Written by:
Simon Hills, Director, Prudential Policy, UK Finance


At the end of last week I submitted UK Finance’s response to the PRA’s consultation paper CP13/17. This is the second of three consultation papers that explain the PRA’s proposed approach to liquidity supervision in Pillar 2, which complements the Pillar 1 regulatory minima of the Liquidity Coverage Ratio (LCR) and Net Stable Funding Ratio (NSFR).

In particular CP 13/17 makes proposals about how Cash Flow Mismatch Risk (CFMR) should be addressed.

The LCR requires banks to have sufficient low risk High Quality Liquid Assets (HQLA) that can be sold to meet liquidity needs arising from cash outflows calculated over a 30-day period of liquidity stress.

CFMR is the risk that a bank has insufficient liquid assets and other liquidity inflows that could be sold to cover stressed liquidity outflows on an individual daily basis, in contrast to the LCR’s cumulative 30-day requirement. This peak liquidity need is a risk that banks should, and already do, assess. It arises as they perform one of their fundamental roles for society of maturity transformation, by borrowing from depositors for shorter periods such as overnight in the case of a current account balance, but on-lending those funds for longer periods – perhaps for a ten-year mortgage.

UK Finance’s response welcomed the PRA’s intention to address CFMR in Pillar 2 but challenged some of the assumptions about rates of outflow compared to the inflow realisation profile of HQLA that the draft guidance is proposing.

The supervisory assumption that all non-maturing deposits (NMDs) can be withdrawn at any time without advanced notice by the depositor is unrealistic. Depending on a bank’s business model, NMDs comprise nearly half of an ordinary retail bank’s deposits. The experience during the financial crisis demonstrated that that any retail deposit outflow is parabolic, starting as a trickle and building over the next week or so.  Similarly, the liquidity outflows from wholesale market transactions, for instance arising from investors exercising put options in debt securities issued by the bank, are unlikely to occur on day one of the stress, in part because they are subject to a two or three-day settlement cycle.  The industry’s preferred solution to the issue of liquidity outflows is to permit banks to model them, with the results being subject to regulatory scrutiny.

UK Finance’s response also expressed concern about the potential overlap between the CFMR framework and recovery and resolution planning.  This may happen because, as well as requiring the holding of HQLA above the Pillar 1 minimum, the PRA is also mandating banks to undertake enhanced liquidity assessments over a 90-day period of stress. This analysis should not result in extra HQLAs being held, which would be super-equivalent to internationally agreed standards, but be undertaken as a monitoring exercise as part of ongoing recovery and resolution planning.

The PRA will reveal its thoughts about calibration –  how much extra HQLAs banks should hold against CFMR in its final CP early next year – but wherever the quantitative CFMR requirement is set, the assumption that all NMDs leave the banks on the first day of a liquidity stress must be revised.

Pillar 2: What’s next?
Tagged on: