Methodology

BriMind AI Score

The BriMind AI Score is a directional, educational signal — not a recommendation to buy or sell. It is designed to give investors a quick, structured view of how a stock compares to its peers across multiple dimensions simultaneously.

The score is built from six sub-scores, each measuring a different aspect of a company's fundamentals and market performance:

• Value — how a company is priced relative to its earnings, cash flow, revenue, and enterprise value. A higher value score indicates more attractive pricing relative to peers.
• Growth — trends in revenue and earnings, combined with long-term price appreciation and analyst consensus estimates of future upside.
• Profitability — margins and returns on capital, measuring how efficiently a company converts revenue into profit and generates returns for shareholders.
• Financial Health — balance sheet strength, liquidity, and debt levels — factors that affect a company's ability to weather downturns and fund future growth.
• Momentum — recent price performance relative to moving averages and historical returns, reflecting market sentiment and trend strength.
• Risk — historical volatility and drawdown measures, indicating how much price variation investors have historically experienced.

Each metric within a sub-score is percentile-ranked against a peer group. Where a company's sector has enough members, ranking is done within that sector; otherwise we fall back to the full universe of covered stocks. This peer-relative approach means a score of 75 indicates a company ranks in the 75th percentile among its peers on that dimension — not that the company is universally "good" in an absolute sense.

Sub-scores are then combined into a single composite score using proprietary weights. Outlier values are normalized before scoring to prevent a single extreme data point from dominating the result.

Limitations: The AI Score is a simplified, automated summary of publicly available data. It does not account for qualitative factors (management quality, competitive dynamics, regulatory environment), does not reflect a reader's personal financial situation, and can be wrong. It should be treated as one input among many, not a standalone signal to trade on.

Does it work? Every score is logged the day it is issued and graded once its horizon has fully elapsed, against the stock and against SPY over the same window. The results — good and bad — are published on our AI Score track record.

1-Year Price Target

Where we display a 1-year price target, it is derived from a proprietary machine learning model trained on historical financial and market data. The model produces a point estimate (target price) and a confidence score. The confidence score influences the width of the displayed target range — lower confidence produces a wider range, reflecting greater uncertainty.

When Wall Street analyst consensus data is available, our model's target is blended with the analyst mean price target to produce the final displayed figure. This blending is designed to ground the output in professional analyst expectations while contributing our model's independent signal.

Limitations: Price targets rely on assumptions about future growth, earnings, and market conditions that may not hold. They are most useful as a rough estimate of relative valuation and directional expectation — not as a prediction of where a stock will trade on a specific date. We do not guarantee the accuracy of any price target, and past accuracy does not indicate future results.

Short-Term Price Prediction

Some analysis pages display a short-term (5–6 trading day) price projection. This is computed using technical price indicators — including short-term moving averages and the Relative Strength Index (RSI) — derived from recent trading history.

Limitations: Short-term price projections are based on recent technical patterns and are one directional signal among several that inform our buy/sell indicators. They are highly speculative — markets can be significantly influenced by news, earnings surprises, and macroeconomic events that no technical model can anticipate. They should not be the sole basis for any investment decision, and do not constitute a guarantee of future price movement.

Insider Trading (SEC Form 4)

Corporate insiders — officers, directors, and 10% owners — must report their trades in company stock to the SEC on Form 4, generally within two business days. We read those filings directly and aggregate them per ticker over a rolling 90-day window.

What we count. Only open-market purchases (transaction code P) and sales (code S) are included. Awards, grants, gifts, tax withholding, and option exercises are excluded, because they do not reflect a discretionary decision to put personal money at risk or take it off the table. An insider who buys the same stock on four separate days appears as one buyer and four filings; we report both numbers because they say different things.

Cluster buys. When three or more distinct insiders at the same company file open-market purchases within 60 days, we flag it as a cluster buy. Independent purchases by several people who see the internal numbers first is one of the more durable patterns in the Form 4 data, and it is harder to explain away as one person's liquidity event.

Unusual activity. We compare the dollar value of recent buying against the same ticker's trailing 11-month baseline. A large multiple of that baseline, or the first insider buying in the whole lookback period, is flagged as unusual. This measures a change in behavior, not the size of the trade in absolute terms.

Credibility score. The 0–100 score weights who is buying (officers rank above directors, who rank above 10% holders), how much the purchase changes that person's own stake, how many insiders are involved, and the buyer's prior Form 4 track record where we have enough history to measure it. It is a measure of signal quality, not of expected return.

Unusual selling. Sales are flagged when they are large relative to the insider's holding and were not executed under a pre-arranged Rule 10b5-1 plan, so the timing was the insider's own choice. Selling is a weaker signal than buying in both directions: insiders sell to diversify, pay tax, and buy houses, and most sales mean nothing.

Limitations: Form 4 data is backward-looking and arrives with a reporting lag. A single purchase, however well-timed historically, is not a forecast. Insider flow is one input among many and should be read alongside valuation, fundamentals, and recent price action — not in place of them. See the forward-return track record for how these signals have actually performed.

Data Sources

Figures used in our analysis and tools are drawn from:

• Company filings and investor relations releases — 10-K and 10-Q annual and quarterly reports, earnings press releases, and official investor relations disclosures.
• Market data providers — stock prices, trading volume, and fundamental financial metrics are sourced from established financial data providers.
• Analyst consensus data — where displayed, analyst price targets and recommendations are sourced from aggregated Wall Street consensus estimates.
• ETF and index provider data — ETF holdings, expense ratios, index methodology, and fund performance figures are sourced from official fund prospectuses and index provider publications.

We do not use anonymous, unattributed, or scraped data without a clearly named upstream source.

Update Cadence

• AI Scores — recomputed nightly based on the latest available financial and price data.
• Price data and technical indicators — updated nightly; intraday prices are also refreshed during market hours on select pages.
• 1-Year price targets — model targets are refreshed periodically; the analyst consensus component is pulled live when the page is loaded.
• Editorial articles and comparison pages — reviewed and updated on a rolling basis. Articles with materially stale data are flagged or corrected per our Editorial Policy.

What This Is Not

Nothing on this page, or produced using this methodology, constitutes personalized investment advice. Our tools and content are educational in nature. Investing involves risk, including the potential loss of principal. Please see our full Disclaimer and consult a licensed financial adviser for advice specific to your situation.

Questions

If you have questions about how a specific score, target, or figure was derived, contact us and we will do our best to explain.