Response to consultation Paper on draft Guidelines on loan origination and monitoring.
Go back
A second level check on the credit risk assessment would already be undertaken according to the principle established by the Guidelines and is also currently undertaken in the context of small ticket leasing with vendor programmes. In these programmes a “four eyes” principle is applied to grant an independent judgment of credit risk. The risk management function evaluates the risk assessment and the commercial function takes the final decision on the basis on both client creditworthiness and business evaluations.
The CCD, supported by the European Court of Justice’s case-law (e.g. Case C-449/13), provides lending institutions with a degree of flexibility to take due consideration of the specific credit application at hand and the operational reality in the various markets, e.g. relating to the availability of data, etc.
The Court recognised that the CCD “affords the creditor a margin of discretion for the purposes of determining whether or not the information at its disposal is sufficient to demonstrate the consumer’s creditworthiness and whether it is necessary to check that information against other evidence”.
There is a real danger that this flexibility would be undermined by the Guidelines in their current form. The application of the proportionality principle needs to be clearly emphasised, particularly for consumer-related activities, and further enabled by amending the wording of Guidelines in relation to the data to be assessed, removing overly commanding words such as “at least” and replace them with more appropriate terminology such as “if relevant”, which is more adequate within the context of non-binding guidance.
Whilst the information listed under Annex 2 of the Guidelines may be relevant for the assessment of an applicant borrower’s creditworthiness, this may not hold true for all lending activities and situations. In line with the relevant provisions of the CCD and supporting case law, discretion should remain with the lender based on their expertise. The collection of the full set of data, as prescribed in the Guidelines, would create major practical and technical challenges for large-scale consumer credit operations. Moreover, it would also raise concerns as to the necessity of the data and, thus, the institution’s adherence to relevant data protection obligations. It should be further clarified that the annexes set out possible indicators, and not strict requirements.
A very diverse landscape exists across Europe for credit bureaus as well as other public registers, with regards to their existence, organisation and the type of data held. Indeed, in some Member States it is legally not possible to establish a credit register with both positive and negative credit data. Some obligations provide for substantial operational challenges beyond the control of lending operators. For example, point 11 of Annex 2 calls for the use of “data from credit registers or credit information bureaus, covering at least the information on financial liabilities and arrears in payment”. Indeed, while the use of external sources can be encouraged, one cannot rule out that in certain circumstances this is simply not relevant or necessary. Additionally, it may not necessarily be compatible with the available resources for lenders.
We would also like to stress the principle of responsible borrowing, echoed in existing legislation, and that the customer should remain responsible for information that he/she is best placed to provide, e.g. on their tax status or whether the borrower has loans with other providers. We should reiterate that not all credit registers in Europe, whether held by public or commercial entities, provide this information.
The concept of a single customer view is generally favoured in the draft Guidelines. However, we want to stress that, although information may be collected by the institutions during the loan origination process, or existing within a company group, this data might not be necessarily be available on a “main” system. In practice, there may be a series of different single customer views within different entities of a group; but not centrally available without significant adjustments to relevant IT systems.
In this context, we also want to raise practical issues with verification of information with third parties. From both a legal and practical view point, it raises a number of potential issues, partially due to the general reluctance for processing of personal data for fear of potential non-compliance with the GDPR, irrespective of whether this fear is well-founded. In practical terms, this processing may prove difficult to carry out. In our view, the EBA’s reference to “reasonable” in paragraph 88 should be interpreted to acknowledge this potential situation and, thus, not forcing such a demand when not possible, e.g. when the data subject does not provide his / her consent for such processing.
We also want to suggest that the final guidance, in relation to consumer credit origination, provide for an opportunity to base the assessment by the means of statistic tools, e.g. rating and scorecard, to evaluate the ability of a borrower to meet its obligations, and, thus, limit the obligation to collect and verify information to situations over a defined threshold, etc. As outlined above, this would be in line with consumers’ expectations in the digital world as well as well-established and proven models in place.
Paragraph 90 of the Guidelines sets out a requirement for the documentation of relevant processes. We seek further clarification as to the precise obligations. A duty for the lender to systematically and fully collect and archive documents testifying the borrower’s declared information for low-value lending activities, e.g. consumer credit at the point of sale or online, would not be compatible with a proportionate approach and also surpass the CCD’s requirements.
We welcome the clarification in paragraph 91 that information on the purpose on the loan should be collected and verified “where relevant for the type of product”. For the majority of consumer credit products, e.g. revolving credit, personal loans, overdraft, etc., the gathering and processing of such data is not necessary where the granted amounts are not significantly above the average. Moreover, it is also unclear how such information would be clearly processed, valued and acted upon by a lender.
