Cup and Handle Pattern: What the Handle Limit Is Measured On

A cup and handle carries one number that decides what a reader does with it: how far the handle may pull back before the shape is discarded. Four widely read descriptions of the pattern state that limit, and each measures it against a different quantity. Two of them use two different quantities inside their own page.

The numbers are not where they disagree. What they are dividing by is.

The Four Prices a Cup and Handle Is Measured From

Every rule attached to this pattern is a ratio, and a ratio needs two prices. Four prices exist on the chart, and naming them first removes most of the confusion that follows.

The first is the high that ended the advance before the shape began. The second is the low of the rounded base. The third is the high on the right-hand side, where price returns to roughly the level it left. That third price is the breakout level, and it is the one most descriptions call the rim or the neckline.

The fourth is the low of the handle, the short pullback that forms after price has returned to the rim.

Two distances can then be built, and they are not the same distance. The depth of the base is the rim minus the low of the base. The advance that preceded the shape is the prior high minus wherever that advance started, which on many charts is far below the base. Confusing the two is the root of the problem in the next section, because both get called the height of something.

The rounded base is what separates this shape from a base confirmed by two or three distinct touches of the same level, where the low is a level rather than a curve and the measurement starts from a price the chart printed more than once.

The Handle Limit Is Quoted Against Four Different Denominators

Four full descriptions of the pattern sit behind this section. Every one of them states how deep the handle is allowed to be, no two measure that depth against the same quantity, and not one of them tells a reader which quantity to prefer when the others disagree.

One states the limit against the height of the cup. One states it against the depth of the cup. One states it against the rise on the right-hand side of the cup alone. One sets the depth of the cup itself as a fraction of the height of the whole formation, which makes its handle rule a fraction of a fraction.

Two of the four are inconsistent inside their own page. In one, the body says the limit is measured against the height of the cup and its own question-and-answer section says depth. In the other, depth appears five times and height once, in the section that describes the target.

The middle two rows are where a reader loses the thread. Height and depth are the same distance on a symmetric base and are not the same distance on an asymmetric one, and a base that is rounded rather than pointed is exactly what allows the two sides to end at different lows.

Patterns whose definitions rest on named proportions have to be stricter than this. Shapes defined by fixed retracement ratios name both ends of every measurement for that reason.

Fix the Denominator and the Reward-to-Risk Ratio Falls Out of It

Pick one denominator and hold it. Call the depth of the cup D, measured from the rim down to the low of the base. Let the handle limit be a fraction f of that depth, so the deepest allowed handle low sits at the rim minus f times D.

The textbook trade is described the same way in all four descriptions. Enter on the break of the rim. Place the stop below the handle low. Project the depth of the cup upward from the rim for the target.

Written out, the reward is D and the risk is f times D. The ratio between them is D divided by f times D, and the D cancels.

A handle limited to one third of the cup depth gives three to one. A handle limited to half gives two to one. A handle limited to a quarter gives four to one. Those are not results anyone measured on a chart. They are the arithmetic of the rule that was chosen, and they follow from it before a single trade is taken.

Three conditions hold that cancellation in place: the entry sits at the rim, the stop sits at the handle low, and the target is the cup depth. Move any one of them and the ratio changes. An entry taken above the rim on a confirmed close raises the risk and lowers the reward at once, which is a real cost of waiting rather than a detail, and it is the same trade-off that governs a continuation shape whose target is also a projected distance.

What the depth of the cup does change is size. Risk in price terms is f times D, so a base twice as deep puts the stop twice as far away, and a fixed money risk buys half the position. The ratio survives; the exposure does not.

Why the Published Ratios Disagree by Exactly That Much

Return-to-risk ratios circulate for this pattern as though they were findings. One description prints two of them on the same page, a lower one for ordinary breakouts and a higher one for extended moves, each attributed to a commercial blog rather than to a study.

Run them backwards through the arithmetic above. A ratio of two and a half to one is the reciprocal of a handle limit of two fifths. A ratio of four to one is the reciprocal of a limit of one quarter. Both are handle rules restated as outcomes.

