In short. Building societies produce three narrative artefacts each year: the Annual Review for members, the Pillar 3 disclosure for the PRA, and the strategic report for the auditor. The narrative reconciliation between the three is rarely performed and is the recurring finding in PRA thematic reviews. The fix is a single workpaper, drafted by audit, that ties every quantitative and methodological assertion across the three documents to a single source.
1. The mutual reporting context
Building societies are constituted under the Building Societies Act 1986 and the related orders. They are prudentially regulated by the PRA under the Building Society Sourcebook (Supervisory Statement SS6/16), conduct-regulated by the FCA, and governed by a member-elected board. Their financial reporting framework is FRS 102 with the Building Society SORP overlay, which provides interpretation for mutual-specific items including subordinated capital, permanent interest-bearing shares, retained earnings as the principal capital, and reserves attributable to members.
This framework sits at the intersection of three quite different audiences. Members of the society have the right to read the Annual Review and to expect a narrative they can understand. The PRA reads the Pillar 3 disclosure and expects a narrative consistent with the supervisory return data. The auditor reads the strategic report and expects a narrative consistent with the audited financial statements. Each audience has a slightly different idea of what the society is doing and how it is performing. The narratives drift accordingly.
2. The three documents
The Annual Review is the member-facing communication, typically a glossy document with a chairman’s statement, a chief executive’s review, performance highlights, and selected financial information. The audience is non-specialist. The tone is reassurance and accessibility.
The Pillar 3 disclosure is the regulatory capital and risk disclosure required under the Capital Requirements Regulation. The audience is the PRA and informed market participants. The tone is technical and template-driven, with prescribed tables and risk-management narrative.
The strategic report is part of the statutory accounts under the Companies Act 2006 (as applied to societies through the Building Societies Act). The audience is members again, but in their capacity as voters on the accounts at the AGM, and the FRC by extension. The tone is balanced and disclosure-driven, with explicit requirements under the Strategic Report Regulations.
The three documents describe the same society in the same year. They are written by different people, at slightly different times, in different review chains. The drift is structural rather than intentional.
3. Where the narratives diverge in practice
Six recurring divergence points, in observed order of significance.
Net interest margin commentary. The Annual Review tells members that margins were maintained through prudent pricing discipline. The Pillar 3 tells the PRA that margins compressed due to competitive deposit pricing. The strategic report tells the AGM something between the two. The three are not inconsistent if you read them carefully; they are inconsistent on a quick read by a journalist or a thematic reviewer.
Mortgage book quality. The Annual Review highlights the low-arrears performance of the book. The Pillar 3 discloses the IRB-based or standardised risk weights and the expected credit loss provisioning. The strategic report describes credit risk management. Three different lenses on the same portfolio, with three different summary characterisations.
Liquidity narrative. The Annual Review describes liquidity as comfortably above regulatory minimum. The Pillar 3 discloses the LCR ratio in percentage terms. The strategic report references liquidity risk management. The three pieces of language are coherent only if the reader does the integration; the documents themselves do not perform it.
Capital adequacy. The Annual Review describes the society as well capitalised. The Pillar 3 publishes specific CET1, Tier 1, and total capital ratios. The strategic report discusses capital management policy. In several inspected cases the “well capitalised” language has remained in member communications through periods when the regulatory buffer was being actively managed, which the Pillar 3 documents transparently.
Conduct and culture commentary. The Annual Review describes the society’s culture in member-friendly terms. The strategic report addresses SM&CR accountability and conduct risk. The Pillar 3, where conduct risk is treated, may discuss it under operational risk. Three different framings of the same set of behaviours.
Strategy and outlook. The Annual Review presents a forward-looking statement appropriate for member confidence. The strategic report presents the going-concern assessment and viability statement (where the society opts into the UK CGC adapted regime). The Pillar 3 may include forward-looking elements through the ICAAP narrative. The three forward-looking statements need to be consistent with each other; in practice they are drafted independently and rarely cross-checked.
4. Why this is a finding waiting to happen
FRC thematic reviews of audit quality in the mutual sector have repeatedly flagged the consistency of disclosures across statutory artefacts as a weak area. PRA supervisory feedback has at points raised concerns about whether the member-facing narrative adequately reflects the supervisory position. The Building Societies Association has published guidance on disclosure consistency for the sector.
The recurring problem is that no single function in the society owns the cross-document consistency check. The finance team owns the strategic report and the financial statements. The communications or marketing function owns the Annual Review, sometimes with finance input but rarely with finance control. The risk function owns the Pillar 3, with finance input on the quantitative side. None of these three functions owns the consistency check between the three documents.
The audit team typically reviews the strategic report under ISA (UK) 720 for consistency with the financial statements and the auditor’s knowledge. It does not, in the standard scope, review the Annual Review or the Pillar 3 against the strategic report. The cross-document check is not anyone’s job, which is why it is rarely done.
