Response to consultation on Implementing Technical Standards amending Commission Implementing Regulation (EU) 2021/451 on supervisory reporting
Question 1: Are the instructions and templates clear to the respondents?
Since no information was shared about how z-axis of exposure classes of STA and IRB will look like in the future the instructions and templates are not clear to us. The missing dimension for C07 and C08 templates is essential for understanding the upcoming changes to exposure classes.
Furthermore, the term ‘sub-exposure class’ is not specified nor its nature and purpose. In particular we would request more information on the intention / objective and the consequence of such new exposure classes. Is it similar like (main) exposure classes or something completely different?
New column 020 in C02 is not fully described . The content of the column is clear but how to proceed in the lines of the template is not fully clear: Is our assumption correct that the exposure classes should not be reported in column 020 even though in C10 you transmit from IRB exposure class to STA expo-sure class – so the output floor is shown quite differently in C10 compared to C02?
How is the reporting defined for transitional provisions in C34.02 in columns 230-250: Are those columns only to be filled when Standardised Approach is not applied? Or must those columns be filled in any case (no matter which approach is applied for SA-CCR)?
Why are some templates of FRTB transferred and some are not included in COREP Annex I report? Since some of market risk RWA will come from template C91 (FRTB) we do not understand why this template will not be included in COREP Annex I report.
Additionally, it is still not very clear which exposure classes will be opened for the C07 (STD) and C08.1/C08.02 (IRB) reports because after reviewing the ITS of the changes in the IRP vs COREP the openings of the categories are different when they should be very similar between IRP and COREP.
For example: The IRP report details 31 exposure categories in STD (EUCR4), however COREP's “Solvency Annex document” lists only 16 categories.
Currently the C.02 reports the RWAs under the IRB and STD approaches match the C08.1 and C07 reports respectively. With the regulatory change to CRRIII, C02 will report RWAs applying the Output-floor TREA and STREA, therefore we understand that C02 will no longer coincide with C08.1 generating confusion with the message that appears: "Cell linked to CA".
Furthermore, with regard to templates containing IRB as well as SA requirements, it should be made clearer that reporting requirements (rows, columns) concerning internal models (i. e. output floor; TREA) shall only be reported by institutions using internal models. Otherwise, institutions using the standardized approach would report certain aspects multiple times. For example, see template C 03.00 rows 0070 – 0090 where SA-CR institutions would repeat their capital ratios. We suggest adding “For institutions subject to the output floor…” as in the instructions for column 0020 of template C 02.00.
According to Annex II, there is a column 90 in template C 08.06 (CR IRB 6) where the expected loss amount is supposed to be reported. However, this column is missing in the corresponding reporting templates in Annex I. Could you please clarify whether the expected loss amount should be reported in Template C 08.06 or not?
Question 2: Do the respondents identify any discrepancies between these templates and instructions and the calculation of the requirements set out in the underlying regulation?
We do not understand row 011 in C0902 ‘of which: regional governments or local authorities’ – since this is displayed in row 012 (without of which)- can you please elaborate what is difference between row 011 and row 012?
Template C 02.00
- Columns 0010 and 0020 are supposed to present the amounts of TREA and S-TREA, respectively. At the same time, the headline in row 0036 suggests that the following rows shall present pre-floor REA amounts. However, according to Art. 92(3), TREA is the maximum of the aggregate unfloored REA (U-TREA) and the aggregate standardised REA multiplied by the respective floor factor (x*S-TREA), meaning TREA is the REA amount after application of the output floor.
Template C 05.01/02
- As far as we understand, the transitional provisions concerning templates C 05.01. and C 05.02 do not apply anymore. In this case, those templates could be deleted. If not, please provide further instructions.
Template C 10.01/02
- It is not clear how to report equity positions subject to transitional provisions according to Article 495 CRR3. According to para. 26 of the Consultation Paper equity exposures subjects to transitional provisions must be reported in the existing IRB-templates C 10.01/C 10.02. Nevertheless, according to para. 99 of the corresponding instructions in Annex II, equity exposures grandfathered in accordance with Article 495 (1) of Regulation (EU) No 575/2013 shall not be reported there. Could you please clarify where to report those equity positions? In our opinion it would be appropriate to include those equity positions in template C 07.00 and introduce new rows for the equity risk weights as the IRB-approach is no longer applicable.