That is the whole of the disagreement. Two authors who pick different handle limits will publish different ratios without either of them having observed anything, and a reader comparing the two ratios is comparing two editorial choices about how deep a handle may be.

It also explains why the ratio never appears with a distribution around it. A measured ratio would arrive with a spread, a sample size and a period. A derived one arrives as a single clean number, which is what these are. Distinguishing a breakout that held from one that did not is the part no ratio settles, and it is treated separately in what happens when a break of a level fails.

MetaTrader Has No Chart-Pattern Detection at All

All four descriptions assume a platform that finds the shape. One of them devotes a section to automating detection inside its own charting product, with named settings for depth and for switching to the inverted version.

The MQL5 Reference lists the technical indicator functions a MetaTrader terminal provides. Read on 17 August 2026, that list runs from the Accelerator Oscillator through to the Williams Percent Range and covers moving averages, oscillators, channels, volume tools and the Ichimoku set. The strings cup and pattern do not appear on it at all. Neither does any function that returns a chart formation.

The same reference documents custom indicators, which is where anything not on that list has to come from. A cup and handle detector on MetaTrader is therefore third-party code by definition, written by someone who chose a denominator, a tolerance for how rounded a base has to be and a rule for when a handle has ended.

Those choices are the ones the previous sections showed to be unsettled. A detector that flags shapes on one terminal and a detector that flags shapes on another are answering different questions, and neither of them is answering a question MetaQuotes has defined. Any screenshot of a pattern found automatically carries its author settings with it, whether or not the settings are shown.

What the Volume Rule Assumes, and What a Forex Chart Counts

Volume carries the confirmation in every description read here. The base is supposed to form on falling participation and the break of the rim is supposed to arrive on a rise in it. Two of the four attach a numeric threshold to that rise.

The assumption underneath is that the volume series counts contracts or shares. On a centralised exchange it does. The spot currency market has no central tape, so a currency chart in a retail terminal is usually plotting the number of price changes in the bar rather than an amount transacted.

That substitution keeps the shape of the series and destroys the units. A rise on a tick series says quoting became more active, which is related to participation and is not the same statement. The distinction between the two series, and what each one is capable of measuring, is set out in the page on what a tick volume field actually counts, which carries that argument in full.

One narrow consequence falls out of it for this shape. A volume threshold expressed as a percentage above an average survives the substitution, because it is a ratio of the series against itself. A threshold expressed in contracts, shares or lots does not survive it at all, since the series is not denominated in any of them.

Which Figures This Page Does Not State, and Why

No handle limit is given here as a rule. Four descriptions state four of them, each against a different quantity, and behind each one there is no retrievable document to open. Choosing between them would mean crowning one unsourced convention over three others on no evidence at all.

No success rate appears either. One description publishes rates rising with the length of the lookback window and, on the same page, a baseline figure from a sample of a few thousand instances that is far below all three. Those cannot all describe the same pattern, and the attributions are to commercial sites rather than to a retrievable study.

Also absent: how long the base or the handle should take to form, how rounded a base has to be before it stops qualifying, and the volume thresholds mentioned above. Each of those circulates widely and is supported on none of the pages that print it.

Absence here is not a verdict against any of those numbers. The claim being made about them is narrower: nobody publishing them has shown where they came from, so printing one would hand a reader something this page cannot stand behind. What appears instead is arithmetic, and arithmetic can be checked against the definitions rather than against anyone.

Who This Page Is Not For

A reader who came for the correct handle depth leaves without one. The four descriptions that state a depth are measuring against four different quantities, and nothing on any of them decides which quantity is the right one.

A reader who came for how often the shape works leaves without that too, and for the same reason.

What is left is worth three measurements before the shape is used at all. Establish which two prices the limit is a ratio between, and write both down rather than carrying a percentage in your head.

Establish next whether the entry, stop and target you intend are the three the ratio arithmetic assumes, since moving any one of them changes the reward-to-risk figure you think you are taking. Then establish whether the volume series on the chart is counting transactions or price changes, because that decides which of the confirmation rules is capable of meaning anything.