Technical Methodology
ValueMarkers Composite Index
A cross-stock percentile-ranked composite of 120 fundamental indicators across five investment pillars, designed to surface the most attractive value opportunities globally.
1. Background and Motivation
The academic literature on equity return prediction is rich but fragmented. Fama and French (1992) demonstrated that value (P/B) and size predict cross-sectional returns. Piotroski (2000) showed that a simple nine-signal accounting score separates winners from losers within cheap stocks. Novy-Marx (2013) established gross profitability as the strongest known predictor of future abnormal returns. Sloan (1996) documented the accrual anomaly: stocks with high accruals underperform those with high cash earnings.
Individual practitioners - Graham, Buffett, Munger, Greenblatt - have proposed their own frameworks: intrinsic value estimation, return on invested capital, Magic Formula, owner earnings. These approaches share a common thread: they ask two questions simultaneously. What do you pay? And what do you get?
The challenge for a modern platform is synthesis. A single metric - whether P/E, ROIC, or Piotroski F-Score - captures one dimension of investment quality but misses others. The ValueMarkers Composite Index (VMCI) is designed to combine 120 rigorously sourced fundamental signals into a single, interpretable 0–100 rating that reflects the full dimensionality of investment quality, without sacrificing transparency.
Every score is derived from first principles. No black boxes. Every indicator has a verifiable formula. Every weight has a stated rationale grounded in empirical finance literature.
2. VMCI Architecture
The VMCI is a five-pillar composite. Each pillar represents a distinct investment question. Pillar weights reflect their empirical predictive power for future equity returns and their importance in professional value investing frameworks.
| Pillar | Weight | Indicators | Investment Question |
|---|---|---|---|
| VALUE | 35% | 28 | What do you pay vs. what is it worth? |
| QUALITY | 30% | 41 | How durable and exceptional is this business? |
| INTEGRITY | 15% | 16 | Is the balance sheet trustworthy? |
| GROWTH | 12% | 15 | Is the business compounding forward? |
| RISK | 8% | 20 | What could go wrong? |
| Total | 100% | 120 |
Design Rationale
VALUE receives the largest weight (35%) because valuation is the primary determinant of long-run returns: buying at a discount to intrinsic value is the central tenet of value investing (Graham, 1949; Fama-French, 1992). QUALITY (30%) is weighted second because high-return businesses compound capital over time - cheap low-quality businesses destroy it (Novy-Marx, 2013; Buffett, 1989 letter). INTEGRITY (15%) ensures the numbers can be trusted. GROWTH (12%) captures compounding power but is weighted modestly because growth without quality is worth little. RISK (8%) provides a downside filter without dominating the composite.
3. The Five Pillars
The VALUE pillar asks: relative to what a rational buyer would pay for this stream of cash flows, is the current market price attractive? It covers traditional price ratios (lower = better), yield-based measures (higher = better), intrinsic value estimates, and shareholder yield signals.
| Indicator | Formula | Direction |
|---|---|---|
| P/E | Price / EPS (TTM) | ↓ |
| Fwd P/E | Price / Forward EPS | ↓ |
| P/B | Price / Book Value per Share | ↓ |
| P/S | Market Cap / Revenue | ↓ |
| P/CF | Price / Operating Cash Flow per Share | ↓ |
| P/FCF | Price / Free Cash Flow per Share | ↓ |
| PEG | P/E / 5Y EPS Growth Rate | ↓ |
| EV/EBITDA | Enterprise Value / EBITDA | ↓ |
| EV/Revenue | Enterprise Value / Revenue | ↓ |
| EV/FCF | Enterprise Value / Free Cash Flow | ↓ |
| EV/EBIT | Enterprise Value / EBIT | ↓ |
| Earnings Yield | EPS / Price × 100 (%) | ↑ |
| FCF Yield | FCF / Market Cap × 100 (%) | ↑ |
| Owner Earnings Yield | (NI + D&A − CapEx) / Market Cap × 100 | ↑ |
| Gross Profit Yield | Gross Profit / EV × 100 (Novy-Marx factor) | ↑ |
| Cash Yield | Cash & Equivalents / Market Cap × 100 | ↑ |
| Graham Number | √(22.5 × EPS × Book Value per Share) | ↑ |
| Normalized P/E | Price / 5-Year Average EPS | ↓ |
| DCF Intrinsic Value | Σ(FCFₜ / (1+WACC)ᵗ) + Terminal Value | ↑ |
| Margin of Safety | (DCF Value − Price) / DCF Value × 100 (%) | ↑ |
| Price / Tangible Book | Price / Tangible Book Value per Share | ↓ |
| Price / Graham | Price / Graham Number | ↓ |
| EV / Invested Capital | EV / (Total Equity + Net Debt) | ↓ |
| Dividend Yield | Annual Dividends / Price × 100 (%) | ↑ |
| Buyback Yield | Buybacks / Market Cap × 100 (%) | ↑ |
| Shareholder Yield | Dividend Yield + Buyback Yield (%) | ↑ |
| Graham Net-Net | Current Assets − Total Liabilities vs Price | ↑ |
| Analyst Upside | (Consensus Target − Price) / Price × 100 (%) | ↑ |
The QUALITY pillar asks: does this business have an enduring competitive advantage that will protect and grow returns over time? It spans returns (ROIC, ROE, ROCE), margin analysis, cash flow quality, asset efficiency, moat consistency signals, and two VMCI-exclusive composite metrics.