We want to stress the need for the necessary flexibility to ensure a general applicability of the Guidelines to all product categories, not least commonly distributed consumer credit products. The ability to adjust processes and steps with the characteristics of the individual situation and with the varying local operational realities, is necessary for a proportionate approach and is reflected in the relevant product legislation.
As previously outlined, the legislator provided for a degree of flexibility in the CCD in light of the varying market conditions and operational challenges differing across the Member States, further confirmed by Court of Justice, and for the differentiation of the creditworthiness assessments to be carried under the CCD and the MCD, respectively.
The current EBA guidance on the assessment to be carried out in relation to a mortgage loan (GL/2015/11) also calls for reasonable steps and enquiries to be fulfilled in order to assess an applicant borrower’s income or ability to meet the financial obligations arising from a loan agreement.
The proposed Guidelines do however provide for a considerably more static approach, with prescriptive and detailed steps to be taken, beyond the requirements set out in both the CCD and the MCD. Given that lending practices vary depending on the type of product as well as local market characteristics, we believe that the Guidelines need to appropriately reflect this and be phrased in a sufficiently broad and flexible manner.
A strict and formalised application of the listed metrics and parameters would require an overhaul of lenders’ existing scoring models, despite the fact that their efficiency and relevance have been proven. The Guidelines would ultimately entail a strict and rigid framework that would be disproportionate in relation to the risk reduction that could be expected, increasing costs and diminishing access to appropriate credit products.
The proposed focus of sensitivity analyses risks to lead to onerous, uncertain and intrusive measures. Such projections are difficult to carry out with any great certainty and may require far-reaching and intrusive processing to asses matter beyond than fairly uniform, i.e. likely retirement age/effect, etc. This raise concerns as to the practical application, but also for the interplay with the GDPR. For low value loans of short durations, such measures are generally considered unnecessary and would raise compliance issues. We believe that the requirement should be eliminated or be further clarified as to extent, and as a “best effort” measure, which, in line with the proportionality principle, may not be relevant for all credit activities.
It is our view, the assessment of an applicant borrower’s creditworthiness should provide for the relevant flexibility taking due account of the type, value, risk and complexity. With the proposed wide scope of the Guidelines, this is of even greater importance, to ensure the practical application across the various product categories covered, and to avoid unnecessarily burdensome requirements exceeding the complexity and risk of the relevant loan products.
Additionally, as previously outlined, a comprehensive financial view of a borrower is not always possible, e.g. due to internal IT limitations and/or varying access to, and content of, relevant credit databases.
5. What are the respondents’ views on the requirements for governance for credit granting and monitoring (Section 4)?
-6. What are the respondent’s views on how the guidelines capture the role of the risk management function in credit granting process?
A “one-size-fits-all” approach to the role of risk management, as set out in the draft Guidelines, is not appropriate. The level of risk undertaken when granting credit varies widely between organisations, depending on the size of the organisation and the nature of the financial activity being undertaken within it. In light of this, the requirements set out in paragraph 82 could be regarded as overly burdensome and prescriptive, since they require precise criteria for variable remuneration policies of the staff in charge of credit granting.A second level check on the credit risk assessment would already be undertaken according to the principle established by the Guidelines and is also currently undertaken in the context of small ticket leasing with vendor programmes. In these programmes a “four eyes” principle is applied to grant an independent judgment of credit risk. The risk management function evaluates the risk assessment and the commercial function takes the final decision on the basis on both client creditworthiness and business evaluations.
7. What are the respondents’ views on the requirements for collection of information and documentation for the purposes of creditworthiness assessment (Section 5.1)?
As stated above, in order to ensure the relevant flexibility and enable the effective application of the Guidelines to the many product groups and activities covered, we want to stress the importance of the full application of the proportionality principle in relation to consumer-related creditworthiness assessments, i.e. allowing for due consideration of the type, the value, the risk profile and the complexity of the individual transaction in the process.The CCD, supported by the European Court of Justice’s case-law (e.g. Case C-449/13), provides lending institutions with a degree of flexibility to take due consideration of the specific credit application at hand and the operational reality in the various markets, e.g. relating to the availability of data, etc.
The Court recognised that the CCD “affords the creditor a margin of discretion for the purposes of determining whether or not the information at its disposal is sufficient to demonstrate the consumer’s creditworthiness and whether it is necessary to check that information against other evidence”.
