📐 Platform Methodology

The research that runs
billion-dollar funds — built for every investor.

The valuation tools Wall Street uses — built for every investor.

Quantitative hedge funds and institutional asset managers pay millions to build models on the same 16 academic papers that power The Compound Family. We apply them directly, openly, and transparently — so every retail investor has access to the same analytical rigor. No subscription required to see the scores.

⚠ Educational purpose only. All scores, analyses and AI interpretations are provided for informational and educational purposes only. Nothing on this page or on thecompoundfamily.com constitutes investment advice or a recommendation to buy or sell any security. All investment decisions are solely your responsibility. Full disclaimer →
6
Scoring factors
60+
Financial metrics
23
Academic papers
10Y+
Historical data
5
Included tools
Stock Analysis Terminal

Six factors. One overall score.

Every stock receives a score from 0 to 100 based on six independent factors. Each factor is weighted differently depending on the company's sector — because what matters for a Utility is not the same as what matters for a Software company. Seven sector profiles for standard industries.

For Banks, Insurance, and REITs — a fully separate methodology applies. Standard metrics like FCF Margin or Debt/EBITDA are structurally misleading for these sectors. Instead, sector-specific metrics are used and each score is z-score normalized against sector peer benchmarks, following the AQR Quality Minus Junk (QMJ) methodology (Asness et al., 2019).

📈 Business Quality ~27%

Is this a good business? We measure revenue growth, free cash flow generation, and profit margins over 5–10 years. A great business earns more cash than it spends — consistently. For Banks: NIM, ROTCE, ROA, Efficiency Ratio. For REITs: FFO Yield, P/FFO, Occupancy Rate, AFFO/Share.

Revenue Growth 5Y FCF Margin TTM Net Margin FCF Trend NIM · ROTCE (Banks) FFO · AFFO (REITs)
💰 Capital Allocation ~18%

How does management use the money? We reward buybacks, dividends, and smart reinvestment. We penalise excessive share dilution — the silent tax on shareholders.

Buybacks TTM Dividends TTM CapEx Intensity Shareholder Return
🏷️ Valuation ~18%

Is it cheap or expensive relative to its own history? We compare current multiples against 5-year averages using five complementary metrics — not just P/E.

P/FCF vs 5Y Avg EV/EBITDA Forward P/E P/S · P/B · PEG
🔬 Earnings Quality ~18%

Can we trust the numbers? We apply the Piotroski F-Score (9-point financial health checklist), the Beneish M-Score (8-factor earnings manipulation detector), and measure accruals, ROIC, ROE, and cash-to-earnings conversion.

Piotroski F-Score Beneish M-Score Accruals Ratio ROIC · ROE Earnings Quality %
⚠️ Risk Assessment ~9%

What could go wrong? We measure leverage, interest coverage, short-term debt risk, and SBC dilution — using the Altman Z-Score framework to flag distress signals. For Banks: Tier 1 Capital ratio and NPL Ratio replace debt metrics — leverage is structural, not a risk signal. For REITs: LTV, Debt/EBITDA vs REIT peers, and Interest Coverage.

Net Debt/EBITDA Interest Coverage Short-term Debt Risk SBC/FCF Tier 1 · NPL (Banks) LTV · Coverage (REITs)
🔴 Altman Z-Score — Bankruptcy Prediction
Z > 2.99
Safe Zone
1.81 – 2.99
Grey Zone
Z < 1.81
Distress Risk
Altman (1968) demonstrated 72–80% accuracy in predicting bankruptcy 1 year ahead, 69% accuracy 2 years ahead. Used globally in credit analysis and institutional risk management.
🚀 Momentum ~10%

Is the market moving with or against this stock? We apply the academic 12M-1M price momentum method — excluding the last month to remove short-term noise — combined with EPS surprise trend over the last 4 quarters.

12M-1M Return 6M-1M Return EPS Surprise Trend

* Weights vary by sector. Utilities: Risk increases to 20%, Business to 20%. Healthcare: Earnings Quality increases to 25%. Energy: Valuation increases to 25%. Consumer Discretionary: Momentum increases to 13%. 7 sector profiles total.
** Banks, Insurance, Asset Management, REITs: fully separate metric sets with z-score normalization vs sector peer benchmarks (AQR QMJ methodology). Altman Z-Score and standard debt metrics are not applied to financial sector companies.

Sector Methodology

Banks, Insurance, and REITs scored differently. By design.

Standard financial metrics break down for companies whose core business is finance. A bank with Debt/Equity of 10x is not overleveraged — that is a savings bank. A REIT with near-zero FCF Margin is not failing — it distributes all cash as dividends by law. Applying the same scoring model to JPMorgan as to Apple produces a misleading result.

