All studies
Live Scanner · 1H · Top 25 Most Liquid

Bilbo Box Scanner

The 25 most liquid names in the S&P 500 and Nasdaq-100 (ranked by 20-day average dollar volume), scanned on hourly compression. The Bilbo Box is the range of the first five compression bars; once it locks, the first trade outside the box is the break. Immediate entry on the break was the winning variant in the Bilbo Box Breakout study (50,889 SPY breaks), with the edge concentrated in 1–4 bar boxes.

loading…
First trade outside a locked box within the last 7 hourly bars. RTH breaks first, then newest. The box's opposite boundary is the study's stop reference (1R = box width). Session matters: in the study's 3,477 1h breaks, RTH-session breaks carried the whole edge (median +0.12R, n=1,479); premarket breaks were flat (+0.02R, n=1,471); after-hours breaks were negative (−0.17R, n=527). Only RTH breaks fire alerts; ETH rows are dimmed. Telegram is stricter than this table: a break only alerts once its bar has closed, out of compression, with the CLOSE outside the box (no wick-only alerts), on a box ≤3×ATR wide.
Box locked, no break yet. Break-watch stays live for 20 bars after lock; a fresh compression start kills the pending box. = live price within 0.25×ATR of a boundary — a touch is plausible within the hour. Telegram pre-alerts only for quality boxes: out of compression, ≤3×ATR wide, price within 0.15×ATR, during RTH.
Currently in compression, box still growing (locks at 5 bars or when compression ends, whichever comes first).
Watched names with no live box lifecycle right now. A dot in the comp column means the stock is inside a compression run whose box already resolved.

How to read it. The range bar shows where price sits inside the box ( above midpoint / breaking up, below / breaking down). Width is the box height in hourly ATR-14 multiples — tighter boxes were the better setups in the study. 1–4 marks early-exit boxes, where the measured edge concentrated. RTH breaks are the tradeable ones per the study; PRE breaks showed no edge and AH breaks were net losers — boxes still form and lock on extended-hours bars (exactly as in the backtest), but only an RTH-session break is a signal.

Caveats. Bars are extended-hours hourly (top-of-hour 04:00–19:00 ET starts, 16/day), the same grid as the SPY backtest and TV charts with ETH on. Primary data is Massive 1h aggregates with SIP-filtered wicks; if that source degrades the scanner falls back to Yahoo per ticker, where pre/post wicks are unreliable and the bar body is used as the extreme outside RTH. Wicks are clipped at 2% beyond the body (the standard bad-tick rule from the studies). The last bar is in progress during the session — a "break" can un-print if the bar reverses before it closes. This is a watchlist builder, not a signal service.

Refreshes automatically — next in 60s. Scanner runs server-side on a cron during RTH. Times shown in Central Time. Methodology: Bilbo Box Breakout study.