Response to consultation on draft RTS on IRRBB standardised approach
Go back
Similarly, the cap on the portion of the core component of NMDs could lead, for instance on retail non-transactional deposit, to unintended cliff-effect for banks with a traditional and retail-based business model, not recognizing the actual stability, both in terms of outstanding and interest rates, of these products.
Furthermore, the application of the scalars 0.8 and 1.2 will amplify this impact. We note that, for the Simplified-SA, the calibration of the caps on the core component, according to the increased scalars (0.7 and 1.3), is even more severe, especially in the “increase of short-term rates scenario”, and even when considering the goal of conservative approach as stated in paragraph 5 of art. 84 of Dir. 2013/36/UE.
This might lead to a potential significant mismeasurement of risk, not necessarily a prudent one. We would like to point out that the final IRRBB measures depend not only on the level of “conservativeness” in modelling NMDs, but also on the repricing structure of the corresponding assets, so that a very short-term representation of NMDs in some cases could miss to capture the real repricing mismatch between assets and liabilities.
Finally, regarding the proposed application of 0.8/1.2 scalars to the core component of NMDs, it is not clear how the two scalars have been calibrated to capture the sensitivity of NMDs stability and pass-through with respect to changing interest rates. From our standpoint, 20% volatility implicit in the scalars is quite high, and would have a major impact on risk measures, not comparable to the corresponding impact that the same 0.8/1.2 scalars would produce when applied to prepayment or early redemption rates. According to our estimates, based on a real case simulation, applying the scalars might lead to an increase of the EVE SOT indicator up to 7% (as a percentage of Tier1 capital), which is roughly half of the SOT threshold.
Banks will have to develop models and produce estimations that are not already available for smaller and non-complex institutions. Besides, for institutions that have outsourced their IT system, it will not be straightforward to retrieve the additional information needed to perform such an analysis: for example, to determine if fixed rate deposits shall be considered as term deposits with the risk of early redemption, the banks need to identify if the early withdrawal would result in a sufficient penalty, like described in paragraph 2 article 9.
In this respect, we observe that the materiality threshold of 2% envisaged in Articles 8 and 9 is too low. As observed in relation to the choice of the 0.8/1.2 scalars for core component of NMDs, we question the application of the same parameter to different kind of behavioural options, due to the different impacts on the final risk measures. For instance, the modelling of NMDs, representing 2% of a banking book portfolio, will have a significantly more material impact on IRRBB measures than applying an early redemption rate to a fixed rate term deposit portfolios with the same outstanding.
As for the fair value component of fair value accounted instruments, we point out that the proposed methodology, which calculates market value changes at the beginning of the time horizon (i.e., at analysis date), will result in a double counting of the impact of FV changes and NII changes over the first year, thus leading to potential mismeasurement. As we state also in our response to EBA/CP/2021/36, our preference is to exclude market value changes from the NII measure, or, as a second choice, we suggest adapting the standardized methodology in order to capture the market value changes at the end of the NII time horizon.
According to this, we would suggest introducing a materiality threshold for it. The threshold could be set in terms of minimum outstanding of floating rate positions funded by/invested in other floating positions, but with different benchmark rate indexation.
A possibility in the direction of greater proportionality could be considering a sensitivity approach for the simplified framework for NII, thus excluding the computation of all fixed rate coupons and commercial spreads that are needed for the full NII projection.
As a second possibility, we suggest considering the definition of standardized conditional prepayment rates as well as standardized average early redemption rates, to be applied in the Simplified SA, consistently with the modelling approach envisaged for NMDs. It is questionable that small and non-complex banks should be provided with standardized modelling parameters for NMDs, while would still be required to produce their own estimations for fixed rate loans and term deposits.
As a further simplification, we would also suggest excluding from the metrics the evaluation of interest rate options, when they are bought by the bank, as it is allowed now for smaller banks in Italy.
Besides, we would point out that simplifications should not imply an exacerbation of constraints, resulting in inconsistent increases of the risk measures. We refer, for example, to caps prescribed in art. 23 for the core component of NMDs, where scalars of 0.7/1.3 instead of 0.8/1.2 have been used: we expect some unintended impacts in risk indicators for smaller banks with a traditional retail business model, which might require them to adjust their business strategies accordingly.
