Primary tabs

ASSOCIATION FRANCAISE DES SOCIETES FINANCIERES (ASF)

In France and in Brussels, the ASF represents specialized financial institutions focusing on credit, financial services and investment. These low credit risk activities are regulated and supervised either directly by the European Central Bank or by the ACPR (Prudential control and resolution authority) or by the AMF (Financial markets authority).
The ASF’s approximately 280 members are financing companies, credit institutions, specialized banks and investment companies employing 40 000 people. They cover a wide range of credit activities: leasing, factoring, guarantees, consumer and real estate loans, and also investment services and securitization.
The contribution of ASF members to the financing of the economy represents 247 billion euros in loans outstanding, or nearly 20% of the credits to the private sector in France. They are key players in terms of growth and employment, and some of them are pan-European leaders.
As a whole, the ASF consider that the EBA’s proposed Guidelines (GL) are too prescriptively worded, in a way that would not allow a flexible and risk sensitive implementation according to the proportionality principle as regards to product types, lending channels, amount of borrowing…
As they are drafted, the EBA’s draft Guidelines appear as standardized (“one size fits all”) whatever the nature and amount of credits originated and as to be comprehensively and strictly implemented by all European institutions, disregarding proportionality principle. We understand that it is not the objective of the EBA, which we strongly agree on since otherwise it would drastically raise the cost of credit and would create financial exclusion.

Against this background, the proposed wording about the proportionality principle is to be revised.
The EBA makes very general mention of proportionality in its introduction, proposing that the proportionality principle should apply as regards to the nature, size and complexity of the institutions referring to chapter 4 (governance requirements) and as regards to the type of products referring to chapters 5, 6, 7 and 8 (loan origination procedures, pricing, valuation of immovable and movable properties and monitoring). EBA proposes that the proportionality principle does not apply regarding consumer protection issues.
We consider that these references to proportionality in introduction to the GL are too light and general. It should be made clear that the supervisors may adapt their use of the Guidelines. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand that it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapters 4 to 8 and in Annexes 1,2 and 3. EBA expressed the fact that the requirements are to be understood as examples of requirements, allowing institutions to use only part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at points of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

For instance, we would propose the following wording for §14
14. “Institutions should apply section 4 of these guidelines in line with the proportionality principle described in Title I of EBA Guidelines on internal governance. Institutions should apply sections 5, 6, 7 and 8 of these guidelines including Annexes 1,2,3 in a manner that is proportionate to the type, size, nature complexity and risk profile of the credit facility.”


The ASF understands EBA references to its previous publications in the field of the 2014/17 Mortgage Credit Directive (MCD), and to the recent review of the European Supervisory Authorities that brings also, through an extension of the EBA activities to consumer protection, the 2008/48 Consumer Credit Directive (CCD) into its scope of action.

Two main reserves need to be expressed regarding these two level 1 legislative texts:

- Firstly, MCD and CCD are currently under evaluation. It is very probable that those directives will be revised by the European Commission in the coming months and it seems that the proposed GL should not anticipate these revisions, otherwise there would be a risk that the EBA regulation ends up inadequate to new legislative dispositions.
We also wish to underline that Guidelines are level 3 documentation supposed to respect the scopes defined by levels 1 and 2 legislative texts. This seems no to be the case in the field of consumer credits since for instance the current directive excludes form its scope the loans under 200€ whereas no mention of this threshold is made in the proposed Guidelines. In France, a level 1 law has introduced a threshold of 3 000€ for the collecting of documentation justifying borrower’s declarative information.
- Secondly, the EBA GL on creditworthiness assessment under MCD is presented as a basis to the proposed GL. Mortgage and guaranteed real estate credits are very specific, and we consider that the extension proposed to all kinds of loans, whatever the amount, distribution channels, counterparties consumers or professionals, leasing etc. is not relevant. These activities are very different and suppose different types of risks and business organisations. A single set of loan origination rules, in a “one size fits all” manner would be detrimental to risk sensibility and, finally, to the access to credit.

Regarding the scope of application, we also have a specific comment regarding the specialised business model of factoring.
It should be explicitly confirmed that the Guidelines do not include credit analysis (Origination & Monitoring) of debtors in factoring operations.
In the factoring activity, the factoring company purchases receivables from a client, and there is no contractual relationship between the factoring company and the debtors of the client (the debtors pay the receivables to the factoring company, or to the client by collection mandate). There is no loan or advance made to the benefit of these debtors. Advances are made to the clients of the factoring company and the contractual link is between clients and the factoring company.
In the unexpected case where debtors would be included within the scope of application, we suggest, at least, confirmation that analysis and monitoring of such debtors will follow lightened guidelines or, preferably, customized requirements, especially for sections 5 and 8. This would be in line with EBA work related to the treatment of factoring for AnaCredit where data required for debtors are limited compared to those required for clients.
As stated in the AnaCredit Reporting Manual – Part III – Case studies, chapter 5 Factoring : “ Meanwhile, in cases where the counterparty is the account debtor, some of the counterparty reference data required for AnaCredit reporting (i.e. data on the debtor) are not available to factors as there is generally no contractual relation between the factor and the debtor (i.e. the factoring client’s customer). Therefore, the following data attributes are reported as “Non-applicable” “.

More generally, many of the requirements proposed in the Guidelines are not compatible with large scale, small amounts and short-term credit origination (ex. consumer credits originated at points of sale or online, small lease tickets originated via vendor programmes, factoring programmes). Those types of business models rely on automatized and more and more on technology-based organisations and tools, currently duly supervised in respect of existing and already constraining regulation framework. For those concerned institutions that have demonstrated an accurate control of risks, the flexibility, adaptability, proportionality of the use of the Guidelines by the supervisors – as long as equivalent risk management efficiency is guaranteed – must be more precisely expressed. Otherwise a strict application of the requirements listed in Chapters 4 to 8 and in Annexes 1,2 and 3 would represent not only a regression in risk management but also huge investments in HR and IT systems disproportionate to any potential benefits in the cost of risk. In many cases, the very existence of the distribution channel would be at risk. It is the case for example of point of sale credit which characteristics (low amounts, high cost of acquisition) make it very vulnerable to any increase of administrative costs.

We also consider that the exclusion of “sovereigns, including central governments, regional and local authorities, and public sector entities” (§9) is to be further explained. We question the objective reasons of such an exclusion considering the history of default of some public entities.

We would finally like to underline a lack of definition concerning the types of loans: in the “definition” section (§17) the term « professionals » is defined as « non consumers » but the term « consumers » is not defined.
The ASF estimate that the proposed date of implementation – 30th June 2020 – is too close as regards to:
- The current evaluation processes of Consumer Credit Directive dated 2008 and Mortgage Credit Directive, which would lead to legislative initiatives not expected before end of 2020 at the best;
- The potential impacts induced by the proposed GL in terms of information system and business organisation within the institutions.

A delay compliant with the content of the Guidelines would be at least 18 months to two years.
Any shorter date of implementation could only be justified by an explicit rewording to better express the flexibility, adaptability, proportionality (at equivalent level of risk management efficiency) of the requirements listed in Chapters 4 to 8 and in Annexes 1,2 and 3.
The ASF agree on the importance of risk management function in credit granting processes. Sound credit granting policies and procedures are required, as well as a “top down” culture of risk within the institutions.
But the ASF consider the requirements for credit risk policies and procedures (Section 4.3), worded as they currently are, not future proof to many extents.

