Commission services staff working document: Further possible changes to the Capital Requirement Directive (CRD4)

Transcription

1 HUNGARIAN BANKING ASSOCIATION European Commission DG Internal Market and Services Banking and Financial Conglomerates Unit markt-h1c.europa.eu Commission services staff working document: Further possible changes to the Capital Requirement Directive (CRD4) Dear Sir/Madam, The Hungarian Banking Association appreciates the opportunity to express its views on the EU Commission Services consultative document on the possible further changes to the capital requirements directive (CRD4). We remain at your disposal for any further questions or requests for information. Yours sincerely, Budapest, 16th April 2010 Dr. Rezső Nyers H-1051 Budapest, József nádor tér 5-6. H-1368 Budapest, Pf (36-1) , -1867, -1910, -1913, Fax/ : (36-1) bankszovetseg@bankszovetseg.hu Internet homepage:

2 2 REMARKS OF THE HUNGARIAN BANKING ASSOCIATION TO THE CONSULTATION DOCUMENT ON THE EU COMMISION SERVICES ON POSSIBLE FURTHER CHANGES TO THE CAPITAL REQUIREMENTS DIRECTIVE The Hungarian Banking Association appreciates the opportunity to express its views on the EU Commission Services consultative document on the possible further changes to the capital requirements directive (CRD4). The Hungarian Banking Association basically shares the opinion of the European Banking Federation, however some of our members have felt useful to express our opinion on the issues, which have special importance for our banking community. The whole proposal package is very complex and stricter rules are envisaged for capital adequacy measurement both from the side of the calculation of capital and the assessment of risks. It is very difficult to estimate the impact of the proposed changes on the activities of the credit institutions together with the modifications to the capital adequacy framework already in course. The introduction of the CRD4 package at one point in future time may have a credit squeeze effect and may contribute to another crisis. Therefore probably it would be better if the potential measures were introduced gradually, preceded by a strict monitoring phase, and not only by one or two quantitative impact studies. A harmonised reporting and monitoring could help to minimise the negative impacts by introducing the necessary modifications yet before gradual implementation. In order to assess the impact of the new set of regulations it would be essential to have at least one year test period. Consistency between the different planned measures should be further analysed to avoid a divergent set of targets. In our opinion the tougher requirements for credit institutions without any restriction on the activities of the unregulated financial firms may increase the share of the grey zone in the credit market, which is an undesirable phenomenon. From this point of view especially the introduction and the calibration of the net stable funding requirement deserve specific attention (see later). If the liquidity ratios and the leverage ratio were to be introduced, they should apply to all financial institutions. The pro-cyclical effects of some of the new proposals (liquidity ratios, leverage ratio) are not discussed at all, while altogether the counter-cyclical elements of the proposal may be weaker than the pro-cyclical effect. We think that for the calculation of any kind of ratios all government securities denominated in the domestic currencies of any EU Member State which at present in the CRD bear 0% risk weight under the standardised approach, must be treated as highly liquid assets similarly to the AA rated government papers. This issue is not treated consequently in the proposal.

