Upside Gap Three Methods is traditionally classified as a bullish continuation pattern. Its real-body gap need not be an empty interval between whole candle ranges. A valid three-candle shape also does not predict a profitable continuation. Compare precise recognition rules and price data before defining a separate trade.

What the upside gap three methods looks like

TA-Lib’s CDLXSIDEGAP3METHODS checks two same-colored bodies with a real-body gap, followed by an opposite-colored candle opening strictly inside the second body and closing strictly inside the first. The bullish version has two up bodies and a down third. It does not check a preceding trend or historical body length. thinkorswim adds long-body and prior-trend parameters, so matching names need not yield identical markers.

TA-Lib’s Upside Tasuki instead ends the third body strictly inside the gap and checks that bodies 2 and 3 differ by less than its historical Near allowance. A Close location alone is insufficient. Rising Three Methods is a different five-candle detector with three intervening small bodies; “no gap at all” is not its general recognition rule.

Use the definition in what a price gap is carefully: this article tests the gap between real bodies, not necessarily L2 > H1. A body spanning an interval does not prove trades occurred at every intermediate price.

Pattern direction also depends on the quote unit

For positive price p in B per 1 A, the reciprocal quote is q = 1/p, in A per 1 B. Our hypothetical p OHLC values are (100; 108; 99; 104), (106; 111; 103; 110), (109; 110.5; 101; 102). Two bullish bodies followed by a bearish body become two bearish bodies followed by a bullish body. The body gap also changes direction.

The same positive prices after reversing quote unitsThree hypothetical candles in B per one A. The first two bodies rise and the third falls. Under q equals one divided by p, the first two fall and the third rises. Each reciprocal High equals one divided by the original Low, and reciprocal Low equals one divided by the original High. Panels have different units and separate linear scales.Reciprocal quotes reverse bodiesB per 1 AA per 1 B112233Gap: 104–1061/106–1/104Up: green, green, red1/p: red, red, green
All four OHLC prices are converted for each candle. Scale p spans 90–115; scale q spans 0.0087–0.0103. Each panel has its own linear scale and unit, so heights cannot be compared between panels.

Changing colours while retaining the same numbers is insufficient. For example, the p gap from 104 to 106 becomes a q gap from 1/106 ≈ 0.009434 to 1/104 ≈ 0.009615. O′3 = 1/109 lies inside the second body and C′3 = 1/102 inside the first; check their ordering in the new unit too. Record the pair and quote convention before interpreting the pattern's name.

Assumptions: Entirely hypothetical positive prices without spread, fees or an execution model. Each O and C becomes 1/p, but H′ = 1/L and L′ = 1/H because the reciprocal function is decreasing. No linear transformation of spread, orders or actual trade outcomes is assumed. Official material: TA-Lib 0.6.4: CDLXSIDEGAP3METHODS.

Twenty pips between bodies, overlapping full ranges

Assign three fictional EUR/USD Bid OHLC bars, one pip 0.0001: (1.1000, 1.1080, 1.0990, 1.1040), (1.1060, 1.1110, 1.1030, 1.1100), (1.1090, 1.1105, 1.1010, 1.1020). Bodies are 40, 40, 70 pips. The first body ends 1.1040 and the second starts 1.1060: gap 20 pips. Full ranges overlap 1.1030–1.1080, or 50 pips, so there is no full-range gap.

O3=1.1090 lies strictly inside body 2 and C3=1.1020 strictly inside body 1. No prior history is supplied, so this checks TA-Lib geometry rather than proving a prior uptrend. A separate change H1=1.1050, L2=1.1055 creates a5-pip full-range gap while retaining the 20-pip body gap.

Change only C3:1.1040 touches C1 and fails the strict Three Methods test;1.1050 lies inside the body gap and is a Tasuki candidate;1.1060 equals O2 and fails the strict gap-interior Close test. At C3=1.1050, body 3 becomes 40 pips, equal to body 2, but a full Tasuki classification still needs history for Near. A wick can pass below the gap even when the Close remains inside it.

TA-Lib treats C=O as white, but a zero-width first or second body cannot contain the required strict interior price. Read all conditions together. OHLC does not disclose participant identities, exhausted supply or whether a correction was technical rather than fundamental.

