Primary tabs

European Banking Federation

• The guidelines should apply only to newly originated loans, and not to loans existing before the application date. The regular credit review of a deal should not trigger any of the new requirements. Complying with the requirements regarding the collection of information is operationally unachievable for the stock of operations.
The cost-benefit analysis decision on this is particularly confusing as it says that these guidelines apply to some existing loans (renegotiated and reviewed as in page 11), but it also says that they apply to “all existing credit facilities” (page 77). As the word “renegotiated” might imply a conceptual connection with the definition of forborne exposures (FBE), we underline that the EBA guidelines have only an indirect effect on the current FBE practice, mainly as part of a change in the underlying credit management process rather than a direct application of the prescription to the FBE exposures – whose regulatory point of reference is the “EBA Guidance on NPLs management and FBE”.
The GLs should therefore apply exclusively to credit facilities granted after the application date.

• Since complying with the requirements regarding the collection is operationally not achievable for the stock of operations originated before the guidelines, we would suggest amending paragraph 10 as follows:
o Section 5 and Annex 2 appliesy to loans and advances that are originated after the application date of these guidelines. Section 5 and Annex 2 also apply to loan agreements where terms are renegotiated or which require specific actions triggered by the regular credit review of the borrower after the application date, even if they have been originated before the application date.
Alternatively, renegotiated loan agreements can be included in the scope of the guidelines only when there is a significant impact on borrower’s affordability assessment from the changes in the contract. However, this solution is largely second best, as the problems on collection of informations for stock of operations (dated maybe 20years) remain extremely difficult from an operational perspective

• The scope of application is also unclear with regard to lease agreements. Are these included in the scope? If so, the requirements in terms of collecting information and documentation go far beyond existing market practices and would have a huge impact on the sector and its ability to provide solutions for customers.

• The scope of the excluded exposures in section 5 and 6 should be extended to exposures which are guaranteed by sovereigns, local or regional governments and public sector entities in order to guarantee that promotional loans are duly carved out. Promotional loans support public policies, and as such they obey to criteria and conditions other than those typical of a commercial decision.

• In the scope section, it is also mentioned that Competent authorities should ensure that institutions apply these guidelines on an individual, sub-consolidated and consolidated basis in accordance with Article 109 of Directive 2013 / 36 / EU, unless competent authorities make use of the derogations as defined in Article 21 and Article 109 of Directive 2013/36/EU. Entities operating in third countries might face greater difficulties for complying with these guidelines at a sub-consolidated or individual level. This would harm the level playing field for European banks operating outside the EU. As such, the guidelines should not apply to portfolios located in third countries.


• Regarding the definitions, we reiterate that definitions must align with existing legislation. New terms shall also be avoided if not clearly defined.

o Professional: new definition shall be avoided to prevent confusion, defining professional as non-consumer is ambiguous. This brings many implication, including for example unrealistic information requirements for Private Investment Companies and Charitable organisations and institutions.
o Technology-enabled innovation: Provide a future-proof definition as currently technology is applied to almost any credit-decision process.
o Staff members involved in credit granting, credit risk management and control: firsly, compliance with EBA Guidelines on Internal Governance’ must be provided. Provide an accurate definition so that entities can assess which staff members would have to comply with such independence requirements.
o Total credit obligations: Provide accurate definition.
o Source of repayment capacity: means the borrower's total funds, cash flows and payment behaviour considerations as registered by the credit provider at the moment of the loan origination, covering all sources of cash flows (such as disposable income, regular private transfers - alimonies, rental income from real estate property, income from financial investments, income from private business or partnership, income from other sources), funds (such as saving accounts, investment products) and behaviour considerations.
o Commercial Real Estate: We understand that this definition refers to 2019 ESRB recommendations on closing data gaps. However, it does not allow for institutions to scope this class based on risk drivers. e.g. social housing, which obeys to criteria and decision conditions other than those typical of a commercial decision,or housing associations and the variety in organising those across the EU. The definition of CRE in respect of buy-to-let lending is also not clear. Buy-to-let can be within scope of MCD, although member states can opt to not apply MCD to this type of borrowing. This will introduce confusion and inconsistency.

Also, the guidelines consider assets used by the owners for conducting their own business within the scope of Commercial Real Estate (CRE) and not in Real Estate business. Criteria for evaluating RE professionals (pure CRE counterparties) are very different from the ones to be used for evaluating the granting of credit lines collateralized by RE assets to companies operating in other businesses. We suggest including only professionals specialized in RE activities in the CRE allowing institutions to continue using the ESRB/2016/14 CRE definition and to exclude on page 16 real estate used by the owners of the property for conducting their own business ”including them instead in the respective sectors the customers conduct their business in. “social housing”. The EBA requirements in this section (5.2.6) are not applicable to social housing, property owned by end-user. The ESRB definitions are primarily aimed at macro financial stability - monitoring purposes and not specific for credit granting purposes. ESRB also acknowledges that the different types of CRE vary and have different risk profiles and the new definition therefore requires separate reporting break downs

• According to the proportionality principle, we understand the guidelines as appropriate in relation to significant-amount transactions, which justify the additional costs connected with further detailed creditworthiness analysis and wider information collection required. On the other hand, some of the required information may not be available at all for small and medium enterprises or individuals. We would like EBA to clearly state that for sections 5-8 proportionality can always be applied to the size, nature and complexity of the credit facility, both when the GL state that creditors “should (at least)" and “should (at least) consider ”.
This issue is particularly relevant as Europe business environment is mainly composed by SMEs- in addition, considering that the majority of European SMEs depend on banks for their financing, this will hamper their possibilities to achieve their financing goals. Consequently, the guidelines implementation could have negative effects on credit granting, if they are not properly calibrated to the business portfolio of banks. We understand that all requirements as mentioned by EBA could be relevant in some cases. Therefore, creditors will always consider whether or not additional information or substantiation is needed. However, the requirements are not always relevant for all types of client, taking the complexity of the deal/financing into account.

• We recall that the European Commission is also running an evaluation of the Consumer Credit Directive (which can lead to a review proposal to be agreed by Parliament and Council). The two initiatives widely overlap on the topic of consumer credit worthiness. It may be worthwhile to avoid that – if the two initiatives follow their respective course each in their own timeline - they lead to different two sets of rules having to be introduced in banks processes and IT systems in a relatively short timeframe.

• The draft guidelines take the approach of grouping credit assessment requirements according to the type of borrower, whereas banks’ processes for corporate lending are typically based on industries / activities. Banks need to be able to align the credit granting process with the approach taken for modelling credit risk, and the guidelines should allow for this so that the guidelines do not replace internal rating systems.

• In the area of creditworthiness assessments, strong and effective rules and models are used and are already subject to controls. The policy and credit acceptance processes are linked to the specificities of the credit portfolios and personal files of each lender and are the subject of historically proven know-how. Under the current wording, the guidelines call into question the principles of risk based approach credit worthiness assessment widely tested in banks’ business models. They could have the effect outlined in our initial remarks with regard to limiting access to credit for atypical customers or innovative businesses.

• By setting the same requirements on banks in their approach towards managing risk, it could potentially lead to systemic issues with a reduced view on unexpected events. Further, alternative financing entities (peer to peer lending) do not have to apply these standards which could lead to unintended risks. We would like to stress the need to take into account the diversity of business models of banks (e.g. commercial banks, development banks and promotional banks, etc), in order to sufficiently address differences between bank types and segments of activity, specifics of processes and products in particular regarding customer segments, product offerings, access to information, state aid procedures, risk profile etc.

• It should be ensured that the Guidelines are consistent with the Regulation amending Regulation n°575/2013 as regards minimum loss coverage for non-performing exposures, and with other measures adopted within the European Council Action Plan on tackling the high level of non-performing exposures."
In our opinion the implementation of the guidelines by June 30th 2020 poses several problems mainly related to:
• If the guidelines are to be applied for credit facilities that were originated prior to the application date but renegotiated after the application date, there would be a compliance barrier related to gathering information to meet requirements included in section five of these guidelines.

• If adopted, the requirements in the guidelines significantly impact the credit granting and managing process, with huge investments in all banking organisational procedures. In particular, the greatest impacts will be on IT structure and staff training.
Banks will need sufficient time to adapt their investment and operational structure to the new standards. The set deadline is not realistic in light of the proposed text, which provides considerable changes to existing rules rather than simple harmonisation of interpretation of laws.

• Even after the necessary application of proportionality, the deadline for implementation is not realistic as the rules require some years to be implemented. Additionally, the deadline should be aligned, for example, with other regulatory efforts in the ESG area.