Key VMCI-Exclusive Metrics
- ROIC–WACC Spread: ROIC − WACC. Positive spread means the business creates economic value; negative destroys it. The primary Buffett/Munger test for business quality.
- Sustainable Growth Rate: ROE × (1 − Payout Ratio). The maximum growth rate achievable without new external financing - Buffett's organic compounding measure.
- FCF / Revenue: Free Cash Flow Margin - how much of each revenue dollar converts to real shareholder cash.
- Dividend Coverage Ratio: FCF / Annual Dividends Paid. Measures whether the dividend is sustainably funded from operations, not debt.
WACC Estimation (CAPM)
Kₑ = 4.5% + β × 5.5% (β clamped [0.3, 3.0])
Kd = (Interest Expense / Total Debt) × (1 − Tax Rate)
WACC = (E / (E+D)) × Kₑ + (D / (E+D)) × Kd
The INTEGRITY pillar asks: can we trust the numbers on this balance sheet, and is the company financially sound enough to survive a downturn? It covers solvency (debt ratios, interest coverage), liquidity (current, quick, cash ratio), three academically validated distress and manipulation models, and working capital efficiency.
Piotroski F-Score
9-signal accounting model (Piotroski 2000). Tests profitability, leverage, liquidity, and operating efficiency. Score 0–9; ≥7 = strong.
Altman Z-Score
Bankruptcy prediction model (Altman 1968). Z = 1.2×X1 + 1.4×X2 + 3.3×X3 + 0.6×X4 + 0.999×X5. Z < 1.81 = distress zone.
Beneish M-Score
Earnings manipulation model (Beneish 1999). M = −4.84 + 0.92×DSRI + 0.528×GMI + ... M > −1.78 = possible manipulation.
The GROWTH pillar captures how fast the business is expanding its economic footprint. It covers revenue, EPS, net income, FCF, and book value growth at 1Y, 3Y, and 5Y horizons, plus consistency signals (rev_stability, eps_stability) that reward smoother compounders over lumpy growers.
Growth is weighted at 12% rather than higher because growth without quality is a trap. Fast-growing businesses with poor ROIC and weak free cash flow conversion destroy value, not create it. The QUALITY pillar's ROIC–WACC spread and sustainable growth rate capture whether the growth is value-creating.
The RISK pillar measures downside exposure across four dimensions: market sensitivity (beta, volatility, max drawdown), capital allocation risk (capex intensity, working capital cycle), cost structure risk (operating leverage, SGA burden), and capital return risk (payout ratio, debt service coverage).
Risk is intentionally the smallest pillar (8%) because overly penalising volatility would bias the composite against high-quality growth companies that naturally carry higher beta. The INTEGRITY pillar handles balance sheet risk; this pillar focuses on operational and market risk.
4. Scoring Algorithm
Step 1 - Cross-Stock Percentile Ranking
For each of the 120 indicators, we collect all non-null values across every stock in the coverage universe. Each stock is then assigned a 0–100 percentile rank for that indicator.
percentile_rank(v, population, direction) =
rank = count(x ∈ population where x < v) / |population|
score = rank × 100 if higher_is_better
score = (1 − rank) × 100 if lower_is_better
Indicators with null values for a stock are excluded from that stock's pillar average.
Step 2 - Pillar Score
Each pillar's score is the arithmetic mean of the percentile ranks of all scored indicators in that pillar.
pillar_score(p) = mean(percentile_rank(i) for i in pillar p)
# Result: 0–100 per pillar
Step 3 - VMCI Composite
The VMCI is the weighted sum of the five pillar scores.
VMCI = 0.35 × VALUE + 0.30 × QUALITY + 0.15 × INTEGRITY
+ 0.12 × GROWTH + 0.08 × RISK
# Result: 0–100 (continuous, two decimal places)
Interpretation
0–40
Weak fundamentals relative to peers
40–60
Average - watch for specific signals
60–75
Above-average - investigate further
75–100
Top-tier - strong across all pillars
5. Value Score - Global Ranking System
The Value Score is a separate 0–100 metric designed for a different purpose than the VMCI. Where the VMCI is a nuanced, relative composite across 120 dimensions, the Value Score is a conviction signal: it identifies the stocks where the evidence for value, quality, and safety converges most clearly. Every stock in the universe receives a single rank - so you can say "AAPL is #1,243 of 93,000+ globally-tracked stocks."
