Three Inside Up Candlestick Pattern (and Three Inside Down): How to Trade It and What the Data Shows

The three inside up candlestick pattern is a three-candle bullish reversal pattern that forms after a downtrend: a long bearish candlestick, a small candle that stays within the first one, then a bullish candle that closes above the open of the first. Its mirror image after an uptrend, the three inside down, is a bearish reversal pattern.

This guide explains how to spot both setups, the psychology behind them, how traders use them, and what happened after them in reality. Our test bench applied the same rule to full market history, found 6,665 bullish cases on daily data alone across stocks, crypto, forex, ETFs, indices and commodities, and compared the outcome with random days on the same assets.

What are the three inside up and down candlestick patterns?

These two candlestick patterns are confirmed versions of the harami. The harami is a two-candle signal: a long candlestick followed by a short one that sits within its real body, a sign that the trend has lost momentum. The three-candle version adds a final candle that confirms the change of direction, indicating a potential trend reversal.

Steve Nison, who introduced Japanese candlestick charting to Western traders in Japanese Candlestick Charting Techniques (1991), describes the harami and the need to confirm it. The "three inside" names are a later Western label for the harami plus its confirmation candle.

The three inside up candlestick pattern

What is the three inside up pattern?

In short, the three inside up candlestick pattern is a reversal signal that appears at the end of a downtrend and suggests that buyers are taking control: a bullish harami followed by a bullish candle that closes above the first candle's open.

How to identify the three inside up candlestick pattern

1. A prior downtrend. The pattern needs a decline to reverse. Our bench checks that the close before the pattern is at least one ATR lower than 10 sessions earlier and in the lower half of the last 40 closes.
2. A long bearish candlestick. The first candle is a big black (or red) candlestick that extends the decline.
3. A short second candle. Following the previous candle, it has a small real body contained within the range of the previous candle's real body. Its color does not matter, and it can be a doji or a spinning top.
4. A confirmation candle. The third candlestick is white (or green) and closes above the first opening level, so it recovers the whole bearish move.

On the median daily occurrence our bench found, the three candles together span 5.18% of the asset's price, or 1.78 times its average true range (ATR). The setup is wider than a typical session, as you would expect from three bars that start with a long bearish one.

Market psychology behind the three inside up

The first session shows sellers in control. On the next session, selling pressure fades: the market trades within the previous range and barely moves. On the last session, the bulls push the close above the point where the first decline started. Traders read this as a shift in the balance between buyers and sellers, and as the possible start of an uptrend. Whether the shift lasts is a question for data; we come back to it in the reliability section.

The three inside down candlestick pattern

What is the three inside down pattern?

In short, the three inside down is the bearish reversal version: a bearish harami followed by a bearish candle that confirms it, at the end of an uptrend. It suggests that the advance is ending and that a trend reversal to the downside is possible.

How to spot the three inside down formation

• A prior uptrend.
• A big up candlestick that extends the rise.
• A short candle whose real body sits within the first candle's real body.
• A bearish candle that closes below the first opening level.

What the bearish version shows

Bulls dominate the first session, hesitate on the next, and lose the whole first advance on the last one. The final bar shows that the bears have erased the gains of the first day. As with the bullish pattern, the story is plausible but the data is less clear.

How to trade the three inside up and down patterns

The trading rules are simple. Our bench simulates the same trade on every occurrence, which gives a fair test of the signal alone. Here is the full plan for the bullish version, step by step:

For the three inside down, reverse every step: an uptrend first, a sell stop below the low of the pattern, a stop loss above its high.

Setting entry points for buy and sell signals

There are two ways to enter the bullish setup. The aggressive entry buys at the close of the third session. The conservative entry places a buy stop just above the top of the pattern (or sets alerts there) and waits for the market to trade through it. We tested the second version: on daily data the buy stop was triggered in 86.3% of cases, and in 13.5% of cases the bottom of the pattern broke first, which canceled the setup before any trade was opened; the few remaining orders expired untouched after 20 bars or when the data ended. For the bearish pattern, the sell stop goes below the bottom of the pattern.

