A cup and handle is a bullish continuation chart pattern. After an uptrend, price makes a rounded, U-shaped dip (the cup) and recovers to near its old high. It then drifts lower for a short time (the handle), and the pattern confirms when price breaks above the handle and the cup's right rim. Traders who use the pattern usually enter on that breakout, put a stop at the handle low, and set a target by projecting the cup's depth upward from the breakout. William O'Neil introduced the pattern in his 1988 book How to Make Money in Stocks. The most widely cited statistics come from Thomas Bulkowski, who ranks it 3rd of 39 chart patterns for performance in bull markets.
What the pattern looks like
The pattern has four parts:
- A prior uptrend. StockCharts' ChartSchool says a prior trend should exist, ideally a few months old, and that a mature trend leaves less upside.
- The cup. A rounded, U-shaped decline and recovery. ChartSchool says a sharp V-shaped bottom is too abrupt to qualify. Equal highs on both sides are ideal but not required.
- The handle. A short pullback or gentle downward drift that forms after price returns to the cup's rim. Bulkowski requires it to sit in the upper half of the cup.
- The breakout. Price closes above resistance: the handle's high, or the cup's right rim.
The idea behind the shape is that sellers are gradually absorbed during the cup. The handle shakes out holders who want to sell at the old high before price moves higher. This is an interpretation, not a measured fact.
Identification rules from two references
The two most-used references give slightly different numbers. They are guidelines, not hard limits.
| Element | StockCharts ChartSchool | Bulkowski (thepatternsite.com) |
|---|---|---|
| Cup duration | 1 to 6 months, sometimes longer on weekly charts | 7 to 65 weeks |
| Cup depth | Ideally 1/3 or less of the prior advance; 1/3 to 1/2 in volatile markets; 2/3 maximum in extreme cases | No depth rule given |
| Cup shape | U-shaped, not V-shaped | U-shaped, not V-shaped, with some variation allowed |
| Rims | Equal highs ideal, not required | Near the same price, flexible |
| Handle duration | One week to several weeks, ideally 1 to 4 weeks | At least 1 week, no maximum |
| Handle depth/position | Can retrace up to 1/3 of the cup's advance, usually less; a smaller pullback is more bullish | Must form in the upper half of the cup |
| Volume | Should expand substantially on the breakout | No volume rule given |
ChartSchool adds that capturing the essence of the pattern matters more than strict measurements.
Bulkowski reports one detail on handle length: handles shorter than the median of 22 days showed better performance after the breakout.
Setups that fail the checklist
- V-shaped bottom. The recovery is too fast to count as a cup under either reference.
- Handle in the lower half of the cup. This breaks Bulkowski's upper-half rule.
- Handle deeper than 1/3 of the cup's advance. This breaks ChartSchool's guideline.
- No prior uptrend. The pattern is a continuation pattern, so with no trend there is nothing to continue.
- Breakout without a volume expansion. ChartSchool treats rising volume as part of the breakout.
Entry, stop and target
These are the mechanics as the references describe them. They are not recommendations.
Entry
- Standard (Bulkowski): buy when price closes above the right cup rim.
- Early (Bulkowski): draw a down-sloping trendline across the handle's peaks. A close above that line is an early buy signal.
- Close confirmation (ChartSchool): some traders wait for a close above the resistance line set by the cup's highs, not an intraday break.
Bulkowski also notes that cups often form inside larger cups, so an inner cup can be traded when price rises above its handle.
Throwbacks
A throwback is when price breaks out upward, climbs, and then returns to the breakout price within 30 days (Bulkowski's glossary definition). Bulkowski reports a 62% throwback rate for this pattern in bull markets and says throwbacks hurt performance. Some traders treat the retest as a second entry. Bulkowski's data does not show whether that improves results.
Stop
Bulkowski places the stop at the handle low and raises it as price rises. If price falls below the handle low, the pattern's structure has failed.
Target
There are two versions of the measure rule:
- Full measured move (ChartSchool): measure from the cup's right peak to its bottom, then project that distance upward from the breakout.
