Response to consultation paper amending Guidelines on definition of default
1. Question 1: Do you believe the current guidelines result in some exposures under forbear-ance measures to be incorrectly classified as defaults, thus hindering proactive, preventive and meaningful restructurings given the detrimental effects that defaulted status has for the affected obligors? If so, please further specify the characteristics of the exposures, which you deem as being subject to an incorrect classification of default.
In practice, a loss in net present value may already arise from the deferral of interest payments or from noninterest-bearing deferrals of principal repayments, without this necessarily implying a genuine risk of default of the relevant exposure (e.g. due to the obligor’ short term liquidity issues that cannot be covered with existing measures – like grace periods – but where no other unlikeliness to pay indicator is in play). On the contrary, such deferrals can serve as restructuring measures precisely intended to reduce the probability of default. The current rule distorts the overall picture by treating economically viable exposures as defaulted, especially in cases where the nominal amounts are not affected. Also, since the termination of the default reason “restructuring/refinancing” is usually subject to manual review, and the relevant data fields (3501/3511) capture both the amount and duration of arrears, the introduction of grace periods or de minimis thresholds could be implemented in practice. For low exposure amounts, this is particularly critical because they occur to NPV changes that breach the 1% NPV limit more frequently, as there is no materiality threshold applied when assessing the NPV limit. To remedy the effects of the fixed NPV 1% limit and incorrect default classification resulting from it, we envisage two potential remedial actions: a) increase the NPV limit to 2%; or b) apply at least a materiality threshold of EUR 500 to the NPV change of an exposure.
2. Question 2: Do you think that relaxing the criteria for the minimum period before returning to the non-defaulted status for defaulted forborne exposures could be an appropriate measure to alleviate a higher burden on your institution and clients? How material would the difference be in your case between the amounts of forborne exposures classified as NPE and as defaulted if the minimum one-year probation period in the definition of default were reduced to three-months for certain forborne exposures (with change in NPV below 5% and no loss on the nominal amount)? Would that proposal create additional operational burden or practical impediments? Do you see support such proposal, and if so, for which reasons?
While relaxing probation period could generally be seen as a welcome step to alleviate burden on institutions and clients, we note that the overall implications are more nuanced and much depends on the processes and systems of institutions – including whether they apply SA or IRB. Therefore, it is difficult to establish a priori the overall outcome of a potential relaxation on the criteria for the minimum period before returning to the non-defaulted status for defaulted forborne exposures will affect institutions. Where banks have for instance already aligned the definitions of default and NPE, changing the requirements to shorten probation periods will create a material gap for all affected exposures and generate adjustment efforts in terms of system developments, reporting, internal policies updates. In addition, the list of criteria for a shorter probation period is difficult to operationalise and/or automate, which hinders the continued application of automated probation/curing algorithms. This could apply both in the case of SA and IRB approach. Some members reported that especially in the retail segment a majority of (de)recognition algorithms are automated, therefore, introducing such a gap could become very operationally burdensome. Furthermore, consideration should also be given to the fact that such a modification may have an impact in terms of model material change for institutions applying the IRB approach, whereas the target is to keep the used DoD as stable as possible throughout the time. Overall, it is essential that the definitions of default and non-performing exposure are harmonised to the greatest extent possible and that this is done already in the level 1 text – as alignment within CRR would ensure greater consistency. A lack of harmonisation creates challenges not only for risk classification but also for accounting practices. (See answer to Question 5) Some members, have also pointed out that under the current rule, any agreement to remedy an overdraft is necessarily considered a forbearance measure if it is made only after the occurrence of a 90-day past due. If this measure, however, leads to a quick resolution of the arrear, a one-year probation period would be disproportionate. Finally, we would highlight again that a higher NPV threshold for distressed restructuring could help avoid intensive manual reviews.
3. Question 3: Do you see any alternatives other than those referred to in this section that the EBA should consider under Article 178(7) CRR to update the Guidelines and encourage insti-tutions to engage in proactive, preventive and meaningful debt restructuring to support ob-ligors?