Price Test
Are you being paid well to own this? Scored against absolute thresholds - independent of market conditions.
- FCF Yield: ≥10% = 100, ≥7% = 85, ≥5% = 70...
- Earnings Yield: ≥10% = 100, ≥7% = 85...
- Margin of Safety: ≥50% = 100, ≥30% = 85...
- EV/EBIT: ≤8× = 100, ≤12× = 85...
Quality Test
Is this a durable, high-return business? Conviction thresholds used by professional value investors.
- ROIC: ≥25% = 100, ≥20% = 90, ≥15% = 78...
- Gross Margin: ≥60% = 100, ≥45% = 85...
- FCF Conversion: ≥90% = 100, ≥75% = 85...
- ROIC Consistency: score 0–100 based on stability
Safety Test
Is the financial foundation sound? Academically validated models, not heuristics.
- Piotroski F-Score: 9 = 100, 8 = 90, 7 = 78...
- Beneish M-Score: <−2.5 = 100, <−1.78 = 72...
- Altman Z-Score: ≥3.5 = 100, ≥3.0 = 88...
- Hard floor: M > −1.78 → Safety ≤ 30
Composite Formula
Value Score = 0.40 × Price Test + 0.40 × Quality Test + 0.20 × Safety Test
# If Beneish M > −1.78: Safety Test = min(Safety Test, 30)
# Requires ≥5 of 11 inputs; rescales if a test block is unavailable
The Value Score uses absolute thresholds (not percentile ranking) so it can be computed per-stock without requiring the full universe. All stocks are then sorted by their Value Score to produce the global rank. A stock scoring 80+ has cleared all three tests simultaneously - cheap price, high quality, and clean books.
6. Data Sources and Global Coverage
All fundamental data is sourced from Financial Modeling Prep (FMP), which aggregates SEC and EDGAR filings, IFRS regulatory submissions, and exchange data across 73 global stock exchanges. Data is refreshed daily for active stocks and cached with a 6-hour ISR window for display.
US Markets (Full Coverage)
- Income statement (20Y+ history)
- Balance sheet (20Y+)
- Cash flow statement
- Annual & quarterly filings
- Analyst estimates & price targets
- Short interest (FINRA data)
- Insider transactions (Form 4)
- Institutional holdings (13-F)
International Markets
- Income, balance, cash flow statements
- All 120 core VMCI indicators
- Multi-currency with USD normalization
- IFRS-based fundamental parsing
- Analyst estimates (major exchanges)
- Note: short interest US-only (FINRA)
- Note: insider data varies by jurisdiction
Null Handling
Indicators that cannot be computed due to missing data return null and are excluded from their pillar's average. A stock's VMCI is always computed on available data - the pillar score is the mean of available indicators, not penalised by nulls. This ensures fair cross-exchange comparison without requiring identical data availability.
7. Limitations and Disclaimers
Backward-looking inputs. All indicators are derived from historical financial statements. The VMCI does not predict future performance - it assesses current fundamental quality relative to peers. A high VMCI stock can still be a poor investment if its business deteriorates.
Sector heterogeneity. The cross-stock percentile ranking does not distinguish sectors. Banks, utilities, and software companies are scored against each other on the same 120 indicators. Some indicators (e.g. inventory turnover) are less meaningful for financial firms. We recommend using the individual pillar scores in conjunction with sector context.
DCF model sensitivity. The intrinsic value estimate and margin of safety are highly sensitive to growth rate and WACC assumptions. These should be treated as starting points for analysis, not definitive verdicts. The platform's interactive DCF tool allows users to stress-test assumptions.
Not financial advice. The VMCI, Value Score, and all indicators are analytical tools for research purposes only. They do not constitute investment advice. Past fundamental strength does not guarantee future returns. Always conduct your own due diligence.
Frequently Asked Questions
What is the VMCI (ValueMarkers Composite Index)?+
How is the VMCI different from other composite scores?+
What is the Value Score and how is it different from the VMCI?+
What data does ValueMarkers use and which markets are covered?+
How is WACC estimated?+
What does the Beneish M-Score hard floor mean?+
Apply the VMCI to Any Stock
120 indicators, 5 pillars, global rank. All in one platform.
Explore More
Investing Tools
- Stock ScreenerScreen 85,000+ stocks by 120 indicators
- VM Score LeaderboardTop stocks ranked daily by VMCI
- Guru StrategiesApply Buffett, Graham, Greenblatt strategies
- DCF CalculatorCalculate intrinsic value, free
- Global MarketsMacro overview across 73 exchanges
- Stock CompareSide-by-side company comparison