Valuation
How cheap or expensive a stock currently is relative to its own historical valuation. Calculated from the percentile of the current P/E/P/B multiples within its own multi-year history.
Undervalued stocks, dividend stocks, growth stocks and eight more lists. Each sorted by exactly one metric, not by an overall verdict. A stock at the top of one list can sit near the bottom of the next, and that is what makes the rankings readable.
Recalculated nightly, last on 16/10/2026.
For each score, the top 3 from core and watch. Each ranking shows one dimension, not the full picture.
How cheap or expensive a stock currently is relative to its own historical valuation. Calculated from the percentile of the current P/E/P/B multiples within its own multi-year history.
How strong the current price trend is. Calculated from the distance to the 90-day average and proximity to the 52-week high.
How solid the balance sheet and profitability are. Calculated from the Piotroski F-Score, stability of return on equity, debt (net debt/EBITDA), and balance sheet quality (accruals).
How reliable and well-covered the dividend is. Calculated from years without a cut, the growth streak, and coverage of the payout by free cash flow.
How strong expected growth is and how fairly it's valued. Calculated from consensus growth estimates for earnings and revenue, consistency of revenue growth, and the PEG ratio.
Whether insiders have been net buyers or sellers recently. Calculated from the ratio of insider buys to sells over the last 90 days.
How reliably a company beats expectations on quarterly results. Calculated from the beat/miss rate and the spread of surprises over recent quarters.
The current direction of analyst rating changes. Calculated from the ratio of rating upgrades to downgrades over the last 90 days.
How extreme social media sentiment currently is. Calculated from volume-weighted Stocktwits/Twitter sentiment over recent days.
How high structural short-selling pressure is. Calculated from the daily short volume ratio (FINRA) and a Reg SHO threshold list (Nasdaq) over recent trading days.
How much the business model will benefit from megatrends in the coming years and how robust it is. Calculated from an AI assessment of megatrend exposure (which sub-themes, how directly, tailwind or headwind, revenue share) and business-model quality (recurring revenue, pricing power, moat, capital intensity, customer concentration); the scoring itself is deterministic.
Basics
A ranking is not a selection, it's a sort order. Three things matter for it to be useful.
Each of the eleven lists condenses exactly one dimension into a number: valuation, trend, quality, dividend, growth, insider activity, earnings, analyst revisions, sentiment, short pressure, and business model. A stock leading on valuation can trail on trend. An averaged overall rank would hide that contradiction instead of showing it.
A score is a relative value within the same universe, not an absolute grade. "Strong on valuation" means: cheaper than the large majority of comparable stocks, measured across several metrics at once, not a single P/E.
All eleven scores are recalculated every night for the entire analyzed universe. A stock rises or falls because its data changed, not because someone edited the list. What led yesterday stays traceable in the history.
A ranking shows who currently leads on one metric. It does not say what to buy. Whether a cheap valuation is a good deal or a justified discount is answered only by analyzing the individual stock.
FAQ
How the scores are built, what they sort by, and what a ranking does not tell you.
To the methodologyThrough the valuation ranking. It condenses several valuation metrics of a stock into one value and sorts by that, instead of comparing just one P/E ratio. At the top are the stocks that are cheapest measured against their own numbers and comparable companies. Whether the discount is justified is answered only by the individual analysis.
The dividend ranking sorts by a score that combines years without a cut, the growth streak, and coverage of the payout by free cash flow. A high yield alone isn't enough: it says nothing about how secure the payout actually is.
Once a night, for all eleven scores and the entire analyzed universe. Intraday price moves don't change any placement.
Deliberately not. An average across eleven dimensions produces a number that no longer means anything, and it smooths away exactly the contradictions that are interesting. Anyone who wants to combine several criteria builds their own ranking and sets a minimum strength per dimension.
No. They show who currently leads on a single metric. That's a sort order, not investment advice and not a call to buy or sell.
The stocks from the two upper stages of the screener, i.e. the ones the market scan has confirmed as notable over several days or weeks. The entire tradable universe is scanned; the stocks that made it into closer watch are scored.
You can see the nightly scan without an account. With a free account you see all scores and lists in full and ask the agents yourself, with five credits on the house.
Not investment advice, not a recommendation, not a forecast. The portfolios are virtual, every decision is timestamped and verifiable.