In the context of default of an obligor pursuant to the EBA the draft RTS proposes a materiality threshold to determine the materiality of a credit obligation according to option 1, whereby the obligor is already considered defaulted if either of the two limits is breached. Alternatively, a second option provides that default is to be assumed only in the event of breach of both the absolute and the relative limits.
In some EU Member States, including Germany and Austria, option 2 is currently the procedure prescribed by the supervisory authorities and applied by the institutions to identify the default of an obligor. It has to date correctly addressed the actual risks of default properly and has provided for an exclusion of the above-mentioned wrong handling of default risks. In this manner, the absolute component ensures the exclusion of technically caused defaults, which could not be countered alone by the relative threshold.
The EFBS is therefore of the opinion that a risk-adapted materiality threshold requires imperatively a coaction of the relative and the absolute limit.
Moreover, the concerns of EBA about a significant delay in the identification of defaults of high-volume credits in the framework of option 2 cannot be acknowledged. Article 178(1)(a) of the CRR lists further default criteria, such as in particular an “unlikelihood of payment”, which ensure that there is no delay in identification of default.
The Bausparkassen therefore prefer option 2, according to which default would occur only if both thresholds are breached in combination. Option 1, on the other hand, is strictly rejected.
Additional comments on the definition of credit obligation past due in the context of the absolute component
In this context EFBS would also like to remark that it is not understandable why EBA wants to determine in article 2(2)(a) of the draft RTS the absolute component as the limit to the sum of all amounts that are past due more than 90 days without establishing a reference to their materiality. Accordingly, a credit institution would need to treat an amount of for example one euro which has been past due for 90 days as a “credit obligation past due”. This alternative of definition is listed by EBA on page 23 of the draft RTS as option 3. Unlike this, an option 4 provides that only credit obligations which have exceeded the materiality threshold and then been overdue for 90 days can be classified as “credit obligations past due” in the sense of article 2 (2) (a).
EFBS fails to understand why EBA has chosen option 3 for the definition of credit obligations past due and rejected option 4. By this means, EBA contradicts to the intention of article 178 CRR which requires the obligor, in order to determine his default, to be past due more than 90 days with any material credit obligation to the institution. Obviously, this material credit obligation is identical with the due amount in arrears of the obligor. Consequently, as long as this amount in arrears is not material it should not be subject to the 90 days-counting and can therefore, in the understanding of EFBS, not be considered a credit obligation past due.
EBA has justified its decision for option 3 especially with the results of an inquiry among Member States according to which their majority applies this definition (cf. chapter 5 (a) (page 23) in the “definition of credit obligation past due”/ option 3). However, table 1 on page 21 shows that none of the Member States which have participated in the inquiry applies option 3. Yet, option 4 is the current practice in 8 Member States. In addition, on page 20 EBA is contradictory by stating that most of the Member States apply for the definition of credit obligations past due option 4 („Most common past due definition is “the sum of the amounts past due more than 90 days (or 180 days if applicable) but the calculation of days past due starts when the materiality threshold is breached.”).
EFBS assumes that the choice of option 3 is a mistake and suggests a thorough review and revision.
In Article 2(3) of the draft RTS, EBA proposes an absolute threshold of 200 euros for retail exposures and 500 euros for all other exposures. A relative threshold is to amount to a maximum of 2% of the total amount of all credit obligations of the borrower to the credit institution. Taking into account the expected impacts listed under Q5, the EFBS advocates the retention of a relative threshold of 2.5% which is already applied in some Member States.
According to article 2(1) of the draft RTS the national supervisory authorities should on basis of the EBA provisions set a threshold for all institutions in the respective Member State. Due to the transposition of these supervisory requirements the credit institutions will be subject to various adaptations in their operative systems. The workload for the institutions will naturally be dependent on the option chosen by the EBA under Q1. If the definition of default is amended, it is expected that implementation work will take one to two years for IRBA institutions. In particular, the following steps would be necessary which are presented in detail again under Q5:
- implementation of the amended definition of default in the production systems: three to nine months;
- adjustment of the identification of default in the historical data: four to eight weeks;
- impact assessment of the amended definition of default; massive changes to the historical default rate are expected especially for retail institutions: two to four weeks;
- complete redesign of all scorecards for the probability of default (PD) and the loss given default (LGD): four to nine months;
- repeat of the IRBA authorisation process, since an amendment to the definition of default brings about a material change to the assessment procedures: three to six months.