It would be appropriate to introduce changes to the CRR to ensure the alignment of the criteria for forborne defaulted and forborne non-performing exposures. This would be possible via the amendment of Art. 47a, paragraph 6 of the CRR as well as of the respective paragraphs of Annex V of the EBA ITS on supervisory reporting (e.g. paragraph 157 and 176) would be needed.
It would also be appropriate to reconsider the treatment of the probation period under paragraph 71 of the draft Guidelines so that such exposures are not classified as defaulted. Upon the release of provisions, the schedule under Article 47c CRR remains unchanged, since the exposure continues to be classified as nonperforming. As a result, the position – now lacking coverage (provisions, Article 47c(3)(b)(i) CRR) – may immediately trigger a backstop deduction. This outcome would be illogical, as the release of provisions is substantively justified and therefore does not represent a shortfall that the backstop is intended to address. To avoid extensive net present value calculations, the identification of loans subject to shortened probation periods should rely on qualitative criteria. For instance, minor measures – such as the capitalisation of arrears – could justify a monitoring period of three months, whereas more substantial measures – such as a restructuring involving the transfer of the exposure into a new account – could require a 12-month monitoring period. Finally, point d) of Article 47b, para 2, of the CRR should be amended to clarify that it stipulates which measures must be classified as forbearance measures. In our opinion, point d) should limit forbearance classification to cases when “the measure results in a total or partial cancellation of the debt obligation” due to experienced or expected financial difficulties of the customer, in line with the previous points a), b), c). This amendment will ensure additional flexibility and avoid forbearance classification where the cancellation of debt obligation is not related to individual difficulties, but external factors, like changes in the legal framework or similar.
4. Question 4: Do you use internal definitions of default and NPE that are different from each other? Which differences are these and how material are those differences? Do you have any reasons or observed practical impediment that warrants a different definition of NPE and default? If so, please provide examples where a different definition of NPE and default is appropriate.
This aspect is essentially institution specific, however we would highlight the use of adjusting the NPE definition to better reflect the definition of default (see Q5).
5. Question 5: Would a potential lack of alignment between the default and NPE definition lead to issues in accounting in your case?
Better alignment between the definition of default and of NPE would certainly make many things easier. However, divergence between the two definitions is preferable to adjusting the default definition towards that of NPE – not least in consideration of possible effects on internal models. The optimal solution would be to adjust the NPE definition to better reflect the definition of default.
6. Question 6: Do you agree that no specific provisions should be introduced for moratoria on the grounds of the sufficient flexibility of the revised framework? In case you think the pro-posed alternative treatment for legislative moratoria should be included in these guidelines, do you have any evidence of the definition of default framework being too procyclical in the context of moratoria? Do you agree with the four conditions that need to be satisfied?
With regard to payment moratoria, while the EBA and competent authorities demonstrated flexibility to act during the COVID-19 pandemic in using the framework’s flexibility to accommodate temporary relief measures, the scope for introducing broader clarifications within the overall framework could be further assessed to ensure greater consistency, promptness and legal certainty in future crisis scenarios (e.g. cases when moratoria – public or private – are introduced in cases of natural disasters and questions about forbearance may again arise).
7. Question 7. Do you agree with the revised treatment of technical past due situations in rela-tion to non-recourse factoring arrangements? And if you do not agree, what are the rea-sons? Do you have any comments on the clarifications of paragraphs 31 and 32 in the current GL DoD?
We agree with the amendments proposed. Aligning factoring past-due situations with art. 178 (1)(b) and with other exposure types streamlines and harmonises the “days past due” framework. Furthermore, the additional time to settle possible disputes and responsibilities would most likely decrease the possibility of incorrect defaults. The revision of para. 31 and 32, as indicated in para. 58, would simplify the Guidelines.
