We, as GEB, appreciate the overall balance of the proposal and in particular the limited weights assigned to RWA measures whose differences among banks and systems are only partially explained by differences in the risks posed to the DGS funds (on the subject, please refer to annexes).
Clear and transparent
We basically agree; a symmetric range (50%-150%) would have been more appropriate to allow an optimal calibration and to set an higher incentive for virtuous behaviors)
We basically agree but:
1-We ask for a better specification of NPL measure which, in our opinion, should be net of loan losses allowances as suggested also in the BIS proposal for a new standard approach in credit risk: a gross measure of NPL doesn’t take into account prudential policies set by individual banks on impaired loans which can counterbalance the country biases of the gross NPL measure. A lower weight for this single measure of the credit risk exposure would be also advisable to create room for a less “static” measure such as those assessing the credit loss absorbtion power implied in higher interest margins (our proposal is to use Net Interest Income or Net Revenues cyclically adjusted for credit cost).
2-Given the well known ongoing biases in the RWA framework, we propose to replace the RWA / Total Assets measure in “business and management” section by doubling the weight of the RoA measure. A RWA / Total Asset measure would thus be left to the discretional assessment by national authorities. Given the strong procyclicality of RoA, a cyclically adjusted RoA would also be preferable in our opinion to prevent underprovisioning.
We strongly prefer the capital coverage ratio: the CET1 requirements are dimension-related so ignoring this factor would bias the measure in favour of large banks (required to build up more capital in consideration of the higher risks posed to the system).
Maybe also the quantile specification of each bucket would have been useful at this level.