How to use it

Begin with the environment, then move toward the company

The Fieldbook is most useful when read as a sequence rather than as a collection of isolated pages.

The research structure

Four layers operating at different speeds

No single layer controls the conclusion. Their value comes from reading them together.

Agreement across the layers can strengthen confidence. Disagreement can be even more useful because it reveals where deeper investigation is needed.

Daily Market methodology

How the Daily Market scores are calculated

The Daily Market layer is a short-horizon reading of the market itself. It does not forecast the next trading session. It measures whether current market behavior appears broad, narrow, constructive, defensive, stressed, or internally conflicted.

The process evaluates evidence across several connected categories:

• price trend and broad equity performance;
• market breadth and participation;
• growth, cyclical, and defensive leadership;
• small-cap and equal-weight participation;
• credit behavior and duration demand;
• volatility and stress;
• global confirmation;
• confirmation and divergence across signals.

Returns

Returns are calculated across multiple periods, including one day, five days, one month, three months, six months, year to date, and one year.

Longer-period returns are compounded from the underlying daily observations rather than simply added together.

Relative leadership

Relative-strength relationships help identify which parts of the market are leading or lagging.

Common comparisons include:

• Nasdaq 100 versus the S&P 500 for growth and mega-cap leadership;
• equal-weight S&P 500 versus capitalization-weighted S&P 500 for breadth;
• small caps versus the S&P 500 for participation;
• high-yield credit versus Treasuries for risk appetite;
• cyclical sectors versus defensive sectors;
• momentum, quality, and value proxies versus the broad market.

Breadth

Breadth measures how widely a market move is being shared. A rising index supported by many advancing assets is generally more internally convincing than a rising index driven by only a few large companies.

Breadth is measured across the tracked ETF and market-proxy universe for each lookback period.

Risk-On score

The Risk-On score is a 0–100 composite intended to summarize the market’s appetite for equity and credit risk.

It considers evidence such as:

• the S&P 500 trend;
• growth leadership;
• equal-weight participation;
• small-cap participation;
• high-yield credit behavior;
• cyclical versus defensive leadership;
• volatility pressure;
• overall breadth;
• current drawdown and trend condition.

A higher score indicates broader and more constructive risk-taking. A lower score indicates defensive behavior, weaker participation, deteriorating credit, elevated volatility, or broader market stress.

Internal Health score

The Internal Health score evaluates the quality of the market move beneath the headline index.

It generally rewards:

• broad positive participation;
• equal-weight and small-cap confirmation;
• constructive credit behavior;
• broad sector leadership;
• price above longer-term trend;
• limited defensive or volatility stress.

It generally penalizes:

• narrow mega-cap leadership;
• weak equal-weight participation;
• credit deterioration;
• defensive leadership during headline equity strength;
• large drawdowns;
• rising volatility and stress.

Regime classification

The final Daily Market regime is based on the combination of the Risk-On score, Internal Health score, breadth, leadership, credit, volatility, and important divergences.

Possible classifications include:

• Broad Risk-On;
• Constructive Broadening;
• Narrow Mega-Cap Leadership;
• Correction Inside a Longer-Term Uptrend;
• Defensive Rotation;
• Momentum Unwind;
• Credit Watch;
• Risk-Off;
• Mixed or Transitional.

Weekly Economic methodology

How the Weekly Economic scores are calculated

The Weekly Economic layer organizes macroeconomic evidence into five broad pillars:

Each pillar is translated into a 0–100 reading. The score represents the current condition of that part of the economy, not a direct forecast of GDP, inflation, interest rates, or asset prices.

Growth score

The Growth score organizes evidence such as industrial production, retail activity, real income, real spending, business surveys, manufacturing indicators, and broader output measures.

Higher scores indicate a more expansionary and supportive level of economic activity. Lower scores indicate weaker activity, contraction risk, or a deteriorating growth environment.

