Response to consultation on Regulatory technical standards on specialised lending exposures
d. From an industry perspective, are there any challenges or limitations you experience with the current Supervisory Slotting Criteria Approach—particularly in assessing and quantifying the risk of the exposure? Do the current assignment criteria provided in the Annexes capture all key risk dimensions effectively, or could it be enhanced by adjusting the current factors, sub factors, or sub-factor components? Please provide details, and examples where they apply, and explain the rationale for your feedback.
While the proposed LTV-related enhancements are welcome, very low LTV transactions and strong over-collateralisation are not sufficiently differentiated within the current framework. Further consideration could therefore be given to whether the resulting reduction in loss severity is adequately reflected in the slotting assessment.
Question 2. Do you have any comments on the proposed amendments to Article 1 or on the replacement of the references to ‘real estate’ by references to ‘income-producing real estate’?
ESBG understands that there is a lack of clarity regarding the treatment of exposures in the construction phase (ADC) under IPRE. The replacement of “real estate” by “IPRE” introduces uncertainty on whether:
- What type of construction-phase exposures should be included within IPRE
- A differentiated treatment is expected between construction and operational phases
Question 3. Do you have any comment on paragraph 3 that was newly introduced to Article 3 regarding the consideration of UFCP?
ESBG understands that guarantees can only be considered within specific sub-factors when explicitly mentioned in the evaluation criteria included in the Annexes; but in cases where a full guarantee exists, we believe it should be possible for it to anchor the final rating of the project.
a. Do you have any comment on paragraph 6, on the consideration of ESG risks that is newly introduced to Article 3 of the RTS?
While we support the clarification on ESG risks, additional guidance would be welcome to ensure a consistent implementation and avoid overlaps with existing assessment criteria. Due to the stated challenges, the ESG related factors, that are not covered in the model (other than e.g. market value, insurance coverage) are currently covered via a dedicated ESG override guideline.
c. Specifically for the newly specified sub-factor component on the CPI: Do you think the cut-off values for the CPI allow for a reasonable risk differentiation? If not, please provide cut-off values that would be meaningful for your specialised lending portfolios.
Corruption-related risks are already reflected through the assessment of the political, legal and operating environment. Additional clarification on the interaction between these existing criteria and the newly introduced CPI-based sub-factor would therefore be welcome to ensure a consistent and harmonised assessment.
Question 5. Documentation: Do you have any comments on the clarifications for the documentation of overrides and validation results that were introduced to Article 6 of the RTS, also in view of the supervisory best practices outlined in the Supervisory Handbook on IRB Validation (EBA/REP/2023/29)?
ESBG supports greater transparency regarding overrides and validation. The documentation requirements should remain proportionate to the supervisory nature of the SSCA. They should not gradually evolve towards the governance and validation expectations applicable to IRB rating systems. Otherwise, the operational benefits of the slotting approach would be significantly reduced.
The proposed documentation and validation expectations increasingly resemble those applied to full IRB models. This risks undermining the original purpose of the SSCA as a simplified supervisory approach for specialised lending.
In particular, institutions should not be required to establish extensive challenger models or highly granular validation frameworks where these do not materially improve risk differentiation. Validation should remain risk-based and proportionate, focusing on whether the slotting methodology consistently differentiates risk rather than reproducing the requirements applicable to statistical rating models.
The EBA should therefore clarify that supervisory expectations under the RTS remain proportionate and should not extend beyond what is necessary to ensure consistency and sound governance.
a. would these challenges remain under the proposal to remove the floor for the factor weights and the proposed amendments to the Annexes?
Yes.
The proposed removal of the minimum floor for factor weights is a welcome improvement and should enhance institutions' ability to reflect portfolio-specific risk drivers more accurately. Nevertheless, important limitations regarding risk sensitivity remain.
The current slotting methodology still relies on a largely linear aggregation of qualitative factors. This can dilute the influence of particularly important risk drivers and lead to similar supervisory categories for transactions with materially different underlying risk profiles.
Especially for specialised lending portfolios characterised by long-term stable cash flows, the methodology does not always sufficiently distinguish between genuinely low-risk and merely average transactions.
b. please specify in detail, which amendments (removal of caps, removal of the linear aggregation, structural changes or other) would, in your assessment, most effectively enhance risk sensitivity under the SSCA?