3 3 It is highly questionable to tie the liquidity ratios to the risk weightings of the standardised approach, in which apart from the risk weights applied to the exposures on the sovereigns and perhaps on the bonds of the local authorities (CRD3), the exposures denominated in domestic currencies and foreign currencies bear the same risk weights. In the case of the net stable funding requirement (NSFR) it may have the effect that in the countries rated below A- more stable funds are required for the same type of domestic exposures than in the countries rated A- or higher. The proposal package is based on the ground that credit institutions on the consolidated level as well as on a standalone level use the IFRS and the fair values. References to the Directive 1986/635/EEC related with the annual accounts of banks are not made. The credit valuation adjustments assume that the derivatives appear in the balance sheet, while it is not the case by the standards of the bank accounting directive. It is not clear either, where through the-cycleprovisions (TTC) would be presented on the balance sheet by the standards of the bank accounting directive and what is the relationship between the TTC and the funds for general banking risks. In our view with respect to all proposals a transparent presentation must be established both for the banks using the IFRS and for those that use the national accounting standards based on the bank account directive. It is clear, that the issue of CVA losses is important for all the banks, which use the internal model method (IMM) for the assessment of the exposure-at-default (EAD) for counterparty risk. The issue has also been treated in Annex III of the CRD (Directive 2006/48). However, it is not explained in the document that why the issue of CVA losses are relevant for other banks using the standardised method, the marked-to-market method or the original exposure method, especially if they do not use the fair valuation for the derivatives in their accounts either. By the national standards based on the bank accounting directive counterparty risk may emerge only in the case of OTC derivatives for recognised hedging purposes. Section I Liquidity standards for credit institutions and investment firms General remarks The liquidity regulations and the behaviour of savings are very different in the Member States. There are some Member States where no mandatory one size fits all quantitative liquidity standards exist for market participants. Even, where regulatory quantitative liquidity limits are in force there are considerable differences among them. The quick introduction of mandatory quantitative liquidity standards may obstacle the financing of the economy. For this reason any kind of liquidity ratios should be tested for a longer period, and would be used rather as monitoring tools for the supervisors and not as mandatory ratios the non-compliance with which is sanctioned by the authorities. In our view the stable funding ratio must be treated with special caution, because it is related with the maturity transformation, a core characteristic of the banking business. Maturity transformation is higher in those countries and EU Member States where the capital market is

4 4 less developed and the contractual maturity of the saving deposits is mostly less than one year. A stable funding ratio which is adequate for one country may hinder credit activity and indirectly the economic development in another country. The scope of application of the liquidity ratios is envisaged on a solo and on a consolidated basis. However, sometimes the sub-consolidated group in a Member State manages liquidity on the sub-consolidated group level. In our view banks registered in the EU should have the option to comply with the liquidity ratios on a sub-consolidated basis, instead of a solo basis, provided that the sub-consolidated group members are registered all in the same Member State. We think in this case the risk factors (domestic currency, saving behaviour, etc.) are the same on a sub-consolidated level and on a solo level The intra-group funding should be treated differently from any third-party interbank funding. If the ratios are to be calculated both on the solo/sub-consolidated and on the consolidated level, the treatment of intra-group transactions should be symmetrical. Depending on the composition and the characteristics of the group, the credit institutions should have an option which treatment would apply, namely to consider the unused liquidity line at the institution which extended it, as a cash-outflow and take it into account at the intra-group counterparty as a cash-inflow or not to consider the unused liquidity line at the institution which extended it, as a cashoutflow and not take it into account at the intra-group counterparty as a cash-inflow. Based on our experience in the Hungarian banking sector, where the foreign subsidiaries of foreign (mainly EU) banks have a relevant market share, the distinction of the intra-group funding from any third party interbank funding is justified also by the fact that the short term (less than 1 year) funding from the parent institutions proved to be stable during the crisis. For this reason the short term funds from the parent banks should not be considered as 100% cash out flow in the case of the LCR, because in emergency situation their renewal is highly probable, and short term funds from the parent banks should be taken into account as stable funding in the case of the NSFR, as well. Neither of the two envisaged liquidity ratios do recognize the role of the central institutions with which other credit institutions are associated in a network and which are responsible for cash-clearing operations within the network and/or where the minimum reserves are held in accordance with Regulation (EC) No 1745/2003 of the European Central Bank of 12 September 2003 on the application of minimum reserves or in accordance with national requirements equivalent to that Regulation. In these networks the central institutions usually regard a portion of the deposits with a maturity less than one year from the network as stable funding. During the crisis, it has been demonstrated, that the behaviour of these deposits were fundamentally different from the usual inter-bank funding. However, in case of the central credit institutions of the networks, at the calculation of the proposed ratios the deposits from the members are treated as any other intra-bank deposits from external institutions. We believe, that in case of liquidity measures there should not be an obligation for public disclosure, only reporting requirement towards supervisors. The ratios could be interpreted differently depending country-, institution- or cycle-specific factors, and could be misunderstood and lead to a negative impact on one or other specific institutions. Moreover,