Labor score

The Labor score considers unemployment, payroll growth, jobless claims, labor-force participation, wage growth, and related employment evidence.

Higher readings generally indicate a resilient labor market. Lower readings indicate weakening hiring, rising unemployment pressure, or increasing labor-market stress.

Inflation Pressure score

The Inflation Pressure score measures the degree of ongoing price pressure rather than treating high inflation as economically positive.

Inputs can include:

• consumer-price inflation;
• personal-consumption-expenditure inflation;
• core inflation measures;
• wage and input-cost pressure;
• commodity prices;
• inflation expectations.

Higher values indicate greater inflation pressure. Lower values indicate more subdued or disinflationary conditions.

Credit score

The Credit score evaluates whether financing conditions are supportive, restrictive, or showing signs of stress.

It can include high-yield and investment-grade spread behavior, lending standards, financial-stress indicators, default-sensitive evidence, and market-based credit confirmation.

Higher scores generally indicate loose or supportive credit conditions. Lower scores indicate widening spreads, tightening standards, or increasing financial stress.

Liquidity score

The Liquidity score examines the monetary and financial conditions affecting the availability and cost of capital.

Inputs may include:

• policy rates;
• the yield curve;
• money-supply and reserve measures;
• financial-conditions indexes;
• dollar and market-liquidity proxies.

Higher readings generally indicate more supportive liquidity. Lower readings indicate tighter or more restrictive financial conditions.

Level, direction, and transition pressure

The current economic methodology separates three different questions:

• Current level: Is the pillar presently supportive or restrictive?
• Momentum: Is the pillar improving, stable, or deteriorating?
• Transition pressure: Is the direction of change strong enough to threaten the current regime?

This distinction matters because a pillar may be deteriorating while still remaining supportive in level terms.

For example, credit conditions may weaken for several weekly observations but still remain historically loose. The report should communicate both facts rather than forcing a premature regime change.

Freshness adjustment

Economic indicators update at different speeds. Jobless claims may be current to the latest week, while GDP data may describe a period that ended months earlier.

The system therefore distinguishes:

• the raw pillar score;
• the freshness-adjusted score;
• the availability of required inputs;
• the defensibility of the final classification.

Stale or missing evidence reduces confidence even when the available data points toward a clear conclusion.

Confidence

Confidence reflects data availability, freshness, agreement across indicators, and whether independent evidence supports the same interpretation.

A high-confidence regime requires more than a strong average score. It also requires sufficiently current inputs and reasonable cross-confirmation.

Economic regime classification

The five pillars are read together to form the weekly economic regime.

Possible classifications include:

• Disinflationary Expansion;
• Inflationary Expansion;
• Late-Cycle Expansion;
• Slowdown;
• Stagflationary Pressure;
• Recessionary Pressure;
• Credit Stress;
• Liquidity Tightening;
• Mixed or Transitional.

Mixed or Transitional is reserved for meaningful conflict across pillars rather than a narrow threshold miss in one indicator.

Monthly Factor methodology

How the Monthly Factor scores are calculated

The Monthly Factor layer studies which broad investment characteristics, industries, styles, and regions have been rewarded or penalized.

It primarily uses academic portfolio and factor data, including the Ken French data library, rather than treating daily ETF proxies as exact academic factors.

Core factors

Short-term and long-term reversal factors are also tracked to identify whether recent leadership is unwinding or reversing.

Compounded factor returns

Factor returns are evaluated over one day, five days, one month, three months, six months, year to date, one year, three years, five years, and ten years where sufficient data exists.

Year-to-date return

Year-to-date performance includes all observations from the beginning of the latest factor-data calendar year through the latest available date.

Volatility

Volatility helps distinguish persistent leadership from a return produced through unusually unstable price behavior.

Drawdown

Maximum drawdown is the most negative observation in the trailing drawdown series. It shows the largest loss from a prior peak during the measurement period.

