Quality means an explainable result
SEC guidance on reading a 10-K draws attention to non-GAAP measures and the need to understand how they differ from comparable reported figures. Source: **curr-stat-sec-read**, Item 8, non-GAAP financial measures. This is a U.S. disclosure context, not a universal legal rule. For educational analysis, the principle is to rebuild the reconciliation and assess the meaning of exclusions independently.
Educational reconciliation of one-off effects
Fictional reported profit is 180. It includes an asset-sale gain of 50, a reorganisation expense of 30 and marketing described controversially as “one-off” at 20. Excluding only the sale and reorganisation gives 180 − 50 + 30 = 160. Adding back marketing gives 180. The difference of 20 is a substantive judgement that needs support, not a technical footnote.
If reorganisation expenses recur every year, calling them “one-off” does not automatically make them irrelevant to normal burden. Income and expenses should be treated symmetrically: do not exclude only adverse events while retaining random positive ones. For comparison, keep reported and analytical results side by side without declaring the second to be the only true figure.
Accruals require movement in balances
Profit may rise while receivables grow and cash lags. Explanations differ: longer payment terms, seasonality, a new customer, a changed perimeter or weaker collection. Check ageing and the movement of the loss allowance. For inventory, examine composition and write-offs. For liabilities, ask whether an improvement in cash is only a temporary payment delay.
IAS 7 describes the indirect transition from result to operating cash through non-cash items and accruals. Source: **curr-stat-ias7**, About, indirect method. A low cash-to-profit ratio for one year is therefore not proof of poor quality; the business cycle and multi-year relationship matter. A persistent unexplained gap deserves a separate review rather than dismissal.
Estimates do not necessarily mean manipulation
A change in useful life, provisions or other assumptions can materially affect current result. IAS 8 distinguishes new estimates from errors of prior periods. Source: **curr-stat-ias8**, About. Reconstruct the basis and timing of new information. Saying “profit was managed” requires more evidence than the fact of an estimate change or rising accruals.
A common mistake is using EBITDA as automatic evidence of quality because it excludes some non-cash expenses. Depreciation relates to the consumption of previously paid resources, while cash investment may still be required. Another mistake is declaring operating cash completely free for owners without considering sustaining investment and mandatory payments.
Final question set
Check repeatability of causes, symmetry of adjustments, movement in accruals, policy changes and the connection with cash over several periods. Note disclosure boundaries separately. If an adjustment cannot be reconstructed, do not publish an exact “clean” result as a confirmed fact.
This material does not detect fraud or confirm professional assurance. Continue with [statement adjustments](/guides/financial-statement-adjustments/), [profit versus cash](/guides/profit-vs-cash/) and [income statement](/guides/income-statement/).
Separate a signal from sufficient evidence
A combination of rapidly growing receivables, slowing cash and frequent earnings adjustments deserves attention but permits several explanations. For each signal, state a supporting document and an alternative cause still to be tested. This does not excuse weak disclosure; it prevents an unsupported leap from observation to accusation. If no explanation is found, the result may be limited confidence rather than a categorical diagnosis.
Check how metric definitions change. If a company repeatedly rebuilds adjusted earnings, compare old and new methods on one period where possible. The difference isolates the effect of changing the frame from the operation of the business. Do not combine sums with different exclusions in one series without warning the reader.
Also test forecast-versus-actual symmetry. If management promised that costs would disappear, check whether they disappeared or acquired a new name. This tests repeatability, not people’s intentions. Without enough data, do not infer motives. A reproducible result is a list of confirmed causes, unresolved questions and sensitive assumptions that another reader can test from the same original statements.
E/R: how moving costs makes a ratio look better
For a telecommunications example, let E be channel-rental expense for the period, and R revenue for the same perimeter and period. E/R measures the share of selected expense, not overall profitability. At E = 420 and R = 1,000, the share is 42%. If revenue falls to 800 while the contracted cost of already ordered channels remains 420, the share becomes 52.5%. The deterioration may reflect under-utilised contracted capacity even without higher tariffs.
