Response to consultation on Technical standards on the new Business Indicator framework for operational risk
Question 1: What are your views with regards to the proposal for the ILDC component? Please explain and provide arguments for your answer.
The mapping on ILDC component requires to identify the “assets leasing component” in the FinRep items. Clarification would be greatly appreciated on the indications in terms of references to the phenomenon, what is meant by “leased asset” that are not clear in the current FinRep reporting framework. Given the previous request, it would be more appropriate to use the internal accounting system to identify the “assets leasing component”.
The definition of “interest earning assets” still contains a substantial portion of assets which are actually non-interest earning, primarily from brokerage receivables (predominantly settlement balances) and non-interest earning deposits with central banks and other banks. Specifically, we suggest that for the two relevant FinRep items (F18_010_005 Cash balance at central banks and other demand deposits as well as F18_010_070 Loans and advances), additional clauses (“only those earning/bearing interests”) similar to derivatives (FinRep F_1.1_10_60) shall be introduced in order to only capture the interest earning parts, ensuring an adequate measurement.
Furthermore, we believe that within the Asset Component, the following elements should be considered as "leased assets":
• leases in which the institution is a lessee, falling within the scope of IFRS16;
• operating leases where the institution is the lessor and therefore the asset continues to be represented under its own balance sheet items.
Consequently, with regard to the related items included in the IC component, the inclusion of the two following items of income statement would not seem applicable as they typically are related to financial credits, including the financial lease in which the bank is a lessor (out of scope according to the element to be included in within the asset component explained above):
• the "Modification gains or (-) losses, net" of financial leasing contracts where the institution is a lessor;
• the "Impairment or (-) reversal of impairment on financial assets not measured at fair value through profit or loss" relating to financial leases in which the institution is a lessor.
• Interplay between ILDC, FC and “clean price”: Regarding the determination of the ILDC component and the Financial Component, it appears that institutions using the so called “clean price” to produce the FinRep reporting might be penalized compared to institutions using the “dirty price” approach.
As a reminder the clean and dirty price approaches are detailed in Instructions for FinRep reporting (Annex 5) as detailed bellow:
“Interest income and interest expense from financial instruments measured at fair value through profit or loss and from hedging derivatives classified in the category ‘hedge accounting’ shall be reported either separately from other gains and losses under items interest income’ and ‘interest expense’ (‘clean price’) or as part of gains or losses from these categories of instruments (‘dirty price’) (...)
‘Interest income. Financial assets held for trading’ and ‘Interest expenses. Financial liabilities held for trading’ shall include, where the clean price is used, the amounts related to those derivatives classified in the category ‘held for trading’ which are hedging instruments from an economic but not accounting point of view to present correct interest income and expenses from the financial instruments that are hedged (...)
Dividend income on equity instruments measured at fair value through profit or loss shall be reported either as ‘dividend income’ separately from other gains and losses from those classes of instruments where the clean price is used, or as part of gains or losses from those classes of instruments where the dirty price is used”.
Institutions using the clean price approach reclassify interest income & expenses and dividend incomes from gains and losses from instruments held for trading or instruments designated at fair value through profit or loss to interest incomes & expenses and dividend income within F02.00. This does not constitute an accounting choice or method but a reporting reclassification. These reclassifications are made for FinRep reporting purpose only.
However, in a situation where the P&L of the trading book of an institution would be negative, this reclassification would result to: i) deepen the loss of the P&L of the trading book and ii) increase the amounts of dividend incomes (as for example dividend revenues stemming from equity instruments measured at FV in Trading book would be reclassified in dividend income).
That will mean that an institution using clean price approach for FinRep would be penalized compared to an institution using the dirty price approach as, considering the same P&L profile, in case of negative P&L of the trading book, one institution using clean price would see its FC (based on absolute value) increased by the amounts related to dividend revenues or net interest and its ILDC increased by the same amount, where another one using dirty price would not suffer from this effect.
We consider that institutions using clean price approach should not be penalized by the application of the clean or dirty price approach and ask EBA to allow institutions to neutralize the negative impact of the reclassification made for FinRep purposes (with the same reclassification mechanism that the one used for AA to PBA approach but allowing reclassification of dividend revenue and interest incomes/expenses from ILDC to FC).