Historical percentile

The six-month percentile compares the current six-month factor return with that factor’s own historical rolling six-month return distribution.

A percentile of 90 means the current six-month result is stronger than approximately 90 percent of comparable historical observations.

Factor breadth

Breadth is calculated separately across several universes:

• the institutional factor universe;
• canonical core factors;
• industries;
• style portfolios;
• international and regional portfolios.

Broad positive participation is generally more convincing than leadership concentrated in only one or two factor series.

Outlier controls

Extremely volatile or implausible observations can distort leadership tables. The system applies reasonable-universe filters to exclude unusually extreme returns or volatility from certain summary rankings.

Typical review thresholds include:

• one-year return outside approximately ±80%;
• six-month return outside approximately ±60%;
• annualized one-year volatility near or above 80%.

These filters are intended to reduce noise, not erase valid market history.

Factor-regime scores

The Monthly Factor report converts leadership, breadth, persistence, volatility, drawdown, and cross-universe confirmation into several 0–100 heuristic scores.

Current regime dimensions include:

• Risk-On;
• Defensive Rotation;
• Value Leadership;
• Small-Cap Leadership;
• Momentum Durability;
• Real-Asset and Inflation-Sensitive Leadership.

These scores summarize the strength and breadth of each market theme. They are not expected-return estimates.

Signal confidence

The factor engine assigns broad confidence buckets to its regime scores:

Confidence should also be interpreted alongside breadth, persistence, data coverage, and disagreement among factor universes.

Factor regime classification

The final classification combines the regime scores with factor leadership, breadth, drawdowns, and important divergences.

Possible classifications include:

• Broad Risk-On;
• Defensive Value Rotation;
• Momentum Unwind or Reversal Tape;
• Correction Inside a Longer-Term Uptrend;
• Mixed or Transitional.

Factor leadership is kept separate from the broader daily market regime. For example, weak profitability-factor performance means lower-profitability companies have outperformed more profitable companies over that period. It does not automatically mean the entire market is Risk-Off.

Scoring methodology

How the Company Scorebook works

The Company Scorebook is a research-prioritization system. It converts a group of fundamental and market observations into four component scores:

Each component is scored on a 0–100 scale. The score is not a probability, a target return, or an estimate of intrinsic value. It is a standardized way to compare research evidence across a broad universe.

A high score means the company deserves attention. It does not mean every important question has been answered.

Component one

Quality score

The Quality score examines whether a company appears to have strong economics, attractive margins, durable growth, and disciplined treatment of shareholders.

The current production model uses:

• return on invested capital;
• operating margin;
• free-cash-flow margin;
• four-year revenue growth;
• four-year operating-income growth;
• four-year free-cash-flow growth;
• four-year share-count change.

High returns on capital and strong margins can indicate a valuable business model. Consistent revenue, operating-income, and free-cash-flow growth can indicate that those economics are expanding. Stable or declining share count can indicate that growth is not being achieved through persistent dilution.

Component two

Valuation score

The Valuation score asks how much investors are paying for the company’s current operating earnings and free cash flow.

The current production model uses:

• enterprise value divided by operating income;
• free-cash-flow yield.

Lower enterprise-value multiples generally receive stronger valuation scores. Higher free-cash-flow yields generally receive stronger valuation scores.

Valuation must still be interpreted alongside business quality. A low multiple may reflect genuine opportunity, cyclical pressure, deteriorating economics, a temporary market dislocation, or unreliable data.

Component three

Balance-sheet score

The Balance-sheet score evaluates the company’s debt burden relative to its operating earnings.

The principal production measure is:

Lower values generally indicate that the company has greater operating capacity relative to its debt. Higher values signal that leverage and financial structure deserve closer review.

This is deliberately a simple and conservative measure. It can over-penalize stable recurring-revenue businesses, asset-light companies, acquisition-heavy companies, and firms that intentionally use leverage as part of their capital structure.

