Response to consultation on ITS on amending Commission Implementing Regulation on benchmarking of internal models
Go back
We suggest keeping the reference to COREP for the F-IRBA portfolios or even to declare column 0120 for the F-IRBA portfolios as not required to be reported.
If this suggestion is not considered, clarification is required on how to deal with collaterals that collateralize multiple transactions and are assigned to different portfolios. In this case, should the market value of the collateral be reported multiple times or should a split take place?
Example: market value (MV) of the collateral = 1.000.000
portfolio transaction collateral collateralized exposure MV ??? version 1 (total MV) version 2 (proportional MV)
ABC K_123456 C_123456 200.000 ?? 1.000.000 250.000 ?? 1.000.000 250.000
DEF K_456789 C_123456 600.000 ?? 1.000.000 750.000 ?? 1.000.000 750.000
In addition, we ask for clarification that HGB (German Commercial Code) reporting institutions do not have to record the mentioned portfolios under number 31 of the consultation paper?
IFRS Q1: Do you see any issues or lack of clarity in the definition of the scope of the exercise?
No comments.IFRS Q2: Do you agree with the proposed list of benchmarking portfolios relevant for IFRS9? Do you believe that other dimensions should be used in the level 2 split? Do you have concerns on the alignment with the IRB benchmarking portfolios?
No comments.IFRS Q3: Do you agree with the proportionate approach taken for the geographical area envisaged by the exercise? How should the materiality thresholds be defined?
Regarding the materiality thresholds we propose to use the criteria according to the EBA stress test, see paragraph 103 to 106 of the “2023 EU-Wide Stress Test Methodological Note”.IFRS Q4: For the sake of allowing meaningful benchmarking observations, do you see any issue in not considering any combination of split at this stage? Or do you see merits in combining some dimension? If yes, which combination of split should be considered?
No comments.IFRS Q5: Do you see any issues or lack of clarity in the definition of the data points of template 115.00? Is the definition of IFRS 9 PD TTC/unconditional sufficiently clear?
Column 0120 “Collateral Value”: In our opinion, it is more appropriate to report the liquidation value of the collateral rather than the market value, in line with the ECL calculation under IFRS9 which also uses the liquidation value instead of the market value.IFRS Q6: Do you see any issues or lack of clarity in the definition of the data points of template 116.00 and 118.00?
No comments.IFRS Q7: Do you agree to the envisaged approach to collect the whole set of information only to limited subset of portfolios (L2 geographical split and aggregated asset classes)? Do you see any issue in reporting the PD curves?
No comments.IFRS Q8: Do you see any issues or lack of clarity in the definition of the data points of template 117.00? Would you see merits in collecting information on more granular quantitative triggers and relevant thresholds used for SICR assessment? If yes, in which ways?
Column 0120-0150: How is the "reporting period" defined?CR 1: Does the removal of the reference to COREP for the data field 0120 of templates C101, 102 and103 of Annex III as explained in paragraph 3 create the need to change your data submission?
The omission of the reference to COREP for column 0120 in the reporting templates C101, 102 and 103 from Annex III creates a need for adjustments in the reporting. For the F-IRBA portfolios, this change does not appear to be useful, as no LGD models are used in the F-IRBA, but rather LGD values prescribed by supervisory law. According to subsection 3.2.1 of the consultation paper, the change was explicitly suggested to be able to check the quality of LGD models.We suggest keeping the reference to COREP for the F-IRBA portfolios or even to declare column 0120 for the F-IRBA portfolios as not required to be reported.
If this suggestion is not considered, clarification is required on how to deal with collaterals that collateralize multiple transactions and are assigned to different portfolios. In this case, should the market value of the collateral be reported multiple times or should a split take place?
Example: market value (MV) of the collateral = 1.000.000
portfolio transaction collateral collateralized exposure MV ??? version 1 (total MV) version 2 (proportional MV)
ABC K_123456 C_123456 200.000 ?? 1.000.000 250.000 ?? 1.000.000 250.000
DEF K_456789 C_123456 600.000 ?? 1.000.000 750.000 ?? 1.000.000 750.000
CR 2: Do you agree that analysing the variability caused by deviating interpretation of eligibility of collateral should be analysed?
Which exposures should enter in the new portfolios 0214-0219 included in C103, only exposures in the scope of IFRS9 or the corresponding IRBA-HDP exposures? What should be done with exposures that are treated with alternative treatment and explicitly excluded from the CR SBP Exercise?In addition, we ask for clarification that HGB (German Commercial Code) reporting institutions do not have to record the mentioned portfolios under number 31 of the consultation paper?