Earnings manipulation: where aggressive accounting ends and fraud begins

Earnings manipulation almost never looks like a lie in a specific figure: it looks like a series of formally permissible choices, made one after another and always in the same direction. The warning signs fall into several families: a gap between profit and operating cash flow, working capital swelling faster than revenue, expenses shifted onto the balance sheet as assets, games with provisions, drift in the definitions of "adjusted" metrics, and signals around the report itself, such as a change of auditor, of chief financial officer or of the scope of consolidation. Taken individually, each of these signs has an innocent explanation; diagnostic power comes from several families coinciding in the same period, and from that coincidence persisting from one period to the next.

Where the line lies: aggressive accounting versus manipulation

Accounting standards require estimates: the useful life of an asset, the likelihood that receivables will be collected, the moment control over goods passes to the buyer. Wherever there is an estimate there is a range, and choosing the upper end of that range is legitimate. This is what is known as aggressive accounting: the financial statements stay within the rules but systematically present the best possible picture. Manipulation begins where the economic event disappears: a sale that never happened, an expense that was not recognised, a liability moved outside the reporting perimeter. In practice the line is drawn not by the degree of optimism but by reversibility: an aggressive estimate corrects itself over time through weaker future periods, whereas fraud requires a fresh falsification to conceal the previous one. Hence the rule for reading: look not at a single report but at a series. The financial reports section and the glossary are built for comparing periods, not for admiring just one.

Profit on paper, no cash in the bank

The main sign is a persistent gap between net profit and operating cash flow. Profit is built from accruals, cash flow from actual settlements, and manipulating accruals is cheaper. If profit is growing while cash flow stagnates or turns negative for several quarters in a row, the difference has to be explained by working capital or one-off items. When there is no explanation, or the explanation changes from one report to the next, that is a reason to check the other families. The opposite pattern is informative too: cash flow that holds up better than profit more often points to conservative write-downs than to a problem.

Working capital as the trace of sales that never happened

Revenue recognised without cash leaves a trace in accounts receivable. The signs: receivables growing faster than revenue; the average collection period lengthening; sales concentrating in the last days of the quarter; large counterparties appearing with no disclosed history; a rising share of sales to related parties. Inventories leave a similar trace from the other side: channel stuffing first looks like sales growth, and later shows up as returns and discounts. It is worth reading the notes on the allowance for doubtful debts: an allowance that does not grow alongside receivables is an undisclosed change in estimate.

The expense that moved onto the balance sheet

Capitalisation shifts costs out of the income statement and onto the balance sheet, and from there into future periods through depreciation and amortisation. The signs: a wider list of costs being capitalised, longer useful lives, a growing share of capitalised interest and development costs, abnormally slow depreciation despite heavy capital expenditure. Impairment sits right next to this: goodwill and intangible assets that are not tested for impairment when conditions have clearly deteriorated represent a deferred loss, not the absence of one. Comparison with the sector matters more than absolute values, and the list of stocks with comparable issuers will be useful for that.

Provisions and "adjusted" metrics

Provisions are the most controllable lever: over-provisioning in a strong period builds a cushion that is released in a weak one, smoothing earnings. The sign is a suspiciously steady earnings trend alongside volatile revenue and cash flow. The second lever is definitions. "Adjusted" profit and EBITDA are not standardised, so it pays to check the formula in each report: if a new item has been quietly excluded, or if "one-off" expenses recur every year, the adjustment has turned from an explanation into a tool. The same goes for reclassifications in segment reporting and the merging of segments, after which a weak business line is no longer visible.

Signals around the report

Some of the signs are not in the numbers at all. A change of auditor without a clear reason, a modified or qualified opinion, a delayed publication, restated comparative figures for prior periods, the departure of the chief financial officer or chief accountant just before the report, a change in the scope of consolidation, growth in transactions and loans with related parties, unusual management incentives tied to reported profit. An abrupt change in the language of the commentary, such as a previously disclosed metric disappearing or vague phrasing replacing specifics, is also a signal. Publication dates and corporate events are easy to track in the events calendar, and the accompanying news flow in the news feed.

How to use this

Treat it as a checklist, not a verdict: note which families have been triggered, find the issuer's explanation in the notes, and check whether it holds up against the sector and against past periods. A cash test is useful in its own right: dividends are paid out of cash flow, not out of adjusted profit, so a gap between generous payouts and weak cash flow is a sign in itself. The history and charts of payouts can be found in the dividends section.

What these signs will not show

A fair caveat: since 2022 some Russian issuers have cut back their disclosure, and a number of cross-checks, by segment, by related party, by turnover period, are simply unavailable for them. In such cases the absence of warning signs means an absence of data, not clean accounting, and that is an argument in its own right when assessing risk. This article does not contain data on which companies made use of the disclosure relief, or to what extent.