Our response to PRA consultation, CRD IV: liquidity, CP 27/14


We welcome the opportunity to comment on the PRA’s proposals in CP 27/14, and has had useful dialogue both with the PRA and with BSA members in preparation for making this response.  At a webinar in mid-February 2015, at which more than two thirds of building societies took part, the broad outlines of this response were overwhelmingly endorsed by online voting.  We concentrate in this response on those parts of the consultation that are relevant to building societies, paying particular attention to the concerns of smaller societies.  Abbreviations (LCR, ILAAP, ILSA etc) used in the consultation have the same meaning in this response

General observations

Building societies’ essential savings and mortgage business involves, and requires, extensive maturity transformation.  Building societies are therefore fully aware, and supportive, of the need for prudent liquidity management, and prepared to support all reasonable and proportionate measures to that end.  Societies can in general already comfortably meet the minimum 100% LCR requirement – indeed societies’ actual LCRs may easily be 200% or 300% even before the requirement is formally introduced. Societies currently hold high, and high quality, liquidity for a number of reasons – but not least due to their innate prudence, and awareness of liquidity and funding risks in a retail savings–based business.  

And societies are also fully conversant with the separate need (apart from a liquidity buffer to use in stressed conditions only) for operational liquidity; many societies keep substantial call deposit balances – in many cases an amount equal to the biggest outflow ever recorded - at their principal sterling clearing bank so that normal day to day cash flows, and intraday risks, are well covered.  Even then most building societies have the reserves account on instant access.  Intra day risks are therefore low for the sector; societies tend not to have the complex treasury operations that could attract material unplanned margin calls or other derivative quirks.  The PRA should take note.

Revocation of BIPRU 12 and simplified regime

The LCR regime is, as the consultation explains, established by EU legislation, and must of necessity displace much of PRA’s current regime – including, unfortunately, the simplified ILAS regime introduced by the FSA in 2009. This is in many respects a great pity, as the simplified regime was a good example of differentiated, appropriate regulation: although the calibration of the minimum requirement proved to have way overshot the minimum LCR (hence another reason why many smaller societies will initially report very high LCRs), the simplicity and relevance of the remainder of the regime compensated for this, and the result was that many smaller building societies, and some smaller, simpler retail banks, were able to meet high post-crisis liquidity resilience standards with economy of management effort. This was a good outcome.

The BSA has been clear that the LCR calculation itself would in due course supersede the simplified regime Pillar 1 formula (although by comparison, the LCR is too complex and over-engineered for societies’ purposes, but nothing further can be done about that). We are less clear that CRD 4 mandates uniformity in the wider approach (ILAAP, stress testing etc) – indeed CRD Article 86 is very clear about the need for proportionality.

One of the major concerns for ex-simplified societies is the burden of preparing a full ILAAP, compared with the limited and proportionate obligation under the current regime, to prepare an ILSA. In discussion, PRA was keen to reassure societies that - in practice – and given a proportionate approach to the ILAAP requirement in line with CRD Article 86, what will be required for a small society ILAAP may differ little from what is now required from the same society by way of ILSA. Societies are prepared to be reassured if more concrete language to this effect is added to the new PRA texts – otherwise they remain to be convinced.  So far the evidence is slim: if the ILAA is a good basis for comparison, then the amount of work to convert an ILSA into an ILAAP is considerable for societies.  They would have to:

  1. Produce bespoke scenarios. 
  2. Apply the (wider) 14 key risk drivers.
  3. Produce realistic, measurable and appropriate stress tests.  Removing the prescribed stress tests will likely add to the bespoke stress testing in order to cover the areas the prescribed stress tests already cover.
  4. Link the outcome of these to the risk appetite.
  5. Produce the associated management information to;
  1. compile a detailed ILAAP.
  2. produce and integrate the “monthly” stress testing.
  3. track the outcomes of the testing.
  4. ensure they comply with their “survival days” and liquidity levels going forward.

One of the major burdens for smaller ex-simplified societies is that in preparing an ILAAP they have to start from scratch, reinventing the wheel, based on a diverse regime that has to cater for large internationally active banks.  How can smaller societies that have not needed to do an ILAA for all the years the FSA, then PRA, liquidity regime has been in place suddenly need to do a full blown ILAAP?  Where is the justification for such overkill?