Question 2: What are your views with regards to the proposal for the Services component? Please explain and provide arguments for your answer.
We interpret the function of the FinRep lines F45.3_020_040 and F2_010_370/380/390/430 with the additional clause “not due to operational risk events and not due to leases” as adding the Operational Risk losses to the Service component. In this regard, we believe that the annual amounts from Operational Risk losses are better sourced from the respective events recording process than directly from the financials. Hence, we suggest providing the option to replace these FinRep lines with the total loss amount as recorded in the Operational Risk events recording database.
In our opinion, it is not entirely clear the proposed mapping, which requires to identify in the FinRep item the only component connected to the operational risk. We are referring to the mapping of “Other operating expense” for the following items FinRep: F02.00_r0370 (due to operational risk and not due to leases); F02.00_r0380 (due to operational risk); F02.00_r0390 (due to operational risk) for this information the only source reliable is the operational risk losses data base.
In addition, Article 6, paragraph 2, stipulates that “For the purposes of point (d), the losses, expenses, provisions, and other financial impacts due to operational risk events shall not be net of any related payments received from insurance or reinsurance policies purchased”. We would welcome the confirmation that this is possible to net of payments different from the one received from insurance to reinsurance policies purchased.
By comparing the list of items as for draft RTS “on the components of the BI under Art. 314(6) of the CRR”, Art. 7 (Following: RTS list) with the list of items as for Draft ITS “on the mapping of the BI components with corresponding supervisory reporting under Art. 314(7) of the CRR” (Following: ITS list) we noticed that RTS list does not include the item “fee and commission income from loan commitments given” (which is included in the ITS list, mapped with F22.01r0200_c0010 FINREP item). Should this item be included also within RTS list?
With reference to the calculation of the Other Operating Income, Article 16 of the RTS does not explicitly exclude from the calculation of the Business Indicator recoveries of expenses on behalf of customers (stamp duty, substitute tax and other recoveries), although Article 314 para. 5 let. d provides for such exclusion.
With regard to the calculation of the Fee & Commission Expenses for which the inclusion of the outsourced financial services expenses in required, entered in field F02.00_r0360_c0010 ("Administrative expenses"), we would kindly ask the EBA to clarify the nature and type of financial services, with respect to which the inclusion of these elements in the Fee & Commission Expenses is defined.
The new Standardised Measurement Approach (SMA) is providing a generic model for banks, yet for entities whose income only comes from “fees from payment services” the component can be misleading as it does not take into consideration any risk mitigation, since the Internal Loss Multiplier (ILM) was removed. Thus, it makes the capital charge somewhat insensitive to the quality of the internal control environment.
Furthermore, due to lack of risk sensitivity in the Services component due to this ILM removal and the subsequent impact it has for entities, whose income only comes from fees from payment services, a more appropriate way in the Services component would be to take into account the fee expenses and include that as a net of the commission income. Assuming a particular internal process regardless of the amount of transactions, an offsetting due to the consideration of net commission income as a basis for the SC should be rational; given the process and its operational risk this should not increase linear but rather decrease all the more often as the process is being executed. Hence, this approach would reflect the reality of the business of the said entity in a more realistic way, since fee expenses and fee income usually correlate, which will be reflected in the same way as it does for IC component under ILDC.
Question 3: What are your views with regards to the proposal for the Financial component? To which extent are you carrying out operations or making accounting choices as referred to under paragraph 2, point a) of Article 9 of this draft RTS? Are you carrying out operations or making accounting choices, other than those specified under paragraph 2, point a) of Article 9 of this draft RTS, that could justify the use of the PBA? Please explain and provide arguments for your answer.
According to Article 314.4 of the CRR, the trading book component (TC) of the financial component (FC) should be “defined as appropriate either in accordance with accounting standards or, in accordance with Part three, Title I, Chapter 3” (i.e., the prudential boundary criteria). This requirement is also explicit in the mandate of Article 314.6 which requires that EBA develops the list of typical sub-items of the business indicator by “taking into account international regulatory standards and, where appropriate, the prudential boundary defined in Part three, Title I, Chapter 3”.