More specifically, ASF consider the requirements in relation to technology enabled innovation and environmental factors and green lending globally non future proof.

 Technology enabled innovation for credit granting:

“Technology neutrality” was clearly expressed by the EBA as an objective during the hearing held on September 20th. We strongly agree on this objective and would recommend expressing more explicitly the “technology neutrality” principle in § 47. It should be clearly worded that the prescriptions in this article would apply whatever the type of “technology-enabled innovation for credit granting” is used.
If some technology enabled procedures for credit granting were not in the scope of §47, but then in the scope of §35 and Annex 1, the proposed GL would be highly over-prescriptive, discouraging innovation and would even challenge existing technology-enabled procedures of creditworthiness assessment and credit granting procedures (for instance online credit origination) that have proved their efficiency. Credit risk policies and procedures described in §35 b), c) et d) referring to annex 1, with the term « at least » associated, are much too prescriptive and formal as regard to existing scoring models in small amount / short duration lending, either to consumers or to professionals, that have been largely automatized for a long time in most leasing, factoring and consumer credit institutions.
Changing these models, which have demonstrated their reliability, to strictly respect the proposed GL would be a backward step and would represent high costs in terms of IT developments and business / HR organisations. Moreover, as compliance to the proposed standardized and very formal requirements of §35 and Annex 1 would be a move backward in terms of risk control efficiency, these significant costs would be disproportionate in comparison of the potential benefits that could be expected in terms of costs of risk reduction.


 Green lending :

To start with, the ASF underline that the EBA mandate regarding green finance is mentioned in §5 but not very precisely. It would deserve to be more clearly explained. We understand that “the ESAs [are] being mandated to develop technical standards related to sustainability”, but green lending requirements do not seem to fit easily in the scope of the present Guidelines, considering that the latter refer to an EBA mandate in the context of an action plan against non-performing loans. Against this background, we highly question the draft Guidelines requirements in §48 to 53 concerning green finance, as respecting them would mainly determine whether a loan is green or not but would have little impact - or no impact - on its probability to be defaulted.

Besides, the requirements in §49, before and after the credit granting, are overly prescriptive and constraining. We understand that they could apply to large amounts project finance in the field of energy and environment, but they are not adapted to small scale lending activities in the scope of green lending such as defined §17 (for example, consumer credits distributed at point of sale for home energy-efficient equipment, small amount corporate loans and leases for building renovation, environmentally compliant equipment, or green vehicles financing etc.). Credit granting organisations and distribution channels on which those types of small amounts lending activities rely could not fit in the too standardised and prescriptive proposed Guidelines. The wording of the requirements should be more risk sensitive and allow more flexibility in the supervisors’ analyses. In this field, the proportionality principle should not be considered only as regard to the nature, size and complexity of the institution, but also to the type, nature, size and complexity of the credits and distribution channels.

Finally, the consolidated view required in §30 could raise an issue for subsidiaries of a banking group. IT systems of entities of a single group are not always connected with one another. As a result, a single view on a client at consolidated and sub-consolidated levels is not always technically possible.
The ASF agree on the importance of risk management function in credit granting processes. Sound credit granting policies and procedures are required, as well as a “top down” culture of risk within the institutions.
But the ASF consider the requirements for credit risk policies and procedures (Section 4.3), worded as they currently are, not future proof to many extents.

We consider that the reference to proportionality in introduction to the GL is too light and general. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand that it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapter 4 and in Annexes 1,2 and 3. EBA expressed the fact that the requirements are to be understood as examples of requirements, allowing institutions to use part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at point of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

As they are worded, the EBA draft GL requirements regarding credit risk policies and procedures are too prescriptive, formal and standardised whatever the type of loan (amount, duration, counterparty, complexity, distribution channel…). They do not take into account existing framework and organisations already implemented within institutions in respect of - already constraining - existing prudential rules. The proposed GL raise the issue of the coherence/articulation with existing prudential and regulatory tools, on the basis of which institution’s’ scoring models, internal organisations, data collecting processes… have long been developed.
The question of the articulation / integration of the proposed GL with(in) forthcoming legislative initiatives is also raised. For instance, concerning the Consumer Credit Directive currently under review, we notice that a Consumer Protection Cooperation workshop of the European Commission dedicated to the evaluation of the Consumer Credit Directive, dated June 2016, gathering competent supervisory authorities of several Member states, has listed guidelines on creditworthiness assessment. The ASF does not agree on some sections of the document but underlines the following comment included in the 1st guiding principle: « Creditworthiness assessments should follow the principles of proportionality and data minimisation in accordance with the data protection rules. This means that the use of data should be adequate, relevant for the purpose (i.e. only those type of data should be collected that are really needed to perform a creditworthiness assessment) and proportionate.”

Against this background, the ASF would wish more risk sensitivity and flexibility in the wording of the EBA proposed requirements for loan origination and the mention of thresholds of application introduced in level 1 texts (for instance € 200 for the Consumer Credit Directive).

We also consider that the requirements in §82 are too prescriptive, and even intrusive as regards the human resources (HR) policies of institutions, in requiring precise criteria for variable remuneration policies of the staff in charge of credit granting. In particular, §82 a., which provides that “variable remuneration of the staff involved in credit granting should be linked, among others, to the long-term quality of credit exposures », should be removed as it seems not applicable for consumer credit for instance:
- The long-term quality of the loan cannot be predicted or even controlled by the staff in charge of credit granting. They cannot be responsible for previous non-performing credits, nor for non-predictable events such as the evolution of the macro-economic context, or disease, unemployment, divorce, or other life accidents. Moreover, linking variable remuneration of staff to the long-term quality of the loan has no meaning for consumer credit with an average duration of 60 months.
- Variable remuneration indicators must comply with labour legal framework, under which in France for instance employees cannot be assessed regarding non-quantifiable and non-controllable objectives. Then, usually for this type of credits, the variable remuneration of staff is linked to credit risk through predictable risk indicators, for example as regards to the ratio of the amount of defaulted loans aver a 3 months period after the release of funds compared to the amount of loans granted over the same 3-month period.
The ASF agree on the importance of risk management function in credit granting processes. Sound credit granting policies and procedures are required, as well as a “top down” culture of risk within the institutions.
But the ASF consider the requirements for credit risk policies and procedures (Section 4.3), worded as they currently are, not future proof to many extents.

We consider that the reference to proportionality in introduction to the GL is too light and general. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand that it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapter 4 and in Annexes 1,2 and 3. EBA expressed the fact that the requirements are to be understood as examples of requirements, allowing institutions to use part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at point of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

As they are worded, the EBA draft GL requirements regarding credit risk policies and procedures are too prescriptive, formal and standardised whatever the type of loan (amount, duration, counterparty, complexity, distribution channel…). They do not take into account existing framework and organisations already implemented within institutions in respect of - already constraining - existing prudential rules. The proposed GL raise the issue of the coherence/articulation with existing prudential and regulatory tools, on the basis of which institution’s’ scoring models, internal organisations, data collecting processes… have long been developed.
The question of the articulation / integration of the proposed GL with(in) forthcoming legislative initiatives is also raised. For instance, concerning the Consumer Credit Directive currently under review, we notice that a Consumer Protection Cooperation workshop of the European Commission dedicated to the evaluation of the Consumer Credit Directive, dated June 2016, gathering competent supervisory authorities of several Member states, has listed guidelines on creditworthiness assessment. The ASF does not agree on some sections of the document but underlines the following comment included in the 1st guiding principle: « Creditworthiness assessments should follow the principles of proportionality and data minimisation in accordance with the data protection rules. This means that the use of data should be adequate, relevant for the purpose (i.e. only those type of data should be collected that are really needed to perform a creditworthiness assessment) and proportionate.”