However, the new proposed SA, as well as the simplified SA, is far more complex compared to the standardized model currently used by less significant and small institutions. Therefore, those banks will have to significantly enhance their analytical tools, in order to comply with the new SA, facing considerable investments.
At the same time, it is not clear whether the required costs would be offset by the benefit of a more precise measure of interest rate risk, considering the business models of smaller banks, which traditionally don’t take major exposures on this risk.
We finally highlight that, if banks will have to adjust their IT systems to comply with the new SA, they should be granted a proper time span. We suggest to extend the current provision of 20 days after the publication on Official Journal of the European Union, which appears not adequate to us, to one year.
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?
Most of Italians banks which are currently applying standardized methodologies don’t rely on prepayments models for IRRBB measurement. The phenomenon of prepayment for floating rate instruments could be material per se, however we observe that it would have limited impact on IRRBB metrics. On the other hand, the implementation of dedicated models would pose significant methodological and operational challenges. Therefore, we don’t foster the inclusion of prepayments’ estimation in Article 6.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.
While we understand that the constraints described in art. 7 are driven by a prudent approach when considering behavioural cash flows, we see a significant risk of unintended consequences of this type of provision. In particular, the maturity constraints to the weighted average maturity of the core component (set at 5, 4.5, 4 years, according to the type and counterparty of the non-maturity deposits, NMDs) don’t reflect the empirical evidence in the Italian banking system, where the estimated expected weighted average life for NMDs is over nine years. The misalignment is even more evident in the Simplified SA, where the maturity constraints refer to the maximum maturity instead of the average.Similarly, the cap on the portion of the core component of NMDs could lead, for instance on retail non-transactional deposit, to unintended cliff-effect for banks with a traditional and retail-based business model, not recognizing the actual stability, both in terms of outstanding and interest rates, of these products.
Furthermore, the application of the scalars 0.8 and 1.2 will amplify this impact. We note that, for the Simplified-SA, the calibration of the caps on the core component, according to the increased scalars (0.7 and 1.3), is even more severe, especially in the “increase of short-term rates scenario”, and even when considering the goal of conservative approach as stated in paragraph 5 of art. 84 of Dir. 2013/36/UE.
This might lead to a potential significant mismeasurement of risk, not necessarily a prudent one. We would like to point out that the final IRRBB measures depend not only on the level of “conservativeness” in modelling NMDs, but also on the repricing structure of the corresponding assets, so that a very short-term representation of NMDs in some cases could miss to capture the real repricing mismatch between assets and liabilities.
Finally, regarding the proposed application of 0.8/1.2 scalars to the core component of NMDs, it is not clear how the two scalars have been calibrated to capture the sensitivity of NMDs stability and pass-through with respect to changing interest rates. From our standpoint, 20% volatility implicit in the scalars is quite high, and would have a major impact on risk measures, not comparable to the corresponding impact that the same 0.8/1.2 scalars would produce when applied to prepayment or early redemption rates. According to our estimates, based on a real case simulation, applying the scalars might lead to an increase of the EVE SOT indicator up to 7% (as a percentage of Tier1 capital), which is roughly half of the SOT threshold.
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.
While we agree that the required determination and application of a conditional prepayment rate and term deposit redemption rate as described in Articles 8 and 9 are reflective of the risks, we point out that the implementation would pose significant challenges to institutions, both under a methodological and an operational point of view.Banks will have to develop models and produce estimations that are not already available for smaller and non-complex institutions. Besides, for institutions that have outsourced their IT system, it will not be straightforward to retrieve the additional information needed to perform such an analysis: for example, to determine if fixed rate deposits shall be considered as term deposits with the risk of early redemption, the banks need to identify if the early withdrawal would result in a sufficient penalty, like described in paragraph 2 article 9.
In this respect, we observe that the materiality threshold of 2% envisaged in Articles 8 and 9 is too low. As observed in relation to the choice of the 0.8/1.2 scalars for core component of NMDs, we question the application of the same parameter to different kind of behavioural options, due to the different impacts on the final risk measures. For instance, the modelling of NMDs, representing 2% of a banking book portfolio, will have a significantly more material impact on IRRBB measures than applying an early redemption rate to a fixed rate term deposit portfolios with the same outstanding.
