Response to consultation on Guidelines on technical aspects of the management of interest rate risk
Go back
It is our view that a materiality threshold of 2.0% of (fixed rate) NLE over total loans should be adopted to provide some guidance on the inclusion of NPEs. A measure based upon EV risk could provide some clarity.
As NPE are usually handled by banks in specific systems, designed to manage for the credit status of these loans and potential recoveries and provisions, integrating these systems within the IRRBB calculation engines could present an issue. New significant investment on IT could be required, budgets for which are typically agreed the year before development. To achieve the proposed deadline of 31 December 2018, IT expenditure and project resources, would need to be been approved within the 2017 calendar year. It would also create a new model risk that would need to be included in the model sensitivity analysis and reports. For this reason longer period for implementation is needed.
For banks with limited NPEs, the cost benefit of this integration would be disproportionate. For those banks below the 2.0% threshold, considering NPE either as non-interest bearing or without provisions should be allowed. Also small non-complex banks should be allowed to treat NPE in a more simple way.
However reading chapter 4.2 about capital identification calculation and allocation point 25 says: “Institutions should not only rely on the supervisory assessments of capital adequacy for IRRBB or on the outcome of the supervisory outlier test (see Section 4.5.), but should develop and use their own methodologies for capital allocation, based on their risk appetite, level of risk, and risk management policies”.
Are we supposed to read the “supervisory assessment” as the same as the “SOT number”? If so, we read point 25 that the SOT has a direct link to the quantification of required capital.
We urge the EBA to be very precise about the objective of the SOT, this both for a clear understanding of the EBA guidelines across institutions, but also to secure level playing field across regional supervisors.
Based on historic observations, ECB behavior and communication a floor level of -1,00% for short term EUR rates with a gradual increase to 0% at the 30 year tenor stand out as the appropriate levels to bridge the typical aim of stress testing; combining highly unlikely and yet feasible outcomes. For some currencies a lower bound of even -1.00 % seems too low. We therefore recommend EBA to consider working with currency specific floors.
Below we show the proposed floor combined with the EBA floor suggestion and the historic absolute low at the relevant tenors. We have looked at EUR, GBP and USD. Is shows that a lower bound of -1.00 % is far below the lowest levels we have observed.
Further motivation
Lately as interest rate levels have become negative the ECB has been more and more cautious lowering the rates further with the latest two reductions as small as 0.1%-points. For two years the ECB has kept the rates unchanged even in an environment of continued monetary policy accommodation. Massive quantitative easening (QE) has taken place within the latest years as a substitute for further interest rate reductions. That shift in usage of monetary instruments is a testimony of their view, that further negative interest rates will cause an increasingly amount of unintended consequences, which is translated into the view that the overall cost outpaces the benefits of lower rates.
Reluctance to bring interest rates hugely into negative territory has been a common feature for central banks. At the same time, though there has been moderate difference in their mix of policy instruments. The FED actually never went negative and the Bank of Japan only went marginally negative before the QE was used in big scale. This observation speaks for adaptation of currency specific floors, calibrated around a short term ECB floor of -1%.
As negative short term rates still has to be regarded as an extraordinary situation, the interest rate structure will always to some extent signal expectations of in-creasing rates; i.e. a positive sloping curve inside a foreseeable future. For that reason, theoretically the floor should gradually move upwards as the tenors increases.
Kindly see enclosed file
It is unclear why such a binary approach is being proposed, without an impact study. Our members have a long established banking book risk management practice, where high correlated currencies are used to hedge underlying risks. This is principally the use of EUR denominated instruments to hedge DKK denominated risks. This practice is well known to the local supervisor and driven by local market conditions, where it is not always possible to access the DKK market, in the size or duration required. This practice has proved to be an effective hedge for banking book risks through the economic cycle.
By disallowing such hedging practices within the SOT calculations, the comparability of the outputs, on a pan-European basis, would have to be questioned. Highly hedged Danish banks could potentially become outliers, a classification that might have unintended consequences. This could manifest itself in a reduction of banking book hedging activities, to manage the SOT risk. This is a concern for our members as they would not want to change the current practice within Denmark of maintaining a low banking book risk profile.
The SOT analysis should allow for aggregation of highly correlated currencies, for a comparable analysis to be effective. Such a method is supported by the analysis of historical correlations. The position of the EBF on this matter was articulated within the Industry IRRBB Working Group response dated 11 September 2015 within Appendix 8 (https://www.bis.org/bcbs/publ/comments/d319/igii.pdf). It is also worth noting that the EU Finance ministers recognized the importance of such correlations through the proposed changes to the EU Commission’s FRTB proposal in CRR2 (where increased correlations between GIRR buckets for EUR/ERM2 currencies).