5 5 public disclosure under Pillar 3 follows with a time-lag the situation of the credit institutions, when the liquidity risk may be significantly different than at the time of reference. Meanwhile the two liquidity ratios are defined quite in details, the possible supervisory sanctions in case of non-compliance is not presented. Liquidity coverage ratio (LCR) We fully agree with the approach that the LCR should not be calculated by currencies and that the composition of the liquidity buffer assets should be left to the institutions with regard to their liquidity needs in different currencies. In this respect, we think that the FX swaps also have to be taken into account. In our view in case of the central credit institutions belonging to a network of credit institutions, when assessing the cash outflows, the deposits from the members of the network should be treated equally to less stable deposit from retail and small business customers. It has been demonstrated also during the crisis that in contrast to the inter-bank deposits in general, the inter-bank deposits from the network have been a stable source of funding. (We think that the term central credit institutions in the Annex I refer to the central banks of the Member States and not to the central credit institutions of the networks.) The definition of liquid assets is extremely narrow. All central bank-eligible assets (including bank and corporate bonds) should be recognized (per se) as highly liquid assets taking into account the haircuts defined by the given central bank. The run-off ratios of both the wholesale and the retail funding have not been proved by the last crisis, and are inadequately high. Especially the 100% run-off factor of the unsecured interbank funding seems to be exaggerated, and in our opinion should be reduced substantially. Cash outflow connected with the repayment of loans and other liabilities covered by cash collateral or collaterals qualifying for the liquidity buffer should be left out from the calculation to the extent of the collateralised part, provided that the released collaterals are not shown as cash inflow. Alternatively, if the repayment of these liabilities is presented as cashoutflow, the release of the collaterals should be presented as cash inflow. In this respect cash collaterals and collaterals qualifying for the liquidity buffer should be treated in the same way, as at a given point in time, the high quality securities posted as collaterals are encumbered assets and cannot be a part of the liquidity buffer. Consequently, they should be regarded as cash-inflow. From the point of view of the cash out-flow the stability of the deposits is a crucial issue. The proposal should have a transparent distinction between the stable and the less stable deposits; otherwise the calculations of two institutions with the same characteristics may result in different sizes of the LCR. The minimum regulatory level of the LCR is based on a stress assumption. If one or more stress assumptions are realised, it is not realistic to assume another stress with the same size.

6 6 The consideration of the decay of the stress over time is not reflected in the proposal at all. In our opinion a reduced size of the LCR in stress situation should be treated explicitly in the document. In a deep stress situation it may happen, that a specific institution cannot comply with the LCR requirement. The possible supervisory sanctions should be treated transparently in the proposal. Net stable funding requirement (NSFR) The calibration of the ratio seems to disregard the average effective maturity in the banking sector of the different countries. If introduced without changes, the NSFR may hinder the medium and long-term financing in those countries where the average effective maturity of the deposits is shorter. For this reason, too, the definition of stable deposits is of extreme importance. A quantitative one size fits all approach may have a negative effect in the countries where the average effective maturity of the deposits is shorter. As far as the technical calculation of the ratio is concerned, in case of the central credit institutions the deposits from the network should be treated equally to less stable deposit from retail and small business customers. Furthermore the following assets should be left out from the calculation of the ratio: liquid assets; assets collateralised by liquid assets, to the extent of the collateralised part, as far as the collateral would be eligible for the liquidity buffer and the contract to the counter-party provides the right to the credit institution to use the collaterals for repo transactions; assets re-financed by other banks and which are connected with special development goals. We find the text related to the NSFR and Annex II to the proposal contradicts each other, as far as contingent commitments are concerned. While in point 12 stable funding is required for any contingent contractual and non-contractual obligations, in Annex II only assets are mentioned. The Basel Committee document on International Framework for Liquidity Measurement, Standards and Monitoring leaves the issue of stable funding for contingent commitments to national discretion. In our view, if any harmonisation is envisaged in this respect, the contingent commitment should be taken into account using the CCFs, at least those of the standardised approach. The required funding factors for customer loans should be reduced (both below and above 1 year). It is assumed in the proposal that in the case of loans to non-financial corporate customers 50 per cent of the loans with a remaining maturity of less than one year are subject to renewal, roll-over or substitution with new customer loans. In the case of retail customers the percentage is 85 per cent, even higher. We think that this percentages are influenced to a great extent on the product structure of the given credit institution, but for the institutions with a hypothetical average risk profile, they are too high, also because contingent funding requirement are also taken into account at the calculation.