For these sectors, the platform applies fully separate metric sets, each z-score normalized against sector peer benchmarks — the methodology used by AQR in their Quality Minus Junk (QMJ) factor model (Asness et al., 2019, Journal of Finance). A company is not judged against absolute thresholds, but against its own sector median: z = (value − sector median) / sector std.

z > +1.0
Excellent
Top 16% of sector
z > +0.3
Good
Above median
−0.3 to +0.3
Average
In line with sector
z < −0.3
Below Avg
Below median
z < −1.0
Weak
Bottom 16%
🏦 Banks & Financial Services

Bank profitability is driven by the spread between lending and funding costs, not by FCF generation. Capital adequacy and loan quality are the primary risk indicators — not Debt/Equity. Benchmarks derived from S&P 500 bank constituents and Damodaran NYU sector data.

NIM — Net Interest Margin ROTCE — Return on Tangible Common Equity Tier 1 Capital Ratio (CET1) NPL Ratio — Non-Performing Loans Efficiency Ratio Return on Assets P/TBV — Price / Tangible Book Value
→ Based on: Berger & Bouwman (2013) · Damodaran sector data · Basel III CET1 framework
🏢 REITs — Real Estate Investment Trusts

Net Income is structurally understated for REITs due to large depreciation charges on real assets that don't actually lose value. FFO (Funds From Operations) restores true economic earnings. REITs are legally required to distribute ≥90% of taxable income — making FCF Margin near zero by design, not weakness.

FFO — Funds From Operations AFFO — Adjusted FFO P/FFO — Primary valuation multiple FFO Yield AFFO Payout Ratio LTV — Loan-to-Value Occupancy Rate Interest Coverage
→ Based on: Ling & Naranjo (2015) · NAREIT FFO definition · Gyourko & Keim (1992)
🛡️ Insurance

Insurance profitability is measured by underwriting quality (Combined Ratio) and investment returns — not by free cash flow or CapEx intensity. A Combined Ratio below 100% means the underwriting operation is profitable before investment income.

Combined Ratio Return on Equity Net Margin Revenue Growth 5Y Dividend Yield P/Book
→ Based on: NAIC insurance industry benchmarks · Damodaran sector data
📊 Asset Management & Capital Markets

Asset managers generate high-margin, asset-light revenue from fees on AUM — making FCF Margin and ROE the primary quality signals. Standard debt risk metrics apply only minimally; business model risk is revenue concentration and AUM stability.

Return on Equity FCF Margin Net Margin Revenue Growth 5Y Dividend Yield P/E · P/Book vs peers
→ Based on: Damodaran sector data · S&P Capital Markets constituents
DCF Valuation Suite

Beyond scoring — calculate intrinsic value.

A score tells you whether a company is financially strong. A valuation tells you whether the price is right. The DCF Valuation Suite applies the same academic methodology used by professional analysts — with one critical improvement: instead of a single estimate, we give you a probability distribution.

🧮 WACC Calculator
Calculates the Weighted Average Cost of Capital using the CAPM framework, relevered beta, country risk premium, and size premium table. Auto-populated from live market data when you enter a ticker. D/E ratio is blended (70% company actual + 30% Damodaran sector median) to prevent leverage outliers from distorting cost of equity. Sector-specific ERP adjustments applied for 20 sectors where risk systematically diverges from market average (e.g. utilities, biotech, semiconductors).
→ Modigliani & Miller (1958) · Damodaran Jan 2026
📊 3-Scenario DCF
Bull, Base, and Bear scenarios with blended growth rates: 40% historical CAGR + 45% analyst forward consensus + 15% nominal GDP anchor — the same blending methodology widely used by major investment banks for DCF input construction. Terminal growth rate is anchored to the risk-free rate: tg = min(RFR × 0.65, 3.5%), consistent with Damodaran's standard. Supports both Gordon Growth (perpetuity) and EV/EBITDA Exit Multiple terminal value methods.
→ Graham & Dodd (1934) · Damodaran · Institutional investment bank DCF practice
🎲 Monte Carlo Simulation
Runs 10,000 simulations with Student-t fat tails and Cholesky correlations varying growth rates and WACC within realistic ranges. Returns a probability distribution of intrinsic value — not just one number. Shows the 10th, 50th, and 90th percentile outcomes. The Monte Carlo P50 is blended into the final weighted fair value (85% scenario-weighted + 15% MC P50) for added robustness — consistent with institutional simulation practice.
→ Metropolis & Ulam (1949) · Institutional simulation methodology
🔄 Reverse DCF
Works backwards from the current stock price. Instead of asking "what is it worth?", it asks: "what growth rate does today's price already assume?" If the implied growth is unrealistic, the stock may be overvalued. The Reverse DCF result also automatically adjusts scenario probabilities: if the market-implied CAGR significantly exceeds analyst consensus, the bull case probability is reduced accordingly — a valuation stress overlay consistent with institutional risk frameworks.
→ Mauboussin (2001) · Institutional risk framework methodology
📋 Sensitivity Analysis
Full sensitivity table showing how Fair Value changes across combinations of growth rates and WACC assumptions. Understand the range of outcomes visually.
→ Standard practitioner methodology
📰 Earnings Update Mode
Enter actual quarterly results vs analyst estimates to instantly recalculate fair value after earnings. Built for active investors who track companies through each reporting cycle.
→ Damodaran earnings revision methodology

Most DCF tools give you one number. One number is always wrong — the future is uncertain. Our Monte Carlo simulation gives you a range and a probability, so you know not just what the stock might be worth, but how confident you should be in that estimate. Open DCF Suite →

Portfolio Risk Engine

Institutional-grade risk metrics, built for every investor.

