Response to consultation on guidelines for common procedures and methodologies for the supervisory review and evaluation process (SREP)
Go back
It is clear that more consistency in the supervisory methods across Europe needs to be achieved. However, more consistency should at the same time not impose restrictions on the management bodies by establishing too detailed rules. It might be dangerous that a too uniform and rule-based framework leads to having very similar institutions, which could finally increase systemic risks and procyclical effects.
Moreover, it is important that the size and the complexity of banking institutions are appropriately taken into consideration when establishing the SREP GL. Retail-oriented institutions should not be pushed too hard with e.g. documentation requirements, but rather be treated in suitable accordance with the principle of proportionality.
In general, ESBG fears that the SREP GL will contribute to the fact that supervision becomes more expensive, in particular for those institutions which are not directly supervised by the ECB. In this connection, some estimations and conclusions of the draft-cost-benefit analysis (section 5.1) are not entirely comprehensible to us.
In conclusion, ESBG very much appreciates that the principle of proportionality has been considered in various parts of the draft SREP GL. However, we feel that the proportionality principle could also become a lot more visible in some sections of the GL.
ESBG understands that the business model analysis focuses essentially on the short term. The methodology proposes to assess the viability of a business model by its ability to generate acceptable returns over the following 12 months. However, this approach, which is overly sensitive to cyclical effects, may tend to produce procyclical assessments. A preferable alternative might be to jointly assess viability and sustainability of a business model and an institution’s strategy on the basis of its capacity to generate acceptable returns over the medium term. The assessment should be made based on the likely trajectory of profitability/solvency/risk over the current year and the following 3-4 years. By the end of this projection period, cyclical effects should not be expected to play a significant role in the results.
In point 59a, “materiality of business lines” shall be defined by their strategic importance, no matter how profitable these business lines are. Strategic importance is not only measured in terms of profit/loss generation.
The “considerations” in point 76 might result in institutions following universal banking models having better scores than specialised institutions. In order to be scored with a “1”, no “material asset and funding concentrations” are allowed. ESBG sees this statement fairly critical as specialised institutions often offer services of higher quality due to their degree of specialisation. At the same time they can definitely have an excellent business model. As a consequence, ESBG suggests introducing the element of specialisation in the “considerations” in point 76.
Concerning the IRRBB itself, ESBG calls on the EBA to deal with this topic in the specific guidelines on the IRRBB which are being developed by the EBA. Nevertheless, ESBG would like to highlight that in the current interest rate environment, a 200 basis points shock as proposed by EBA in point 293 of the consultation paper, is not necessarily tailored to all circumstances and rate environments, and therefore it would be more suitable to change the wording of +/- 200 basis point for “standard shock”.
ESBG strongly believes that it would be better to continue to allow for the use of the internal model approach. On the metrics debate, ESBG supports that the importance of both the earnings as well as the economic value approach should be recognised in any future regime. The choice of the metrics should be consistent with the risk management strategy and risk appetite of the institution.
To ESBG’s mind, the subcategories listed in point 244 are contradictory to the Advanced Measurement Approach (AMA)-categorisation. For instance, conduct risk does not play a role in the AMA-categorisation.
It is indeed difficult to compare IT-systems and IT-architectures among banking institutions as there are no harmonised standards to measure their complexity. Thus, if IT-complexity is included in the SREP GL, common and comprehensive assessment criteria will be more than useful.
There is no need to have reputational risk as a separate category. As reputational risk is regularly considered a part of operational risk, it should be treated as such within the SREP GL.
ESBG doubts whether it is necessary to differentiate between the 4 different types of risk in point 280. This might be a suitable theoretical approach. In practice, however, these risks are usually not dealt with in a separate way.
More precisely, as we understand it, the new approach would require a comparison per risk between the pillar 1 capital charges, the capital requirement calculated using the bank’s internal risk model and an amount calculated using a prudential benchmark model. We presume that the highest of these three amounts would be taken as a (worst-case) yardstick. If it was above the pillar 1 capital charge, an add-on for the difference would be imposed. These additional capital requirements for credit risk, market risk and operational risk would be added together and then further supplemented with a capital requirement for interest rate risk, model risk and reputational risk. No diversification effects would be taken into account. The objective of this exercise would be to calculate the worst possible overall risk scenario facing the bank. We do not believe that this would be a viable approach suitable for the management of a bank. Pillar 2 would be diverted from its intended purpose and downgraded to a “pillar 1 plus”.
We consider a more appropriate approach that the bank calculates its overall need for economic capital (taking into account diversification effect) and to discuss the outcome with supervisors. Then the results of this process must be compared with the overall capital requirements determined under pillar 1. This approach is already in place in a couple of EU countries.
Specifically, the CRD IV explicitly uses the term “internal capital” (Article 73), while the SREP GL apply the term “regulatory own funds” to pillar 2. While article 104 (1a) CRD IV gives competent authorities the power to determine additional own funds requirements, we construe this power to be restricted to singular cases only. In other words, article 104 (1a) CRD IV does not affect the internal capital definition of article 73. In summary, the proposed limitation to own funds is not in line with the CRD IV’s internal capital definition and point 337 should be amended accordingly.