• As the scope of guidelines overlaps with other regulatory texts ( level 1 and 2 ) under development (DCC, MCD, CSR, sustainable financing) that should first come to conclusion before full application of guidelines can be envisaged.

• Development of credit-risk metrics incorporating ESG factors, especially regarding the forward-looking approach of risk measurement methodologies would require more time.

• Development of sustainableproduct catalogues and establishment of processes inside credit risk policies and procedures for evaluating the proper use of sustainable credit facilities would require more time.

• As including the proposed changes (even when applied in a proportionate way) would require years to implement, we suggest a phased-in implementation period with some requirements to be implemented earlier, and those more time consuming to be applied later.
Section 4.3.3. Technology-enabled innovation for credit granting:

• From our point of view section 4.3.3 (Technology-enabled innovation for credit granting) should be removed since regulation should be technology neutral and not impose a higher compliance burden when using a specific technology. Moreover, we believe that the majority of requirements listed in this section are already addressed in other parts of the document (need to understand and manage risks, involvement of the management body etc.).

• The main part of the guidelines deal with the “traditional” way of doing the creditworthiness assessment, while technology-enabled innovation for credit granting processes is underexposed. When these data models are adequately governed and back-tested, and these measures show that model outcomes are sufficiently robust and prudent, their use should be allowed.

As a de minimis proposal, requirement d) in paragraph 47 should be at least amended as it could hinder the application of innovative technology to credit grantingby obliging financial institutions to maintain multiple methods/tools only when using innovative technology. The comparison requested could also be impracticable for financial institutions as it is impossible to know the potential output of a credit granted via innovative technology if it would have been granted with a traditional method. More broadly on paragraph 47, it requires explaining the outcome, understand the underlying model used and ensure traceability, auditability, robustness and resilience. The requirements should be consistent with the main principles identified by the European Commission High Level Expert Group on Artificial Intelligence, as described in the Ethics Guidelines for Trustworthy AI, published on April 2019. In those Guidelines, a list of seven key requirements have instead being identified. Consistency across these two sets of guidelines is needed also to account for future legislative developments. It is also worth noting that in many cases there is generally not a clear divide with on the one hand a fully AI model, and on the other hand a fully expert based model. Models are very often a mix of both.

• In addition, we believe that most of the requirements in section 5 and criteria established in annex 1 and 2 will hinder innovation in credit granting as they are too prescriptive and do not allow companies to develop alternative procedures to determine the creditworthiness of a consumer or a professional. Thus, it seems that the creditworthiness assessment will always require financial institutions to collect specific information and documentation and have details on income, cashflow or financial commitments, for instance. This dismisses the possibility of developing alternative creditworthiness procedures that , for example, minimize the information required from borrowers or even though such procedures could be more accurate than traditional ones. Therefore, although we understand the rationale behind this section, the EBA should provide less prescriptive requirements to firms applying alternative procedures, which could improve customer access to credit or improve the accuracy of the creditworthiness assessment.

Section 4.3.4. Environmental factors and green lending:

General comments:

• We would like to stress that the Guidelines must be aligned with the Commission’s Action Plan and regulatory initiatives on sustainable investments/finance.
• As a general comment, we believe that the terms used in the Guidelines should be aligned with those in the regulatory framework for classification system of sustainable economic activities (taxonomy) currently under development. While credit provision activities are not included in the scope of the EU taxonomy and disclosure regulations, the EBF and UNEP-FI are launching a joint project to assess the applicability of the EU taxonomy to the banking activities , including loan portfolios and whether potential changes are necessary. To facilitate the application of the EU taxonomy for bank that would like to apply it to their activities on a voluntary basis.

• In addition, we caution against a temptation to adopt a too binary approach. The taxonomy is meant to be a common reference for sustainable activities (starting with environmentally sustainable ones). Activities, that are not in the scope of taxonomy do not necessarily pose an environmental risk. They may contribute to the environmental objectives either immaterially or could be environmentally neutral. Also economic activities that are currently identified as sustainable over time may no longer belong to this definition based on taxonomy update.

• We would like to also recall that the EBA was tasked in the CRR with a mandate to incorporate ESG factors in risk management via revised technical standards.
While we support integration of sustainability-related financial risk into risk management framework and risk management practices, we would like to stress that this is a complex issue and banks are currently at different stages with differences in terms of tools they are using or developing in relation to risk management, governance pricing or measurement of ESG risks.

• While, as an ultimate goal, all aspects of sustainability, including environmental, social and governance should be considered, currently the sustainability principle in banking is based primarily on the assessment of the environment and to a certain extent social aspect. Governance risk is often evaluated as part of the operational and reputational risk. Despite emergence of some methodologies in the area of climate risk, computing ESG-related financial risk remains a complex issue, given the lack of data, long term nature of the climate risk, its forward looking nature and number of assumptions that need to be made as well as “know-how” to assess the future ESG risk profile of the counterparties.

• The speed of negative climate change impact, potential new climate regulation or taxation and changing consumer behaviour represent a structural breach. The traditional retrospective approach does not capture the risk and sound forward-looking techniques capturing the longer term nature of environmental risks are neither available, nor will they be easily incorporable into the prudential framework given the different time horizon. With regards to physical risks, it is of vital importance that society as whole works on preventing these extreme risks. However, prudential regulation should focus on steering away from the financing of harmful activities which contribute to the chances of physical risk. This is a better approach for financial, economic and social stability than charging the potential physical risk itself.

• The EBF is willing to take an active part in the development of proper new methodologies, risk assumptions and analysis and the collection of experience necessary for a better integration of the ESG dimensions into the risk management systems.

• As the ESG considerations targeted in these Guidelines will be embedded into the loan origination process under the new technical standards which has to be developed, we recommend to await completion of the mandate given to EBA in the CRR and seek alignment with the EU classification system of sustainable economic activities.
As a minimum, we propose EBA to specify that ESG risk related requirements (including climate risks) are aligned with the timeline set in CRR2 or required on a best effort and proportionate basis, focusing on material risks in corporate lending .

Specific comments on the ESG related requirements in the Guidelines

• The requirements are especially demanding for banking groups operating in developing financial markets to which the Guidelines does not apply and for which the perception of the considerations on green lending and sustainable finance vary widely across countries. This implies difficulties at applying a homogenous criterion for classifying operations as “green” and monitoring climate-related risks in all jurisdictions. For this reason, we propose that the EBA should consider applying a two-speed requirement model for consolidated banking groups having subsidiaries in developing countries. Developing countries are usually out of scope of the guidelines but the European Banks have loans directly- or through foreign branches - with developing countries.

• Paragraph 48: We propose the following addition (underlined ) to the latter sentence: “Pending the finalization of the EBA Guidelines (CRD5 (Article 98 (7 c)) based on June 2021 EBA report, institutions should include, on a best effort basis and according to the proportionality principle, environmental, social and governance (ESG) factors as well as risks and opportunities related to ESG in their risk management policies, credit risk policies and procedures. Institutions should adopt a holistic approach, and incorporate ESG considerations in their credit risk policies and procedures for corporate lending using a risk-based approach. This emphasizes that banks should focus on most material ESG related financial risk cases and industries. This is particularly important in lending to SMEs where banks often have very large numbers of customer companies operating at industries with small ESG related financial risks.

• The scope of ESG related requirements should be limited based on proportionality. We therefore suggest:
o giving banks the possibility to adjust the approaches to ESG risks (including climate risks with the objective to focus on material ESG risk cases/industries and manage the most relevant risks.
o Limit the scope of the requirements to non-SME corporates.

• While project-based transactions can be easier to measure and assess against their environmental performance (Equator Principles, etc.), general use financing to companies (loans to companies that are not affected to a specific activity or project) and smaller transactions or SME lending are more complex to assess. It is going to be very data-challenging and operationally complex for banks’ systems to segregate and weigh each borrowing company’ activities, with the exact same classification format of the taxonomy, based on revenues or expenditures. It is also worth recalling the difficulties for smaller companies to have the relevant expertise to provide this kind of data to lenders. For project financing there are technical skills that allow the guidelines application, however, the requested strong specialization of relationship managers and CRM are not suitable to lending to SME.

• Paragraph 49 - the requirements as in the current version are too burdensome and difficult to track given also the scope of the Non-Financial Reporting Directive and difficulties to gather data from SMEs – who constitute a huge loan portfolio for banks. The requirements should also work in a situation where a bank originates large number of sustainable loans (e.g. hundreds or thousands) in a year (compared to small numbers nowadays). We like to point out that documentation and monitoring work repeats in each loan’s case. Monitoring should therefore not be more frequent than once a year unless special circumstances would require more regular monitoring. Without adequate data for origination and monitoring, sustainable loans cannot be offered.

