Response to consultation on Draft Guidelines on the methodology to estimate and apply credit conversion factors under the Capital Requirements Regulation

1. Question 1: How material are the cases for your institution where you would have to assign an SA-CCF to exposures arising from undrawn revolving commitments and thus restrict the use of own estimates of LGDs within the scope of application for IRB-CCF in the CRR3? For which cases would you not have enough data to estimate CCFs but have enough data to es-timate own estimates of LGDs?

The consultation paper on GLCCF introduces the concept of a “Fixed CCF” for exposures that fall within the scope of the IRB-CCF framework, but for which institutions are unable to assign an IRB-compliant CCF due to not meeting the minimum estimation requirements. This may occur, for example, when LGDs are robustly estimated, but CCFs are not, or when subsegments of the application portfolio lack observed defaults in the RDS. In such cases, institutions are required to apply a Fixed CCF, which incorporates a MoC quantified by the institution. The minimum level of final CCF including MoC must be no less than 100%.

First and foremost, the introduction of the Fixed CCF does not appear to be fully aligned with the provisions of CRR III - Article 166(8b), which explicitly permits the use of SA-CCF for exposures where the minimum requirements for estimating IRB-CCF, as outlined in Section 6, are not met (e.g., due to data scarcity). Notably, this article makes no reference to the LGD estimation approach, which in our view supports the continued applicability of the IRB-LGD combined with SA-CCF approach.

Regardless of the point above, while the introduction of a Fixed CCF for portfolios affected by data scarcity may offer a pragmatic solution—enabling institutions to apply their own IRB-compliant LGD estimates—the imposition of a minimum CCF of 100% appears overly conservative. This threshold may disproportionately disadvantage institutions using IRB-CCF compared to those applying the SA-CCF, particularly in the case of unconditionally cancellable commitments, where the gap between Fixed CCF and the corresponding SA-CCF can be significant for the same type of product.

In addition, the requirement to apply a “sufficiently conservative” MoC presents practical challenges, as its calibration can be complex and highly judgmental. The fact that the MoC may explicitly push final CCF values above 100% could further disadvantage institutions opting for the Fixed CCF approach instead of the SA-CCF. This issue is particularly relevant given that Supervisory Authorities tend to adopt a cautious stance when it comes to model estimates and may therefore require additional prudential uplifts, further amplifying the conservatism embedded in the final figures. Disproportionate MoCs are expected for portfolios with low volumes of data or when the starting level of the realised CCFs is very low (though not sufficiently robust), conflicting with the concept of MoC that should serve only to cover for the residual uncertainty. 

When CCFs significantly exceed 100% this may lead to a counterintuitive outcome in the application portfolio: fully drawn credit lines could be perceived as less risky—or even preferable—compared to undrawn lines subject to the Fixed CCF approach, despite the latter showing no evidence of utilisation. This paradox arises from the disproportionate conservatism embedded in the Fixed CCF, which may not accurately reflect the actual risk profile of undrawn exposures.

In general, there is also a lack of clear transitional provisions regarding the application of the Fixed CCF, as well as no defined mechanism for its periodic review or adjustment as more data becomes available. Furthermore, the back testing requirement set out in Recital 132 appears challenging to implement in practice.

As an alternative it could be suggested to:

• Allow for a fixed but conservative CCF that needs to be defined and regularly reviewed by the institution
• Specify a dedicated CCF input floor for these cases that is higher than the one in Art. 166(8c) CRR but still linked to the level of SA-CCF. 

With such an approach, the different levels of risk associated with different types of products could still be sufficiently acknowledged

Additional considerations

Banks have already invested heavily in the development and maintenance of IRB models, including robust LGD estimates. Replacing the current IRB-LGD + SA-CCF approach with a Fixed CCF would undermine these investments and create an uneven playing field compared to institutions applying SA-CCF. For subsidiaries of international banking groups, such treatment could also lead to intra-group inconsistencies and competitive disadvantages within the EU.

For these reasons, it is important to preserve the flexibility explicitly allowed under CRR III Article 166(8b), i.e. permitting the continued use of IRB-LGD in combination with SA-CCF in cases of data scarcity.

2. Question 2: Do you have any comments related to guidance on the identification of a relat-ed set of contracts which are connected such that they constitute a facility?

We acknowledge that the Guidelines build on the CRR3 definition of a credit facility as a “contract or set of contracts,” and that the EBA cannot deviate from this Level 1 provision. We welcome the clarification that consistency with risk management practice and the presence of overarching arrangements should guide the identification of a “set of contracts.” At the same time, we see a risk of divergent interpretations in practice, particularly in cases such as forward-starting loans or roll-overs of existing credit lines, where treating separate contracts as a single facility could lead to double-counting of exposures.

Furthermore, in cases of restructuring of connected facilities, both pre-default and post-default, it may occur that a revolving facility is used to repay a non-revolving facility (e.g., a term loan) via interfacility balance transfers. In these cases, no additional funds are provided to the obligor, however the drawing on the revolving facility increases, which affects the estimation of pre- and post-default drawings. We suggest that the EBA clarifies that such interfacility transfers should not impact the realised CCF or LGD. Providing this clarification would further reduce divergent interpretations and simplify the treatment of connected facilities.

As EBA leverages on the CRR3 definition of facility, which allows to consider a set of contracts for CCF calculation, we would like the EBA to be more explicit in the Guidelines that LGD and CCF can be calculated at distinct granularities. Indeed, the CRR definitions clarify that both the LGD and CCF parameters are related to a “single facility” as per articles 4.1(55) and 4.1(56), leaving the door open to an unintended reading of the regulation that both parameters need to be calculated at the same level of granularity. It is worth highlighting that supervisory reading as per paragraphs 260/316 of the ECB Guide to Internal Models (July 2025) still allows in CRR3 distinct granularities between LGD and CCF. Indeed, the rationale for calculating LGDs at a more aggregated level than the facility level is linked with the recovery processes, while the CCF parameters are linked with the way limits are granted and managed within the bank. In addition, the restriction of IRB-CCF models to revolving commitments only makes it all the more difficult to align LGD calculation granularity (as LGD models would cover both revolving and non revolving products) with the CCF calculation granularity.

3. Question 3: Do these GL cover all relevant aspects related to the definition of revolving commitments that you consider relevant for the scope of the IRB-CCF? Have you identified any product that should be in the scope of the IRB-CCF that is currently excluded in the GL? In terms of off-balance sheet exposures, how material are the exposures that fall within the defined scope of the IRB-CCF for your institution?