Against this background, the ASF would wish more risk sensitivity and flexibility in the wording of the EBA proposed requirements for loan origination and the mention of thresholds of application introduced in level 1 texts (for instance € 200 for the Consumer Credit Directive).

We also consider that the requirements in §82 are too prescriptive, and even intrusive as regards the human resources (HR) policies of institutions, in requiring precise criteria for variable remuneration policies of the staff in charge of credit granting. In particular, §82 a., which provides that “variable remuneration of the staff involved in credit granting should be linked, among others, to the long-term quality of credit exposures », should be removed as it seems not applicable for consumer credit for instance:
- The long-term quality of the loan cannot be predicted or even controlled by the staff in charge of credit granting. They cannot be responsible for previous non-performing credits, nor for non-predictable events such as the evolution of the macro-economic context, or disease, unemployment, divorce, or other life accidents. Moreover, linking variable remuneration of staff to the long-term quality of the loan has no meaning for consumer credit with an average duration of 60 months.
- Variable remuneration indicators must comply with labour legal framework, under which in France for instance employees cannot be assessed regarding non-quantifiable and non-controllable objectives. Then, usually for this type of credits, the variable remuneration of staff is linked to credit risk through predictable risk indicators, for example as regards to the ratio of the amount of defaulted loans aver a 3 months period after the release of funds compared to the amount of loans granted over the same 3-month period.
The ASF agree on the importance of risk management function in credit granting processes. Sound credit granting policies and procedures are required, as well as a “top down” culture of risk within the institutions.
But the ASF consider the requirements for credit risk policies and procedures (Section 4.3), worded as they currently are, not future proof to many extents.

We consider that the reference to proportionality in introduction to the GL is too light and general. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand that it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapter 4 and in Annexes 1,2 and 3. EBA expressed the fact that the requirements are to be understood as examples of requirements, allowing institutions to use part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at point of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

As they are worded, the EBA draft GL requirements regarding credit risk policies and procedures are too prescriptive, formal and standardised whatever the type of loan (amount, duration, counterparty, complexity, distribution channel…). They do not take into account existing framework and organisations already implemented within institutions in respect of - already constraining - existing prudential rules. The proposed GL raise the issue of the coherence/articulation with existing prudential and regulatory tools, on the basis of which institution’s’ scoring models, internal organisations, data collecting processes… have long been developed.
The question of the articulation / integration of the proposed GL with(in) forthcoming legislative initiatives is also raised. For instance, concerning the Consumer Credit Directive currently under review, we notice that a Consumer Protection Cooperation workshop of the European Commission dedicated to the evaluation of the Consumer Credit Directive, dated June 2016, gathering competent supervisory authorities of several Member states, has listed guidelines on creditworthiness assessment. The ASF does not agree on some sections of the document but underlines the following comment included in the 1st guiding principle: « Creditworthiness assessments should follow the principles of proportionality and data minimisation in accordance with the data protection rules. This means that the use of data should be adequate, relevant for the purpose (i.e. only those type of data should be collected that are really needed to perform a creditworthiness assessment) and proportionate.”

Against this background, the ASF would wish more risk sensitivity and flexibility in the wording of the EBA proposed requirements for loan origination and the mention of thresholds of application introduced in level 1 texts (for instance € 200 for the Consumer Credit Directive).

We also consider that the requirements in §82 are too prescriptive, and even intrusive as regards the human resources (HR) policies of institutions, in requiring precise criteria for variable remuneration policies of the staff in charge of credit granting. In particular, §82 a., which provides that “variable remuneration of the staff involved in credit granting should be linked, among others, to the long-term quality of credit exposures », should be removed as it seems not applicable for consumer credit for instance:
- The long-term quality of the loan cannot be predicted or even controlled by the staff in charge of credit granting. They cannot be responsible for previous non-performing credits, nor for non-predictable events such as the evolution of the macro-economic context, or disease, unemployment, divorce, or other life accidents. Moreover, linking variable remuneration of staff to the long-term quality of the loan has no meaning for consumer credit with an average duration of 60 months.
- Variable remuneration indicators must comply with labour legal framework, under which in France for instance employees cannot be assessed regarding non-quantifiable and non-controllable objectives. Then, usually for this type of credits, the variable remuneration of staff is linked to credit risk through predictable risk indicators, for example as regards to the ratio of the amount of defaulted loans aver a 3 months period after the release of funds compared to the amount of loans granted over the same 3-month period.
A “one size fits all” approach for procedures covering loan origination (creditworthiness assessment and collection of information) is not relevant. Prudential requirement in loan origination procedures cannot be the same for large scale, small amount and short-term credits or leases as for a large amount project finance operation.

We consider that the reference to proportionality in introduction to the Guidelines is too light and general. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapter 5 and in Annexes 1,2 and 3. EBA expressed the fact that they are to be understood as examples of requirements, allowing institutions to use part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at points of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

The ASF wish the EBA to consider the following points referring to different sections of Title “5. Loan origination procedures”.

 Collection of information and documentation

§85 and 126 f) - A comprehensive view on borrower’s financial/credit position is not always possible. The IT systems of the different entities of a banking group might not be connected to each other. In addition, in some members states, such as France, there does not exist any positive credit register.

§88 à §90 - The wording should be clarified to determine more precisely what means for the creditor “the documentation of information”. We estimate that it 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. In addition, it would be overpassing the Consumer credit directive requirements, and some member states level 1 legislative texts: in France for instance, a level 1 law has introduced a threshold of 3 000€ for the collecting of documentation justifying borrower’s declarative information.
From an operational point of view, it would not be compatible with small ticket vendor lease to professionals either.

§91 to 94 and Annex 2 – The requirements concerning the collection of information for consumer credit activity seems to be far beyond what is currently admitted - and practiced - in application of the Consumer Credit Directive. Besides, the collection of some required personal data would have to be challenged against the background of the GDPR.
It would be technically impossible to collect some of the required data on a large-scale consumer credit activity. This could lead to the modelling of « budget types », with a risk of standardised creditworthiness assessment processes less risk sensitive than those currently used.
The wording of the EBA proposed GL for creditworthiness assessment can be understood as a “one size fits all” approach, whatever the type of loan (amount, duration, counterpart, nature, distribution channel, etc.). As such, it is not flexible enough and would eventually prevent institutions from adapting their creditworthiness assessment to distribution channels (at points of sale, online...) which require fluid and fast processes in response to « on site / on time » demands of clients who wants to buy goods or services.
Consumer credit worthiness assessment requires flexibility. Prudential framework is necessary but should be proportionate and adequate to the scale and finality of the operation.

