Upside and downside gap three methods: when a gap gets filled

The upside gap three methods is two rising white candles separated by a gap, followed by a black candle that falls all the way back through the gap and closes inside the first white body. The downside gap three methods is the reverse: two falling black candles with a gap between them, then a white candle that rallies back through the gap into the first black body.

At first sight the third candle looks like a reversal, because it erases the gap that the trend had just created. Traditional analysis reads it the other way. The trend made its point with two candles in the same direction, the third candle is a pullback that fills the gap, and once the gap is filled the trend can resume from a cleaner base. Both directions are classed as continuation patterns.

TA-Lib handles both directions in one function, CDLXSIDEGAP3METHODS (the X stands for "up or down"). This page covers the anatomy, the precise rules and how they differ from the Tasuki gap (upside and downside): a gap that holds on a pullback, two real upside cases with opposite outcomes, and data on 194 detections, which makes this one of the most frequent gap patterns.

Anatomy of the gap three methods

For the upside version:

1. First white candle, part of an advance.
2. Second white candle whose body opens and closes entirely above the first body: a gap up between bodies.
3. Black candle that opens inside the second body and closes inside the first body. The gap between the two white bodies is now closed.

For the downside version, read every color and direction in reverse: two black candles with a gap down between their bodies, and a white candle that opens inside the second black body and closes inside the first.

The exact rules Finabase uses

Finabase runs TA-Lib's CDLXSIDEGAP3METHODS on daily bars:

What the rules leave out is as important as what they include:

No minimum gap. One cent between bodies is enough. In the Texas Instruments example below, the gap was 48 cents on a $217 stock.

No size tests. None of the three candles has to be long or short. The textbook drawing has two strong candles; TA-Lib accepts two small ones.

No trend test. In our data 48.4% of upside detections formed after a ten-session decline, nearly half outside the setting the pattern is named for.

Gap three methods versus the tasuki gap

Both start with a candle, a body gap and a second candle in the gap's direction, and both end with an opposite-colored third candle that opens inside the second body. TA-Lib then separates them on four points:

So the same two opening candles can end as either pattern, depending only on where the third candle closes. Close in the gap with a similar-sized body: tasuki. Close through the gap into the first body: gap three methods. Close below the first body entirely: neither.

Why a filled gap can be bullish

Gaps are often described as something the market "wants" to fill. Whatever the truth of that saying, a gap left behind a fast move does attract orders: sellers who missed the top of the first candle, and buyers who wanted in but refused to chase the gap. The black third candle in the upside version is those orders meeting. It fills the gap in a single session.

The continuation argument is that this is a healthy reset. The trend produced two candles in its direction, the overextension was corrected in one session, and the stock now sits back in the first candle's body with the gap gone. If buyers return from there, they are building on a level that has been tested rather than on an empty gap.

The counter-argument is plain on the chart: the third candle is a strong move against the trend that erased a gain in one day. That is exactly why traditional sources insist on confirmation before treating it as continuation.

Confirmation

• Upside: a close above the second white candle's high. Downside: a close below the second black candle's low. Either one says the trend resumed.
• A softer version used in our statistics: the next close beyond the third candle's high (upside) or low (downside).
• The first body holds. For the upside version, a close below the first white candle's open would mean the pullback is more than a gap fill.
• Trend health indicators, such as an Relative Strength Index (RSI): how to read and use it that stays above 50 during the pullback in an uptrend.

Two real upside cases

Texas Instruments, April 15, 2026. TXN had gained 11.4% in ten sessions. A white candle on April 13 closed at $216.71; another white candle on April 14 opened at $217.19, 48 cents above it, and closed at $218.87. On April 15 a black candle opened at $217.56 and closed at $216.29, inside the April 13 body: a tiny gap, filled in one session. The next close was $223.10, above the pattern's high, and ten sessions later TXN was 24.5% higher. The pattern played out as described, though the gap was so small that most traders would not have noticed it on the chart.