7 7 Section II Capital The proposal follows the principles of the Basel Committee document on the resilience of the banking sector, but it is not exactly clear how it matches with the changes in the CRD2. Till now in the Basel capital adequacy framework within the Tier 1 capital only the 15 per cent limit exists for the hybrid instruments. In our understanding the paragraph 1.a. of the replaced Article 66 of the Directive 2006/48/EC introduces a 50 per cent limit within the original own funds for the convertible instruments, which consist of a 15 per cent part for the dated instruments and the instruments with incentives to redeem, and another 35 per cent part for other convertible instruments. It is not clear, whether these latter convertible instruments which are undated and have no incentives to redeem, could be eligible for Core Tier 1 or not. In our view it is not justified to detract the minority interests from own funds on a consolidated basis and to include fully the risks from the subsidiaries in the calculation of the capital requirement and the internal capital need. A proportionate solution would be the pro rata consolidation of the risks of the subsidiaries, but this is not desirable for the following reasons. It would entirely separate the prudential consolidation and the accounting consolidation. It would generate extra work within the institutions. There is no evidence that risks of the subsidiaries would be proportionate to the holdings in the subsidiaries. Following from what has been mentioned above we strongly oppose to exclude the minority interests form the calculation of the own funds. There are some cases where the minority shareholders in the subsidiaries belong to the same network, where the majority of the shares belong to the central credit institution of the same network. In such cases the minority interests are not those of external persons, but of credit institutions belonging to the same network. If in contrast to our opinion the minority interest are to be deducted from capital, in these cases it would be absolutely justified to exempt them from the general rule and include these minority interests in the regulatory own funds, i.e. into Core 1 capital. Apart from minority interests in the proposed capital structure it is not very clear how the items stemming from the consolidation differences mentioned in the Article 65 of the Directive 2006/48/EC would be treated. Stock surplus According to ANNEX V stock surplus should be treated the following way: Stock surplus (i.e. share premium) may be included in Core Tier 1 only if the shares giving rise to the stock surplus qualify as Core Tier 1 capital. Stock surpluses relating to shares excluded from Core Tier 1, e.g. preference shares, shall be included in the same elements of capital as the shares to which they relate. We think that stock surplus (share premium) has been paid in cash, irrespective of how the shares, which have originated it, were classified. This reserve is fully available to cover losses;

8 8 it is never paid back, so we do not see any reason to discriminate between the parts of the stock surplus depending on the classification of the related shares. Large exposures As large exposures are concentration risks, in the CRD regime they are covered under Pillar 2. Concentration risks are inter-connected and large exposures have an influence on industry concentration, geographical concentration, too. The large exposure regime has been just reformed by CRD2. A possible new change is justified only by the adjustment to the going concern principle. In our view the impacts of the CRD2 (e.g. inter-bank exposures, connected clients) are not yet clear and any further changes may be harmful for specifically the smaller institutions where the large exposures generally have a greater share. We would like to stress the large exposures are not included in the quantitative impact study in course, as they are not even a part of the consultative package of the Basel Committee. With regard to these, we think that the further changes in the large exposure regime are not of primary importance. However, if the authorities intend to define the large exposures on a stricter basis, in relation to the original own funds, or equity Tier 1, the limits should be reviewed, re-calculated and increased. If the large exposure basis will be the original own funds, the limits should be multiplied at least by the minimum capital adequacy ratio divided by the minimum own funds ratio. Section III Leverage ratio In general we do not exactly understand the rationale behind introducing the absolutely nonrisk sensitive leverage ratio as a supplemental simple measure for capital adequacy. The intentions and strategic goals of the regulation concerning the calibration process are really missing. The introduction of the leverage ratio may depreciate the power of Basle II based regulation which is much more sophisticated and really risk sensitive and do not eliminate the risks which are not captured well by the Basel II-CRD framework. The regulatory practice of the 60s and the 70s has demonstrated that a non-risk sensitive ratio is not adequate to regulate an industry where risk taking is a basic feature of the business activity. We wonder why the idea to introduce a ratio, which is less risk sensitive than the Basel I capital ratio. Under the Basle II regime the world of capital calculation is quite complex with many dimensions. There are 2 pillars with different methodology applied for capital calculation. Furthermore, the changes proposed in the CRD4 package are based on liquidation ( gone concern ) and a going concern approach, which would result in at least 4 different capital requirement/capital need figures. Introducing one more capital related indicator makes the capital management even more difficult. The result could be that relevant functions are

