One screen for every stock you follow, scored the same way every night.
Most tools tell you what a stock did. Tickr Lab tells you where it stands — on a fixed scale that means the same thing this quarter as it did five years ago — and then names the one rule that produced its verdict. Nothing here is a tip, and nothing is a black box.
A watchlist tells you the price. It does not tell you where a stock stands.
If you follow more than a few dozen companies, the work is not finding data — it is applying the same judgement to all of them, on the same day, without getting tired by the fortieth row.
Tickr Lab was built for that specific problem. Every stock in the universe is scored on the same eight momentum factors and the same fundamental measures, from the same daily bars, every night after the close. The scores are absolute, not ranked against each other — so adding a ticker does not move anyone else's number, a weak quarter looks like a weak quarter instead of "best of a bad lot", and the same stock's score a year ago is directly comparable.
That last property is the quiet one, and it is why the scoring model was rewritten. A percentile-ranked score has a median of a half by construction; measured at six dates spanning fifteen years, the old model's median landed on the same value in every single period. It could not express a weak market at all.
The screener is descriptive rather than predictive — a consistent way to see where every stock stands and how each one got there. That is a legitimate thing for a tool to be, and it is stated here rather than dressed up as something else.
Eight capabilities, one row per stock.
Everything below reads from the same nightly snapshot, so no two surfaces can disagree about what a number is.
The screener
measures per stock across eleven column groups —
momentum, quality, trend, relative strength, volume, size, fundamentals, ownership.
Filter any column with a small grammar (>15, 2..8,
50B..), sort on any of them, and save the set you actually use as a
named view.
Advice that names its reason
Ordered rule lists, first match wins, loss-avoidance rules above every buy rule. Every verdict therefore has exactly one nameable cause, and clicking it opens the full ladder: which rung fired, which were never reached, and where this stock sits on each condition.
Screens and filters
starter screens in six groups — recovering names, breakouts, cheap-and-growing, reporting soon — each one a saved set of filters you can inspect, edit and re-point at any portfolio.
The bar archive
Daily open, high, low, close and volume going back to , held locally rather than re-fetched. Splits are detected and repriced; provisional closes are healed the next session. Every chart, score and backtest reads from it.
Research surfaces
A strategy backtest with costs and cash yield as sliders, a single-stock exposure simulator, an indicator lab that shows the universe-wide result beside whatever you are tuning, and an advice backtest that replays the rules on a past date.
Portfolios and baskets
shared lists, plus personal lists for each account that filter the shared data without adding to it. Any list aggregates into a basket view with its own curve against the universe.
Headlines and earnings
Per-stock headlines with source and timestamp, a ticker of the last twelve hours, and recorded earnings quarters with estimate, actual and surprise.
Phone, cards and an assistant
A separate phone page for browsing screens, a card studio that renders shareable 1080px images from the same rows, and an assistant that answers questions about the data it can actually see — and says so when it cannot.
Four questions, asked in order, made visible as columns.
The intermediate steps are not hidden behind the answer. Each condition family collapses to one word and gets its own column, so the derivation reads left to right across the table and any cell opens the reasoning behind it.
Company type decides which rules apply, not just their thresholds — an early-stage business is judged on growth, gross margin and burn, never on a P/E that would be meaningless. Funds are scored on the technical rules alone.
What is actually held, counted rather than claimed.
Every figure in this exhibit is read from the database when you load the page. None of it is typed into the text, so it cannot go stale between deploys.
| Data type | Stocks | Depth | Reach |
|---|---|---|---|
| Daily price bars Open, high, low, close and volume, split-adjusted | |||
| Five-year history Enough depth for the 5Y return column | The rest fill in as the archive is deepened, one credit a stock. | ||
| Fundamentals history Valuation, margins, growth and balance sheet, per day | Recorded from . Cannot be backfilled — the provider only ever returns today's numbers. | ||
| Earnings history Estimate, actual and surprise, by quarter | Back to . | ||
| Headlines Title, source, link and timestamp — never article bodies | Pruned to the last 21 days per stock. |
Session counts are estimated from each stock's first and last bar, not counted row by row: an exact count reads more than a million rows, and the database bills by rows read. The estimate is deliberately labelled rather than presented as exact. Figures refresh hourly.
The limits are part of the product, not fine print.
Every claim a tool like this could make has been tested here, and most of them came back empty. Those results are kept rather than quietly dropped.
It does not predict
Six framings have been tested against the full archive — range position, mean reversion with falling-knife filters, RSI across forty settings, risk-adjusted velocity across one hundred and forty-four, a faster trend rule, and ranking within a verdict tier. All six came back flat at a one-to-two-month horizon. Nothing in the product claims otherwise.
It does not recommend
The Advice column is a mechanical reading of the rules against the table, not advice tailored to anyone. The assistant will report what the rules said and attribute it, and will decline to give a buy, sell or hold verdict of its own.
It does not track positions
There is no portfolio accounting anywhere in this product — no cost basis, no profit and loss. Portfolios are groupings of stocks to look at, nothing more.
It holds US listings only
NASDAQ, NYSE and AMEX. A bare ticker can silently resolve to a foreign listing of a different company, so candidates are checked against the published US listing file before anything is added.
The universe is survivorship-lite, not survivorship-free
A deliberate ballast of laggards sits in the universe so that backtests have losers in the pool. Companies that went bankrupt or were acquired are still absent, and those are the tail that matters most. Any result that improves when the universe shrinks is treated as a fact about the sample.
Refreshed after the close, and through the session.
A full pass runs nightly at 4:15 PM New York, fifteen minutes after the bell: prices, rotating fundamentals, scores, verdicts and the snapshot every reader sees. Prices refresh again every thirty minutes through the trading day, and the server decides whether the market is actually open rather than trusting a calendar.
Runs are logged with their rounds, credits and outcome, and a watchdog on a separate scheduler reports a night that did not happen — because a watchdog sharing the scheduler it watches would miss the same failure.
Last refreshed —.
See where your list stands.
The guest preview opens the full screener on a handful of stocks — no account, nothing to cancel.