Most Common IFRS Errors Found in Big 4 Audit Files
Author
Sai Manikanta Pedamallu
Published
Reading Time
16 min read
Table of Contents
Most Common IFRS Errors Found in Big 4 Audit Files
By Sai Manikanta Pedamallu (ACCA, CMA US, CSCA US, CGMA, ACMA, Dip IFRS, M.Com, MBA, MA)
Lead Instructor, Global Fin X | www.globalfinx.in/manikanta
IFRS errors are not randomly distributed. The same mistakes appear across different standards, different clients, and different years, because they arise from the same structural causes rather than from ignorance of the requirements.
Someone who knows IAS 36 perfectly can still get the impairment wrong, in exactly the way someone who knows IAS 12 perfectly still gets deferred tax wrong, and for the same underlying reason.
This post is organised by error pattern rather than by standard. Posts 41, 68 and 82 covered impairment, deferred tax and related party findings in their own right. What follows is what those findings have in common, and where else the same pattern shows up.
Why the Same Errors Recur
Four structural causes explain most of what auditors find.
The work is done at the end. Deferred tax, impairment, provisions, and disclosure all depend on numbers that are only final in the closing days of the reporting cycle. Work performed under time pressure, reviewed briefly, is where errors survive.
Prior year is the starting point. Almost every schedule in a finance function is built by rolling forward last year's version. That captures last year's transactions perfectly and this year's new ones not at all.
Systems categorise by form, not substance. A ledger knows a balance is in foreign currency. It does not know whether the balance is monetary. A payables system knows a counterparty name. It does not know whether that counterparty is related.
Nobody outside finance reviews the answer. A wrong revenue number gets challenged by someone in the business. A wrong deferred tax balance, an incorrect CGU boundary, or a missing related party looks identical to a right one to almost every reader.
Pattern One: The Roll-Forward Trap
The most common completeness failure in IFRS application is not a misunderstanding of a requirement. It is a schedule that was correct when it was built and has not been rebuilt since.
Deferred tax, as Post 68 set out, is the paradigm case. A temporary difference schedule constructed three years ago captures the differences that existed three years ago. A first lease, a first acquisition, a first revaluation, a first ESOP grant, a new jurisdiction, or a decommissioning provision will not appear unless someone adds it.
The same pattern operates elsewhere.
Leases. A lease register built at Ind AS 116 transition, rolled forward each year, misses new contracts that contain embedded leases and misses modifications, extensions, and terminations that require remeasurement.
Related parties. A related party register updated for board changes but not for entities newly controlled by key management personnel or their close family members, and never adjusted for relationships that existed during the period but ended before the reporting date, as Post 80 explained.
Segments. Segment identification performed at transition and never revisited, while the internal reporting structure has evolved, produces reported segments that no longer correspond to what the chief operating decision maker actually reviews.
Uncertain tax positions. A schedule of disputed matters carried forward without reassessment, when IFRIC 23 requires reassessment at each reporting date and specifically states that the absence of action by the authority is not in itself a change in facts and circumstances.
The control: rebuild from the current position rather than roll forward from the prior schedule. Every asset and liability on the trial balance either has a tax base difference or does not. Every contract either contains a lease or does not. The only reliable way to know is to look at this year's population, not last year's answer.
Pattern Two: Populations That Miss What Never Hit the Ledger
Identification processes built on ledger data cannot find items that never generated a ledger entry.
Related party transactions with no consideration. Interest-free loans, guarantees given without charge, rent-free use of premises, and services provided without invoice are related party transactions requiring disclosure. None of them appears in a payables or receivables extract.
Transactions with unrelated parties that benefit related parties. NFRA has specifically identified this pattern, described in Post 82. Counterparty screening against a related party master list cannot detect it, because every leg involves a genuinely unrelated counterparty.
Embedded leases. A contract for logistics, data centre capacity, power supply, or equipment-based services may contain a lease. It sits in the procurement system as a service contract and nothing flags it.