• Since banks wish to collect the relevant data and steer their finances towards sustainable activities, we call for opening up Eurostat data sources and make them re-usable for finance purposes (for example to apply in deterring the carbon footprint of an activity following the PCAF method).

• The same applies for the technology aspect. The guidelines require the acquisition of a large amount of data and specific expertise to evaluate risks of the like of “technology risk”. It can be very difficult for banks to gather and assess this for a large number of clients (small and fragmented in different industries).

• Paragraph 50: We propose the following addition (underlined):
Institutions should position their environmentally sustainable lending policies and procedures within the context of their overarching objectives, strategy and policy related to sustainable finance. In particular, institutions should set up qualitative and quantitative targets to support the development and the integrity of their environmentally sustainable lending activity and to assess the extent to which this development is line with or is contributing to their overall climate-related and environmentally sustainable objectives.

• Paragraph 51: We propose the following addition (underlined):
Institutions should in particular take into account, on a best effort basis and according to the proportionality principle, risks associated with environmental factors and climate change in their credit risk policies and procedures. The risks of climate change for the financial performance of borrowers can be classified as physical risks or transition risks.

• Finally, we would like to mention that it should not be required that banks disclose publicly information about their sustainable lending targets etc. that is commercially sensitive. Disclosure should be aligned with Article 449 a CRR2 (June 2022).
• We consider that the requirement for governance and for credit granting are too standardized and too prescriptive. EBA must take into account existing framework and organisations already implemented in compliance with prudential rules. This existing framework does work well.

• In several instances, the Guidelines state the criteria listed are to be applied on a “at least” basis which would seem to imply they are binding. However, the criteria listed for example in Annex 1 may not apply in certain situations. As such, the expression at least" is not appropriate. The approach adopted in Paragraph 132 (d) - “at least considering which metrics [in Annex 3] would be applicable...” - is in our view more appropriate.

• Harmonisation between these guidelines and the guidelines on “Reporting instructions on Credit Underwriting data collection” recently issued by ECB (April 2019) shall be ensured to avoid possible overlaps / mismatches, also to better understand some definitions.

• Regarding the implementation of automatic process of decision making, it should be clarified which kind of analysis are required to perform the comparison between automatic and manual processes.

Paragraph 34 - it should not be necessary to establish specific and more granular policies for different sectors, unless justified by the fact that their frequencies of default significantly deviate from the average.
• Paragraph 40 to paragraph 42 - the controls should only cover the search of potential third party and never cover the source of funds, beside in exceptional cases (politically exposed persons in case of high risk defined in a risks classification). A risks classification should be defined with the aim to have a thorough knowledge of the borrower in case of high risk (creditworthiness, analysis of the investment’s profitability, the use of funds if possible…) and a detection of unusual transactions (particularly complex transactions, unusually large transactions or transactions which have no apparent economic or visible lawful purpose). The vigilance measures should not apply if during the relationship (which cannot be limited to the credit), there isn’t any payment incident, any new application for credit, any change of IBAN or any material change. The specificity of non-purpose loans and of specialised providers of consumer credit which are not account holders should be taken into account. A risk-based approach and the implementation of risk-proportionate measures are necessary.

• Paragraph 54 on artificial intelligence/ data infrastructure - The scope of the data to be input and available in systems seems very wide. These types of requirements should once again be applied proportionally as their implementation cost would be significant for no concrete benefit. The benefit of such data recording should not be overestimated, as good origination and loan monitoring requires teams of expert front officers and portfolio managers who currently do it on the basis of in depth analysis. Data recording in IT systems will never provide a comprehensive view of the risk taken, as reality cannot be reduced to a few number of drivers. The risk here would be that banks management decide to reduce the credit activities that require unnecessary excessive data recording.

• Paragraph 56 - Data requirements for NPL are extended to performing loans which seems excessive. We suggest amending the text to ‘…institutions should consider if any data elements from the EBA’s NPL templates are also appropriate to include for performing loans’.

• The credit granting criteria described in Annex 1 containing limit ratios cannot always be imposed in their entirety. In particular, acceptable loan-to-income and debt-to-income ratio limits may sometimes reduce the effectiveness of risk systems compared to using net income.
In general, we caution the systematic implementation of these ratios, as they don’t really translate to the situation of each customer, and are not necessarily predictive of the risk of non-payment.

• Annex 2 imposes too much information with no consideration of of the type of credit. Clients may consider these information requests to be intrusive when disproportionate compared to the type of loan facility. This also seems to contradict the principle of minimization in data collection enshrined in the GDPR."
• As stated in our general comments, we see that the requirement on credit decision making may limit proven and well-functioning lending activity. Defining the organizational control, monitoring structures, policies and procedures on conflicts of interest based on the proposed requirements is extremely challenging.
In more detail:
o limitations in credit decision making in terms of time and number should be removed. The number of delegated credit decision is not correlated to an increase in terms of risks undertaken by the bank (e.g. para 59 limitations on the number of delegated approvals)
o Paragraph 60-61, this seems to imply that the risk function should be involved in the credit decision making process at a granular level. Whereas the risk function is expected to be involved in setting the policy for credit decisions, in some banks it does not have a mandate to decide on individual credit files.
o Paragraph 63, allowing individual approval authorities only for small and non-complex transactions could significantly increase the complexity of the lending process. This could decrease the level of efficiency of banks. We propose to eliminate point a) para 63 for all the banks that , in case of individual decision made only by business functions, can ensure a credit process that includes an independent opinion released by the risk management function (that ensures the independency of the overall judgment, limiting the discretion of the delegated role). Moreover, criteria laid down in para. 63 on conflict of interest are extremely wide and go beyond what is provided for in CRD, CRR and in the EBA Guidelines on Internal Governance. In particular, the Guidelines seem problematic with regards to:
 professional relationship with the borrower to which para. 63 b. i. refers e.g. is professional relationship considered between a client and a relationship manager at a branch?
 It is not clear the meaningpolitical influence to which para. 63 b. ii. refers.
To ensure legal clarity, the guidelines should simply cross-reference existing legislation (CRD, CRR) and guidelines (EBA guidelines on internal governance), instead of reporting them or adding divergent/additional requirements compared to what is already in place to ensure that conflicts of interests are avoided.
• It is important to introduce in Section 4 the concept of materiality" at portfolio level for a Bank. Individual credit files decisions ensure individual credit file quality and compliance with risk strategy and credit policies. In addition to decisions on individual credit files, credit risk limits ensure risk diversification and prevents concentration on portfolio with shared risk characteristics. Credit risk limits are only meaningful for material credit risk portfolios, when smaller, non material, diversified portfolios should not require specific RAF limits. Applied to a large diversified generalist bank, the RAF does not separately cover every single credit portfolio of the bank with dedicated limits. Limits are set-up for material portfolios with shared risk characteristics risk profile, concentrations and performance,.

• Regarding section 4.3.1:
AML & CTF topics should be addressed exclusively in the Directives on Anti-Money Laundering and ESA Risk Factor Guidelines. This would ensure legal clarity and consistency. These guidelines shall simply refer to relevant legislation without adding different/ new/ or inconsistent requirements.

• Regarding section 4.4:

o In our opinion, independence in credit decision-making section (4.4.1) is ambiguous and EBA should revise it. In the existing Risk Committees where risk granting is decided, several stakeholders act as challengers of the decision-making process. For instance, when the business area proposes a new credit risk facility, the risk management area performs a challenge analysis, which is then supported by a technical contrast. Then, under a delegation scheme a new responsible person in risk management area authorizes the facility.
o From our point of view lending to affiliated parties’ section (4.4.3) should be removed from this section given that the section is independent and, in a way, disconnected from the purpose of the document. EBA should take into account that the only reference to such operations can be found in paragraphs 67 to 69 related to credit decision-making, while we do not observe additional mentions in the remaining sections of the document. Should EBA decide to keep the above-mentioned paragraphs on “lending to affiliated parties”, we would the propose additional comment: A definition of the term “affiliated parties” should be included in section 2 of these guidelines consistent with those already provided by local regulations

• Regarding session 4.5:

o Paragraph 76 (g) - The risk management function should be reviewed so that it is not confused as a second opinion of a nature identical to the first, otherwise there is a complete duplication of functions.

o Paragraph 76 (k) - should allow sufficient leeway for stress tests to be part of other regular stress tests that banks perform, notably in the ICAAP, and not specific to credit portfolios.

