Response to consultation on draft Guidelines on resubmission of historical data
Go back
However, we believe that the EBA's objective to develop a "one size fits all" approach and to reduce ongoing reporting costs would not fully address the main concern.
The proposed tolerance limits would lead to a considerable effort according to the large absolute values to be reported and the absence of materiality will result in excessive levels of resubmissions which is far from efficient for both banks and supervisors. In the light of this, a materiality threshold should be defined as relying on the EBA's "filling rules" would mean using thresholds that are too low and would be operationally costly in terms of data resubmission volumes and documentation of these resubmissions, notably for large institutions.
The EBA proposes to limit the time periods to be considered for the submission of correction by defining 1 year resubmission. The proposed resubmission policy has a positive side by limiting the time periods to be considered for the submission of corrections. However, resubmitting such number of periods by going back one year prior to the current date for all reporting frequencies would be burdensome for institutions and we consider the limit could well be shorter.
Thus, regarding the general approach to the resubmission of historical data, the industry proposes to have a limited timeframe with a uniform methodology for all cases (annual, semi-annual, quarterly and monthly reporting) and to resubmit only until the last year-end period. It would aim to have a simplest solution and a harmonized approach for all frequencies. Indeed, having several rules depending on the frequencies and the reference dates would add more complexity.
In addition, it would add no supervisory value to cover earlier periods as the data the most important from a supervision standpoint are the previous year-end data. In this sense, covering only the previous year-end data instead of resubmitting one year back would be the right balance between the supervisory needs and the capacity constraints for institutions in terms of proportion of costs and time spent on resubmissions.
There are also various relevant aspects in the proposed Guidelines that need to be further clarified:
The impact by changes to the taxonomy and how these would affect the resubmission of historical data. It remains unclear how the EBA will handle the changes in DPM and in Validation Rules to previously submitted returns. This will imply difficulties in reconstituting historical data which would be not available if they no longer apply in the current taxonomy, or they were not part of the previous taxonomy.
Regarding this specific topic, the opinion of the industry is that dealing with two different taxonomies would be unmanageable as it would require to manage two sets of formats and two specific data collections. Current reporting tools and IT solutions are not able to restore such historical data and it would lead to manual processes.
In this sense, we strongly suggest to limit the historical resubmissions to the current taxonomy alive. It would be a derogation from the general principle the industry proposes to apply based on resubmissions until the last annual closing.
Example: The errors or corrections affect quarterly data for Q3 2022 that has been submitted to the competent or resolution authorities and which is considered as the current data in the meaning of the Guidelines. On Q4 2021, the taxonomy X.0 applied and on Q1 2022 the taxonomy X.2 started to apply.
In the theoretical principle proposed, the financial institutions should in theory resubmit data for Q3 2022 and the following reference dates: Q2 2022, Q1 2022, Q4 2021. But as the taxonomy changed structurally, the financial institutions would only resubmit until Q1 2022 and not consider Q4 2021 in the batch of resubmissions as Q4 2021 dealt with a former taxonomy.
It is also not addressed in the Guidelines which it may be the time framework for updating the historical data. A sufficient and reasonable time should be granted to enable institutions to answer to supervisors’ questions, retrieve the data, reprocess it and filling the templates again.
The draft Guidelines envisage specific role and processes for the assessment of resubmitted data by the authorities (supervisor, resolution authorities and the EBA). However, the interactions between the institutions, the national supervisors, and the ECB in the context of the resubmission process are not developed and specified. The interactions between all the stakeholders during the resubmission process need to be further clarified. For example, in terms of timing, do they access to the data at the same time? Are the questions sent to the institutions shared between the authorities? Is there any coordination in the resubmission/question process between NCAs, JSTs and the ECB?
While the proposed guidelines aim to introduce a uniform approach regarding resubmissions aiming for greater consistency, certainty and practicality to the resubmissions of data, we consider they fail to address the core concern that was at the origins of these EBA Draft Guidelines on historical data resubmissions themselves i.e., to reduce the costs related to data resubmissions. Under the proposed approach, it is not only that they could lead to more resubmissions by financial institutions, but these would, additionally, need to allocate resources to comply with the extensive requirements on the resubmissions of historical data.
As a general comment, the Guidelines emphasize the idea that a data point by data point approach should prevail over a risk-based approach, which from our view, would be more relevant in the specific context of resubmissions. This would be aligned with the ECB strategy, as this risk-based approach has been developed by the ECB as part of its work on significant resubmissions, along with the determination of targeted key risk indicators. In this sense, we suggest that it should be also followed through the prism of the general approach to resubmission of historical data.