9 9 overwhelmed with reconciliation issues of different measures, understanding different figures, not focusing on the detailed analysis of risk profile and its development. It may be confusing to the public, because among the increased number of capital related ratios it will be more and more difficult to judge which reflects the best the risks of the credit institution. As pointed out under paragraph 79 and 80, leverage ratio is intended to be a "backstop" measure. It is not very clear why such a regime would be necessary for the well-capitalised banks, the capital ratio of which is high, even taking into account the Pillar 2 capital needs and the proposed capital buffer. Therefore, we think that, if the leverage ratio is introduced, it should be only a monitoring ratio for supervisors. Furthermore, it would be crucial to calibrate it with great caution and it seems to be important to include some risk sensitivity, otherwise the new measure may have a negative impact on the development of the entire sector. Our point is that, even if for monitoring purposes, the use of any kind of leverage ratio similar to the one proposed is based on the assumption that the accounting standards used are identical. Moreover additional ratios must be evaluated too, in order to reflect the size of the asset groups with different risk characteristics. The leverage ratio in its proposed form could be equal for two institutions where the risks of entirely different, e.g. because the ratio does not treat at all the risks of the underlying assets. Therefore the ratio of an institution where the majority of the assets consists of derivative contracts could be equal to the ratio of an institution, where the majority of the assets is composed from loans. The same limit for the two institutions favours the institution with the derivative business to the detriment of the institution with the lending business characteristics. This is because in contrast to the stable loan value, the gross fair value of the derivative contracts does not reflect that the value may increase in the future. For the calculation of the ratio it is assumed that all institutions can calculate the gross fair value of the derivative contracts or the replacement cost of the derivative contracts using the marked-to-market method of Annex III in the CRD. However, it has to be considered that unlike the Basel Capital Accord, the CRD has not abolished the possibility to calculate the CCR of the derivative contracts by the original exposure method. Several small institutions, which use derivative contracts only for hedging purposes, still apply the original exposure method for assessing the EAD of the derivatives. In any case, we think that if the ratio is to be calculated, then the liquid assets should be left out from the calculation of the ratio; the inclusion of the off-balance sheet items listed in Annex II of the CRD should be calculated by applying the standardised regulatory CCF for capital adequacy purposes. Section IV Counterparty credit risk Capital add-on for CVA losses

10 10 The credit valuation adjustment in relation with the OTC derivatives is used in the IFRS accounts, but in the national accounting standards based on the Directive 1986/635/EEC they can be related only to the derivatives for hedging purposes. The national accounting standards treat the OTC derivatives as off-balance sheet items, and the value adjustments are related to the value of the hedge, if positive. As many small institution in the EU do not prepare their accounts according to the IFRS, and the annual accounts according to the IFRS are available sometimes only on a consolidated level, we think it would be important to describe the proposal also with the terms of the Directive 1986/635/EEC on the annual accounts of the banks. It is not very clear, whether the banks using the standardised approach for credit risk should also calculate the capital requirement for the credit valuation adjustment, as they have to calculate the capital requirement for the net exposure, i.e. net of value adjustments. It is not explained either that in case of credit institutions where neither the CDS spread nor the bond spread is available, how the credit valuation adjustment should be calculated. In point 113 for the calculation of the additional capital charge to cover mark-to-market unexpected counterparty risk losses for the bond equivalent a holding period of 1 year to be applied, which results 5 times the market risk capital charge of the bond equivalent. Taking into account that for the VAR models at least one year stress period have to be used for the volatility estimation, this may increase considerably the capital requirement for OTC derivatives. In our opinion this treatment is a far too conservative. Asset value correlations The new proposal includes higher asset correlation factor compared to the current requirements for financial institutions. The reasoning behind the change is logical, but the choice of the exact figures is rather arbitrary. On the one hand, this rule results in approximately 30 per cent higher capital requirement for financial institutions in the relevant PD ranges. On the other hand, for rating migration the capital increase is also 30 per cent higher by applying the new correlation factors than it was by using the old ones. This is of high importance for the point-in-time rating systems, which are usually applied internally by banks. In other words, beyond requiring more capital the new asset value correlations cause also higher fluctuations in the capital requirement. This circumstance is definitely contrary to the intention of smoothening pro-cyclicality of the capital requirement. Section V. Counter-cyclical measures General remarks