A weak balance-sheet score is therefore a research prompt, not an automatic rejection.

Component four

Market-confirmation score

The Market-confirmation score examines whether the stock’s recent behavior supports, challenges, or complicates the fundamental setup.

The current production model uses:

• three-month price return;
• three-month relative strength versus the S&P 500;
• whether the price is above its 200-day moving average;
• one-year drawdown from the prior peak.

Strong market confirmation can indicate that investors are recognizing improving or durable evidence. Weak confirmation can indicate company-specific concerns, sector pressure, valuation compression, cyclical weakness, or a broader market regime.

Price behavior does not determine whether a company is good. It helps identify whether the market agrees with the current research case and whether additional questions need to be answered.

Combined output

Overall research score

The Overall Research Score combines the four component scores:

Quality, Valuation, and Balance-sheet scores are required for a company to receive a normal overall score. The model does not allow a company with only market data to rank as a fully scored research candidate.

When a component contains missing inputs, the remaining available metrics within that component are equally weighted. Material data-quality concerns may reduce confidence, cap a contribution, or route the company into a verification queue.

The Overall Research Score is best understood as:

How compelling and research-ready does the current combination of quality, valuation, financial resilience, and market evidence appear?

It is not an expected-return forecast and should not be read as one.

Metric glossary

What the underlying variables mean

Limitations

Where the model requires additional interpretation

Asset-light businesses

Companies that require little tangible capital can produce unusually high ROIC. That may reflect excellent economics, but the accounting denominator can also make comparisons less straightforward.

Goodwill, intangible assets, and acquisitions

Acquisition-heavy companies may carry substantial goodwill and intangible assets. The Fieldbook preserves those accounting effects and may also examine acquisition-adjusted quality measures to determine whether reported returns are being distorted.

Intentional leverage

Some durable companies operate comfortably with more debt than the standard model rewards. Debt relative to operating income should therefore initiate a balance-sheet review rather than force a conclusion.

Financial companies and specialized industries

Banks, insurers, real estate businesses, development-stage companies, and certain other industries require specialized analytical models. They may be excluded from the general operating-company score or treated with reduced confidence.

Vendor timing differences

Market capitalization updates continuously, while balance sheets, cash, debt, and some enterprise-value fields update periodically. The system identifies material timing mismatches and routes uncertain cases into data verification.

Historical growth

Historical growth does not guarantee future growth. High CAGR can reflect a low starting base, acquisitions, temporary demand, cyclical peaks, or exceptional conditions that may not persist.

Market confirmation

Market behavior may lead fundamentals, lag fundamentals, or reflect factors unrelated to the company. The Market score is evidence about investor behavior—not proof of business value.

False precision

A score of 82 is not meaningfully exact in the same way as a physical measurement. Scores standardize evidence so that companies can be compared consistently. They do not eliminate uncertainty or judgment.

Methodology governance

Versioning, auditability, and model changes

The Fieldbook treats methodology changes as part of the research record.

Material scoring changes should be:

• identified explicitly;
• tested against the covered universe;
• reviewed for unintended effects;
• assigned a methodology version;
• dated when they become effective;
• preserved rather than silently rewriting history.

Historical reports generally remain as originally published. When a model changes, the current methodology may differ from the methodology used in an older report.

This is intentional. The goal is to maintain an honest record of what the process indicated at the time.

What it is not

Research support, not personalized investment advice

The Market Fieldbook does not provide:

• personalized investment recommendations;
• automatic buy or sell instructions;
• guaranteed forecasts;
• price targets;
• promises of future performance;
• portfolio allocations tailored to an individual reader.

A high-quality company can still be too expensive. A low valuation can still reflect a deteriorating business. A sound thesis can still produce a disappointing investment outcome.

The Fieldbook exists to make the reasoning process more visible, organized, repeatable, and open to revision.

Browse the research archive →

Explore the Company Scorebook →