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?
Fixed rate loan commitments do not represent material exposures in the Italian market. However, we believe that the provision of Article 11 would benefit from further clarification regarding the expected representation within the SA framework, also by means of numerical examples.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?
The provision of calculating the impact of a 25% increase in implicit volatility is operationally implementable. However, it would pose significant challenges to less significant institutions, since this approach requires a more sophisticated methodology than that most frequently used, which is based on a delta equivalent approach. As for other topics discussed, the challenge is not only limited to the enhancement of the applied financial models, but relates consistently to the evolution of current analytical tools.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?
As mentioned before in relation to other topics, we believe that the implementation of the additional slotting of repricing cash flows according to original maturity/reference term is operationally feasible, however it will require banks to adopt more sophisticated analytical tools, with a considerable effort in terms of time and investments. This holds true also for other requirements, such as for the determination of the commercial margin components, broken down by product, geography and so on.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)?
In the same vein of previous answer, we observe that the SA for NII calculation is very complex. In this sense, its implementation is feasible, but it will be at a considerable cost for banks.As for the fair value component of fair value accounted instruments, we point out that the proposed methodology, which calculates market value changes at the beginning of the time horizon (i.e., at analysis date), will result in a double counting of the impact of FV changes and NII changes over the first year, thus leading to potential mismeasurement. As we state also in our response to EBA/CP/2021/36, our preference is to exclude market value changes from the NII measure, or, as a second choice, we suggest adapting the standardized methodology in order to capture the market value changes at the end of the NII time horizon.
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 believe that the add-on component for basis risk would add further complexity to an already complex framework, while not adding much value in terms of risk measurement accuracy, given the limited materiality of this type of risk for Italian banks.According to this, we would suggest introducing a materiality threshold for it. The threshold could be set in terms of minimum outstanding of floating rate positions funded by/invested in other floating positions, but with different benchmark rate indexation.
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?
We find that the adjustments in the Simplified Standardised Approach are operationally implementable. Nonetheless, we believe that the complexity of the structure of SA, even after those simplifications, will still require major enhancements in the current tools used by small and non-complex institutions, so that a significant change of approach for the measurement of IRRBB will be needed. From this point of view, it seems that proportionality has not been properly considered.A possibility in the direction of greater proportionality could be considering a sensitivity approach for the simplified framework for NII, thus excluding the computation of all fixed rate coupons and commercial spreads that are needed for the full NII projection.
As a second possibility, we suggest considering the definition of standardized conditional prepayment rates as well as standardized average early redemption rates, to be applied in the Simplified SA, consistently with the modelling approach envisaged for NMDs. It is questionable that small and non-complex banks should be provided with standardized modelling parameters for NMDs, while would still be required to produce their own estimations for fixed rate loans and term deposits.
As a further simplification, we would also suggest excluding from the metrics the evaluation of interest rate options, when they are bought by the bank, as it is allowed now for smaller banks in Italy.
Besides, we would point out that simplifications should not imply an exacerbation of constraints, resulting in inconsistent increases of the risk measures. We refer, for example, to caps prescribed in art. 23 for the core component of NMDs, where scalars of 0.7/1.3 instead of 0.8/1.2 have been used: we expect some unintended impacts in risk indicators for smaller banks with a traditional retail business model, which might require them to adjust their business strategies accordingly.
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 find that the standardised and simplified standardised approaches are clear, and that are operationally implementable.However, the new proposed SA, as well as the simplified SA, is far more complex compared to the standardized model currently used by less significant and small institutions. Therefore, those banks will have to significantly enhance their analytical tools, in order to comply with the new SA, facing considerable investments.
At the same time, it is not clear whether the required costs would be offset by the benefit of a more precise measure of interest rate risk, considering the business models of smaller banks, which traditionally don’t take major exposures on this risk.
We finally highlight that, if banks will have to adjust their IT systems to comply with the new SA, they should be granted a proper time span. We suggest to extend the current provision of 20 days after the publication on Official Journal of the European Union, which appears not adequate to us, to one year.
Upload files
2022-03-31_CP-RTS-SA.pdf
(135.93 KB)