Response to consultation on draft RTS on the identification of a group of connected clients (GCC)
Go back
This RTS would thus cause significant implementation costs due to the necessary technical adaptations and will require a renewed adjustment of the documentation and processes in the institutions. This should be avoided if possible. In addition, we would like to note the following:
1) Scenario C 1, 51 i): We would like to suggest to further clarify letter i) as follows:
“The absence of economic interdependence or any other factors that could be indicative of a material positive correlation risk of contagion between the credit quality of the parent undertaking A and the credit quality of the SPE/SPV (B, C or D). Among other factors, the absence of a potential reliance on parent undertaking A for funding sources and circumstances of deconsolidation of the SPE/SPV under the applicable accounting rules have to be assessed as potential signs of economic independency.”
The wording proposed in the draft RTS is ambiguous because the word “correlation” is a statistical term. The concept of a single risk, however, is based on the risk of contagion between clients that are exposed to the same idiosyncratic risk factor.
2) Scenario E7, 65: The wording "it cannot be ruled out" means that if the refutation of the single risk cannot be ruled out (after a possibly elaborate examination and documentation), economic dependency is to be assumed and a GCC is to be formed due to economic dependence (burden of proof on the bank). This is a tightening of the current regulation and should be reworded for this situation ("if the economic dependence cannot be disproved").
3) Scenario E7, 65: The scenario in 65 builds on the example in 64, i.e. we assume equal participation of P1 and P2. In our opinion, this means that the connections between P2 and A/B/C are missing from the diagram in 65. If these have been omitted for the sake of clarity, this should be clarified by a comment. Furthermore, the companies are incorrectly labelled B/C/D and should be renamed A/B/C (analogous to the text in 65).
4) Scenario E8, 67: The previous indicator of consolidation is now made a mandatory criterion, contrary to recital 6: "that at least should be considered" (see also answer to Q5). Furthermore, the facts do not need to be presented in an example, as it does not provide any further insight. If the example were to remain, then the heading of 67 Example E8 would have to be adapted, as it describes the facts of a GCC due to economic dependence and the content refers to a GCC due to control (Art. 22 (7) includes the indicators of a control unit).
5) The wording in 11 "It has to be noted that the term "legal person" used in the RTS should be understood and applied to any entity other than a natural person irrespective of its national legal structure (also encompassing e.g. civil law associations) would also include civil-law associations. This would mean that the current alternative treatment of civil-law associations would no longer apply and that a GCC would always have to be formed on the basis of control between a shareholder of the civil-law association and the civil-law association itself, irrespective of any entrepreneurial activity of the civil-law association.
As noted in the introduction, specifically referring to Chapter 4 of the GL we miss an extensive elaboration on why the ‘consolidated’ large exposure rules can and should likewise be applied on an individual basis.
Therefore, in our view, when looking at exposures to subsidiaries in the individual scope, situations and elements in the draft RTS should be treated as indicative elements for the analysis on connectedness whereby reporting institutions should look at both the control relationship and the economic dependency to determine whether there is a case for connectedness.
In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
Art. 1 (3) appears to suggest that some of the former indicators may become mandatory criteria. At least point (c) should remain indicative based on practical experience to date.
We understand that the rationale underpinning Article 1 followed the consolidated perspective as indicated in Basel Large Exposure Framework SCO10.1. In that vein, it however ignores that such an approach cannot be applied at the individual level on a 1:1 basis. Please see also our comments under General Comments and Q2.
In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
As already mentioned, some of the requirements of the existing GLs would become mandatory under the wording of the draft RTS. In the draft RTS, the wording “shall constitute a single risk” indicates that, in any case where there are consolidated financial statements, institutions have to form a group of connected clients. In our view, that goes too far because there are different possibilities as to why consolidated financial statements have to be set up and this would not take into account the particularities of the internal intra-group situation.
In addition, there is need for a lot of time and effort to scrutinize documents in order to correctly form groups of connected clients. Time and effort would increase significantly if Art. 1 (1) would enter into force in the proposed wording and extent (Compliance cost to identify and maintain all relevant exposures for the large exposure reporting). The drafting does not give due consideration to such costs which are also not taken into account in Part 5.1 on the cost-benefit analysis.
