The self-confessed objective of the guidelines is to ensure that newly originated loans are of good credit quality and thus contribute to a more stable and resilient EU financial system. While this goal is easy to accept on a general level, many if the individual suggestions in the guidelines are problematic, especially if applied to credit facilities granted to SMEs and consumers.
The guidelines should only apply to loans granted after their entry into force. If the guidelines were to cover changes to credits originating before the application date, institutions can be met with excessive data collection challenges. Moreover, the review of an existing credit facility should not cause the facility in question to fall within the scope of the guidelines.
While the principle of proportionality is mentioned several times, we feel it is not adequately followed in several individual provisions included in the guidelines. These concerns are more specifically addressed in answers to the questions below. In general, many of the suggested provisions are recommendable for larger corporate customers and exposures, but excessively exhaustive for SMEs and, in some cases, for consumers as well.
The applicability of the proportionality principle to consumer lending is rather unclear. It is stated in background section of the guidelines (Proportionality and implementation, paragraph 15) that “consumer protection framework should be applied regardless of the size and complexity of the institutions or the loan.” This formulation leaves it open whether e.g. the data mentioned in Annex 2 should be collected for all consumer loans regardless of type and risk. As the objective of the guidelines is primarily macroeconomic, this should not be the case.
Several definitions given in the guidelines seem to differ somewhat from the ones used for the same concepts in existing regulation. Apart from the terminology connected to sustainable finance (see question 3 for details), this applies especially to the definitions of CRE and RRE, which are not aligned with the text of CRR.
As a rule, the guidelines should not automatically apply to credit facilities outside the banking union, as this would hamper banks’ efforts to compete with local institutions on a level playing field in third countries.
The contents of the guidelines necessitate significant changes in the day-to-day operations of banks, not least in IT systems and staff training. Given the wide scope of the guidelines, the suggested application date approaches too soon. It would be advisable not to apply the new requirements before the end of the year 2021.
The compatibility of the Sections with existing and other upcoming regulation should be ensured. Banks will have to comply to further requirements in the future, when the EBA revises its technical standards for risk management to include ESG factors, as mandated in the CRR. In addition, EBF and UNEP-FI will assess the applicability of the EU taxonomy for sustainable activities to the banking operations. The functions under assessment will include loan portfolios and an analysis of whether changes are necessary. The objective is to facilitate the application of the EU taxonomy for banks which would like to apply it to their activities on a voluntary basis.
In the name of coherence, ESG-related criteria should be referred to as “sustainable” rather than “green”, as sustainability is a wider concept. The definition of “green lending” in the guidelines is acceptable but could be branched out as suggested above.
The requirements for green lending policies set out in paragraph 49 would excessively increase the workload connected to green credit facilities, effectively excluding many SMEs from the scope of these products due to challenges in data collection. Any requirements should be realistically set for large numbers of sustainable credit facilities (e.g. hundreds of thousands a year), to each of which the requirements apply.
The definitions of “transition risk” and “physical risk” in paragraphs 52 and 53 should be checked to ensure their compliance with the ones used by the Task Force on Climate-Related Financial Disclosures.
For technology-enabled innovation, the guidelines should remain consistent with other related initiatives, such as the Commission’s ethics guidelines for trustworthy AI.
The suggested data requirements can in many cases be unattainable without state-sponsored or otherwise comprehensive data collection frameworks, which currently are not available in all countries. Examples include credit and income registers open to banks for creditworthiness assessment purposes.
To avoid any eventual regulatory conflict or overlap, the contents of the guidelines should be checked against the ECB’s reporting instructions on credit underwriting data collection.
We support the general requirements set in the guidelines, such as clarity and documentation. For the sake of just clarity, procedures related to anti-money laundering and counter-terrorist financing should mainly be addressed in guidelines dedicated to these purposes, chiefly, the ESA Risk Factor Guidelines. In addition, some of the more detailed suggestions could result in overtly rigid and complicated procedures replacing well-functioning current practices. These include at least the following:
Paragraph 59: There is no need to limit the time period for delegated decision-making bodies or the number of delegated approvals. If the bodies follow the bank’s internal risk policies as is expected from them, risks are not likely to increase despite the absence of the suggested limitations. Eventual problems connected to cases of malpractice can be addressed according to existing regulation and internal procedures. In smaller institutions, any such limitations will be rendered useless by the small amount of staff.
Paragraphs 62 and 63: When delegating credit decision making powers, institutions do consider the expertise and experience of the individual employee or management body member as required in point 62. When these requirements are appropriately followed, the principle of proportionality should allow institutions to independently determine “small” and “non-complex” facilities, as these are relative concepts depending on e.g. the size and clientele of the institution.