Managing risk with stop-loss placement

Taking profit: exit strategies for traders

The pattern does not give a target. Classic exits are a fixed reward-to-risk ratio, the next resistance level, a trailing stop, or an opposite signal on the upcoming candles. Our simulation uses a target at twice the initial distance to the stop (2:1). On daily data, trades that reached it took a median of 18 days; trades that were stopped out took a median of 9 days. Of the daily trades that hit their stop, 24.1% had first been up 1 R.

Combining with technical analysis

The three inside up candlestick can be combined with technical indicators and other free tools that charting platforms offer. Here is what we could measure and what we could not.

Using moving averages to filter the trend

We split daily occurrences by the 200-day moving average. Above it, the 2:1 setup reached its target in 37.5% of cases with an expectancy of +0.14 R; below it, 34.5% and +0.05 R. The moving average filter barely changed the result.

Incorporating RSI for overbought and oversold signals

Oscillators are another option. One approach is to take the bullish setup only when the RSI indicator is oversold (below 30), and the bearish one when it is overbought (above 70). We have not tested an RSI filter; if you use this indicator, test it on your own market first.

Combining patterns with support and resistance levels

A bullish pattern at a known support level, such as a prior swing bottom or a rising trend line, has a logical reason behind it: demand appeared at that level before. We have not measured support levels, because their definition varies from one trader to another.

Utilizing volume to confirm breakout strength

Classic analysis reads higher volume on the third session as confirmation. On daily stocks, bullish cases printed on volume above 1.5 times the 20-day average reached the 2:1 target in 35.8% of cases with an expectancy of +0.13 R, against 35.6% and +0.08 R on normal volume. These splits are descriptive and not corrected for multiple testing: volume made little difference.

Comparison with similar candlestick patterns

Several other candlestick patterns tell the same story of a trend that stalls and turns. Here is how each candlestick pattern compares on the daily timeframe, all markets combined.

Three inside up candlestick pattern vs. morning star

Both are three-bar bullish reversal patterns, and each candlestick pattern needs a prior decline. In the three-bar star reversal, the middle candle gaps away from the first body; in the harami version, it stays within it. The star also needs its last candle to close well into the first body, while our setup needs a close above the first opening level. Neither beat chance on direction by a margin our tests separate from it.

Three inside up vs. three outside up

The three outside up is the engulfing version: a down bar, an up bar that engulfs it, then a higher close. One starts from a harami (a short candle within a big one), the other from a bullish engulfing (a big candle around a short one). The outside version was much more frequent on daily data (11,258 occurrences against 6,665).

Three inside down vs. evening star

The evening star is the bearish counterpart of that bullish formation. On daily data it was followed by a decline after 10 days in 46.9% of cases, against 46.0% for the three inside down: both close to random days.

Three inside down vs. three outside down

Same logic as on the bullish side: one is a confirmed bearish harami, the other a confirmed outside reversal. On daily data, the three inside down did not differ measurably from the rest of the market on the average excess return, or from random dates on the average result per trade. The three outside down did not either: its average excess return of −0.09% 10 days later is within the margin of error, and its 2:1 trade stayed close to random dates.

Is it just a confirmed harami?

Yes, and the third candle changes the trade more than the odds. With the bullish harami alone, our buy stop above the pattern was triggered in only 51.5% of daily cases; with the confirmed version, 86.3%, because the third candle has already closed near the top. The direction after 10 days was close to chance in both cases.

Advantages of using three inside patterns

• Clear rules: three bars, a short one within a big one, a close beyond the first open.
• A built-in check, so fewer setups are canceled before entry than with the harami alone.
• A natural stop at the other end of the pattern.
• Frequent: about one occurrence every 1,272 daily candles for the bullish version.

