Response to consultation on draft RTS on IRRBB standardised approach
Go back
But the additional distinction between “stable” and “non-stable” should be dropped, in our view, as the two are not unequivocally distinguishable and the approach mixes up interest and liquidity. It is confusing, moreover, that the definition of the stable part on page 18 mentions “under the current level of interest rates” while Article 7(2) on page 23 requires the consideration of “upward and downward movements” over the last ten years. Furthermore, the specification of specific parameters for modelling in no way reflects our understanding of an appropriate model.
Banks’ individual models reflect the customer interest rate adjustments intended by the bank. The synchronisation of pricing and risk mapping will be significantly restricted by concrete specifications of what constitutes “core” – be they strict upper limits and scenario-dependent factors (under the standardised approach) or direct mandating of precise proportions (under the simplified standardised approach). This will lead to incorrect risk measurement. We would like to reiterate our view that, in general, standardised approaches can be no substitute for appropriate internal models in the IRRBB environment.
Moreover, the exclusion from the core category of wholesale NMDs from financial customers is neither appropriate nor consistent with the Basel standard.
Throughout the document, it should be made clear that the estimation has to be applied consistently over time (cf. Article 9) and not that the estimator itself has to be consistent (cf. page 9 vs page 26).
Examples of products that currently fall under the definition:
◼ Floating rate products with an implicit floor of 0% either on the total costumer rate or the reference rate
◼ Wholesale fixed term deposits with an early redemption right under Article 9(3)
◼ Implicit 0% floors on non-maturing retail deposits
We would also appreciate it if the EBA provided details of the empirical information on which the assumption of the 25% increase is based.
In particular, the cash flow slotting under shock scenarios is far too complex and the economic rationale is not clear since the core component is the part of the NMDs that “is unlikely to reprice even under significant changes in the interest rate environment”.
Furthermore, we do not understand the rationale behind the structure of the reference term time buckets and would appreciate more detailed explanation of the economic background on why this is deemed appropriate.
Also, the EBA should clarify how to treat non-contractual cashflows such as early redemptions and prepayments regarding their repricing term. If, for example, 50% of a fixed deposit is modelled to be redeemed O/N, it should be clarified whether it should be reinvested with the O/N shock or at the initial maturity.
We would nevertheless like to point out that many small and medium-sized banks in Germany prepare German GAAP accounts. Yet the current definition of fair value effects only applies to banks using IFRS, which will make it difficult for banks using German GAAP to implement the requirements. Smaller banks, in particular, generally have few positions with effects on P&L in different interest rate scenarios. This should be taken into account by setting an appropriate threshold to ensure that such a resource-intensive calculation will only be performed if the underlying risk is actually material for the bank. If the narrow definition of NII is selected for the regulatory outlier test, this should also be taken into account concerning the standardised models.
◼ General approach:
We understand the general approach and welcome the clarification of ambiguities in the existing standardised approach of the Basel Committee, especially with regard to the modelling of demand and savings deposits and the definition of automatic options. It is nevertheless apparent in important areas that the standardised approach and simplified standardised approach are not suitable for small, medium-sized or non-complex institutions and are generally too complex. This is particularly evident when it comes to the requirements for considering option and basis risks, demand and savings deposits and early repayments. For details, see below.
◼ Automatic options:
It will not be possible for many small banks to calculate the value of automatic options in accordance with Article 12 using a scenario-based full valuation as they do not have the technical capability or expertise to carry out such a valuation. For small, non-complex institutions, on the other hand, the simplified standardised approach offers a feasible way of taking these options into account. We support such an approach but would point out that the results thus obtained will be correspondingly imprecise. We also believe the standardised approach should contain a materiality threshold for automatic options below which they do not need to be considered.
◼ Early repayment:
We welcome the EBA’s proposal to enable a clear standardised approach to dealing with early repayment. Here too, however, the time and effort involved in implementation should be in proportion to the materiality of influencing factors. The proposed thresholds for the consideration of early repayment are not appropriate in our view. The consultation paper currently envisages that early repayments should be modelled as soon as 2% of total assets consist of fixed-interest assets with early repayment rights in accordance with Article 8. This would affect a lot of German banks given the scale of their fixed-income lending. Yet this threshold relates only to the volume and not the impact of such repayment rights, which is not appropriate in our view. We would recommend setting thresholds that relate to the expected impact on the risk figure instead of the absolute volume of products with options.
