Our response to PRA consultation on regulatory reporting of financial statements, forecast capital data and IFRS 9 requirements

We urge the PRA to use Brexit as an opportunity to review its regulatory reporting requirements, particularly for non-systemic institutions.   FINREP through the back door is not the answer.

Executive summary

We wish to record our thanks to the PRA team who took the time and trouble to discuss their plans with a group of societies on 10 June 2016.  At the time, the referendum on UK membership of the EU had not yet happened so “Brexit” was not considered.  But Brexit is now a reality – and also an opportunity. 

Brexit means that EU regulations that currently determine the landscape – such as the regulatory reporting aspects of CRR and its delegated Acts – may cease to be binding at the point of exit, and this could – and indeed, should - create scope for greater proportionality in the UK’s requirements.   We therefore urge the PRA to use the UK’s exit from the EU to pause its reporting plans and reflect on what data is appropriate and demonstrably necessary.  We doubt that every single piece of data from every building society, some of which is duplicated, is actually used by the PRA.  A rigorous cost-benefit analysis is surely needed.

We do understand why the PRA wishes to update and standardise its returns.  Our sector has been questioning for some time why, for example, it should still be completing QFS1 and FSA 014, not to mention the large number of ad hoc reporting requests.  To incorporate the information from these old returns into a new suite sounds sensible and desirable.  But we consider FINREP (itself an EU artefact) to be fundamentally flawed – and imposing the templates beyond the reach of Article 99 of the CRR seems like FINREP through the back door.

So we question the proposed further use of FINREP templates when the PRA has previously told us the information they contain does not help its supervision.  This is particularly galling for IFRS building societies (and most probably banks) that currently complete the templates as they do not use the data internally at all.   They are compiled solely for regulatory compliance. 


We have long advocated a more proportionate approach to prudential regulation than is currently possible under the EU’s Single Rule Book.  Regulatory reporting is one such area ripe for reform.  Our argument for proportionality is now explicitly supported by the Bank of England/ PRA.  This can be seen in their response to the EC’s call for evidence on the EU Regulatory Framework for Financial Services in January 2016:

But a “one size fits all” approach of common binding rules for all banks, no matter what their size, complexity or level of cross-border activity, can cause distortions given that the costs of regulation tend to bear more heavily on smaller banks……More proportionate, differentiated rules are more likely to enable banks of different size and business model to compete on an equal footing across the EU than the same rules applied to all banks.

The costs of regulation must be proportionate to the benefits.  The benefits and costs vary across banks of different size and business model.  Often the benefits of regulation are proportionately bigger for larger or more complex banks, while the extent that regulation imposes fixed costs those will tend to bear more heavily on smaller banks.

Consultation questions

Question 1: Does your firm envisage any difficulty in providing the balance sheet and/or P&L data via the specific FINREP templates proposed?

Only the very large building societies are required to complete FINREP templates under Article 99 of the Capital Requirements Regulation.  This is because they use IFRS and report on a consolidated group basis.   These larger societies will be familiar with the templates and the significant level of detail required.  Others will not be.  They have been using the PRA’s templates (“FSA 0XX”), which are less detailed and tailored to domestic supervisory need.

Crucially, the FSA 0XX returns were introduced as part of the then FSA’s implementation of integrated regulatory reporting (“IRR”), itself a consequence of the Capital Requirements Directive.   With IRR, the FSA consulted industry and ran pilot programmes.  The FSA also provided advice.  This did not happen with the EU-mandated COREP, with many societies (and banks) being left stranded by the inability of the PRA to help with the templates.  Instead, they were referred to the EBA, which took a long time to answer – and in a few cases, referred the society back to the PRA.

FINREP building society reporters had a similar experience.  Furthermore, they say that the data is compiled separately and solely for the purpose of regulatory compliance.  While societies may, in some cases, draw on existing data used for annual financial statements, or other regulatory reporting, there are many areas of significant divergence.  These are due to differences in aggregations, definitions, and areas where data is required at a more granular level.  For example, FINREP forbearance disclosures are different to the statutory forbearance disclosures.   This extra work consumes valuable resources. 

FINREP building societies use financial information internally which more appropriately reflects the way in which they manage the business. The differing nature of information across the reports from different regulatory requirements further leads to the need to reconcile across the different data uses by the business.

Within PRA itself doubts on the value of FINREP templates must linger.   The templates have not been put forward for use by credit unions in the July consultation on credit union reporting.  While credit unions are out of scope of CRD 4, we assume if FINREP templates were considered valuable they might have been applied to the sector (assuming, of course, the European authorities were to agree to their use beyond the CRR).