Non-monetary foreign currency balances retranslated. As Post 87 explained, ledger systems flag all foreign currency balances for retranslation without distinguishing a monetary payable from a non-monetary prepayment. Both look identical to the system; only one should move.
Commitments. Related party disclosure extends to outstanding commitments, and lease disclosure extends to leases not yet commenced. Neither appears on the balance sheet, so neither appears in a balance-sheet-driven identification process.
The control: identification must run from contracts and arrangements, not from ledger balances. At least once, for at least the material population, someone has to read the agreements.
Pattern Three: Form Driving the Accounting
IFRS is a substance-based framework, and the most consequential errors occur where the legal form of an arrangement has been allowed to determine its treatment.
Preference shares classified as equity because they are called shares, when mandatory redemption or mandatory dividends make them financial liabilities under IAS 32, as Post 23 explained.
Corporate wrapper transactions treated as business combinations because they were structured as share purchases, when the legal structure was chosen for stamp duty efficiency and the substance is an asset acquisition, as Post 55 covered.
Related party arrangements disclosed leg by leg without describing the arrangement they collectively constitute. The NFRA case in Post 82 is the clearest example: each component was accurately disclosed and the arrangement was invisible.
Service contracts that contain leases, and lease contracts that are in substance service arrangements, classified by their title rather than by the identified asset and control tests.
Concession arrangements where the operator recognises the infrastructure as its own property, plant and equipment because it built it and physically holds it, when IFRIC 12 says the grantor controls it, as Post 84 explained.
The control: for every material arrangement, ask what the entity has actually obtained and what it has actually given up, before asking what the document is called.
Pattern Four: The Wrong Unit of Account
Many IFRS requirements produce a defensible answer at one level of aggregation and a wrong answer at another. Getting the unit of account wrong is often more consequential than getting the measurement wrong.
CGUs defined too broadly, so that a deteriorating business is absorbed into a healthy one and no impairment arises. Post 40 covered this, and it remains the most challenged judgment in IAS 36.
Segments aggregated where economic characteristics genuinely diverge, concealing an underperforming business, as Post 81 set out.
NRV tested on aggregate inventory rather than item by item, allowing surpluses on some items to absorb deficits on others. Post 69 covered this.
Contracts not unbundled into their lease and non-lease components, or performance obligations not separated, so that consideration is allocated to the wrong things.
Uncertain tax positions assessed individually where they are interdependent, or collectively where they are not, contrary to the IFRIC 23 requirement to choose the basis that better predicts resolution.
Business combination goodwill allocated to a level larger than an operating segment, which IAS 36 prohibits precisely because it masks impairment.
The control: the unit of account is a decision, not a default. It should be documented with a reason, and it should not change between periods without an operational justification.
Pattern Five: The Wrong Date
A surprising number of IFRS requirements turn on identifying a specific date, and using the wrong one produces an error that is invisible on the face of the accounts.
| Requirement | The correct date | The date frequently used instead |
|---|---|---|
| Share-based payment fair value | Grant date, being the date approval is obtained | Date terms were agreed or announced |
| Business combination consideration | Acquisition date | Agreement date |
| Debt-for-equity swap consideration | Date the liability is extinguished | Date the restructuring was agreed |
| Foreign currency advance consideration | Date the advance was paid or received | Date of delivery or revenue recognition |
| Non-monetary asset at fair value | Date fair value was determined | Closing rate at reporting date |
| Levy recognition | Date the obligating event occurs | Payment date, or spread across the period |
| Agricultural produce cost | Point of harvest | Date of sale, or accumulated growing cost |
| Held for sale classification | Date all criteria are met | Date of board decision |
Each of these was covered in the relevant post in this series, and each recurs.
Pattern Six: The Wrong Rate
Where a calculation requires a rate, IFRS is frequently specific about which rate, and the specification is frequently ignored.
Pre-tax versus post-tax. IAS 36 requires a pre-tax discount rate applied to pre-tax cash flows. Using a post-tax WACC without grossing up understates the discount and overstates value in use, as Post 41 set out.