The reason for non-aggregation of positive changes, within the EVE metrics, is not explained and would have a direct impact for our members. As such, we would seek for highly correlated currencies to be allowed to be included, subject to approval and oversight from our local supervisor.
Proposed implementation in EBA guidelines
We propose a factor of 80 % is used for the purpose of aggregating any positive EV or expected earnings between EUR denominated positions and ERM2 country denominated positions, though only where a formal agreement about a lower fluctuation band than the standard 15 % exits. As Denmark, for now, is the only ERM2 country, this would in practice mean between EUR and DKK risk.
When calculating SOT for EUR and DKK risk, institutions would for each of the 6 scenarios, when there is loss in one currency and a gain in the other currency, be allowed account for 80 % of any positive EV or expected earnings. When there is a loss in both currencies there will naturally be no reduction.
Justification
As stated above, there is, due to very close link between the Danish and the European Monetary Policy, herein the formal agreement on a very low fluctuation band for the EUR-DKK currency rate, a very high and stable correlation between DKK and EUR interest rates. This is recognized by the EU Finance Ministers in the FRTB proposal, with an 80 % correlation between GIRR buckets for EUR/ERM2 currencies (Study shows an average correlation of 0.85 between DKK and EUR rates).
The FRTB is a VaR like setup and consequently using correlations directly in the calculations. IRRBB is one the other hand working with specific stress scenarios. For that reason, we can’t just plug in a correlation. Instead we need a nominal hedge ratio reflecting the allowed amount of aggregation between the two currencies. As we are in a stress scenario, the relevant analysis is to quantify the nominal covariance between DKK and EUR rates, when we have large movements in the general rate level. This is the reason for “only” looking for the hedge ratio in stressed periods.
Using 16 years of data and looking at periods with large movements in the general rate level, study show that the hedge ratio between DKK and EUR is at or above 80 %.
Underlying analysis
Below is the 1 % and 5 % largest changes in DKK rates (x axis) and the corresponding change in EUR rates (y axis). Monthly changes are shown for 2Y and 10Y swap rates.
There is clearly a high nominal covariance between the two currencies. We define the hedgeratio as:
Hedge ratio: 1-(∑▒ABS(〖∆spread〗_t ) )/(∑▒〖〖ABS(∆DKK〗_t)〗)
The table below shows the hedge ratio for different maturities looking at both the 1% and 5% largest outright movements. The analyses is done on monthly changes.
From the above it seems reasonable to use a general hedge ratio of 80 % for all maturities.
We will be happy to present the proposal in more details for the EBA.
Kindly see enclosed file
Question 1: Are the definitions sufficiently clear? If not, please provide concrete suggestions and justify your answer.
Kindly see EBF responseQuestion 2: Are the guidelines in section 4.1. regarding the general provisions sufficiently clear? If not, please provide concrete suggestions.
Kindly see EBF responseQuestion 3: Do you agree that cash flows from non-performing exposures (NPEs) should be net of provisions and treated as general interest rate sensitive instruments whose modelling should reflect expected cash flows and their timing for the purpose of EV and earnings measures? If not, please provide concrete suggestions and justify your answer.
We agree in principal, but feel that a materiality threshold would also be needed. The proposed guidelines have a relatively short implementation period within which to develop new systems. If NPEs are a relatively small part of the overall Banking Book, and new system developments are required to capture the risk, our members would want to prioritize other areas that potentially had higher levels of risk.It is our view that a materiality threshold of 2.0% of (fixed rate) NLE over total loans should be adopted to provide some guidance on the inclusion of NPEs. A measure based upon EV risk could provide some clarity.
As NPE are usually handled by banks in specific systems, designed to manage for the credit status of these loans and potential recoveries and provisions, integrating these systems within the IRRBB calculation engines could present an issue. New significant investment on IT could be required, budgets for which are typically agreed the year before development. To achieve the proposed deadline of 31 December 2018, IT expenditure and project resources, would need to be been approved within the 2017 calendar year. It would also create a new model risk that would need to be included in the model sensitivity analysis and reports. For this reason longer period for implementation is needed.
For banks with limited NPEs, the cost benefit of this integration would be disproportionate. For those banks below the 2.0% threshold, considering NPE either as non-interest bearing or without provisions should be allowed. Also small non-complex banks should be allowed to treat NPE in a more simple way.
Question 4: Are the guidelines in section 4.2. regarding the capital identification, calculation, and allocation sufficiently clear? If not, please provide concrete suggestions and justify your answer.