The draft RTS in the consultation does indeed provide the two required approaches to calculate the financial component in the proposed accounting approach (AA) and prudential boundary approach (PBA). There are, however, several issues with the requirements envisaged by EBA.
First and foremost, the AA has been made the default approach while the PBA can only be used by way of derogation after meeting some conditions. This requirement goes beyond that of the CRR which does not favour one approach over another but, rather, requires that the PBA be available where and as appropriate. An institution should therefore be able to choose the PBA on a permanent basis if it considers such approach as appropriate. It should be noted that the CRR already imposes very strict requirements for the management of the trading book including for the inclusion of positions (Articles 102, 103 and 104), and equally strict rules to reclassify a trading book position (Article 104a) which contributes to the robustness of the PBA approach. However, to avoid continuous changes from one approach to another, it appears reasonable that when an institution has made the decision to apply the PBA, it would only be permitted to revert to the AA (or the other way round) if such change is triggered by material evolutions of its activity, environment or risk management (for example a change of business model) and after approval from the competent authority only. This would provide the consistency required for having a sound framework for the PBA and would ensure that no regulatory arbitrage is possible which, as previously mentioned, is already prevented by trading book framework of the CRR.
Second, in the approach proposed in the RTS the application of the PBA is conditional on several criteria including the presence of certain operations or accounting choices that result in an “unwarranted increase” of the FC when using the AA. Once again, this would limit the usage of PBA while CRR does not favor one approach over another nor intend to limit the usage of PBA. Furthermore, an unwarranted increase in the TC’s P&L over a certain period can be volatile by definition as it can be impacted by several market factors. An institution can therefore experience an unwarranted increase in a given reporting period and not experience any in a following reporting period while having similar operations and accounting choices. The application of the PBA should, therefore, not be based on an unwarranted P&L increase in the TC nor be subject to any limitation. In any case, the potentiality of such increase demonstrated ex ante shall be sufficient and would avoid volatility.
Finally, the notification process seems very cumbersome, especially as all the requirements (points (a) to (h) of Article 13.2 of the RTS) should be reviewed annually. All these requirements should only be required for the initial notification of the intention to use the PBA and the annual review should be limited to the independent review on the fulfilment of the conditions to use the PBA (point (h) of Article 13.2 of the RTS).
Question 4: What are your views with regards to the proposal for the specification of the items to be excluded from the BI? Please explain and provide arguments for your answer.
As a general remark, we note that Chapter 4 (Article 16) is titled “Elements to be excluded from the business indicator”. However, the Chapter enlists what is not excluded, which can be quite confusing to apply. Thus, it would be more helpful to have an actual list of exclusions – otherwise, it would be challenging to ensure the completeness of exclusions.
In this sense, income and expenses from insurance or reinsurance business are to be excluded from the calculation of the business indicator under CRR3. However, Art.16 of the RTS specifies that where these are “resulting from the distribution of insurance or reinsurance products or services”, they shall not be excluded. Given that EU credit institutions are not able to act as insurance providers[1], the clarification included in recital 12 and Article 16 paragraph 1 point (a) of the RTS would invalidate CRR3 exclusion from the BI of income and expenses from insurance or reinsurance business, as no amounts comply with such characteristics (i.e.: income or expenses from the insurance business in a credit institution not resulting from the distribution of insurance products). Therefore, eliminating this clarification introduced by the RTS would invalidate Article 314 paragraph 5 point (a) clause in the CRR.
Additionally, we would like to note that the Solvency II Directive already includes an operational risk capital charge for insurance and reinsurance products as well as services. That being said, by not excluding these categories from the Business Indicator (BI) would lead to a capital double counting situation.
Finally, income and expenses from the distribution of insurance products are not exempted from the BI exclusions in the UK Regulation issued by the PRA.
Therefore the proposal for the specification of the items to be excluded from the BI should be modified as follows:
“The exclusions referred to in Article 314(5) Regulation (EU) No 575/2013 shall be applied as follows:
- for the purposes of Article 314(5) point (a) of that Regulation, income and expenses resulting from the distribution of insurance or reinsurance products or services manufactured or provided by an insurance company falling outside under Directive 138/2009 (:Solvency II Directive) or under third country regulations that are considered equivalent to it under Articles 172, 227 and 260 of the said Directive, shall be excluded from the calculation of the business indicator”.