Adequacy is also at stake as a strict application of the Guideline’s requirements would represent a regression in risk management for the concerned institutions that have demonstrated a permanent high-quality level in granting process. For instance, a very quick and automatized decision-making process is fully part of the good quality of small ticket lending as it allows selecting the customers in the first instance instead of having to deal with customers which have seen their financing request refused by competitors.

Annex 2 referred to in §91 to 94 – Annex 2 requirements are much too restrictive, as associated with the term “at least”. For some large-scale factoring, leasing or consumer credit activities, a strict application of the listed data requirements in Annex 2 would severely challenge current concerned institutions’ scoring models, sometimes technology-enabled, despite they have proved their efficiency.
Such a strict application would represent disproportionate costs in comparison to the benefits in terms of cost of risk reduction that could be expected.

Referring to our previous response to Question 1 we also wish to underline that the Guidelines and especially the Annex 2 should respect the scope of the consumer credit directive in explicitly excluding from their scopes the loans under 200€.


 Assessment of borrower’s creditworthiness

Requirements for lending to consumers

§99 – The ASF estimate that a strict application of this requirement would be much too prescriptive and would lead to regressive results if implemented as such. The section should be supressed or the term “where appropriate” should be more precisely developed.
A strict application of the listed metrics and parameters would severely challenge current concerned institutions’ scoring models, sometimes technology-enabled, despite they have proved their efficiency.
Such a strict application would represent disproportionate costs in comparison to the benefits in terms of cost of risk reduction that could be expected.
This too standardised and formal requirements would create a rigid creditworthiness framework detrimental to more risk sensitive existing processes. The proposed metrics and parameters would finally become factors of credit exclusion as they could become “standards” for indebtment ratios, without any reduction in the cost of risk.

§101 and §121 - We consider that the requirement to analyse potential negative scenarios in the future should be suppressed: by definition, unexpected events cannot be anticipated. This is the role of insurance activity and not credit activity.
Besides, such prospective analyses make little sense for small amount short term consumer credit for instance.
At least a wording allowing a more proportionate application should be introduced, with a requirement based on best efforts, and/or limited to large amount specific types of loans, but it cannot be a requirement for all credit activities.

§104 – A “third party verification” in order to document borrower’s income capacity is not always available. This requirement seems to be too formal and standardised.

§114 – We underline the extreme complexity to carry out “sensitivity analyses reflecting potential negative market and idiosyncratic scenarios in the future, including, for example, deterioration in the marketability of the immovable property, increase in vacancy rates, reduction in the rental prices for similar properties”.


§109 and §118 – The requirement to collect data on expenditures including “appropriate substantiation” and “consideration of the living expenses” should be challenged against GDPR framework.

For instance, we would suggest rewording the following paragraphs:

§85 “Institutions and creditors should have a sufficiently comprehensive view of the borrower’s financial position, including if relevant the most accurate and comprehensive view of all the borrower’s credit commitments (for example, if possible, the single customer view).”

§86 “Information and data should be accurate when they could be collected, timely and relevant to the asset class and specific product, and proportionate given the purpose, size, complexity, and potential risk associated with the loan. “

§91 “Institutions and creditors should collect and verify information in relation to the following:
a. purpose of the loan, where relevant for the type of product;
b. employment;
c. income;
d. financial commitments;
e. collateral (for secured lending); and
f. other risk mitigation factors, where available. “

§92 “For the purposes of the collection and verification of information, institutions and creditors should consider collecting, where appropriate the information and data as set out in Annex 2. Information and data contained in the Annex 2 are not binding and should be proportionate to the type, size, nature, complexity and risk profile of the loan.”

§93 “For the purposes of the creditworthiness assessment of professionals, institutions should collect and verify information in relation to the following, where it is relevant:
a. purpose of the loan, where relevant for the type of product;
b. income and cash-flow;
c. financial position and commitments, including assets pledged and contingent
liabilities
d. business model and corporate structure;
e. business plans;
f. financial projections;
g. collateral (for secured lending);
h. other risk mitigation factors, where available; and
i. product type specific legal documentation (e.g. permits, contracts etc.).
j. other relevant data and information »

§94 “For the purposes of the collection and verification of information, institutions should where appropriate consider collecting the information and data as set out in Annex 2. This Annex is not binding and must be proportionate to the type, size, nature, complexity and risk profile of the loan.”

§99 “Institutions and creditors should apply the metrics and parameters that are relevant from the perspective of assessing the individual borrower’s ability to repay the loan. Where appropriate, these metrics and parameters could include the following:
a.loan to income ratio;
b. loan service to income ratio;
c. debt to income ratio;
d. debt service to income ratio.
e. other relevant criteria and metrics.”

§100 “When relevant and technically possible, institutions and creditors should apply metrics and parameters to have an accurate single customer view that enables the assessment of the borrower’s ability to service and repay all its financial commitments.”

§101 “When assessing the borrower’s ability to meet obligations under the loan agreement, the institutions and creditors should carry out, if possible, analyses reflecting potential negative scenarios in the future, for example, a reduction of income; an increase in interest rates in the case of variable rate loan agreements; negative amortization; balloon payments, or deferred payments of principal or interest.”


Requirements for lending to professionals

§125 – We agree that the borrower’s payment capacity is a key feature of credit analysis, but the case of leasing is to be underlined as regards to the following formulation: « Collateral should only be considered as the institution’s second way out in case of default and not as the primary source of repayment, with the exception where the loan agreement envisages that the repayment of the loan is based on the sale of the property pledged as collateral”.
Indeed, in the case of leasing, whereby the financed asset is the property of the credit institution, all contracts rely on the ability of the lessor to dispose of the asset in case of non-payment of the loan. The property of the asset by the lessor is the main feature of a lease, and a central parameter for credit risk analysis and creditworthiness assessment .

§129- The requirement to analyse political, economic and legal contexts of foreign counterparties in cross border lending would not be compatible with some large scale / small tickets lending activities such as some factoring activities.

§130 – The assessment of climate-related and environments risks would not be possible in any contexts. The redaction is too standardised and would deserve a more proportionate approach. For example, it would not be suitable with the large scale / small tickets activities in the fields of leasing and factoring.

§131 à 135 et § 138 à 141 (SME specific) + Annexe 3 - We insist on the importance of an explicit rewording to better express the flexibility, adaptability – as long as an equivalent level of risk management efficiency is guaranteed - of the requirements listed in Chapter 5.
Otherwise, the ASF estimate that imposing to refer to a list of metrics to analyse a professional borrower’s financial position, especially considering SMEs, is over-prescriptive and regressive.
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 at least partially automatized and technology-enabled, that have long been developed and have proved their efficiency. It would represent a step backward to less risk sensitive credit granting analyses, for a disproportionate cost in terms of HR and IT in comparison with the potential expected reduction of cost of risk.
Besides, we underline that a very quick and automatized decision-making process is fully part of the good quality of small ticket lending as it allows selecting the customers in the first instance instead of having to deal with customers who have seen their financing request refused by competitors.