The same core risk methodologies used by institutional asset managers — portfolio VaR, factor exposure analysis, stress testing, and correlation structure — are the foundation of professional risk management. The Compound Family Portfolio Risk Engine applies these academically grounded techniques directly to your own positions.

📐 Concentration Risk — HHI & Effective N

The Herfindahl-Hirschman Index measures portfolio concentration as the sum of squared position weights × 10,000. A score below 1,000 is well-diversified; above 2,500 signals dangerous concentration. Effective N = 10,000 / HHI — the true number of independent positions. A portfolio with 15 holdings where one is 60% of value has an Effective N of ~2.8, not 15. This is the single most honest measure of diversification.

HHI (0–10,000) Effective N Max single-stock weight Top-3 concentration
→ Woerheide & Persson (1993), Financial Services Review
📉 Historical VaR & CVaR — Expected Shortfall

Value at Risk (VaR) is the maximum expected 1-day portfolio loss at a given confidence level — 95% or 99%. Calculated using historical simulation over 252 trading days: actual daily returns are weighted by current portfolio weights and the 5th/1st percentile of that distribution defines VaR. CVaR (Conditional VaR / Expected Shortfall) is the average loss in scenarios worse than VaR — the answer to "when things go wrong, how wrong do they get?" CVaR is a coherent risk measure per Artzner et al. (1999) and is the standard required under Basel III for internal risk models.

VaR 95% (daily %) VaR 99% (daily %) CVaR 95% — Expected Shortfall CVaR 99% Annualised volatility σ√252
→ Rockafellar & Uryasev (2000), Journal of Risk · Artzner et al. (1999), Mathematical Finance · Basel III (BCBS 2019)
🔗 Diversification Ratio & Correlation Matrix

The Diversification Ratio (DR) = Σ(wᵢ × σᵢ) / σ_portfolio. It measures how much diversification benefit the portfolio actually receives. DR = 1 means all positions move together — holding more of them adds no protection. DR = √N (number of positions) means they are completely uncorrelated — maximum diversification. Most retail portfolios have a DR far closer to 1 than √N, because tech-heavy portfolios hold highly correlated positions. The pairwise correlation matrix shows which pairs of positions provide redundant exposure and which provide genuine diversification.

Diversification Ratio Max theoretical DR (√N) Average pairwise correlation Top correlated pairs
→ Choueifaty & Coignard (2008), Journal of Portfolio Management
⚡ Stress Testing — Historical Scenario Analysis

Four pre-built stress scenarios apply historical sector-level drawdowns to current portfolio weights to estimate P&L impact. Each scenario uses peak-to-trough drawdowns calibrated to SPDR sector ETF historical data and Damodaran sector betas. The result is a dollar-and-percent estimate of portfolio loss under each scenario, broken down by position — so you can see exactly which holdings are your primary risk concentration in each crisis type.

2008 Financial Crisis (peak-to-trough by sector) 2020 COVID Crash (Feb–Mar, 33 days) Rate Shock +200bps (2022-style) Tech Selloff −30% (dot-com style)
→ Damodaran sector betas (Jan 2026) · SPDR sector ETF peak-to-trough historical data
📊 Portfolio Beta — Market Exposure Proxy

Portfolio beta = Σ(wᵢ × βᵢ) — the weighted average of position betas, using Damodaran's January 2026 unlevered sector betas relevered to each sector's median D/E ratio. A beta below 0.7 indicates a defensive portfolio; above 1.15 is aggressive. This tells you how much the portfolio is expected to move relative to the S&P 500 — crucial for understanding market exposure before a drawdown, not after.

Portfolio β (weighted) Damodaran sector betas (Jan 2026) Defensive / Conservative / Market-like / Aggressive
→ Damodaran (2026), NYU Stern — betas.xls · Fama & French (2015) factor framework
📐 VaR Method: Historical Simulation (non-parametric)
252 days
Price history window (1 trading year)
5th pct
VaR 95% — worst day in 20
1st pct
VaR 99% — worst day in 100
Historical simulation makes no assumption about the return distribution — it uses actual past returns, preserving fat tails and skewness that parametric (Gaussian) VaR ignores. This is the method preferred by Basel III for internal risk model validation. CVaR (Expected Shortfall) addresses VaR's key weakness: it measures the average loss beyond the threshold, not just the threshold itself — satisfying the coherence axioms of Artzner et al. (1999).

