Accruals are the portion of reported earnings that is not backed by cash — the bookkeeping that converts cash flows into earnings under accrual accounting. Working-capital signals examine how the operating-asset and operating-liability accounts (receivables, inventory, payables, deferred revenue) move relative to revenue. The core tension is simple but powerful: earnings can be inflated almost without limit through accruals (booking a sale before cash arrives, capitalizing costs, deferring expense recognition), but cash is much harder to fake. A persistent gap between accounting profit and operating cash flow — and abnormal swelling of working-capital accounts — is one of the most empirically validated red flags in fundamental analysis. It serves two related purposes: detecting outright manipulation, and, more broadly, gauging earnings quality (how persistent and cash-backed reported profit is).

How it's calculated / formed

There are two ways to measure total accruals, and the distinction matters:

Balance-sheet approach (Sloan 1996). Accruals = (ΔCurrent Assets − ΔCash) − (ΔCurrent Liabilities − ΔShort-term Debt − ΔTaxes Payable) − Depreciation & Amortization, typically scaled by average total assets. This captures the change in non-cash net working capital, less D&A.

Cash-flow approach (Hribar & Collins 2002). Accruals = Earnings (income before extraordinary items and discontinued operations) − Cash Flow from Operations, both taken directly from the cash-flow statement, then scaled by average total assets. Hribar and Collins showed the balance-sheet method introduces measurement error for firms with mergers, acquisitions, divestitures, discontinued operations, or foreign-currency translation, because those events move balance-sheet accounts without corresponding accruals (a "non-articulation" error). The cash-flow method is now the academic standard.

Targeted working-capital ratios are used alongside the aggregate measure. The diagnostic intuition is that working-capital accounts should grow roughly in line with sales; divergence is the flag:

• Days Sales Receivable Index (DSRI) — a component of the Beneish M-Score — compares days-sales-outstanding year over year. A reading well above 1.0 flags receivables outrunning revenue.
• Sales Growth Index (SGI) and inventory growth vs. COGS/sales — inventory rising faster than sales suggests overproduction or unsalable goods.
• CFO-to-Net-Income ratio — persistently below 1.0 means earnings are not converting to cash.
• Cash conversion cycle (DSO + DIO − DPO) — a lengthening cycle signals deteriorating working-capital efficiency.

How it's used in practice

Analysts use these signals in two modes. As a forensic screen, the classic pattern is revenue recognized faster than cash collected: receivables (or contract assets / unbilled revenue) balloon, the CFO/NI ratio falls below one, and net income keeps rising while operating cash stagnates. Channel stuffing — overselling to distributors to pad a quarter — shows up exactly this way, as does premature or fictitious revenue. Inventory growing faster than sales points to demand weakness or capitalized costs propping up gross margin. Deferred-revenue declines while revenue rises can signal pulling future sales forward.

As an earnings-quality filter, investors prefer firms whose profits are cash-backed (low accruals, CFO ≥ net income). High-accrual earnings are mechanically less persistent — they reverse — so a quality-conscious investor discounts them even absent any fraud. This is the heart of the Sloan finding below, and it is also embedded in composite scores like Piotroski's F-Score (which rewards CFO > net income) and the Beneish M-Score (which weights DSRI, asset-quality, and accruals).

Adoption, debate & evidence

The intellectual anchor is Richard Sloan's 1996 paper ("Do Stock Prices Fully Reflect Information in Accruals and Cash Flows About Future Earnings?", The Accounting Review). Sloan documented that the accrual component of earnings is less persistent than the cash component, yet investors price the two as if equally durable. A hedge portfolio long low-accrual and short high-accrual firms earned a size-adjusted abnormal return commonly cited as roughly 10% in the first year after portfolio formation (the often-quoted figure is ~10.4%; some replications and lower-cost reconstructions report less — Quantpedia, for example, reports ~7.5% per annum). This "accruals anomaly" is one of the most studied effects in accounting.

The honest landscape is mixed on tradability:

• As an earnings-quality concept, the evidence is robust. The negative relation between accruals and future earnings persistence is repeatedly confirmed and is mainstream in accounting curricula and CFA material.
• As a standalone trading anomaly, the edge has decayed. Multiple researchers report the long-short accrual strategy weakened after roughly 2002–2003, consistent with the pattern of anomalies fading once published and arbitraged. Some studies (e.g., post-2008) note partial rebounds, and the result is sensitive to transaction costs, small-cap and illiquid-stock concentration, and idiosyncratic-risk limits to arbitrage. Out-of-sample and international results are inconsistent, raising data-mining concerns.

So: working-capital/accruals signals are a strong quality and red-flag tool with weak standalone alpha today. The forensic application (catching deterioration in a specific firm) remains highly valuable; the mechanical factor application is contested.

Strengths & limitations

When it works: Best as a confirming screen on a single firm over several periods, where you can see the trend in CFO/NI, receivables-to-sales, and inventory-to-sales together. The signal is hardest for management to fake because it triangulates the income statement against the cash-flow statement and balance sheet simultaneously.

When it fails — and the #1 misuse: Treating high accruals as automatic evidence of fraud. Accruals legitimately spike for fast-growing firms (working capital must expand to fund growth), seasonal businesses, and firms with genuine credit-term changes. The signal must always be normalized against growth and peers. Subscription/SaaS and other deferred-revenue models invert the usual logic — there, rising deferred revenue (a liability) is a healthy sign. Balance-sheet-derived accruals are corrupted by M&A and FX (use the cash-flow method). And any single flag is noisy: these signals are inputs to composite scores (Beneish, Piotroski, Dechow's F-score), not verdicts on their own.

Sources

• Sloan, R. (1996), "Do Stock Prices Fully Reflect Information in Accruals and Cash Flows About Future Earnings?", The Accounting Review 71(3):289–315 — original accruals anomaly; ~10% size-adjusted hedge return in year one (Quantpedia's reconstruction reports ~7.5%/yr).
• Hribar, P. & Collins, D. (2002), "Errors in Estimating Accruals: Implications for Empirical Research", Journal of Accounting Research 40(1):105–134 — cash-flow vs balance-sheet measurement error.
• Quantpedia, "Accrual Anomaly" — formula, post-2002 weakening, decay/data-mining caveats: https://quantpedia.com/strategies/accrual-anomaly
• AnalystPrep (CFA L1), "Accounting Warning Signs" — receivables/inventory vs sales, CFO/NI: https://analystprep.com/cfa-level-1-exam/financial-reporting-and-analysis/accounting-warning-signs/
• AAII, "16 Red Flags to Watch for in a Company's Financial Statements": https://insights.aaii.com/p/16-red-flags-to-watch-for-in-a-companys
• Beneish, M. (1999), "The Detection of Earnings Manipulation" — DSRI and accrual components of the M-Score.

Disputed: The tradable accruals anomaly's persistence is genuinely contested (decay post-2002 vs. claimed rebounds); the earnings-quality relation is not. Quantpedia rates the standalone anomaly's confidence as "weak."