet)
Svein Gjedrem (Governor, Norges Bank)
- Confer Norges Bank's letter of 20 Januar 2010 to The European Commission, DG Internal Market and Services: Comments on the Commission Communication [COM(2009) 561/4]: An EU framework for Cross-Border Crisis Management in the Banking Sector
Copy: The Norwegian Ministry of Finance