French Banking Federation (FBF)

We welcome the EBA initiative to issue guidelines on NPL management to be applicable to all banks in the EU. We also appreciate the rationale of distinguishing between high and low NPL banks.
However, as detailed below (question 2), we think that chapters 4 and 5 of the draft guidelines should not have to be automatically applied when a credit institution displays a NPL ratio above the threshold. In particular at portfolio level, the approach should rather rely on a supervisory dialogue, through the “comply-or-explain” process.
In its draft guidelines, EBA says that: “Credit institutions should apply chapters 4 and 5 when their level of non-performing exposures is elevated. An NPL ratio above 5% should be considered as an elevated level of NPEs” (§10). The wording also implies that NPL ratios are defined at a granular level and also that NPLs reduction measure should be automatically taken when the threshold are triggered.
First, we think that the ratio should only be indicative of the need for more scrutiny by the bank senior management and the Competent Authorities. Let us just say that we see a contradiction in the fact of calculating it both at the consolidated level and at the portfolio level: a bank can have a portfolio with a NPL ratio >5% because this portfolio is related to a business activity where defaults are more frequent (consumer finance, for instance) and at the same time, its consolidated NPL ratio would be < 5% because the other business activities are less risky. There would not be any legitimate reasons for the bank to apply chapters 4 & 5 of the guidelines to the portfolio which displays high level of NPL.
So, we think that looking at the NPL ratio alone (especially when it is computed without taking into account the level of coverage of doubtful loans by provisions ) is not enough to require a credit institution to apply chapters 4 and 5 of the draft guidelines. Competent Authorities also need to take into account specific features of the credit institution or the portfolio of loans. Indeed some activities have “by definition” a business model that demands high rates of gross NPLs without necessarily bearing high losses: loan restructuring department, consumer finance, social housing, etc. The realized loss, i.e. the net amount lost, is the true parameter to be considered (no need to remind that the Basel Committee combines the probability of default (PD) and the loss given default (LGD) to really assess the risks at stake).
For example, in the case of residential loans to borrowers with low incomes, probability of default is of course higher when compared to average-income borrowers (and in addition own initial contribution from low income borrowers are limited or even absent). The gross NPL ratio is therefore expected above 5%. However, taking this data alone would be completely misleading. There are indeed multiple remedies to address this issue and the actual losses experienced by the lender are quite low. That is thanks to the guarantee scheme put in place at origination. In France for instance, banks distributing housing loans to borrowers with low incomes (and which meet certain eligibility criteria) can benefit from a State-guaranteed scheme called FGAS. If the value of the loan is > 15.000 €, then a first-rank lien must be taken out by the lender. It means that, at the end of the day, those kinds of loans are twice guaranteed: first by the mortgage and second by FGAS. As a result, “expected losses” are quite low (as the LGD is low, while the PD is higher). So, in that case, the NPL ratio isn’t the relevant feature to be considered by Competent Authorities.
It is therefore necessary to watch for net ratio as well as expected losses over the long run: in France, historical series from specialized brands show that these client categories do not suffer high losses. This shows that such business model, when properly managed, beats the predictions of the “usual” and in fact partial return indicators.
Finally, it is not useless to remind that the business model described here is necessary to continue serving clients that would not have any access to credit if they had to turn to the “usual” financing providers. If such business models were to strictly apply the rules elaborated by the EBA, then credit institutions would be required to elaborate and execute a plan to reduce the NPL ratio of the portfolio below the threshold, which would be counterproductive: since the Expected Loss is low, margin are also relatively low, and operating a NPL portfolio according to the principles described in the EBA guidelines would put additional pressures on the operational margin of the banks.
To sum up, we think that chapters 4 and 5 of the draft guidelines should not have to be automatically applied when a credit institution displays a NPL ratio above the threshold. Credit institutions should have the possibility to provide explanations to competent authorities as regard their portfolio with an elevated level of NPL.
No specific remarks.
For comparability purposes and effectiveness of NPE reduction strategies, the breakdown between “past due” and “Unlikely to Pay” (UTP) needs to be taken into consideration. Some of the criteria for the UTP part may not be sustained from a contractual, legal or judicial execution point of view which may constitute an effective impediment to NPE reduction.
No specific comments.
No specific comments.
No specific comments.
No specific comments.
In its draft guidelines, in the chapter about collateral valuation for immovable and movable property, EBA requires the use of a property specific appraisal (individual property valuation) instead of an indexed valuation when the gross carrying amount of a NPE is > EUR 300,000: “indexed valuation should not be used to update the valuation of immovable property collateral for a NPE the gross carrying amount of which is larger than EUR 300,000 or a lower amount defined by the compentent authority”.
We think that this threshold is too low, and that the rule will substantially increase the cost of operating non-performing exposures, and we would recommend both to increase the threshold (in order to take into account the high value of real estate in great European cities) and to require a less frequent individual property valuation (for instance, banks could use indexed valuation every year and property specific appraisal every 3 years).
No specific comment.
Damien VAUDE