The proposed specification of the items to be excluded from the BI introduces an undue burden on banks’ capital requirements, causing them to lose their risk sensitive attribute (driving unexpected loss quantification beyond confidence level for banks adopting a fee-based business model) and introducing very high volatility on changes in expected T1 Monthly Recurring Charge (hereinafter: MRC), as testified by most recent QIS studies.[2]
From a systemic point of view, operational risk capital charge aimed at covering unexpected losses arising from insurance or reinsurance products and services is already fairly set aside by EU-based insurance companies falling within Solvency II Directive scope or third country regulations that are considered equivalent to it. These fees coming from the distribution of insurance products are already subject to capital requirements under Solvency II Directive, hence laying down additional operational risk capital requirements would lead to a capital double counting situation.
From a bank’s specific point of view, under the new standardized approach the service component of the Business Indicator is now defined as the maximum amount between “Fee Income” and “Fee Expenses”, while in the current methodology the same component is given by Net Fee margin. Taking into account the fact that insurance products distribution drives lower operational risk than financial products distribution because of their very nature of protecting customers’ unforeseen events, we see the proposed change as an effective way of reducing unlevel playing field driven by excessive MRC volatility, while keeping unchanged the original Basel 3 goals in revising current framework.
Interest income (Article 1 of the RTS on BI items): we agree with the mapping proposed. For the leasing component requested in clear in the mapping, as it’s not in clear in the FinRep reporting framework item mentioned, we will use the available information in the accounting system (for example FR_45.03 row 010 column 010 only from leased assets).
Interest expenses (Article 2 of the RTS on BI items) we agree with the mapping proposed. For the leasing component requested in clear in the mapping, as it’s not in clear in the FinRep reporting framework item mentioned, we will use the available information in the accounting system (for example FR_45.03 row 010 column 020 only from leased assets).
The elements "Modification gains or (-) losses, net" and the "Impairment or (-) reversal of impairment on financial assets not measured at fair value through profit or loss" will not be considered as related to financial leasing where the bank is the lessor as indicated in Question 1.
Also, one of the items to be excluded from the calculation of BI should be the realised profits/losses from the sale of non-trading book items or income from extraordinary or irregular items. We believe that these items do not reflect the institutions’ “business as usual” and are not reflective of their operational risk profile. For instance, the sale of an Assets and Liabilities Committee (ALCO) portfolio, which, in principle, is expected to provide income along a certain period of time.
We would like to highlight that there seems to be a discrepancy between Article 314.5 of CRR3, which lists the elements that should be excluded from the calculation of the Business Indicator, and some of the items being requested in Template 16.03.
So, while Article 314.5 states that “institutions shall not use any of the following elements in the calculation of their Business Indicator”:
(a) “Administrative expenses, including staff expenses, oursourcing fees paid for the supply of non-financial services, and other administrative expenses”
(f) “depreciation of tangible assets and amortisation of intangible assets, except the depreciation related to operating lease assets, which shall be included in financial and operating lease expenses”;
… the Template 16.03 is not consistent with this Article and requests the following fields:
- 0020 “Administrative expenses to operational risk events”
- 0030 “Depreciation due to operational risk events”
It would be necessary to amend fields of Template 16.03, by:
- Removing 0020 to comply with the exclusion of Administrative expenses
- Amending the content of 0030 so that it corresponds to “Depreciation related to operating lease assets”.
[1] Art.4.1.1CRR Excludes insurance undertaking from the definition of credit institution.
[2] See for instance EBA/REP/2023/32, Chapter 5.
Question 5: What are your views with regards to the proposed mapping of the BI items to the FINREP cells? Please explain and provide arguments for your answer.
The mapping carried out here between the BI components in accordance with Article 316 (6) CRR III-E and the FinRep templates in accordance with Annex III and IV of the reporting ITS is based on the reporting items of a full FinRep report. However, certain institutions report FinRep as data point or simplified approach users. A review of the mapping has shown that a large proportion of the reporting items referenced here cannot be reported by data point or simplified approach users on the basis of this mapping. This applies in particular to template C 16.02, which requires a very high degree of granularity. A corresponding mapping should therefore also be provided for these institutions.
