Foundation
Chart patterns
Chart patterns are names for how swings arrange themselves. Use them to describe structure and to plan a break or a failure. Do not use them as a forecast the market owes you.
On this page
This page is the photo atlas, drawn on purpose. Every figure is a schematic. The numbers are not a ticker. The captions are the lesson. If you only look at the shape and skip the sentence under it, you will leave with the superstition this page is trying to retire.
Section 01
Structure first
An uptrend is a sequence, not a feeling
Each pullback ends higher than the pullback before it, and each push makes a higher high. That is the whole definition of an uptrend in swing structure. A downtrend is the mirror: lower highs and lower lows. When the sequence breaks, the structure has changed. It has not issued a forecast.
Mark swings the same way every time. One workable rule: a swing high is a bar whose high is higher than the bars on either side of it, and you ignore wiggles smaller than a fraction of the recent average range. The exact rule matters less than freezing it. If you mark swings differently after you know where price went, your patterns will look brilliant and teach you nothing.
Section 02
Support and resistance
Areas, not a magic price
Price stalls more than once near the upper line and more than once near the lower line. Traders call the upper area resistance, or supply, and the lower area support, or demand. The lines on this drawing are rulers for the eye. Real trade prints across a band, not on a single tick. After a break, the old ceiling sometimes acts as a floor. Sometimes it does not.
A level becomes more interesting when several reactions stall in the same area, when it lines up with a prior break, or when it is obvious enough that many people will have orders there. Obvious does not mean reliable. It means crowded. Crowded places produce both clean turns and violent breaks, because the orders stacked on the idea become fuel when the idea fails.
Section 03
Head and shoulders
Head and shoulders, top
Three peaks. The middle peak, the head, is the highest. The two shoulders are lower and roughly similar. The neckline joins the reaction lows between them. A break is a close through that neckline, not a wick that pokes it. Technicians often measure the distance from the head to the neckline and project it down from the break as a planning target. Price does not owe you that distance. The inverse pattern, a head and shoulders bottom, is this picture flipped.
- Top: an advance, a higher peak, a lower peak, then a break of the neckline. It is a reversal pattern only if there was a trend to reverse.
- Bottom: the same shape after a decline, often called an inverse head and shoulders. The neckline break is upward.
- Shoulders do not need to match. They need to be visibly lower than the head, and the whole structure needs enough bars to be more than a squiggle.
- Volume is often described as heavier on the left shoulder than the right. That is part of the classic commentary. Do not invent volume you cannot see. In spot foreign exchange, a retail platform's volume is often tick counts from that venue, not the world's trade.
The measured move takes the height from the head to the neckline and projects it from the break. Use it to ask a planning question: if the stop belongs back inside the pattern, is the projected distance large enough, after costs, to justify the risk? If the answer is no, you skip it. You do not move the stop closer to make the ratio pretty.
Section 04
Double top and double bottom
Double top
Two peaks form at a similar price, with a valley between them. The pattern is confirmed, in the usual definition, only when price closes through the valley. Peaks do not need to match to the tick. If the two highs are minutes apart on a chart you meant to read as a daily pattern, you are looking at noise, not a base. A double bottom is the same shape upside down.
Confirmation is the close through the valley, for a top, or through the peak between two lows, for a bottom. Buying the second peak because it "looks double" is not the pattern. It is an early guess. Time between the two tests matters. Two highs in the same impulse are one swing. Two highs separated by a real reaction can be a pattern. Write the minimum reaction in your own rules.
A triple top or triple bottom is the same idea with three tests. It is less common. It is not more magical. Each extra test is another chance for the level to hold or to fail.
Section 05
Triangles
Ascending triangle
Highs stall near the same ceiling while the lows climb. The drawing breaks upward. That is the textbook bias, not a requirement. A close through the rising line would have been a failed pattern, and failed patterns are part of the distribution you must expect if you trade breaks.
Descending triangle
Lows keep finding a floor while the highs step down. The textbook lean is a break lower. Treat the lean as a hypothesis. Your rule still has to say what counts as a break, where you are wrong, and what you do when price springs back inside the coil.