Locked-in versus current. The IFRS 17 contractual service margin accretes at the rate determined at initial recognition, while the fulfilment cash flows are discounted at current rates. Post 75 covered this.
Original versus revised. Under IFRS 16, a lease modification uses a revised discount rate at the modification date, while an index-linked remeasurement uses the original incremental borrowing rate. Post 28 covered the distinction.
Historical versus closing versus average. IAS 21 requires monetary items at closing rate, non-monetary items at historical cost at the transaction date rate, income and expenses of a foreign operation at transaction date or average rates, and all amounts of a hyperinflationary operation at the closing rate.
Market versus coupon. A compound instrument's liability component is discounted at the market rate for equivalent non-convertible debt, not at the instrument's coupon, as Post 24 explained.
Statutory versus expected. Deferred tax is measured at the rate expected to apply on reversal, which for an Indian company that has elected a concessional regime is not the default statutory rate.
Pattern Seven: Follow-the-Item Allocation Failures
Several standards require an amount to be recognised in the same place as the item it relates to, and this is one of the most consistently mishandled presentational requirements in IFRS.
Deferred tax follows the underlying item: to OCI where the item was in OCI, to equity where the item was in equity.
Exchange differences on non-monetary items follow the gain or loss on the item under IAS 21.
IAS 19 remeasurements go to OCI and are never recycled, while service cost and net interest go to profit or loss.
Revaluation decreases offset any existing surplus for that specific asset in OCI before affecting profit or loss, tracked asset by asset rather than pooled.
The symptom is usually visible: an OCI statement showing amounts gross rather than net of tax, and an effective tax rate reconciliation that will not balance without an unexplained item.
Pattern Eight: Estimates That Are Never Tested Against Outcomes
The most consequential judgment errors involve forecasts that are prepared, used, and never compared to what actually happened.
Value in use cash flows. As Post 41 set out, the single most effective audit procedure is comparing last year's forecast to this year's actual. Where forecasts have consistently overstated, the current forecast requires substantially more support than a board approval.
Expected credit loss models. Forward-looking scenarios and probability weightings that are never backtested against realised losses.
Deferred tax asset recognition. Recovery forecasts supporting recognition of losses, where prior recovery forecasts did not materialise.
Attrition and salary escalation in defined benefit valuations. Assumptions carried forward from the prior actuarial report without checking them against the entity's own recent experience, as Post 57 explained. A zero attrition assumption is not conservative; it systematically overstates the obligation.
Warranty and provision estimates. Expected values that are never compared to actual claim experience.
The control: every recurring estimate should carry a backcast. Show last year's estimate alongside this year's actual before presenting this year's estimate.
Pattern Nine: Spreading What Should Be Recognised at a Point, and Vice Versa
A recurring family of errors involves the timing profile of recognition.
Levies spread evenly where the legislation identifies a single obligating date. A property tax triggered on a specific date is recognised in full on that date, not accrued monthly, as Post 90 explained.
Interim tax computed discretely where IAS 34 requires an estimated annual effective rate, as Post 67 covered.
Restructuring costs provided for on a management decision, before the constructive obligation conditions in IAS 37 are met.
Onerous contract losses deferred rather than recognised in full when identified.
IFRS 17 onerous group losses deferred rather than recognised immediately, in a framework that defers profit but never defers loss.
Cancellation of share-based payment awards treated as ceasing future expense, rather than accelerating all remaining unrecognised expense into the current period.
Pattern Ten: Disclosure That Complies Without Informing
The final pattern is not a measurement error at all, and it appears in regulatory findings as consistently as any of the others.
NFRA has observed that disclosures in notes are sometimes neither relevant nor useful to users and have the potential to obscure material information.
The characteristic features:
Sensitivity analysis with implausible ranges, showing that an extreme movement would be required before impairment arises.
Key assumptions described qualitatively without disclosing the actual rates used.
Seasonality notes carried forward unchanged each interim period.
Arm's length assertions made as a drafting convention without evidence capable of substantiating them, which IAS 24 does not permit.
Aggregated related party disclosure that prevents identification of any individually significant transaction.
