Stock Analysis Methodology

1. How Does the MoonshotScore Work?

MoonshotScore V2 is the only current scoring model. MoonshotScore V2 covers eligible common stocks and ADRs on NASDAQ, NYSE and AMEX. OTC listings are outside the scheduled universe; funds, ETFs, warrants, units, SPACs, preferreds, and notes carry no MoonshotScore. A company without a usable V2 result has no current score; an archived V1 value is never substituted.

The V2 five-pillar model. Companies are assessed against sector peers where there are enough observations. Smaller sector samples are pooled. Available pillar inputs form a weighted composite, financial-fragility penalties are applied, and the result is ranked across the scored universe. Data coverage and company size then draw the result toward 50.

• Business Quality: Is this a genuinely good business? Profitability, margins, and returns on capital versus sector peers.
• Financial Safety: Could this blow up? Debt load, cash position, and overall balance-sheet strength.
• Valuation: Am I paying a fair price? Price measured against fundamentals, relative to peers.
• Growth Durability: Is the growth real and likely to last? Quality of the revenue and earnings trend, not one-off spikes.
• Momentum: Is the market already moving on this? Recent price trend.

Score bands: Exceptional (80-100), Strong (65-79), Fair (45-64), Weak (below 45). The result is adjusted after ranking, so the final bands do not contain a fixed share of companies. The five displayed pillar percentiles are explanatory breakdowns, not values to average into the headline. The score is not a probability of future returns.

2. Historical V1 Records (9-Factor Model)

The retired V1 model used the nine signals and fixed weights below. Its Operating Leverage factor reused revenue growth; the signals were not all independent. These weights explain historical records and are not the current V2 formula.

Legacy V1 nine-signal calculations are retained as historical records only. They are not used as a fallback when a current V2 score is unavailable.

• Revenue Growth (20%): Year-over-year revenue growth scored against fixed thresholds. Faster-growing companies score higher — the single most important factor.
• Gross Margin (10%): Gross profit as a percentage of revenue. Higher margins indicate pricing power, competitive moats, and scalable business models.
• Operating Leverage (10%): Scaling-efficiency proxy computed from the revenue-growth figure above — it takes no separate cost or margin input. Because of that, revenue growth effectively drives 30% of this model, not 20%.
• Cash Runway (10%): How long the company can operate at its current burn rate — enough cash to fund the growth strategy.
• R&D Intensity (10%): Research and development spending as a percentage of revenue. High intensity signals commitment to innovation and a future pipeline.
• Insider Activity (15%): Net insider buying or selling. When insiders buy their own stock, it signals confidence in the company's prospects.
• Trading Liquidity (5%): Daily volume turnover relative to the share float. Healthy turnover means smoother entry and exit. This is a liquidity measure, not short interest — we do not use short-interest data.
• Price Momentum (10%): Price performance across 1-, 3-, and 6-month windows.
• News Sentiment (10%): AI-analyzed sentiment from recent news articles and analyst reports.

Display letter grades: A+ (80-100), A (70-79), B+ (60-69), B (50-59), C (40-49), D (30-39), F (below 30).

3. What Data Does the Score Use?

• Financial Modeling Prep (FMP) — Primary: Fundamentals, financial statements, analyst ratings, insider transactions, technical indicators, and intraday quotes.
• Yahoo Finance — Fallback: Supplementary market data, ETF data, earnings estimates.
• Alpaca Markets — Fallback: Intraday and historical price data for charting and technicals.
• SEC EDGAR: Corporate filings (10-K, 10-Q, 8-K) and insider filings for fundamental analysis.
• News: Headlines from financial news providers, used for sentiment and catalysts. Volume varies; the newsroom pipeline that fed the Market Intelligence Journal is currently paused.

4. The 7-Dimensional Quality Gate

Every stock page is scored across 7 dimensions. This gate measures the page, not the company: whether there is enough data and content to be worth publishing, whether the page is distinct from its siblings, and whether it carries the right disclosures and markup.

• Data Completeness (25%): Profile, financials and price data present.
• Content Uniqueness (20%): How much of the page is specific to this company rather than shared template text.
• Content Freshness (15%): How recently the page's data and written content were updated.
• Trust Labeling (15%): Disclosures, data-source attribution and the not-advice notice.
• Internal Linking (10%): Whether the page is reachable from, and links out to, the rest of the site.
• Schema Validity (10%): Structured data present and well-formed.
• SSR Sufficiency (5%): Whether the server-rendered page carries the content on its own.

A page scoring below 10 is not generated at all. A page scoring below 30 is generated but not submitted for indexing. In practice every active stock page currently clears 30, so the indexing threshold is not excluding anything today.

5. Validation Pipeline

Translated pages are checked against their English source through a 6-layer pipeline before publication. This is a translation-fidelity check — it verifies that the translation still says what the English said. It does not re-verify figures against the data provider:

1. Numeric carry-over (figures in the English text reappear in the translation)
2. Ticker consistency (symbol mentions match the page entity)
3. Length ratio (translated content stays in proportion)
4. Paragraph preservation (≥60% structural fidelity)
5. HTML tag validation (no malformed markup)
6. Title sanity (5-300 char range, no template leakage)

A translation that fails is regenerated automatically, up to a retry limit; one that keeps failing is left unpublished rather than shipped. Figures themselves are validated where they are produced, against the provider response, not here.

6. Update Frequency

• MoonshotScore: The V2 five-pillar model is scheduled to recalculate once daily at 10:15 UTC (13:15 in Türkiye). The score's computation date shows its last successful calculation; a newer price does not mean the score has refreshed. A failed update leaves the previous dated score in place.
• Stock prices: Updated throughout U.S. market hours.
• Dossiers: Written once and not currently on a refresh cycle — the automated rewrite was switched off in June 2026. Each dossier carries its own date; read it as of that date.
• Financial statements: After quarterly and annual filing processing.

7. Limitations & Disclaimers

MoonshotScore is an analytical signal, not a prediction. Past performance does not guarantee future results. Scores reflect the data used in the dated calculation; market conditions change continuously. AI may occasionally produce inaccurate outputs despite validation; report errors to sedat@stockexpertai.com.

Not financial advice. Stock Expert AI is an educational research platform, not a registered investment advisor. Consult a licensed financial professional for personal investment decisions.

8. Related Documents

• AI Methodology & Limitations — Technical detail of AI models, validation, hallucination prevention.
• Editorial Policy — Content creation, error correction, accountability.
• MoonshotScore Detail — Each factor explained.
• Full Disclaimer — Legal disclosures.