There is a real danger that this flexibility would be undermined by the Guidelines in their current form. The application of the proportionality principle needs to be clearly emphasised, particularly for consumer-related activities, and further enabled by amending the wording of Guidelines in relation to the data to be assessed, removing overly commanding words such as “at least” and replace them with more appropriate terminology such as “if relevant”, which is more adequate within the context of non-binding guidance.
Whilst the information listed under Annex 2 of the Guidelines may be relevant for the assessment of an applicant borrower’s creditworthiness, this may not hold true for all lending activities and situations. In line with the relevant provisions of the CCD and supporting case law, discretion should remain with the lender based on their expertise. The collection of the full set of data, as prescribed in the Guidelines, would create major practical and technical challenges for large-scale consumer credit operations. Moreover, it would also raise concerns as to the necessity of the data and, thus, the institution’s adherence to relevant data protection obligations. It should be further clarified that the annexes set out possible indicators, and not strict requirements.
A very diverse landscape exists across Europe for credit bureaus as well as other public registers, with regards to their existence, organisation and the type of data held. Indeed, in some Member States it is legally not possible to establish a credit register with both positive and negative credit data. Some obligations provide for substantial operational challenges beyond the control of lending operators. For example, point 11 of Annex 2 calls for the use of “data from credit registers or credit information bureaus, covering at least the information on financial liabilities and arrears in payment”. Indeed, while the use of external sources can be encouraged, one cannot rule out that in certain circumstances this is simply not relevant or necessary. Additionally, it may not necessarily be compatible with the available resources for lenders.
We would also like to stress the principle of responsible borrowing, echoed in existing legislation, and that the customer should remain responsible for information that he/she is best placed to provide, e.g. on their tax status or whether the borrower has loans with other providers. We should reiterate that not all credit registers in Europe, whether held by public or commercial entities, provide this information.
The concept of a single customer view is generally favoured in the draft Guidelines. However, we want to stress that, although information may be collected by the institutions during the loan origination process, or existing within a company group, this data might not be necessarily be available on a “main” system. In practice, there may be a series of different single customer views within different entities of a group; but not centrally available without significant adjustments to relevant IT systems.
In this context, we also want to raise practical issues with verification of information with third parties. From both a legal and practical view point, it raises a number of potential issues, partially due to the general reluctance for processing of personal data for fear of potential non-compliance with the GDPR, irrespective of whether this fear is well-founded. In practical terms, this processing may prove difficult to carry out. In our view, the EBA’s reference to “reasonable” in paragraph 88 should be interpreted to acknowledge this potential situation and, thus, not forcing such a demand when not possible, e.g. when the data subject does not provide his / her consent for such processing.
We also want to suggest that the final guidance, in relation to consumer credit origination, provide for an opportunity to base the assessment by the means of statistic tools, e.g. rating and scorecard, to evaluate the ability of a borrower to meet its obligations, and, thus, limit the obligation to collect and verify information to situations over a defined threshold, etc. As outlined above, this would be in line with consumers’ expectations in the digital world as well as well-established and proven models in place.
Paragraph 90 of the Guidelines sets out a requirement for the documentation of relevant processes. We seek further clarification as to the precise obligations. A duty for the lender to systematically and fully collect and archive documents testifying the borrower’s declared information for low-value lending activities, e.g. consumer credit at the point of sale or online, would not be compatible with a proportionate approach and also surpass the CCD’s requirements.
We welcome the clarification in paragraph 91 that information on the purpose on the loan should be collected and verified “where relevant for the type of product”. For the majority of consumer credit products, e.g. revolving credit, personal loans, overdraft, etc., the gathering and processing of such data is not necessary where the granted amounts are not significantly above the average. Moreover, it is also unclear how such information would be clearly processed, valued and acted upon by a lender.
8. What are the respondents’ views on the requirements for assessment of borrower’s creditworthiness (Section 5.2)?
As outlined above, the information and documentation mandated to be assessed in the creditworthiness assessment go further than the requirements set out by both the CCD and MCD, as well as established case-law. The current framework already provides for a strong focus on responsible lending.We want to stress the need for the necessary flexibility to ensure a general applicability of the Guidelines to all product categories, not least commonly distributed consumer credit products. The ability to adjust processes and steps with the characteristics of the individual situation and with the varying local operational realities, is necessary for a proportionate approach and is reflected in the relevant product legislation.