• Regarding the section 4.7:

o The wording of this section is ambiguous and leads to confusion. Current remuneration policies in banks already cover the independence measures which forbid the link between remuneration schemes associated with the growth of new business. The performance targets of the staff involved in credit granting and their variable remuneration are based on many factors and they are aimed at creating long-term value and rewarding the achievement of results on the basis of prudent, responsible risk bearing. The sector already applies "MIFiD" rules of conduct relating to the variable remuneration of its staff. The principles derived from this Directive require institutions to respect the interests of consumers, to adopt compensation policies that eradicate conflicts of interest and to provide fair and equitable remuneration not based solely on file acceptance quotas. The current legal provisions are intended to prevent the remuneration of the staff of the lender from being solely dependent on the number or proportion of accepted credit applications and therefore appear sufficient.

o In addition, the variable remuneration of the staff involved in credit granting should be consistent with and promote sound and effective risk management and should not encourage risk-taking that could exceed the level of tolerated risk of the institution. The quality of credit analysis will anyway depend on the criteria and requirements defined by the bank in its credit risk policy, guidelines and procedures for credit granting process and its monitoring. Besides, the definition of the population is large/not precise and includes staff with different roles and levels of responsibilities (involved in credit proposals, credit approvals, credit administration, controls…) Apart from the above, it should be in line with the business strategy, objectives, values and long- term interests of the institution, and should incorporate measures to avoid conflicts of interest.

o Also, the link of variable remuneration of the staff involved in credit granting to the long-term quality of credit exposures appears more as a theoretical concept rather than a practical one, since the credit cycle in some products can be very long and dependent on the economic cycle. For instance, would this mean that a mortgage credit variable remuneration must be disbursed over a 30 year period? It is better to monitor the correct application of the rules by the acceptance teams, than the solution proposed. Also, the scope of population seems too large and not sufficiently precise.It should be coherent with provisions of CRD 4/CRD 5 on material risk takers.

o While remuneration policies and practices should be consistent with the overall credit risk appetite and not create conflict of interest, it is also essential that performance management and reward of employees involved in credit activities is be based on several criteria and on indicators linked to their activities and the quality of their credit risk analysis but not based on the quality of credit exposures which are independent and disconnected from the employee him/herself, his/her individual performance and the way he or she conducts his/her activity."
• Paragraph 76 - The requirement to provide an “independent risk opinion to the credit decision takers” (par 76c) and an “independent/second opinion to the creditworthiness assessment” (par. 76g) seems to require an ex-ante supervision of the risk management function within the credit process.
This approach, implying an active role performed by the risk control function during the lending phase, might be hardly applicable as:
o the prior involvement of the risk control function appears not fully coherent with the separation of responsibilities between the ex-ante first line of defence (lending functions) vs the ex-post second line of controls (risk management) and, ultimately, with the regulatory principle of segregation of duty;
o the need to have second opinion to the creditworthiness assessment might trigger process inefficiencies related to the duplication of activities and skills in charge of different functions, entailing inter alia also additional staff costs.
o in case an institution implements such an active role by the risk control function (for example for the large corporate clients), such a review should be considered consistent with par. 76n, because it means that the second line of defence has performed a complete coverage of credit assessments rather than sample-based one. In case further sample-based checks are deemed necessary, these should be executed by a third line of defence (typically internal audit) rather than another risk-management unit, in line with par. 75c.

• Risk management should not be interpreted as a function to be uniquely performed by the Risk Office. It can also be performed by other areas that have the required functional competencies provided segregation / independence with the commercial area is ensured. It could be areas such as credit management, rating, etc.
• Information (even if documented) like income, financial commitments (i.e. expenses) and employment, reflect only the past (in best cases also presence) but loans (more precisely instalments) are commitments in the future.
On the other hand, predictive analytics encompasses a variety of statistical techniques from data mining, predictive modelling, and machine learning that analyze current and historical facts to make predictions about future or otherwise unknown events. One of the best-known applications is credit scoring which is used throughout financial services. This efficient tool which (compared to just simple equation: „income – expenses“): a) provides better protection of the client; b) leads to higher stability of banking sector and in the same time c) creates lower costs for the client (time/effort).
The requirements in the guidelines do not account for the changing market environment and technological progress in the field of financial services. This requirement considers collecting paper documents as evidence of declared information (only) which impose another requirement for verification (which generate costs not only for the bank but also for other institutions like .e.g. employer of the client) – because documents submitted by clients can be modified/faked (exactly the same as declared data).
This approach can be easily replaced by implementation of efficient scoring model. Unfortunately, this approach is not in line with this guideline. Level of collection and verification of information should be adequate and reasonable to reflect defined situation (profile of client, loan amount, total exposure…etc.).

• We reiterate the importance that compliance with the collection of information and data as set out in Annex 2 should be proportionate to the type, size, nature, complexity and risk profile of the credit facility and of the client. Business plans (93.e) or financial projections (93.f) for instance are not available for all “professional firms” and their analysis not always relevant in case of short terms facilities (eg. Trade Finance or Export Finance). It should be made clearer that Annex 2 is not to be understood as a prescriptive list to be applied always and for all types of lending. Again, wording at least" in para 92 to 94 introduces ambiguity in the applicability of Annex 2 and are understood as directly contradicting the principle of proportionality.

• If the client does not accept the use of a third party to collect the information (by consent), what means will be granted to the institutions to fulfill the various verification obligations under laws in the field of combating fraud, money laundering and terrorist financing? Employer surveys of suspicion of payroll fraud are one example. Employer queries respect the principles of confidentiality of data and respect for privacy and are used in cases of suspicion. Requiring systematic consent from the consumer will be tantamount to depriving the lender of a reliable source of information without another current audit solution being available on the market. Additionally, GDPR provides several legal grounds for the processing of personal data. These guidelines should not further reduce the scope of GDPR but should instead cross-reference it by simply stressing that data collection and processing should comply with GDPR.

• We would like to note that the European Court of Justice has already ruled on the application of the principle of proportionality in the granting of consumer credit. The Court of Justice has ruled that the Consumer Credit Directive gives the lender leeway in deciding whether to verify the information received from the consumer with other data.
According to the Court, the lender must, when assessing its audit duty, take into account:
o the fact that simple unsubstantiated statements made by the consumer cannot be qualified as sufficient if they are not accompanied by supporting documents. For example, new clients / prospects whose lender does not have prior knowledge of their financial situation; and
o the fact that the Consumer Credit Directive does not require lenders to systematically check the veracity of the information provided by the consumer. In other words, depending on the specific circumstances of each case, the lender may either be satisfied with the information provided by the consumer or decide that it is necessary to obtain confirmation of this information - whether from the consumer or not.
The question whether the information communicated by the consumer is sufficient can therefore, in application of the principle of proportionality, concretely differ according to:
o The circumstances in which the credit agreement is concluded and the digitization of processes;
o The personal situation of the consumer. When, for example, the consumer's salary is paid into the lender's account, the lender can verify this himself without having to ask for supporting documents;
o The amount of the credit agreement. The smaller the credit amounts, the more limited a verification requirement is, where the creditworthiness of the customer can be evidenced through other avenues (e.g. knowledge of payment history on a credit bureau and/or knowledge of asset base) to support the customer declared information.

• The only distinction made by the guidelines in this section is between consumers and professionals, establishing general criteria for these segments and some additional specific criteria applicable to certain portfolios. But, for example, the process of granting a card or a preapproved small loan with respect to a Lombard credit is very different among consumers. Regarding professionals, a freelancer and a large corporation seem completely different.

• For most loans granted to SMEs - which represent a significant proportion of the loan portfolio for many banks - some information listed in Annex 2 lead to disproportionated collecting costs compared to the economic value of the financing transaction or to the added value in the creditworthiness analysis. We reiterate the need for the guidelines to clearly state that information listed in Annex 2 are provided as examples and should be collected and verified only if they are relevant for the type of product, according to the proportionality principle. Again, the expression “at least” seems not accurate as it implies that this information has always to be collected and does not allow for the application of the proportionality principle. Flexibility for very easy and small loans must be considered. E.g. banks may have a simplified credit process for overdraft and credit card limits up to e.g. EUR 1.000, where they may ask customers for their income, but do not verify it. We deem, moreover, that the Guidelines should better state the possibility for info packages differentiated for individual firms, small corporates, mid-sized companies, large firms (for example by introducing thresholds, also driven by loans’ sizing and borrowers’ risk profile) and accept a certain degree of flexibility.

