Response to consultation on draft Guidelines on IRRBB and CSRBB
Go back
Nonetheless, there might be specific type of retail deposits (e.g., passbook savings accounts), held mainly for savings purposes, showing very low elasticity to market rates and high stability, entailing a longer-term average repricing (see A. Castagna. Risk and Profitability of Sight Deposits in the Italian Banking Industry. Iason Research Paper Series, 2019). In such cases, posing a cap would not lead to a prudent management of the interest rate risk exposure, but to a structural underhedge (or shorter investment of assets), which exposes the bank to the risk of decreasing interest rates.
Furthermore, we would like to highlight that this regulatory constraint is further strengthened in the provision of the proposed Standardized Approach (SA), which sets the five-year cap to the core fraction only of the non-financial NMDs (5 years for retail transactional, 4.5 years for retail non-transactional and 4 years for wholesale non-financial). This stricter constraint could be detrimental to specific assets class such as residential mortgages or project financing, because the difference would create a material cliff effect, and a consequently significant impact on the bank’s ability to support real economy, in case a bank should be asked to change its approach from the internal measurement system (IMS) to the SA. Hence, we would suggest to align the SA proposal to the general framework presented in the Guidelines.
However, a point of concern is that, given their high-level formulation, these criteria might be interpreted and applied with significant differences by different authorities and/or different supervisory teams.
In particular, it might be useful to provide a clearer explanation of the criteria related to the use of the methods for IRRBB described in Annex I. It might be worthwhile, for example, to specify that an institution using one or two of the most sophisticated methods/approaches, with adequate calibration, validation and backtesting, does not necessarily need to use also other methods/approaches to be compliant with the criteria. Moreover, in the current formulation, the statement “When measuring their exposure to IRRBB, institutions should not limit themselves to the methods listed in Annex I in order to ensure that material aspects of interest rate risk are captured adequately” can be interpreted in different ways and would benefit from further clarifications.
Furthermore, it should be clarified if, in case of a non-satisfactory internal model, the SA should be applied only to the specific element deemed non-satisfactory or to the whole IRRBB framework. In addition, in case of a non-satisfactory internal system or model, it’s not specified the timeline to move from the IMS to the SA. Given the potential impact on the banking business, we would consider beneficial to limit the implication of the SA only to the specific item or Legal Entity of the banking group deemed non-satisfactory during the supervisory assessment, and to define an adequate timeline for the shift from IMS to SA. Moreover, it might be worth considering the possibility for the bank to improve its internal system to be compliant within the defined timeline, as an alternative to the use of the standardised methodology.
Similar considerations apply also to the distinction between “market liquidity spread” and “idiosyncratic liquidity spread”, with the further issue that liquidity spreads tend to be more instrument-specific, therefore even more difficult to be reliably measured.
We highlight that a clearer definition of market credit spreads and market liquidity spreads would be even more important in case the monitoring of CSRBB should be applied to the portfolio of loans classified held-to-collect and without an active secondary market. In this case, in the absence of observable market prices, the definition of the methodology for the identification of the market credit spread, as distinct from the idiosyncratic spread, to assess and monitor CSRBB, could require introducing arbitrary assumptions, thus leading to heterogeneous approaches, with detrimental impacts on results comparability or even on business strategies.
Should different accounting practices across banks and jurisdictions suggest not to rely on accounting categories to identify the perimeter of the analysis, the CSRBB should be assessed with reference to instruments traded in liquid markets and with observable prices.
Mainly, we are concerned that extending the perimeter to the whole banking book could lead to a significant impact in terms of Pillar 2 internal capital, while at the same time figures could be very different from bank to bank, given the lack both of well-established common practices and of reliable market data for the evaluation of credit spread risk arising from non-marketable assets portfolios.
As mentioned in the answer to question 3 above, it is not clear, from a methodological point of view, how the approach for assessing CSRBB could be implemented in the risk management process for instruments that are not marketable, such as loans. Besides, from an operational point of view, implementing the proposed Guidelines according to an extended banking book perimeter would require significant changes in the IT systems, in terms of input data, such as rating classes, and analytical tools, thus requiring a considerable effort in time and investments.
Finally, we observe that the whole CSRBB framework envisaged in the proposed Guidelines is quite complex and demanding, also in terms of governance and related processes. In front of it, however, it is not clear how the risk measures and controls that will have to be put in place to monitor CSRBB on the “wider” perimeter, would be beneficial in the steering process of the banking book. For instance, in case of limit breaches, as it is also reported in the accompanying documents, banks would reduce their positions, but this would not be feasible when the exposure arises from a loan portfolio.
