Triangles, wedges and rectangles are consolidation patterns formed by converging or parallel boundaries. An ascending triangle has a flat top and rising lows; a descending triangle has a flat bottom and falling highs; a symmetrical triangle narrows from both sides; a rectangle has parallel boundaries. All describe contracting or contained volatility, and all are traded on the break of a boundary rather than on the shape itself.
Triangles, wedges and rectangles are consolidation patterns defined by their boundaries. They are the multi-candle version of the compression covered in Part 12 — a market narrowing toward a decision.
Four shapes, distinguished only by the slope of their boundaries.
The two triangles
Ascending triangle. Flat resistance above, rising lows below. Buyers pay progressively more each time; sellers defend one fixed price.
Descending triangle. Flat support below, falling highs above. The mirror.
Symmetrical triangles and rectangles
Symmetrical triangle. Both boundaries converge. Neither side is advancing on the other — the range simply narrows.
Rectangle. Parallel boundaries. A range, in the Part 4 sense, given a pattern name.
A wedge is a variant where both boundaries slope the same way while converging: a rising wedge tilts upward and tends to resolve down, a falling wedge tilts down and tends to resolve up.
The honest version of the textbook claim
Textbooks assign directional bias: ascending triangles break up, descending break down, rising wedges break down.
The reasoning is sound. In an ascending triangle, buyers are demonstrably willing to pay more each time while sellers hold one price — that asymmetry is genuine information, and it does tilt the odds.
But two cautions belong alongside it.
The bias is a tilt, not a rule. Ascending triangles break downward often enough that trading one before the break is speculation. The pattern tells you where the decision will happen, not what it will be.
Context outranks shape. An ascending triangle forming within a larger downtrend is a counter-trend setup regardless of its own bias. The higher-timeframe trend matters more than the pattern’s internal geometry.
The practical stance: use the shape to identify where the break will occur and to prepare both directions, then trade whichever happens.
Time inside the pattern
Boundaries converge toward an apex. Where price sits within that convergence matters.
Most patterns resolve in roughly the middle-to-late portion of their range — after enough compression to build energy, before the boundaries meet. A pattern that drifts all the way into the apex has typically exhausted itself, and the eventual break tends to lack force.
The reason is participation. Compression works because positions accumulate while nobody forces the issue. If compression continues long enough, participants lose interest and drift away — and there is no accumulated imbalance left to release.
Key takeaway These patterns are most valuable for what they tell you about timing, not direction. A market in a tightening triangle is a market approaching a decision on a roughly known schedule — which is exactly the situation to set a Market Watcher alert on both boundaries and stop watching.
Trading the break
Everything from breakouts and false breakouts applies directly, because that is what this is.
Wait for a close beyond the boundary, not a wick. Prefer breaks in the direction of the higher-timeframe trend. Expect false breaks, particularly in symmetrical triangles where both boundaries are heavily watched. Consider the retest for a tighter stop and better ratio.
One addition specific to compression patterns: because the range has narrowed, the stop can sit relatively close — often the opposite boundary. That produces a favourable position size, which is part of why these setups are popular. But it also means the stop sits inside a well-watched area, so a buffer matters.
Next: Measuring moves from patterns — how to turn any of these into a target.
- An ascending triangle has a horizontal resistance level and a rising support line.
- A descending triangle has a horizontal support level and a falling resistance line.
- A symmetrical triangle converges from both sides and carries no directional bias on its own.
- Compression patterns precede expansion, but do not reliably predict the direction of it.