Response to consultation on Guidelines on disclosure of non-performing and forborne exposures
Go back
Moreover, a review of the existing EBA Pillar 3 CR1-E template is welcome as the Template #1 and CR1-E present very similar information. We also recommend that Template #1 should be kept aligned with Finrep 19.
It should also be made clear that Template #3 will replace CR1-D which presents similar information; we propose once again to delete column m “of which impaired” (cf. question 1).
Moreover, it seems that template #5 is simply an adaptation of EU CR1-C (which has already entered into force. EBA GL/2016/11) due to IFRS9 by substituting concepts such as defaulted exposures by non-performing exposure, thus, further breakdown should be avoided in order to maintain consistency between reports. We would appreciate confirmation that the future guidelines will replace CR1-C when they will be implemented. Also, in Template # 6 information requested are already those of the existing template EU CR1-B. The EBF emphasizes the need of a consistent disclosure framework which avoids undue duplications.
It would be appreciated if more institution specific and country specific characteristics could be taken into consideration. From an EU-perspective, it would be more cost effective to implement regulatory NPL-actions in EU countries and for institutions where it is actually needed, based on supervisory evaluation, SREP, etc. In particular, more clarity is sought on whether these guidelines of disclosure should be in line with CRR requirements or with FinRep requirements.
We also suggest improving the articulation of these guidelines with GL 2016/11 (Guidelines on disclosure requirements under Part Eight of CRR 575/2013"), in particular, for the quantitative information templates mentioned in point 4.8 in order to avoid duplication of mandatory disclosure on the same topics. Confirmation by EBA that the future guidelines will replace the current ones will be upmost welcome as it will provide visibility for banks.
Regarding template #5 - Quality of NPEs by geography and template #6 - Quality of loans and advances by industry, we would appreciate confirmation that they will replace current CR1-E and CR1-C from 31 December 2019. Similarly, template #1 should replace CR1-E template #3 should replace CR1-D, template #4 should replace CR1- E and template #8 should replace CR2-B.
Regarding template #4, we have some additional comments on column m “accumulated partial write-offs”. We believe that it should be deleted as
- it will not help investors comparing write-off amounts from institutions based in different jurisdictions mainly due to significant differences in terms of tax treatment
- moreover, the disclosure of the stock of write off amounts may be relevant for supervisors having a good knowledge of the institutions’ background, but it is misleading for the large part of investors as they confuse write off with provisions and impairments. Accumulated write off amounts will not impact the future financial performance of the institutions, and therefore is not useful for investors.
- Write off amounts are more relevant when they are related to a period to explain the variation of the stock of non-performing exposures (cf. Template 8).
Regarding template 2, the information does not seem relevant for an investor. First of all, forbearance policies depend on an institution’s business model, hence related information may lead investors to a wrong interpretation of this information regarding the institution’s risk profile. Secondly, forbearance is a complex concept for the vast majority of the investors, with technical information such as the exit criteria and long probation periods. At best, they will not use this information, at worst, they will be mistaken and will not be able to compare institutions.
For these reasons, we suggest keeping this template for supervisors only, who can fully understand the quality of an institution’s forbearance methodology.
Specifically in relation to template #7, we have serious reservations with regards to the prevalent content. This template provides for the disclosure of very detailed information about the NPL’s delinquencies and, for this reason, may lead to extrapolations to the existing NPL stock of the capital requirements that would result from the eventual application of the prudential backstop to it (independent from the European Commission’s proposal being only applicable to new loans that become non-performing).
Regarding template #9, the detailed disclosure on collateral obtained by taking possession does not propose any valuable information to investors, but, on the contrary many misleading interpretations, with the following unintended consequences:
- Bank image / reputation risk as residential houses, autos, SMEs properties…. Takings are very unpopular
- Difficult Benchmarking as the takings systems in Europe are different across jurisdictions
- Wrong negative interpretation of the figures as there is no link either with the risk management policy when the bank grants the loan or with the recovery policy;
- For instance, the reader could wrongly interpret low amount figures as the bank runs laxest processes when granting the loan or when selling the collateral, whereas in reality the bank runs its business in a jurisdiction where a huge share residential loans is guaranteed by an insurance scheme and not by residential properties or the bank loan policy is based on income to value and not on loan to value.
- Finally, in the case of low NPL institutions (representing the majority of the case), institutions help the customer to sell himself the asset given as collateral, in order to limit the cost and timing of legal procedure and to preserve the potential future relationship with this customer. Therefore, if this template would be maintained as public disclosure, we strongly recommend that it becomes mandatory only for high NPL entities.
In addition, collateral taking possession policy is bank proprietary and confidential information. Disclosing it publicly would undermine banks competitive position.