§ 142 to 156 - Requirements concerning sensitivity analyses are over-prescriptive, especially concerning « financial projections » (management evolution, analysis of the strategy, market events, etc.). These types of analyses are operationally not possible to produce for each borrower in the context of large scale, small amounts and short-term activities, such as part of the leasing or factoring industries activities.
As mentioned in §144, proportionality measures given the purpose, size, complexity, term and potential risk associated with the loan would at least be necessary, in a more precise manner, with an explicit wording within the Guidelines.
We particularly consider that the very precise requirement of §146d should only be performed at the institution’s level and not at the borrower level.

We also have a specific comment regarding the specialised business model of factoring.
It should be explicitly confirmed that the Guidelines do not include credit analysis (Origination & Monitoring) of debtors in factoring operations.
In the factoring activity, the factoring company purchases receivables from a client, and there is no contractual relationship between the factoring company and the debtors of the client (the debtors pay the receivables to the factoring company, or to the client by collection mandate). There is no loan or advance made to the benefit of these debtors. Advances are made to the clients of the factoring company and the contractual link is between clients and the factoring company.
In the unexpected case where debtors would be included within the scope of application, we suggest, at least, confirmation that analysis and monitoring of such debtors will follow lightened guidelines or, preferably, customized requirements, especially for sections 5 and 8. This would be in line with EBA work related to the treatment of factoring for AnaCredit where data required for debtors are limited compared to those required for clients.
As stated in the AnaCredit Reporting Manual – Part III – Case studies, chapter 5 Factoring : “ Meanwhile, in cases where the counterparty is the account debtor, some of the counterparty reference data required for AnaCredit reporting (i.e. data on the debtor) are not available to factors as there is generally no contractual relation between the factor and the debtor (i.e. the factoring client’s customer). Therefore, the following data attributes are reported as “Non-applicable” “.
A “one size fits all” approach for procedures covering loan origination (creditworthiness assessment and collection of information) is not relevant. Prudential requirement in loan origination procedures cannot be the same for large scale, small amount and short-term credits or leases as for a large amount project finance operation.

We consider that the reference to proportionality in introduction to the Guidelines is too light and general. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapter 5 and in Annexes 1,2 and 3. EBA expressed the fact that they are to be understood as examples of requirements, allowing institutions to use part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at points of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

The ASF wish the EBA to consider the following points referring to different sections of Title “5. Loan origination procedures”.

 Collection of information and documentation

§85 and 126 f) - A comprehensive view on borrower’s financial/credit position is not always possible. The IT systems of the different entities of a banking group might not be connected to each other. In addition, in some members states, such as France, there does not exist any positive credit register.

§88 à §90 - The wording should be clarified to determine more precisely what means for the creditor “the documentation of information”. We estimate that it 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. In addition, it would be overpassing the Consumer credit directive requirements, and some member states level 1 legislative texts: in France for instance, a level 1 law has introduced a threshold of 3 000€ for the collecting of documentation justifying borrower’s declarative information.
From an operational point of view, it would not be compatible with small ticket vendor lease to professionals either.

§91 to 94 and Annex 2 – The requirements concerning the collection of information for consumer credit activity seems to be far beyond what is currently admitted - and practiced - in application of the Consumer Credit Directive. Besides, the collection of some required personal data would have to be challenged against the background of the GDPR.
It would be technically impossible to collect some of the required data on a large-scale consumer credit activity. This could lead to the modelling of « budget types », with a risk of standardised creditworthiness assessment processes less risk sensitive than those currently used.
The wording of the EBA proposed GL for creditworthiness assessment can be understood as a “one size fits all” approach, whatever the type of loan (amount, duration, counterpart, nature, distribution channel, etc.). As such, it is not flexible enough and would eventually prevent institutions from adapting their creditworthiness assessment to distribution channels (at points of sale, online...) which require fluid and fast processes in response to « on site / on time » demands of clients who wants to buy goods or services.
Consumer credit worthiness assessment requires flexibility. Prudential framework is necessary but should be proportionate and adequate to the scale and finality of the operation.

Adequacy is also at stake as a strict application of the Guideline’s requirements would represent a regression in risk management for the concerned institutions that have demonstrated a permanent high-quality level in granting process. For instance, a very quick and automatized decision-making process is fully part of the good quality of small ticket lending as it allows selecting the customers in the first instance instead of having to deal with customers which have seen their financing request refused by competitors.

Annex 2 referred to in §91 to 94 – Annex 2 requirements are much too restrictive, as associated with the term “at least”. For some large-scale factoring, leasing or consumer credit activities, a strict application of the listed data requirements in Annex 2 would severely challenge current concerned institutions’ scoring models, sometimes technology-enabled, despite they have proved their efficiency.
Such a strict application would represent disproportionate costs in comparison to the benefits in terms of cost of risk reduction that could be expected.

Referring to our previous response to Question 1 we also wish to underline that the Guidelines and especially the Annex 2 should respect the scope of the consumer credit directive in explicitly excluding from their scopes the loans under 200€.


 Assessment of borrower’s creditworthiness

Requirements for lending to consumers

§99 – The ASF estimate that a strict application of this requirement would be much too prescriptive and would lead to regressive results if implemented as such. The section should be supressed or the term “where appropriate” should be more precisely developed.
A strict application of the listed metrics and parameters would severely challenge current concerned institutions’ scoring models, sometimes technology-enabled, despite they have proved their efficiency.
Such a strict application would represent disproportionate costs in comparison to the benefits in terms of cost of risk reduction that could be expected.
This too standardised and formal requirements would create a rigid creditworthiness framework detrimental to more risk sensitive existing processes. The proposed metrics and parameters would finally become factors of credit exclusion as they could become “standards” for indebtment ratios, without any reduction in the cost of risk.

§101 and §121 - We consider that the requirement to analyse potential negative scenarios in the future should be suppressed: by definition, unexpected events cannot be anticipated. This is the role of insurance activity and not credit activity.
Besides, such prospective analyses make little sense for small amount short term consumer credit for instance.
At least a wording allowing a more proportionate application should be introduced, with a requirement based on best efforts, and/or limited to large amount specific types of loans, but it cannot be a requirement for all credit activities.

§104 – A “third party verification” in order to document borrower’s income capacity is not always available. This requirement seems to be too formal and standardised.

§114 – We underline the extreme complexity to carry out “sensitivity analyses reflecting potential negative market and idiosyncratic scenarios in the future, including, for example, deterioration in the marketability of the immovable property, increase in vacancy rates, reduction in the rental prices for similar properties”.


§109 and §118 – The requirement to collect data on expenditures including “appropriate substantiation” and “consideration of the living expenses” should be challenged against GDPR framework.

For instance, we would suggest rewording the following paragraphs:

§85 “Institutions and creditors should have a sufficiently comprehensive view of the borrower’s financial position, including if relevant the most accurate and comprehensive view of all the borrower’s credit commitments (for example, if possible, the single customer view).”

§86 “Information and data should be accurate when they could be collected, timely and relevant to the asset class and specific product, and proportionate given the purpose, size, complexity, and potential risk associated with the loan. “

§91 “Institutions and creditors should collect and verify information in relation to the following:
a. purpose of the loan, where relevant for the type of product;
b. employment;
c. income;
d. financial commitments;
e. collateral (for secured lending); and
f. other risk mitigation factors, where available. “

§92 “For the purposes of the collection and verification of information, institutions and creditors should consider collecting, where appropriate the information and data as set out in Annex 2. Information and data contained in the Annex 2 are not binding and should be proportionate to the type, size, nature, complexity and risk profile of the loan.”