Besides, setting limits would impact the business strategy of the bank. For the time being, given the lack of consolidated market practices, and hence given the potential heterogeneity of the approaches implemented across banks, and of the limits that would be set, we are concerned about potential distortions in the way banks will define their strategies and provide credit to the market. In other words, heterogeneity in the adopted frameworks could pose issues in terms of comparability and of unintended consequences on banks strategies.
We observe that while general provisions for identification, assessment and monitoring of CSRBB appear to be very demanding both conceptually and operationally, in line with what is requested for IRRBB, there are however no detailed annexes describing measurement methods or sophistication matrix, unlike what has been done for IRRBB. In our view this denotes that the current evolution of measurement methods for CSRBB is still at an early stage and should lead to reconsidering regulatory expectations for IMSs capabilities.
Question 1: In the context of the measurement of the impact of IRRBB under internal systems, paragraph 111 envisages a five year cap repricing maturity for retail and non-financial wholesale deposits without a specified maturity. Would you foresee any unintended consequence or undesirable effect from this behavioural assumption in particular on certain business models or specific activities? If this is the case, please kindly provide concrete examples of it.
Considering the average business mix of Italian commercial banks, we observe that the five-year cap, applied on the weighted average repricing date of the full amount of non-financial non-maturity deposits (NMDs), should not introduce unintended consequences, also considering that it is consistent with the current Guidelines’ prescriptions for SOT.Nonetheless, there might be specific type of retail deposits (e.g., passbook savings accounts), held mainly for savings purposes, showing very low elasticity to market rates and high stability, entailing a longer-term average repricing (see A. Castagna. Risk and Profitability of Sight Deposits in the Italian Banking Industry. Iason Research Paper Series, 2019). In such cases, posing a cap would not lead to a prudent management of the interest rate risk exposure, but to a structural underhedge (or shorter investment of assets), which exposes the bank to the risk of decreasing interest rates.
Furthermore, we would like to highlight that this regulatory constraint is further strengthened in the provision of the proposed Standardized Approach (SA), which sets the five-year cap to the core fraction only of the non-financial NMDs (5 years for retail transactional, 4.5 years for retail non-transactional and 4 years for wholesale non-financial). This stricter constraint could be detrimental to specific assets class such as residential mortgages or project financing, because the difference would create a material cliff effect, and a consequently significant impact on the bank’s ability to support real economy, in case a bank should be asked to change its approach from the internal measurement system (IMS) to the SA. Hence, we would suggest to align the SA proposal to the general framework presented in the Guidelines.
Question 2: Do respondents find that the criteria to identify non-satisfactory IRRBB internal models provide the minimum elements for supervisors’ assessment?
In our view the Guidelines provide the elements to identify non-satisfactory IRRBB internal models, although the criteria are expressed in very general terms.However, a point of concern is that, given their high-level formulation, these criteria might be interpreted and applied with significant differences by different authorities and/or different supervisory teams.
In particular, it might be useful to provide a clearer explanation of the criteria related to the use of the methods for IRRBB described in Annex I. It might be worthwhile, for example, to specify that an institution using one or two of the most sophisticated methods/approaches, with adequate calibration, validation and backtesting, does not necessarily need to use also other methods/approaches to be compliant with the criteria. Moreover, in the current formulation, the statement “When measuring their exposure to IRRBB, institutions should not limit themselves to the methods listed in Annex I in order to ensure that material aspects of interest rate risk are captured adequately” can be interpreted in different ways and would benefit from further clarifications.
Furthermore, it should be clarified if, in case of a non-satisfactory internal model, the SA should be applied only to the specific element deemed non-satisfactory or to the whole IRRBB framework. In addition, in case of a non-satisfactory internal system or model, it’s not specified the timeline to move from the IMS to the SA. Given the potential impact on the banking business, we would consider beneficial to limit the implication of the SA only to the specific item or Legal Entity of the banking group deemed non-satisfactory during the supervisory assessment, and to define an adequate timeline for the shift from IMS to SA. Moreover, it might be worth considering the possibility for the bank to improve its internal system to be compliant within the defined timeline, as an alternative to the use of the standardised methodology.
Question 3: Is there any specific element in the definition of CSRBB that is not clear enough for the required assessment and monitoring of CSRBB by institutions?
