Based on the Sloan (1996) accrual anomaly — firms whose earnings lean on accruals (non-cash book entries) rather than operating cash tend to underperform. A higher QoE score means higher-quality, cash-backed earnings. Composite of CFO/NI (40%), FCF conversion (30%), inverted accrual ratio (20%), and inverted one-time items (10%).
What this is
The Quality of Earnings Screener ranks ~300 US large-cap stocks (market cap > $5B, approximating the S&P 500 + Nasdaq 100 overlap) by how much of reported net income is backed by real operating cash. Firms with low-quality earnings tend to underperform in the 1–3 years after filing — a robust anomaly documented by Sloan (1996) and replicated widely since.
Each component is percentile-ranked within the universe (0–100), then combined with fixed weights below. All scores are relative to today's basket — the #1 stock has the best cash-backed profile among its peers, not on an absolute scale.
Score formula
QoE Score = 0.40 · CFO/NIpct + 0.30 · FCF/NIpct + 0.20 · (1−Accrual)pct + 0.10 · (1−OneTime)pct
• CFO / NI ratio (40%) — TTM operating cash flow ÷ TTM net income. > 1.0 means every dollar of reported profit is backed by more than a dollar of cash.
• FCF conversion (30%) — TTM free cash flow (CFO − CapEx) ÷ TTM net income. > 0.8 is healthy.
• Accrual ratio (20%, inverted) — (NI − CFO) ÷ average total assets. Low or negative = earnings backed by cash. Sloan's original anomaly measure.
• One-time items (10%, inverted) — proxied as |non-operating income / pretax income|. Low = earnings mostly from core operations.
Red flag rules
Flags are informational — they annotate rather than exclude. Two or more flags tag a ticker as HIGH RISK, one as Watch, none as Clean.
• CFO < NI for 2+ consecutive quarters — earnings and cash are diverging the wrong way.
• FCF negative while NI positive — classic capex-driven or working-capital-driven accrual gap.
• Accrual ratio > 10% — Sloan's top-decile accrual zone, historically a shorting candidate.
Config schema
• Universe — top ~300 US stocks by market cap (MC > $5B)
• TTM window — last 4 reported quarters (calendar-end-dated)
• Normalization — per-component percentile rank within today's universe
• Weights — CFO/NI 40 · FCF/NI 30 · (1−Accrual) 20 · (1−OneTime) 10 (sum = 100)
• Direction — composite always ↑ = better earnings quality
• Refresh — weekly, Monday 22:00 UTC (Tue 06:00 Beijing); quarterly fundamentals don't need daily refresh
Caveats — what this does NOT capture
• Valuation — a high-quality earner can still be overpriced; pair with P/E or P/FCF.
• Banks & insurers — CFO is structurally noisy for financials (deposit/premium flows inflate CFO). Treat sector = Financial Services reads with care.
• One-time items proxy — we use |other non-op income| / |pretax| as a rough stand-in for SEC Item 2.02 restatements; aggressive re-segmentations won't show here.
• Point-in-time fidelity — we use calendar-end-dated quarters; a filing lag of 30–45 days is typical. Real-time PIT users should consult FILING_DATE-stamped data.
• No forward-looking info — entirely backward-looking. Guidance revisions, insider trading, analyst expectations are orthogonal signals.
Data source
Company income / cash flow / balance sheet statements — Alva SDK partition equity_fundamentals. Universe from stock_screener MARKET_CAP screener. Price bars from equity spot OHLCV.