§93 “For the purposes of the creditworthiness assessment of professionals, institutions should collect and verify information in relation to the following, where it is relevant:
a. purpose of the loan, where relevant for the type of product;
b. income and cash-flow;
c. financial position and commitments, including assets pledged and contingent
liabilities
d. business model and corporate structure;
e. business plans;
f. financial projections;
g. collateral (for secured lending);
h. other risk mitigation factors, where available; and
i. product type specific legal documentation (e.g. permits, contracts etc.).
j. other relevant data and information »

§94 “For the purposes of the collection and verification of information, institutions should where appropriate consider collecting the information and data as set out in Annex 2. This Annex is not binding and must be proportionate to the type, size, nature, complexity and risk profile of the loan.”

§99 “Institutions and creditors should apply the metrics and parameters that are relevant from the perspective of assessing the individual borrower’s ability to repay the loan. Where appropriate, these metrics and parameters could include the following:
a.loan to income ratio;
b. loan service to income ratio;
c. debt to income ratio;
d. debt service to income ratio.
e. other relevant criteria and metrics.”

§100 “When relevant and technically possible, institutions and creditors should apply metrics and parameters to have an accurate single customer view that enables the assessment of the borrower’s ability to service and repay all its financial commitments.”

§101 “When assessing the borrower’s ability to meet obligations under the loan agreement, the institutions and creditors should carry out, if possible, analyses reflecting potential negative scenarios in the future, for example, a reduction of income; an increase in interest rates in the case of variable rate loan agreements; negative amortization; balloon payments, or deferred payments of principal or interest.”


Requirements for lending to professionals

§125 – We agree that the borrower’s payment capacity is a key feature of credit analysis, but the case of leasing is to be underlined as regards to the following formulation: « Collateral should only be considered as the institution’s second way out in case of default and not as the primary source of repayment, with the exception where the loan agreement envisages that the repayment of the loan is based on the sale of the property pledged as collateral”.
Indeed, in the case of leasing, whereby the financed asset is the property of the credit institution, all contracts rely on the ability of the lessor to dispose of the asset in case of non-payment of the loan. The property of the asset by the lessor is the main feature of a lease, and a central parameter for credit risk analysis and creditworthiness assessment .

§129- The requirement to analyse political, economic and legal contexts of foreign counterparties in cross border lending would not be compatible with some large scale / small tickets lending activities such as some factoring activities.

§130 – The assessment of climate-related and environments risks would not be possible in any contexts. The redaction is too standardised and would deserve a more proportionate approach. For example, it would not be suitable with the large scale / small tickets activities in the fields of leasing and factoring.

§131 à 135 et § 138 à 141 (SME specific) + Annexe 3 - We insist on the importance of an explicit rewording to better express the flexibility, adaptability – as long as an equivalent level of risk management efficiency is guaranteed - of the requirements listed in Chapter 5.
Otherwise, the ASF estimate that imposing to refer to a list of metrics to analyse a professional borrower’s financial position, especially considering SMEs, is over-prescriptive and regressive.
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 at least partially automatized and technology-enabled, that have long been developed and have proved their efficiency. It would represent a step backward to less risk sensitive credit granting analyses, for a disproportionate cost in terms of HR and IT in comparison with the potential expected reduction of cost of risk.
Besides, we underline that a very quick and automatized decision-making process is fully part of the good quality of small ticket lending as it allows selecting the customers in the first instance instead of having to deal with customers who have seen their financing request refused by competitors.

§ 142 to 156 - Requirements concerning sensitivity analyses are over-prescriptive, especially concerning « financial projections » (management evolution, analysis of the strategy, market events, etc.). These types of analyses are operationally not possible to produce for each borrower in the context of large scale, small amounts and short-term activities, such as part of the leasing or factoring industries activities.
As mentioned in §144, proportionality measures given the purpose, size, complexity, term and potential risk associated with the loan would at least be necessary, in a more precise manner, with an explicit wording within the Guidelines.
We particularly consider that the very precise requirement of §146d should only be performed at the institution’s level and not at the borrower level.

We also have a specific comment regarding the specialised business model of factoring.
It should be explicitly confirmed that the Guidelines do not include credit analysis (Origination & Monitoring) of debtors in factoring operations.
In the factoring activity, the factoring company purchases receivables from a client, and there is no contractual relationship between the factoring company and the debtors of the client (the debtors pay the receivables to the factoring company, or to the client by collection mandate). There is no loan or advance made to the benefit of these debtors. Advances are made to the clients of the factoring company and the contractual link is between clients and the factoring company.
In the unexpected case where debtors would be included within the scope of application, we suggest, at least, confirmation that analysis and monitoring of such debtors will follow lightened guidelines or, preferably, customized requirements, especially for sections 5 and 8. This would be in line with EBA work related to the treatment of factoring for AnaCredit where data required for debtors are limited compared to those required for clients.
As stated in the AnaCredit Reporting Manual – Part III – Case studies, chapter 5 Factoring : “ Meanwhile, in cases where the counterparty is the account debtor, some of the counterparty reference data required for AnaCredit reporting (i.e. data on the debtor) are not available to factors as there is generally no contractual relation between the factor and the debtor (i.e. the factoring client’s customer). Therefore, the following data attributes are reported as “Non-applicable” “.
A “one size fits all” approach for procedures covering loan origination (creditworthiness assessment and collection of information) is not relevant. Prudential requirement in loan origination procedures cannot be the same for large scale, small amount and short-term credits or leases as for a large amount project finance operation.

We consider that the reference to proportionality in introduction to the Guidelines is too light and general. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapter 5 and in Annexes 1,2 and 3. EBA expressed the fact that they are to be understood as examples of requirements, allowing institutions to use part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at points of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

The ASF wish the EBA to consider the following points referring to different sections of Title “5. Loan origination procedures”.

 Collection of information and documentation

§85 and 126 f) - A comprehensive view on borrower’s financial/credit position is not always possible. The IT systems of the different entities of a banking group might not be connected to each other. In addition, in some members states, such as France, there does not exist any positive credit register.

§88 à §90 - The wording should be clarified to determine more precisely what means for the creditor “the documentation of information”. We estimate that it 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. In addition, it would be overpassing the Consumer credit directive requirements, and some member states level 1 legislative texts: in France for instance, a level 1 law has introduced a threshold of 3 000€ for the collecting of documentation justifying borrower’s declarative information.
From an operational point of view, it would not be compatible with small ticket vendor lease to professionals either.

§91 to 94 and Annex 2 – The requirements concerning the collection of information for consumer credit activity seems to be far beyond what is currently admitted - and practiced - in application of the Consumer Credit Directive. Besides, the collection of some required personal data would have to be challenged against the background of the GDPR.
It would be technically impossible to collect some of the required data on a large-scale consumer credit activity. This could lead to the modelling of « budget types », with a risk of standardised creditworthiness assessment processes less risk sensitive than those currently used.
The wording of the EBA proposed GL for creditworthiness assessment can be understood as a “one size fits all” approach, whatever the type of loan (amount, duration, counterpart, nature, distribution channel, etc.). As such, it is not flexible enough and would eventually prevent institutions from adapting their creditworthiness assessment to distribution channels (at points of sale, online...) which require fluid and fast processes in response to « on site / on time » demands of clients who wants to buy goods or services.
Consumer credit worthiness assessment requires flexibility. Prudential framework is necessary but should be proportionate and adequate to the scale and finality of the operation.