Segment measures explained without describing the differences from IFRS measurement or the asymmetrical allocations.
The test is whether a reader could reach an independent view from what is disclosed. Where the disclosure lets the reader only confirm that the entity considered the matter, it satisfies the checklist and not the standard.
The Documentation Dimension
An error in an audit file is not always an error in the financial statements. A file can be deficient where the accounting is correct.
NFRA's inspection findings have repeatedly identified documentation that is incomplete, insufficiently linked to identified risks, or finalised too close to or after the audit sign-off stage.
Three specific failures:
Risk identified, procedures not responsive. The file identifies an area as a significant risk and then contains standard procedures that do not address that risk specifically.
Conclusion without evidence. A working paper concludes that a position is supportable without documenting what was examined.
Documentation assembled after the conclusion. A file completed after the opinion was formed does not demonstrate that the work informed the opinion.
For a preparer, the equivalent is a technical memo written to justify a decision already taken, rather than a memo that records the analysis that led to it.
The Materiality Trap
Two materiality errors recur.
Interim materiality assessed against annual expectations. IAS 34 requires materiality to be assessed in relation to the interim period data. Applying an annual threshold to a quarter systematically suppresses disclosure, as Post 67 explained.
Individual immateriality used to avoid aggregate assessment. A series of individually immaterial items sharing a common cause may be collectively material. Related party transactions, uncorrected misstatements, and disclosure omissions all behave this way.
A Diagnostic Checklist
The following questions surface most of the patterns above.
Was this schedule rebuilt from the current position, or rolled forward from last year?
Could an item requiring recognition or disclosure exist without generating a ledger entry, and if so, how would we find it?
For each material arrangement, does the accounting follow the substance or the legal form?
At what level of aggregation was this assessed, and why that level?
Which specific date does this requirement turn on, and did we use it?
Which specific rate does this requirement specify, and did we use it?
Does anything recognised in OCI or equity have a corresponding tax, exchange, or remeasurement effect that should also be there?
For every recurring estimate, how did last year's estimate compare to what actually happened?
Does the timing profile of recognition match what the standard requires, rather than what is convenient?
Could a reader reach an independent view from what we have disclosed?
Does the effective tax rate reconciliation balance without an unexplained item?
FAQ
Are these errors caused by not knowing the standards?
Rarely. Most arise from process rather than knowledge: schedules rolled forward, populations built from the wrong source, work performed under time pressure at the end of the cycle, and estimates never tested against outcomes.
Which single control catches the most errors?
Rebuilding rather than rolling forward. Completeness failures are the largest category, and they are almost entirely a consequence of starting from last year's answer instead of this year's population.
Why is the effective tax rate reconciliation such a useful diagnostic?
Because it is arithmetically constrained. If every temporary difference is identified, measured, rated, and allocated correctly, the reconciliation balances with only genuine reconciling items. An unexplained plug is the size of an error somewhere in the computation.
Can the accounting be right and the audit file still be deficient?
Yes, and regulators find this regularly. A file where risks are identified but procedures do not respond to them, where conclusions are recorded without supporting evidence, or where documentation is assembled after the opinion was formed, is deficient regardless of whether the financial statements are correct.
How do you catch errors in items that never touch the ledger?
By running identification from contracts and arrangements rather than from balances. For related parties, embedded leases, commitments, and non-cash transactions, there is no alternative to reading the agreements for the material population.
Enroll with Global Fin X
Technical knowledge prevents the errors that come from not knowing the requirement. The errors in this post mostly come from somewhere else, and recognising the pattern is what prevents them. Our programme covers the full IFRS syllabus with detailed lectures, worked examples grounded in Indian practice, exam-style MCQs, and a dedicated LMS for working professionals.
Enroll Now: Dip IFRS Programme
Faculty profile: www.globalfinx.in/manikanta
This is Post 94 of the Global Fin X IFRS Series. Previous: IFRS in Big 4 Audit Practice: What Associates and Senior Associates Actually Do. Next: Post 95: NFRA Inspection Reports and IFRS: What Indian Auditors Are Getting Wrong.