But most societies believe they will be able to provide most of the data set out in the consultation, although as previously noted, their resources are already stretched.  What could be a challenge is the method of submission.  At our meeting on 10 June 2016, PRA officials said they were not sure about the method of submission.  This is an unfortunate gap in knowledge.  Current returns FSA 001 and FSA 002 allow data to be entered directly into GABRIEL online at no cost to the society.  Since these new returns are FINREP based, we would be concerned if they needed to be created using a system tool and submitted as XBRL.  Societies consider XML to be a more appropriate and proportionate solution.

A further challenge for the PRA is to establish whether the EBA will grant it permission to apply its FINREP templates beyond the statutory population.  This is even more pertinent (and urgent) since Brexit.

Question 2: Does your firm foresee any issues with providing balance sheet and P&L data at the above frequencies?

The smaller the society the harder it will find providing the data at the required frequencies.  This is because a smaller society typically has fewer staff to prepare, review and sign off the returns.   Such concerns are reduced in larger societies but present nonetheless. 

We welcome the proposal that firms will be able to request to report on a schedule linked to their ARD.  That is most helpful.  But we believe the PRA should reduce further the frequency of reporting for smaller firms such as building societies, particularly the PRA 104-108 returns.  Similarly, we recommend the submission date for the PRA 101 return should be extended to 30 working days after the end of the reporting period.

Given the high number of competing regulatory initiatives – not just PRA, but also FCA and HMRC - the more notice building societies (and banks) have the better.  They need time to absorb the new changes then plan, test, implement and review them.  Often the process lasts up to two years particularly when outside resources are needed.   We urge the PRA to start the clock running only when it has adequate guidance in place.  To help all firms, taxonomies should be made permanent and not changed frequently as under the EBA.
We understand from our June meeting that the PRA does not intend to require parallel reporting.  We welcome this pragmatic move but would appreciate formal confirmation.  Introduction of the new returns should mean the previous returns cease to be required.

Question 3: Do the instructions for PRA 104 to PRA 107 give your firm sufficient guidance on how to complete the returns?

These instructions effectively link to the EBA website and relevant data point model.   This is an inadequate response, particularly in light of Brexit.  We strongly advise – as we did during discussions with BoE staff in October 2012 -  the regulator to produce guidance of its own backed up by named contact points, a helpline and countrywide training events.  

Experience with COREP shows how limited and unhelpful EBA guidance can be.  The Q&A format can be difficult to navigate.  Societies with questions on COREP are routinely referred to the EBA – which, in a few cases, refers the society back to the PRA.   The whole process can take months leaving societies in the dark when returns are due. 

EBA guidance is intended for all jurisdictions and is therefore not always relevant in a UK situation, let alone the mutual sector.    There is no one place to seek an answer either, it could lie, for example, in guidelines, models or in the Q&A section.  We understand the PRA is mounting a review of COREP returns in Q3 2016 to help it understand reported inaccuracies and consistencies.  One society explains: “Legislation is difficult to interpret with little produced in terms of management guidance.  Whilst we only have to complete a small proportion of the cells compared to a large active international banks we are still required to interpret the same legislation and confirm the data cells are not relevant to us.”

A further example is the implementation of the new ALMM returns.  There is lack of guidance over the detail to be included in the returns, which returns will actually be introduced and when they will be introduced.  Societies need time to ensure their MI systems are capable of producing the information required; there is a strong perception that the PRA expects societies to be able to complete the returns at the press of a button.  The rollover of funding return in particular is one better collated daily than retrospectively and so requires some notice

Therefore, to prevent a re-run of the COREP and ALMM situations, we oppose any move to provide only EBA guidance.

Question 4: What is your firm’s view on forecast data and the level of granularity at which the PRA proposes to collect balance sheet and P&L information (PRA104 to PRA 107)?

The forecast requirements are slightly more granular than even large societies produce for their own internal forecasts.   For example, few forecast market fair value movements or by instrument grouping ie they forecast a net return on their liquidity portfolio not by instrument type.  Similarly, few, if any societies, forecast actuarial gains or losses.
We understand that forecasting information may be provided on a best efforts basis.  Written confirmation of this helpful position would be most helpful.

Question 5: Does your firm forecast elements of the statement of comprehensive income (SOCI)? Which elements of the SOCI could your firm forecast on a regular basis to provide a more comprehensive picture of your firm’s future profitability?

We have concerns about the SOCI return.  Even the largest societies would only be completing 6-8 lines (including totals).  Echoing our response to the question above, items such as the actuarial gain/ loss and deferred tax would only be obtained once a year and that is at year end.  The only figure they would report quarterly is movement on the available for sale reserve.   As one society points out, this proposed forecasting – even on a best efforts basis - is becoming dangerously close to guesswork.  This leads us to question how much use the data will be to the PRA.

