The Bank of England (BoE) has published a speech by Andrew Bailey, BoE Deputy Governor, Prudential Regulation, and PRA CEO, on the subject of bank capital adequacy.
In his speech, Mr Bailey draws out the failings of the old capital adequacy regime and why those failings were so critical, comments on the major progress made over recent years, and then draws out design questions that remain to be settled. In doing so Mr Bailey examines the key objectives of the regime being that it:
- creates good incentives for the behaviour of regulated firms;
- is appropriately forward-looking – at the risks in the future; and
- encourages and requires appropriate transparency which fosters market discipline.
Mr Bailey makes a number of points in his speech including the following:
- further work needs to be done on asset quality reviews;
- resolution and gone-concern loss absorbency is the key outstanding issue;
- a major principle of the new framework is that there is no single right approach to assessing capital adequacy as stress testing, risk-based supervision and the leverage ratio all have merits and weaknesses;
- using models and stress tests requires intensive development and maintenance by firms and a highly skilled body of supervisors, together with a regime where judgement can be used. Supervisors must have a credible capacity to withdraw the permission given to a firm to use a particular model if the model is considered inadequate or the firm has not demonstrated the capacity to use it safely. The long term consideration is whether the relevant risks can be modelled and how easy it is for supervisors and firms to observe and police the performance of models; and
- the question of whether supervisors should supplement the use of firms’ models by developing their own models is an important one. To date, in the UK, the approach has been to urge banks to improve their models, and to be more realistic about what their models can and cannot do. There may eventually be greater international convergence of supervisory approaches on the use of models.