Where the numbers come from
Fundamentals are drawn from company filings via SEC EDGAR and from market-data vendors (Yahoo Finance, Finnhub), with the listed-security universe taken from NASDAQ Trader. Prices and market data are delayed and are not suitable for trading decisions. Where a figure is derived rather than reported — a score, a fair-value estimate, a growth rate — the methodology page documents how it is calculated, including the choices we rejected.
How it gets checked
The universe is re-scored after every US close, and each run passes a quality gate before anything publishes: coverage thresholds, sanity bounds on individual metrics, and reconciliation against filed figures. A run that fails the gate does not ship. Model changes are tested out-of-sample rather than tuned until the backtest looks good — we would rather publish a smaller honest edge than a flattering over-fitted one.
Scores shown in ranked lists come from the most recent completed run and carry an as-of date; a stock's own page re-scores against live data, so the two can differ slightly for a name that has moved since the last close. That is a freshness difference, not a disagreement.
Corrections policy
If a published figure is wrong, we fix it. Data errors are corrected at the source so the fix flows to every page that uses the figure, and because the universe re-scores daily, a corrected input is reflected in the next run. Where an error materially changed a published conclusion — not a rounding difference, but a wrong call — we say so in the article rather than editing it silently.
Errors can be reported at any time via the contact page. Reports about a specific ticker are most useful with the figure you believe is wrong and, where possible, the filing that contradicts it.
Where automation is used, and where it isn't
The rankings are quantitative: a model produces them, not an opinion. Most of the writing is automated too. Per-stock analysis and the daily and weekly recaps are generated from the same audited figures displayed on the page, then checked against those figures before publication, so the prose cannot claim something the data does not support. The explanatory guides are drafted with AI assistance and published under our editorial responsibility — we decide what they cover, and we answer for what they say.
We would rather state that plainly than imply a newsroom that doesn't exist. What we do not do is present generated text as the work of a named person: every page says which applies to it, and a personal byline is used only where someone genuinely wrote the piece.
We do not submit the automated daily recaps to search indexes. They remain readable and linkable as an audit trail of what the model said on a given day, which is the point of keeping them.
Independence and how we make money
No company can pay to be covered, to rank higher, or to be removed. We have no business relationship with any company we score and hold no positions we promote to readers.
The site is funded by advertising, optional paid subscriptions, and affiliate links to brokerages — disclosed on the pages that carry them. None of these influence the model's inputs, its scoring, or which stocks appear. Advertising is served by third parties and is not selected by us on a per-stock basis.
What this site is not
Nothing here is personalised financial advice, and no ranking is a recommendation to buy or sell. The model does not know your goals, your tax position, or your risk tolerance, and a high score is a statement about a business's measured quality, growth and valuation — not a prediction. See the terms for the full disclaimer.