As we read the guidelines, we understand from chapter 3, point 17. (page 8), that the objective of the SOT is “to inform supervisors about the exposure of institutions to IRRBB..”. That is, as we read this, a kind of warning indicator comparable across entities.However reading chapter 4.2 about capital identification calculation and allocation point 25 says: “Institutions should not only rely on the supervisory assessments of capital adequacy for IRRBB or on the outcome of the supervisory outlier test (see Section 4.5.), but should develop and use their own methodologies for capital allocation, based on their risk appetite, level of risk, and risk management policies”.
Are we supposed to read the “supervisory assessment” as the same as the “SOT number”? If so, we read point 25 that the SOT has a direct link to the quantification of required capital.
We urge the EBA to be very precise about the objective of the SOT, this both for a clear understanding of the EBA guidelines across institutions, but also to secure level playing field across regional supervisors.
Question 5: Do you agree with the list of elements to be considered for the internal capital allocation in respect of IRRBB to earnings in paragraph 30? If not, please provide concrete suggestions and justify your answer.
Kindly see EBF responseQuestion 6: Are the guidelines in section 4.3. regarding the governance sufficiently clear? If not, please provide concrete suggestions and justify your answer.
Kindly see EBF responseQuestion 7: Are the guidelines in section 4.4. regarding the measurement sufficiently clear? If not, please provide concrete suggestions and justify your answer.
Kindly see EBF responseQuestion 8: Do you consider the comparison between EV metrics calculated using contractual terms for NMDs with the EV metrics calculated with behavioural modelled assumptions sensible and practical? Please justify your answer.
Kindly see EBF responseQuestion 9: Are the guidelines in section 4.5. regarding the supervisory outlier test sufficiently clear? If not, please provide concrete suggestions and justify your answer.
Kindly see EBF responseQuestion 10: Is the proportionality adequately reflected in the guidelines, in particular in relation to the transitional period for SREP category 3 and 4 institutions and the frequency of calculation for the additional outlier test under paragraph 112?
No, although the guidelines make comments on proportionality it is not sufficiently reflected in the guidelines. Banks with limited interest rate risk in the baking book due to limited activity should be allowed to use simple and standardized stress tests with less frequent calculations and reporting.Question 11: If relevant, do you manage interest rate risk arising from pension obligations and pension plans assets within the IRRBB framework or do you cover it within another risk category (e.g. within market risk separately from IRRBB, etc.)?
Kindly see EBF responseQuestion 12: Which treatment of commercial margins cash flows do you consider conceptually most correct in EV metric, when discounting with risk free rate curve: a) including commercial margins cash flows or b) excluding commercial margins cash flows? Please justify your answer.
Kindly see EBF responseQuestion 13: Are your internal systems flexible enough to exclude margins for the purpose of calculating EV measures for the supervisory outlier test? If not, what would be the cost to adapt your systems (high, medium, low)? Please elaborate your answer.
Kindly see EBF responseQuestion 14: Do you consider the level of the proposed linear lower bound as described in paragraph 113 (k) appropriate? If not, please provide concrete suggestions and justify your answer.
The introduction of an absolute floor for interest rate is sensible, but we would like to use the opportunity to challenge the appropriateness in the suggested calibration.Based on historic observations, ECB behavior and communication a floor level of -1,00% for short term EUR rates with a gradual increase to 0% at the 30 year tenor stand out as the appropriate levels to bridge the typical aim of stress testing; combining highly unlikely and yet feasible outcomes. For some currencies a lower bound of even -1.00 % seems too low. We therefore recommend EBA to consider working with currency specific floors.
Below we show the proposed floor combined with the EBA floor suggestion and the historic absolute low at the relevant tenors. We have looked at EUR, GBP and USD. Is shows that a lower bound of -1.00 % is far below the lowest levels we have observed.
Further motivation
Lately as interest rate levels have become negative the ECB has been more and more cautious lowering the rates further with the latest two reductions as small as 0.1%-points. For two years the ECB has kept the rates unchanged even in an environment of continued monetary policy accommodation. Massive quantitative easening (QE) has taken place within the latest years as a substitute for further interest rate reductions. That shift in usage of monetary instruments is a testimony of their view, that further negative interest rates will cause an increasingly amount of unintended consequences, which is translated into the view that the overall cost outpaces the benefits of lower rates.
Reluctance to bring interest rates hugely into negative territory has been a common feature for central banks. At the same time, though there has been moderate difference in their mix of policy instruments. The FED actually never went negative and the Bank of Japan only went marginally negative before the QE was used in big scale. This observation speaks for adaptation of currency specific floors, calibrated around a short term ECB floor of -1%.
As negative short term rates still has to be regarded as an extraordinary situation, the interest rate structure will always to some extent signal expectations of in-creasing rates; i.e. a positive sloping curve inside a foreseeable future. For that reason, theoretically the floor should gradually move upwards as the tenors increases.