* VaR, CVaR, Diversification Ratio, and correlation require 252-day price history (KV-cached 24h — zero additional API cost after first load per ticker per day). HHI, Effective N, Beta, and Stress Tests are computed instantly from existing portfolio weights. All metrics are for educational purposes only and do not constitute investment advice.

Stock Screener

Find alpha with academically grounded factor analysis.

The TCF Screener goes beyond typical financial metrics. In addition to fundamentals like P/E, ROIC, and margins, it exposes four academically proven composite Alpha Scores — each combining multiple factors into a single actionable signal. Every score is calculated from the same underlying data used in the Stock Analysis Terminal, cached daily.

⚡
Momentum Alpha Score
12-1 Price Momentum + Earnings Surprises + EPS Growth
0–100

Combines three academically distinct momentum signals into one composite score. The 12-1 price momentum (12-month return excluding last month) is the strongest single return predictor documented in academic literature, outperforming all other factors on a standalone basis across 40+ markets.

Earnings momentum — standardised unexpected earnings (SUE) — exploits Post-Earnings Announcement Drift (PEAD): stocks with positive surprises continue to outperform for 3–6 months after the announcement. EPS growth YoY anchors the signal to fundamental momentum, reducing false positives from purely price-based signals.

Jegadeesh & Titman (1993) · Bernard & Thomas (1989) · Asness, Frazzini & Pedersen (2013)
🛡
Defensive Alpha Score
Quality + Low Beta (BAB) + Low Volatility + Interest Coverage
0–100

The most counterintuitive finding in modern finance: low-risk stocks outperform high-risk stocks on a risk-adjusted basis. The Betting Against Beta (BAB) factor demonstrates that low-beta portfolios deliver superior Sharpe ratios due to leverage constraints among institutional investors who cannot lever low-beta assets.

Combined with our Quality Score (ROIC, earnings quality, accruals) and interest coverage, this score identifies companies with fortress balance sheets that hold up in market downturns while participating in upside. The MSCI Quality+MinVol combination has historically delivered the best risk-adjusted returns of any two-factor combination in developed markets.

Frazzini & Pedersen (2014) · Baker & Haugen (1991) · MSCI Factor Research
📈
Value + Momentum Score
Valuation Score + 12-1 Price Momentum + FCF Yield
0–100

Value and momentum are negatively correlated by construction — cheap stocks tend to have poor momentum, and momentum stocks tend to be expensive. This makes their combination particularly powerful for diversifying alpha sources. Asness et al. (2013) showed that the combined Value+Momentum strategy delivers a Sharpe ratio of 0.9+ historically — higher than either factor alone.

FCF yield anchors value in real cash terms rather than accounting earnings, making it robust to earnings manipulation and accruals distortions. A stock scoring high on all three signals — undervalued by multiples, high FCF yield, and positive price momentum — represents the classic "durable alpha" setup.

Fama & French (1992, 2015) · Asness, Moskowitz & Pedersen (2013) · Damodaran (FCF Yield)
🏗
Compounder Score
FCF Yield + Earnings Quality + Revenue CAGR 5Y + Gross Margin Trend
0–100

Compounders are businesses that generate durable, growing, real free cash flow — and reinvest it at high returns. This score identifies them through four lenses: FCF yield (real cash return to shareholders), earnings quality via Sloan accruals (OCF/Net Income ratio — cash-backed earnings outperform accruals-heavy earnings), 5-year revenue CAGR (durability of growth), and gross margin trend (widening moat).

Novy-Marx (2013) demonstrated that gross profitability — gross profit divided by assets — predicts returns as powerfully as book-to-market value, but with opposite sign: profitable companies outperform unprofitable ones regardless of valuation. The gross margin trend captures whether a company's competitive advantage is widening or eroding.

Novy-Marx (2013) · Sloan (1996) · Piotroski (2000)
Important: Alpha Scores are composite signals, not buy/sell recommendations. They identify statistical patterns that have historically predicted excess returns. Past factor performance does not guarantee future results. Always combine with your own fundamental analysis and risk assessment. Open Screener →
Academic foundation

19 papers. Decades of research. All applied here.

Every factor, every formula, every scoring decision is grounded in peer-reviewed academic research. Unlike platforms that hide behind proprietary black boxes, we cite the original paper for everything we do. You always know exactly why a stock scored the way it did.

