Response to consultation on ITS on Supervisory Reporting amendments with regards to FINREP
Go back
There are mixed views on this topic. At some banks, these contributions are part of “other operating expenses”. However, other banks include this item in “administrative expenses”.
We generally believe that it is unnecessary to show this line separately because these contributions should already be known to supervisory authorities. Should separate disclosure of the contributions to resolution funds and deposit guarantee schemes nevertheless be considered imperative, we propose that the contribution be presented not as an of which" position, but in a separate line item in template F 02.00 between lines 355 and 360. In any case, this line should not contain any elements that have not yet been recognised in the income statement and should only be shown as an off-balance sheet liability. In our understanding, such off-balance-sheet obligations should not be a subject of template F 02.00 (P&L).
We also consider it necessary to clarify the term "payment commitments" in this context. It is not clear whether this refers to the maximum 30% share of payment obligations according to the Directive on Deposit Guarantee Schemes (2014/49/EU) or to other commitments and contingent liabilities to the guarantee fund respectively."
A further problem is of a more technical nature: the lack of a one-to-one allocation of collateral and loans. Collateral may secure several loans and even borrowers that are assigned to different exposures.
We also note that the establishment of a validation rule between the initial value, the difference (resulting from negative developments) and the current value of the collateral is not mathematically possible due to the exclusion of positive developments.
The current EBA proposal for a definition of commercial real estate could cause further confusion, because owner-occupied company headquarters, for example, which generate no rental income per se, would have to be classified as residential property according to the EBA criteria. This does not seem appropriate. As stated above, we do not consider a separate definition of commercial real estate to be necessary. However, if such a definition is imperative, it should depend on the actual or intended use rather than on the generation of revenue, etc. We therefore also consider the definition of “income-producing real estate” to be superfluous; moreover, such a criterion is not included in existing IT systems for every type of property/collateral.
In addition, we would like to point out that template F 18.00 is unnecessarily complex. It is laden with detailed maturity bands and “of which” items or sub-items, at times without any clear allocation (see line 130 ff.). It is not clear, for instance, whether line 140 should represent the sum of lines 132 and 135, or which breakdown affects all “non-financial corporations” and which only the “SME”. This template belongs to Module 1 and would therefore also have to be drawn up each quarter by small and non-complex institutions, preceded by a very cost-intensive IT changeover. We therefore generally suggest that the required scope of information and the actual need for such granular information, especially regarding the LTV and CRE, be reconsidered against the background of the principle of proportionality. One possible solution would be to reduce the number of maturity bands, then also analogously for the NPL disclosure templates (EBA/CP/2018/06). We advocate limiting the granular presentation (42 columns) to particularly relevant positions and (for technical reasons) moving this additional data to a new template. Providing account extensions/sub-accounts for all “of which” positions leads to an overburdened IT data set with no recognisable additional benefits, especially for institutions with a low NPL ratio.
In addition, it must be examined in more detail whether financial reporting can, in technical terms, provide the required LTV ratio-related data. So far, this data is not technically linked to the individual transaction data or it is questionable whether it can be evaluated at all (for reporting purposes). The technical effort required to implement such a link is estimated to be very high.
With a view to the complexity of template F 18.00, we refer to our answer to question 3. We also see the following challenges facing implementation:
• Collaterals or its values are not relevant for certain types of loans. This statement especially applies to retail business (e.g. motor vehicle financing). In such cases, Credit risk management does not include any concrete collateral values in its management. At present, neither the IT data sets nor (in part) the loan files record collateral values for these portfolios. This means that reporting these values would not be realistic.
• Due to their development mandate and the fact that development banks do not intend to make a profit, the property values are also at times irrelevant for their loan decision. In these cases, real estate collateral is provided on a subordinated basis and is therefore not taken into account as risk mitigating in either internal or external management (as defined by the CRR). Therefore, property values are not always available, for instance, in the context of social housing development. Reporting would also not be possible for such loans.
• The assignment of one collateral to several loans (with different repayment schedules and residual maturities) and, if applicable, to several borrowers leads to technical challenges, e.g. with regard to the maturity band assignment for the different overdue maturities of secured loans. The problem in this case is that banks would have to work with assumptions, and an IT-based estimation/distribution model would be required for reporting.
