Earnings quality and accounting red flags

Key takeaways

• Profit that persistently grows faster than operating cash flow is the most general sign of weakening earnings quality.
• Balance-sheet and cash-flow accruals ratios scale the non-cash part of earnings by net operating assets, so companies of different size can be compared.
• Receivables growing faster than sales (rising DSO), capitalized operating costs and releases of prior reserves are the most common ways earnings are pulled forward.
• The Beneish M-Score combines eight financial-statement indices; scores above -1.78 are commonly treated as a flag, but it is a screen with false positives and false negatives.
• The Altman Z-Score for public manufacturers combines five ratios with weights 1.2, 1.4, 3.3, 0.6 and 1.0; below 1.81 signals distress and above 2.99 relative safety.
• No single metric proves manipulation or distress; a red flag is a reason to ask a specific question and test it against source documents.

What earnings quality means

Analysts judge earnings quality along three lines. The first is cash backing: over time, operating cash flow should track net income, and a widening gap needs an explanation. The second is persistence: earnings from recurring operations are worth more than gains on asset sales, tax settlements or reserve releases, which will not repeat. The third is estimate conservatism: the more of reported profit that depends on management's forecasts of useful lives, collectibility, warranty costs or contract margins, the more room there is for those forecasts to be optimistic.

The tests in this topic run on one example company. Vesper Components is a hypothetical supplier of electronic parts. Over three years its revenue grows steadily and its net income grows faster, which on the surface is an attractive story. The figures below, in $m, show why a diligence team would not accept it at face value.

Cash conversion

The simplest test divides operating cash flow by net income. A ratio around or above 1.0 is normal for a mature business, because depreciation is a non-cash charge that lifts CFO above net income. A ratio falling well below 1.0 means an increasing share of reported profit is sitting on the balance sheet as receivables, inventory, capitalized costs or other accruals rather than arriving as cash.

One year of weak conversion can be innocent: a large customer paid a week late, or the company built inventory ahead of a launch. A trend across several years is harder to explain. Private equity buyers often run the same test on EBITDA, dividing EBITDA less capex and the change in working capital by EBITDA, because EBITDA is the figure they are paying for.

Accruals ratios

Accruals are the part of earnings not yet realized in cash. Scaling them by the size of the business gives a ratio that can be compared across years and companies. Two versions are standard, and both scale by net operating assets (NOA): operating assets (total assets less cash and short-term investments) minus operating liabilities (total liabilities less debt). NOA is the capital tied up in running the business.

The balance-sheet version measures the growth in NOA over the year. If NOA grows much faster than the business, the growth is made of accruals: receivables, inventory, capitalized costs. The cash-flow version measures net income minus the operating and investing cash flows, the portion of profit not explained by cash. Because the change in NOA equals net income less operating and investing cash flow when there are no acquisitions, currency effects or other non-cash transactions, the two versions give the same answer in a clean case. When they diverge in a real filing, the non-cash transactions explain the gap.

Higher accruals ratios mean lower earnings quality. There is no fixed cutoff; the ratio is most useful compared with the company's own history and with peers in the same industry, because a growing business naturally invests in working capital.

Revenue red flags: channel stuffing and DSO creep

Channel stuffing is shipping more product to distributors or retailers than end demand supports, usually near a period end, to meet a revenue target. The distributor may be offered extended payment terms, generous return rights or a promise that unsold goods can be sent back. Under ASC 606 and IFRS 15, rights of return and price concessions should reduce the revenue recognized, but only if they are known and estimated properly, and side agreements are exactly what is not disclosed.

The fingerprint is in receivables. If customers are taking goods they cannot sell, they pay slowly, so days sales outstanding (DSO) rises. Other signs include revenue concentrated in the last weeks of each quarter, rising returns or credit notes early in the following period, falling deferred revenue alongside rising revenue, and inventory building up at distributors when the company discloses channel data.

Costs, reserves, related parties and auditors

On the cost side, the most common technique is capitalizing costs that should be expensed. Operating costs relabeled as software development, customer acquisition, or major repairs move to the balance sheet, lifting EBITDA and operating cash flow at once. The trace is capitalized costs, or capex, growing faster than revenue, and intangible assets or other non-current assets rising without an acquisition.

Reserves are the second. Companies estimate allowances for doubtful debts, warranty provisions, inventory obsolescence and restructuring liabilities. Over-reserving in a good year and releasing the excess in a weak one, sometimes called a cookie jar reserve, smooths earnings. Cutting the rate at which new provisions are made does the same thing more quietly. The trace is a provision balance falling as a share of the activity it covers, or a release shown in the notes.

Related-party transactions, such as purchases from or sales to entities owned by management or major shareholders, can be priced away from market and shift profit between the company and its insiders. They are disclosed in the notes under US GAAP and IFRS, and their size and terms deserve reading in full. A change of auditor, especially one following a disagreement, a qualified opinion, a delayed filing, or a move to a much smaller firm, is a governance signal rather than an accounting one, and it always deserves an explanation. Frequent changes of CFO send the same signal.

The Beneish M-Score

The Beneish M-Score, developed by Messod Beneish from a study of companies subject to enforcement actions for earnings manipulation, combines eight indices built from two consecutive years of financial statements into a single score. Each index captures a pattern associated with manipulation, such as receivables growing faster than sales, deteriorating gross margins that create pressure to inflate results, rapid growth, slowing depreciation, and high accruals. The model was estimated on public US companies, excluding financial firms.

Scores above -1.78 are the threshold most commonly used to flag a company as a possible manipulator in the eight-variable model; some practitioners use a lower cutoff of -2.22, which flags more companies. The score is a probability model, not a detector. It produces false positives, especially for fast-growing companies whose indices naturally resemble the manipulation pattern, and false negatives, when manipulation takes a form the eight indices do not capture. It is a screening tool for deciding where to look harder.

The Altman Z-Score

The Altman Z-Score, developed by Edward Altman, predicts financial distress rather than manipulation. The original model was estimated on publicly traded US manufacturing companies and combines five ratios: liquidity (working capital to total assets), cumulative profitability (retained earnings to total assets), operating returns (EBIT to total assets), market-based solvency (market value of equity to total liabilities) and asset turnover (sales to total assets). A low score indicates a profile similar to companies that went on to fail.

The original model divides results into three zones. Above 2.99 is the safe zone, 1.81 to 2.99 is the grey zone, and below 1.81 is the distress zone. Altman later published revised versions for private companies, which use book rather than market equity, and for non-manufacturers and emerging markets, which drop the sales ratio. Each has its own coefficients and zones, so apply the version that fits the company rather than the original formula to every business.

A red flag checklist

The checklist below gathers the signals in this topic with the question each one should prompt. In practice, diligence teams work through it with the company's own monthly data, which is more revealing than annual figures, and follow each flag to source documents: contracts, bank statements, aged receivables listings, provision workings and board minutes.