The feed and gap definition determine what is counted

MetaTrader documents OTC bars based on Bid quotes and exchange instruments using Last data. A Bid bar is not a record of all FX transactions. BIS’s survey reports average daily FX turnover of 7.5 trillion USD in April 2022 across surveyed instruments; that historical total is neither a current figure nor evidence that a provider’s prices are continuous.

Trading breaks and interrupted data require separate handling: record provider, clock, interval and missing observations before labeling a gap. The weekend forex trading article covers reopening risks. A weekend or news label alone does not establish this body pattern’s frequency or a universal pip jump.

The relevant question is which gap definition is being tested. A real-body gap can exist even when candle shadows overlap, so continuous weekday trading alone does not rule this pattern out. Count occurrences on the selected pair, timeframe and feed rather than declaring that it can only occur at the Sunday open.

Tasuki needs body similarity as well as the Close

An independent source-condition calculation assigns five prior OHLC bars(100,100.5, 99.5,100.2), then(100,105, 99,104),(106,111,105,110). The third has H/L/C111/102/105; only O3 changes. For Near referenced to candle 2, the preceding five ranges are four 1-unit ranges plus 6:0.2 × (4 × 1 + 6)/5 =0.4. O3=109 gives bodies 4 and 4, difference 0 <0.4, so the Tasuki conditions pass. O3=107 gives bodies 4 and 2, difference 2, and fails. Both C3 values remain in the same gap.

Both variants return no Three Methods match because C3=105 is outside body 1. These hypothetical calculations inspect source predicates; they are not a new compiled-library run or a success rate. The separate published 120-bar Kraken EUR/USD H1 classification for 21–25 September 2026 contains zero nonzero CDLXSIDEGAP3METHODS codes. That one-feed interval does not establish general FX rarity or a trading result.

How to trade the upside gap three methods step by step

The complete three-bar geometry is known after candle 3 closes. Requiring a later Close above H2 is an added entry rule, not a fourth candle needed to recognize the pattern or proof buyers returned. A trend filter likewise defines another version. Record its exact history, data availability, order and expiry before evaluation.

Specify trigger, order, units and exits separately. A close-below-pattern rule is different from an intrabar stop; neither defines maximum loss for an ordinary order. Fix the offset and cost assumptions rather than prescribing “a few pips”. The stop loss versus take profit article distinguishes these instructions.

Retain a separate fictional geometry from the original example, now with complete assigned Bid OHLC: (1.0845, 1.0885, 1.0840, 1.0880), (1.0895, 1.0940, 1.0890, 1.0930), (1.0925, 1.0930, 1.0855, 1.0862). Bodies 35, 35, 63 pips have a15-pip gap, O3 inside body 2 and C3 inside body 1. Their lowest Low is 1.0840. These assigned extra extremes complete the illustration, not observed market data or a verified trend.

Assign a later Close 1.0945 above H2=1.0940. Only after that decision, separately assume an actual next available Ask fill exactly 1.0945; equality with the signal Close is an illustration assumption, not automatic execution. Chosen SL1.0830 and Bid target 1.1060 give 115-pip price distances on either side before costs. Planned 1:1 does not establish expectancy, and the stop can fill worse.

Strict numerical comparisons are reproducible once the feed and detector are specified; the width of a drawn candle is not a tolerance rule. A prior-trend filter needs its own definition. Store rejected boundary cases and missing history rather than treating judgment or a protective stop as proof of a known maximum loss.

The mirror image of this pattern on the downside is the downside gap three methods — identical logic, the other way round. And if you want to understand gaps in themselves, start with the piece on trading price gaps, because they are the heart of this setup.

What to do tomorrow

  1. Specify the detector version, trend/length parameters, source and time interval. Validate complete OHLC and compare strict body versus full-range gaps.
  2. Check Tasuki body similarity and its historical Near reference, not only the Close. Keep equality and insufficient-history cases.
  3. Record recognition time separately from later orders/fills and include fees, funding and adverse execution. Do not infer pattern frequency by counting only Opens beyond the preceding full range.
  4. Test complete rules on data not used to select them. A handful of stock charts, a week on demo or an added indicator does not establish profitability on FX.

To organise your candlestick knowledge within a broader course, a good starting point is the technical analysis section in our library.