Piotroski (2000)
Journal of Accounting Research
Earnings Quality
9-point financial health scoring system based on profitability, leverage, and operating efficiency signals. Demonstrated that high F-Score stocks outperformed low F-Score stocks by 23% annually in the original study. One of the most cited papers in quantitative equity investing.
→ Used in: Earnings Quality factor (F-Score, 9 signals) · WACC quality adjustment (F-Score 8–9: −0.25% WACC; F-Score ≤3: +0.40% WACC)
Beneish (1999)
Financial Analysts Journal
Earnings Quality
The Beneish M-Score is an 8-factor statistical model for detecting earnings manipulation. It combines financial ratios measuring changes in receivables, gross margins, asset quality, sales growth, depreciation, SG&A, leverage, and accruals into a single score. An M-Score above −1.78 indicates a likely earnings manipulator. Used by forensic accountants, short sellers, and institutional risk teams worldwide. The model correctly identified Enron and WorldCom as manipulators before their collapse.
→ Used in: Earnings Quality factor (Beneish M-Score, 10 pts · threshold −1.78)
Altman (1968)
Journal of Finance
Risk Assessment
The Altman Z-Score — a 5-variable linear discriminant model predicting bankruptcy probability from financial ratios. Demonstrated 72–80% accuracy in predicting bankruptcy 1 year ahead and 69% accuracy 2 years ahead. Z > 2.99 = safe zone; 1.81–2.99 = grey zone; < 1.81 = distress risk. Used globally in credit analysis and institutional risk management for over 50 years.
→ Used in: Risk Assessment factor (bankruptcy prediction)
Jegadeesh & Titman (1993)
Journal of Finance
Momentum
Foundational momentum research demonstrating that stocks with strong 6–12 month price performance continue to outperform over the next 3–12 months. Excluding the most recent month reduces mean-reversion noise — a refinement that improves momentum signal quality.
→ Used in: Momentum factor (12M-1M, 6M-1M price returns)
Sloan (1996)
Accounting Review
Earnings Quality
Demonstrates that earnings with high accrual components predict lower future returns. Cash-backed earnings are more reliable than accounting-based earnings. The accruals ratio is now a standard measure of earnings quality in institutional analysis.
→ Used in: Earnings Quality factor (Accruals Ratio)
Fama & French (1992, 2015)
Journal of Finance
Sector Weights
Multi-factor asset pricing models establishing the role of size, value, profitability, and investment factors in explaining stock returns. Basis for our sector-adjusted weighting system — different factors matter differently across industries.
→ Used in: All 7 sector-adjusted weight profiles
Damodaran
NYU Stern — ongoing
Valuation · DCF
Aswath Damodaran's publicly available WACC methodology, FCF yield framework, CapEx intensity analysis, country risk premium tables, Revenue Exit Multiple approach, sector median D/E ratios (Jan 2026), and sector-specific ERP adjustments for 20 industries. The global standard reference for practitioner valuation — used by investment banks, PE funds, and asset managers worldwide.
→ Used in: DCF Suite (WACC + sector D/E blend + sector ERP), Valuation factor, Capital Allocation
Modigliani & Miller (1958)
American Economic Review
DCF · WACC
Foundational capital structure theory — the theoretical basis for WACC calculation. The cost of capital depends on the blend of debt and equity financing and their respective risk profiles. Nobel Prize in Economics (Miller, 1990).
→ Used in: DCF Suite (WACC calculation)
Graham & Dodd (1934)
Security Analysis
Valuation · DCF
The original framework for fundamental analysis and value investing. Discounted cash flow valuation, margin of safety, and earnings normalisation all originate from this work. The foundation of modern security analysis — still directly applicable 90 years later.
→ Used in: DCF Suite, Valuation factor, Business Quality
Metropolis & Ulam (1949)
Journal of the American Statistical Association
Monte Carlo
Original Monte Carlo simulation methodology — developed for nuclear physics, now applied across finance, engineering, and risk management. Used in our DCF model to generate probability distributions for intrinsic value across thousands of scenarios instead of a single point estimate.
→ Used in: DCF Suite (Monte Carlo simulation)
Mauboussin (2001)
Credit Suisse First Boston
Reverse DCF
Reverse DCF methodology — inferring the growth rate the current market price implies, rather than projecting a growth rate to estimate value. Useful for understanding what the market expects and whether those expectations are realistic.
→ Used in: DCF Suite (Reverse DCF)
Institutional Investment Bank DCF Practice
Practitioner methodology
DCF · Scenarios
Institutional DCF input construction: blended growth rates combining historical CAGR (40%), analyst forward consensus (45%), and nominal GDP anchor (15%). Terminal growth anchored to the risk-free rate: tg = min(RFR × 0.65, 3.5%). Standard methodology for equity research departments at major investment banks.
→ Used in: 3-Scenario DCF auto-population (blended growth + terminal growth via RFR)
Institutional Risk & Simulation Frameworks
Institutional risk framework
Monte Carlo · Scenario Weights
Institutional approach to simulation-based valuation and scenario probability adjustment. Monte Carlo P50 blended into weighted fair value (15% weight) for robustness. Reverse DCF valuation stress overlay: when market-implied growth significantly exceeds analyst consensus, bull case probability is automatically reduced.
→ Used in: Monte Carlo P50 blend in final fair value · Reverse DCF scenario weight adjustment
Asness, Moskowitz & Pedersen (2013)
Journal of Finance
Momentum + Quality
Demonstrates that value and momentum strategies are negatively correlated across asset classes and geographies, making their combination exceptionally powerful for diversifying alpha. The combined Value+Momentum portfolio delivers a Sharpe ratio of 0.9+ historically. Also establishes that quality factors (profitability, earnings stability) independently predict returns beyond the classic three-factor model.
→ Used in: Momentum Alpha Score · Value+Momentum Alpha Score · TCF Overall Score weighting
Frazzini & Pedersen (2014)
Journal of Financial Economics