We consider the Guidelines broadly adequate but would welcome further clarifications to ensure consistent application across jurisdictions. In particular:

• Revolving commitments clearly in scope: overdrafts, credit cards, and re-advanceable working-capital lines, as illustrated in our case studies, are genuine revolving commitments and materially relevant for both retail and corporate portfolios.
• Items that should not be in scope: certain instruments risk being misclassified as revolving commitments when they are not:
  • Trade-finance instruments such as documentary credits and guarantees: These are contingent commitments already captured under Annex I (Bucket 4, 20% CCF) and should not be recast as revolving commitments.
  • Factoring “undrawn amounts”: Most factoring lines are uncommitted (outside CCF scope) or, if in scope, count as UCC (Bucket 5). They do not fit revolving-commitment logic.
  • Binding mortgage offers: These are agreements to lend (not credit lines) and must remain treated as financing commitments (Bucket 3, 40% CCF).
  • Contractual arrangements offered by an institution, but not yet accepted by the client, that would become a commitment if accepted by the client: Article 5(10) CRR defines commitments for the purposes of Part Three, Title II only as contractual arrangement that an institution offers to a client, and is accepted by the client, to extend credit, purchase assets or issue credit substitutes. Contractual arrangements offered but not yet accepted are not part of the definition of commitment. Only for the calculation of the exposure value under the Standardised Approach, these items are to be considered in accordance with Article 111(4) CRR. This interpretation seems to be aligned with the interpretation of the ECB, included in the revised ECB Guide to Internal Models, Credit Risk, paragraph 313(b).

Overall, no new products are proposed for inclusion, though further analysis is needed to assess whether certain trade finance instruments (e.g., Letters of Credit) or committed UCC factoring lines could exhibit revolving characteristics, as suggested by some members. The main need remains to ensure exclusions are explicit, avoiding misclassification and capital overstatement. Materiality of the exposures in scope is significant for banks with large retail and SME portfolios, while the impact is lower for trade-finance-focused institutions given the contingent nature of those exposures. This materiality assessment focuses on drawn exposures, but off-balance sheet products (e.g., guarantees, documentary credits) with not-yet-issued portions may also fall within IRB-CCF scope if revolving, an aspect currently underexplored in all the EBA Guidelines.

EBA should nuance its position regarding deferred debit cards. The revolving definition is intepreted with 3 cumulative criteria being met:

1. The obligor has flexibility on drawings, within a given limit; AND
2. The obligor has flexibility on repayments : the obligor decides the timing of its reimbursements; AND
3. The drawing capacity is replenished by the amount reimbursed.

This reading is based on article 166(8b) of CRR : “a commitment shall be deemed “revolving” where it lets an obligor obtain a loan where the obligor has the flexibility to decide how often to withdraw from the loan and at what intervals, allowing the obligor to drawdown, repay and redraw loans advanced to it. Contractual arrangements that allow prepayments and subsequent redraws of those prepayments shall be considered revolving.”

Banks should determine the revolving feature of products based on CRR3. Some deferred debit cards will not meet for instance the criteria (B) / (C) and will not be considered as revolving products. Banks should determine the revolving feature of products based on CRR3/European law definition. In overall, we think that considering increase of EAD for deferred debit cards is less of an issue because their drawings are off-set by the current account balance which is most of time positive. 

4. Question 4: Are there products that have an advised limit of zero but a nonzero unadvised limit that should be included in the scope of the IRB-CCF GL? How material are these cases for your institution?

No. If drawdown depends on conditions tied to non-credit events—as is typical for trade-finance guarantees—these are contingent, not revolver-type exposures. They should remain under the contingent-commitment classification per Annex I, not swept into revolving-commitment scope.

5. Question 5: Do you think that dynamic limits (e.g. limits the extent of which is dependent on the market value of financial collateral pledged by the obligor in relation to the revolving loan) warrant a specific treatment in the IRB-CCF GL? How material are these cases for your institution?

No special revolving-commitment treatment is warranted. Facilities with dynamic limits (e.g., Lombard, secured lending) are financing commitments by nature and should be treated accordingly—i.e., under commitment CCF rules—regardless of limit fluctuations. This avoids over-broadening the scope under Articles 1.1/1.2.

6. Question 6: Have you identified any unwarranted consequences of including fully drawn revolving commitments in the scope of the IRB-CCF. How material are these cases for your institution?

We are concerned that the inclusion of fully drawn revolving commitments in the scope of IRB-CCF modelling is not fully aligned with the definition of credit conversion factors under Article 4(56) CRR. By definition, CCFs apply to the undrawn portion of a commitment; applying them to facilities that are already fully drawn appears inconsistent with the Level 1 text. The same concern arises when comparing the Guidelines to paragraph 11 of the Background and Rationale, which likewise does not foresee CCFs being applied to fully drawn facilities.

We [VG1] do not see in level 1 text any ground for developing an approach for fully drawn revolving commitments which would estimate the EAD above the full on-balance sheet drawn amount. As article 166.8 of CRR indicates "An institution that uses IRB-CCF shall calculate the exposure value for undrawn commitments as the undrawn amount multiplied by IRB-CCF", if the undrawn amount is 0, the exposure value for undrawn part will be 0. In article 182.1c of CRR, only facilities which are close to be fully drawn are concerned by the fact that CCF should be effectively quarantined ("Institutions shall ensure that their IRB-CCF are effectively quarantined from the potential effects of region of instability caused by a facility being close to being fully drawn at the reference date"). We understand that such specific treatment does not include fully drawn commitments.

Additionally, we support further EBA guidance on the calibration of alternative realized CCFs, particularly regarding flooring. Options such as flooring at 0% (where drawings are not negative) or at the reference date utilization rate (to align with standard CCF flooring logic) could be considered. Clarity on this would prevent future inconsistencies, though we note the potential for over-conservatism, including bias in the long-run average (LRA) CCF when zero-flooring over-utilized facilities, and the material impact of number-weighted averaging on small-exposure outliers. We will assess member views, including those highlighting substantial shares of overdrawn cases, to determine the most appropriate approach during implementation.

Beyond these technical clarifications, we would also stress the broader modelling impacts, which we see articulated in three main considerations:

1. the floor introduces an unnecessary upward bias and implies a significant impact in terms of increase of the final estimated parameters
2. the floor prevents models from reflecting improvements in proactive management practices and processes
3. the floor increases the divergence between CCF estimates used for regulatory capital and those used for IFRS9 accounting purposes.

The combination of the quantification floor (highly impacting on the left tail of the CCF distribution) and the overly conservatism introduced by some elements of Draft GL CCF (highly impacting on the right tail of the CCF distribution) is pushing CCF estimates well beyond the expectations of financial institutions. This would result in an unjustified (and artificial) increase in allocated regulatory capital, a consequent reduction in lending capacity to the real economy, and a decline in banks’ appetite for revolving commitments, whose levels are increasingly driven towards the 100% threshold that characterizes full utilization (i.e., similar to “on balance sheet” exposures).

8. Question 8: Are there cases for your institution where the calibration samples should be shorter than the sample used to calculate the long run average (LRA) CCF

Question 7: Do you have any concerns on the introduction of the notion of the different samples that constitute the RDS for CCF estimation? Do you have a modelling practice implemented that deviates from this approach? 

The clarification and formalization of development, testing, and quantification samples in the EBA Guidelines is welcome, as it aligns with established modelling practices in IRB-compliant institutions. The creation of different samples in the risk differentiation (training versus testing out-of-sample and out-of-time) and in the risk quantification is consistent with the practices (and requirements) of PD and LGD parameters as well even before CRR3 finalization. 

Question 8: Are there cases for your institution where the calibration samples should be shorter than the sample used to calculate the long run average (LRA) CCF? 