If alternative reporting items are available for non-full FinRep users, these will be used. If this alternative is not available, the reporting item will not be filled. This will have the consequence that a higher value could tend to be used in the reporting items, as totals items are reported, among other things.
If reporting items are named in the FinRep reporting items that are not relevant for HGB users (see completion instructions for FinRep), the reporting item cannot be completed in the Operational Risk reporting form.
Concerning the “Asset component” (Article 3 of the RTS on BI items, page 34):
The described “total assets” comprise gross carrying amount positions and carrying amount positions. However, total assets are defined as carrying amount. In order to deliver reliable data, all required positions which relate to template F18.00 should include column 0130 “accumulated impairment”.
“Cash balance at central banks and other demand deposits” (the position asset component (a) does not contain “cash on hands” (ref. F01.01. Cash, cash balances at central banks and other demand deposits)
We propose to provide a clearer correspondence between the items included in the calculation of the business indicator and the according positions of the FinRep templates to narrow down the scope of interpretation and therefore achieve better harmonization level across all institutions.
Double counting: Interest income “(k) profits from leased assets including gains from lease modifications” F45.03_r0040_c0010 (only from leased assets) versus other operating income “(b) income from other income” F45.03_r0040_c0010 (in total).
Moreover, the calculations of operational risk requirements are based on the average of the values of the last three financial years. Taking also into account that operational risk losses will be materialised in the currency in which the company operates, we consider that the exchange rates to be applied in the case of subsidiaries with a currency other than the € for the purposes of calculating own funds requirements for operational risk should be the exchange rate at each reporting date (use of fixed exchange rate)
Otherwise, should FinRep exchange rates were used, capital requirements would be:
-outdated: up to a 3-year gap since historic FX rates would be used; and
-inconsistent with credit and market risk frameworks, as these two are calculated with up-to-date FX rates.
Question 6: What are your views with regards to consider the financial statements used for the final valuation as the only reference for the acquisition of activities under the baseline approach (i.e. full historical data)? Please explain and provide arguments for your answer.
With a view to Article 1, paragraph 1, “Institutions shall include items of merged or acquired entities or activities in their business indicator for the last three financial years based on the audited financial statements of those entities or the financial information used for the final valuation of those activities”. We consider including the term “external” audited, for the sake of clarity.
In addition, we believe that financial statements alone are not a good reference in this case as they may not be available in all cases.
Question 7: What are your views with regards to the proposed three alternative calculation approaches instead of a unique alternative approach to be defined? Please explain and provide arguments for your answer.
With reference to Article 1, paragraph 2, “If the data referred to in paragraph 1 is not available or accurate, institutions shall adopt, among the below alternative approaches, the one that results in the highest own funds requirements for operational risk:[…]”.
In our view, requiring financial institutions to apply among three different methods the most conservative constitutes an excessive effort. Therefore, we propose to define a ranking among these three methods, in order to apply only the first one which is feasible. The reported order c), a), and b) can be applied for this purpose.
As an alternative, we want to highlight the possibility to use the M&A factor method (according to Article 1 (2) (b) generally (e.g. as first resp. standardized approach), for example up to an M&A factor of 1.1. This would be a great relief for the institutions, yet it will only have a minor impact on the business indicator. Up to an M&A factor of 1.1, a notification according to Article 1 paragraph 4 of the afore-mentioned RTS should be omitted as well as the comparison of the three approaches according to Article 1 paragraph 2 of the RTS.
Accordingly, the three alternative calculation approaches should be replaced by a unique alternative approach based on the acquired company’s first financial statements after its acquisition. The historical data should be rebuilt based on an annualised quarterly financial statement as shown below (adjusted with proposal on issue regarding the timeline for calculation of the operational risk capital requirements in the general comments):
[a table/Annex is inserted in this part but not supported in this section - for this purpose, please kindly consult Question 7 of the attached PDF document hereunder]
With reference to Article 1, paragraph 4, “Institutions shall notify their competent authority when including acquired or merged entities or activities items in accordance with paragraph 2. This notification shall provide evidence of the unavailability or inaccuracy of data referred to in paragraph 1 and present the own funds requirements for operational risk as calculated in accordance with paragraph 2, points (a), (b) and (c).”
