Whilst the aim of the consultation is to reduce non-performing exposures (NPE), we are concerned that they proposed requirements could lead to conflict of interest issues to the detriment of customers, which would not be appropriate. This point is covered in more detail in the reply to Q2.
Business models impact NPL ratios
Different banks have different business models. Some more specialised banks provide loans to non ‘mainstream’ borrowers and provide an important service to some segments of society that typically have less access to finance. Such banks may have priced for higher default rates so their higher levels of NPLs do not necessarily pose a systemic threat to the financial system. For near/sub-prime lenders a 5% threshold may not be relevant, so we encourage the EBA to recognise the importance of proportionality and enable supervisors to set thresholds at higher levels, where appropriate, taking into account the bank’s strategy and appropriate levels of capitalisation. This would limit impacts of this guidance on better performing institutions. So, we recommend that the level of NPLs is considered holistically as part of a firm’s overall risk management and addressed through the regular dialogue between firms and regulators, using a firm’s risk appetite and stress testing framework to set a more meaningful target, recognising the differing product mix and risk profile of each firm.
Setting hard targets may be counterproductive
Experience is that any setting of targets brings the risk of unintended consequences. Management of exposures in difficulty is a key part of a bank’s duty of care to its customers and any target threatens this delicate relationship with clear conduct risks. This is especially the case if there are “time-bound quantitative targets” for reduction. We are concerned that setting such a threshold may encourage some institutions to target their NPL ratios, which may not be compatible with best practice with respect to NPL recognition and management. Further, if the measures also include forbearance accounts, then this might mean that deserving customers in financial distress would not be granted forbearance to prevent reaching the threshold.
The NPL ratio can be affected (in both directions) by changes in the denominator that might reflect good risk management of the overall business and have no significance for the appropriateness of the level of NPLs. Setting a materiality threshold in this way has potential implications, including:
o Sub portfolios > 5% - the threshold potentially sets a disincentive to establish a Run Off or Non-Core portfolio to address a pocket of non-performance, when that may be the best course of action to address issues promptly
o Readiness – whilst currently well under 5%, we recognise that the economic environment can change over time, making it more difficult for some of our clients to service their loans. Therefore, as part of being a prudent and conservative lender, firms may be bound to be compliant with sections 4 & 5, or at least have readiness plans in place, in case such an eventuality arises. It therefore does not necessarily reduce any compliance burden on firms, even if they are not the actual target of these.
o Inappropriate targets - for foreclosed accounts; waiting to foreclose / dispose of a property purely to meet a target (e.g. para 38 of 4.3.3) could erode customer equity in a declining house price market and would be detrimental. We note that paragraph 50 mentions staff being incentivised to reach targets set by NPE reduction strategy – historically these types of incentives have proved counter-productive to customer outcomes.
At what level should the threshold be set?
If the 5% target is maintained, then this should be set at the level of the overall firm to maintain the benefits of large diversified financial groups where the ability to lend is determined by the overall capital position rather than any specific (potentially small) portfolio.
We suggest that there should be a de minimus on loan size, as on high volume, low value cases these requirements could be quite onerous.
The proposed implementation date of 1st January 2019 is challenging as:
o The finalisation of requirements is unlikely until H2 2018, leaving limited time for banks to implement them;
o The complexity inherent in larger banks, with a range of underlying portfolios, will be subject to a range of different requirements.
o Additional clarification on final requirements is likely to be required, particularly in terms of requirements in Chapter 4 re setting strategy if banks need to comply due to high NPL ratio.
It should be made clear that the requirements under this article do not and are not intended to supplement nor amend the relevant Securitisation/Credit Risk Mitigation regulations currently within the CRR (in particular but not exclusive to CRR 119(5), CRR 243/244) and the Securitisation Regulation.
NPE Strategy (para 56) references the setting of incentives for NPE workout activities. It should be noted the utmost caution is needed in setting of individual incentives so as to not drive inappropriate behaviours and/or poor customer treatments. We believe treating customers fairly should take precedent over prudential aspects of managing NPLs.
Many banks already have operational structure, enabling them to comply with the spirit of the guidelines, but the requirements as documented are extensive if they were required to comply (i.e. NPL of 5%+), introducing a disproportionate compliance burden for banks. For some banks a one size fits all approach would be difficult to implement as producing and reporting MI against some of the requirements in the Guidelines would be challenging in relation to some portfolios the analysis of which is not highly automated. This could require time for development.
It should be made clear in section 188.8.131.52 that NPE workout units are a specific requirement only for the management of non-performing exposures, and asset disposals for any underlying security/collateral of the loans is not covered under this paragraph.
Care should be taken in the wording of Section 184.108.40.206 as the introduction of the concept of “avoidance of foreclosure” is generally not included within language of risk participation/credit risk hedging transactions and this requirement within the GL might introduce unwanted accounting implications. We recommend providing clarity that minimising expected losses is the overriding principal in this article and “avoidance of foreclosure” is only relevant insofar as legally required (in the case of consumer protection rules for example) or under as a secondary consideration to minimisation of expected losses.