This situation could be possible in the modelling experience: the backward replication of all the risk drivers may not always be feasible for very old data, whereas the ECB Guide to Internal Models (EGIM) encourages making all reasonable efforts to recover historical experience of losses and drawings (thus favoring the longest time series possible), though some members suggest shorter samples are unnecessary if representativeness is addressed by existing mechanisms. We deem that this requirement shall be set in an equilibrated way because ultimately the CCF model shall be calibrated to a downturn level: if the downturn falls within the period where there are observed data (or, in contrary case, if the downturn LGD is extrapolated) that period represents the calibration target for the CCF estimates, no matter if the calibration sample is shorter or equal to the LRA. Therefore, stressing the full alignment, envisaging even the application of MoCs looks excessively penalizing. 

We note and support the EBA’s approach to calculate the long-run average (LRA) CCF as a sample average (and not as annual average as per par. 322(d) of 2025 EGIM) and suggest aligning this expectation with the ECB Guide to ensure consistency across regulatory frameworks.

Furthermore, we propose to clarify that the “good vs. bad years” variability analysis specified in par. 322(e) of the 2025 EGIM is no longer relevant given that the LRA is now defined as a sample average (and not as annual averages). In our view, the representativeness framework already provides sufficient mechanisms to address the risk of a non-representative LRA period.

9. Question 9: Do you have any concerns with the requirements introduced to analyse and mitigate a lack of representativeness for CCF? Do the requirements on the different data samples when observing a lack of representativeness impede your ability to model CCF portfolios?

We welcome the intention to simplify representativeness requirements for CCF modelling and do not see provisions that would, in themselves, prevent the ability to model CCF portfolios. At the same time, we note that in practice the training sample will still need to ensure full representativeness with the application portfolio, since out-of-sample and out-of-time datasets—where full representativeness is required—are typically derived through stratification of the training sample.

It is also not fully clear how the requirement to prioritise model performance over representativeness should be applied. Representativeness is generally assessed after sample selection, while model performance can only be assessed once the risk differentiation function has been defined. Linking these two concepts’ risks creating practical challenges for model development, even if it may be useful for the review of estimates. To make the simplification more effective, the dimensions and methods for assessing representativeness should be reviewed and streamlined, for example in line with the adjustments already suggested for defining the scope of application. The creation of different samples in the risk differentiation (training versus testing out-of-sample and out-of-time) and in the risk quantification is consistent with the practices (and requirements) of PD and LGD parameters as well, even before CRR3 finalization. Nonetheless, one potential concern is for non-retail CCF models covering low default portfolios : due to low number of observations, it may not be possible to meaningfully split into the required samples.

10. Question 10: Do you have any concerns with linking the fixed CCF to the lack of historical data available to the institution in relation to the coverage by the RDS of material subseg-ments of the application portfolio? How is your institution currently treating these cases?

We have concerns with linking the fixed CCF to the lack of historical data and RDS coverage of material subsegments of the application portfolio, as it imposes a disproportionately conservative 100% floor on IRB-CCF estimates. This is overly punitive compared to SA-CCF values (e.g., 40% for Bucket 3 revolving commitments), creating a significant unlevel playing field between IRB and SA institutions. Instead, in these cases, we recommend adopting SA-CCF as the direct fallback (or as a floor on punctual estimates) that can be used in conjunction with an IRB-LGD (i.e. not forcing a reversion from the A-IRB to the F-IRB approach for the relevant portfolio subsegment) and extending this to other situations where reliable modeling is not feasible (e.g., as outlined in paragraph 131 of the Background and Rationale). Currently, institutions typically treat such cases by falling back to SA-CCF or applying conservative overrides to ensure compliance.

11. Question 11: Are there any concerns with requiring consistency in the analysis of changes in the product mix with the institution’s definition of facility? Are institutions able to identify and link contracts (partially) replacing other contracts where the closing or repayment of one contract is related to the origination of a new contract? Are institutions able to link new contracts that are originated after the reference date to related contracts existing at refer-ence date? In particular, is it possible in the case contracts that are revolving commitments are replaced by contracts that are non-revolving commitments (e.g. by a term loan)?

Yes, there are significant concerns with requiring consistency in the analysis of changes in the product mix with the institution’s definition of facility, as this could lead to disproportionately conservative and economically implausible CCF estimates exceeding 100% (due to including only revolving facilities in the denominator while capturing all facilities, including non-revolving like term loans, in the numerator). The criteria for identifying "distressed financing" as related facilities (e.g., the five points in paragraph 56 of the Background and Rationale) are overly broad and prone to interpretation, often triggering automatically (e.g., point c for any new finance in the 12-month period), requiring substantial effort to rebut and risking supervisory conservatism. Points a, b, and d could reasonably indicate distress (with a clearer threshold for d, e.g., 3 months benchmarked to past-due classification), but points c and e should not systematically qualify as such, as new finance under ordinary approval processes without risk signals (e.g., no watchlist or poor rating) lacks inherent linkage to existing revolving facilities, even if the obligor later defaults.

Institutions generally lack explicit links to identify and connect (partially) replacing contracts where closing or repayment relates to new origination, relying instead on complex algorithms analyzing balance and limit evolutions, which yield sub-optimal results for clients with multiple products and are particularly challenging for legacy portfolios with old historical data.

Linking new post-reference date contracts to existing ones is similarly difficult without explicit mappings (unless from source system changes), and especially so when revolving commitments are replaced by non-revolving ones (e.g., term loans), as this requires distinguishing ordinary originations from restructurings amid varying institutional definitions of "facility," leading to inconsistent materiality of mixed products across banks.

To address these issues, the guidelines should incorporate materiality thresholds (as in paragraph 54 of the Background and Rationale) to allow simplified approaches for low-materiality mixed products or legacy data, and consider substitution/consolidation cases (e.g., terminated facilities replaced by similar or fewer new ones) as transformations for realized CCF calculation.

In addition, in practice, restructuring is often assessed at client level rather than facility by facility. Even if only part of a client’s facilities are replaced, one could reasonably argue that the entire credit structure has been restructured, making it difficult to draw a strict one-to-one link between facilities before and after the restructuring. In such cases, an analysis at client level—mapping all facilities at default to those existing at the reference date—may provide a more economically sound approach.

Finally, EBA is not consistent in the Guidelines when asking for instance consistency between application and estimation on fully drawn commitments but the selection of perimeter does not ensure such consistency. Indeed, paragraph 60, page 71 of the Guidelines implies notably that if the facility is non-revolving at the reference date but it becomes revolving between 12 months before default and default, it is included in the IRBA CCF modelling perimeter. However, in the CCF application, at a certain snapshot/reference date of RWA calculation, the bank cannot anticipate future change of product type for a specific facility and can only observe the product nature at the snapshot/reference date. It would be then be inconsistent with the estimation of CCF.

12. Question 12: Do institutions consider it proportionate to the risks of underestimation of CCF to perform the identification analysis and allocation procedure? If it is deemed not propor-tional, what would be an alternative approach that is still compliant with Article 182(1b) CRR?