We would like to point out that not all banks use loan-to-value ratios in their internal management strategies. Banks often use loan-to-collateral-value ratios instead, since these make better economic sense. Thus, the existing reference to the Recommendation of the European Systemic Risk Board of 31 October 2016 on closing real estate data gaps (ESRB/2016/14) should be properly supplemented, so that the use of the value of the collateral in accordance to Article 229 (1) subparagraph 3 CRR meets also the definition of the “current value of the property”.
Sensitive HR and remuneration data should not be included in FINREP. As EBA itself documented in the consultation paper, this information is already collected under EBA Guideline 2014/08 (remuneration benchmarking exercise) and on a national level in the German Remuneration Regulation for Financial Institutions. It would be redundant and disproportionate to include it in FINREP as well. This would not be in line with current EU efforts to establish an efficient reporting framework.
The experience made by some institutions with so-called NPE stocktake templates" shows that collecting the new data required in templates F 23, F 24, F 26, F 47 is very time-consuming and mostly carried out manually. In our view, if the NPE stocktake templates were integrated into Module 2 of the FINREP report while, at the same time, partially reducing the scope of information, this would help to optimise the flow of information to the supervisory authority."
Most of the information is already reported annually as part of monthly national balance sheet statistics (BISTA) of Deutsche Bundesbank and is also included in the consolidated annual reports for the purposes of consolidation under commercial law. Expanding the required statistical data to the supervisory scope of consolidation can only be ensured at present by way of manual reconciliations. This means that there is a need to carefully evaluate whether these redundancies serve a useful purpose.
Nevertheless, the formulations on inflows and outflows in Annex V appear to require explanation in two respects.
On the one hand, it is not fully clear whether the changes refer to the allocation of loans to/out of the NPE portfolio (inflows and outflows of individual transactions) or inflows and outflows to loans, which are already part of the NPE portfolio (repayments and interest payments) or to both types of change. So do in-/outflows relate to the composition of the portfolio or to the cash movements or both? We suggest that the definition be sharpened.
On the other hand, Annex V (e.g. section 239iii) requires that: “outflow shall reflect exposures that cease to be non-performing at any point in time during the reporting period”. From our point of view, it is unrealistic for change data to be subject to complete daily monitoring with documented daily retrieval. We understand the term at any point in time" to mean that data, such as access to the NPE portfolio, deferral measure or recovery date, is recorded once a quarter and used for quarterly reporting purposes. In the event that the EBA sees this differently, an explanatory statement with justification will have to be included."
Question 1 (on template F 02.00) Do respondents agree with the proposed FINREP representation of “Contributions to resolution funds and deposit guarantee schemes” as part of “other operating expenses”? If not, which representation would you suggest?
We also refer to our general remarks contained in our comments which are attached as a file upload.There are mixed views on this topic. At some banks, these contributions are part of “other operating expenses”. However, other banks include this item in “administrative expenses”.
We generally believe that it is unnecessary to show this line separately because these contributions should already be known to supervisory authorities. Should separate disclosure of the contributions to resolution funds and deposit guarantee schemes nevertheless be considered imperative, we propose that the contribution be presented not as an of which" position, but in a separate line item in template F 02.00 between lines 355 and 360. In any case, this line should not contain any elements that have not yet been recognised in the income statement and should only be shown as an off-balance sheet liability. In our understanding, such off-balance-sheet obligations should not be a subject of template F 02.00 (P&L).
We also consider it necessary to clarify the term "payment commitments" in this context. It is not clear whether this refers to the maximum 30% share of payment obligations according to the Directive on Deposit Guarantee Schemes (2014/49/EU) or to other commitments and contingent liabilities to the guarantee fund respectively."
Question 2 (on templates F 13.2.1 and F 13.3.1) ‘Accumulated negative changes’ tries to capture, in a generic manner and independently from the type of collateral, the accumulated decrease in the value of a collateral item obtained, where the value of the collateral decreased since it was obtained by the institution (i.e. cases of increases in the collateral value are excluded). In this regard, 'accumulated negative changes' captures the net impact of changes in market prices, impairments and reversals of impairments, write-offs, depreciation and appreciation, change of accounting policies and similar on the carrying amount of an individual collateral item, where the difference between the value at initial recognition and carrying amount, as influenced by this factors, is negative. The comparison between value at initial recognition and the carrying amount at the reference date shall be done for each collateral item separately. In order to obtain the aggregate figure that is reported in the template, only the negative differences shall be aggregated, while cases of positive differences are to be neglected. Is this definition clear?