Suppose an accountant improperly moves 84 of these current expenses into an asset, although no separately recognisable resource was created under the assumptions. Reported expense becomes 336, E/R returns to 42%, profit before tax is overstated by 84 and assets are also overstated by 84; the cash payment for channels does not fall. This is an arithmetic signal of reclassification, not proof of fraud from one stable percentage. Actual capitalisation is determined by the criteria of the applicable standard: **curr-stat-ias16-text**, 7, and **curr-stat-ias38**, About. Intent and personal responsibility cannot be inferred from this exercise.
An outside reader can compare expense movement with capital additions, contract periods and utilisation. A constant share during weaker demand deserves the question “what changed in the numerator?” but also permits lawful explanations: price revision, contract termination or improved use. Without primary documents, none can be declared established. The numbers and events in this model do not concern WorldCom.
Internal audit: test the exclusion route, not only the average ratio
Internal audit addresses risk management, internal controls and governance; its independent assessment is not the same as the accounting team performing the controls themselves. IIA describes it as an independent third line with objective assurance and advisory roles: **ca-iia-position**, questions on purpose and the Three Lines Model. External financial-statement audit, internal process audit and the finance director have different tasks and recipients; one does not replace the others.
In our model, accounting must first support why 84 meets the conditions for recognising an asset. A control may require technical confirmation of the resource and separate approval of an unusual manual entry. Without becoming the entry’s author, internal audit tests whether the control exists and operated: select the entry from the original journal, link it to the contract and supplier invoice, inspect date and approval authority, and reconcile the accounting breakdown with the asset register. Testing a final export prepared by the same department does not replace testing the entry’s origin.
To assess how the function operates, ask who receives significant findings, who can restrict access to data and how disagreements are escalated. These are questions about the particular function, not a promise that every department called “audit” is independent. If the manager who approved the entry can also prohibit its review, disclose the conflict and discuss it with the oversight body. Correcting the entry personally would create a self-review risk.
The result should distinguish fact, basis and open question. Preserving the original document version and later testing the corrected control is more useful than a general assurance that “a review was performed”. This route is an independently written exercise, not an account of WorldCom employees’ system access, conversations, court findings or external auditor history.
Acquisitions and disposals: start with the same perimeter
Growth in total revenue after an acquisition is not organic growth. In a separate model, the buyer has revenue of 1,000 for the year, while the business acquired on 1 July has 120 in the first half and 140 in the second. Under the stated conditions, the consolidated annual report includes 1,000 + 140 = 1,140. The following year contains the buyer at 1,060 and the acquired business at 280, for 1,340. Reported growth is 17.54%, but different inclusion intervals for one business are being compared.
Build an explicit “as if acquired at the beginning of the first year” base: 1,000 + 120 + 140 = 1,260. Suppose the buyer sold goods worth 20 to the acquired business in the first half, included in those separate revenues; after consolidation this is intra-group turnover. The compatible conditional base is 1,240; the next year has no intra-group turnover in the stated figures. Comparable growth is therefore (1,340 / 1,240 − 1) × 100% = 8.06%. It cannot be called verified organic growth without controlling for other price, currency and business-composition changes.
Profit needs its own bridge. Let the buyer’s actual profit, already including the subsidiary’s second half, be 100. The acquired business’s pre-acquisition profit is 12; additional half-year depreciation on revalued acquired assets is 4; financing expense for the purchase is 3. Conditional annual profit on the agreed basis is 100 + 12 − 4 − 3 = 105 before tax effect, not 112. Assume the relevant second-half expenses are already included in 100, and intra-group goods were sold to external customers before year-end, so the model has no additional unrealised profit. Synergies are not added as an established fact.
IFRS 3 B64(q) provides disclosures of the acquired business’s post-acquisition revenue and result and the combined organisation as if acquired at the beginning of the annual period: **ca-ifrs3-text**. Our example explains comparability; it does not claim that all required disclosures were made. For a disposal, use the symmetric method: remove the departed perimeter’s revenue and result from the comparative period, show sale income separately and show remaining common expenses. This is an analytical base, not an official discontinued-operations classification or a forecast of future result.