To fulfil the requirements as set out by the draft GLs might be difficult and disproportionate to achieve in practice and seem therefore to be at odds with the goal of these GLs to reflect the low materiality into more simplified approaches.

Institutions do not consider the identification analysis and allocation procedure proportionate to the risks of underestimation of CCF, as it requires substantial resources (e.g., engineering historical relational databases with manual mapping and forensic reconciliation) for marginal accuracy gains, especially in low-materiality portfolios where pre-default restructurings are rare and there is limited evidence of systematic underestimation. This approach risks disproportionate overestimation leading to economically implausible CCF estimates (e.g., exceeding 100%), and contradicts the GL's stated goal of simplified methods reflecting CCF's lower materiality relative to PD and LGD.

Moreover, we note that under CRR III and the proposed EBA GLs, the risk of underestimation is in practice non-existent, as the framework already embeds a high degree of conservatism. This arises, for instance, from the artificial flooring of CCF to 0% in the calibration step and from the requirement to ignore terminated contracts before default (which are not considered observed EaD of zero but instead removed from the analysis). Such mechanisms structurally bias outcomes upwards, and would not be compatible with a genuine best-estimate objective.

Alternative approaches compliant with Article 182(1b) CRR, ensuring a detailed understanding of customer product mix changes, include:

• Grounding justifications for rebutting the "related facilities" presumption in risk classification and rating at the time new finance granted in order to let the bank to set out a clear criterion for differentiating what is qualifiable as “ordinary finance” from “distressed” one.
• Checks on facility articulation at reference date vis-à-vis default date with the focused control on the presence of terminated facilities and issuance of new one during the 12 months period, e.g. by checking the consistency of the overall outstanding amount at client level before and after the change.
• To the extent possible (and likely just on non-retail perimeters), assessment of historical single file.
• Incorporating proportionality via materiality thresholding: Apply the identification procedure only when the affected segment represents a material share of EAD or demonstrates historical patterns likely to cause significant CCF underestimation.
• Targeted linkage: Focus on high‑risk restructuring types (e.g. distressed restructurings, performing forbearance) where empirical evidence shows distortion risk.
• Proxy adjustment with MoC: Where linkage is infeasible, apply a documented Margin of Conservatism calibrated to observed bias in similar segments.

13. Question 13: Do you have any concerns on the proposed approach for the treatment of so-called ‘fast defaults’? In case you already apply a 12-month fixed-horizon approach, do you apply a different treatment for ‘fast defaults’ in practice, (and if so, which one)? Is the ‘fast default’ phenomenon material according to your experience? If yes, for which exposures, exposure classes or types of facilities?

We see the issue of “fast defaults” as closely interlinked with the treatment of product type transformations, as elaborated under Questions 11 and 12. In practice, when the customer–product relationship is properly defined, almost all new financing to existing clients is considered related to existing exposures at T–12, so that the occurrence of “fast defaults” becomes very limited and largely residual.

Nevertheless, when it can be demonstrated that a new facility is not related to any existing exposure at T–12, such exposure should be treated as a separate facility with a shorter reference period, and included in the risk estimation process as a “fast default”.

Given their expected immateriality under the current interpretation, a proportionate approach would be appropriate – for instance, allowing institutions to exclude fast defaults where they represent less than a defined materiality threshold (e.g. 10% of total defaults). This would balance prudential completeness with operational feasibility, while maintaining consistency with the broader framework on product transformation.

14. Question 14: Do you have any concerns on the multiple default treatment? To what extent are your current models impacted by the application of a multiple default treatment?

We have no significant concerns with the multiple default treatment, as aligning it with the LGD treatment is a logical and consistent approach, enhancing coherence between CCF and LGD estimates. The specification for files re-defaulting within less than 10 months after a cure is particularly appreciated for added clarity. For some banks, current models will require review and potential adjustments due to the shift from the ECB Guide on Internal Models (EGIM) expectations, necessitating ECB alignment with the new EBA GLs. The impact on CCF appears to be of low materiality, suggesting manageable model updates.

15. Question 15: Do you agree with the three principles for the calculation for realised CCF in the context of consumer product mix, and their implications for the cases mentioned as ex-amples? In case of disagreement, what is the materiality of the cases with unwarranted re-sults, in particular in relation with the definition of facility applied in your institution? In case of material unwarranted results, can you describe your alternative practice to this CP?

We partially agree with the three principles for calculating realized CCF in the context of consumer product mix, recognizing their conceptual merit in ensuring a consistent approach to credit risk estimation. However, their application is heavily reliant on a detailed understanding of customer product mix, which can be challenging due to data limitations and varying facility definitions, adding unnecessary complexity. A significant concern is the lack of coverage for cases involving contingent liabilities, such as Multipurpose Credit Lines (MPCLs) that can shift from revolving commitments to guarantees like financial or trade instruments. The current principles, particularly the disaggregation rule, exclude drawn guarantees from CCF realization. Yet, if these guarantees are executed and convert to on-balance items, they increase the drawn amount without a corresponding adjustment to the undrawn limit, potentially leading to CCF estimates exceeding 100%, which is economically unfeasible.

Let the following illustrative example for better understanding: 

• Limit of MPCL = 100 at T-12
• Drawn at T-12
  • 50 Cash
  • 20 Financial Guarantees (for sake of simplicity falling under Full Risk attracting SA-CCF = 100% according to Annex I CRR3)
• Undrawn at T-12 = 30

Within T-12 and T the 20 Guarantees are executed and converted to cash thus contributing to the 110 OBS at time of default T. 

The following chart shows the calculation of the CCF1 and consequent EAD1 in application according to the current frame of the EBA GL, whereas CCF2/EAD2 reports, in our opinion, the alternative calculation. Specifically under CCF1 the 20 guarantees shall be excluded from the denominator (that would be equal to 30 undrawn amount despite the 20 would be still an off-balance item at that moment) but in the numerator the 20 of drawn guarantees meantime executed will contribute to the 110 of OBS at default time T leading to 200% CCF ((110-50)/30) and, in the stylized example of application, to EAD1 of 130 (=200%*30+50+20*100%)

However, and also pursuant to principle c. of par. 79 of Section BR of EBA paper applied to the contingent liabilities instead of term loans, the Drawn Guarantees amount, in case of conversion, shall be excluded from the 110 (i.e. from the numerator) otherwise we are considering in the CCF estimation a non-revolving item already present at time T and we would have a double counting in application (20 Delta EAD in this stylized example) stemming from inflated CCF combined with adoption of SA-CCF on contingent liabilities in the application. Specifically, the CCF2 should be equal to 133% ((110-20-50)/30) with and EAD2 in application of 110 (133%*30+50+20*100%). 

MPCLs with contingent liabilities represent a notable portfolio segment, with impact varying based on whether a facility is defined as a single contract or a set of related contracts. This can skew realized CCF calculations, especially in portfolios with mixed products. Additional concerns include the conservative bias of the allocation mechanism, which may overestimate risk for limit increases or new loans approved through explicit credit decisions, particularly where such changes predate default. This can distort Long-Run Average (LRA) CCF, especially in low-volume IRB-CCF segments, and lead to unreasonable results in certain cases. The emphasis on mixed products with sublimits also seems disproportionate for portfolios where these are rare, potentially reducing client flexibility (e.g., for SMEs) or increasing exposure, contrary to their risk-mitigating purpose. Data availability assumptions further complicate adherence, particularly for legacy portfolios.