11 11 The proposal concentrates on the institutions using the IFRS, but does not make clear the relationship of the proposed counter-cyclical measures in terms of the Directive 1986/635/EEC on the annual accounts of banks. It is not clear what would be the relationship between the funds for general banking risks and the proposed measures, and how the capital conservation buffer would be treated from an accounting and a legal point of view. Through-the-cycle provisioning for expected losses in banks using the standardised approach for credit risk Purpose and rationale of risk parameters for through-the-cycle provisioning materially differs from parameters for capital requirement calculation, thus these two sets of risk parameters should be clearly distinguished. For bank using the standardised approach the proposal envisages the embedded PD-s in the risk weights. In our view the proposal is not justified for the following reasons: the standardised approach is not based on expected and unexpected losses concept, the CRD (Basel II) risk weights in the standardised approach were intentionally simplified in order to be not too complicated and to be similar in many respects to the Basel I capital adequacy framework, therefore risk weights reflect very little, if at all, the expected losses; the risk-weights based not differentiate between the real risks of the market participants, which should be reflected in the EL in the case of o the risk weighs based on the sovereign applied for institutions and regional/local governments; o the risk weights for the corporate sector; o the risk weights of the retail sector; the PDs of the same rating grade are not equivalent for the sovereigns, institutions and corporates, and the methodology for the embedded PDs is not transparent; nobody has ever demonstrated that the proposed embedded PDs would be characteristic in any of the Member States. Moreover, we have serious doubts, that in a country, which is in the Credit Quality Step 3 bucket, the realistic EL for institutions would be the same as the EL for corporates. While all exposures to institutions with a maturity over three months bear the same 100 per cent risk weights than corporates. It has to be mentioned that neither the quantitative impact study by CEBS measures the impacts of such a change for the banks using the standardised approach. We should like also to call the attention that in all Member States the credit institutions using the standardised approach have an important role in SME finance. To establish a general provision on the embedded PDs in the risk weights would increase the financing costs to the SME sector. As the standardised approach is not based on an EL-UL approach, we think that it is difficult to establish a general rule for a kind of anti-cyclical general provisions. Therefore we think that the management of the banks using the standardised approach should have a possibility to set aside general provisions, if they think it necessary due to specificities of the portfolio. If supervisors are not satisfied with the size of the general provisions, they have a supervisory tool under the Pillar 2 to require more capital to cover credit risk, provided that they