5. The single workpaper that fixes it
The fix is a narrative consistency workpaper, drafted by audit, that takes every material quantitative and methodological assertion in the Annual Review and the Pillar 3 and ties each back to a source in the audited financial statements or the supporting workpapers.
The structure is a three-column table. Column one: the assertion as it appears in the source document (Annual Review or Pillar 3). Column two: the corresponding figure or assertion in the audited financial statements or working-paper file. Column three: the auditor’s conclusion on consistency, with reference to the underlying evidence.
The first time this workpaper is drafted, it takes two senior audit days. Inconsistencies surface. Some are resolved by the society amending the relevant document before publication; others are resolved by accepting a documented difference (typically a presentation difference rather than a substantive one). After the first cycle, the workpaper updates with marginal effort in subsequent years.
The workpaper is technically outside the standard scope of an ISA 720 review, which is restricted to other information in the document containing the audited financial statements. The workpaper extends the principle voluntarily to documents in the same communication suite. Audit committees with experience of mutual reporting view it as a value-add. In observed cases, supervisory feedback on the society’s next thematic engagement is materially better after the workpaper has been in place for a cycle.
6. The Building Society SORP specifics
Beyond the narrative consistency point, FRS 102 reporting under the Building Society SORP introduces specific technical areas that recur in audits.
Permanent Interest Bearing Shares. PIBS are deferred shares that rank as Tier 2 capital. Their classification under FRS 102 paragraph 22 depends on the specific terms of the instrument and the issuing society’s rules. The recurring audit issue is that the technical accounting classification has been carried forward from prior years without re-evaluation against the current substance of the rules. Material societies have, on occasion, found that an instrument classified as equity has features that suggest a liability classification under current interpretation.
Effective interest rate accounting. Mortgage books accounted under FRS 102 paragraph 11 require EIR application. The systems implementation in many smaller societies has known limitations — for example, the treatment of fees and incentives on a specific product family may diverge from theoretical correctness. The audit position requires either remediation of the system or an explicit override workpaper with quantified impact.
Pension scheme accounting. Many societies operate defined-benefit pension schemes accounted under FRS 102 Section 28. The recurring issue is whether the multi-employer scheme exemption applies, what the basis of measurement is, and how movements in the scheme deficit interact with the income statement and equity.
Member redress and complaints provisions. The provisions for member redress under FRS 102 Section 21 require careful evidence of the unwind history, the basis for current estimates, and the consistency of methodology over time. Mutual sensitivity to redress narrative makes this an area where audit-management tension is highest.
7. The board interaction
The board of a building society is unlike a listed-company board in one important respect: most of the members are member-elected non-executives who do not have prior banking experience. The audit committee is the usual exception, where the chair is normally a chartered accountant or comparable professional. But the wider board reads the audit findings with the framework of a member rather than the framework of a banker.
This has implications for how audit findings are communicated. A statement that “disclosure consistency has improved this year following implementation of the narrative reconciliation workpaper” is opaque to a member-elected director. A statement that “the society’s public communications now match what is in its regulatory filings, in a way they did not last year” is clearer.
The translation from audit language to mutual-board language is part of the audit job for a society. It is not a soft skill; it is a substantive part of how the audit committee discharges its responsibility to the wider board. Auditors who do not translate effectively leave findings under-actioned, which becomes a remediation issue in subsequent years.
8. The supervisory cycle and how it interacts with the audit
The PRA’s supervisory cycle for building societies includes periodic thematic reviews, the SREP (Supervisory Review and Evaluation Process), and ad hoc engagement on specific matters. The audit cycle is annual; the supervisory cycle is rolling. The two intersect in three places.
First, the ICAAP submission, typically due in the spring, draws on the same forecast data that informs going concern in the next audit. Inconsistency between the ICAAP base case and the going-concern base case is an audit issue.
Second, supervisory thematic findings often surface in the audit committee within months of the audit. The audit firm should be aware of recent thematic findings affecting the sector and should incorporate them into the planning risk assessment. A finding by the PRA in the autumn that affects half the sector should not arrive at the audit firm in the spring as a surprise.
Third, the conclusion of the SREP — particularly any change in Pillar 2 capital requirement — affects the going-concern and viability assessment in the next audit. Audit teams should request the most recent SREP letter at planning and integrate it into the assessment.
9. Closing observation
Building societies are a quieter audit territory than listed banks. The technical issues are no less demanding; the regulatory expectations are no less rigorous. What differs is the absence of a price-sensitive market reaction to disclosure inconsistency. That absence is a feature, not a bug, but it also removes a discipline that listed entities have without trying. Mutuals must impose the equivalent discipline themselves, which is what the narrative consistency workpaper does.
A building society audit done well leaves the society with a strategic report, an Annual Review, and a Pillar 3 disclosure that say the same things in different registers. A building society audit done badly leaves three documents that say slightly different things and a thematic review waiting to happen. The work to move from the second state to the first is unspectacular and recoverable. The senior thinking required is exactly the kind that mutual-sector auditors who have done the work before bring to a file.
— DK Buledi, August 2025