8. Question 8. Do you agree with the other changes to the guidelines to reflect updates from Regulation (EU) 2024/1623?
Paragraph 32
We notice that the information related to paragraph 32, which has been deleted as a whole in the draft GLs. Currently, it stipulates the moment when the “Days Past Due” counting should start in case of undisclosed factoring. We have seen that an additional point f) is suggested to be added to paragraph 23, which specifies the definition of a technical past due situation in case of undisclosed factoring. However, the current paragraph 32 contains an important reference as to which moment has to be considered to start DPD counting in case of undisclosed factoring that has not been reflected in point f), para. 23
Paragraph 53
In paragraph 53, the description of one of the possible “Unlikeliness to Pay” indications in point d) changed from “(d) the exposures to the obligor have been subject to distressed restructuring more than once” to “(d) the exposures to the obligor have been subject to a forbearance measure more than once”. The new wording might be misleading because it suggests that if there were several forbearance measures – without default – granted to exposures of a debtor, this might indicate unlikelihood to pay and would lead to default. We consider that this change makes criterion d) much stricter than it is currently, which would contradict one of the stated aims of the GL revision: to provide more flexibility to institutions in providing supporting restructuring measures to customers. It is not clear why a debtor, who had more than one forbearance measure in the past, but never defaulted, should be assessed as unlikely to pay on the basis of successful forbearance measures granted to such debtor in the past, which in many cases might also mean a very long timespan (e.g. in case of mortgage or other loans with long maturities). We consider the wording “(d) the exposures to the obligor have been subject to a distressed forbearance measure more than once” more reflective of the aim of this requirement because it includes only those exposures that were in default (due to distressed restructuring)
Paragraph 54
The EBA should clarify why the first sentence in paragraph 54 has been deleted (“Any concession extended to an obligor already in default should lead to classifying the obligor as a distressed restructuring”). Does this mean that the EBA considers that if a bank grants a concession to an obligor in default, this will not lead to additional recognition of the indicator “Diminished financial obligation due to a forbearance measure”, and therefore, no respective probation period of 1 year will be required? In our understanding, granting a concession to a debtor already in default represents a forbearance measure and as it is already the case that a debtor is considered unlikely to pay, being in default, should lead to recognition of the indicator “Diminished financial obligation due to a forbearance measure”.We think that the existing sentence should be rather re-phrased to: “Any concession extended to an obligor already in default, should lead to classify the obligor as defaulted due to diminished financial obligation due to a forbearance measure”. In addition, the EBA should clarify whether it is still expected that all the “forborne non-performing” exposures are also “defaulted due to diminished financial obligation due to a forbearance measure”, given that the draft GLs do not contain anymore the provision “All exposures classified as forborne non-performing in accordance Article 47a of Regulation (EU) No 575/2013 should be classified as defaulted due to diminished financial obligation due to a forbearance measure”.
Paragraph 55
In paragraph 55, the phrasing “institutions should also assess whether a forbearance measure has been consented” seems not fully in line with the current framework:
- If a modification of the schedule was caused by financial difficulties of the obligor, then the granted measure is by definition a forbearance measure, i.e. the requirement to assess whether it represents a forbearance measure is redundant/contradictory to the forbearance definition;
- If an institution consents to forbearance measure, this does not mean default automatically.
We suggest that the EBA review this paragraph to avoid any potential legal uncertainty or misinterpretation.
Paragraph 73(f)
We understand the need to align the EBA GL with the latest CRR wording for this indicator, but substituting the wording “distressed restructuring” with the word “forbearance” might be misunderstood in a way that a forbearance measure is treated as an equivalent to a default indicator (currently called “distressed restructuring”), whereas a forbearance measure is not by itself a default indicator/reason, but becomes such only if diminished financial obligation results from this forbearance measure or if additionally other “Unlikeness to Pay” indications are observed by the institution due to the granted forbearance measure. To avoid a misunderstanding and treating any forbearance measure as default in this context, it seems more appropriate to use a formulation “distressed forbearance” instead of only “forbearance” in the wording of point f) of this paragraph.
Paragraph 117 (newly added section on legislative moratoria)
As the currently proposed text only refers to Member States, EBA might also include recognised third countries in the considered framework for legislative moratoria. In the banking groups where several units are located outside of the EU but in recognised third countries, we consider that a harmonised approach should be applied to debtors in scope of legislative moratoria so that if all the eligibility conditions are met, also legislative moratoria in recognised third countries and respective debtors could be covered by the suggested framework. Otherwise, these groups might face a different treatment of debtors in scope of the legislative moratoria on the same conditions as in a Member State, which will be caused solely by the geographic location / different jurisdiction of those debtors, even though the banking or financial regulatory and supervisory framework is generally deemed to be comparable/equivalent to the EU.