Regarding template #10, the above concerns remain valid If important stocks of foreclosed assets are publicly disclosed in specific geographical areas, it could lead to a negative pro-cyclical effect with price decreases and difficulties for institutions to sell their foreclosed assets. Therefore, disclosing template 10 would go against both institutions’ economic interest and the action plan released by the Council of the European Union aiming at reducing the stock of NPL in institutions’ balance-sheets. From an operational point of view, the template is very burdensome to build as it is not related to any FINREP / COREP.
The same concern applies to template #8, column “Related net cumulated recoveries” from the lines 5 to 12 that could provide to potential buyers of non-performing loans portfolios sensitive information such as the price/discount the institution is ready to accept for the sale.
Templates #9 and #10 should only be submitted to supervisors and potential buyers of NPL portfolios, but not disclosed to the public. The reason for doing so is that this information may be particularly sensitive or even confidential and of limited use to the public. As mentioned above, the information that these templates entail are better suited for regulatory usage. Therefore, it could be detrimental to the sale of NPLs."
Question 1: Could you provide your views on whether adding an “of which” column to column ‘f’ of template 1 - “Credit quality of forborne exposures”, including the information on non-performing forborne exposures that are impaired (i.e. “of which impaired”) would be useful?
No, it would not add any more relevant information. From a general point of view, we would like to emphasize that the current financial information is already far too complex to be analysed by a very large part of investors. The tables presenting the technical typologies of exposures (including multiple “of which …” columns) can be useful for the supervisors. However, they are complex and confusing for most of the investors that do not have the full risk and accounting knowledge to understand all the subtleties of tables. Therefore, we recommend not adding any additional column, but also deleting from public disclosures (all the templates in this consultation) the distinction between “defaulted” and “impaired” representing very similar information.Moreover, a review of the existing EBA Pillar 3 CR1-E template is welcome as the Template #1 and CR1-E present very similar information. We also recommend that Template #1 should be kept aligned with Finrep 19.
Question 2: Could you provide your views on whether adding the columns with the breakdown of provisions for non-performing exposures by buckets of the number of days that the exposure has been past due to template 3 - “Credit quality of performing and non-performing exposures by aging of past due days” would be useful?
Adding the columns with the proposed additional breakdown would introduce further complexity. This information can be already found in FINREP template #18 but with a slightly different breakdown. Since the information is the same, it would be more useful to either remove the breakdown by buckets in this template or to align the buckets with FINREP template #18. Additionally, if the purpose is to disclose information reported in FINREP templates, there already exist transparency exercises made by the EBA each year where some regulatory information is disclosed. If entities also need to disclose them as part of financial statements or Pillar III disclosures, wouldn’t this imply duplicating sources of the same public information?It should also be made clear that Template #3 will replace CR1-D which presents similar information; we propose once again to delete column m “of which impaired” (cf. question 1).
Question 3: Could you provide your views on whether the breakdown between “on balance sheet exposures” and “off balance sheet exposures” included in template 5 – “Quality of Non-performing exposures by geography” is useful?
This breakdown is not even required in the supervisory reporting package (COREP) nor in the BCBS consultative document mentioned above. Therefore, we believe that the breakdown is of little value to users and at a higher cost to credit institutions since the information is not subtracted directly from existing reports.Moreover, it seems that template #5 is simply an adaptation of EU CR1-C (which has already entered into force. EBA GL/2016/11) due to IFRS9 by substituting concepts such as defaulted exposures by non-performing exposure, thus, further breakdown should be avoided in order to maintain consistency between reports. We would appreciate confirmation that the future guidelines will replace CR1-C when they will be implemented. Also, in Template # 6 information requested are already those of the existing template EU CR1-B. The EBF emphasizes the need of a consistent disclosure framework which avoids undue duplications.
Question 4: Could you provide your views on whether the information on loans and advances secured with immovable property with a loan-to-value higher than 60% and lower than 80% included in row 3 of template 7 – “Collateral valuation - Loans and advances at cost or amortised cost” is useful?
We propose to delete template #7 as the information required is very detailed, in particular w.r.t. NPL delinquencies as it may lead to extrapolations to the existing NPL stock of the capital requirements that would result from the eventual application of the prudential backstop to it (independent from the European Commission’s proposal being only applicable to new loans that become non-performing).. Further, not all jurisdictions use Loan-To-Value ratios.Question 5: Do you agree with the overall content of these guidelines and with the templates proposed? In case of disagreement, please outline alternatives that would help to achieve the purpose of the guidelines.
Overall, the EBF suggests to harmonise disclosure requirements already requested by EBA for Pillar 3 disclosure and with BCBS requirements. The proposed disclosure of detailed templates based on a specified NPL condition (which is yet to be in force as it is still subject to public consultation) is in principle considered adequate. Nevertheless, as previously mentioned, the use of a single threshold for the identification of institutions with high NPL levels may in same cases not be appropriate as it does not consider the existence of significantly different economic conditions within the EU countries or any additional risk mitigating factors (e.g. coverage levels, expected credit loss, supervisory backstops, ...).It would be appreciated if more institution specific and country specific characteristics could be taken into consideration. From an EU-perspective, it would be more cost effective to implement regulatory NPL-actions in EU countries and for institutions where it is actually needed, based on supervisory evaluation, SREP, etc. In particular, more clarity is sought on whether these guidelines of disclosure should be in line with CRR requirements or with FinRep requirements.