Question 6: What is your firm’s view on the timing of the implementation of the new balance sheet and P&L related data items, in particular coordination with the forthcoming introduction of IFRS 9 (2018, subject to EU endorsement) and the implementation of Structural Reform (2019)?
The proposed implementation date for the capital+ proposal is 1 July 2017; for the other proposals the date is 1 January 2018.  

Since societies are already completing capital+ returns, the July 2017 deadline seems achievable.   But even if there were adequate and responsive guidance in place, January 2018 for the other returns could be a challenge for many, particularly the smaller and medium-sized societies where the burden of changing regulation falls on the same few people.  In any case, January is a busy time for many societies with resources stretched to manage financial year end work.  Perhaps a more convenient start point would be a mid-year but only in a year where the outcome of Brexit in relation to financial reporting is known and already tested in larger institutions. 

Question 7: Does your firm have any suggestions how the PRA’s proposals for balance sheet and P&L reporting could be made more proportionate?

We have indicated elsewhere in the response where a more proportionate approach could be made.  For the PRA only, we attach the appendix of a March 2016 survey on the burden of regulation which has a section on regulatory reporting including suggestions to remove duplication and address proportionality issues.

Question 8: Does your firm foresee any issues with providing Capital+ at the above frequencies and/or by the specified due dates?

Although societies have been completing capital+ returns for some time, making them regulatory returns means they are now formalised and subject to more intense review and scrutiny.  They represent another return to fit into the already busy year end schedule of regulatory reporting and accounts

Currently societies are required to prepare current plus one year forecasts ie two years.  But the new returns specify current plus two years’ forecasts ie three years.  We question the overall benefit of such an extension – it will be more onerous for societies, but will add little to the PRA’s understanding, which will presumably have seen multi-year forecasts as part of the business plan/ ICAAP.  The current uncertainty surrounding Brexit shows that such long range forecasts will end up as guesswork.

Question 9: Does your firm foresee any issues if the reporting date for Capital+ continues to be aligned to end of accounting period for the applicable period?

A reporting date aligned to the end of accounting period works well for societies and helps to reduce reconciliation work. 

Question 10: Other than FSA015, which other existing PRA Rulebook reporting requirements may require revision to ensure compatibility with firms’ implementation projects for IFRS 9?
Question 12: Does your firm have views on how a limited extension of the scope of certain data items from certain IFRS 9 FINREP templates on impairment could be undertaken proportionately?
Question 13: Does your firm have views on any particular challenges that would arise from an extension of FINREP data on impairment to firms using UK GAAP?

The following response addresses the above three questions:

We agree that IFRS 9 is a more appropriate standard than IAS 39.  The changes on credit loss provisioning should contribute in addressing concerns about the issue of “too little, too late” recognition of credit losses and improve the accounting recognition of loan loss provisions by incorporating a broader range of credit information.  But the standard’s implementation in the UK is not yet certain: when this consultation was first published, IFRS 9 had not been adopted in the EU.  Brexit muddies the water further.

In view of this, we urge the PRA to put on hold its plans to implement IFRS 9 FINREP impairment templates to IFRS reporters such as the larger building societies.  The situation is so uncertain that any work building societies (and banks) do could be subject to change, which is always resource-consuming.  The PRA should make an announcement as soon as possible.

The situation is more acute with UK GAAP societies.  The FRC is conducting a fundamental review of FRS 102 with minor changes envisaged for adoption on or after 1 January 2019 adoption and more substantive changes, such as the introduction of a modified IFRS 9, envisaged for adoption on or after 1 January 2022.  This staggered approach is designed to give UK GAAP firms adequate time to prepare and to learn from larger firms.  This move is an acknowledgement by the FRC that IFRS 9 expertise is scarce and expensive, meaning smaller firms are disadvantaged if they have to compete for it with larger firms. 

Since the earliest adoption date of a modified IFRS 9 will be 2022 for UK GAAP societies, we therefore consider any attempt by PRA to force through a form of IFRS 9 for these societies to be premature and harmful.  

Question 11: Does your firm have views on how the limitations of FSA015 for firms using IFRS 9 identified above should be addressed? Specifically, the arguments for and against adding granularity to FSA015 and/or deleting FSA015 and extending FINREP data on impairment to firms not currently submitting FINREP.

As stated above, we urge the PRA to put on hold its plans to implement IFRS 9 FINREP templates to IFRS reporters such as the larger building societies given the uncertainty surrounding Brexit.  On a separate note, MLAR table F does provide data on arrears and credit quality that is similar to FSA 015.  We would be interested to know if the PRA plans to retain this table once IFRS 9 is eventually adopted.