We seek clarity on the treatment of other fully off-balance sheet products, such as guarantees . Guarantees that are revolving should be modelled consistently with other revolving products, ensuring that conversions to on-balance items are correctly reflected in realised CCF. 

Similarly, guidance on the treatment of term loan repayment behaviour is needed, as inappropriate handling could bias CCF estimates. We welcome illustrative examples from the EBA to ensure consistent application across institutions. 

This approach ensures compliance with demonstrating product mix impact while addressing unwarranted results across affected segments.

In general, we suggest that EBA does not impose the allocation methods. It should be up to the banks to ensure consistency between CCF numerator and CCF denominator and detail in procedures the allocation methodologies used in case of product mix or product transformation. Such allocation methods should be in line with the bank's granted process and monitoring of facilities.

16. Question 16: Are there any concerns related to the allocation mechanism described in these GL?

The approach remains overly generic and prone of interpretations in a conservative drift in the context of Internal Model Investigations. In particular, the case IX would require further elaboration in the approach of allocation: indeed, looking to that Example the logical conclusion would be that out of the 200 outstanding of the Instruments X-A 150 are related to Terms Loans I-A and 50 to Revolving Instruments II-A (thus leading to a CCF = 0% that in this specific example would make absolutely sense). Therefore, the approach based on the outstanding/drawn amount of the consolidated facilities seems to be most logical approach.

Furthermore, regarding “Example box 3 – consumer product mix,” Case I and Case II refer to instances when “new finance” is granted to the obligor, either by extending the limit of an existing revolving commitment or by adding a new revolving contract to the client facilities’ pool. Conceptually, consolidating such new finance within a single calculated CCF may be sound, particularly when it relates to the same facility. However, if the denominator only accounts for the original facility and excludes new financing, the resulting CCF ratio can be artificially inflated, overstating risk and distorting the LRA-CCF. This is evident in examples where the calculation yields a CCF of 200%. Combined with the floor imposed by Article 182 CRR, this produces a double hyper-prudential effect, pushing estimated CCFs further from their actual realizations. Outliers are created, inflating the right tail of the CCF distribution and restricting the left tail, while potentially discouraging banks from granting additional limits even to high-quality clients. To address this, the CCF formula should also incorporate new finance into the denominator.

17. Question 17: Where credit lines are kept open even if the facility is in default, the alterna-tive option described in this consultation box could lead to high realised CCF values. Is this a relevant element for your institution and if yes, why and how material are these cases with-in the scope of IRB-CCF models?

Yes, this issue is relevant and material. Where credit lines remain open after default, the consultation’s “alternative option” (adding total debit turnover to OAD, sometimes without netting) produces extremely high realised CCFs that distort conversion risk and destabilise estimation. 

The preferred approach, as outlined in the draft GLs, is to calculate additional drawings as the maximum drawn amount during default, discounted to the date of default, which avoids spurious CCF inflation while maintaining coherence with LGD measurement.

Some flexibility may be warranted to reflect portfolio diversity and alignment with LGD, such as short-horizon netting of drawings and repayments or punctual recognition of drawings when already applied in LGD, provided that consistency between CCF and LGD is maintained and model validation standards are met. All alternative implementations should be subject to governance safeguards, including explicit linkage to LGD definitions, back-testing to avoid systematic bias, and evidence of stable realised CCF distributions with appropriate risk-driver differentiation.

18. Question 18: In case of multiple defaults, the CCF might also be driven by drawings while the obligor was in its default probation period or in the dependence period between the merged defaults. Do you expect this to be material for your CCF models?

This effect may be relevant considering that in the course of the probation / dependence period the customer is factually like-performing. Anyway, even in presence of higher drawings it is expected to have also higher recoveries from repayment on LGD side as such it is valid what reported in the previous Question.

19. Question 19: Do you see any unwarranted consequences of the proposed approach for in-corporating additional drawings after default? In particular, in order to maintain consistency between the realised CCF calculation and the calculation of the denominator of the realised LGD as described in paragraph 140 of the GL PD and LGD, would this require a redevelop-ment of your LGD models?

Yes, there could be consequences, but their materiality depends on the current modelling framework. For institutions applying punctual recognition of drawings in line with EGIM supervisory steering, moving to a single prescribed approach would require significant redevelopment of LGD models. Because the definition of the target variable would change, this could affect not only risk quantification but also risk differentiation, particularly where exposure size is a key driver. 

To avoid unnecessary model disruption, it would be preferable to retain a spectrum of admissible approaches, provided consistency between CCF and LGD is ensured. This would allow institutions to align with the proposed CCF framework while maintaining coherence with existing LGD recovery rules.

In this context, we recognise the relevance of the Maximum Exposure at Default (Max EAD) concept and agree that consistency with it should be preserved. However, we recommend clarifying the definition of drawn amounts and permitting separation of pre- and post-default drawings to achieve alignment without undue redevelopment or operational burden.

At any rate, approach in B&R Recital 84 should present definitive approach towards which new  LGD and CCF models should converge in the long-run perspective, i.e., that means that any redevelopment of the LGD model should conform to proposed CCF approach in Recital 84.

In sum, the proposed approach risks unnecessary redevelopment and operational inconsistency. We recommend clarifying the definition of drawn amounts, permitting separation of pre- and post-default drawings, and allowing alignment with existing LGD recovery rules to achieve consistency without undue burden.

20. Question 20: Do you think that the relative threshold is an appropriate approach to restrict the use of the alternative CCF approach for those facilities in the region of instability? Do you think it is appropriate to define a single relative threshold per rating system or are there circumstances where multiple relative thresholds would be warranted? Do you see a need to use an absolute threshold in addition to the relative thresholds?

The absolute threshold is particularly relevant in cases where very small undrawn amounts are observed at the reference date. Such situations often lead to disproportionately high realised CCFs, which are typically the primary focus of the alternative CCF methodology. For this reason, the introduction of an absolute threshold—alongside a relative one—is especially valuable for this application. Relying solely on relative thresholds risks capturing a broader range of commitments within the area of statistical instability, including exposures whose CCFs are not problematic and could meaningfully contribute to the core IRB CCF estimation. In other words, the ROI needs to be broadened by means of the relative threshold in order to achieve the same level of variability containment that would be ensured if an absolute threshold were also applied. 

Illustrative examples could be provided to show that, when absolute thresholds are not allowed, the proportion of observations subject to the alternative CCF treatment increases significantly relative to the total. This outcome underscores the importance of retaining absolute thresholds to avoid the undue inclusion of exposures that do not present issues for IRB-CCF estimation.

With respect to the observed CCFs in the ROI, we propose that the EBA clarifies that fundamental statistical techniques – such as the treatment of outliers using percentile or absolute caps – are permissible to enhance the robustness of model estimation. This proposal considers paragraph 322(c) of the 2025 EGIM, which is typically interpreted by assessment teams of the Supervisory Authority as prohibiting statistical outlier treatment for risk quantification. Such an interpretation conflicts with the overarching objective of achieving robust and reliable risk-quantification levels that are unbiased by extreme outliers.