◼ (Simplified) standardised approach to NII
◼ We appreciate the attempt to simplify the NII standardised approach for small, non-complex institutions. However, the requirements are still rather complex (data requirements, options, margins, basis risk, fair value changes).
◼ General approach:
In addition to the following NII-specific challenges, our above comments on Articles 7 to 12 relating to the EVE standardised approach also apply.
◼ Consideration of margin:
Based on the EBA’s assumption in the explanatory box under Article 18 that margins are not materially sensitive to interest rates, the effect of margin payments on delta NII is not relevant except for minimal changes caused by scenario-specific cash flow changes. To determine the effects of margin adjustments, however, banks will have to prepare the term linkage table introduced in Article 17, which will require substantial additional time and effort. Of primary relevance for regulatory purposes is the risk measure of delta NII, provided option A is selected in the SOT for NII, which we believe is the metric that makes most sense. We therefore recommend only prescribing the delta NII calculation in the standardised approach provided that it is sufficiently conservative and ignoring the absolute NII along with margin effects.
◼ It should be made clear that no breakdown into counterparties is required in the empirical determination of commercial margins.
◼ Like in the EVE simplified standardised approach, the treatment of NMD does not reflect the actual behaviour of small banks, where deposits are one of their core competences.
◼ We also suggest only considering the reinvestment of the principal in line with the constant balance sheet definition as the priority and focusing on delta NII in a narrow sense. This will simplify the currently very complex calculation and make implementation much more feasible. This is particularly important as the approach should also be implementable by banks whose interest rate risk management has been deemed inadequate.
We would like to point out that inconsistencies may arise if internal systems are used for one perspective (EVE/NII) while the (simplified) standardised methodology has to be used for the other. With respect to NMDs, for instance, different cash flows could be modelled in the two perspectives: one cash flow that appropriately maps the institution’s planned interest rate adjustment policy and one cash flow constructed according to prudential regulations. In this case, differing risk management incentives could arise, not only from the differences between the EVE and NII methods but also from the diverging cash flows. This would significantly complicate the interpretation of the results. Solutions to this problem should be explored. One option would be the simultaneous application of the (simplified) standardised methodology in both perspectives – even if a satisfactory internal system exists for one of them.
Question 1: What is the materiality of prepayments for floating rate instruments and what are the underlying factors? Would you prefer the inclusion of a requirement in Article 6 for institutions to estimate prepayments for these instruments?
We welcome the exclusion of prepayments for floating rate products given their lack of materiality. Besides the already small impact on the overall risk metrics in the different scenarios, prepayments for floating rate products are typically independent of the interest environment and therefore do not have a significant impact on the delta EVE and delta NII risk metrics which are the key result of this standardised approach.Question 2: Do respondents find that the required determination of stable/non-stable deposits, and core/non-core deposits as described in Article 7 is reflective of the risks and operationally implementable? In case of any unintended consequence or undesirable effect on certain business models or specific activities, please kindly provide concrete examples.
We generally consider the modelling of demand and savings deposits in the standardised approach on the basis of the distinction between “core” and “non-core” volumes applied by many banks to be a sensible choice even for smaller institutions. This will enable banks to take account, within a clearly defined framework, of bank and customer-specific characteristics of their deposits even in the standardised approach.But the additional distinction between “stable” and “non-stable” should be dropped, in our view, as the two are not unequivocally distinguishable and the approach mixes up interest and liquidity. It is confusing, moreover, that the definition of the stable part on page 18 mentions “under the current level of interest rates” while Article 7(2) on page 23 requires the consideration of “upward and downward movements” over the last ten years. Furthermore, the specification of specific parameters for modelling in no way reflects our understanding of an appropriate model.
Banks’ individual models reflect the customer interest rate adjustments intended by the bank. The synchronisation of pricing and risk mapping will be significantly restricted by concrete specifications of what constitutes “core” – be they strict upper limits and scenario-dependent factors (under the standardised approach) or direct mandating of precise proportions (under the simplified standardised approach). This will lead to incorrect risk measurement. We would like to reiterate our view that, in general, standardised approaches can be no substitute for appropriate internal models in the IRRBB environment.
Moreover, the exclusion from the core category of wholesale NMDs from financial customers is neither appropriate nor consistent with the Basel standard.
Question 3: Do respondents find that the required determination and application of a conditional prepayment rate and term deposit redemption rate as described in Article 8 and 9 is reflective of the risks and operationally implementable? In case of any unintended consequence or undesirable effect on certain business models or specific activities, please kindly provide concrete examples.