Symmetrical triangle
Highs fall and lows rise. The shape alone does not choose a direction. This drawing happens to break upward so you can see a completed example. A downward break would have been just as consistent with the coil.
A triangle is compression. Ranges get smaller. That can precede expansion. Compression does not pick a direction unless your rule adds something else, such as the trend that led into an ascending or descending triangle. Even then, you are stacking a tendency on top of a tendency. The trade, if you take one, is the break you defined, with a loss defined if price re-enters the coil.
Section 06
Flags and pennants
Pole, then a flag
A sharp move is the pole. The flag is the small drift against that move, usually on quieter trade. A pennant is the same idea with a tiny triangle instead of a channel. The continuation is the part people remember. The failure, a flag that rolls over and eats the pole, is the part that has to be in the plan.
The pole should be obvious without squinting. If you have to argue that a drift was a pole, it was not. Flags that slope against the pole are the tidy version. A flag that slopes with the pole is often just the trend continuing, and the "break" is a late entry. Pennants are small symmetrical triangles on the pole. They inherit the same warning: the failure of the continuation is a real outcome, and it often travels fast because late entries are crowded.
Section 07
Wedges
Rising wedge
Both the highs and the lows are rising, and the two boundaries converge. The pushes are getting less impressive relative to the space they use. The common reading is bearish, and this drawing breaks down. A falling wedge is the mirror and is commonly read as bullish. Commonly read is not the same as reliable. Write the break rule before you need it.
Do not confuse a wedge with a triangle. In a rising wedge both boundaries rise. In an ascending triangle the top boundary is flat. The names are not interchangeable, and a chart guru's arrow does not make them so. A falling wedge has two falling, converging boundaries and is commonly read as a bullish compression. Test the reading. Do not marry it.
Section 08
Cup and handle
Cup and handle
William O'Neil described a rounded cup, then a shallower pullback near the highs, the handle, then a breakout. He wanted the handle relatively tight and preferred heavier volume on the break. That is his specification, from his stock-selection work. It is not a shape the market is obliged to finish. If you change the depth and the duration until every dip looks like a cup, you no longer have his pattern.
O'Neil's cup is typically a longer base, on the order of many weeks in his stock work, not a shape you force onto a five-minute chart because the curve rhymes. The handle should be shallower than the cup. A handle that cuts deep into the cup is a different, weaker structure under his rules. Volume on the break was part of his confirmation. If your market does not have trustworthy volume, do not pretend a broker's tick count is the New York tape.
Section 09
Ranges, breaks, and failures
A range, then a close outside it
Price oscillates between two areas and then closes outside. A breakout system buys or sells that close. A mean-reversion system does the opposite and fades the edge of the range. Both people are looking at this picture. They cannot both be trading a complete plan unless each one has defined the failure. A wick through the edge, followed by a close back inside, is the usual picture of a failed break.
A poke is not a break
Price pushes through a prior high and comes back. If your rule requires a close beyond the level, this poke never qualified. If your rule buys the poke, this is a failed break and the stop has a job to do. Decide which rule you use before the bar prints. Changing the definition after you see the result is how a journal learns nothing.
Pick one job for a range before you click.
Breakout
What it is. Enter when a close leaves the range, in that direction.
Watch for. Many breaks fail and snap back. A string of small losses is the normal cost.
Retest
What it is. Wait for a break, then for price to return to the broken area and hold.
Watch for. The best moves never retest. You will miss them. That is the fee for a tighter entry.
Fade
What it is. Bet that a push to the edge returns toward the middle.
Watch for. The one time it does not fade is the trend that pays the other camp. You need a hard exit.
Section 10
Before you call it a pattern
- Can you state the definition without pointing at this one chart?
- Is there a prior trend if the pattern claims to reverse one?
- What printed event will count as confirmation? A close, a percent, a number of ticks?
- Where is the idea wrong? If that price is absurdly far, the pattern is not tradable at your size. Skip it.
- What would a failed pattern look like, and will you exit when you see it, or will you rename it?
Check yourself
When is a double top usually considered confirmed?