Our members’ prime focus remains housing finance (for owner occupation) so we concentrate on the proposed treatment of residential real estate. We have related interests in credit conversion factors, credit risk mitigation, and also the treatment of bank exposures – since our members both take and place funds from/ in the wholesale market.
We strongly support the overall objective set by the G-20 GHOS for the BCBS in this, and related exercises, of not causing a significant overall increase in the level of capital requirement. But we doubt whether the calibration in the second consultation will achieve this: rather it will push up capital requirements for standardised approach users, and widen (subject to the imposition of any capital floors) the gap with capital requirements derived from IRB approaches.
Many of the standardised approach risk weights in the second consultation are still far too high. The resulting increase in Pillar 1 RWAs, as well as directly increasing the Pillar 1 capital requirement, also increases the Pillar 2 buffer requirements, as these are based on RWAs. And so far – in the absence of any calibration of IRB capital floors – the widening gap with IRB outcomes has anti-competitive effect by favouring large incumbent banks. This discourages smaller lenders, including many mutuals, for whom IRB is not a realistic option.
Excessive standardised approach risk weights, combined with adverse features such as the insistence on retention of LTV at origination, and the loss of loan splitting ("tranching"), make this problem most acute in the residential real estate area. Indeed, the effective RWs for high LTV mortgage loans are also far too high in relation to actual risk when compared with the RWs for unsecured retail, SME and corporate portfolios.
to read the response in full.