Moreover, national discretions have to be taken into account, where a consolidation required is based on banking law.
It should furthermore be clarified that a “supervisory consolidation” for certain purposes (e.g. credit institutions permanently affiliated to a central body according Art. 10 CRR) should not be covered under Art. 1 (1). National law in various Member States defines and regulates Associations according to Art. 10 CRR and requires a consolidated financial statement. Thus, there is a contractual group but no control according to IFRS 10. (For instance, in Austria Para 30a (8) BWG stipulates that, for the purposes of full consolidation, the central organsisation is to be treated as a superordinate institution and each assigned institution and, under certain conditions, each contributing legal entity as a subordinate institution.) In the absence of an ultimate controlling parent company, a consolidated presentation can only be prepared in the sense of a group of equals.
In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
In general, we think that it is important to have the possibility to explain, why certain listed circumstances or indicators nevertheless do not lead to a single risk. Banks may still face the problem that there is uncertainty on the extent to which arguments made to refrain from economic dependency are accepted. Moreover, the assessment whether an economic dependency exists can be time and resource intensive for banks.
With regard to the criterion mentioned under a), we would like to point out that in our view this is already included in the definition of single risk due to economic dependence ("Two or more natural or legal persons constitute a single risk because they are so interconnected that, if one of them were to experience financial problems, in particular funding or repayment difficulties, the other person(s) would also be likely to encounter funding or repayment difficulties") and is not a type of economic connection but a consequence of the economic connection or dependence.
A corresponding dependency can exist both for a share of < 50 % and for a share of > 50 % due to substitutability. Threshold values are not suitable for determining dependencies due to a lack of sufficient validation. Furthermore, they are too general and thus arbitrary. Too often, they would simply not fit the concrete case. We therefore advocate leaving the wording at "significant" but continuing with these cases on a purely indicative basis.
In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
Question 1. Could you please indicate, if the approach of sections 4, 6 and 7 of the existing EBA guidelines, now transposed in the Articles of the draft RTS, remains sound and is implementable with no major challenge or unduly high costs. Please elaborate.
The requirements of the Guideline from Chapters 4, 6 and 7 have been completely transferred to the RTS. However, they have been tightened up – even though, according to the summary, no significant changes are involved.This RTS would thus cause significant implementation costs due to the necessary technical adaptations and will require a renewed adjustment of the documentation and processes in the institutions. This should be avoided if possible. In addition, we would like to note the following:
1) Scenario C 1, 51 i): We would like to suggest to further clarify letter i) as follows:
“The absence of economic interdependence or any other factors that could be indicative of a material positive correlation risk of contagion between the credit quality of the parent undertaking A and the credit quality of the SPE/SPV (B, C or D). Among other factors, the absence of a potential reliance on parent undertaking A for funding sources and circumstances of deconsolidation of the SPE/SPV under the applicable accounting rules have to be assessed as potential signs of economic independency.”
The wording proposed in the draft RTS is ambiguous because the word “correlation” is a statistical term. The concept of a single risk, however, is based on the risk of contagion between clients that are exposed to the same idiosyncratic risk factor.
2) Scenario E7, 65: The wording "it cannot be ruled out" means that if the refutation of the single risk cannot be ruled out (after a possibly elaborate examination and documentation), economic dependency is to be assumed and a GCC is to be formed due to economic dependence (burden of proof on the bank). This is a tightening of the current regulation and should be reworded for this situation ("if the economic dependence cannot be disproved").
3) Scenario E7, 65: The scenario in 65 builds on the example in 64, i.e. we assume equal participation of P1 and P2. In our opinion, this means that the connections between P2 and A/B/C are missing from the diagram in 65. If these have been omitted for the sake of clarity, this should be clarified by a comment. Furthermore, the companies are incorrectly labelled B/C/D and should be renamed A/B/C (analogous to the text in 65).
4) Scenario E8, 67: The previous indicator of consolidation is now made a mandatory criterion, contrary to recital 6: "that at least should be considered" (see also answer to Q5). Furthermore, the facts do not need to be presented in an example, as it does not provide any further insight. If the example were to remain, then the heading of 67 Example E8 would have to be adapted, as it describes the facts of a GCC due to economic dependence and the content refers to a GCC due to control (Art. 22 (7) includes the indicators of a control unit).
