RADEX MARKETS Editorial Team
The RADEX MARKETS editorial team consists of seasoned financial professionals and market observers. Dedicated to delivering objective market summaries, macroeconomic insights, and educational content, the team strives to keep traders well-informed in a fast-paced financial environment.
ACTIONS
The shape looked right. You drew the neckline, waited for what seemed like the break, took the entry, and price turned straight back through your level. What was missing was never the ability to spot the pattern. It was the absence of rules for what counts as confirmation, where the stop belongs, and how far the target sits.
This guide covers how to recognise 15 of the patterns that appear most often on forex charts, one trading process that applies to every one of them, and what to do when a pattern fails. Failure is part of the job here, not an afterthought.
Chart patterns are recurring structures that price forms on a chart, and they fall into three groups: reversal, continuation and bilateral. Each group tells you whether the current trend is likely to keep running or turn. A pattern on its own is not a reason to enter, though, because the breakout still has to be confirmed, and it is the height of the pattern that sets both the target and the stop.
Open your trading account today with Radex Markets.
A chart pattern is a recognisable structure that price movement forms on a chart, reflecting the tug of war between buyers and sellers across a particular price area. These shapes repeat because the behaviour behind them repeats: traders hesitate at the same kinds of levels, take profit in the same kinds of places, and give up at roughly the same point. A pattern is a record of crowd behaviour, not a forecasting tool.
That distinction runs through the whole of this guide. The pattern tells you what might happen; the confirmation conditions tell you whether to act. Treat those two as the same thing and you will spend a lot of money finding out they are not.
Every chart pattern belongs to one of three groups, defined by what it implies about the trend already in place, and that grouping is the first filter you apply when reading a chart. The table below sets the three side by side.
Category |
What it suggests |
Example patterns |
What to wait for |
|---|---|---|---|
| Reversal | Current trend is losing control | Head and shoulders, double top, double bottom, cup and handle | Break of the neckline or support level |
| Continuation | Trend is pausing, not ending | Bull flag, bear flag, pennant, rectangle | Break in the direction of the prior trend |
| Bilateral | Direction undecided until price picks one | Symmetrical triangle, rising wedge, falling wedge | Break of either boundary, with volume |
A bilateral pattern deserves particular care, because the temptation is to decide which way it will resolve while it is still forming. It has not chosen yet, and neither should you. The group a pattern belongs to also decides which side of the structure your stop sits on: reversal patterns put it beyond the extreme of the pattern, continuation patterns put it on the far side of the consolidation.
Each of the fifteen sections below covers one pattern. Jump straight to the one you want from the list here.
The table below gathers all fifteen onto a single screen, so you can compare where the signal sits and how each one tends to go wrong without scrolling through every section.
Pattern |
Category |
Signal level |
Target calculation |
Most common failure |
|---|---|---|---|---|
| Head and shoulders | Reversal | Neckline break | Head to neckline, projected down | Price recovers above the neckline |
| Inverse head and shoulders | Reversal | Neckline break | Head to neckline, projected up | Break lacks follow-through |
| Double top | Reversal | Break of middle low | Peak to middle low, projected down | Second peak exceeds the first |
| Double bottom | Reversal | Break of middle high | Trough to middle high, projected up | Lows too close together in time |
| Rising wedge | Bilateral | Break of lower boundary | Widest point, projected down | Breaks upward in a strong uptrend |
| Falling wedge | Bilateral | Break of upper boundary | Widest point, projected up | Mistaken for a falling channel |
| Bull flag | Continuation | Break of flag upper edge | Flagpole length, projected up | Flag retraces more than half the pole |
| Bear flag | Continuation | Break of flag lower edge | Flagpole length, projected down | Consolidation drifts sideways too long |
| Pennant | Continuation | Break in trend direction | Flagpole length, projected in trend direction | Consolidation extends beyond the usual window |
| Ascending triangle | Continuation | Break of horizontal resistance | Widest height, projected up | Repeated failed tests weaken the level |
| Descending triangle | Continuation | Break of horizontal support | Widest height, projected down | Support holds and price breaks upward |
| Symmetrical triangle | Bilateral | Break of either boundary | Widest height, projected in break direction | Direction guessed before the break |
| Cup and handle | Reversal | Break of handle high | Cup depth, projected up | Handle retraces too deeply |
| Rectangle | Continuation | Break of either boundary | Range height, projected in break direction | Range breaks against the prior trend |
| Diamond | Reversal | Break of lower boundary | Widest height, projected down | Structure misread while still forming |
Three peaks form in sequence with the middle one standing highest, and a neckline joins the two lows that sit between them. The signal is a close below that neckline, and the target is the vertical distance from the top of the head down to the neckline, projected downward from the break.