21. Question 21: Do you consider the guidance sufficiently clear in relation to the requirement for institutions to set up a policy to define a threshold value?

Broadly speaking, yes.

22. Question 22: Do you consider it appropriate to set a prescribed level or range for the de-fined threshold, and if so, what would be an appropriate level for the threshold? In case an absolute threshold is warranted, what would be an appropriate prescribed level for an abso-lute threshold?

We deem more appropriate not to fix a predefined threshold but rather elaborating harmonic regulatory guidelines to be then applied for calibrating on the local portfolio of the institution the most representative threshold.

23. Question 23: Do you think that, for the facilities in the region of instability, and/or for fully drawn revolving commitments, a single approach should be prescribed (e.g. one of the ap-proaches above defined in the Basel III framework), or that more flexibility is necessary for institutions to use different approaches they deem most appropriate for these facilities?

We recognise the importance of ensuring convergence and reducing interpretative divergence in the region of instability. A single prescribed approach could therefore be acceptable, provided it is method (c.), which is most consistent with the standard CCF logic and differentiates facilities more appropriately by utilisation rate. By contrast, method (b.) is conceptually weak, as it measures increases in exposure against the drawn amount, producing unintuitive results especially where the drawn balance is small.

At the same time, calibration should remain segmented at product level to reflect material differences in conversion behaviour—trade guarantees typically show very low CCFs, while consumer credit lines behave quite differently. A “one size fits all” calibration risks misrepresenting risk across portfolios.

24. Question 24: If such flexibility is indeed warranted, what is the technical argumentation why prescribing a single alternative approach for these facilities is not suitable? Which cases or which types of revolving commitments could not be modelled under the approaches pre-scribed? Are there types of revolving commitments that could not be modelled by any of the approaches described in the Basel III framework?

It would be relevant to allow use of other Basel approaches or adaptation of such approaches. As a reminder, Basel approaches consist of:

• Limit Factor approach: the predicted balance at default is expressed as a percentage of the total limit that is available to the obligor under the terms and conditions of a credit facility
• Balance Factor approach: the predicted balance at default is expressed as a percentage of the current balance that has been drawn down under a credit facility
• Additional Utilisation Factor approach: the predicted additional drawings in the lead-up to default are expressed as a percentage of the total limit that is available to the obligor under the terms and conditions of a credit facility 

Each Basel approach may raise the following issues in the calculation:

• The Basel approaches were not written in the context of a 12-month fixed horizon approach. However, one strong assumption is the reference date especially for the denominator in the calculation of each Factor. Such reference date is 12 months before default, thus when applying the Basel approaches, the denominator is the drawn/balance amount or total limit at the reference date. However drawn/balance amounts or total limit amounts could evolve before 12 months before default and the default (thus with a mismatch with amounts in the numerator). Such asymmetry between the numerator and the denominator is structurally an issue in all Basel approaches
• Low total limits at reference date could create extreme values in the calculation under Limit Factor and Additional Utilisation Factor approaches, as the denominator is rather low
• The Balance Factor approach could create instability issues due the denominator being the drawn/balance amount at reference date, which could be disconnected with the drawn/balance amount at default date.

We describe below relevant adaptations of Basel approaches that banks could think of, to illustrate the issues encountered (rather than supporting them as a prescriptive approach):

• In the Additional Utilisation Factor approach for instance, we understand that the calculation of an Additional Utilisation Factor is necessary (being the predicted additional drawings in the lead-up to default calculated as the difference in the drawn amount between default date and reference date, divided by the total limit at reference date). However, to circumvent issues related to low limits, we could express directly the EAD as the drawn amount at reference date + an additional drawing factor. Thus, the EAD will be an equivalent to what appears in the CCF numerator.
• Another way to circumvent the difficulties could be to simplify significantly the approach where the scope is rather limited. In this case, we can think of expressing the EAD as the drawn amount at reference date + X amount. The X amount could be either calibrated or a fixed value.

As there could be different ways to best estimate the calculation depending on the cases as illustrated above, we would favour the EBA providing sufficient flexibility (use of Basel approaches or adaptations) in the calculation approaches in the Region of Instability so that banks can take the most relevant approach for their portfolios.

25. Question 25: Which of the three approaches described in the Basel III framework is pre-ferred in case a single approach would be prescribed?

We support convergence across institutions to ensure consistency and comparability, provided the prescribed approach reflects the most risk-appropriate logic — namely, method (c) under the Basel framework. However, given the structural limitations and data asymmetries inherent in all Basel approaches, flexibility should be maintained to allow institutions to adapt or refine the method most suitable to their portfolios and products. Calibration should remain product-specific to ensure risk sensitivity. (please, refer to answers to Q23 and Q24).

26. Question 26: For the purpose of the long run average calculation, are there any situations where such intermediate exposure weighted averaging at obligor level would lead to a dif-ferent outcome (that is unbiased) with regard to the CCF estimation? How material is this for your portfolio?

Situations can arise where obligor-level averaging leads to different outcomes than facility-level averaging, particularly when multiple facilities of the same obligor fall within the same pool or grade. In practice, we consider facility-weighted averages sufficient and do not see a strict need for intermediate obligor-level steps.

At the same time, Article 182(1)(a) CRR requires the long-run average CCF to be a default-weighted measure, which implies that aggregation should give equal weight to each default at obligor level. This creates a tension with the draft EBA Guidelines, which stipulate that institutions should not use averages across subsets of observations, including facilities of the same obligor, while the ECB EGIM (July 2025, par. 322(d)) allows yearly averaging.

To avoid uncertainty and divergent practices, clarification is needed on how institutions should reconcile the CRR requirement, the draft EBA GLs, and the EGIM when calculating the long-run average CCF.

27. Question 27: Do you have any comments on the condition set to use the simple approach to estimate additional drawings after default. Do you consider that the simple approach is also relevant for retail portfolios?

A simple approach can be a pragmatic expedient but should be allowed only under strict, evidence-based conditions. In particular:

• When it may be used. The simple approach is acceptable where the population of currently-defaulted facilities with a remaining undrawn amount is immaterial for the resulting LRA CCF. “Immaterial” should be assessed by the potential impact on the LRA CCF, not by an absolute procedural label: either a very small number of cases or a very small available headroom (i.e. potential drawdown is economically immaterial) are acceptable bases for the simple approach.
• Retail portfolios. For most retail portfolios (where historical data are plentiful) the simple approach will typically not be necessary. It may nevertheless be used for retail segments only where the same materiality test (potential impact on LRA CCF) is satisfied.
• Required pre-checks (to address the concern in Input C). Before adopting the simple approach an institution should: (i) run the modelling approach as an exploratory test (or equivalent sensitivity analysis) to confirm that using the simple approach does not materially change the LRA CCF; and (ii) document why the simple approach is an appropriate simplification for that segment. This addresses the point that the simple approach cannot be a shortcut that avoids the modelling work needed to demonstrate its acceptability.
• Operational and governance safeguards. Use of the simple approach should be subject to clear policy, periodic review, back-testing and reversion triggers (e.g. if the share of affected exposures or the sensitivity of LRA CCF increases materially). The institution must retain the ability and data to move to the full modelling approach if conditions change.
• Design of the materiality test. The test should be based on quantitative sensitivity (how much the LRA CCF would change if unresolved cases were imputed as resolved) and on the economic headroom available to draw; it should be proportionate and documented so supervisors can replicate the assessment.