As previously outlined, the legislator provided for a degree of flexibility in the CCD in light of the varying market conditions and operational challenges differing across the Member States, further confirmed by Court of Justice, and for the differentiation of the creditworthiness assessments to be carried under the CCD and the MCD, respectively.
The current EBA guidance on the assessment to be carried out in relation to a mortgage loan (GL/2015/11) also calls for reasonable steps and enquiries to be fulfilled in order to assess an applicant borrower’s income or ability to meet the financial obligations arising from a loan agreement.
The proposed Guidelines do however provide for a considerably more static approach, with prescriptive and detailed steps to be taken, beyond the requirements set out in both the CCD and the MCD. Given that lending practices vary depending on the type of product as well as local market characteristics, we believe that the Guidelines need to appropriately reflect this and be phrased in a sufficiently broad and flexible manner.
A strict and formalised application of the listed metrics and parameters would require an overhaul of lenders’ existing scoring models, despite the fact that their efficiency and relevance have been proven. The Guidelines would ultimately entail a strict and rigid framework that would be disproportionate in relation to the risk reduction that could be expected, increasing costs and diminishing access to appropriate credit products.
The proposed focus of sensitivity analyses risks to lead to onerous, uncertain and intrusive measures. Such projections are difficult to carry out with any great certainty and may require far-reaching and intrusive processing to asses matter beyond than fairly uniform, i.e. likely retirement age/effect, etc. This raise concerns as to the practical application, but also for the interplay with the GDPR. For low value loans of short durations, such measures are generally considered unnecessary and would raise compliance issues. We believe that the requirement should be eliminated or be further clarified as to extent, and as a “best effort” measure, which, in line with the proportionality principle, may not be relevant for all credit activities.
It is our view, the assessment of an applicant borrower’s creditworthiness should provide for the relevant flexibility taking due account of the type, value, risk and complexity. With the proposed wide scope of the Guidelines, this is of even greater importance, to ensure the practical application across the various product categories covered, and to avoid unnecessarily burdensome requirements exceeding the complexity and risk of the relevant loan products.
9. What are the respondents’ views on the scope of the asset classes and products covered in loan origination procedures (Section 5)?
As set out above, we believe that the requirements set out in Section 5 of the Guidelines should be subject to the application of the proportionality principle, taking account of the characteristics of the specific activities. Such a uniform approach for the processes relating to loan origination is not practical nor proportionate. Consumer credit, and potentially leasing covered, activities are generally characterised by the large-scale distribution of small value/tickets of typically short durations, and cannot be compared to other, more complex lending activities.10. What are the respondents’ views on the requirements for loan pricing (Section 6)?
The parameters listed in paragraphs 186-190 of the Guidelines carry relevance both in relation to prudential and consumer protection measures. However, serving very different purposes, they are not suitable to form part of a “one-size-fits-all” approach. A strict requirement to consider the parameters on individual loan basis, would lead to pricing standardisation and commoditisation, thereby reducing the incentive to innovate. It effectively risks raising the cost of lending and increasing financial exclusion. Ultimately it would be very close to the parameters of what can be done in prudential supervision, and risks encroaching on the commercial judgement of operators. To achieve a more proportionate wording, we suggest to replace “inter alia” in paragraph 187 by “for example”.11. What are the respondents’ views on the requirements for valuation of immovable and movable property collateral (Section 7)?
The requirements for the valuation of immovable and moveable property collateral, as set out in the Guidelines, are not compatible with the business model of many leasing firms. Leasing firms’ employees are typically responsible for ascertaining the value of assets they fund, and this is particularly prominent in certain types of leasing. For example, this is very common practice in vendor financing, where finance is provided by equipment manufacturers, who are best placed to assess the value based on their market expertise. The requirements set out in section 7, requiring the party responsible for valuing the asset to be independent from the party financing the asset, would not allow for this practice to continue. In addition, performing immovable property full appraisals for revaluation purposes as set out in paragraph 213 instead of the customary drive-by appraisal paired with desk research, would significantly increase the appraisals’ annual cost, and likely delay delivery time.12. What are the respondents’ views on the proposed requirements on monitoring framework (Section 8)?
The processes set out by the Guidelines are elaborate and advanced, and of particular relevance in relation to complex and high-value lending activities. However, in relation to typical consumer credit activities they appear disproportionate and unjustified. Clarifications on a proportionate application of the requirements would be necessary.Additionally, as previously outlined, a comprehensive financial view of a borrower is not always possible, e.g. due to internal IT limitations and/or varying access to, and content of, relevant credit databases.