Template C 07.00
- There is need for a clarification of the reporting of retail exposures which must be multiplied by 1.5 according to Art. 123a CRR3. According to Art. 123 (3) CRR3 retail exposures are assigned a risk weight of 75 %. If the multiplier of 1.5 is applied the final risk weight would be 112.5%. In template C 07.00 no specific row is provided for this risk weight and according to Annex II the row 0280 “other risk weights” is not available for the exposure class retail. Could you please clarify how to deal with this issue?
EU CMS2:
Looking into column a “RWEAs for modelled approaches that institutions have supervisory approval to use” and row 3 “Equity” (blank data point) is not included in the total cell row 9 “Total” showing datapoint Sum of rows 1, EU1a, EU1b 2, 5, 6, 6b and 8. Does this mean that values within IRB should not be included, for example of C10.01?
The correlation between row EU 7b “Equity” of table CR4 datapoint is {C 07.00, r0070, c0040, s0014} and of the CMS2 the datapoint is {C 07.00, r0070, c0220, s0016} + {C 07.00, r0080, c0220, s0016}. Which is the correct sheet to extract data from s0014 or s0016?
In row 6 “Retail”, column a “RWEAs for modelled approaches that institutions have supervisory approval to use”, the first data point {C 08.01, r0010 - r0040 - r0050 - r0060, c0260, sum(s0013-s0017)} indicates that it contains the sum of sheets numbered 0013 to 0017. However, upon examining the subsequent data point {C 08.01, r0010 - r0040 - r0050 - r0060, c0260, s0015}, it appears that sheet 0015 is also included. Isn't it already encompassed in the initial statement?
In row EU 5.2a, under the exposure class "Of which: Corporates – General" and in column C labeled "Total actual RWEAs" the data displays the total IRB exposures. However, it appears that standard exposures may be absent from this specific data point. Is this correct?
The correlation between the “Total actual RWEAs” in column c and the “RWEAs for modelled approaches, which institutions have supervisory approval to use”, in column a of row “EU 5.2a, specifically for Corporates – General”, involves two data points. The first data point {C 08.01, r0010 - r0040 - r0050 - r0060, c0260, s0020} +{C 08.01, r0010 - r0040 - r0050 - r0060, c0260, s0025} gathers data from sheet 0025, while the second one {C 08.01, r0010 - r0040 - r0050 - r0060, c0260, s0020} + {C 08.01, r0010 - r0040 - r0050 - r0060, c0260, s0024} extracts data from sheet 0024, both derived from the second equation. Could you please confirm if this understanding is correct?
For “Total actual RWEAs” in column c and row “EU 6.1b, specifically for Retail – Other”, only data points from the IRB exposures are included {C 08.01, r0010 - r0040 - r0050 - r0060, c0260, s0016} + {C 08.01, r0010 - r0040 - r0050 - r0060, c0260, s0017}. Therefore, there isn’t exposure of SA approach. For example, this applies to data extracted from s0010 of C07. Is this clarification correct?
The correlation between “RWEAs calculated using full standardised approach” column d and the “RWEAs for modelled approaches, which institutions have supervisory approval to use”, in column a of row “EU 6a, Of which: Retail - Categorised as secured by mortgages on immovable properties and ADC exposures in SA”, involves two data points. The first data point gathers data from row 250, while the second one extracts data from row 160. Could you please if the correct row to be taken into account is r0160 and not r0250?
Question 3: Do the respondents agree that the amended ITS fits the purpose of the underlying regulation?
ITS for C0902 row 011 refer to Article 147 (3a) which is new in CRR III. Still the purpose of row 011 versus row 012 is totally unclear. What is the difference between Article 147 (3a) and 147 (2)? Why is the reference in Article 147 (3a) taking care of regional governments, local authorities and public sector entities whereas in Annex I the row 011 is named: ‘Of which: Regional governments or local authorities. Is the purpose showing regional governments, local authorities, and public sector entities in separate exposure classes and at the same time having some of these exposures remaining in exposure class central governments and central banks depending on application of Art. 115 or 116? This regulation does not provide any clarity or more transparency it is rather confusing and raises many questions.