Based on all the above considerations, it is our view that the proposed framework is rather limited on reducing the current complexity from a cost-benefit perspective. It requires to be further developed and designed along the comments provided and taking greater account of the cost-benefit considerations, toned to find the right balance between what would be useful for the supervisor needs, and what would be feasible for institutions, by among others, avoiding time consuming processes which require to spend huge time involvement on corrections of minor errors from last closings and on remediation plans.
The approach requires extending the correction of the historical data for one reference date to the other reference dates. We question the relevance of the request given the cumbersome nature of the IT processing to extend the corrections and the expected benefits.
In addition, there are some specific cases where the regulator requests to submit data with wrong figures due to erroneous templates (and which are consequently not aligned with the consistency checks developed through the EBA validation rules). As they derive from requests from the regulator to override the issues due to wrong templates, these specific cases should not be considered as errors on historical data for the next reporting periods in the context of these guidelines, and resubmissions of such wrong historical data should not be requested accordingly. This specific example also underlines the importance of providing some flexibility in the proposed approach to:
- take into consideration specific situations, and
- grant waivers when such cases are met.
Moreover, there are also some specific cases where the regulator requests to resubmit data of several reporting periods due to non-official EGDQs. As institutions are not pre-informed and aware of these additional consistency checks, these non-official EGDQs should also not be considered in the context of these guidelines. These guidelines should apply in the same way for the institutions and the regulator and the specific resubmission requests should be limited.
We would suggest also that a change in an interpretation or assumption would not constitute an error.
Therefore, the scenarios when resubmissions of historical data may not be required should be extended to regulatory and taxonomy changes and should not be limited to the cases already presented in the EBA consultation.
It should be noted that banks will have to implement at the same time the changes brought about by the implementation of Basel 4 and the new guidelines for the resubmission of historical data, which would apply from 31 December 2023. However, at present, given the stakes of regulatory projects, IT departments are prioritising developments linked to the implementation of Basel 4-induced changes. Thus, in order to ease the implementation of all the on-going regulatory projects and notably the EBA's guidelines on the resubmission of historical data, we would suggest a transition period during the implementation of Basel 4 during which data resubmissions will be limited to the current reference period.
In this context, and since both EBA and ECB have stressed in various opportunities an open dialogue between both institutions about their respective workstreams on the resubmission field, we strongly advise the EBA to consider the ECB work undertaken with key risk indicators (KRIs) and applicable thresholds as part of the ECB management report on data governance and data quality, as well as the ECB significant resubmissions policy with the ongoing pilot exercise ending by mid next year.
We propose accordingly to put in place an alternative methodology that would rely on the initial ECB principles on significant resubmissions using the 26KRIs deployed by the ECB and depending on the level of risks according to a red-amber-green classification. The ECB methodology is then adapted with lower thresholds to determine whether a resubmission of a specific reporting or template or a full reporting is required. For other reporting not covered by the 26 KRIs and that may be potentially affected by errors, we propose to complete the proposed methodology by thresholds based on a percentage of variation of the cell impacted by the error and / or a percentage of Prudential Own Funds impacted (for balance-sheet items and off balance-sheet items) and a percentage of Net Operating Income impacted (for profit and loss items).
We will be in position to present this industry proposal in the course of September for which we will appreciate to start coordinating with the EBA for a meeting to present and explain our proposal.
The industry proposes to have a uniform methodology for all cases (annual, semi-annual, quarterly, and monthly reporting) and to resubmit only until the last year-end period.
This general approach to resubmission of historical data would apply to all types of institutions (large, medium-sized and SNCI). It would aim to have a simplest solution and a harmonized approach for all frequencies and all types of institutions regardless of their size.
1. What are your general views on the proposed approach to the resubmission of historical data?
We appreciate the EBA’s intention to propose a common approach to the resubmission of historical data to resolve one of the main issues resulting in high costs for financial institutions when currently resubmitting historical data.However, we believe that the EBA's objective to develop a "one size fits all" approach and to reduce ongoing reporting costs would not fully address the main concern.
The proposed tolerance limits would lead to a considerable effort according to the large absolute values to be reported and the absence of materiality will result in excessive levels of resubmissions which is far from efficient for both banks and supervisors. In the light of this, a materiality threshold should be defined as relying on the EBA's "filling rules" would mean using thresholds that are too low and would be operationally costly in terms of data resubmission volumes and documentation of these resubmissions, notably for large institutions.
The EBA proposes to limit the time periods to be considered for the submission of correction by defining 1 year resubmission. The proposed resubmission policy has a positive side by limiting the time periods to be considered for the submission of corrections. However, resubmitting such number of periods by going back one year prior to the current date for all reporting frequencies would be burdensome for institutions and we consider the limit could well be shorter.