- Adjusted measure (Bulkowski): measure from the right cup lip to the lowest valley, multiply by 61%, and add the result to the breakout price. The 61% is the share of patterns in his sample that met the full target.
Worked example with round numbers:
| Input | Value |
|---|---|
| Right rim (breakout level) | $100 |
| Cup bottom | $75 |
| Cup height | $25 |
| Full measured-move target | $100 + $25 = $125 |
| Bulkowski adjusted target | $100 + ($25 × 0.61) = $115.25 |
The live version of this article computed the target by applying a percentage to the breakout price ($101 × 1.25 = $126.25). Neither reference describes that method. Both use the price height of the cup.
Where does the cup and handle target land?
Enter the breakout level (right rim) and the cup bottom, then choose the full measured move or Bulkowski's 61% version to see the target price.
| Right rim (breakout level) | Cup bottom (lowest valley) | Share of cup height projected upward (1 = full measured move, 0.61 = Bulkowski adjusted) | Measured-move target price |
|---|---|---|---|
| 100 $ | 75 $ | 0.61 × | 115.25 $ |
| 100 $ | 75 $ | 1 × | 125 $ |
| 100 $ | 75 $ | 0.61 × | 115.25 $ |
- Rim and bottom prices are your own chart readings, not figures from any source.
- Both methods use the cup's price height added to the breakout level, not a percentage applied to the breakout price.
- The 0.61 factor is the share of patterns in Bulkowski's sample that met the full target; it is not a probability for your chart.
This only does the target arithmetic on your inputs: it does not check whether the pattern is valid, set a stop at the handle low, or cover the inverted pattern, whose target uses the handle height instead.
What the statistics say
Bulkowski's bull-market figures for the cup with handle, based on 913 trades he labels "perfect":
| Measure | Value |
|---|---|
| Overall performance rank | 3 of 39 (1 is best) |
| Break-even failure rate | 5% |
| Average rise after breakout | 54% |
| Throwback rate | 62% |
| Share meeting the full price target | 61% |
Bulkowski defines the break-even failure rate as the share of patterns that fail to rise at least 5% after an upward breakout. The page gives no date range for the 913 trades and does not say which markets they cover, and the "perfect trade" label is not defined in his glossary. The page publishes no bear-market figures. These numbers come from one researcher's dataset and should not be read as odds for any chart in front of you.
The inverted (bearish) cup and handle
The inverted version is the bearish mirror image. Price makes a rounded top, the two rims bottom out near the same price, and a handle forms to the right. Bulkowski's rules:
- The handle must not rise above the cup top. It often retraces 30% to 60% of the cup's height.
- The pattern confirms when price closes below the right cup lip, which is also his short-entry signal.
- Target: measure the handle height and subtract it from the right rim low. This uses the handle, not the cup.
His bull-market figures for the inverted pattern, from 556 "perfect" trades:
- Overall rank: 6 of 36
- Break-even failure rate: 18%
- Average decline: 17%
- Share meeting the target: 62%
The page gives no stop-loss rule for this version. On a spot market you can only sell what you hold. Opening a short position needs a derivative such as a futures contract.
Using the pattern on crypto charts
The rules above come from stock-market research. They can be applied to crypto charts, but two differences matter:
- Volume is split across venues. Crypto trades on many venues around the clock, so the volume on any one chart is a partial view. That makes ChartSchool's volume-expansion test harder to read.
- Volatility is higher. Deeper and faster swings make ChartSchool's "1/3 to 1/2 in volatile markets" range the more relevant cup-depth guide. They also make V-shaped recoveries, which do not qualify, more common.
If you are new to chart reading, start with our guide to reading crypto charts. The falling wedge pattern is another continuation setup confirmed by a breakout. Position sizing and stop discipline are covered in our crypto risk management guide.
Limitations: the identification rules and statistics come mainly from two references, both built on stock data, and Bulkowski does not publish the period his sample covers. This article is educational and is not trading or investment advice.