We think that some specific elements in the current definition are not clear enough for assessment and monitoring purposes. In particular, the distinction between “market credit spread” and “idiosyncratic credit spread” would benefit from further clarifications, maybe by means of some specific numerical examples. To this regard, we notice that idiosyncratic credit spread is defined in footnote 14 as including assessment of risks “arising from the sector and geographical location of the borrower”, which is quite a “wide” definition of idiosyncratic credit spread. In this vein, it could be useful to make clear whether credit spread volatility driven by sovereign spread volatility should be considered as “idiosyncratic” or “market” credit spread.Similar considerations apply also to the distinction between “market liquidity spread” and “idiosyncratic liquidity spread”, with the further issue that liquidity spreads tend to be more instrument-specific, therefore even more difficult to be reliably measured.
We highlight that a clearer definition of market credit spreads and market liquidity spreads would be even more important in case the monitoring of CSRBB should be applied to the portfolio of loans classified held-to-collect and without an active secondary market. In this case, in the absence of observable market prices, the definition of the methodology for the identification of the market credit spread, as distinct from the idiosyncratic spread, to assess and monitor CSRBB, could require introducing arbitrary assumptions, thus leading to heterogeneous approaches, with detrimental impacts on results comparability or even on business strategies.
Question 4: As to the suggested perimeter of items exposed to CSRBB, would you consider any specific conceptual or operational challenge to implement it?
We observe that most common banking practices for CSRBB focus on assets accounted at fair value (typically bonds in the Held to Collect and Sell category), according to an Economic Value approach. This seems to be also more in line with the 2016 Basel Standards on IRRBB. Thus, we think that introducing a different framework encompassing a much wider perimeter for CSRBB would lead to an unlevelled playing field for European banks.Should different accounting practices across banks and jurisdictions suggest not to rely on accounting categories to identify the perimeter of the analysis, the CSRBB should be assessed with reference to instruments traded in liquid markets and with observable prices.
Mainly, we are concerned that extending the perimeter to the whole banking book could lead to a significant impact in terms of Pillar 2 internal capital, while at the same time figures could be very different from bank to bank, given the lack both of well-established common practices and of reliable market data for the evaluation of credit spread risk arising from non-marketable assets portfolios.
As mentioned in the answer to question 3 above, it is not clear, from a methodological point of view, how the approach for assessing CSRBB could be implemented in the risk management process for instruments that are not marketable, such as loans. Besides, from an operational point of view, implementing the proposed Guidelines according to an extended banking book perimeter would require significant changes in the IT systems, in terms of input data, such as rating classes, and analytical tools, thus requiring a considerable effort in time and investments.
Finally, we observe that the whole CSRBB framework envisaged in the proposed Guidelines is quite complex and demanding, also in terms of governance and related processes. In front of it, however, it is not clear how the risk measures and controls that will have to be put in place to monitor CSRBB on the “wider” perimeter, would be beneficial in the steering process of the banking book. For instance, in case of limit breaches, as it is also reported in the accompanying documents, banks would reduce their positions, but this would not be feasible when the exposure arises from a loan portfolio.
Besides, setting limits would impact the business strategy of the bank. For the time being, given the lack of consolidated market practices, and hence given the potential heterogeneity of the approaches implemented across banks, and of the limits that would be set, we are concerned about potential distortions in the way banks will define their strategies and provide credit to the market. In other words, heterogeneity in the adopted frameworks could pose issues in terms of comparability and of unintended consequences on banks strategies.
Both IRRBB and CSRBB
Yes, the separation of IRRBB and CSRBB seems to be sufficient to understand where the Guidelines apply to IRRBB only, CSRBB only or both. In our understanding, only paragraph 4.1 applies to both risks, paragraphs 4.2 to 4.4 apply to IRRBB only and paragraphs 4.5 and 4.5 apply to CSRBB only, except for the treatment of diversification effects, concerning the combination of IRRBB and CSRBB.We observe that while general provisions for identification, assessment and monitoring of CSRBB appear to be very demanding both conceptually and operationally, in line with what is requested for IRRBB, there are however no detailed annexes describing measurement methods or sophistication matrix, unlike what has been done for IRRBB. In our view this denotes that the current evolution of measurement methods for CSRBB is still at an early stage and should lead to reconsidering regulatory expectations for IMSs capabilities.
Upload files
2022-03-31_CP-GL-IRRBB-CSRBB.pdf
(152.17 KB)