Adequacy is also at stake as a strict application of the Guideline’s requirements would represent a regression in risk management for the concerned institutions that have demonstrated a permanent high-quality level in granting process. For instance, a very quick and automatized decision-making process is fully part of the good quality of small ticket lending as it allows selecting the customers in the first instance instead of having to deal with customers which have seen their financing request refused by competitors.

Annex 2 referred to in §91 to 94 – Annex 2 requirements are much too restrictive, as associated with the term “at least”. For some large-scale factoring, leasing or consumer credit activities, a strict application of the listed data requirements in Annex 2 would severely challenge current concerned institutions’ scoring models, sometimes technology-enabled, despite they have proved their efficiency.
Such a strict application would represent disproportionate costs in comparison to the benefits in terms of cost of risk reduction that could be expected.

Referring to our previous response to Question 1 we also wish to underline that the Guidelines and especially the Annex 2 should respect the scope of the consumer credit directive in explicitly excluding from their scopes the loans under 200€.


 Assessment of borrower’s creditworthiness

Requirements for lending to consumers

§99 – The ASF estimate that a strict application of this requirement would be much too prescriptive and would lead to regressive results if implemented as such. The section should be supressed or the term “where appropriate” should be more precisely developed.
A strict application of the listed metrics and parameters would severely challenge current concerned institutions’ scoring models, sometimes technology-enabled, despite they have proved their efficiency.
Such a strict application would represent disproportionate costs in comparison to the benefits in terms of cost of risk reduction that could be expected.
This too standardised and formal requirements would create a rigid creditworthiness framework detrimental to more risk sensitive existing processes. The proposed metrics and parameters would finally become factors of credit exclusion as they could become “standards” for indebtment ratios, without any reduction in the cost of risk.

§101 and §121 - We consider that the requirement to analyse potential negative scenarios in the future should be suppressed: by definition, unexpected events cannot be anticipated. This is the role of insurance activity and not credit activity.
Besides, such prospective analyses make little sense for small amount short term consumer credit for instance.
At least a wording allowing a more proportionate application should be introduced, with a requirement based on best efforts, and/or limited to large amount specific types of loans, but it cannot be a requirement for all credit activities.

§104 – A “third party verification” in order to document borrower’s income capacity is not always available. This requirement seems to be too formal and standardised.

§114 – We underline the extreme complexity to carry out “sensitivity analyses reflecting potential negative market and idiosyncratic scenarios in the future, including, for example, deterioration in the marketability of the immovable property, increase in vacancy rates, reduction in the rental prices for similar properties”.


§109 and §118 – The requirement to collect data on expenditures including “appropriate substantiation” and “consideration of the living expenses” should be challenged against GDPR framework.

For instance, we would suggest rewording the following paragraphs:

§85 “Institutions and creditors should have a sufficiently comprehensive view of the borrower’s financial position, including if relevant the most accurate and comprehensive view of all the borrower’s credit commitments (for example, if possible, the single customer view).”

§86 “Information and data should be accurate when they could be collected, timely and relevant to the asset class and specific product, and proportionate given the purpose, size, complexity, and potential risk associated with the loan. “

§91 “Institutions and creditors should collect and verify information in relation to the following:
a. purpose of the loan, where relevant for the type of product;
b. employment;
c. income;
d. financial commitments;
e. collateral (for secured lending); and
f. other risk mitigation factors, where available. “

§92 “For the purposes of the collection and verification of information, institutions and creditors should consider collecting, where appropriate the information and data as set out in Annex 2. Information and data contained in the Annex 2 are not binding and should be proportionate to the type, size, nature, complexity and risk profile of the loan.”

§93 “For the purposes of the creditworthiness assessment of professionals, institutions should collect and verify information in relation to the following, where it is relevant:
a. purpose of the loan, where relevant for the type of product;
b. income and cash-flow;
c. financial position and commitments, including assets pledged and contingent
liabilities
d. business model and corporate structure;
e. business plans;
f. financial projections;
g. collateral (for secured lending);
h. other risk mitigation factors, where available; and
i. product type specific legal documentation (e.g. permits, contracts etc.).
j. other relevant data and information »

§94 “For the purposes of the collection and verification of information, institutions should where appropriate consider collecting the information and data as set out in Annex 2. This Annex is not binding and must be proportionate to the type, size, nature, complexity and risk profile of the loan.”

§99 “Institutions and creditors should apply the metrics and parameters that are relevant from the perspective of assessing the individual borrower’s ability to repay the loan. Where appropriate, these metrics and parameters could include the following:
a.loan to income ratio;
b. loan service to income ratio;
c. debt to income ratio;
d. debt service to income ratio.
e. other relevant criteria and metrics.”

§100 “When relevant and technically possible, institutions and creditors should apply metrics and parameters to have an accurate single customer view that enables the assessment of the borrower’s ability to service and repay all its financial commitments.”

§101 “When assessing the borrower’s ability to meet obligations under the loan agreement, the institutions and creditors should carry out, if possible, analyses reflecting potential negative scenarios in the future, for example, a reduction of income; an increase in interest rates in the case of variable rate loan agreements; negative amortization; balloon payments, or deferred payments of principal or interest.”


Requirements for lending to professionals

§125 – We agree that the borrower’s payment capacity is a key feature of credit analysis, but the case of leasing is to be underlined as regards to the following formulation: « Collateral should only be considered as the institution’s second way out in case of default and not as the primary source of repayment, with the exception where the loan agreement envisages that the repayment of the loan is based on the sale of the property pledged as collateral”.
Indeed, in the case of leasing, whereby the financed asset is the property of the credit institution, all contracts rely on the ability of the lessor to dispose of the asset in case of non-payment of the loan. The property of the asset by the lessor is the main feature of a lease, and a central parameter for credit risk analysis and creditworthiness assessment .

§129- The requirement to analyse political, economic and legal contexts of foreign counterparties in cross border lending would not be compatible with some large scale / small tickets lending activities such as some factoring activities.

§130 – The assessment of climate-related and environments risks would not be possible in any contexts. The redaction is too standardised and would deserve a more proportionate approach. For example, it would not be suitable with the large scale / small tickets activities in the fields of leasing and factoring.

§131 à 135 et § 138 à 141 (SME specific) + Annexe 3 - We insist on the importance of an explicit rewording to better express the flexibility, adaptability – as long as an equivalent level of risk management efficiency is guaranteed - of the requirements listed in Chapter 5.
Otherwise, the ASF estimate that imposing to refer to a list of metrics to analyse a professional borrower’s financial position, especially considering SMEs, is over-prescriptive and regressive.
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 at least partially automatized and technology-enabled, that have long been developed and have proved their efficiency. It would represent a step backward to less risk sensitive credit granting analyses, for a disproportionate cost in terms of HR and IT in comparison with the potential expected reduction of cost of risk.
Besides, we underline that a very quick and automatized decision-making process is fully part of the good quality of small ticket lending as it allows selecting the customers in the first instance instead of having to deal with customers who have seen their financing request refused by competitors.