Defensive Alpha · BAB Factor
Introduces the Betting Against Beta (BAB) factor — the empirical finding that low-beta assets deliver higher risk-adjusted returns than high-beta assets across 20 countries and multiple asset classes. The mechanism: leverage-constrained investors (mutual funds, pension funds) bid up high-beta assets, creating a persistent mispricing that low-beta assets exploit. The BAB factor has annualised returns of 7–12% historically with a Sharpe ratio exceeding the market.
→ Used in: Defensive Alpha Score (Low Beta component, 25% weight)
Novy-Marx (2013)
Journal of Financial Economics
Gross Profitability · Compounder
Demonstrates that gross profitability (Gross Profit/Assets) predicts cross-sectional returns as powerfully as book-to-market value, but with opposite sign — profitable companies outperform regardless of valuation multiples. This finding challenges the traditional value-investing intuition: quality at any price outperforms low quality at a discount. The gross profitability factor has survived out-of-sample testing across multiple decades and markets, and is now incorporated into AQR's and MSCI's Quality factor definitions.
→ Used in: Compounder Alpha Score (Gross Margin Trend component, 20% weight)
Baker & Haugen (1991)
Journal of Finance
Low Volatility Anomaly
One of the earliest and most persistent anomalies in finance: low-volatility stocks deliver higher returns than high-volatility stocks — the opposite of what CAPM predicts. Replicated in 33 countries over 50+ years of data by subsequent researchers. The anomaly is explained by behavioral bias (investors prefer "lottery ticket" stocks), agency problems (fund managers benchmark-hug high-beta names), and leverage constraints. The MSCI Minimum Volatility index has outperformed its parent index on a risk-adjusted basis since inception.
→ Used in: Defensive Alpha Score (Low Volatility component via 3M return std, 25% weight)
Asness, Frazzini & Pedersen (2019)
Journal of Finance
Sector Normalization · QMJ
Quality Minus Junk (QMJ) — the definitive academic framework for measuring company quality across profitability, growth, safety, and payout dimensions. Critically, QMJ normalizes all factors within sector peer groups using z-scores rather than absolute thresholds. This approach is now the standard in institutional factor investing and is the methodological foundation for the platform's financial sector scoring. The paper demonstrates that quality factors predict returns across 24 countries and 30+ years of data.
→ Used in: Sector-specific scoring for Banks, Insurance, REITs, Asset Management (z-score normalization vs peer benchmarks)
Ling & Naranjo (2015)
Real Estate Economics
REIT Methodology · FFO
Demonstrates that Funds From Operations (FFO) is a significantly stronger predictor of REIT returns and dividend sustainability than GAAP Net Income. Because depreciation is a non-cash charge on assets that often appreciate in value, Net Income systematically understates REIT earning power. FFO and AFFO (Adjusted FFO) are now the standard valuation metrics used by institutional REIT analysts and the NAREIT industry body worldwide.
→ Used in: REIT sector scoring (FFO, AFFO, P/FFO, AFFO Payout Ratio as primary metrics)
Berger & Bouwman (2013)
Review of Financial Studies
Bank Risk · Capital Adequacy
Establishes that bank capital (Tier 1 / CET1 ratio) is the primary predictor of bank survival and performance across all phases of the economic cycle. Banks with higher capital ratios are significantly more likely to survive financial crises and continue lending. This finding — replicated in the 2008 financial crisis — is the basis for the Basel III regulatory framework and confirms that Tier 1 Capital Ratio, not Debt/Equity, is the correct risk metric for bank analysis.
→ Used in: Bank sector Risk Assessment (Tier 1 Capital as primary risk factor, NPL Ratio as secondary)
Rockafellar & Uryasev (2000)
Journal of Risk
Portfolio Risk · CVaR
Introduces the formal optimization framework for Conditional Value at Risk (CVaR), also known as Expected Shortfall. CVaR is the average loss in scenarios beyond the VaR threshold — answering "when things go wrong, how wrong do they get?" Unlike VaR, CVaR is a coherent risk measure satisfying subadditivity, monotonicity, translation invariance, and positive homogeneity (Artzner et al., 1999). CVaR became the required risk measure under Basel III for internal risk model validation, replacing VaR as the primary metric in 2019 (BCBS 2019). The paper provides a computationally efficient linear programming approach to CVaR minimization.
→ Used in: Portfolio Risk Engine (VaR 95%/99%, CVaR 95%/99%, historical simulation over 252 trading days)
Choueifaty & Coignard (2008)
Journal of Portfolio Management
Portfolio Risk · Diversification
Introduces the Diversification Ratio (DR) as a formal measure of portfolio diversification: DR = Σ(wᵢσᵢ) / σ_portfolio. DR = 1 means assets are perfectly correlated and no diversification benefit is received. DR = √N (for N positions) is the theoretical maximum when all assets are uncorrelated. The paper proves that the Most Diversified Portfolio (MDP) — which maximises DR — is a superior portfolio construction approach compared to equal weighting, minimum variance, and market cap weighting. The methodology is now used by Tobam Asset Management and referenced by institutional factor investors worldwide.
→ Used in: Portfolio Risk Engine (Diversification Ratio, pairwise correlation matrix, max theoretical DR)
Woerheide & Persson (1993)
Financial Services Review
Portfolio Risk · Concentration
Applies the Herfindahl-Hirschman Index (HHI) — originally an antitrust measure of industry concentration — to portfolio diversification measurement. HHI = Σ(wᵢ²) × 10,000, where wᵢ are portfolio weights. Introduces the Effective N = 1/Σ(wᵢ²) as an intuitive measure of the "true" number of independent positions: a portfolio with 15 holdings where one position is 60% of value has the same HHI and Effective N as a 3-stock equally-weighted portfolio. Demonstrates that portfolio count is a misleading diversification proxy — HHI and Effective N are the correct measures.
→ Used in: Portfolio Risk Engine (HHI score, Effective N, concentration warnings, 7-metric strip)
Data sources