Question 4 - Cost of compliance with the reporting requirements: Is or are there any element(s) of this proposal for new and amended reporting requirements that you expect to trigger a particularly high, or in your view disproportionate, effort or cost of compliance? If yes, please: ▪ specify which element(s) of the proposal trigger(s) that particularly high cost of compliance, ▪ explain the nature/source of the cost (i.e. explain what makes it costly to comply with this particular element of the proposal) and specify whether the cost arises as part of the implementation, or as part of the on-going compliance with the reporting requirements, ▪ offer suggestions on alternative ways to achieve the same/a similar result with lower cost of compliance for you. ▪ what are your views on introducing more granular reporting in Step 2 in credit risk IRB templates C 08.XX to include obligor or loan level reporting? Explain the nature/source of the cost and the benefits.
In general, more details of (sub)exposure class do cost additional effort and therefore we ask to re-consider this if it really has an additional benefit. We do not see this benefit.
Moreover, the details on output-floor including the application of the transitional provisions related to the output-floor delivers more complexity and will create huge efforts for the implementation in the internal systems and statutory reports.
- Prioritization and eliminate less relevant data points
The number of templates and data points requested for supervisory reporting has been considerably increased in recent years (e.g., FRTB templates, IRRBB templates, new C 10.00, CVA template C 25, ESG-Add-hoc etc.). We understand the fundamental need for information on supervisory relevant data for supervisory authorities. However, there should be a focus on truly essential information for regulators, so that resources are not used for non-relevant or insignificant data points. Therefore, with each introduction of new reporting requirements, a review of the existing data points should be carried out. If new requirements are considered as more important than existing data points, old templates/data points with lower priority should be removed. To avoid a disproportionate increase of cost of compliance we consider it necessary to remove existing data points with every new requirement.
- In this regard we propose the following amendments to the ITS:
- Large institutions according to Art. 4 (146) CRR should have the general option to report all values in millions (e.g. million EUR) and all validation rules should accept small deviations due to rounding.
- Delete template C 05.01 - TRANSITIONAL PROVISIONS (CA5.1) – In practice, there should be little relevant information here, as the vast majority of transitional provisions have been expired.
- Delete template C 05.02 - GRANDFATHERED INSTRUMENTS: INSTRUMENTS NOT CONSTITUING STATE AID (CA5.2) – In practice, there should be little relevant information here, as the vast majority of transitional provisions have been expired.
- Waiver of template C 32.01 - Prudent Valuation: Fair-Valued Assets and Liabilities (PRUVAL 1) for all institutions that exceed the relevant threshold and fill in templates C 32.02-04
- Waiver of C 90.00 - Trading book thresholds (TBT) – for institutions that exceed the relevant threshold and provide the relevant detail templates for market risk
- Delete template C 43.00 - ALTERNATIVE BREAKDOWN OF LEVERAGE RATIO EXPOSURE MEASURE COMPONENTS (LR4) – With introduction of parallel calculation and reporting of TREA, S-TREA and S-TREA output floor there should be a relevant backstop in place and sufficient information on EAD and RWA available according to different approaches (C 02.00 and C 10.00) so that C 43.00 would lead to double reporting.
The more (sub)exposure classes the more effort/costs are generated for those classes. The more reporting positions the more effort is given. You need to identify the fine granularity to identify those classes and let the countries deliver it to your data pool having those specifications transported and processed in your reporting system. Moreover, the complexity also triggers costs in terms of data quality steps and validation rules.
A more granular reporting on obligor or loan level would massively increase the cost for the reporting. As can be seen from the discussion on the introduction of granular reporting in the context of IReF, many technical detailed questions have to be clarified beforehand. A parallel expansion of COREP reporting at the contract level would result in parallel efforts that cannot be covered by the institutions. The move to a more granular reporting of supervisory data should be integrated into the IReF initiative in a second step after this approach proved to be feasible in practice.
Do you identify any issues regarding the introduction of this template? Would it be more useful to report the information in C 08.01 to directly compare between capital requirements determined by the IRB approach and the SA?