12 12 demonstrate the standardised approach underestimates the real risks. In our view more steps in the case of the standardised banks are not necessary. A similar approach could be applied in case of IRB banks applying the foundation IRB approach. For these institutions the regulatory LGD is an arbitrary estimation and may not reflect the real risks. In our opinion the management of these institutions should have the choice to decide whether for the TTC they use the regulatory LGD or their own LGD assessment that are used internally. If in specific cases supervisors are not satisfied with the assessment, they could require more capital under the Pillar 2. Tax treatment of through-the-cycle provisions Through-the-cycle provisions, if introduced should be before tax in all Member States, since this is an ex-ante provision for specific losses. The tax treatment should be the same in all EU Member States. For tax differences the proposed treatment of deferred tax assets could only be a partial solution, since this item on the solo level does not exists. Capital conservation buffer It is not exactly clear how the capital conservation buffer would work. 1. In our understanding the capital conservation standards are bank specific, and they are built on the minimum capital requirement, which is disclosed. However, it is not exactly clear, what is the relationship between the capital requirement taking into account the SREP and the level of the capital conservation standard. For capital conservation buffers Pillar I and Pillar II capital requirements should be clearly distinguished. 2. The capital conservation standard would be applied on a consolidated level. We do not understand how the dividend payment restriction on a solo level could be reconciled with the standard on a consolidation level. This is an important issue, also because in the subsidiaries the general meeting is usually earlier than in the parent company. Therefore it would be very difficult to calculate exactly how much dividend should be retained in the subsidiaries. 3. The counter-cyclicality of the buffer would be expressed by an add-on, which means in our view the increase or decrease of the capital conservation standard. Even if clear rules are defined when and to what extent the capital conservation level is to be increased or decreased, capital planning would be very difficult because the uncertainty of the required regulatory capital level in the medium term. 4. From an accounting point of view the undistributed profit should be a part of the other reserves, because it is a legal reserve which could be used only for covering losses. However, since it is a reserve, it is an element of own funds, too, and if so, it can generate more business. The counter-cyclical effect of the buffer for this reason is not clear. 5. According to the proposal, capital should be increased after the first signs to a coming downturn, which may strengthen pro-cyclicity. As far as the time limit for reaching the capital target level in case of non-compliance, we think that harmonisation across EU would help to encourage fair and equivalent conditions in all Member States.

13 13 Cyclicality of the minimum capital requirement The use of a downturn PD (one candidate would be the historically highest, as proposed in paragraph 164.) in the capital calculation can mislead the risk consciousness of banks. The Pillar 1 model is designed to require a long run average PD as input, which is - under the model's assumptions - transformed to a 99.9% confidence level PD in the formula. If the formula is fed with a downturn PD this way, the confidence level of the capital requirement becomes meaningless - it cannot be determined on which confidence level the bank's capital can absorb losses. Furthermore, the corresponding confidence level is increased significantly, which results in further unnecessary appreciation of conservatism regarding capital calculation. Long run average PDs themselves are stable enough in time. However point-in-time rating systems are usually applied by banks, because they are in congruence with business usability and limited sample for development. The current Basle II approach also permits to apply point-in-time systems. Based on this circumstance, even if long run average PDs are stable in time for a certain rating class, the rating grade of a certain customer can change significantly, therefore the corresponding capital requirement also highly fluctuates. We think that the current proposal does not address this issue appropriately, the detailed approach of macroeconomic model based adjustment, or moving average based smoothing can be the solution. These concepts are just mentioned in the document, so in our view they should be definitely developed further. The proposal does not address the issue of using point-in-time versus through-the-cycle rating systems. Since the time horizon is quite different on which these grades are indicative, therefore the approach of capital requirement calculation also defers. However, there is no distinction in the formula used under Basle II depending on the type of rating system. (For a through-the-cycle rating system the same PD level would indicate higher capital requirement, since the default rate fluctuation is much higher compared to a point-in-time system, which reflects mainly short run behaviour.) The pro-cyclical effect is relevant for capital requirement, provisioning and market value calculation. These issues should be jointly treated to mitigate the fluctuation of financial condition. Section VII. Single rule book It is not very clear, what does the Income in the Loan To Income (LTI) ratio mentioned in point 175 mean. Does it mean annual/monthly income of the private individual? In addition, RRE covered loans can be extended even to non-retail customers. How would be the income here defined? In our opinion any LTV limit should not mean that a loan, as a whole, can not be treated as real estate covered, if the LTV based on the total exposure of the deal and the value of the RE is above the LTV limit indicated. Applying LTV limits should only define that proportion of the exposure, which can be treated as covered, while the remaining part as uncovered.

14 14 For the same reason we suggest to eliminate even C* defined in Annex VIII, Part 3, paragraph 69, as it causes an inconsistency between the following two possible calculation method: Exposure=100 RRE=14 C*=14%, i.e. below the threshold to be treated the exposure as being collateralised Treatment 1: The exposure is treated as one deal, where C*=14%, so the total exposure (100) is uncovered Treatment 2: The exposure is split in two smaller parts e.g. 10 and 90, so the exposure=10 will be fully recovered, and only the remaining exposure of 90 will remain uncovered. The rule of C* results in different RWA for the two cases, which is an inconsistency.