The way this one most often goes wrong is that price slices through the neckline, everyone piles in short, and then price climbs straight back above it within a couple of candles. A neckline reclaimed that quickly is not a delayed signal. It is a cancelled one.
Read more: Head and shoulders pattern trading strategy for beginners
This is the mirror image, appearing at the end of a downtrend: three troughs with the middle one deepest, and a neckline across the two highs between them. The signal is an upward close through the neckline, and the target is measured the same way, projected up from the break.
What separates it from a double bottom is the middle low. In an inverse head and shoulders that middle trough sits clearly lower than the two beside it; in a double bottom the two lows are roughly level. Get that wrong and you will measure the target from the wrong point.
Two peaks form at a similar height with one low between them, and the signal is a close below that middle low. The target is the distance from the peaks down to the middle low, projected down from the break.
If the second peak is clearly higher than the first, the pattern does not stand. That is not a variation on a double top; it is an uptrend making a higher high, and trading it as a reversal means shorting into strength.
The mirror image of the double top: two troughs at a similar depth with one high between them, forming the W shape most traders recognise on sight. The signal is a close above that middle high, and the target is the distance from the lows up to it, projected upward.
The two lows need enough time between them to mean anything. Two troughs three candles apart are noise inside a single move, not a market testing a floor twice and failing to break it.
Both boundaries rise, but they converge as they go, so the structure narrows while it climbs. It looks bullish and usually resolves bearishly, which is exactly why it catches people out. The signal is a close below the lower boundary, and the target is the widest part of the wedge projected down.
The convergence is the tell. Each new high is arriving with less room above it than the last, which is what fading buying momentum looks like drawn on a chart.
The mirror image: both boundaries fall while converging, and it usually resolves bullishly. The signal is a close above the upper boundary, with the target taken from the widest point and projected up.
The thing to check is whether the boundaries actually converge. A falling channel has parallel boundaries and carries no such implication, and the two look similar enough at a glance that traders regularly treat one as the other.
A sharp advance forms the pole, then a small counter-trend consolidation drifts down or sideways to form the flag. The signal is a close above the flag's upper edge, and the target is the length of the pole projected from the breakout point.
Quality drops sharply once the flag retraces more than half the pole. At that point the move you were planning to join has given back most of its ground, and what you are looking at has more in common with a reversal than a pause.
Read more: Bull flag pattern: A practical guide for forex traders
The mirror image, appearing after a sharp decline: a steep drop forms the pole, then a shallow upward drift forms the flag. The signal is a close below the flag's lower edge, with the pole length projected down from there.
Watch how long the consolidation runs. A bear flag that drifts sideways for a long stretch stops being a pause in a downtrend and starts being a base.
A pennant differs from a flag in the shape of the consolidation: a small converging triangle rather than a parallel channel. The signal is again a close in the direction of the prior trend, and the target is the pole length projected that way.
Pennant consolidations are usually shorter than a flag's. Once the pause stretches well beyond the handful of candles you would normally expect, the momentum that created the pole has probably gone.
A horizontal resistance level sits above a series of rising lows, so the structure squeezes upward into the ceiling. The signal is a close above that horizontal level, and the target is the height at the widest part of the triangle projected up from the break.
Rising lows are the point of this pattern. Buyers are willing to pay progressively higher prices to get in, which is a different market from one merely bouncing off a floor.
The mirror image: horizontal support below a series of falling highs, with sellers accepting progressively lower prices. The signal is a close below that support, and the target is the widest height projected down.
If support holds and price breaks upward instead, the read was wrong from the start, which happens often enough that the upside break deserves its own plan rather than surprise.
Both boundaries converge at once, neither of them horizontal, and nothing in the structure implies a direction. It sits squarely in the bilateral group, and the signal is a close beyond whichever boundary gives way first.
Calling the direction early on this pattern is one of the more reliable ways to lose money with chart analysis. The whole information content of a symmetrical triangle is that the market has not decided, so an entry placed before the break is a guess wearing a technical costume.
A rounded base forms the cup, then a small pullback on the right-hand side forms the handle. The signal is a close above the handle's high, and the target is the depth of the cup projected up from the break.