The definition is basically clear in itself. However, the question arises as to whether the mere value development of the collateral or also the adjustments of the collateral value to the value of the outstanding loan after repayments are be taken into account (e.g. at the beginning, loan 100 and collateral 100, then repayment of the loan 20 then adjustment of the value of the collateral to 80 due to limitation to the value of the outstanding loan, although the object value may still be 100 or 90).A further problem is of a more technical nature: the lack of a one-to-one allocation of collateral and loans. Collateral may secure several loans and even borrowers that are assigned to different exposures.
We also note that the establishment of a validation rule between the initial value, the difference (resulting from negative developments) and the current value of the collateral is not mathematically possible due to the exclusion of positive developments.
Question 3 (on several templates, see F 18.00) The ESRB recommendation defines CRE as follows: ‘Commercial real estate’ (CRE) means any income-producing real estate, either existing or under development, and excludes (a) social housing; (b) property owned by end-users; (c) buy-to-let housing. If a property has a mixed CRE and RRE use, it should be considered as different properties (based for example on the surface areas dedicated to each use) whenever it is feasible to make such breakdown; otherwise, the property ca be classified according to its dominant use. ‘Commercial reals estate (CRE) loan’ means a loan aimed at acquiring a CRE property (or set of CRE properties) or secured by a CRE property (or set of CRE properties). ‘Income-producing real estate’ means all immovable properties with income generated by their rents or profits from their sale. Is this definition clear? To which extent is compatible with, for example, your internal classification? Which challenges with regard to the practical application of this definition do you envisage?
In our view, the definition is ambiguous. The CRR contains a definition for residential property (point 75 of Article 4 (1) CRR). This definition implies a negative delimitation for commercial properties. From a banking perspective, this negative delimitation is sufficient and clear. It is even perceived as being sufficient for the very granular requirements of AnaCredit (analytical credit datasets). It would be useful to adopt the same approach for FINREP purposes. The purpose of financing, which was recently introduced for AnaCredit reporting, is now available in IT data sets for individual transactions and could therefore provide a definition basis for both disclosures without any unnecessary effort. In principle, we advocate harmonised definitions within the European reporting system which shall be based as far as possible on existing terms.The current EBA proposal for a definition of commercial real estate could cause further confusion, because owner-occupied company headquarters, for example, which generate no rental income per se, would have to be classified as residential property according to the EBA criteria. This does not seem appropriate. As stated above, we do not consider a separate definition of commercial real estate to be necessary. However, if such a definition is imperative, it should depend on the actual or intended use rather than on the generation of revenue, etc. We therefore also consider the definition of “income-producing real estate” to be superfluous; moreover, such a criterion is not included in existing IT systems for every type of property/collateral.
In addition, we would like to point out that template F 18.00 is unnecessarily complex. It is laden with detailed maturity bands and “of which” items or sub-items, at times without any clear allocation (see line 130 ff.). It is not clear, for instance, whether line 140 should represent the sum of lines 132 and 135, or which breakdown affects all “non-financial corporations” and which only the “SME”. This template belongs to Module 1 and would therefore also have to be drawn up each quarter by small and non-complex institutions, preceded by a very cost-intensive IT changeover. We therefore generally suggest that the required scope of information and the actual need for such granular information, especially regarding the LTV and CRE, be reconsidered against the background of the principle of proportionality. One possible solution would be to reduce the number of maturity bands, then also analogously for the NPL disclosure templates (EBA/CP/2018/06). We advocate limiting the granular presentation (42 columns) to particularly relevant positions and (for technical reasons) moving this additional data to a new template. Providing account extensions/sub-accounts for all “of which” positions leads to an overburdened IT data set with no recognisable additional benefits, especially for institutions with a low NPL ratio.