In our view Section 220.127.116.11 contradicts most industry standard language included in guarantees, risk participation agreements and insurance contacts whereby the workout team should not consider the effects of the credit hedge in making decisions about forbearance, or recoveries in general. Rather to reduce the possibility of adverse selection against the investors in the credit hedge transactions, the beneficiary of the credit hedge is usually required to act as if the guarantee did not exist to maximise the recoveries in forbearance. Workout teams should still consider the effects of any collateral.
The guidance sets out a view that firms should distinguish between short and long-term forbearance measures, and whether forbearance measures are viable or not, together with suggesting monitoring metrics. In some circumstances this may be too prescriptive. For instance, it assumes a portfolio approach, which is not appropriate for a Commercial Banking portfolio where concessions are considered on an individual, case by case basis. Whilst generally we would comply with the suggested criteria noted in Section 6.1 there may be circumstances to go beyond those, in the best interests of both the client and the bank, or where, for example the ‘Information requirements for assessing viability of a forbearance measures’ (6.2.1c) could not be readily set out in advance.
Section 6.2.2 describes specific monitoring of forbearance measures. However, this anticipates the measures to be standard and comparable across a portfolio of obligors which is not considered appropriate for a Commercial lending portfolio, where each case is considered individually.
Similarly, Section 6.1.1 describes short-term vs long-term measures with different requirements and again we feel that this is overly prescriptive for a Commercial portfolio.
Section 6.2.2 describes the risk that an impairment may be masked, but again this is not reflective of how a non-portfolio exposure would be managed. Furthermore, it is not clear to us that portfolio level metrics of forbearance would be at all helpful where exposures are individually considered.
We note that Section 6.2.4 begins to recognise that there are differences between homogenous borrowers and more complex ones, but this is not reflected throughout the guidance.
As we note above the requirements are overly onerous for some portfolios where impairments are assessed on a bespoke, manual basis without the use of an NPV model.
We think that estimating future cash flows should be a matter for accounting guidance under IFRS9, not regulatory guidance, so the crossover/duplication with the accounting standard is not helpful. Specific timescales for instance are also considered counter-productive when considering a Commercial portfolio on an individual assessment basis.
Whilst the vast majority of properties above the proposed €300,000 threshold for property-specific valuation would be subject to individual specific valuation it is neither feasible nor necessary to complete physical valuations of all NPE properties, even introducing a threshold of €300k would result in significant volumes of physical valuations.
Our members do not wish to be bound by an arbitrarily set threshold which may be too low in some circumstances.
In our view Para 214 overstates the requirement of Article 208(3) of Reg (EU) No 575/2013 concerning monitoring property valuations. Its implication is that these should be updated annually. In reality the requirement is for annual internal review, and for external update every 3 years for property securing facilities greater than EUR3m. We understand the term ‘appraiser’ in para 201 to refer to an individual valuer within a Valuation firm rather than the Valuation firm itself, which may have one or many valuers. We believe the valuer rotation proposed in para 201 to be too prescriptive. In many cases, there is merit in maintaining the valuer as they have historic details and local knowledge, enabling them to assess any deterioration in condition. And there may also be cost implications for the customer. The competitive impact of this amendment on smaller Valuation firms should also be considered, as banks would probably move their business to larger multi-valuer firms to meet these valuation/re-valuation requirements. Therefore, a uniform rule on this is not considered appropriate and lenders should be free to make judgements in this area, linked to the specifics of the customer situation the level of risk, and the required abilities / expertise of the valuer.
We believe there should also be consideration of an LTV threshold. Increased frequency of valuations has the potential to increase the cost of managing NPEs leading to additional charges for customers. This cost is not justified for lower e.g. sub 50% LTV and could lead to unnecessary erosion of equity.
Many FBEs are only in short term financial difficulty so these need not be included in the requirement for a physical evaluation. For long term FBE where there is a strong rehabilitation performance hence we would disagree with imposing a physical valuation requirement on these collaterals.
In order to manage these potential issues, we recommend that physical valuations should be limited to collateral which is in repossession only, as these exposures are more likely be the result of no contact with a customer and/or their inability to engage in a repayment plan.
The annual valuation requirement for non-residential immoveable property is too short given that the sale process can take longer than this. We recommend that only in the event of a sale not being in progress (i.e. no offer post 1 year from repossession) would a further annual assessment be made.
Section 9.7.249 may be in conflict with some securitisation agreements, especially in the case of Reverse Mortgage Backed Securities (RMBS) or Commercial/Residential Development Commercial Mortgage Backed Securities (CMBS), where the originator may be required to complete renovations/construction/development on properties to realise market value for investors in the notes in an event of default of the borrowers. Noteholders (especially in a publicly offered RMBS/CMBS) would often chose this route despite it being less economically optimal (i.e. the land plus development cost is less than the market value of the developed property) due to the differences in liquidity of the underlying and the operational burden of the former.