Official filings. Not estimates.

Our primary data source is SEC EDGAR — the official US Securities and Exchange Commission database of company filings. Every number you see is sourced from official 10-K and 10-Q reports, not third-party estimates or analyst projections. This is the same data that institutional analysts work from.

🏛️
SEC EDGAR — Primary source
Official 10-K annual reports and 10-Q quarterly reports filed directly with the US Securities and Exchange Commission. XBRL-structured financial data going back 10+ years per company. Completely free, public, and government-maintained. This is the ground truth — the same filings auditors and institutional investors use.
How we compare

More depth. Open methodology. Less cost.

Most platforms show you data or hide their scoring in a proprietary black box. We score every metric using published academic methods, explain every formula on this page, and give you professional-grade tools with no account required to start.

Feature The Compound Family Typical Screeners
Free–$10/mo		 Premium Platforms
$20–35/mo
Academic scoring model
19 peer-reviewed papers		 ✓ Fully transparent ✗ Raw data only ~ Proprietary / hidden
Altman Z-Score bankruptcy prediction
72–80% accuracy 1 year ahead		 ✓ Full Z-Score ✗ ~ Limited
Sector-specific methodology
Banks · Insurance · REITs · Asset Mgmt		 ✓ Full sector scoring + z-score normalization ✗ ~ Limited
10+ years historical data
From official SEC filings		 ✓ SEC EDGAR ~ Varies ✓
Full DCF + Monte Carlo
Thousands of scenarios		 ✓ Included ✗ ~ Basic / paid
Reverse DCF
What growth is priced in?		 ✓ Included ✗ ✗
Momentum factor
12M-1M academic method		 ✓ Academic ✗ ~ Limited
No account required
Full scores visible instantly		 ✓ Always ~ Often limited ✗ Login required
AI interpretation
Plain-English explanation		 ✓ 1/day included · increased with Premium ✗ ~ Some platforms
Methodology fully disclosed
Every formula explained		 ✓ This page ✗ ✗ Black box

Unlike platforms that hide their scoring logic, every formula we use is documented here — referenced to the original academic paper. You always know exactly why a stock scored the way it did. See full pricing comparison →

Important notice

A model is a tool. Not a verdict.

⚠ Model limitations & continuous improvement. The Compound Family scoring model is continuously developed and improved. Like any quantitative model, scores may be affected by data quality issues, sector-specific accounting differences, extraordinary items, companies in transition, or factors not captured by the underlying metrics. The model does not account for qualitative factors such as management quality, competitive moat, geopolitical risk, or pending litigation.

Scores should never be the sole basis for any investment decision. Always conduct your own comprehensive research, review primary sources including official company filings, and consider consulting a qualified financial advisor before making any investment. Past model performance does not guarantee future accuracy. Full disclaimer →
Common questions

