Direct answer
A “triangle” in forex refers to a recurring chart geometry where price swings create two lines that converge—one representing repeated high points and the other representing repeated low points. The idea is not a guaranteed forecast; it is a way to describe structure that you can measure and verify on your own chart using consistent rules.
To explain how it “works,” it helps to separate (1) stable mechanics—what you are measuring on the chart—from (2) variable conditions—how market behavior, liquidity, spreads, and order execution can change outcomes.
What “triangles” means on a price chart
A triangle is typically defined by two sets of swing points:
- A “top line” drawn through repeated swing highs.
- A “bottom line” drawn through repeated swing lows.
If the lines converge, the chart is forming a narrowing range. In practice, people may label variants such as contracting triangles, ascending triangles, or descending triangles, but the common measurement step is the same: identify multiple touches that define two boundaries, then check whether the boundaries converge.
Key terms used for verification:
- Swing high/low: a local peak or trough on the chart.
- Boundary line: a line fitted through selected swing points.
- Convergence: the distance between the boundary lines shrinks over time.
- Breakout (or rejection): a price move that exits a boundary or turns back inside it.
Simple model: inputs → outputs → sequence
Inputs you choose
Because different charting platforms can show different candles and different zoom levels, your inputs matter. A self-contained triangle check usually specifies:
- Chart timeframe (example assumption: using a 1-hour chart for the identification step).
- How swing points are selected (example assumption: use the most recent three to five clear swing highs and lows that form the narrowing structure).
- Line-fitting rule (example assumption: draw the top line to touch highs, and the bottom line to touch lows, using the same rule throughout the pattern).
- What counts as a “touch” (example assumption: the candle body or wick touches within a small tolerance, like “near the line,” though your tolerance definition must be consistent).
Outputs you can measure
From those inputs, the pattern outputs are:
- Geometry: whether the top and bottom boundaries converge.
- Levels: the approximate boundary prices at different times (the top boundary price and bottom boundary price).
- Potential exit zone: the region near the point where boundaries meet (often called the “apex” in chart language).
- Invalidation/continuation check: whether subsequent price action respects the boundaries or exits them.
Sequence you can run on historical data
A practical sequence for independent verification looks like this:
- Locate candidate structure: scan for a phase where highs and lows alternate and the range narrows.
- Draw boundaries using fixed rules: apply the same swing selection and line-fitting rules every time.
- Confirm convergence: check that the boundary gap decreases rather than just trending.
- Track subsequent behavior: observe whether price repeatedly “interacts” with the boundaries.
- Define your decision criteria in advance: for example, record whether price closes outside a boundary, or whether it rejects the apex region.
Even with a careful sequence, you should treat the result as an observation, not a prediction.
Evidence or example (with explicit assumptions)
Below is a generic example you can replicate without needing live prices.
Assumptions:
- You use a single timeframe (for example, 1-hour candles).
- You fit two lines to three swing highs (top line) and three swing lows (bottom line).
- You label the triangle when the boundary gap decreases across those touches.
Example observation:
- Over multiple sessions, the highest candle wicks rise less than earlier peaks, so your top boundary slopes downward or flattens depending on the variant you label.
- Simultaneously, the lowest candle wicks fall less than earlier troughs, so your bottom boundary slopes upward or flattens.
- The chart becomes a tightening range with alternating reactions at the two boundaries.
What you output:
- The top and bottom boundary lines you drew.
- The apex region where the lines would meet (an approximate time and price area).
- A record of what happened next (for example: did price exit the range and hold outside, or did it move back inside?).
Material limitation revealed by the example: Two different analysts can draw different swing points, which can shift the boundary lines. This can change whether price “breaks out” under your chosen criteria.
Limitations and risks (including failure modes)
1) Pattern ambiguity
A triangle label is a description of geometry, not a unique identifier. Different selection rules for swing points can produce different triangles from the same chart.
2) False breakouts and “chop”
Price can exit a boundary briefly and then return inside the triangle. Even if a triangle is geometrically valid, follow-through is not guaranteed.
3) Costs and execution effects
Forex trading involves transaction costs and execution conditions. A move that appears meaningful on a chart may be less meaningful once spreads, commissions, and order execution timing are considered.
4) Timeframe dependence
Triangles may appear on one timeframe but not another. The same underlying market activity can produce different structures depending on how you aggregate candles.
5) Jurisdiction and provider differences (verification reality)
Markets are influenced by venue infrastructure, and providers may differ in how instruments are quoted or executed. If you compare two accounts or feeds, the visible chart interaction with your drawn lines might differ.
How to verify facts and decide what to learn next
To independently verify what “triangles work like,” focus on measurable facts you can check:
- Can you reproduce the same convergence using the same swing-selection and line-fitting rules?
- What proportion of your historical samples show the behavior you recorded after the apex region?
- How sensitive is your conclusion to changing the timeframe or the number of swing points you include?
A useful next question is not “Will it work?” but: “Under my exact chart rules, what outcomes do I observe historically, and what are the main failure cases?”
If you want, share your triangle definition in plain terms (timeframe, how you choose swing points, and what counts as a breakout), and I can help you turn it into a checkable observation workflow without treating it as a signal.