• The same concerns apply to consumer loans are typically for smaller amounts and shorter duration. It is disproportionate to ask the same kind of information and verification for a short-term loan of € 300, as for a loan of € 250.000. Loans under € 200- or of a duration less than 3 months are outside the scope of the Consumer Credit Directive which governs the provision of unsecured consumer loans in Europe. Moreover, the systematic collection of documentation is not fit to some distribution channels, such as credit at a distance or linked credits in shops. Consumers do not go shopping with their tax notice or pay slip.

• Additionally, some information is not at all available to banks, as it cannot be obtained by either financed companies or official sources. Alternatively, although information is collected by the bank during the loan origination process, this may not be available on its main data system; therefore, it cannot be used without significant IT adjustments, which would take a long period of time for implementation. Therefore, we believe the guidelines should consider alternative procedures for obtaining the information which allows to assess debtor creditworthiness, and minimize the information required from borrowers or which do not take into consideration specific individual pieces of information, even though such procedures could prove to be more accurate than traditional ones. In this sense we propose to include a specific paragraph that allows financial institutions, where necessary, to obtain the information by consulting the relevant database as it is already the case for 2008/48/EC (Consumer Directive) and 2014/17/EC (Mortgage Directive) refers. This would also apply to the actions and assessments in paragraph 90 which may be recorded, where appropriate, in internal databases rather than documents.

• Paragraph 85 – It is not clear whether the reference in paragraph 85 to have a ‘view of all the borrower’s credit commitments (‘single customer view’) refers to all commitments or only those towards the lender. If it is the former case, then due to national requirements on data protection (amongst others) it may be extremely difficult to comply with this requirement. With regard to the comprehensive view of all the borrower’s credit commitments, the guidelines should consider the constraints to obtain the information about external debts in certain markets (e.g. local bureau provides the exposure but will not always show the instalment amount, local bureau does not provide information if the financial institution exposure is below certain amount). The legal framework in some countries may forbid the transfer of specific information on borrowers even between separate legal entities belonging to the same banking group within the same country. Therefore, borrowers’ own declaration is key to have a comprehensive view of its credit commitments of which some elements may be verified by the lender, some may not. Additionally, in recent years and especially in consumer financing, the admission process has evolved from clients applying for credit in a branch and assessed by a reactive scoring model (i.e. model that assesses the application based on the information and documentation generally provided by the client); to a massive evaluation process which make available pre-screened credit limits to clients in digital channelsassessed by a behavioural scoring model (i.e. model that assesses clients based on information automatically fed from internal or external database).
This automatic process significantly reduces operational risk. However, it is highly dependent on the wealth of database information, especially in external databases which sometimes do not evolve as the financial institutions and business models demand. We propose to include in the definitions in section 2 the term “total credit obligations” and add a clarification that external debts will be collected if that information is available in a public register or database...

• Asking for the mandatory availability of business plans and projections from all clients is in clear contrast with the proportionality principle and the evidence that smaller (and therefore less structured) counterparties do not usually have managerial ability to develop such detailed documents. In such cases banks’ assessment should be allowed to rely on most recent historical performances and few key budgeted figures (where available) with the aim to understand their future sustainability. Involvement of internal specialist functions for all transactions is in fact not sustainable..
• Paragraph 83-88: It is stated that information gathered should be verified. It is however not clear how this shall be performed given the concerns on verification of information outlined above, notably that the Courte of Justice has already clarified that the principle of proportionality shall apply. The guidelines should reinforce good lending practices to deliver good consumer protection. They should not however introduce unnecessary burdensome and disproportionate requirements which would have the effect of driving up the cost of borrowing or exclusion from access to credit by certain customer segments “The documentation of information” cannot be considered as an obligation for the creditor to collect systematically documents testifying the borrower’s declarative information. Such a requirement would not be compatible with small amounts lending business models such as consumer credit at points of sale or online for example. It would not be compatible with small ticket vendor lease either

• Paragraph 87: We have reservations with regard to the requirements about the in-depth assessment of a guarantor for private persons. The processes of many institutions are primarily focused on the assessment of repayment capacity of the borrower, not the guarantor.

• Paragraph 88 - before making enquiries to third parties on borrower’s personal data, institutions and creditors can put a limiting factor on reliability and certainty as there is a certain interdependence on the client in ensuring the reliability and certainty, particularly regarding the information that is being asked in the GL guidelines for loan origination. A good example is IFRS16 whereby banks are depending on accountants and clients in providing information on operational lease contracts that could impact leverage ratios. We propose to replace the term “should make any necessary checks” with the term “should make reasonable checks”. This would ensure the application of the proportionality principle with the requirement. In addition, bank secrecy limits the possibility to conduct enquiries among third parties to verify the information and data collected. Inconsistences are an alert to further investigate within the allowed legal framework.

• Paragraph 89 - we propose to replace the term “on all related connected clients” with the term “on all relevant connected clients”. This would ensure the application of the proportionality principle with the requirement. Information to be gathered for all connected clients, is considered not to be feasible due to information overload and time-consuming process. This is currently only done for the most important connected clients. This approach strikes the balance between cost and complexity on the one hand, and benefits on the other.

• Paragraph 86 - states that the information and data can be proportionate given the purpose, size, complexity and potential risk associated with the loan but the requirement under paragraph 92 to collect and verify at least the information as set out in Annex 2 for consumer loans goes against proportionality. Consequently, paragraph 92 has also to be adjusted to allow differences in information and data collection/verification depending on for example whether the nature, the amount and duration of the loan.

• Paragraph 91-94 - proportionality principle but also appropriateness should once again be included in the text as the strict application of the guidelines would represent a regression in risk management for the concerned institutions that have demonstrated a permanent high-quality level in granting process. We would like to remind that quick and automatized decision making are fully part of the good quality of small ticket leasing as they allow providing customers with prompt feedback.
To account for duly application of proportionality principle , the EBF suggests rephrasing “should collect and verify information” to “should take reasonable steps to collect and verify relevant information”. For both sections (5.1.1 and 5.1.2) . For annex 2, which is a more detailed list of information, a statement that clarifies that these information lists are indicative and not compulsory should be included.

• Paragraph 101 (and 114,121). The supervisor’s expectations should be clarified regarding the sensitivity analysis. The requirements must be properly delimited by the proportionality principle, as for limited proposals and retail SME customers the requirements are not proportionate to the risk.

• Specifically on Annex 2 on professionals:
o We reiterate the need to clearly state that that all annexes should be read as ‘should consider’ and therefor are recommendations
o Point 3: Financial statements covering a reasonable period: in the case of specialized lending, where a new asset is being financed, there would be no existing financial statements covering the previous years.
o Point 10: Information on existing covenants, and borrower’s compliance with them, where relevant. Having this information available in annual reviews memos, in PDF version, should be considered as sufficient, and banks should not be obliged to record these covenants in IT Data systems.
o Point 14 requests evidence of the value of collateral: this point is linked to Section 7, so please see our response to question 11 of the consultation paper.
o Point 16 requests information on the enforceability of collateral is disproportionate if requested for any loan origination. Depending on the nature of the collateral (mortgage, privilege of the money lender – PPD, guarantee given by an insurance company or a financial institution) the terms for calling the collateral into play within a Member State should be sufficient while complementing information on the collateral itself requested by point 12. Also regarding lending to professional, information on the enforceability of collateral, in the case of specialized lending, substantial control of the collateral is achieved through different security packages. The power of this security package is notably to enable lenders to put a strong pressure on sponsors (who brought the equity), which makes a restructuring easier. The recovery generally best obtained through a restructuring is based on the future cash flows to be generated by the collateral on which the lenders have a substantial through different structures and security packages. The rating and LGDs based notably on the efficiency of such security package, in terms of future cash flows benefit, is assessed by the internal legal teams and front officers and validated by the risk department. Therefore, regarding the point 16, we suggest adding “in the case of specialized lending, description of the structure and security package of the transaction”.
o Any point mentioning “evidence of” : as long as these information are in the credit applications or in the annual reviews memos, this should be considered as sufficient evidence. There should be no request of recording this information in IT data systems. For example, regarding item 6, banks should not be obliged to record the financial projections (balance sheet, profit or loss, cash flow) in data IT systems. Having the financial accounts of the borrower, as published by it, in a PDF version for example, should be considered as sufficient. However, financial statements should not be mandatory, even in a pdf format, as some small companies may have not audited those accounts and therefore those data could be of poor quality and could lead to wrong decisions."
• We consider that the assessment is way too prescriptive. The CCD doesn’t precise specific metrics to assess creditworthiness to account for the flexibility that is needed to consider the different situations. We reaffirm that it is not adequate to treat in the same way all credits. EBA must take into account the differences among types of credit the different duration and amount. -A strict application of the listed metrics and parameter would challenge current scoring models. These models do work well, and we are able to limit risks and indebtedness. In addition, the use of the metrics proposed could eventually become factor of exclusion as their rigidity would not be able to account for the differences at member states’ level. More broadly, we would like to stress the importance of keeping consumer credit and mortgage credit as separated and not treated under the same rules. These types of credits are fundamentally different and require different treatment. These two different types of loans are indeed duly treated under two different Directives specifically because of their differences. So the credit worthiness assessment for these two types of loans cannot be the very same.