In ESBG’s opinion it is not necessary to also include credit concentration risk in point 335b as it is already part of the internal models determining credit risk.
With regard to point 320, ESBG would like to state that comparability of institutions in the context of ICAAP will not always be possible. For instance, there are differences in the parameterisation of the risk calculation concerning the risk capacity as “going concern” (in contrast to a “gone concern” approach).
Moreover, the additional capital requirements should not be determined by simply adding the capital needs of each risk category without allowing for diversification.
Q2. Do you agree with the proportionate approach to the application of the SREP to different categories of institutions? (Title 2)
Proportionality is a crucial element to be followed in the establishment of the SREP GL: In the past, good and efficient supervision has always been linked to proportional supervisory principles. Bearing mind the need for a consistent and coherent regulatory framework, in pillar 2, principle-based approaches have often proven to be more suitable than rule-based approaches as the former can easier adapt to all different types of banking institutions. It is to be feared that rule-based approaches might lead to regulatory gaps, which would be beneficial to some institutions (lower capital requirements), but also unfavourable to many other institutions (higher capital requirements). The very diverse banking sector in Europe calls for proportionality and a principle-based approach in order that all institutions, having different business models and risk profiles, are covered in an appropriate way by one common supervisory framework, namely the SREP GL.It is clear that more consistency in the supervisory methods across Europe needs to be achieved. However, more consistency should at the same time not impose restrictions on the management bodies by establishing too detailed rules. It might be dangerous that a too uniform and rule-based framework leads to having very similar institutions, which could finally increase systemic risks and procyclical effects.
Moreover, it is important that the size and the complexity of banking institutions are appropriately taken into consideration when establishing the SREP GL. Retail-oriented institutions should not be pushed too hard with e.g. documentation requirements, but rather be treated in suitable accordance with the principle of proportionality.
In general, ESBG fears that the SREP GL will contribute to the fact that supervision becomes more expensive, in particular for those institutions which are not directly supervised by the ECB. In this connection, some estimations and conclusions of the draft-cost-benefit analysis (section 5.1) are not entirely comprehensible to us.
In conclusion, ESBG very much appreciates that the principle of proportionality has been considered in various parts of the draft SREP GL. However, we feel that the proportionality principle could also become a lot more visible in some sections of the GL.
Q3. Are there other drivers of business model / strategy success and failure that you believe competent authorities should consider when conducting the BMA? (Title 4)
In general, the SREP GL will introduce many more competences for the supervisory bodies. They will gain several additional powers to intervene. It is crucial to strike the right balance. In this regard, ESBG would like to emphasise that the NCAs’ competence to conduct a BMA should not lead to the result that they end up having a say in the institutions’ business policy. Supervisors should not take responsibility for the institutions’ business policies – this shall rather remain a task of the institutions’ management bodies. The scoring of business model and strategy should therefore be left out or used only as support when considering the other SREP components and proportionality.ESBG understands that the business model analysis focuses essentially on the short term. The methodology proposes to assess the viability of a business model by its ability to generate acceptable returns over the following 12 months. However, this approach, which is overly sensitive to cyclical effects, may tend to produce procyclical assessments. A preferable alternative might be to jointly assess viability and sustainability of a business model and an institution’s strategy on the basis of its capacity to generate acceptable returns over the medium term. The assessment should be made based on the likely trajectory of profitability/solvency/risk over the current year and the following 3-4 years. By the end of this projection period, cyclical effects should not be expected to play a significant role in the results.
In point 59a, “materiality of business lines” shall be defined by their strategic importance, no matter how profitable these business lines are. Strategic importance is not only measured in terms of profit/loss generation.
The “considerations” in point 76 might result in institutions following universal banking models having better scores than specialised institutions. In order to be scored with a “1”, no “material asset and funding concentrations” are allowed. ESBG sees this statement fairly critical as specialised institutions often offer services of higher quality due to their degree of specialisation. At the same time they can definitely have an excellent business model. As a consequence, ESBG suggests introducing the element of specialisation in the “considerations” in point 76.
Q4. Does the breakdown of risk categories and sub-categories proposed provide appropriate coverage and scope for conducting supervisory risk assessments? (Title 6)
In several sections, the SREP GL establish the need for the NCAs to review risk concentrations (bond portfolio, deposit accounts, etc.). It would be necessary to clarify whether supervisors, when measuring Interest Rate Risk in the Banking Book (IRRBB), will analyse concentrations apart from accumulations of re-pricing/maturity at different points on the curve that determine interest rate risk (points 288-304). If the NCAs evaluated concentrations in fixed-income portfolios in the IRRBB risk assessment and product concentrations (e.g. deposits), this would not be evaluating IRRBB risk, but rather evaluating credit risk, business risk or liquidity. Therefore, we believe that the EBA should consider providing definition of concentration risk in the area of interest rate risk.Concerning the IRRBB itself, ESBG calls on the EBA to deal with this topic in the specific guidelines on the IRRBB which are being developed by the EBA. Nevertheless, ESBG would like to highlight that in the current interest rate environment, a 200 basis points shock as proposed by EBA in point 293 of the consultation paper, is not necessarily tailored to all circumstances and rate environments, and therefore it would be more suitable to change the wording of +/- 200 basis point for “standard shock”.