For institutions that use the standardised approach there will be at least the adjustment of the identification of default and corresponding consideration in the internal management, as well as impact assessments.
So far, the EBA has undertaken only a qualitative analysis of the proposed changes. It becomes clear from its explanations that the benefits of a change to the definition of default are limited in comparison with the expected costs. This is already shown by the length of the passages which the EBA has devoted to the respective explanations: four lines of comments on the benefits on page 30 and 1.5 pages of comments on the costs on pages 29 and 30.
The EFBS lists as benefits easier comparability of the thresholds by the European supervisory authority and the harmonised procedure when determining obligor default by cross-border credit institutions. The EFBS gets the impression that the focus of the paper is definitely on the latter, whereas the national supervisory authorities and the credit institutions which will continue to report to them in the future too are unable to draw any significant benefit from the RTS, but will be heavily burdened by implementation costs. As the EBA correctly states on page 15 in the explanatory text on question 3, the operational impact is substantial especially for IRBA institutions. This also applies for credit institutions that have applied to use the IBR approach pursuant to Article 143 CRR. The uncertainties they will experience in the application procedure as a result of the change in the definition of default are not however dealt with in the EBA paper.
There is no quantitative cost estimate yet in the EBA paper. The limits finally set by EBA are nevertheless decisive for the national supervisory authorities in determining the thresholds. If when analysing the specifications an imbalance arises at national level between the benefits of changing the definition of default and the associated costs, only limited room for action remains for the national authorities. They would not succeed in creating an excess of the benefit of the new regulation over the costs.
For this reason, the EFBS considers it essential for a quantitative impact assessment to be carried out before the thresholds are definitively determined by the EBA.
The EFBS expects substantial conversion efforts to arise for institutions if the definition of default is amended, especially in the following fields:
- The institutions must draw up time-consuming, costly impact assessments on the new definition of default. This is the case especially for IRBA institutions. However, this is likely to apply to a limited extent for the institutions that use the standardised approach as well, since on account of the resulting higher frequency of default, they must make adaptations with regard to provision for risks and risk-bearing capacity.
- By tightening up the default criteria, the probability of default would rise. In the event of amendment of the definition of default, also an interaction between PD and LGD could develop. Thus, the large number of artificial defaults caused, could allow LGD to decrease. Since however own estimates of loss given default for collateral in the form of property under the IRBA may not be lower than the limits of 10% and 15% respectively provided for in Article 164(4) and (5) CRR, a significant increase in the expected loss relevant for provision for risks is to be counted on, especially for credit institutions specialising in business secured by residential property, such as notably the Bausparkassen. An offsetting effect of “increasing PD against reducing LGD” is possible only to a limited extent as a result of the lower limit. We are asking EBA to consider this consequence in its remarks on the impacts of the materiality threshold on the possible default and thus the calculation of expected loss (EL) and own funds requirements.
- What the expected impacts will be on provision for risks and risk-bearing capacity and, in the case of IRBA institutions, on the calculation of the expected loss and necessary capital adequacy requirements depends on the portfolio of each individual institution. However, it is in any case certain that in individual institutions massive additional own funds requirements will arise.
- The EFBS is expecting a decisive impact on the estimates of risk parameters. The time-consuming and costly steps to be carried out by IRBA institutions and IRBA applicants in the context of the development of estimation processes must be repeated on the basis of the new definition of default. New parameter estimation processes must be developed.
- The estimation processes newly developed by the IRBA institutions must again be subject to the inspection for use of the IRB approach by the competent authorities in accordance with Article 144 CRR. The costs invested are on the same scale as the costs of the initial inspection for use of the IRBA. In addition, the delay in the completion of the inspection prior to the grant of authorisation also gives rise to additional costs for IRBA applicants.