5) The wording in 11 "It has to be noted that the term "legal person" used in the RTS should be understood and applied to any entity other than a natural person irrespective of its national legal structure (also encompassing e.g. civil law associations) would also include civil-law associations. This would mean that the current alternative treatment of civil-law associations would no longer apply and that a GCC would always have to be formed on the basis of control between a shareholder of the civil-law association and the civil-law association itself, irrespective of any entrepreneurial activity of the civil-law association.
As noted in the introduction, specifically referring to Chapter 4 of the GL we miss an extensive elaboration on why the ‘consolidated’ large exposure rules can and should likewise be applied on an individual basis.
Question 2. Have you identified any additional aspect(s) that would require clarification? In this vein, would you see the need for further illustrative examples (and if yes, on which precise situation or specific case)? Please elaborate.
Yes. Throughout the wording of the draft RTS it emerges that some requirements and/or elements become mandatory where there are situations and circumstances which would be more appropriate to treat as indications, e.g. in intra-group relations of the reporting institution. In the baseline scenario “Control relationship” at page 15 of the consultation document, the example only refers to SPEs/SPVs when talking about the possibility that the assumption of single risk can be refuted despite the existence of a control relationship. In our view one should also include an example when the reporting institutions has exposures to its subsidiaries as there would be very good arguments that no single risk may prevail despite the circumstances of paragraph 1, 2 and 4 of Article 1 of the draft RTS. For example, when the subsidiary is supervised, the reporting institution which is fully consolidating on the basis of IFRS 10 might be confronted with the refusal by the ‘local’ supervisor to execute powers, i.e. when the soundness of the supervised entity will be endangered by executing control powers.Therefore, in our view, when looking at exposures to subsidiaries in the individual scope, situations and elements in the draft RTS should be treated as indicative elements for the analysis on connectedness whereby reporting institutions should look at both the control relationship and the economic dependency to determine whether there is a case for connectedness.
In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
Question 3. After considering the circumstances set out in Article 1 that constitute a single risk by means of control, could you please indicate if the described circumstances are sufficiently clear? Please elaborate.
Art. 1 (2)(c) should be limited to specific cases in which natural persons or legal entities can exercise a controlling influence over legal entities. Otherwise, questions arise about legal relationships between two natural persons (guardianship, care, trusteeship, execution of wills, etc.), but these have nothing to do with the control concept.Art. 1 (3) appears to suggest that some of the former indicators may become mandatory criteria. At least point (c) should remain indicative based on practical experience to date.
We understand that the rationale underpinning Article 1 followed the consolidated perspective as indicated in Basel Large Exposure Framework SCO10.1. In that vein, it however ignores that such an approach cannot be applied at the individual level on a 1:1 basis. Please see also our comments under General Comments and Q2.
In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
As already mentioned, some of the requirements of the existing GLs would become mandatory under the wording of the draft RTS. In the draft RTS, the wording “shall constitute a single risk” indicates that, in any case where there are consolidated financial statements, institutions have to form a group of connected clients. In our view, that goes too far because there are different possibilities as to why consolidated financial statements have to be set up and this would not take into account the particularities of the internal intra-group situation.
In addition, there is need for a lot of time and effort to scrutinize documents in order to correctly form groups of connected clients. Time and effort would increase significantly if Art. 1 (1) would enter into force in the proposed wording and extent (Compliance cost to identify and maintain all relevant exposures for the large exposure reporting). The drafting does not give due consideration to such costs which are also not taken into account in Part 5.1 on the cost-benefit analysis.
Moreover, national discretions have to be taken into account, where a consolidation required is based on banking law.
It should furthermore be clarified that a “supervisory consolidation” for certain purposes (e.g. credit institutions permanently affiliated to a central body according Art. 10 CRR) should not be covered under Art. 1 (1). National law in various Member States defines and regulates Associations according to Art. 10 CRR and requires a consolidated financial statement. Thus, there is a contractual group but no control according to IFRS 10. (For instance, in Austria Para 30a (8) BWG stipulates that, for the purposes of full consolidation, the central organsisation is to be treated as a superordinate institution and each assigned institution and, under certain conditions, each contributing legal entity as a subordinate institution.) In the absence of an ultimate controlling parent company, a consolidated presentation can only be prepared in the sense of a group of equals.
