Strategy Library
The three setups the scanner actually checks for, and the reasoning behind each.
MA Bounce
Rejection off the 8, 21, or 50 EMA
Price is trending up, dips down, taps one of the key moving
averages — the 8, 21, or 50 EMA — and rejects: closes back above
that same line on a green candle. Any of the three counts. The idea:
in a healthy uptrend, these lines tend to act as dynamic support —
each tap is buyers stepping back in before the trend continues,
rather than a break in the trend.
What the code checks
Stock is trading above its own 50 EMA (uptrend context). Signal
fires if the day's low touches within 1% of the 8, 21, OR 50
EMA, and the candle closes back above that same EMA on a
green candle.
Why it can fail
A tap-and-reject that loses the EMA again the very next day usually
means the bounce wasn't backed by real buying — just a brief wick
back above the line. Follow-through over the next day or two matters
more than the single signal candle.
See historical win rate for this setup on any ticker →
Breakout
Clearing yesterday's high out of a tight base, on volume
Price has been consolidating in a tight range — tight relative to
that stock's own typical daily range, not a flat percentage that
would be meaningless across tickers with very different volatility
— then closes above the previous day's high on real volume. This
doesn't require clearing some older high from earlier in the base;
the level that matters is the one right in front of it (PDH), not a
stale high from days ago.
What the code checks (daily)
The prior 4 days' high-low range stays under
2× the stock's own 20-day average daily range (the
consolidation, sized to that stock's own volatility rather than a
fixed percentage). Signal fires when the close is above the
previous day's high, on volume at least 1.2× the
20-day average. This daily version is what decides breakout on the
Daily Scan, the single-ticker scanner, and the historical stats
below.
Intraday chart confirmation
The interactive chart's intraday views (1m through 4h) work
differently. Each timeframe independently checks its own candles
against the same PDH-based level — requiring a close above it with
volume above 1.2× that timeframe's own 20-bar average —
without waiting for the daily candle to close, and without requiring
the full daily rule above to have fired that day. A breakout
confirmed on 1m does not imply it's confirmed on 5m, 15m, or any
other timeframe — each is evaluated on its own. A candle that trades
above the level but closes back below it is marked as a sweep (a
muted gray dot), shown for context but never counted as a signal.
Why it can fail
Breakouts without real volume behind them ("wick breakouts") often
reverse quickly — price pokes above the level intraday but can't
hold it by the close, or holds for a day and then fails on the next.
A breakout on unremarkable volume is a weaker signal even if the
price technically cleared the level.
See historical win rate for this setup on any ticker →
Undercut & Rally
A failed breakdown below the previous day's low
Price dips below the previous day's low — often running the
stop-losses sitting right below it — but reclaims that low and
closes back above it by end of day. The dip below isn't
automatically bearish on its own; what matters is whether sellers
can actually keep price down there, or whether it gets bought back
up almost immediately. A failed breakdown is often a sign control
has shifted back to buyers.
What the code checks
Signal fires when today's low drops below the previous day's
low, but the candle's close finishes back above that
previous day's low.
Why it can fail
Not every dip-and-recover is meaningful — since this checks only the
single prior day's low, it can fire on ordinary day-to-day noise
around a level with no real significance, not just at levels other
traders are actually watching.
See historical win rate for this setup on any ticker →
A note on all three
These are pattern definitions, not predictions. The historical stats
on the scanner page show how each setup has actually performed on a
specific stock — some tickers respect these patterns much more
consistently than others. See
Methodology for
the full disclaimer on backtested results.