Thus, regarding the general approach to the resubmission of historical data, the industry proposes to have a limited timeframe with a uniform methodology for all cases (annual, semi-annual, quarterly and monthly reporting) and to resubmit only until the last year-end period. It would aim to have a simplest solution and a harmonized approach for all frequencies. Indeed, having several rules depending on the frequencies and the reference dates would add more complexity.
In addition, it would add no supervisory value to cover earlier periods as the data the most important from a supervision standpoint are the previous year-end data. In this sense, covering only the previous year-end data instead of resubmitting one year back would be the right balance between the supervisory needs and the capacity constraints for institutions in terms of proportion of costs and time spent on resubmissions.
There are also various relevant aspects in the proposed Guidelines that need to be further clarified:
The impact by changes to the taxonomy and how these would affect the resubmission of historical data. It remains unclear how the EBA will handle the changes in DPM and in Validation Rules to previously submitted returns. This will imply difficulties in reconstituting historical data which would be not available if they no longer apply in the current taxonomy, or they were not part of the previous taxonomy.
Regarding this specific topic, the opinion of the industry is that dealing with two different taxonomies would be unmanageable as it would require to manage two sets of formats and two specific data collections. Current reporting tools and IT solutions are not able to restore such historical data and it would lead to manual processes.
In this sense, we strongly suggest to limit the historical resubmissions to the current taxonomy alive. It would be a derogation from the general principle the industry proposes to apply based on resubmissions until the last annual closing.
Example: The errors or corrections affect quarterly data for Q3 2022 that has been submitted to the competent or resolution authorities and which is considered as the current data in the meaning of the Guidelines. On Q4 2021, the taxonomy X.0 applied and on Q1 2022 the taxonomy X.2 started to apply.
In the theoretical principle proposed, the financial institutions should in theory resubmit data for Q3 2022 and the following reference dates: Q2 2022, Q1 2022, Q4 2021. But as the taxonomy changed structurally, the financial institutions would only resubmit until Q1 2022 and not consider Q4 2021 in the batch of resubmissions as Q4 2021 dealt with a former taxonomy.
It is also not addressed in the Guidelines which it may be the time framework for updating the historical data. A sufficient and reasonable time should be granted to enable institutions to answer to supervisors’ questions, retrieve the data, reprocess it and filling the templates again.
The draft Guidelines envisage specific role and processes for the assessment of resubmitted data by the authorities (supervisor, resolution authorities and the EBA). However, the interactions between the institutions, the national supervisors, and the ECB in the context of the resubmission process are not developed and specified. The interactions between all the stakeholders during the resubmission process need to be further clarified. For example, in terms of timing, do they access to the data at the same time? Are the questions sent to the institutions shared between the authorities? Is there any coordination in the resubmission/question process between NCAs, JSTs and the ECB?
While the proposed guidelines aim to introduce a uniform approach regarding resubmissions aiming for greater consistency, certainty and practicality to the resubmissions of data, we consider they fail to address the core concern that was at the origins of these EBA Draft Guidelines on historical data resubmissions themselves i.e., to reduce the costs related to data resubmissions. Under the proposed approach, it is not only that they could lead to more resubmissions by financial institutions, but these would, additionally, need to allocate resources to comply with the extensive requirements on the resubmissions of historical data.
As a general comment, the Guidelines emphasize the idea that a data point by data point approach should prevail over a risk-based approach, which from our view, would be more relevant in the specific context of resubmissions. This would be aligned with the ECB strategy, as this risk-based approach has been developed by the ECB as part of its work on significant resubmissions, along with the determination of targeted key risk indicators. In this sense, we suggest that it should be also followed through the prism of the general approach to resubmission of historical data.
Based on all the above considerations, it is our view that the proposed framework is rather limited on reducing the current complexity from a cost-benefit perspective. It requires to be further developed and designed along the comments provided and taking greater account of the cost-benefit considerations, toned to find the right balance between what would be useful for the supervisor needs, and what would be feasible for institutions, by among others, avoiding time consuming processes which require to spend huge time involvement on corrections of minor errors from last closings and on remediation plans.
2. How do you see the proposed approach in relation to your existing resubmission policies set out in your institutions, agreed with internal audit and control functions?
Whilst the proposed approach is setting a common resubmission framework in terms of thresholds and reference dates, it should provide for the possibility for competent authorities or JSTs to continue to provide guidance to institutions on the corrections of errors and the data to be resubmitted according to the specific situations encountered (i.e. evolution of the taxonomy making it difficult to resubmit historical data, non-material errors, )The approach requires extending the correction of the historical data for one reference date to the other reference dates. We question the relevance of the request given the cumbersome nature of the IT processing to extend the corrections and the expected benefits.
3. How do you see the proposed approach in relation to actual practices for the resubmission of data also considering the legal requirements set out in existing legislation (e.g. Article 3(5) of Commission Implementing Regulation (EU) 2021/451)?