• Requirement for lending consumers: § 96 a 99. The guidelines for assessment of a borrower's credit worthiness rely heavily on the calculation of different ratios, which is not standard practice for all types of lending (e.g. credit facilities fully collateralised on the basis of a marketable securities portfolio such as lombard lending).

• Paragraphs 101, 114, 121, 145, 146 – In our view, the requirement to carry out sensitivity analyses go against the principle of proportionality, also in light of the limited duration of most consumer credits. The most important aspect is to calculate a break-even point. The assessment should consider any known or reasonable foreseeable changes to the customers circumstances which might affect their ability to sustainability meet the repayments over the term of the loan. In the CCD no such requirement is foreseen, therefore before the introduction of any such requirement, a proper quantitative cost benefit analysis should be performed by EBA. We believe that the costs brought by such requirements outweigh by far the benefits.
In addition, the time, effort and resources that this kind of exercises need and the fact that could lead to perfectly viable operations being rejected in some unlikely stress scenarios. Application of severe idiosyncratic events in combination with macroeconomic downturn and other adverse changes (e.g. increase of interest rate on funding by 200 bps) would undoubtedly lead to drastic worsening of the financial position of the borrower and most probable rejection of the loan application, notwithstanding the fact that such scenario is very unlikely.
Moreover, the implementation of these analyses is based on the existing suitability tests for so-called complex investment products under the Mifid rules. It goes without saying that this type of adequacy test is impossible to implement for consumer credit. Solvency analysis and scoring models already take into account potential negative future scenarios. The prescriptive proposals introduced excessively complicates the models with no concrete benefit. On the contrary, they risk hampering the availability of credit to consumers.
Indeed, stress exercises in origination, their reporting and the analysis per operation introduce complexity, delay the loan concession and do not provide added value for consumers. As an alternative to the need to analyzing each operation, the Guidelines should allow these analyses to be carried out centrally per portfolio, vintage, etc. In conclusion, the guidelines should leave room for a “standardized analysis”, especially regarding “professionals”.

In general, we understand it could be more practical to contemplate serious prudent parameters that can cover possible future contingencies (for example, measuring effort, LTV - which is, actually, what entities use nowadays).

• Paragraphs 98 and 116 – Repayment capacity. This term should be considered as in a wider approach, where applicable, to assess the ability to meet obligations. Based on some of our members positive experiences: the savings account, investment products, etc. are other sources of repayments beyond the income that may be considered on the ability to meet obligation assessment. Therefore, we propose to include in the definitions of section 2 the term “source of repayment capacity” which would not only consider the disposable income but also those alternatives sources of repayment such as saving accounts, investment products, etc. which allows paying the credit. The references to ”income” should be replaced by “source of repayment capacity”.

• Paragraph 99 (et al.) - Ability to meet obligation assessment. The guidelines try to limit concession decisions to a series of ratios. But the quality of decision making cannot be judged solely by the value of certain ratios, which could end up rejecting perfectly viable operations . This evaluation has to be made by the joint evaluation of all elements identified as relevant factors at the time of the concession. In addition, the guidelines establish specific procedures for self-employed, seasonal or irregular income, terms that pass the borrower’s expected retirement age, no consideration of potential income increases and foreign currency loans. These specificities are based on past experiences where loans with these characteristics showed a significant worse behavior. To account for proportionality, it must be clearly stated in the guidelines that these lists are indicative and not compulsory. This would also apply to Annex 1 which sets some credit granting criteria. This would allow also to consider that the assessment may be adjusted due to the existence of information constraints (e.g. access to external debts in database).

• The definitions of “consumers” and “professionals” are also not adequate. For instance, for couples of consumers or joint accounts, it should suffice to fulfil the creditworthiness criteria on consolidated level. The same should apply for professionals who have a joint account. The category of “professionals” now appears to include entities from small businesses to multinational corporations. The requirements must respect the principle of proportionality and the level of granularity must be proportionate to the risk profile of the counterparty or class of counterparties. Once again, the proportionality principle should be better reflected into the text of the guidelines.

• For example, the proportionality principle should be better integrated by replacing wording at least" with "where relevant" and "list non exhaustive" to avoid confusion. We ask EBA to amend the wordings “Should at least” and “should” for example in the following articles:
o 129, 134, 136, 139, 143-146, 154, 169: Please replace “should” by “should consider”
o 118: Please rephrase to ‘“Institutions should take reasonable effort to consider living expenses.. etc”,
o Paragraphs 126, 127, 132 and 134 – Proportionality is not appropriately applied due to the term “at least.

• Paragraph 112 b) and c) – the requirements (similar to those included in paragraph 166 and 173 and 177) are excessively burdensome and difficult to fulfil. As a matter of fact, lenders have no data and cannot be responsible for assessing the quality of architects, engineers who take part in the property development. Furthermore, the certification of the costs associated with the development is not easy to obtain and it could be very expensive for the borrower. We ask EBA to allow the use of proportionality.

• Paragraph 122 & 166 Projection of all costs associated with the development certified by a qualified and reputable quantity surveyor (or similar) - for smaller projects, typically in SME this is not done by a quantity surveyor, due to the costs that normally are associated with this exercise.

• Paragraph 125 - "Collateral by itself should be under no circumstance a criterion for approving a loan…” is too restrictive. For example, a financial pledge of a deposit on the same amount of the loan largely removes any concern with the loan repayment. To allow for flexibility it can be allowed ‘under extreme circumstances’.

• Paragraph 127 – These considerations should be linked to the amount of the loan, existing outstandings and type of customer.
• Paragraph 130 – Assessment of the borrower’s exposure to climate-related and environments risk should be allowed to apply proportionally.

Paragraphs 131 to 135 and 138 to 141 (SME specific) + Appendix 3: imposing to refer to a list of metrics to analyse a professional borrower’s financial position, is over-prescriptive and regressive, especially considering SMEs. The requirement to « consider at least » some of the metrics listed, such as “cash conversion cycles”, “cash flow generation”, “projected capital expenditure” is not compatible with credit granting processes of some leasing and factoring large scale, small amounts and short-term activities (for instance small equipment leases through vendor programmes). It would have disproportionate impacts on the organisation and tools on which these activities currently rely, often partially automatized and technology-enabled, that have long been developed and have proved their efficiency. In addition, banks need to be flexible enough in order to assess complex business models of SMEs, in order to be able to use innovative financing strategies to allow them to get the financing they need. It would represent a backward move to less risk sensitive credit granting analyses, for a disproportionate cost in terms of HR and IT compared with the potential expected reduction of cost of risk. We would like to remind that very quick and automatized decision making are fully part of the good quality of small ticket leasing as it allows selecting the customers in the first instance.

• Paragraph 131 – Clients setting up new activities cannot evidence future cash flows. They can set up business cases, but evidence is different. Banks should sufficiently challenge the provided projections and (financial) information instead of always make their own projections.

• Paragraph 132 – Propose to add that “if relevant” at the end of the first paragraph. The list of assessment requirements is too detailed and while relevant to larger customers and credit exposures e.g. 132 a and c are excessive requirements for limited SME exposures

• Paragraph 135 – It is confusing that EBA uses “if relevant” and “at least” in the same sentence. “at least” should be deleted.

• Paragraph 135 – Some of the listed financial metrics may not be available or can be hard to perform by the bank or in the third country where the credit operation is carried out. For example, difficulties may arise from the calculation of DSCR (Debt service coverage ratio): both with reference to cash flow available for debt service (business plans are not always provided) and to amortization profile of third parties debt. Another example is the capitalization rate which is difficult to assess for most clients as those companies are not listed on the stock exchange. Consequently their market value might be unknown or very difficult to assess. The approach adopted in Paragraph 132 (d) - “at least considering which metrics [in Annex 3] would be applicable...” - is in our view more appropriate.
• Paragraph 138, 139, 141 and 149 – The assessments mentioned in these paragraphs should be carried out only where relevant.

• Paragraph 140 – This should not be required in member states where published financial figures are available.