We find the determination and application of a conditional prepayment rate as described in Article 8 operationally implementable. However, we do not agree with the definition of the exception/threshold in Article 8(2). Rather than defining a threshold based on 2% of total fixed rate loans, we would suggest a threshold based on the impact such options will have on the results. Under the currently envisaged requirements, a bank that allows a full loan repayment for 1.9% of the positions referred to in Article 2(2) would not have to model its prepayment whereas a bank that allows a 5% repayment for 2% of its positions would have to include the impact. We believe it would be more appropriate to determine materiality based on the percentage of possible prepayments.Throughout the document, it should be made clear that the estimation has to be applied consistently over time (cf. Article 9) and not that the estimator itself has to be consistent (cf. page 9 vs page 26).
Question 4: Is the treatment of fixed rate loan commitments to retail counterparties clear and are there other instruments with retail counterparties where a behavioural approach to optionality should be taken?
Yes, the approach is clear. However, we propose including a materiality threshold under which such instruments need not be included.Question 5: Do respondents find that the required determination of the impact of a 25% increase in implicit volatility as described in Article 12 is operationally implementable?
First of all, we would like to point out that not every institution is able to perform full revaluations, which may make it impossible for this approach to be applied by all institutions. Even if a full revaluation is possible, there are significant operational challenges. The current definition of products that fall under this full revaluation requirement is too broad, making it impossible for such banks to implement without disproportionate time and effort.Examples of products that currently fall under the definition:
◼ Floating rate products with an implicit floor of 0% either on the total costumer rate or the reference rate
◼ Wholesale fixed term deposits with an early redemption right under Article 9(3)
◼ Implicit 0% floors on non-maturing retail deposits
We would also appreciate it if the EBA provided details of the empirical information on which the assumption of the 25% increase is based.
Question 6: Do respondents find that the required slotting of repricing cash flows in accordance with the second dimension of original maturity/reference term as described in Article 13 is operationally implementable?
The approach is comprehensive. However, it will be challenging to collect the relevant data and operationally challenging to perform this calculation on a regular basis.In particular, the cash flow slotting under shock scenarios is far too complex and the economic rationale is not clear since the core component is the part of the NMDs that “is unlikely to reprice even under significant changes in the interest rate environment”.
Furthermore, we do not understand the rationale behind the structure of the reference term time buckets and would appreciate more detailed explanation of the economic background on why this is deemed appropriate.
Also, the EBA should clarify how to treat non-contractual cashflows such as early redemptions and prepayments regarding their repricing term. If, for example, 50% of a fixed deposit is modelled to be redeemed O/N, it should be clarified whether it should be reinvested with the O/N shock or at the initial maturity.
Question 7: Do respondents find it practical how the determination of several components of the NII calculation, with in particular the fair value component of Article 20 and the fair value component of automatic options of Article 15, is generally based on the processes used for the EVE calculation (in particular Article 16 and Article 12)?
Yes, this process consistency makes sense.We would nevertheless like to point out that many small and medium-sized banks in Germany prepare German GAAP accounts. Yet the current definition of fair value effects only applies to banks using IFRS, which will make it difficult for banks using German GAAP to implement the requirements. Smaller banks, in particular, generally have few positions with effects on P&L in different interest rate scenarios. This should be taken into account by setting an appropriate threshold to ensure that such a resource-intensive calculation will only be performed if the underlying risk is actually material for the bank. If the narrow definition of NII is selected for the regulatory outlier test, this should also be taken into account concerning the standardised models.
Question 8: Do respondents find that the calculation of the net interest income add-on for basis risk is reflective of the risk and operationally implementable
We understand the selected approach and the underlying assumptions. It is true that, in principle, basis risk also theoretically exists in an NII approach. We nevertheless consider the inclusion of this risk to be problematic and believe that one-size-fits-all requirements do not serve a useful purpose. We therefore welcome the bank-specific definition of scenarios, as this is the only way to take account of the actual situation of a bank. As with the other elements, however, we feel it would make good sense to introduce thresholds. The diversity of interest-linked financial instruments usually found at smaller banks and specialised institutions with a limited product range is low. In the retail banking environment, for example, there are many institutions with a negligible proportion of variable-rate products with different reference curves. It is important as a general principle that any existing option and basis risks do not have to be measured separately but can be included in an integrated measurement system. This is also true for the EVE approach.Question 9: Do respondents find that the adjustments in the Simplified Standardised Approach as set out in Article 23 and 24 are operationally implementable, and do they find that any other simplification would be appropriate?
