Cup and Handle Pattern

What the pattern is

The cup and handle pattern is one of the classic chart patterns. It has four parts.

• The left rim. A prior high where price starts to fall.
• The cup. A fall to a low, then a gradual rise back toward the rim. The shape is meant to be rounded, like a bowl or a U. A sharp V is usually not counted.
• The right rim. The point where the recovery reaches roughly the level of the left rim.
• The handle. A short, shallow pullback after the right rim. It is meant to be smaller and shorter than the cup.

The breakout is a move above the rim level after the handle. Chart readers see this as the pattern completing. The story behind it is that holders who were stuck at the old high have sold or given up during the cup, so there is less supply left when price returns to it. That is an explanation people give. It is not proven here, and it does not mean the pattern works.

The rim, the handle and the breakout level

The two rims rarely match exactly, so a rule needs to say how close is close enough. Most people take the breakout level as the higher of the two rims. In the made-up path, the left rim is 100.0 and the right rim is 98.5, a gap of 1.50 percent of the left rim. The breakout level is 100.0.

The handle is the part that makes the pattern look different from a plain recovery. It is a pause just under the rim. Here it falls from 98.5 to 95.5, a drop of 3.0 points. That is 15.0 percent of the cup depth, and it lasts 3 bars against 13 bars for the cup. Many descriptions say the handle should give back no more than about a third to a half of the cup. Sources vary, so treat any such number as a setting, not a fact.

The usual measured target

A measured move projects the size of the pattern past the breakout. For this pattern, the common version is the cup depth added to the breakout level.

It is a way to mark where a move of the same size would end. Price may stop short of it, pass it, or never break out at all. The target is arithmetic on a drawing.

A worked example with made-up numbers

Take the made-up closes in the picture. Bar 1 closes at 100.0. The low close is 80.0 on bar 8. The right rim is bar 14 at 98.5. The handle low is 95.5 on bar 17. Bar 20 closes at 99.0. Suppose, for the arithmetic only, that a trader treated a close of 99.0 as the breakout, entered there, and put a stop at the handle low. The rows below show the numbers.

All numbers are made up, calculated by us on 31 Oct 2026, exact. The ratio of reward to risk here is 21.0 to 3.5, or 6.00 to 1. That ratio only holds if the target is reached, and it ignores spread, fees and slippage. See the risk reward ratio page for why a ratio alone says little.

One more point. The breakout level here is 100.0, set by the left rim. The close on bar 20 is 99.0, which is under that line, so a strict rule that waits for a close above 100.0 would not have fired yet. A looser rule that uses the right rim of 98.5 would have. The same chart gives a different answer depending on the definition.

Why it is subjective

The pattern is easy to see after the fact and hard to define before. Each of these is a judgment.

• Which high is the left rim? A wick or a close, the last high or a bigger one?
• How round is round? No standard test separates a cup from a V or a wide base with a flat floor.
• How deep and how long? A cup might last a few weeks or many months. Depth ranges are a matter of opinion.
• How big may the handle be? The limit is a setting. So is the number of bars.
• What counts as the breakout? A touch, a close, a close with extra volume, or a close that holds for several bars?

Because there are so many choices, a person looking back through charts can usually find some set of choices that makes a given shape qualify. That is the hindsight problem behind overfitting.

How to define it in code with tolerances

A bot cannot look at a picture. It needs a list of numeric tests with limits. A simple version has five steps.

• Find the pivots. A pivot high is the highest close with M lower closes on each side. Use only bars that have already closed. A pivot is only known M bars later, and using it earlier is look-ahead bias.
• Rim test. The two rim closes must be within a tolerance, for example 2 percent of the left rim.
• Cup test. Depth as a share of the left rim must sit between a minimum and a maximum. A roundness check might require the bottom to be roughly in the middle of the cup rather than at either end.
• Handle test. The handle drop must be positive, at most some share of the cup depth, and last no more than some share of the cup length.
• Breakout test. A close above the higher rim.

Each limit is a number someone chose. Here is the effect on the made-up path. The measurements do not change: the rim gap is 1.50 percent, the cup depth is 20.0 percent of the left rim and the handle gives back 15.0 percent of the cup. Only the limits do.

Made-up path, calculated by us on 31 Oct 2026, exact. One chart is a pass under one setting and a fail under three others. A backtest of this rule would trade different bars for each setting, and each setting can be defended. That is why one good backtest says little. The sensitivity to settings is the first thing to measure. See how to backtest and walk-forward testing.

How the pattern fails

• No breakout. Price stalls under the rim, rolls over and never closes above it. The pattern then never completes.
• A false breakout. Price closes above the rim, then falls back below it within a few bars. A stop at the handle low takes the loss of 3.5 points in this example. See false breakout.
• The handle deepens. If the handle gives back more than the limit, the shape turns into a second leg down or a double top, and the label no longer fits. See also double bottom for a related shape.
• Target not reached. The measured target is arithmetic. Price can turn around well before it, and a move that stalls halfway never reaches 120.0.
• Costs and gaps. A breakout order can fill worse than the trigger when price jumps. That is slippage, and it eats into a 21.0-point reward.
• Too few trades. A strict definition finds a handful of patterns across years of data. A handful of trades cannot separate skill from luck.

Where it fits with other pages

A cup and handle breakout is a form of breakout trading: a trade placed when price leaves a range or passes a marked level. It shares the same risks, such as false breaks and costs at the trigger. The rim is also a horizontal line of the kind covered under support and resistance.

What this does not tell you

It does not say how often the pattern appears, how often it completes, or whether trading it earns anything after costs. Every price and result on this page is made up to show the arithmetic and the problem of definition. This is education, not a signal.