Is the purpose to show S-TREA in C02 in IRB exposure classes structure whereas in C10 the STA exposure class structure is given – is this assumption correct?
Against the background of still not completely harmonized asset classes for the standardized approach (Art. 112 CRR) and the internal ratings-based approach (Art. 147 CRR) we consider the integration of detailed information regarding the calculation of the output floor as very complex. We are concerned that the integration in the C 08.01 - template would inflate the template C 08.01 or make an additional template necessary to report the asset classes only existing in the standardized approach.
Therefore, we appreciate the idea of a separate template to trace the calculation of the output floor. This seems to be more transparent than the integration of the information in the C 08.01 – templates.
However, we understand rows 0030 to 0240 of the template C 10.00 as a breakdown by SA asset classes. The integration of IRB asset classes as “of which”-rows does not seem consistent to us and leads to an increased complexity, especially in cases of deviating asset classes in SA and IRB. Therefore, we would appreciate a breakdown exclusively by SA asset classes.
We prefer introducing a separate template C 10.00 instead of integrating this information in the existing IRB-templates. According to our current understanding, the relevant information is generally available, as it is required for the calculation of risk-weighted assets anyway.
Is the design for the reporting of transitional provisions for the output floor clear enough? If you identify any issues, please specify the related templates and instructions.
The details on output-floor including the application of the transitional provisions related to the output-floor delivers more complexity and will create huge efforts for the implementation in the internal systems and statutory reports. Furthermore, it is unclear how to proceed in C34.02 in case you are already applying standard approach what to fill in in output floor columns – the same values or different values due to transitional provisions which could bring additional reporting burdens because calculation is differing to the regular standard approach.
Question 7 – group solvency template C06.02: Do you identify any issues with the new column 0075 introduced in the group solvency template C06.02 to report the floor adjustment of group entities subject to own funds requirements?
Yes, information transfer about possible local floor adjustments is additional effort.
Do you identify any issues related to the introduction of this new subset? Is this proposal clear enough? If you identify any issues, please suggest how to clarify the reporting.
Missing exposure information in z-axis (dimension) is not helping for consultation phase. Scope of application in all templates would be useful for implementing CRR III regarding sub-exposure classes. Any further delay in this respect is a risk for a timely implementation of Basel IV.
Question 10: Do you have any comment on the other changes included in the C 07.00 template? Other changes include a separate exposure class for “Corporates – Specialised lending, an “of which” row for exposures to central banks, revised memorandum item rows to align with the breakdown for exposures secured by immovable property, a new column “other” for transitional CCFs for UCC, and a last column to report the impact of transitional provisions on CCFs for UCC.
The transitional provisions in C07 column 241 are creating more complexity without any burden relief effects.
As equity exposures form a single exposure class, in Annex I in template C 07.00, row 0015 it should read “exposure class” instead of “exposure classes”.
Regarding the reporting of the Memorandum Items (rows 0290 – 0340) we suggest sorting the rows by topic and do not mix up “Exposures secured by real estate” with “Exposures in default”.
The breakdown within the exposure class “Exposures secured by immovable property” is very detailed. It would reduce reporting costs and burden to scale down this breakdown and only keep a division by IPRE, Commercial Real Estate, Residential Real Estate, Other and ADC.
Question 11: CIUs under the SA approach – Please also refer to question 16 on the reporting of CIU positions and underlying exposures under the IRB: Do institutions have information readily at their disposal on underlying exposures of CIUs in order to be reported as it is proposed to be done in C 08.01? Would this add substantial reporting costs?
It is not clear if rows 190-210 are memorandum items or not. Naming of these rows do differ in templates compared to consultation paper.
Example in CP is also not clear since example 3 (150 Euro CIU exposure class) is the same as example 7 (150 Euro as FBA to central banks) – so you don’t see what 150 EUR is meant in the displaying of the templates. Furthermore a FBA to central banks is also not understandable – why fallback when you have look-through to central banks? Is the assumption correct that header of example 7 is incorrect?
Example 6 is named as Exposure to Corporates – Others (EUR 50) but in C07 it is displayed as CIU exposure and look-through-approach. It is not called as a CIU anywhere in example 6.