A handle that retraces too deeply no longer counts as a handle. The handle is supposed to be a shallow shake-out near the rim, and once it drops back toward the base of the cup the structure has lost the thing that made it worth watching.
Price oscillates repeatedly between a horizontal upper boundary and a horizontal lower one, touching each several times. It belongs in the continuation group, and the signal is a close beyond a boundary in the direction of the prior trend.
A rectangle that breaks against the prior trend is still a tradable break, but it is no longer a continuation pattern, and the measured move should be projected from the boundary that actually gave way rather than the one you expected to.
The structure broadens first and then converges, producing the diamond outline, and it turns up most often at a trend top. The signal is a close below the lower boundary, with the widest height projected down.
This is the hardest of the fifteen to identify honestly. It occurs rarely, the broadening phase looks like disorder while it is happening, and a diamond is far easier to name in hindsight than to trade in real time.
These three steps apply to all fifteen patterns above. The pattern decides what you are looking at; these steps decide whether you act on it. This section matters more than any individual pattern section, because a trader who knows three patterns and applies these steps will do better than one who can name all fifteen and applies none of them.
Confirmation has three conditions. Wait for a candle to close beyond the boundary rather than acting on an intraday touch, because a wick through a level and a close beyond it are different events. Watch whether activity expands on the break. Then check whether the boundary holds when price comes back to retest it.
There is a limitation here that most pattern tutorials skip, because most of them are written from a stock market background. Forex has no central exchange, so there is no true volume figure to read. What your platform shows is tick volume, which counts how many times the price updated in a period. It works as a proxy for activity, and it is not the same as executed volume. Treat a volume spike on a forex chart as evidence of attention, not of committed capital.
The stop level is decided by the structure of the pattern, not by the amount you are willing to lose. For a reversal pattern it goes beyond the extreme point of the structure: past the head, past the second peak, past the cup's base. For a continuation pattern it goes on the far side of the consolidation, beyond the flag or the pennant.
The amount you are willing to lose sets a different variable, which is position size. Structure decides where the stop goes, then risk decides how big the trade is. Doing it the other way round produces stops placed at convenient round numbers that the market has no reason to respect.
Read more: 10 Forex risk management tips every trader should know
The measured move is the target method every one of these patterns shares. Take the vertical height at the widest part of the pattern, then project that same distance from the breakout point in the direction of the break. One rule, fifteen patterns.
Worked GBP/USD Example
Take an ascending triangle forming on the GBP/USD four-hour chart. Resistance sits flat at 1.2750, tested three times, while the lows underneath climb from 1.2640 toward it. The widest height of the triangle is therefore 1.2750 − 1.2640 = 110 pips.
A four-hour candle closes above resistance and the entry fills at 1.2755. Projecting the pattern height from the break gives a target of 1.2755 + 110 pips = 1.2865. The most recent higher low sits at 1.2700, so the stop goes just beneath it at 1.2695 — a risk of 60 pips.
- Pair and time frame: GBP/USD, four-hour
- Pattern height: 110 pips
- Entry: 1.2755
- Target: 1.2865
- Stop: 1.2695
- Risk: 60 pips | Reward: 110 pips
At 0.20 lots on GBP/USD, each pip is worth roughly $2, so the trade risks $120 to make $220. That is a risk-to-reward ratio of 1.83:1. The figures are illustrative rather than a recommendation to trade those levels, but the sequence is the point: the height came from the chart, the target came from the height, and the position size came last.
A measured move is a target set by taking the vertical height at the widest part of a chart pattern and projecting that same distance from the breakout point in the direction of the break. It gives the target a source in the structure rather than in a round number.
A pattern has to be drawable before it can be read. Whether you can put the same neckline in the same place two days running depends less on your eye than on the tool you are drawing with, which makes the charting package a practical decision rather than a cosmetic one. Four things are worth checking.
The table below compares three common options.
Tool |
Best use |
Pattern features |
|---|---|---|
| MT4 | Familiar layout with a light footprint | Core drawing objects, custom indicators, price alerts |
| MT5 | More timeframes and object types than MT4 | Wider drawing set, built-in alerts, extended analytical objects |
| Third-party charting packages | Detailed annotation and sharing | Large drawing toolsets, automatic pattern detection on paid tiers |
Automatic detection is worth having, with one honest caveat. It can tell you that a shape matching a descending triangle exists on the chart. It cannot tell you whether that triangle is forming into a level that has already broken twice this month, or whether the break that just printed had any conviction behind it. Treat it as a way to find candidates, not a way to decide entries. Platform features also change between versions, so check the current documentation before you build a routine around one.