Question 4 (on several templates, see F 18.00) The ESRB recommendation defines the current loan-to-value ratio as follows: ‘Current loan-to-value ratio’ (LTV-C) means the sum of all loans or loan tranches secured by the borrower on a property at the reporting date relative to the current value of the property; ‘Current value of the property’ means the value of the property as assessed by an independent external or internal appraiser; if such assessment is not available, the current value of the property can be estimated using a real estate value index sufficiently granular with respect to geographical location and type of property; if such real estate value index is also not available, a real estate price index sufficiently granular with respect to geographical location and type of property can be used after application of a suitably chosen mark-down to account for the depreciation of the property; Is this definition clear? Which challenges with regard to the practical application of this definition do you envisage?
The definition is basically clear, however, it raises questions, some of which have already been raised in the previous answer. For example, whether loans, which are not used to finance CRE property but are collateralised by a CRE property, should be included in the calculation. It is also unclear how differences in the LTV ratio definitions among different EU states are to be dealt with.In addition, it must be examined in more detail whether financial reporting can, in technical terms, provide the required LTV ratio-related data. So far, this data is not technically linked to the individual transaction data or it is questionable whether it can be evaluated at all (for reporting purposes). The technical effort required to implement such a link is estimated to be very high.
With a view to the complexity of template F 18.00, we refer to our answer to question 3. We also see the following challenges facing implementation:
• Collaterals or its values are not relevant for certain types of loans. This statement especially applies to retail business (e.g. motor vehicle financing). In such cases, Credit risk management does not include any concrete collateral values in its management. At present, neither the IT data sets nor (in part) the loan files record collateral values for these portfolios. This means that reporting these values would not be realistic.
• Due to their development mandate and the fact that development banks do not intend to make a profit, the property values are also at times irrelevant for their loan decision. In these cases, real estate collateral is provided on a subordinated basis and is therefore not taken into account as risk mitigating in either internal or external management (as defined by the CRR). Therefore, property values are not always available, for instance, in the context of social housing development. Reporting would also not be possible for such loans.
• The assignment of one collateral to several loans (with different repayment schedules and residual maturities) and, if applicable, to several borrowers leads to technical challenges, e.g. with regard to the maturity band assignment for the different overdue maturities of secured loans. The problem in this case is that banks would have to work with assumptions, and an IT-based estimation/distribution model would be required for reporting.
We would like to point out that not all banks use loan-to-value ratios in their internal management strategies. Banks often use loan-to-collateral-value ratios instead, since these make better economic sense. Thus, the existing reference to the Recommendation of the European Systemic Risk Board of 31 October 2016 on closing real estate data gaps (ESRB/2016/14) should be properly supplemented, so that the use of the value of the collateral in accordance to Article 229 (1) subparagraph 3 CRR meets also the definition of the “current value of the property”.
Question 5 (on F 40.01, F 40.02) The information included in the two group structure templates is currently collected on an annual basis. Without prejudice to notification obligations under national laws, a more frequent collection (quarterly) would improve the timely reflection and awareness of changes to institutions’ group structures. Which benefits and challenges with regard to the compilation and reporting of this information on a more frequent basis do you envisage?
Although the group of consolidated companies is determined every quarter and adjusted if necessary, these adjustments are, however, merely restricted to the number of additions/disposals of companies. This means that regular data on group entities during the year is not yet available. As a rule, only information from the final annual accounts is collected. The quarterly reports would therefore continue to contain the data from the end of the year and would not represent any significant added value compared to the yearly report. Since there is often a delay in the information which, moreover, is then entered manually, a higher reporting frequency poses a greater challenge which is disproportionate to its benefits and, in the worst case, would lead to late reporting. We therefore do not see that more frequent reporting will deliver any additional benefit to supervisors.Question 6 (on F 44.04, F 48.00) Some of the items included in templates F 44.04 and F 48.00 are also collected for the purposes of benchmarking in accordance with EBA’s Guideline on the remuneration benchmarking exercise (EBA/GL/2014/08). The items requested in FINREP are of high-level nature and full alignment has been sought to keep the reporting burden limited. What is your view on the inclusion of this information in FINREP? Do you see any inconsistencies between this data and the data collected in accordance with the GL on remuneration benchmarking exercise that impact the reporting burden?
