Quality Score Framework

The Quality Score Engine evaluates public equities across twelve fundamental and market structure dimensions. Each company receives a composite score out of 100 that is mapped to a letter grade (A+ through D) and paired with risk and confidence signals.

Core Purpose

• Create a repeatable, explainable view of business quality rooted in Benjamin Graham-style defensive investing principles.
• Blend fundamental health, capital efficiency, and market sentiment into a single benchmark.
• Highlight predictive risk factors so analysts can gauge upside resilience and downside exposure.

The 12 Dimensions

1. Earnings Quality — Stability of operating income, revenue quality, and accruals ratio.
2. Balance Sheet Strength — Debt-to-equity trend, interest coverage, and liquidity runway.
3. Capital Efficiency — Return on invested capital (ROIC), asset turnover, and reinvestment discipline.
4. Cash Flow Reliability — Free cash flow margin, conversion of earnings to cash, and payout sustainability.
5. Margin Durability — Gross/operating margin trend, variance, and industry-relative positioning.
6. Growth Trajectory — Revenue and EPS CAGR normalized against sector baselines.
7. Valuation Fairness — Forward EV/EBITDA, price-to-free-cash-flow, and earnings yield reasonability.
8. Market Position — Competitive moat strength, customer retention, and pricing power indicators.
9. Innovation Velocity — R&D intensity, patent issuance, and product launch cadence.
10. Governance & Stewardship — Insider alignment, capital allocation track record, and board independence.
11. Risk Controls — Volatility regime, downside deviation, and drawdown recovery speed.
12. ESG Resilience — Carbon intensity, social impact scoring, and governance controversies monitor.

Weighting Model

• Baseline weights favor durable quality: balance sheet (15%), cash flow (12%), capital efficiency (10%), and margin durability (10%).
• Growth, valuation, and market position share 30% to balance upside with defensibility.
• ESG resilience and governance carry 8% to reflect long-horizon risk mitigation.
• Analyst overrides can tune weights ±5% per dimension; overrides are logged for auditability.

Normalization Pipeline

1. Collect raw metrics and standardize to rolling five-year z-scores.
2. Winsorize the top/bottom 2.5% outliers for stability.
3. Scale each dimension to a 0–100 score using sector-specific percentiles.
4. Apply weights, then blend into an overall score with confidence interval estimation.

Confidence Levels

• High (≥75%) — Data coverage is complete, signals are directionally consistent, and the model produced low variance across simulation runs.
• Medium (50–74%) — Some metrics required interpolation or conflicting signals increased model variance.
• Low (<50%) — Sparse data or high dispersion; treat directional calls as exploratory rather than investable.

Risk Interpretation

• Risk Level is calculated from downside deviation, liquidity stress tests, and valuation froth indicators.
• A rising risk level with stable quality implies resilience but shrinking margin of safety.
• High risk combined with low confidence flags names that need deeper analyst review before allocation.

Benchmarking & Peer Percentiles

• Sector Percentile compares each score to the company’s direct peer group, helping analysts spot outliers.
• Predicted Quality Change simulates next-quarter movement based on momentum signals and macro drivers.
• Combining percentile rank with predicted change surfaces improving franchises early.

Recommended Workflow

1. Start with the overall grade to triage coverage lists.
2. Review the dimensions scatterplot to spot strengths vs. drag factors.
3. Check confidence and risk levels to size positions appropriately.
4. Use predicted quality change to stage follow-up research or rebalance alerts.

Compliance & Data Sources (Snapshot)

• Fundamental data: Polygon.io, Alpha Vantage, and company filings.
• Market structure data: IEX Cloud and Tiingo streaming feeds.
• ESG overlays: MSCI sector composites and Sustainalytics summaries.
• Model assumptions reviewed quarterly; audit trail stored for regulatory review.