◼ (Simplified) standardised approach to EVE◼ General approach:
We understand the general approach and welcome the clarification of ambiguities in the existing standardised approach of the Basel Committee, especially with regard to the modelling of demand and savings deposits and the definition of automatic options. It is nevertheless apparent in important areas that the standardised approach and simplified standardised approach are not suitable for small, medium-sized or non-complex institutions and are generally too complex. This is particularly evident when it comes to the requirements for considering option and basis risks, demand and savings deposits and early repayments. For details, see below.
◼ Automatic options:
It will not be possible for many small banks to calculate the value of automatic options in accordance with Article 12 using a scenario-based full valuation as they do not have the technical capability or expertise to carry out such a valuation. For small, non-complex institutions, on the other hand, the simplified standardised approach offers a feasible way of taking these options into account. We support such an approach but would point out that the results thus obtained will be correspondingly imprecise. We also believe the standardised approach should contain a materiality threshold for automatic options below which they do not need to be considered.
◼ Early repayment:
We welcome the EBA’s proposal to enable a clear standardised approach to dealing with early repayment. Here too, however, the time and effort involved in implementation should be in proportion to the materiality of influencing factors. The proposed thresholds for the consideration of early repayment are not appropriate in our view. The consultation paper currently envisages that early repayments should be modelled as soon as 2% of total assets consist of fixed-interest assets with early repayment rights in accordance with Article 8. This would affect a lot of German banks given the scale of their fixed-income lending. Yet this threshold relates only to the volume and not the impact of such repayment rights, which is not appropriate in our view. We would recommend setting thresholds that relate to the expected impact on the risk figure instead of the absolute volume of products with options.
◼ (Simplified) standardised approach to NII
◼ We appreciate the attempt to simplify the NII standardised approach for small, non-complex institutions. However, the requirements are still rather complex (data requirements, options, margins, basis risk, fair value changes).
◼ General approach:
In addition to the following NII-specific challenges, our above comments on Articles 7 to 12 relating to the EVE standardised approach also apply.
◼ Consideration of margin:
Based on the EBA’s assumption in the explanatory box under Article 18 that margins are not materially sensitive to interest rates, the effect of margin payments on delta NII is not relevant except for minimal changes caused by scenario-specific cash flow changes. To determine the effects of margin adjustments, however, banks will have to prepare the term linkage table introduced in Article 17, which will require substantial additional time and effort. Of primary relevance for regulatory purposes is the risk measure of delta NII, provided option A is selected in the SOT for NII, which we believe is the metric that makes most sense. We therefore recommend only prescribing the delta NII calculation in the standardised approach provided that it is sufficiently conservative and ignoring the absolute NII along with margin effects.
◼ It should be made clear that no breakdown into counterparties is required in the empirical determination of commercial margins.
◼ Like in the EVE simplified standardised approach, the treatment of NMD does not reflect the actual behaviour of small banks, where deposits are one of their core competences.
◼ We also suggest only considering the reinvestment of the principal in line with the constant balance sheet definition as the priority and focusing on delta NII in a narrow sense. This will simplify the currently very complex calculation and make implementation much more feasible. This is particularly important as the approach should also be implementable by banks whose interest rate risk management has been deemed inadequate.
Question 10: Do respondents find that all the necessary aspects are covered and the steps and assumptions for the evaluation of EVE and NII as laid out in the standardised approach and simplified standardised approach clear enough and operationally implementable?
We would like to stress that standardised approaches should not be used to challenge internal measurement systems (by benchmarking them against standardised approaches). On no account should institutions be required to implement standardised and internal systems in parallel. In any event, the data collection for a parallel calculation would be far too time-consuming (cf. our reply to question 6). We would ask the EBA to take a concrete position on this point both in the introductory section of the RTS and in section 4 and to define the precise scope of application and spell out its boundaries.We would like to point out that inconsistencies may arise if internal systems are used for one perspective (EVE/NII) while the (simplified) standardised methodology has to be used for the other. With respect to NMDs, for instance, different cash flows could be modelled in the two perspectives: one cash flow that appropriately maps the institution’s planned interest rate adjustment policy and one cash flow constructed according to prudential regulations. In this case, differing risk management incentives could arise, not only from the differences between the EVE and NII methods but also from the diverging cash flows. This would significantly complicate the interpretation of the results. Solutions to this problem should be explored. One option would be the simultaneous application of the (simplified) standardised methodology in both perspectives – even if a satisfactory internal system exists for one of them.