§ 142 to 156 - Requirements concerning sensitivity analyses are over-prescriptive, especially concerning « financial projections » (management evolution, analysis of the strategy, market events, etc.). These types of analyses are operationally not possible to produce for each borrower in the context of large scale, small amounts and short-term activities, such as part of the leasing or factoring industries activities.
As mentioned in §144, proportionality measures given the purpose, size, complexity, term and potential risk associated with the loan would at least be necessary, in a more precise manner, with an explicit wording within the Guidelines.
We particularly consider that the very precise requirement of §146d should only be performed at the institution’s level and not at the borrower level.

We also have a specific comment regarding the specialised business model of factoring.
It should be explicitly confirmed that the Guidelines do not include credit analysis (Origination & Monitoring) of debtors in factoring operations.
In the factoring activity, the factoring company purchases receivables from a client, and there is no contractual relationship between the factoring company and the debtors of the client (the debtors pay the receivables to the factoring company, or to the client by collection mandate). There is no loan or advance made to the benefit of these debtors. Advances are made to the clients of the factoring company and the contractual link is between clients and the factoring company.
In the unexpected case where debtors would be included within the scope of application, we suggest, at least, confirmation that analysis and monitoring of such debtors will follow lightened guidelines or, preferably, customized requirements, especially for sections 5 and 8. This would be in line with EBA work related to the treatment of factoring for AnaCredit where data required for debtors are limited compared to those required for clients.
As stated in the AnaCredit Reporting Manual – Part III – Case studies, chapter 5 Factoring : “ Meanwhile, in cases where the counterparty is the account debtor, some of the counterparty reference data required for AnaCredit reporting (i.e. data on the debtor) are not available to factors as there is generally no contractual relation between the factor and the debtor (i.e. the factoring client’s customer). Therefore, the following data attributes are reported as “Non-applicable” “.
The ASF agree on the importance of the parameters listed in §186 to 190 for the pricing of loans, especially in the field of prudential regulation and consumer protection.
But a requirement to consider systematically the listed parameters, for each single loan, would lead to “pricing” standardization and very formal conduct of business. It would be over-prescriptive and eventually would raise the cost of lending and lead to credit exclusion.
Such a requirement would be disproportionately intrusive in the ordinary course of business that concerns primarily the institutions’ management. It would be at the very limit of the prudential field since it would impose regulation on what is currently part of a commercial judgement.

We insist on the importance of an explicit rewording to better express the flexibility, adaptability – as long as an equivalent level of risk management efficiency is guaranteed - of the requirements listed in Chapters 6.

We also estimate that the requirement that “all transactions below costs should be reported and properly justified” (§190) is over-prescriptive.

For instance, we would suggest rewording the following paragraphs:

187. “Institutions should consider and reflect in loan pricing for example :…”

189. “Institutions should transparently document and review the underlying cost allocation framework. Institutions should establish a fair distribution of costs within the organization in order to ensure that business lines reflect the correct expected return corresponding to the risk assumed.”

190. “Institutions should implement a regular monitoring linking together transaction risk, pricing and expected overall profitability at business line level “

As mentioned by the EBA itself during the 20th September public hearing, it should be clearly expressed that: The guidelines do not prescribe any specific pricing strategies, as these remain business decisions of the institutions. »"
§ 191 à 200 – Concerning valuation of immovable property, the ASF consider that the proposed requirements regarding independent / external valuation of the properties are far over-prescriptive. They would considerably raise the cost of doing business and would finally lead to credit exclusion.
The question is raised of the coherence / articulation with the Mortgage Credit Directive, on the issue of property valuation.

§ 197 - The requirement to establish a panel of accepted external valuers is inapplicable in practice. For instance, as regards to the car market, there is only one single reliable external valuer for the market (ex. Argus in France or Autovista in Europe).

§ 199 that creates a requirement regarding the format of the valuation reports is over-prescriptive.

§ 201 – Same comment regarding the valuation of movable property, the ASF consider that the organisational requirements related to independence of valuers are over-prescriptive and would raise the cost of doing business.

§ 213 - Additional constraints for revaluation purposes on desktop (mainly) or drive-in (negligible) towards performing full appraisals would significantly increase the appraisal’s annual cost, as well as delivery time.
We consider that the reference to proportionality in introduction to the GL is too light and general. From the exchanges during the EBA hearing on the subject matter on September 20th, we understand that it is not the idea of the EBA to impose a comprehensive and strict respect, for all institutions, regarding all types of loans, of each of the very detailed requirements included in Chapter 8 and in Annexes 1,2 and 3. The EBA expressed the fact that the listed requirements are to be understood as examples, allowing institutions to use part of it depending on the context, or other equivalently efficient parameters and processes, as there is no point to issue universal requirements that would fit all types of lending and institutions in Europe. We strongly agree with this principle. For instance, small consumer loans with short maturities distributed at points of sales, as well as small factoring tickets and small amount leases, cannot possibly refer to the same requirements as those concerning large amount credits in project finance.
We exhort the EBA to express this view more explicitly in the drafting of the Guidelines so that national competent supervisory authorities find the margin of manoeuvre to agree sound governance, loan origination, pricing, valuation of immovable and movable properties and monitoring procedures and organisations as of equivalent value, even if they diverge from the strict list of requirements.
For instance, we notice a high use of the term “at least” or similar (eg : “should” interpreted as mandatory), which can be read as contradictory with any proportional application of the Guidelines. Using a wording such as “for example” and “where relevant” and “or equivalent” to introduce the requirements would probably ease a reasonable application of the proportionality principle.

§ 234 - A comprehensive view on borrower’s financial/credit position is not always possible. The IT systems of different entities of a banking group might not be connected to each other. In most cases, a comprehensive view could be achieved only at a central level but not in each subsidiary of a banking group.
In addition, in some member states, such as France, there does not exist any positive credit register, and the French Constitutional Council has explicitly expressed reserves on such disposal based on personal data protection considerations.
At least, such a requirement would have to be challenged in the context of GDPR.

§ 238 - The requirement to collect and monitor qualitative information on the borrowers (such as quality of management, company’s pricing power, cost structure, etc.) is too detailed and prescriptive. We understand that it could be relevant for large amount project/investment finance regarding large corporates, but it is not suitable with some “large scale/ small amount/ short term” leasing or factoring activities.
We insist on the importance of an explicit rewording to better express the flexibility, adaptability, proportionality – as long as an equivalent level of risk management efficiency is guaranteed - of the requirements listed in Chapters 8.

§243 – Regarding exposures on debtors in Factoring, that are not loans provided to these debtors but trade receivable, where events of dilution risk may generate delays in payment, strong requirements about past due relevant review should probably only focus on 90+.

§255 – Institutions including numerous subsidiaries should run the process of stress testing at banking group level. Where relevant, Banking group’s teams which achieve stress testing exercise may ask subsidiaries for specific data or inputs to perform the process.

§ 263 s – We understand that the requirement about 30+ past due borrower-related facilities is to be considered at individual institution level. In other words, an institution itself can’t review all borrower-related facilities in its banking group.
M. Yves-Marie LEGRAND
+33 (0)1 53 81 51 67