Furthermore, displaying of exposure class (e.g. Corporates Other) is not easy to understand since 0 EUR is total exposure whereas in Memorandum row 50 EUR is shown. This is consequence from this concept but hard to comprehend and not self-explainable.
Which option would be preferable taking into account the ready available data and reporting costs? Which one would be more advantageous for data analysis?
Both Options do bring extra burden to reporting and are not preferred by us. If we would have to choose one of them we would go for Option 1.
Question 13 – IRB retail: Is the breakdown of exposure class ‘Retail’ clear and unambiguous? Would an “of which” approach analogous to option 2 described in question 12 but referring to “Secured by immovable property” instead of “Large Corporates” be advantageous for data analysis and preferable taking into account the ready available data and reporting costs?
Given the new structure of C0801 and C0902 the breakdown of Retail is not clear but rather confusing. C0902 has eleven Retail-rows which do provide a breakdown and a further sub-breakdown which seems to provide all details of a retail-sub-breakdowns but then it also requests ‘of-which’ rows in some cases (e.g.: SME of which: secured by a commercial real estate).
Whereas in C0801 the ‘of-which’ rows do concentrate on secured mortgages we are facing different views for IRB exposure classes in different IRB tables.
Would it be less costly to report the whole breakdown of exposure classes of Art. 147 (2) c) CRR3, i.e. including ‘Corporates-other’ instead of reporting ‘of which’ items for Specialised Lending exposures and purchased receivables?
Yes, if the new ‘of which’ lines (i.e.: specialised lending rows 46-49, purchased receivables row 55 and large corporates row 56) are dismissed and not requested in reporting then we would appreciate this approach.
• Question 15.1: Is it clear how positions of exposure class CIU (Art. 147 (2) e1) CRR3 are to be reflected in the CR-IRB templates (C 08.01 to C 08.07)?
We assume new exposure class CIU is used for reporting in IRB templates and no longer underlying exposure classes but in C0801 there will be both requested: underlyings and CIU exposure classes.
• Question 15.2: Regarding CIU positions whose underlying are securitisations or equity exposures, would it be clearer and easier to report these underlying exposures under the securitisation and equity templates (C 13.01 and C 10.01, respectively)? Inversely, should they be reported under the credit risk templates?
The current approach under credit risk is preferred.
• Question 15.3: If you identify any issues, please suggest how to clarify their treatment in the templates and/or instructions.
Examples in consultation paper are not correctly displayed and described – see comments to question 11.
• Question 15.4: Do institutions have information readily at their disposal on underlying exposures of CIUs in order to be reported as it is proposed to be done in C 08.01? Would this add substantial reporting costs? If so, how are those underlying exposures currently reported?
It does not support the analytical view on template C0801 because you switch between memorandum items (also in C0801) to reporting items (C0902) with the same rows. Moreover, naming of row CIU at the end of C0801 is not the same as in consultation paper ‘memorandum item’ is missing in Annex I) but usually in the case of look-through information is available.
• Question 15.5: Would it add substantial reporting burden for institutions if these exposures would be reported under a separate template where both the CIU positions and the underlying exposures would be reported under the corresponding exposure class? Would this approach be clearer?
Introducing an additional new template would bring implementation costs depending on the details requested in this template it could bring substantial efforts. In the long run reducing reporting needs of equity and CIU exposure classes to only one template and removing it from all other credit risk templates (STA and IRB) it could bring more clarity and reduced reporting burdens.
In template C 08.01 a breakdown on mortgages is added for covering supervisory information needs on residential and commercial real estate as well as IPRE and ADC exposures. In this context, a breakdown for non-IPRE exposures into “secured” and “unsecured” (risk weighted as not secured by immovable property) is introduced referring to Articles 125 (1) respectively 126 (1) CRR3 in order to further align reporting for SA and IRB exposures. Do institutions – in particular the ones applying own LGD estimates – have information readily at their disposal for providing this further split into “secured” and “unsecured”. Would this add substantial reporting costs?