Point 82: For some credit facilities, the variable remuneration is very difficult to disperse evenly for the duration of the credit, which, in case of mortgages, can be decades.
Creditworthiness assessment should be mainly carried out the first line of defence. The chance to consult risk office for second opinion is useful but should not be a mandatory part of the assessment.
The principle of proportionality, while mentioned in paragraph 86, could be highlighted in connection to other requirements as well. The list of required types of information in Annex 2 is partly over-comprehensive for smaller consumer loans. In some countries it is not customary to collect verified data e.g. on the exact purpose of the loan or financial commitments for non-complex small credits, such as cards and revolving credits. In addition, some data, such as third-party information on financial liabilities, is currently not available in all countries.
As far as professional borrowers are concerned, simple business plans and financial projections for small and medium-sized enterprises should suffice (paragraph 93 e-f). In general, few SMEs can produce these in any detail, which we suggest could be considered e.g. in the paragraphs concerned with SME specificities in Section 5.2.
The rules set out for sensivity analysis in paragraphs 101, 114 and 121 should be amended to include references to the proportionality principle, as they are disproportionately heavy for smaller exposures.
If interpreted rigidly, paragraph 125 could prove too restrictive in cases where a sum equivalent to the loan is deposited in and pledged to the institution. It could also cause trouble for vehicle finance in some countries, where loans granted for such purposes are comparatively small and the vehicle is used as a collateral for the loan.
Of the ratios mentioned in paragraph 99, loan to income and debt to income are the most useful ones.
The mention of “reasonability” in connection to the verification of the customer’s repayment ability in paragraph 103 is very recommendable, especially regarding future events. The borrower’s prospects cannot be predicted with certainty. However, the requirement not to rely on expected increase in the borrower’s income (paragraph 108) might be too stern for student borrowers about to graduate.
The heavy use of the term “at least” seems to conflict with the demands for proportionality at certain points. For example, the requirements set out in paragraph 126 are necessary for larger customers and exposures, but too burdensome for SMEs and limited exposures. The same is true for paragraph 132. We recommend that the text is modified to allow greater flexibility.
Creditworthiness assessment procedures for professional borrowers should reflect the great differences in size, field of business and resources between these customers. For instance, the assessment of ESG risks as set out in Paragraph 130 should not be obligatory for smallest professional customers. The principle of proportionality is especially important as far as sensitivity analysis is concerned, as stated above in connection with question 7. If applied categorically, the requirements in Paragraphs 144-146 would compel institutions to turn down an overwhelming majority of otherwise sound applications.
For real estate development, the nature of such projects makes it practically impossible for the bank to collect much background information on all parties taking part in the project, as e.g. sub-contractual details referred to in Paragraph 166 might not be entirely agreed on when the decision to finance the project is preliminarily made. Moreover, institutions should be able to entrust on-site visits chiefly to experts in the field of building (Paragraph 169).
The scope is very wide, possibly leading to problems mentioned in our answers to questions 1 and 2 of the consultation. These worries can nonetheless be in part alleviated by emphasizing the principle of proportionality in selected paragraphs, as suggested in our answers to questions 6-8.
While the requirements themselves are mostly reasonable, they might be unattainable within the suggested transition period.
In general, we value the EBA’s recognition of statistical models as an important toll of collateral valuation, as they often produce a better, more predictable and cost-effective outcome than valuation by staff or by an external expert.
We hope, however, that this approach would extend to the valuation of immovable property as well, as valuers nonetheless employ these models even currently, especially in desktop evaluation. The currently proposed requirements for on site evaluation of immovable property are too rigid, as in many cases there is good data available on the collateral even without a visit.
Paragraph 198: Neither law nor market practice demand an indemnity insurance from external valuers in all countries. While it can be good for a valuer to have one, this could substantially increase the cost of valuation.
If there is a recognized external framework or body authorizing the expertise of external valuers for certain types of collateral, the institution should not have an obligation to establish a separate panel (Paragraph 203).
The requirements for monitoring and revaluation set out in paragraphs 207-213 would necessitate comprehensive changes to evaluation processes, which only recently have been updated to comply with the recent EBA NPL Guidelines. A new round of updates is very difficult to complete within the suggested timeframe. Additionally, the limitations for the use of desktop revaluation proposed in paragraph 213 would lead to increased cost without much added value. The same applies to the rotation suggested in paragraph 214.
The framework is excessive and will in many cases lead to additional reporting without added value. It also could curtail disproportionately institutions’ freedom to determine their risk appetite.