ESBG strongly believes that it would be better to continue to allow for the use of the internal model approach. On the metrics debate, ESBG supports that the importance of both the earnings as well as the economic value approach should be recognised in any future regime. The choice of the metrics should be consistent with the risk management strategy and risk appetite of the institution.
To ESBG’s mind, the subcategories listed in point 244 are contradictory to the Advanced Measurement Approach (AMA)-categorisation. For instance, conduct risk does not play a role in the AMA-categorisation.
It is indeed difficult to compare IT-systems and IT-architectures among banking institutions as there are no harmonised standards to measure their complexity. Thus, if IT-complexity is included in the SREP GL, common and comprehensive assessment criteria will be more than useful.
There is no need to have reputational risk as a separate category. As reputational risk is regularly considered a part of operational risk, it should be treated as such within the SREP GL.
ESBG doubts whether it is necessary to differentiate between the 4 different types of risk in point 280. This might be a suitable theoretical approach. In practice, however, these risks are usually not dealt with in a separate way.
Q5. Do you agree with the use of a standard approach for the articulation of additional own funds requirements to be used by compete authorities across the Union? (Title 7)
As originally intended by the Basel Committee of Banking Supervision, pillar 2 of the Basel framework serves to acknowledge and evaluate institutions’ own risk assessments and their corresponding allocation of internal capital. This reasoning has been adapted throughout the implementation of Basel II and Basel III by the European Union. ESBG believes that a standard approach is difficult to align with the CRD IV’s and the Basel framework’s understanding of pillar 2. Under the proposed standard approach for the assessment of an institution’s capital adequacy and for the articulation of additional own funds requirements, pillar 2 as such would cease to exist and be replaced by a “pillar 1+” or “pillar 2A” perspective. We strongly believe that institutions should not be restricted to formal own funds definitions within pillar 2. In our view and in line with the Basel framework, institutions should be allowed to allocate internal capital as long as both its economic ability to absorb losses and its availability in a going-concern situation are ensured. ESBG therefore pleads for maintaining the original and internationally agreed upon definition of pillar 2.More precisely, as we understand it, the new approach would require a comparison per risk between the pillar 1 capital charges, the capital requirement calculated using the bank’s internal risk model and an amount calculated using a prudential benchmark model. We presume that the highest of these three amounts would be taken as a (worst-case) yardstick. If it was above the pillar 1 capital charge, an add-on for the difference would be imposed. These additional capital requirements for credit risk, market risk and operational risk would be added together and then further supplemented with a capital requirement for interest rate risk, model risk and reputational risk. No diversification effects would be taken into account. The objective of this exercise would be to calculate the worst possible overall risk scenario facing the bank. We do not believe that this would be a viable approach suitable for the management of a bank. Pillar 2 would be diverted from its intended purpose and downgraded to a “pillar 1 plus”.
We consider a more appropriate approach that the bank calculates its overall need for economic capital (taking into account diversification effect) and to discuss the outcome with supervisors. Then the results of this process must be compared with the overall capital requirements determined under pillar 1. This approach is already in place in a couple of EU countries.
Specifically, the CRD IV explicitly uses the term “internal capital” (Article 73), while the SREP GL apply the term “regulatory own funds” to pillar 2. While article 104 (1a) CRD IV gives competent authorities the power to determine additional own funds requirements, we construe this power to be restricted to singular cases only. In other words, article 104 (1a) CRD IV does not affect the internal capital definition of article 73. In summary, the proposed limitation to own funds is not in line with the CRD IV’s internal capital definition and point 337 should be amended accordingly.
In ESBG’s opinion it is not necessary to also include credit concentration risk in point 335b as it is already part of the internal models determining credit risk.
With regard to point 320, ESBG would like to state that comparability of institutions in the context of ICAAP will not always be possible. For instance, there are differences in the parameterisation of the risk calculation concerning the risk capacity as “going concern” (in contrast to a “gone concern” approach).
Moreover, the additional capital requirements should not be determined by simply adding the capital needs of each risk category without allowing for diversification.
Q6. Do you agree that competent authorities should be granted additional transition periods for meeting certain capital and liquidity provisions in the guidelines (Title 12)?
Yes, ESBG strongly supports the idea that additional transition periods are granted to the competent authorities. Longer transition periods should be granted not only for the specified capital and liquidity provisions, but in general, as the guidelines will trigger significant implementation processes for both NCAs and supervised institutions.Upload files
SREP - ESBG position paper - FV.pdf
(262.69 KB)