A chart pattern is a compressed description of an inventory situation. A flag is not magic geometry; it is "the move paused, sellers did not show up, and buyers are still there". Read them that way and they stop being superstition.
Continuation patterns
Bull flag / pennant
A sharp move (the pole), then a shallow drift against it on declining volume (the flag), then a break in the original direction. The most reliable intraday pattern there is, and the backbone of the first-pullback setup.
- ·Trigger: break of the flag's upper boundary, ideally with volume above the flag's average.
- ·Invalidation: a close below the flag's low, or below the pole's midpoint.
- ·Fails by: time. A flag that takes longer than the pole that made it is not a flag, it is a top. If the drift exceeds roughly the duration of the pole, stand down.
Ascending triangle
Flat resistance, rising support. Buyers pay more each attempt while a fixed seller defends one price. It resolves when the seller is exhausted or cancels.
Fails by: the seller being an ATM programme rather than a trader. An algorithmic seller does not get exhausted; it sells every rally all day. This is the specific failure that makes ascending triangles unreliable on diluting microcaps, and it is checkable in the filings.
Cup and handle
A rounded base, then a small pullback near the rim, then a breakout. Genuinely useful on a daily chart. Intraday on a low float it usually needs more time than the session has. Fails by being called too early: most intraday "cups" are just a range.
One of these is a bull flag. Which one, and how do you know before the break?
Reversal patterns
Double top / double bottom
Two attempts at a level, the second failing to exceed the first. Its meaning is straightforward: the second wave of buyers was smaller than the first.
- ·Trigger: break of the intervening low (the neckline), not the second peak.
- ·Invalidation: a new high above the second peak.
- ·Fails by: being the second test of a level that then breaks on the third. Double tops on strong low floats are frequently just consolidation before continuation.
Head and shoulders
Three peaks, the middle one highest, with a neckline across the intervening lows. Well defined and widely watched, which is most of why it works. Fails by: the right shoulder being higher than you allowed. Define the neckline first and let it invalidate rather than adjusting the pattern to keep it alive.
Rising wedge / parabolic exhaustion
Higher highs and higher lows, but the range compressing and volume falling: the move is accelerating on less and less participation. This is the shape behind the parabolic-fade archetype, and the one worth learning best because the reversal is violent.
Fails by: being early. A parabola can add another 40% while you are being correct about it. The climax is only identifiable afterwards, which is why the fade is the lowest win rate of the six archetypes and needs the tightest risk.
The pattern nobody teaches: the halt continuation
Unique to this corner of the market, and more actionable than half the classical patterns. A volatility halt on a thin float, a reopening auction that prints near the halt price, then continuation. The halt itself is the pattern: it tells you the book was too thin to absorb an ordinary order.
Fails by: the reopening cross printing well away from the halt price, which repricings on news-pending halts routinely do. A T1 reopen is a repricing, not a continuation. Know which code you are in.
Done with this module?
Close it and the candle prints on your tape · stored in this browser