A triangle forms when price makes lower highs and higher lows at the same time. The range narrows until the two converge.
What it records is straightforward: volatility is contracting. Buyers and sellers are agreeing on a narrower and narrower price. Neither side is winning.
What it does not record is which way it will break — and this is where most triangle teaching goes wrong.
The three types, and the honest version
Type
Shape
Conventional claim
Ascending
Flat top, rising lows
Bullish
Descending
Flat bottom, falling highs
Bearish
Symmetrical
Both converging
Continuation
The directional claims are weak. The evidence that an ascending triangle predicts an upward break is thin, and in practice they break both ways often enough that trading the shape alone is close to a coin flip with extra steps.
What is reliable is the compression itself. A narrowing range means an expansion is coming, because volatility mean-reverts. That is a genuine, tradeable observation — it just says *something will happen*, not *what*.
Trade the break, not the shape
Since the direction is not predictable from the triangle, the method follows:
Mark both boundaries. They are the levels; the triangle is just what they look like together.
Wait for a close beyond one, with displacement — a decisive body, not a probe.
Enter on that close or its retest.
Stop beyond the opposite boundary, or beyond the last swing inside the triangle.
Direction bias comes from the higher timeframe, not from the triangle's type.
Point 5 is the useful one. An ascending triangle inside a daily downtrend is not bullish. It is a pause in a downtrend that happens to have a flat top.
The apex problem
As the triangle narrows, two things happen and both are bad:
False breaks multiply. With the boundaries close together, small moves cross them. Late in a triangle you get repeated breaks that immediately fail.
The stop gets tight, which is a trap. A narrow triangle offers a very tight stop, which looks attractive. But the tight stop is inside the noise the compression is about to release. You are taking the smallest stop at exactly the moment of the largest impending move.
Practical rule: trade breaks from the first two-thirds of the triangle. Once price is near the apex, stand aside and wait for the expansion to resolve, then trade the retest of whichever boundary broke.
A symmetrical triangle. Range has compressed from 74 points to 26.
Daily context: uptrend intact, last swing low 3,381. Bias: long.
Price closes at 3,455 — above the descending upper boundary, on a candle with a 31-point body and small wicks. Displacement. It agrees with the daily.
Entry 3,455. Stop 3,404 — below the last swing low *inside* the triangle, not just below the upper boundary. Risk 51 points, so size reduced accordingly.
Target: the widest part of the triangle (74 points) projected from the break → 3,529. Reward 74. Ratio about 1:1.45.
Marginal. The trader takes it at reduced size with a first target at the prior high of 3,489 (1:0.67, partial) and the remainder to 3,529.
Now the version that loses money. Two days earlier, with the range at 26 points, price closed at 3,440 — barely above the boundary, on a 9-point body. A trader entering there with a 14-point stop below the boundary would have been stopped within four hours as price fell back to 3,418, then watched the real break happen without them.
Same triangle. Same eventual direction. One entry had displacement and a structural stop; the other had a tight stop near the apex and no confirmation.
What to take away
A triangle records volatility compression, not direction.
The bullish/bearish labels for ascending and descending triangles are weak — use the higher timeframe for bias.
Require displacement on the break, not a marginal close.
Stop beyond the last swing inside the triangle, not just beyond the boundary.
Avoid trading near the apex. The tight stop there is the trap, not the opportunity.
Next: the patterns that appear inside a move rather than at the end of one.