Kindly see enclosed file
Question 15: Do you consider the minimum threshold for material currencies included into the supervisory outlier test (5% for individual currency and minimum 90% of the total non-trading book assets or liabilities) sufficient to measure IRRBB in term of EVE? If not, please provide concrete suggestions and justify your answer.
Kindly see EBF responseQuestion 16: When aggregating changes to EVE in the supervisory outlier test, does the disregarding of positive changes to EVE have a material impact on the calculation of the supervisory outlier test?
The proposed change marks a significant change from the existing guidelines and would have an impact on Danish banks. The core principle of the supervisory outlier test (”SOT”) is to identify outlier banks and the exclusion of positive changes would represent a diversion from the methodology proposed with the BIS Standards in April 2016 . For Danish banks, the proposed change could potentially distort the results of the SOT and present a position that is not representative of the true underlying banking book risk.It is unclear why such a binary approach is being proposed, without an impact study. Our members have a long established banking book risk management practice, where high correlated currencies are used to hedge underlying risks. This is principally the use of EUR denominated instruments to hedge DKK denominated risks. This practice is well known to the local supervisor and driven by local market conditions, where it is not always possible to access the DKK market, in the size or duration required. This practice has proved to be an effective hedge for banking book risks through the economic cycle.
By disallowing such hedging practices within the SOT calculations, the comparability of the outputs, on a pan-European basis, would have to be questioned. Highly hedged Danish banks could potentially become outliers, a classification that might have unintended consequences. This could manifest itself in a reduction of banking book hedging activities, to manage the SOT risk. This is a concern for our members as they would not want to change the current practice within Denmark of maintaining a low banking book risk profile.
The SOT analysis should allow for aggregation of highly correlated currencies, for a comparable analysis to be effective. Such a method is supported by the analysis of historical correlations. The position of the EBF on this matter was articulated within the Industry IRRBB Working Group response dated 11 September 2015 within Appendix 8 (https://www.bis.org/bcbs/publ/comments/d319/igii.pdf). It is also worth noting that the EU Finance ministers recognized the importance of such correlations through the proposed changes to the EU Commission’s FRTB proposal in CRR2 (where increased correlations between GIRR buckets for EUR/ERM2 currencies).
The reason for non-aggregation of positive changes, within the EVE metrics, is not explained and would have a direct impact for our members. As such, we would seek for highly correlated currencies to be allowed to be included, subject to approval and oversight from our local supervisor.
Proposed implementation in EBA guidelines
We propose a factor of 80 % is used for the purpose of aggregating any positive EV or expected earnings between EUR denominated positions and ERM2 country denominated positions, though only where a formal agreement about a lower fluctuation band than the standard 15 % exits. As Denmark, for now, is the only ERM2 country, this would in practice mean between EUR and DKK risk.
When calculating SOT for EUR and DKK risk, institutions would for each of the 6 scenarios, when there is loss in one currency and a gain in the other currency, be allowed account for 80 % of any positive EV or expected earnings. When there is a loss in both currencies there will naturally be no reduction.
Justification
As stated above, there is, due to very close link between the Danish and the European Monetary Policy, herein the formal agreement on a very low fluctuation band for the EUR-DKK currency rate, a very high and stable correlation between DKK and EUR interest rates. This is recognized by the EU Finance Ministers in the FRTB proposal, with an 80 % correlation between GIRR buckets for EUR/ERM2 currencies (Study shows an average correlation of 0.85 between DKK and EUR rates).
The FRTB is a VaR like setup and consequently using correlations directly in the calculations. IRRBB is one the other hand working with specific stress scenarios. For that reason, we can’t just plug in a correlation. Instead we need a nominal hedge ratio reflecting the allowed amount of aggregation between the two currencies. As we are in a stress scenario, the relevant analysis is to quantify the nominal covariance between DKK and EUR rates, when we have large movements in the general rate level. This is the reason for “only” looking for the hedge ratio in stressed periods.
Using 16 years of data and looking at periods with large movements in the general rate level, study show that the hedge ratio between DKK and EUR is at or above 80 %.
Underlying analysis
Below is the 1 % and 5 % largest changes in DKK rates (x axis) and the corresponding change in EUR rates (y axis). Monthly changes are shown for 2Y and 10Y swap rates.
There is clearly a high nominal covariance between the two currencies. We define the hedgeratio as:
Hedge ratio: 1-(∑▒ABS(〖∆spread〗_t ) )/(∑▒〖〖ABS(∆DKK〗_t)〗)
The table below shows the hedge ratio for different maturities looking at both the 1% and 5% largest outright movements. The analyses is done on monthly changes.
From the above it seems reasonable to use a general hedge ratio of 80 % for all maturities.
We will be happy to present the proposal in more details for the EBA.
Kindly see enclosed file