We question the possibility given to competent authorities (on paragraph 29 of the Draft guidelines) to require financial institutions to resubmit historical data for additional reference dates compared to those defined in the guidelines (), whereas the proposed approach aims to harmonise the resubmission of data.In addition, there are some specific cases where the regulator requests to submit data with wrong figures due to erroneous templates (and which are consequently not aligned with the consistency checks developed through the EBA validation rules). As they derive from requests from the regulator to override the issues due to wrong templates, these specific cases should not be considered as errors on historical data for the next reporting periods in the context of these guidelines, and resubmissions of such wrong historical data should not be requested accordingly. This specific example also underlines the importance of providing some flexibility in the proposed approach to:
- take into consideration specific situations, and
- grant waivers when such cases are met.
Moreover, there are also some specific cases where the regulator requests to resubmit data of several reporting periods due to non-official EGDQs. As institutions are not pre-informed and aware of these additional consistency checks, these non-official EGDQs should also not be considered in the context of these guidelines. These guidelines should apply in the same way for the institutions and the regulator and the specific resubmission requests should be limited.
We would suggest also that a change in an interpretation or assumption would not constitute an error.
4. Would the proposed approach be feasible from the technology perspective considering the current reporting solutions?
To be able to restore historical data going back in time to several previous reference dates, the data must be historized in the systems upstream that would need to be re-configured.. However, the reporting tools and IT solutions currently in place in the institutions are not capable of storing the historical reporting parameters in order to resubmit the historical data when there are changes in taxonomy or regulations. What is more, additional investments and workload would occur for editors to rework former taxonomies.Therefore, the scenarios when resubmissions of historical data may not be required should be extended to regulatory and taxonomy changes and should not be limited to the cases already presented in the EBA consultation.
It should be noted that banks will have to implement at the same time the changes brought about by the implementation of Basel 4 and the new guidelines for the resubmission of historical data, which would apply from 31 December 2023. However, at present, given the stakes of regulatory projects, IT departments are prioritising developments linked to the implementation of Basel 4-induced changes. Thus, in order to ease the implementation of all the on-going regulatory projects and notably the EBA's guidelines on the resubmission of historical data, we would suggest a transition period during the implementation of Basel 4 during which data resubmissions will be limited to the current reference period.
a. If not, please provide concrete and realistic proposals for improving the proportionality element that can be efficiently implemented in the reporting systems without unreasonable costs or increasing the overall complexity.
Fixing materiality thresholds at unnecessarily low levels would increase the reporting burdens and costs on institutions without delivering material benefits to supervisors.In this context, and since both EBA and ECB have stressed in various opportunities an open dialogue between both institutions about their respective workstreams on the resubmission field, we strongly advise the EBA to consider the ECB work undertaken with key risk indicators (KRIs) and applicable thresholds as part of the ECB management report on data governance and data quality, as well as the ECB significant resubmissions policy with the ongoing pilot exercise ending by mid next year.
We propose accordingly to put in place an alternative methodology that would rely on the initial ECB principles on significant resubmissions using the 26KRIs deployed by the ECB and depending on the level of risks according to a red-amber-green classification. The ECB methodology is then adapted with lower thresholds to determine whether a resubmission of a specific reporting or template or a full reporting is required. For other reporting not covered by the 26 KRIs and that may be potentially affected by errors, we propose to complete the proposed methodology by thresholds based on a percentage of variation of the cell impacted by the error and / or a percentage of Prudential Own Funds impacted (for balance-sheet items and off balance-sheet items) and a percentage of Net Operating Income impacted (for profit and loss items).
We will be in position to present this industry proposal in the course of September for which we will appreciate to start coordinating with the EBA for a meeting to present and explain our proposal.
b. If such additional proportionality proposals are to be based on any threshold(s), please provide examples of such thresholds (relative and absolute) in relation to the size and complexity of your institution, and the reasoning behind that threshold.
We will be in position to present the industry proposal in the course of September for which we will appreciate to start coordinating with the EBA for a meeting to present and explain our proposal.6. If such additional proportionality proposals are to be based on less historical reference dates to be resubmitted (compared to those set out in paragraph 17), then what could these be for different types of institutions (large, medium-sized, SNCI)?
We consider the ECB approach and its declination for historical resubmissions would fit well institutions regardless of their size.The industry proposes to have a uniform methodology for all cases (annual, semi-annual, quarterly, and monthly reporting) and to resubmit only until the last year-end period.
This general approach to resubmission of historical data would apply to all types of institutions (large, medium-sized and SNCI). It would aim to have a simplest solution and a harmonized approach for all frequencies and all types of institutions regardless of their size.
Upload files
20230731_EBA_resubmission_data_FBF_comments.pdf
(239.08 KB)