28. Question 28: It was considered that requiring institutions to exclude unresolved cases from the long run average CCF, if their realised CCF is lower than the LRA of the corresponding fa-cility grade, could be seen as too conservative. Do you have any comments on this treat-ment introduced in the simple approach? Do you have specific examples when this treat-ment would not be appropriate?

With reference to recital 118, we contend that the proposed methodology is overly simplistic. It fails to adequately consider the underlying reasons for the lower realised CCF observed for unresolved cases when compared to the resolved ones.

For example, in the presence of a decreasing trend in realised CCF (e.g., resulting from an improved recovery process or a more favourable macroeconomic scenario), it is logical that incomplete workout positions would predominantly feature in the most recent part of the historical series, which inherently displays lower realised CCFs. In our opinion, it is incorrect to disregard these data within the LRA CCF calculation by applying the simple approach, particularly when the lower realised CCFs are attributable to distinct economic factors not related to the additional drawing topic.

A more appropriate, yet still straightforward, alternative could be a comparative analysis of closed recovery processes. This would involve comparing the realised CCF with and without considering additional drawings. The resulting differential could then be applied to the realised CCF of unresolved cases to ensure a more accurate estimation.

29. Question 29: Do you have any comments on the modelling approach to estimate additional drawings after default for unresolved cases?

The modelling approach seems to be aligned with the principles applicable to the estimation of future costs and recoveries for the purpose of LGD estimation. 

Implementation of the EBA approach raises operational issues. Indeed, our current LGD and CCF models are currently developed with possible misalignement between LGD and CCF calculation granularities. There would be a first issue of mapping if we were to reuse extrapolation from LGD model line by line. Moreover, as CCF grades are not defined on the same basis that LGD, the mapping for the maximum recovery period could be difficult when it is calibrated by LGD grades for example. One additional difficulty is that the LGD is developed for revolving and non-revolving lines which complexifies the issue. If we were to use extrapolated AD estimates both for CCF and LGD as requested by EBA, this could imply redeveloping LGD models

It is worth noting that there might be situations where the drawing period for a product is shorter than the maximum length of the recovery process; in this case, we would expect this to be appropriately reflected in the estimation methodology. 

Furthermore, we would also expect that drawings should be estimated from one reference date to the next considering closed as well as incomplete workouts available until the next reference date. Paragraph 96(a) seems to suggest that the estimation should be based on closed cases only.

30. Question 30: Do you have any concerns with the requirement to use as a maximum drawing period the maximum recovery period set for LGD?

Members are divided on if the MRP used on LGD side representative of a maximum drawing period. Indeed, the MRP is based on historical observations of exposures that have migrated to Liquidation status where the credit lines are terminated and no additional drawings is possible. From a theoretical standpoint a dedicated maximum drawings period estimation would be needed based on the realizations of drawings observable only on period before the Liquidation status (when the customer is still in a going concern situation). 

This could exacerbate the redevelopment risks noted in Q19, particularly for LGD models reliant on historical Liquidation data, and contrasts with the controlled post-default drawings in Q17.

31. Question 31: For CCF estimation, do you use estimation methods that incorporate portfolio-level-calibration of the estimates? What are the main reasons to use a calibration at a level that is higher than the grade-level calibration?

Calibration should occur at a product-type level (e.g., trade guarantees, overdrafts), not blindly at aggregated levels. Given the wide spectrum of drawing behaviors across product categories, segmented calibration ensures both accuracy and proportionality.

In this context, we propose that EBA clarifies that off-balance sheet claims that convert into on-balance items (e.g., executed guarantees) do not impact the  drawn amount when calculating the realised CCFs.

In some cases, for certain portfolios – typically non-retail ones – a continuous scale and not a grade-level scale is used. It is proposed that the GL specify how the references to “grade-level” calibration must be interpreted when continuous-scale model predictions are used.

There can be several reasons for calibration at levels higher than the grade level: 

1. Limited observations per grade level, e.g. especially for non-retail portfolios.
2. Additional model complexity caused by too granular segmentation (if calibration segments are defined at a grade-level) because of the requirement to estimate downturn, appropriate adjustments and margins of conservatism at the calibration-segment level.
3. A too granular calibration segmentation causes the estimation error per segment to increase since this error scales proportional to the number of observations per segment.

32. Question 32: Do you have any comments on the guidance for the CCF estimation of default-ed exposures?

We have mixed views on the guidance for CCF estimation of defaulted exposures.

From a regulatory standpoint, the CRR does not introduce any distinction in exposure value between defaulted and non-defaulted exposures (Article 166). The same exposure value definition applies to both performing and defaulted assets under Articles 153 and 154, and no such differentiation exists in the calculation of Expected Loss (Article 158.5) or under the Standardised Approach. Therefore, we consider that applying performing CCFs to all defaulted exposures is fully consistent with the CRR framework and should represent the default regulatory approach.

Introducing a specific “CCF-in-default” model would add unnecessary complexity to the modelling landscape, given limited data availability and minimal incremental drawdowns post-default, especially in non-retail IRB portfolios where drawdowns are often blocked. The benefit of a dedicated model for in-default CCFs would thus be marginal, while creating proportionality and operational challenges for a narrow scope (revolving commitments only).

Nevertheless, for institutions that identify material additional drawings post-default, simplified approaches could be used to capture incremental exposure in a consistent and operationally efficient way. In such cases, the use of a Limit Factor–based approach (as in the “region of instability”) could serve as an alternative target variable, ensuring conceptual alignment with CCF’s pre-default linkage.

33. Question 33: Do you have any comments on the determination of the low share of ob-served additional drawings after default in the historical observation period relative to the observed undrawn amount at default date? Do you consider it appropriate to set a pre-scribed threshold to determine what constitutes this low share? If so, what would be an ap-propriate value for such a materiality threshold?

More than a prescribed threshold, the adoption of the simple approach could be linked to the adoption of the simple approach on open default considering that the underlying reasons are basically the same (i.e. limited materiality of the phenomenon and/or restrictive policies on additional drawings). 

34. Question 34: Are there examples where the haircut approach should be considered the most appropriate approach for estimating the downturn CCF?

While consistency with LGD downturn treatment is conceptually sound, the practical application of downturn CCF estimation should account for the limited responsiveness of CCF to macroeconomic conditions. Institutions should be permitted to justify a zero or minimal downturn adjustment where empirical evidence shows no significant increase in undrawn utilisation during downturns (e.g. by means of their proactive role in manage limits in stressed situations). The EBA should provide clear materiality thresholds or safe harbours for portfolios where drawdown behaviour is stable, to prevent unnecessary model complexity and over-conservatism.