We therefore call on the PRA to carry over into the new regime other good features of the simplified regime, by giving ex-simplified societies, and smaller banks that also benefited from that regime, a clear steer as to the important risk drivers, and an intelligent selection of a few standard stress scenarios, that are relevant[1] to them. We see no reason why this cannot be accommodated within the requirements of CRD Article 86.

Glide path

Since societies already have high liquidity buffers, the BSA has no practical objection to the PRA’s proposed glide path, notwithstanding the element of modest goldplating (or rather, front-running) involved (as the glide path lies above the minimum in CRR Article 460).

Pillar 2 approach

The BSA appreciates that, given the short timescales now involved, it is not practical for the PRA to perform L-SREPs before setting initial Pillar 2 add-ons under the new regime, and the proposal to transition the absolute amount of existing ILG add-ons ( for full ILAS firms) into the new interim Pillar 2 add-ons is reasonable. That has the benefit of giving those firms greater certainty about their total buffer requirements. For simplified ILAS societies, however, there is no existing add-on, and therefore material uncertainty as to their Pillar 2 add-on under the new regime. Such societies may be reluctant to reduce their current liquidity buffer – even if their actual LCR is 200% or 300% - when to do so would be unambiguously beneficial to themselves and their customers (and, to quote from paragraph 2.9 of the CP, mutatis mutandis, would not  “risk temporarily constraining [societies’] ability to extend credit to the real economy”). To some extent this is unavoidable in the short term, but we think PRA needs to do two things:

  • First, it should accept explicitly that it will not use Pillar 2 to “recapture” all, or even most, of the quantum of liquidity buffer requirement released by a society’s move from the simplified formula to the LCR.
  • Second, it should give some clearer steer as to the order of magnitude of Pillar 2 add-ons likely to be imposed – since the existing ILG add-ons which will transition to Pillar 2 add-ons for full ILAS firms are confidential to each individual firm, currently simplified ILAS societies have no way of knowing whether Pillar 2 could conceivably take their total buffer to 150%, 200% or 500% LCR. This uncertainty puts them at a seriously unfair disadvantage going forward. Nor is it plausible for PRA to maintain that those societies will have to maintain high liquidity anyway to meet the OLAR – since the other elements of the OLAR can be easily met by non-HQLA resources such as call deposits at the principal clearing bank, or prepositioned assets.

Theoretically, there should not be much, if any, Pillar 2 add-on, for many societies however.  If the A/B split is a big driver of the ILAA and the LCR takes account of the (new) A/B split in the new five a to e designated buckets, and societies are well covered in their LCR calculations, what is left?

Daily LCR reporting capacity

The BSA strongly opposes this rule as a uniform measure applicable to all societies, and we criticise the PRA for not taking advantage of the explicit discretion in CRR Article 414 to modify this approach for small, simpler societies. CRR Article 414 requires daily LCR reporting where an institution does not meet, or expects not to meet, its Pillar 1 LCR. But Article 414 clearly permits competent authorities to reduce the reporting frequency and/ or permit a longer reporting delay based on the circumstances of the individual institution and (most importantly) taking into account the scale and complexity of the institution’s activities. The clear meaning here is that daily reporting may not be necessary for small, simple institutions, like many/ most building societies, under the circumstances envisaged.

The practical reason why this rule is inappropriate for small societies can be seen from the scenario below:


A small society experiences serious retail outflows traceable to the spreading of a malicious rumour in its locality. The society invokes its contingency funding plan (under whatever name) and implements the necessary actions.

Among the matters that the society’s senior managers must have urgent regard to are: supply of banknotes to cope with enhanced cash withdrawals; ensuring the society’s payments continue to be satisfactorily cleared; ascertaining the position with any wholesale counterparties with whom the society deals; arranging to monetise liquid assets by repo or outright sale; drawing on reserve accounts or against pre-positioned assets – as well as dealing with reputational issues, local media etc. Completing a full LCR template comes a long way down the list of actions that are a priori desirable in such circumstances.  Worse, for a senior manager to have to spend the time needed to finalise a daily LCR report risks the neglect of more urgent tasks. Neither will the society, nor we suspect, even the PRA, actually use the final output to any appreciable extent.  Instead, attention will rightly focus around simpler, more concrete and immediate measures such as actual outflow risks and number of likely survival days[2]. (In this context we also observe that the LCR calculation is prone – as the PRA is now well aware – to generate certain perverse outcomes. For instance, where a society has substantial holdings of call money, and transfers some of these to its reserves account, the effect on the LCR calculation is totally perverse: such a transfer substantially reduces the reported LCR. This instance has been drawn to the PRA’s attention. It underlines the folly of depending on the full LCR calculation in a crisis.)