• Paragraph 152 – These checks are often outsourced e.g. to notaries.

• Paragraph 156 - the provision to perform a due diligence of the agent in a syndicated loan facility may be exaggerated depending on whether existing / prescribed procedures (KYC, AML) or any additional procedures are interpreted (what exactly and how this affects time and cost aspects).

• Paragraph 169 – We have reservations about an assessment of the real estate and follow-up of all phases of development done by a reputable estate agent for consumer finance clients. Again, given the costs linked with such requirement for consumer finance clients this is fulfilled from the relationship manager and/or decision taker having experiences.

• Paragraph 176 - the pledge of shares in an SPV is not always reasonable or possible, so it is necessary to add "and where applicable" after the words “In case where a special purpose vehicle is established for the project”.

• Paragraph 177 – verification of the costs is not reasonable for all the projects so we propose to add the words “when applicable” at the beginning of point c.

• Section 5.2.6 – The CRE loan definition that is presented in Page 16 also includes loans collateralised by real estate that is not necessarily income-producing real estate. Some of the recommendations made on Chapter 5.2.6 don´t make sense for such loans. It must be coherent with existing legislation. We recommend using the same definitions as art.4 CRR via cross-reference."
• The scope of the asset classes, products and clients covered is very wide. The proportionality principle is key to section 5.2 and certain wordings should be modified for clarity sake on their applicability (eg. at least" replaced by "where relevant" and "list not exhaustive"). Once again, the “ one size fits all” approach is not appropriate for this topic.

• In particular, for consumer loans, using a common framework to regulate the loan origination for mortgage loans and for consumer loans is not adapted to those loans’ characteristics, which are completely different in terms of amount, duration and impact on the borrower financial situation. The creditworthiness assessment of borrowers significantly differs from consumer credit (industrial approach where the human decision is often mainly based on the result of a scoring) and mortgage credit (tailor-made approach).

• Paragraph 180 – The requirements seem to impose on lenders a responsibility for the possible misrepresentation of information provided by the borrower. We propose an alignment of this with the requirements of the art. 18, (4) of the Mortgage Credit Directive.

• Paragraph 183 - It will be burdensome to document every decision in detail. If the key issues are weighed correctly in the presentation, then repetition is redundant. We suggest rewording as follows: “Credit decision should be clear and well documented”.

• Paragraph 185 – In some cases, pre-disbursement (before all conditions are met) is granted, upon decision of the appropriate decision level. We in principle agree with the introduction of the requirement, however, there should be flexibility with decisions to be taken by an appropriate credit committee.

• Paragraph 181 - The information needed for the decision of the credit committee is generally included in the credit application. Which we consider as sufficient regarding documentation. The recording of such credit applications should be considered as sufficient and no requirement of recording these detailed information in a data infrastructure should be considered here.

• Ch 5.2. Specifically in scope are: ‘Commercial Real Estate Lending, Shipping Finance and Project and Infrastructure Finance’. It is unclear why specific separate requirements are proposed.. The selection seems quite random and puts additional burden on these specific groups. Furthermore, Project and Infrastructure Finance are grouped under one header, however Project Finance is much broader that Infrastructure Finance. While typically Infrastructure Finance transactions have the character of a Project Finance, the latter also includes transactions in e.g. Oil & Gas Midstream and in Power & Utilities. It would make more sense in this case to remove the specific references to these sectors, as they are specialized enough as to follow their own specific financing strategies. On top of this, Export Credit Agency backed deals can fall in many of these categories, while being ruled by the OECD Arrangement on Export Credits. Extra requirements could disrupt the usual way of conducting business.

• General consistency is very important. It seems that chapter 4, 7 and 8 apply to all exposures whereas some exposures are excluded from chapter 5 and 6. It should be clear that the exemption from the requirements in chapter 5 and 6 would not be nullified by the requirements in chapter 4, 7 and especially 8.

• Regarding section 5 on “Loan origination procedures”: there are no additional mentions to loans between affiliated parties. We only observe in article 89 a reference to the concept of “group of connected clients”, which we believe is related to lending facilities to an affiliate belonging to an economic group for which the provision of collateral is done by the group of “connected clients”. The EBA should therefore include the necessary amendments to adapt the content of section 5 so that lending to affiliated parties is also taken into account

• Please refer to the responses above for similar issues"
The overall Section 6 represents a considerable interference with the conduct of business rules. We could only understand this section if seen as a set of best-practices. This must be clarified in the text. Pricing (methodology and profitability calculation) should continue to be flexible and based on individual methods / approaches. Pricing is the result of the business strategy of each loan provider/bank and it follows his/its business objectives. To make a profit in short or medium term is the basic interest of any borrower and there is no need to specify the composition of the prices.
Further:
 the profitability can be defined on the transaction level, customer level or portfolio level so that the requirement stated in Art. 190 „All of the transactions below costs should be reported and properly justified.“ is completely inadequate.
 the Art 187 c. (“Operating and administrative costs resulting from cost allocation processes that involve all group entities”) doesn´t make sense since the operating and administrative costs don´t include costs of all group entities (i.e. “that involve all group entities” should be deleted).

• The expertise of each institution should not be overlooked, as well as the global customer relationship, and the risk of exclusion. We should avoid that a single distribution model is promoted by supervisory expectations. Moreover, the EBA requirements would be very difficult to implement when dealing with practices, for instance, such as promotional rates in consumer credit, as these do not necessarily cover the various components listed by the EBA, including the cost of refinancing. Generally, how pricing is calculated and organized does not belong to the scope of the regulator or supervisor in a competitive market. Chapter 6 under its current form ought to be erased. Although we understand that the list intends to provide only examples, supervisory expectations may still require to apply these lists.
• Regarding section 6 “Pricing”, EBA should also consider that the determination of prices between related parties must comply with the international rules on transfer pricing and BEPS, as well as the internal policies on Group Prices to which entities belong
• Paragraph 186 - We underline that the determination of margins is not done in a mechanical way but is the result of the process of bidding for a transaction, taking into account the risk of a given loan, and as well taking into account competition from other bidding financial institutions.
• In general we fear that the combination of Basel IV requirements and the requirements set in chapter 7 of this GL could push lenders to enter into lending activities with less collateral because of the administrative burden, or to remain competitive with other parties to whom these guidelines do not apply.

• In the Guidelines no distinction is made between cash flow-based lending and asset-based lending. It feels disproportionate to apply all requirements to cash flow-based lending in which the collateral is mainly held for fall back purposes. In our view, the strict requirements for valuation are also not needed to the same extent when the collateral is not considered eligible for internal managerial purposes (such as provisioning and pricing) and RWA calculation

• The Guidelines mention that the use of models at the stage of loan origination may create shortcomings in risk management. However, the concern of insufficient level of transparency and inadequate governance in relation to these methodologies employed can be relieved by requiring adequate mitigating measures regarding model performance and quality assurance. We consider therefore that there is no need to disallow the use of these models when they have proven to produce reliable results.
• We also concur with the argument that in the future, it is expected that the progress in information technology and development of large property and transaction databases will increase the precision of advanced statistical models. A strict restriction on the use of those models can hamper the development of this market and the overall progress of the valuation market.

• Paragraphs 191 to 200 - the required valuation of the collateral as defined paragraphs 191 to 200 is not compliant with article19 of the Mortgage Credit Directive. The MCD requires the valuation to respect specific standards when the lender decide to do the valuation, but this valuation is not requested for granting the mortgage credit. We also highlight that requirements of regular monitoring of the collateral valuation are already into force if institutions want the collateral to be recognized as a guarantee for prudential indicators computation, or for re-financing as a covered bond.
• Paragraph 192 - If the provisions of paragraph 201 allow movable property to be valued by advanced statistical models for a wide range of very different movable property types (e.g. vehicles, vessels, aircraft, industrial machinery, production lines, construction and agricultural machinery etc.), it would be suggested that such models are also allowed to be used for apartments and residential houses which homogeneous and stable markets
• More broadly, it is not clear why the CP does not allow for the use of advanced statistical models for valuation purposes at origination, even though these models produce reliable results. Disallowing its use would in our view does not contribute in making banks’ standards more robust. Instead, it would result in additional costs, without direct benefits to the client and the bank. For asset classes where advanced statistical models have proven to be reliable, we advise EBA to allow a continuation of its use. As pointed out in the consultation paper, advanced statistical models are already in place in several countries and are used as a good and credible source of information. When advanced statistical models provide valuation proven to be of the same quality as physical valuations, they have the benefit of capturing changes in the market prices and present them in an objective manner for the credit institutions. This is achieved on the basis of large databases of property transactions and public registries.
In addition, in those cases where national legislation allows for the use of other methods for valuation (e.g. model-based valuations), this should be allowed as national legislation supersedes the EBA guidelines.)