Moreover, in determining DT period for CCF estimation, par. 146 explicitly asks for the inclusion of the “yearly default rates of the portfolio in scope of application of CCF model” in the indicators’ set as per art 2(1)(c) of CDR 2021/930. In the Background and Rationale recital 141 it is clarified that the requirement of art. 182 CRR to analyse the correlation between default frequency and magnitude of conversion factor can be covered in this way, under Downturn framework. This approach appears to be inconsistent with the original intent of the CDR on Downturn, which requires that the indicators selected to identify downturn periods reflect the evolution of the economic cycle independently of the institution’s own (default, loss or drawdown) experience. In fact, the CDR 2021/930 explicitly requires the assessment of external default rates (to which institution’s internal default experience contributes but it does not restrict the analysis to it). Moreover, the use of internal default rates solely for CCF downturn estimation could lead to misalignments in the identification of the downturn period, which should—by design—be consistent across LGD and CCF parameters, as both are meant to reflect macroeconomic or system-wide conditions. Such inconsistencies may ultimately trigger increased supervisory scrutiny and requests for alignment or justification of divergent downturn periods across risk parameters.

Regarding par. 162 – steps to calculate Reference Value - is not clear the reference to “economic loss” at point “(iii) selecting the two individual years with the highest annual realised CCF as described in (ii) as the two individual years with the highest observed economic losses”. It could be a typo, being the framework inherited from LGD. 

Question 34: Are there examples where the haircut approach should be considered the most appropriate approach for estimating the downturn CCF? 

We do not consider the haircut approach the most appropriate method for estimating downturn CCF in most cases, given the weaker link between macroeconomic factors and drawdown behavior compared to LGD. However, niche scenarios might justify its use, such as portfolios with highly stable drawing patterns—where flooring effects (per Q36) further minimize downturn sensitivity—along with a statistically significant correlation with specific economic indicators, or situations with limited data where a conservative haircut is supported by expert judgment and historical stress scenarios. Generally, extrapolation appears more suitable for downturn CCF estimation due to its adaptability to available data and behavior patterns.

35. Question 35: Do you think the add-on of 15 percentage points is adequately calibrated when the downturn impact cannot be observed nor estimated? Could you provide clear examples or reasons why this add-on should be higher or lower than 15 percentage points?

The 15 percentage point add-on for downturn CCF, when impact cannot be observed or estimated, appears broadly conservative but may not be adequately calibrated across all contexts. As a fallback for exceptional cases, it aligns with past EBA GL on PD-LGD, yet its suitability varies by portfolio. For low-baseline CCF portfolios (e.g., credit cards with strong limit management), a 15pp increase could overstate risk, while high-volatility drawdown portfolios or those with historical stress indicating sharp utilization rises may require a higher add-on. We propose a risk-sensitive approach with a minimum add-on, adjustable based on observed volatility or expert judgment. Additionally, the add-on’s arbitrariness and lack of verifiability suggest it should be tested against industry-wide realized CCF time-series data to determine an average downturn increase, ensuring a data-driven calibration.

36. Question 36: Have you observed, or do you expect a (statistically significant) correlation be-tween economic indicators and realised CCFs? If so, do you expect higher or lower levels of CCFs observed in the downturn periods compared to the rest of the cycle? Do you have pol-icies in place that restrict or, on the other hand, relax the drawing possibilities in the down-turn periods?

Yes, we may observe some statistical correlation between economic indicators and realized CCFs, with higher CCF levels typically occurring during downturn periods due to increased obligor drawdowns, though the magnitude varies by product type and may be lagged. However, this correlation can be obscured by proactive credit management, such as limit reductions or covenant enforcement, which may lead to negative CCFs floored at 0%. This flooring inflates the baseline Long-Run Average (LRA) CCF, reducing the relative increase in downturn CCF, and weakens statistical significance when testing macroeconomic relationships or quantifying downturn impact. We request regulatory clarification on flooring negative CCFs in these contexts, especially in stable-drawing portfolios where the haircut approach might apply, as discussed in Q34.

Our policies generally restrict drawing possibilities during downturns through measures like limit reductions and enhanced monitoring, aiming to mitigate utilization spikes. However, these measures may not fully offset behavioral tendencies to draw more, potentially smoothing CCF behavior over the cycle and complicating downturn identification.

Specifically, we propose that the EBA clarifies that observed CCFs should not be floored for the purposes of testing the relationship between CCFs and macroeconomic factors. Furthermore, we recommend that the relevant downturn impact be assessed by comparing unfloored downturn CCFs with LRA CCFs (which are floored). The rationale for using unfloored downturn CCFs is that applying a floor penalizes banks for implementing risk management practices aimed at reducing or limiting spending on revolving credit—particularly during economic downturns, when such practices are most critical. Effective credit risk management can influence customer behaviour, potentially leading to reduced utilization of available credit limits. This may result in lower exposure levels and, in some cases, negative CCFs. The alternate option of applying a floor to downturn CCFs would inadvertently penalize banks for demonstrating the effectiveness of their risk management capabilities. This issue has been a point of contention in recent internal model inspections, and we therefore request clarification from the EBA on this matter.

With respect to the Reference Value specified in par. 162 of the CP (that is used as a challenger value for the downturn CCF), the CP is silent whether the zero-flooring is also required for the Reference Value. This detail is relevant since the Reference Value is compared to the downturn CCFs, for which it is also unclear if zero-flooring is required (refer to the proposal in previous paragraph). To align with said proposal, it is proposed that EBA clarifies that the Reference Value must not be zero-floored.

At the same time, it should be acknowledged that in practice the statistical relationship between economic indicators and realized CCFs may be difficult to observe in a consistent and robust manner. We therefore see merit in considering pragmatic approaches that balance analytical effort with proportionality, avoiding requirements that may place undue burden on banks without materially improving risk capture.

37. Question 37: The possibility to have no downturn effect on CCF estimates is restricted to the case where observations are available during a downturn period. Which alternative meth-odologies could be used to prove the non-existence of a downturn effect on CCF estimates, in the case where no observation is available during a downturn period?

While we recognize that downturn evidence is inherently difficult to obtain and often produces inconclusive results, in cases where banks still need to demonstrate the absence of a downturn effect, the following alternative methodologies could be applied

These include: a statistical analysis of economic factors on available time series, corroborated by expert evidence and assessment of policy rules; proxy-based comparative analysis using periods with similar macroeconomic stress indicators (e.g., elevated credit spreads, GDP contraction); drawing stability assessment across non-downturn economic conditions using long-run data; macroeconomic sensitivity analysis of CCF to risk drivers (e.g., unemployment, interest rate volatility); and benchmarking with pooled datasets from similar portfolios. While European-level downturn data is often available, favouring the observed approach, these methods provide robust indirect evidence when direct observations are lacking.

As a general comment on downturn estimation, once a common downturn window is established for LGD and CCF, the selection of impact years should be coherently justified within that window. This does not imply identical impact years for LGD and CCF, but rather macro-consistent choices reflecting potential lead/lag effects and portfolio characteristics. The aim is to ensure internal coherence between the two parameters and to avoid artificial combinations unrelated to the defined downturn period.

Upload files

Name of the organization