What Happens Under the Hood When a Scanner Runs?
The scanner ingests a data feed—prices, volume, bids, asks, and often fundamental fields—for each symbol in its universe. For every symbol, it evaluates boolean and numeric conditions: Is price above five dollars? Is relative volume greater than two? Did price cross above the twenty-day high? Symbols passing all required filters enter the result set. Optional ranking sorts by gap percent, dollar volume, or custom scores. The process repeats on a schedule—every tick, every minute, or once after the close—depending on platform and scan type.
Latency and data quality determine usefulness: delayed quotes produce delayed opportunities.
How Are Filter Conditions Structured?
Filters combine with AND logic—all must pass—or grouped OR logic for alternatives. Range filters specify min and max: price between ten and eighty dollars. Comparison filters test against a value or another field: close above fifty-day moving average. Cross filters detect events: fast MA crossed above slow MA today. Time-window filters restrict to the current session bar or last N days. Complex scans nest groups—large cap AND (gap up OR high relative volume)—to balance breadth and precision.
Test each filter's contribution by removing one at a time; often a single overly tight condition empties results.
How Do Alerts Differ From Static Scan Results?
A static scan shows what qualifies now. An alert fires when a symbol newly qualifies or when a numeric threshold crosses—first time relative volume exceeds two, or price breaks opening range high. Alerts reduce screen watching: you review charts only when the scanner notifies you. Effective alert design includes cooldown rules so the same symbol does not spam every bar. Pair alerts with sound or visual cues prioritized by watchlist tier.
Define whether alerts trigger on every tick or on bar close—intraday strategies often need close confirmation to avoid wick fakeouts.
What Data Fields Do Scanners Typically Access?
Price and change: last, open, high, low, percent from prior close, percent from open. Volume: current, average, relative volume ratio, dollar volume. Volatility: ATR, beta, intraday range. Technical: moving averages, RSI, MACD, new highs/lows. Fundamental: market cap, float, sector, earnings date. Session context: time of day, minutes since open. Platform-specific scores may combine multiple inputs into one rank—useful for sorting large result sets before manual review.
Know whether your scanner uses consolidated or primary exchange volume—thin names can look distorted on one venue.
How Do You Build Scans That Produce Tradable Output?
Anchor filters to your strategy's non-negotiables: minimum average daily volume for liquidity, minimum price to avoid penny-stock noise, maximum spread proxy if available. Add one catalyst or pattern filter tied to edge. Cap results with a sort—top twenty by relative volume—so you are not overwhelmed. Backtest informally by saving daily results and noting which names would have met your chart rules. Iterate weekly. A scanner that returns fifty random movers is a distraction; one that returns five qualified names is an edge tool.
Name scans after the setup they hunt—ORB breakout, gap and go, pullback to VWAP—so you run the right tool for the session regime.