• Paragraph 194 – EBA seems to require that all immovable property collateral is assessed by an independent qualified valuer. This is not a market practice in some member states, where the purchase price, market price or construction price is allowed as valuation of the real estate, without a qualified valuer or statistical model.

• Paragraph 197 and 203 – The implementation of a panel of accepted external appraisers should be seen as a recommendable option to ensure the quality and independence of the valuation and the valuers but not as mandatory precondition for eligibility. The required quality could be ensured in different alternative ways. Inter alia for movable assets a panel of independent valuers is impracticable if possible at all due to lack of availability of qualified valuers and the heterogeneous nature of the collateral

• Paragraph 198 - An indemnity insurance is not market practice in every country and is not needed to ensure valuation quality. A mandatory insurance would increase costs that would be passed over to clients if not incurred by big valuation companies who could afford some extra expenses.

• Paragraph 199 – The Paragraph should be adjusted to accommodate the situations where the legal framework allows the debtor to deliver an independent valuation report. “Valuation should be carried out (internal valuation) or ordered (external valuation) by the institution, unless it is subject to a request from the borrower under certain circumstances” – seems to allow borrowers to choose the valuers, also if the responsibility for the real estate evaluation belongs to the lenders. This point should be clarified to avoid possible conflicts of interest

• Paragraph 204 – a reference to paragraph 2000 is wrong

• Paragraph 201 – 206 : Collateral valuation (a valuer, statistical models and indices) for movable property is not a common practice in most member states and we are not convinced of the added value of such valuation. In particular, the systematic valuation of movable property by appraisers is deemed neither feasible nor necessary given the volume and the additional associated cost generated, with limited benefits for most of the case. As a point of consideration, we propose that in respect of movable asset/property purchase financings, one may use as an appropriate initial value also the value of asset evident from an authentic document (e.g. value stated in the purchase contract or invoice), and not only the estimated market value of an asset determined by a certified asset valuer or calculated on the basis of an advanced statistical model

• Regarding moveable property collateral, in our view, the book value of collateral items can serve as the primary source of information. Financial data is – by nature – sufficiently reliable, as the processes that generate them are bound by numerous assurance measures, including independent external audits. Moreover, the appropriateness of these valuations can be evidenced (via back-testing) and it cannot be concluded that any other method (i.e. valuer or advanced statistical model) will result in more accurate valuations.
To further underline this, we would like EBA to clarify the scope of section 7.1.2. Where the requirements are relevant for the explicitly mentioned examples of moveably collateral (e.g. vessels, aircrafts, etc.), they are less relevant for others types of moveable collateral (in particular less standardized / comparable types as other means of transportation, livestock or inventories (sometimes frequently changing during the term of the loan) but also for commodities.
An example is the valuation with respect to leasing. A lease company generally finances on the basis of physical assets, and provides financing for the purchase amount of the asset. So, in principal a valuation is not deemed necessary, as the financing amount equals the purchase amount (whereby checks are performed on the market conformity of the purchase price).

• Paragraphs from 207 to 213 – the requirements would overhaul the current monitoring applied to collaterals subject to revaluation and the frequency of the update. Many banks have just modified their evaluation processes on the basis of the recent NPEs guidance. Any new changes would require high IT disburses and longer time for their implementation than what proposed in the guidelines. For example, performing full appraisals for revaluation purposes as set out in paragraph 213 instead of the current desktop ones, would significantly increase the appraisals’ annual cost, and delivery time could be delayed. Additionally, mainly in case of NPE, the debtor/asset owner wouldn’t permit an internal visit of the Real Estate asset. Moreover desktop or drive-by valuations will be mostly particularly useful and adequate for updating the previously established value of the real estate (i.e. on the basis of the most recent full visit with internal and external assessment of the property). Furthermore, statistical model-based revaluation (Paragraph 209 and 216) used by banks generally update the real estate assets value every 6 months. Also, the proposed parameters in para 208 to be used to structure the frequencies of monitoring are not necessarily the best. Market volatility and risk of deterioration regarding industry, technical infrastructure and location as well as respective market price developments are deemed more suitable. The institutions do have enough experience and market knowledge to judge on the best parameter reflecting the risk structure of their portfolio. Hence, parameter for determining different monitoring frequencies should not be predetermined by the EBA.

• Paragraph 208 - Asking for more frequent valuation of collateral in case of high LTV does not seem appropriate as generally a high LTV is granted for low risks assets.

• Paragraph 214 – This is difficult when working with external valuers. The question is also whether it relates to a valuer or an agency (2 different valuers of the same agency, or 2 different agencies?). The requirement for appraiser rotation applies already to non-performing loans via EBA NPL Guideline and the processes of banks have just been updated to accommodate for this new rule. The existing requirement for only NPLs is assessed as appropriate and an expansion to all exposures should not be pursued. It is essential in ensuring high quality valuations that the valuers obtain a deep knowledge of specific local real estate markets – hence the same immovable property could be monitored / revalued by the same valuer over a significant time period in order to maintain and enhance knowledge of the market. Therefore, we propose to remove this paragraph.

• Paragraph 219 - For standardized assets such as aircrafts, these assets don’t necessarily require a visit and external appraisers valuation of half-life assets are considered as sufficient.

• Paragraph 223 - The payment of valuers follows different approaches. This includes models where market price of the property is taken as indicator for the complexity and the needed effort for the valuation. This requirement should be deleted as it excludes customary pricing models that show no risk for the appraiser´s neutrality – also due to sufficient quality assurance. We recommend allowing member states using their current practices which are monitored and approved by the national supervisors.

• For immovable and movable collateral valuation must be independent, whether internal or external. The external valuations used are often indices (e.g. including statistics of notaries) or independent valuers, in some countries for outstanding amounts exceeding € 3 million, which does not seem compatible with the proposals of the draft GL. For movable collateral the requirement of the rotation of valuers is problematic. This change in market practices is particularly difficult to implement and would come at a huge additional cost for the activities affected.
• Overall, the ongoing monitoring proposed in the guidelines appears overly complex. This framework represents a burden that is not justified in relation to the average size of the banks' portfolio loans. The monitoring activity shouldn’t lead to undue additional reporting or disproportionate increase of the administrative obligations for banks, also considering that the requirements are already met through existing processes and reporting.

• Paragraph 229: Provisions according to which the credit risk monitoring framework and data infrastructure should provide the capability to gather and automatically compile data regarding credit risk without undue delay and with little reliance on manual processes are costly and problematic in terms of IT developments

• Paragraph 240: In several Entities the framework does not aim to monitor the initial recognition of the credit exposure (IFRS 9 concept) but its actual performance over time

• Paragraph 243: There are not Monitoring rules/signals based on the CRS; to be clarified if the reference is to the Monitoring of the Portfolio or single file.

• Paragraphs 256, 257 and 263 - they need to be framed by the proportionality principle, otherwise, in a transaction-by-transaction approach, there is the risk of burdensome procedures, information and reporting requirements. For para 263 is is to be clarified what has to be monitored at portfolio level (by Risk Management) or at single client level (by Credit Monitoring and/or Underwriting”)

• Paragraph 266 - The same applies to Paragraph 266 as otherwise the reporting will be unmanageable under the escalation process defined in the paragraph. As a general consideration we would suggest EBA to better specify whether and in which situation the Warning on Monitoring should be performed at portfolio level or at loan level. In particular, we deem important to clarify the supervisory expectations related to the watch list.

• The proposed approach to early warning indicators and watch lists does not seem suitable for adequately taking into account the different characteristics of institutions and customer groups. The requirements appear in many parts only practical for corporate customers.The monitoring requirements do not clarify what exactly is expected from business and/or risk. The level of monitoring covers both portfolio and individual borrower level, which is in practice the responsibility partly of business and partly of risk. Room shall be left to apply the requirements according to the own organization.

• Section 8.6: The requirements on Early Warning Signals expect that for each EWS that is triggered, a formal decision is taken. While we agree with the principle, we believe that relationship managers should be empowered to take any actions based on the Know Your Customer approach and submit the file to an appropriate credit committee if relevant. We suggest that this requirement should give some flexibility.

• As already mentioned in par. 76, there should be made a clear distinction with respect to responsibilities / roles of the credit function (1st line) and the risk function (2nd line) in the credit decision-making part. The risk function should not take an active role in the lending phase.
Lucia Pecchini
0032470908358