We do not agree with the request for additional information that goes far beyond the scope and objective of FINREP to date. FINREP should continue to focus solely on financial information. As mentioned in our general remarks, we see no reason to supplement or overload FINREP with data on other matters.Sensitive HR and remuneration data should not be included in FINREP. As EBA itself documented in the consultation paper, this information is already collected under EBA Guideline 2014/08 (remuneration benchmarking exercise) and on a national level in the German Remuneration Regulation for Financial Institutions. It would be redundant and disproportionate to include it in FINREP as well. This would not be in line with current EU efforts to establish an efficient reporting framework.
Question 7 (on several templates, see F 23.01 – F 23.03) The following templates (templates F 23, F 24, F 26, F 47) request information on loans and advances subject to definition of non-performing and forborne exposures (with the exception of loans and advances classified as held for sale), in contrast to F 18 and F 19 that cover ‘exposures’ in a broader sense, e.g. also debt securities. The rationale behind applying these additional templates to loans and advances only is that the majority of exposures in credit institutions’ balance sheets that turned non-performing are loans and advances. To have risk based focus to monitor evolution of asset quality and to balance reporting burden the templates focus on loans and advances only. Are the definitions and instructions on the definition of the scope clear?
Yes, the definition is clear.The experience made by some institutions with so-called NPE stocktake templates" shows that collecting the new data required in templates F 23, F 24, F 26, F 47 is very time-consuming and mostly carried out manually. In our view, if the NPE stocktake templates were integrated into Module 2 of the FINREP report while, at the same time, partially reducing the scope of information, this would help to optimise the flow of information to the supervisory authority."
Question 8 (on F 48.00) The information collected in this template is different in nature from the information collected in the remainder of FINREP, i.e. it is mostly of non-financial nature. It is valuable as contextual information to understand core elements of fixed costs of institutions. Similar information, where applicable potentially with regard to a different scope of consolidated entities, is collected, for example, by monetary authorities. Which benefits and challenges with regard to the compilation and reporting of this information do you envisage?
The statistical information requested here far exceeds the scope of FINREP. The proposed extension of the scope would make it increasingly difficult for banks to retrieve the requested data without delivering any matching benefit. While we can imagine that reporting this information in FINREP might cause the EBA to drop certain data requests for benchmarking purposes, we nevertheless have serious reservations about including statistical data in FINREP because of the additional strain it would place on banks. We are also very critical of any expansion of the requested data in view of the already tight delivery deadlines.Most of the information is already reported annually as part of monthly national balance sheet statistics (BISTA) of Deutsche Bundesbank and is also included in the consolidated annual reports for the purposes of consolidation under commercial law. Expanding the required statistical data to the supervisory scope of consolidation can only be ensured at present by way of manual reconciliations. This means that there is a need to carefully evaluate whether these redundancies serve a useful purpose.
Question 9 Are the scope of application of the revised reporting requirements as set out in the draft ITS above, the reporting templates as set out in Annexes 1 and 2 to this Consultation Paper (Annexes III and IV to the ITS) and the related instructions in Annex 3 to this Consultation Paper (Annex V to the ITS) sufficiently clear? In case of uncertainties with regard to scope of entities subject to the reporting obligation or the information that shall be reported, please provide clear references to the applicable provision respectively the relevant columns/rows of a given template as well as specific examples that highlight the need for further clarifications.
The scope and the template-related instructions are generally clear.Nevertheless, the formulations on inflows and outflows in Annex V appear to require explanation in two respects.
On the one hand, it is not fully clear whether the changes refer to the allocation of loans to/out of the NPE portfolio (inflows and outflows of individual transactions) or inflows and outflows to loans, which are already part of the NPE portfolio (repayments and interest payments) or to both types of change. So do in-/outflows relate to the composition of the portfolio or to the cash movements or both? We suggest that the definition be sharpened.
On the other hand, Annex V (e.g. section 239iii) requires that: “outflow shall reflect exposures that cease to be non-performing at any point in time during the reporting period”. From our point of view, it is unrealistic for change data to be subject to complete daily monitoring with documented daily retrieval. We understand the term at any point in time" to mean that data, such as access to the NPE portfolio, deferral measure or recovery date, is recorded once a quarter and used for quarterly reporting purposes. In the event that the EBA sees this differently, an explanatory statement with justification will have to be included."