Frequently asked questions

The Piotroski F-Score is a 9-point financial health checklist developed by Joseph Piotroski in 2000 and published in the Journal of Accounting Research. It evaluates a company across three categories: profitability (4 signals), leverage and liquidity (3 signals), and operating efficiency (2 signals). A score of 8–9 indicates a financially strong company; 0–2 signals potential distress. The original study showed high F-Score stocks outperformed low F-Score stocks by 23% annually. We apply this directly in the Earnings Quality scoring factor.
Altman (1968) demonstrated 72–80% accuracy in predicting bankruptcy 1 year ahead, and 69% accuracy 2 years ahead in the original study. The model uses 5 financial ratios (working capital, retained earnings, EBIT, market cap, and sales — all scaled by total assets) combined into a single Z-Score. A Z-Score above 2.99 indicates a safe zone; between 1.81 and 2.99 is the grey zone; below 1.81 signals high distress risk. The model has been used in credit analysis and institutional risk management for over 50 years.
We use the academic 12M-1M momentum method from Jegadeesh and Titman (1993): the 12-month price return excluding the most recent month. Removing the last month reduces short-term reversal noise that is well-documented in academic literature. We also calculate 6M-1M return and combine both with an EPS surprise trend measured over the last 4 quarters.
DCF stands for Discounted Cash Flow — a method of valuing a company based on its projected future free cash flows, discounted back to today using the WACC. A standard DCF gives a single estimate — which is always wrong because the future is uncertain. Our Monte Carlo model runs 10,000 simulations with Student-t fat tails and Cholesky correlations, varying growth rates and WACC assumptions within realistic ranges, to produce a full probability distribution of intrinsic value. The Monte Carlo P50 (median outcome) is also blended into the final weighted fair value — 15% Monte Carlo P50 + 85% scenario-weighted — for added robustness, consistent with institutional simulation practice. This tells you not just what a stock might be worth, but how confident you should be in that estimate — and what the downside scenario looks like.
Reverse DCF works backwards from the current stock price. Instead of asking "what is this company worth?", it asks: "what growth rate does today's price already assume?" Based on Mauboussin (2001), this is one of the most powerful valuation tools for growth stocks — it tells you what the market is pricing in, and whether those expectations are realistic given the company's track record. If the implied growth rate is higher than the company has ever achieved, the stock may be overvalued.
Different sectors operate under fundamentally different business models. A utility company is valued on stability, regulated income, and dividend yield — so balance sheet risk matters more (weight 20%). A software company is valued on growth and margin expansion — so business quality matters more (weight 27%). Applying the same fixed weights to all sectors produces systematically misleading scores. We apply 7 sector profiles so each stock is judged by criteria that actually matter for its industry.

For Banks, Insurance, and REITs — we go further. Standard metrics like FCF Margin or Debt/EBITDA are structurally wrong for these sectors: banks trade in money by design, and REITs carry structural leverage. These sectors receive fully separate metric sets (NIM, ROTCE, Tier 1 Capital for banks; FFO, AFFO, P/FFO, LTV for REITs), each z-score normalized vs sector peer benchmarks following the AQR QMJ methodology (Asness, Frazzini & Pedersen, 2019).
The core platform requires no account. All 6 scoring factors, all metric cards with tooltips, all historical charts, Piotroski F-Score detail, Altman Z-Score, the full DCF Suite (Monte Carlo, Reverse DCF, Sensitivity, Earnings Update Mode), and Save Model are included. Users also get 1 AI analysis per day after signing in with Google. Premium ($19/month or $149/year) unlocks Fair Value shown directly on the stock page and increased AI analysis. See the full pricing page →
VaR (Value at Risk) at 95% means: on 95% of trading days, your portfolio will not lose more than this amount. Calculated using historical simulation — your actual weighted daily returns over the past 252 trading days, with the 5th percentile defining VaR 95% and the 1st percentile defining VaR 99%.

CVaR (Conditional VaR, also called Expected Shortfall) answers the follow-up question: when things are bad enough to breach VaR, how bad is the average outcome? CVaR is the mean of all returns worse than VaR — a more complete picture of tail risk. It is the risk measure required under Basel III for internal risk model validation, because unlike VaR, it satisfies the mathematical coherence axioms: subadditivity, monotonicity, translation invariance, and positive homogeneity (Artzner et al., 1999; Rockafellar & Uryasev, 2000).
The Herfindahl-Hirschman Index (HHI) = Σ(wᵢ²) × 10,000, where wᵢ is each position's weight. A portfolio of 20 equal positions has HHI = 500 (well diversified). A portfolio with one position at 60% and 14 small positions has HHI ≈ 3,640 — highly concentrated, despite 15 holdings on paper.

Effective N = 10,000 / HHI gives the number of equal-weight positions that would produce the same concentration. In the example above, Effective N ≈ 2.7 — you effectively have fewer than 3 independent positions despite 15. This is why position count is a misleading diversification metric: what matters is the weight distribution, not the count. Thresholds: HHI below 1,000 = well diversified; 1,000–2,500 = moderate; above 2,500 = concentrated. Based on Woerheide & Persson (1993).
No. All scores, analyses, and AI interpretations are provided for informational and educational purposes only. Nothing on thecompoundfamily.com constitutes investment advice or a recommendation to buy or sell any security. The platform is a research and education tool — it helps you understand a company's financial profile. All investment decisions are solely your responsibility. See our full disclaimer.