- In our opinion the integration of the suggested very detailed breakdown of Exposures secured by real estate into the IRB-template C 08.01 (CR IRB 1) would cause substantial reporting costs due to the following reasons:
- The content of template C 08.01 is only based on IRB calculation methods which differ substantially from the standardized approach. The breakdown of the new rows refers exclusively to the rules of the standardized approach. Furthermore, there are different methods for the allocation of collateral under the IRB and SA and different requirements for the eligibility of collateral. It would therefore be an immense burden to reconcile these two calculation methods.
- If this breakdown is mandatory for supervisory purposes, we suggest integrating the breakdown in template C 10.00 which was specifically introduced to provide a view of the IRB portfolio under the calculations of the standardized approach. In this template it would be much less costly to show this breakdown as we are within a template which should be fully populated with figures calculated according to the rules for the standardized approach.
Question 17 – revised instructions for template C 15.00: The instructions have been updated to align with the legal references with the new articles introduced in Regulation (EU) No 575/2013 for exposures secured by immovable property and the revised [Article 430a] on specific reporting obligations. The instructions have been clarified on certain aspects. The template has been amended to remove the two columns referring to the mortgage lending value. Are the revised instructions clear enough? If you identify any issues, please suggest how to clarify the reporting.
The amendment of Art. 430a is bringing different limits up to the losses which have to be reported in the columns 010 and 030. Bringing examples would bring additional support in better understanding the requested reporting positions.
We suggest amending Art. 13 of regulation (EU) 2021/451 so that the IPLOSSES report shall be submitted only at highest level of consolidation within a member state. For calculating the overall loss rates for a single member state which are published on NCA level, the report at highest consolidation level should be sufficient.
Could you please clarify if estimated losses should be reported furthermore? Under para. 13 a) of Annex VII the part regarding estimated losses is deleted, while under para. 12 it is maintained.
Are the reporting template C 25.00 and related instructions clear enough? If you identify any issues, please suggest how to clarify the reporting.
Template and instructions are almost clear enough.
Could you please clarify whether direct exposures to Central Counterparties should be reintegrated as well (direct clearing). According to Art. 382 (3) CRR those direct exposures are exempted but in Annex II only client transactions are mentioned.
a) Did you identify any issues regarding the representation of the (policy) framework for the boundary in templates C 90.05 and C 90.06?
General remarks: We question the rationale behind the request for this information in general, particularly in light of the exceptionally high costs anticipated for implementing such reporting frameworks.
The CRR3 already imposes strict and comprehensive rules for distinguishing between banking and trading book items. These regulations, along with other requirements for classifying instruments, are incorporated into internal guidelines and policies. Supervisors can access these on an ad hoc basis, and they are subject to regular audits. Thus, the criteria for classifying new portfolios as either banking or trading book items are clearly defined. Furthermore, classifications of new portfolios are subject to supervisory scrutiny through the NPP as well as regular audits. Factors such as current risk classes and quantitative measures like market value do not influence the distinction between the trading and banking book items, nor do they affect the monitoring of classification criteria.
Therefore, we suggest a pragmatic and proportionate approach to overseeing the categorization of instruments into the banking or trading books, as already outlined in the feedback to question 7 in the Final Report On The Amending ITS On Specific Reporting Requirements For Market Risk (EBA/ITS/2024/02). The existing monitoring of the classification criteria for banking and trading book items in line with the CRR3 would suffice for supervisory purposes.
b) Are the scope of application of the requirement to report the different templates, the scope of positions/instruments/profits and losses etc. included in the scope of every template, the template itself and the instructions clear? If not, please explain the issues needing clarification, and make a suggestion on how to address them.
The general intention of the template is clear. Given the high level of complexity of required data, partial overlaps, etc. issues will most likely be discovered in the implementation and testing process. Given the high level of attention put on interest rate hedges in the CRR, it is surprising that there is no dedicated column for reporting such hedges in the templates.
Question 21: Do you agree with the changes to the Leverage ratio reporting as implementing the new CRR3 provisions? Do you see any further amendments needed?
In general, we see no legal basis for maintaining templates C 40.00 and C 43.00. These templates were introduced for the reporting of data required for the preparation of the report under Article 511 of the CRR. The EBA has already submitted this report in 2016. As the retention of data reports that are no longer required for supervisory purposes is unduly burdensome for institutions, we advocate the removal of both templates.