We also suggest improving the articulation of these guidelines with GL 2016/11 (Guidelines on disclosure requirements under Part Eight of CRR 575/2013"), in particular, for the quantitative information templates mentioned in point 4.8 in order to avoid duplication of mandatory disclosure on the same topics. Confirmation by EBA that the future guidelines will replace the current ones will be upmost welcome as it will provide visibility for banks.
Regarding template #5 - Quality of NPEs by geography and template #6 - Quality of loans and advances by industry, we would appreciate confirmation that they will replace current CR1-E and CR1-C from 31 December 2019. Similarly, template #1 should replace CR1-E template #3 should replace CR1-D, template #4 should replace CR1- E and template #8 should replace CR2-B.
Regarding template #4, we have some additional comments on column m “accumulated partial write-offs”. We believe that it should be deleted as
- it will not help investors comparing write-off amounts from institutions based in different jurisdictions mainly due to significant differences in terms of tax treatment
- moreover, the disclosure of the stock of write off amounts may be relevant for supervisors having a good knowledge of the institutions’ background, but it is misleading for the large part of investors as they confuse write off with provisions and impairments. Accumulated write off amounts will not impact the future financial performance of the institutions, and therefore is not useful for investors.
- Write off amounts are more relevant when they are related to a period to explain the variation of the stock of non-performing exposures (cf. Template 8).
Regarding template 2, the information does not seem relevant for an investor. First of all, forbearance policies depend on an institution’s business model, hence related information may lead investors to a wrong interpretation of this information regarding the institution’s risk profile. Secondly, forbearance is a complex concept for the vast majority of the investors, with technical information such as the exit criteria and long probation periods. At best, they will not use this information, at worst, they will be mistaken and will not be able to compare institutions.
For these reasons, we suggest keeping this template for supervisors only, who can fully understand the quality of an institution’s forbearance methodology.
Specifically in relation to template #7, we have serious reservations with regards to the prevalent content. This template provides for the disclosure of very detailed information about the NPL’s delinquencies and, for this reason, may lead to extrapolations to the existing NPL stock of the capital requirements that would result from the eventual application of the prudential backstop to it (independent from the European Commission’s proposal being only applicable to new loans that become non-performing).
Regarding template #9, the detailed disclosure on collateral obtained by taking possession does not propose any valuable information to investors, but, on the contrary many misleading interpretations, with the following unintended consequences:
- Bank image / reputation risk as residential houses, autos, SMEs properties…. Takings are very unpopular
- Difficult Benchmarking as the takings systems in Europe are different across jurisdictions
- Wrong negative interpretation of the figures as there is no link either with the risk management policy when the bank grants the loan or with the recovery policy;
- For instance, the reader could wrongly interpret low amount figures as the bank runs laxest processes when granting the loan or when selling the collateral, whereas in reality the bank runs its business in a jurisdiction where a huge share residential loans is guaranteed by an insurance scheme and not by residential properties or the bank loan policy is based on income to value and not on loan to value.
- Finally, in the case of low NPL institutions (representing the majority of the case), institutions help the customer to sell himself the asset given as collateral, in order to limit the cost and timing of legal procedure and to preserve the potential future relationship with this customer. Therefore, if this template would be maintained as public disclosure, we strongly recommend that it becomes mandatory only for high NPL entities.
In addition, collateral taking possession policy is bank proprietary and confidential information. Disclosing it publicly would undermine banks competitive position.
Regarding template #10, the above concerns remain valid If important stocks of foreclosed assets are publicly disclosed in specific geographical areas, it could lead to a negative pro-cyclical effect with price decreases and difficulties for institutions to sell their foreclosed assets. Therefore, disclosing template 10 would go against both institutions’ economic interest and the action plan released by the Council of the European Union aiming at reducing the stock of NPL in institutions’ balance-sheets. From an operational point of view, the template is very burdensome to build as it is not related to any FINREP / COREP.
The same concern applies to template #8, column “Related net cumulated recoveries” from the lines 5 to 12 that could provide to potential buyers of non-performing loans portfolios sensitive information such as the price/discount the institution is ready to accept for the sale.
Templates #9 and #10 should only be submitted to supervisors and potential buyers of NPL portfolios, but not disclosed to the public. The reason for doing so is that this information may be particularly sensitive or even confidential and of limited use to the public. As mentioned above, the information that these templates entail are better suited for regulatory usage. Therefore, it could be detrimental to the sale of NPLs."