Stock Screener narrows thousands of names into a scored shortlist using a two-phase architecture — a coarse filter on sector / cap / financial constraints, then detail enrichment and 5-factor scoring on the top candidates. Style presets tune the factor weights for value, growth, or quality hunts.
Updated—
Built-in skill
Driven
Built-in skill
A single call to the provider screener narrows thousands of names via sector, market cap, and basic financial constraints. Results sort by market cap, and the top 10 advance to detail enrichment — the coarse filter is about investability, not ranking.
The top 10 candidates get key metrics, ratios, and price-change data fetched in parallel via background jobs in a single command. Ten stocks enrich in roughly the time of one sequential fetch — the parallelism is what makes a fully-scored screen fast enough to iterate.
Each candidate is scored on five dimensions — Valuation, Profitability, Growth, Momentum, Safety. Valuation is ranked relatively within the result set (PE and EV/EBITDA against peers in the scan) rather than against absolute thresholds — a PE of 25 scores well among growth stocks but poorly among utilities. Growth is capped so turnaround outliers do not dominate the factor. Missing dimensions (e.g., banks lacking EV/EBITDA) get a neutral score, redistributing weight to what is available.
Four preset weight schemes tune the model for different hunts: balanced (general-purpose), value (heavy valuation + safety), growth (heavy growth + momentum), and quality (heavy profitability + safety). Swapping the preset reweights the same 5 factors — no recomputation needed — so the same universe reveals different leaders under different investment styles.