Continuation of FSA 047/ 048

We understand the PRA’s need to access the FSA 047/ 048 information until LCR reporting is more settled, accurate and reliable.  But a two year transition period is too long.  We see this need as implicit confirmation of the BSA’s critique of much of COREP and its implementation, and the wasted costs and misdirected effort that it has entailed. Nevertheless, since societies are now in the unfortunate position of being required to report COREP/ LCR to challenging timetables – with PRA apparently unable to provide any relief by reducing that burden – we do not think it is reasonable for societies to be asked also to continue full reporting of FSA 047/ 048 on the same timetables as before.

The cumulative reporting burden, especially for smaller societies, is simply too great. We therefore call on the PRA to modify its proposal, either by reducing the volume of reporting requested, and/ or by lengthening the submission timetable.  Although FSA 047/ 048 is familiar to societies, it still takes appreciable effort to finalise under quality control and submit. The PRA may need to ask itself – which is more important, timely completion of LCR, or continued completion of FSA 047/ 048 - and why?

Operational requirements in Article 8

We agree with the general observations in the CP on buffer diversification– such as at paragraph 5.38 – though we point out that many of the assets societies typically hold (e.g. gilts, and even multilateral development banks such as the EIB) are explicitly excluded from the scope of any restriction the competent authorities may impose under Article 8.

Pre-positioned assets

We agree that non-HQLA pre-positioned assets cannot be included in the buffer for either Pillar 1 or Pillar 2 as they are not HQLA.  However, what is more important is that the access to liquidity represented by such pre-positioned assets helps societies meet the OLAR – and we therefore agree with the general advice given in paragraph 5.41 of the CP.

One of our members has pointed out an inconsistency.  The BoE encourages the pre-positioning of loan collateral and own-name securitisations, and active use of the Sterling Monetary Framework in liquidity planning.  Yet pre-positioned collateral (post haircut) may not be included as HQLA for Pillar 2 purposes.  The same society also argues that the value of pre-positioned loan collateral (post haircut) can be readily calculated and therefore included in the LCR calculation in Pillar I, even if a further factor were to be applied,

Reporting debacle

We note that the COREP/ LCR reporting debacle requires some common sense interim solutions, including alternative methods such as spreadsheets. The BSA is not in a position to comment in detail on these practical issues – we leave that to experienced practitioners. We reiterate instead our observation that the relevant authorities have consistently failed to foresee, take seriously, forestall, or make genuinely proportionate, the burden of reporting and especially the burden of changes in reporting. COREP as a whole, and the LCR elements, serve as object lessons in how not to approach these matters.

Further engagement

We are, quite naturally, happy to engage further with the PRA, after the close of the consultation period, to explore any of these issues further.

About us

The Building Societies Association represents all 44 UK building societies. Building societies have total assets of over £330 billion and, together with their subsidiaries, hold residential mortgages of over £240 billion, 19% of the total outstanding in the UK. They hold over £240 billion of retail deposits, accounting for 19% of all such deposits in the UK. Building societies account for about 28% of all cash ISA balances. They employ approximately 39,000 full and part-time staff and operate through approximately 1,550 branches.

Jeremy Palmer

27 February 2015

[1] At a meeting with society chief executives on 17 February 2015, PRA mentioned stressed situations such as reduction in credit ratings and closure of a major employer in a society’s home town.  Smaller societies are not rated and few are the major employer in their home town.  The key risk is an outflow generated by specific reputational damage.

[2] In a real time, stressed environment, societies might be better off calculating survival days as follows:

  1. £240 million liquidity at start of stress
  2. Day 1, £10 million out.  £230 million remains.
  3. At that outflow, there are 23 survival days.

Would need to add in the committed inflows/ outflows.