The most effective improvements would be:
- the removal of mandatory minimum factor weights, allowing institutions to better reflect transaction-specific risk characteristics;
- greater flexibility in weighting the individual factors where justified by the underlying risk profile;
- avoiding situations where linear aggregation masks particularly strong or particularly weak risk characteristics;
- further refinement of the criteria for specialised lending segments with structurally low default rates, such as infrastructure and renewable energy finance.
Overall, the revised framework should aim to improve risk sensitivity rather than increasing methodological complexity. The objective of the revised slotting criteria should not be to increase granularity for its own sake, but to achieve a meaningful improvement in the differentiation of economically distinct risk profiles.
The removal of factor-weight floors is a positive development although the practical implementation of this flexibility is unclear. Some factors do not contribute to risk differentiation in certain portfolios, and it is case-dependent. In this regard, dynamic weighting could be more appropriate for SSCA.
More fundamentally, the objective of the revised slotting criteria should not be to increase granularity as an end in itself, but to achieve a meaningful improvement in risk sensitivity. Additional methodological complexity should only be introduced where it demonstrably improves the differentiation of economically distinct risk profiles.
ESBG furthermore considers important to clarify and align Article 3(4) in the light of the removal of the factor-weight floor in Article 2(2). Currently, Article 3(3) practically (and in the interpretation of several competent authorities) requires using all sub-factors and sub-factor components such that they are relevant in the overall slot assignment, which implies strongly positive weights. This is because of the wording “exceptionally” and “…is not relevant for all specialised lending exposures belonging to a certain type of specialised lending exposures…”, the latter being practically impossible to verify if taken literally. Particularly, this wording of Article 3(4) for the sub-factor and sub-factor components is much stronger than the wording of Article 2(2) for factors, where even a 0% weight is possible if empirical evidence can be provided. At the same time, more flexibility for sub-factor and sub-factor components should be expected than for factors. Consequently, ESBG members propose to align Article 3(4) by offering the additional possibility to have a lower weight for sub-factors and sub-factor components in case of empirical evidence. In this context, adding the following sentences to the existing Article 3(4) is proposed: “Where institutions can provide empirical evidence for determining the weight of a sub-factor or sub-factor component for a type of specialised lending exposures, they may apply this weight for any of the specialised lending exposures belonging to that type. Specifically, where institutions can provide empirical evidence demonstrating that a sub-factor or sub-factor component does not contribute to the risk differentiation for a type of specialised lending exposures, they may decide not to apply that sub-factor or sub-factor component for any of the specialised lending exposures belonging to that type.”
a. on the newly introduced sentence to Article 3(1) of the RTS to clarify the attribution of categories to sub-factors or sub-factor components,
ESBG generally welcomes the additional clarification of factor descriptions and category allocation, as these should improve supervisory consistency.
However, category 4 should remain an exception rather than becoming the default outcome whenever information is temporarily unavailable.
Institutions frequently encounter situations where certain information cannot immediately be obtained although the overall credit quality is well understood. Automatically assigning the weakest category in such situations may create overly conservative outcomes that are unrelated to the actual credit risk.
The RTS should therefore explicitly distinguish between temporary information gaps and genuine uncertainty regarding the borrower's risk profile.
The treatment of Category 4 as a residual category and automatic penalisation in case of missing data may result in potential over-conservatism and a limited ability to incorporate expert judgement. More clarity on what constitutes “insufficient information” would be appreciated as well as some expectation to allow the use of external references or proxies as well as expert judgement where appropriate when no detailed/specific data for a project is available.
c. to the implementation of category 4 as residual category, as well as to the quantitative threshold introduced for the assessment of the DSCR as criterion for category 4, in particular on the need of providing further clarifications on the treatment of missing information and its documentation?
Regarding the threshold for DSCR, ESBG notes that the proposed threshold of 1.15 for Category 4 may not always reflect the underlying risk profile. In some cases, a DSCR below 1.15 may still be consistent with investment-grade risk, depending on the characteristics of the project. In this context, a fixed threshold may constrain the slotting evaluation. ESBG therefore considers that the current criterion in Delegated Regulation (EU) 2021/598, which takes into account the level of project risk, may be more appropriate.
a. Do you have any comments on the specification of the new sub-factors and the amendments introduced to the sub-factors and sub-factor components? Do you expect an improvement in risk differentiation by the specification of new sub-factors and sub-factor components, and do you expect material model changes introduced by newly specified sub-factors or sub-factor components (related to ESG aspects or other)? Specifically for CF, given the current low usage of the CF criteria: Do you support the specification of additional sub-factors for CF or would it be more commensurate to maintain the current criteria?
ESBG supports the introduction of additional sub-factors only where they demonstrably improve risk differentiation.
The objective of the revised slotting criteria should not be to increase granularity as an end in itself, but to achieve a meaningful improvement in risk sensitivity. Additional complexity should be introduced only where it demonstrably enhances the differentiation between economically distinct risk profiles.
Additional qualitative criteria inevitably increase operational complexity. Therefore, every new sub-factor should be supported by empirical evidence that it contributes to a better assessment of credit risk.
In particular:
- ESG-related sub-factors should only be introduced where sufficient data and empirical evidence exist demonstrating their relevance for default risk.
- Reducing financial ratios to the DSCR alone may not sufficiently capture the financial strength of all specialised lending transactions. Depending on the transaction type, additional indicators may continue to provide relevant information.
- Loan-to-value concepts should remain fully consistent with existing CRR valuation concepts to avoid parallel valuation methodologies.
- The proposed market-price-volatility factor should be limited to asset classes where short-term price movements are genuinely relevant risk drivers. For many infrastructure and project finance transactions, long-term contractual cash flows rather than short-term market prices determine repayment capacity.
Overall, ESBG supports greater consistency but recommends introducing any additional sub-factor only where it materially improves risk differentiation without unnecessarily increasing operational complexity.
Any introduction of additional sub-factors should therefore be supported by robust empirical evidence demonstrating that they improve predictive performance. Otherwise, additional criteria risk increasing implementation costs while providing only limited supervisory benefit.
ESBG would like to highlight the importance of maintaining stable factor, sub-factor and sub-factor component definitions. The introduction of new or the adjustment of existing sub-factors and sub-factor components should generally be considered carefully as data on these elements have yet to be gathered by institutions to determine the respective weights. It can therefore be expected that, for an extended period of time, these changed sub-factors and sub-factor components are collected but do not play an empirical role and will, therefore, not be effectively applied. ESBG would appreciate further clarification on this point being reflected in the RTS.
b. Specifically for the clarifications to the DSCR for the sub-factor ‘financial ratios’, do you have any concerns about reducing the financial ratios to consider just to the DSCR? Are there other products than operational leases, where the DSCR does not allow unbiased risk assessment?
The exclusive reliance on DSCR may result in loss of relevant information (e.g. we believe LLCR in mini-perm structures is more adequate in IPRE or OF transactions).
While the DSCR is a well-established metric, its sole use may not always fully reflect the risk profile of certain specialised lending transactions. Allowing the consideration of additional financial metrics, such as Project Life Coverage Ratio (PLCR) or other life-of-project coverage measures, where appropriate could enhance the risk sensitivity of the assessment.
c. Specifically for the sub-factors that refer to the loan-to-value, could you please provide insights in inhowfar the determination of the loan-to-value used is aligned with the exposures-to-value according to Article 124(6) CRR?
Regarding the “resale value vs debt” sub-factor, ESBG notes that it may partially overlap with existing metrics such as the LTV/advance ratio, which could limit its additional contribution to the assessment framework. At the same time, we consider that the current set of factors, including design and location, may not always be sufficient to fully capture the value of the underlying asset.
d. For the newly specified sub-factor market price volatility, how could the different levels of volatility (low/ moderate/ high) be further specified? For IPRE and OF, is the short-term volatility of market prices a risk driver that is frequently considered in practice as additional information when assessing the assignment criteria? Do you support the specification of a new sub-factor or sub-factor component to capture short-term volatility of market prices for IPRE or OF and if so, why?
Similarly, regarding the Cash-flow Predictability (IPRE) factor, ESBG notes that there may be limited differentiation between strong private tenants and public entities, as both would generally lead to a strong evaluation. In this context, it is not clear that this distinction would provide significant additional value. ESBG also notes that there is limited clarity on how yield should be calculated and benchmarked, which may result in methodological inconsistencies across institutions.
Regarding the remaining proposed changes in factor/sub-factor criteria, the proposed amendments are reasonable and can improve the risk differentiation on the transactions under the SSCA approach.