Read more: MetaTrader 4
You will have seen the numbers. Head and shoulders 89%, double bottom 88%, bull flag 85%, passed between sites as though they were settled facts. They are not, and the reason is worth understanding, because it changes how you read every reliability claim you meet afterwards.
Almost none of those figures state a sample size, a period, a market or a time frame. None of them account for spread, slippage or commission, which is where a theoretical edge goes to die. And they contradict each other: one set ranks head and shoulders near the top, another puts horizontal channels at 68% and double tops at 66%, with no way to reconcile the two because neither shows its working.
The one body of systematic statistics with a published methodology is Thomas Bulkowski's Encyclopedia of Chart Patterns, which devotes a chapter to how the measurements were taken and whose third edition rests on roughly four times the sample of the second. It is the only source in this area we are willing to cite. It also has a limitation that has to be stated alongside it: the performance figures are calculated on price running to its ultimate high or low, which is a theoretical ceiling rather than a result any trader will achieve, because nobody exits at the exact extreme.
So rather than memorising a success rate for a pattern, remember the three conditions that actually make patterns work, which are the confirmation, the stop and the measured target covered above.
Patterns fail often enough that knowing the common causes is part of using them properly.
Read more: How to Read Forex Charts for Traders
Four honest limitations sit underneath this whole method.
Knowing these four makes you better at using patterns rather than worse, because each one points at a specific place to be more careful: draw the boundaries the same way every time, size the position for a possibility rather than a certainty, confirm the break with price behaviour as well as activity, and pick a time frame before you start looking.
There is no single most successful chart pattern, though historical studies in technical analysis frequently rank the head and shoulders, double bottom and rectangle among the more reliable and profitable ones. Rankings of that kind almost all trace back to Bulkowski's statistics, which are measured on price running to its ultimate high or low. What they give you is the relative order of the patterns, not the result you should expect to achieve on a live account.
The ten most widely used are head and shoulders, inverse head and shoulders, double top, double bottom, rising wedge, falling wedge, bull flag, bear flag, pennant and symmetrical triangle. This guide adds ascending triangle, descending triangle, cup and handle, rectangle and diamond to make fifteen, because those five turn up on forex charts just as often as the first ten.
A reversal pattern suggests the current trend is losing control, while a continuation pattern suggests the trend is pausing rather than ending. The judgement rests on where in the trend the pattern appears and whether the eventual break runs with or against the prior direction.
No. Failure is the norm rather than the exception, and false breakouts are especially common in thin liquidity and around major data releases. Every entry therefore needs a price defined in advance at which the pattern is proven wrong, which is what the stop is for.
The structures themselves carry across, but the confirmation does not. Forex has no central exchange and therefore no true volume, so what the platform shows is tick volume, and the general rule of confirming a break with volume has to be supported another way — by the size of the move on the break and by how price behaves on the retest.
Higher time frames carry less noise and produce more reliable patterns, with daily and four-hour charts being the common choices. Lower time frames produce more patterns but a higher proportion of false signals, which suits traders who already have a defined trading plan rather than those still building one.
We have covered the three categories every pattern falls into, how to recognise fifteen of them and the way each one typically fails, the three-step process that applies to all fifteen, and how to read the reliability claims you will meet elsewhere.
Three criteria are worth taking to your chart today:
One closing view, offered as opinion rather than fact. The patterns are the easy part of this. Anyone can learn to spot a head and shoulders in an afternoon, and the fact that fifteen of them fit into one article should tell you how little of the work sits in recognition. What separates traders who use patterns well is that they decided in advance what would count as confirmation and where the idea would be proven wrong, and then did not renegotiate either of those once money was on the line.
The RADEX MARKETS editorial team consists of seasoned financial professionals and market observers. Dedicated to delivering objective market summaries, macroeconomic insights, and educational content, the team strives to keep traders well-informed in a fast-paced financial environment.
Reviewed by RADEX MARKETS
Risk Warning
Trading derivatives and leveraged products carries a high level of risk, including the risk of losing substantially more than your initial investment.
Deeper into chart patterns:
Related concepts:
Le trading de produits dérivés et à effet de levier comporte un risque élevé et peut entraîner des pertes supérieures à votre investissement initial. Veuillez lire nos Conditions générales (T&C) ainsi que le Product Disclosure Statement (PDS) avant de trader.
CRÉER UN COMPTE