Qualcomm, May 27, 2026. QCOM had risen 11.0% in ten sessions, including a $23.73 white body on May 22 that closed at $238.16. On May 26 it gapped up, opening at $240.72 and closing at $248.82. On May 27 a $12.03 black candle opened at $245.43 and closed at $233.40, inside the May 22 body. The early follow-through looked like continuation: the stock closed at $251.02 two sessions later, above the pattern. Then it broke down, and by June 10, ten sessions after the pattern, it closed at $191.20, 18.1% below the pattern close. A filled gap after a near-vertical run can mark the top as easily as a pause.

What the data says

Over two years of daily bars (October 30, 2024 to October 29, 2026) for 60 large US stocks and ETFs, CDLXSIDEGAP3METHODS fired 194 times, at least once on every symbol: 92 upside and 102 downside.

• Upside: no better than a random day. The direction count was 1.8 points above the baseline, and the average returns were slightly below it.
• Downside: near the baseline at five sessions, with a slightly negative ten-session average against a positive baseline. That is a small lean in the textbook direction.
• The confirmed subsets moved more, 7 of 11 confirmed downside cases lower a week later, but 15 and 11 cases are anecdotes.

Related patterns

Despite the name, the gap three methods has little in common with the rising and falling three methods, which are five-candle consolidation patterns with no gap.

An educational outline

How the pattern is generally approached. It isn't a recommendation to trade.

1. Filter the noise. Look for a visible gap and two candles of real size in the trend direction. A 48-cent gap on a $217 stock may qualify for TA-Lib and still mean little.
2. Use the first body as the line. For the upside version, a close below the first white candle's open says the pullback is a reversal, not a gap fill.
3. Wait for resumption. A close above the second white candle's high is the strongest confirmation.
4. Watch the extension. After a very steep run, like Qualcomm's, a filled gap deserves more suspicion.

How Finabase shows the gap three methods

• Chart markers. The Upside/Downside Gap Three Methods study marks the third candle of each detection on daily charts.
• Scanner. TA-Lib scores the pattern by the first candle's color: upside (+100) on the bullish list, downside (−100) on the bearish list. A stock on the bullish list will show a black latest candle, which is how the pattern ends.
• Signals tab. The Candlestick gauge on a symbol page such as AAPL's signals gives each of 61 patterns a vote from its latest reading in the last eight bars; upside is Buy, downside is Sell. The score (buy − sell) / (buy + sell + min(neutral, max(buy, sell))) reads Strong Sell below −0.6, Sell below −0.2, Neutral below 0.2, Buy below 0.6 and Strong Buy above.

Frequently asked questions

Is the upside gap three methods bullish even though it ends with a black candle?

Yes. The black candle is treated as a pullback that fills the gap. TA-Lib scores the upside version +100 and the downside version −100.

What is the difference from the tasuki gap?

Where the third candle closes. In the tasuki gap it closes inside the gap, leaving it partly open, and TA-Lib requires similar body sizes. In the gap three methods it closes inside the first candle's body, so the gap is completely filled.

How big does the gap have to be?

Any size. TA-Lib only requires the second body to sit entirely beyond the first. Judging whether a gap is meaningful is left to the reader.

Why is it so common in the scanner?

Because the rules are loose: tiny gaps count and there are no size or trend tests. It fired 194 times in 29,760 daily bars, on all 60 symbols in our sample.

Summary

The gap three methods is a trend that gaps, then fills the gap in one session, with the traditional expectation that the trend resumes from the filled level. TA-Lib's version is loose, with no gap minimum and no size or trend test, so it appears often and many detections are trivial. On 194 cases in two years of large-cap data, it behaved much like an ordinary day. Texas Instruments and Qualcomm show the two ways it resolves. The useful part is the structure, a first body that should hold and a second-candle high or low that marks resumption, not the label itself.