Question 4. Is the additional Scenario C 0 related to the determination of a group of connected clients by means of control, listed in Section 3.4.1 (Groups of connected clients based on a control relationship), sufficiently clear? Would you see need for further illustrative examples of a control relationship?
See answers under Q2 and Q3.Question 5. After considering the circumstances set out in Article 2 that constitute a single risk by means of economic dependency, could you please indicate if the described circumstances are sufficiently clear? Please elaborate.
See also our comment to Q1 point 3: the indicative character incorrectly moves into a mandatory application. This is specifically relevant when looking at individual Large Exposure analysis.In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
In general, we think that it is important to have the possibility to explain, why certain listed circumstances or indicators nevertheless do not lead to a single risk. Banks may still face the problem that there is uncertainty on the extent to which arguments made to refrain from economic dependency are accepted. Moreover, the assessment whether an economic dependency exists can be time and resource intensive for banks.
With regard to the criterion mentioned under a), we would like to point out that in our view this is already included in the definition of single risk due to economic dependence ("Two or more natural or legal persons constitute a single risk because they are so interconnected that, if one of them were to experience financial problems, in particular funding or repayment difficulties, the other person(s) would also be likely to encounter funding or repayment difficulties") and is not a type of economic connection but a consequence of the economic connection or dependence.
Question 6. In point (c) of Article 2(1), would you prefer following a quantitative approach by replacing the term “significant part” with a threshold of “50% or more” as envisaged in point 1 of LEX 10.16? What would be the advantages or disadvantages? Please elaborate.
In our opinion, the introduction of a threshold value should be avoided. The introduction of such a threshold could quickly lead to a situation where economic dependencies are no longer assessed below the threshold and a CC is automatically formed above the threshold. However, the examination of economic dependence is always a case-by-case consideration.A corresponding dependency can exist both for a share of < 50 % and for a share of > 50 % due to substitutability. Threshold values are not suitable for determining dependencies due to a lack of sufficient validation. Furthermore, they are too general and thus arbitrary. Too often, they would simply not fit the concrete case. We therefore advocate leaving the wording at "significant" but continuing with these cases on a purely indicative basis.
Question 7. What is your view on the wording “that cannot be replaced in a timely manner without excessively increased costs” compared to the wording used in the GL “that cannot be easily replaced”? What do you think about this change, is it more comprehensible? Please elaborate.
The requirement is now somewhat more concrete and can be better made operational in practice, which is why we welcome the change. However, it is still not specific enough for a mandatory application; therefore, an indicator would be more suitable than a mandatory requirement.Question 8. Is the additional Scenario E 8 related to the determination of a group of connected clients by means of economic dependencies, listed in Section 3.4.2 (Establishing interconnectedness based on economic dependency), sufficiently clear? Would you see need for further illustrative examples of an economic dependency relationship? Please elaborate.
See answer to Q1. Furthermore, example E8 is sufficiently clear. We see no need for further examples on this.Question 9. After considering the circumstances set out in Article 3 that constitute a single risk by means of the combined existence of control and economic dependencies, could you please indicate if the described circumstances are sufficiently clear? Please elaborate.
The explanations are sufficiently clear, but very technical. It would be better to present general principles than trying to describe them only in technical terms. Recital 7 is also far too technical.In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.
Question 10. Is the additional Scenario E 7 related to the determination of a group of connected clients by means of the combined existence of control and economic dependencies, listed in Section 3.4.3 (Relation between interconnectedness through control and interconnectedness through economic dependency), sufficiently clear? Please elaborate.
The example is sufficiently clear. However, it must also be clear in the future that the understanding of control and risk contexts remains decisive for GCC formation and cannot be replaced by case groups.In our view a separate section should be included on how to apply the connectedness at the individual level taking into account the particularities of intra-group situations and relations, instead of applying the consolidated approach to the solo level on a 1:1 basis.