We are wondering how institutions shall notify this inclusion; based on Article 3, it appears to be an ex-ante notification. Would it be enough to send emails with the data/information required in Article 4?
Question 8: What are your views with regards to not providing any alternative method but adjustment to the effective perimeter of the disposal? Please explain and provide arguments for your answer.
Please refer to Question 7.
Question 9: What are your views with regards to the inclusion of a threshold? Please explain and provide arguments for your answer, as well, if applicable, further evidence on situations where BI adjustments as set out under articles 1 and 2 would not be feasible or deemed excessively cumbersome and identify potential consequences on the dynamics of the European financial markets.
According to the new framework, the data of the new companies is supposed to be integrated into the FinRep figures based on the (audited) financial statements of the respective companies. Adjusting FinRep figures in this way for past years is very burdensome, especially if there have already been business relationships with the respective companies before the M&A transaction took place. On the other hand, the impact of the adjustments on the business indicator is likely to be of minor impact in most of the M&A transactions and does not justify the efforts for the re-computation of FinRep figures. It should also be noted that any such M&A, even if it is below the threshold, is already reflected in the next YE Business Indicator / RWA calculation following the M&A, resulting in full recognition only very few quarters after the M&A, so the focus should only be on material intra year impacts. Against this background, we consider facilitations for M&A transactions with minor impact necessary and appreciate the inclusion of a threshold to have a de minimis criterion for focusing on the M&A that create a substantial impact on the BI.
We suggest that the threshold be a relative measure below which the past data are not modified, and to be aligned with other out of cycle adjustments to financial statements (e.g. late adjustments to YE financials).
Again, it would be more useful to have a threshold under which the disposed contributions could be excluded without a prior notification (or submitting just an ex-post notification). For example, if total operating income, net (FR 02 r355) of the disposed Company/ total operating income, net (FR 02 r355) of the Institution <=2%, it should be possible to exclude the related contributions by submitting an ex-post notification (like the criteria applied in the EU Regulation 529/2014 for the AMA extensions).
Question 10: What are your views with regards to the basis for the calculation of the threshold? Please explain and provide arguments for your answer.
As a reasonable measure of a simple basis for the threshold can be the total operating income, in order to avoid recalculating all the business indicator items.
As an alternative, immateriality could also serve the M&A factor already introduced by the EBA. Up to an M&A factor of 1.1, M&A adjustments could be omitted. We also want to highlight the possibility to use the M&A factor method as first response standardized approach for adjustments to the business indicator; i.e., up to an M&A factor of 1.1, a notification according to Article 1 paragraph 4 of the above - mentioned RTS should be omitted as well as the comparison of the three approaches according to Article 1 paragraph 2 of the RTS.
In case of a M&A, it might be useful to set materiality alternative thresholds (regardless of the level – individual, sub group or consolidated):
- TOTAL OPERATING INCOME, NET (FR 02 r355 c010) of the merged Company/ TOTAL OPERATING INCOME, NET (FR 02 r355) of the Institution <=2%
- operational risk requirement for of the acquired company (from the last PILLAR III of the company) / the total of the purchaser operational risk requirement summed up with that of the acquiree>=2%.
Question 11: What are your views with regards to the level you consider would be appropriate for the threshold? Please explain and provide arguments for your answer.
It is worth mentioning that, within a banking group, the capital requirement is calculated at several levels, e.g. group, sub-group, legal entity level. In the definition of a possible threshold, we suggest considering the peculiarities of the calculation level.
In case an activity is transferred among legal entities or a merger occurs among different legal entities within the group, the modification of past data should not be required at individual level, if the total operating income of the activity is transferred or the one of the merged company is lower than 5% of the group. Eventually, there should be no impact at consolidated level.
In case an activity is transferred among legal entities, below some threshold, it should not be required to modify the past data (legal entity level).
At group level, in case of disposal, we do not see the need to define a threshold under which past data have not to be modified. It will be up to the banking group to decide if the disposal is significant enough to ask for immediate exclusion of contributions related to disposed parts.