Limitations and common mistakes to avoid

• On its own, the pattern did not clearly beat chance in any market or timeframe we tested.
• The stop is wide because the first bar is tall.
• The confirmation comes late: by the time the third bar closes, part of the move is already done.
• Against a strong trend, a reversal candle can fail, and a rate policy decision or earnings news can overwhelm any candle.

The importance of context in chart analysis

Avoiding false signals in sideways ranges

In sideways conditions, short candles within big ones appear constantly. Without a clear prior decline (or advance for the bearish version), skip the signal. Also wait for the third session to close: a candle that looks strong at midday can close lower.

Best timeframes for trading these patterns

No timeframe stood out once we compared with random dates. Weekly bullish cases had the best raw results, 63.4% higher after 10 weeks and a 2:1 hit rate of 47.9%, but random weeks on the same assets rose in 57.1% of cases and the same setup hit its target in 40.8%. The average excess return does not differ significantly from the rest of the market, nor the average result per trade from random dates. On hourly crypto, which suits day trading, the up version moved up after 10 hours in 49.8% of cases against 48.0% for random hours.

Reliability: what the data shows

The three inside up candlestick pattern is not reliable on its own: across every market and timeframe we tested, its average excess return never clearly beat the rest of the market, and on daily commodities it moved against the expected direction relative to its market (−1.56%), by a gap that passes a simple test but not our correction: weak, below chance. On weekly ETF charts, the 2:1 setup reached its target in 68.4% of 19 trades against 40.4% on random dates, rated Not enough data: fewer than 30 trades, so the gap is shown, not interpreted. Hourly and 4-hour forex charts are missing from the table: our bench finds no three inside pattern there.

How to read the table

Read it in pairs of columns; the details of each test follow from the same rules. On daily data, 54.0% of bullish cases were followed by a gain 10 days later. That sounds bullish until you see that 54.2% of random days on the same assets were too: the average stock price rises a little more often than it falls. The excess return over the rest of the market after 10 days is −0.03%, with a 95% confidence interval from −0.23% to +0.17%. Zero is inside the interval: no clear edge.

The 2:1 setup tells the same story. Its share of targets hit was 35.8%, against 37.6% for the same entry, stop and target replayed on random dates, for an expectancy of +0.09 R against +0.16 R for the random version. Expectancy is measured in R (the initial risk) before fees and slippage, which would lower it further. The bearish version was in the same situation: 46.0% of daily cases were followed by a decline, against 45.2% for random days.

Neither half of the history showed a clear edge: the excess return was +0.09% in the first half of the history and −0.15% in the second. These results agree with academic work: Marshall, Young and Rose, in "Candlestick technical trading strategies: Can they create value for investors?" (Journal of Banking & Finance, 2006), found that candlestick strategies did not add value on large US stocks. You can explore every step in our methodology.

Do these patterns work for forex and crypto?

Not better than elsewhere, and on hourly and 4-hour forex charts not at all: our bench finds no occurrence there. On a market that trades around the clock, the second candle almost always opens where the first one closed, so its body touches the edge of the first one and never sits strictly inside it. On daily crypto the pattern is rare for the same reason (38 occurrences on 38 coins since 2020): it moved up in 56.8% of cases against 40.7%, and the 2:1 trade hit its target in 43.3% of cases against 35.0% on random dates, rated no clear edge: on so few cases, gaps of that size do not pass our tests. Hourly crypto, the largest intraday sample, is in the table above.

Real-world examples of the patterns in action

In concrete terms, here are real three inside up and down patterns, picked by our bench (best and worst results, textbook shapes, most recent, random draws), with their actual outcome. The three case studies below are on daily crypto charts.

A bullish setup that worked

From September 6 to 8, 2024, BNB/USDT printed a three inside up: a bearish candle from 502.4 to 487.2, a short candle within it, then a close at 503.1, above the first open. The buy stop at the top of the pattern (509.9) was triggered on September 9, with a stop at 471.3 and a 2:1 target at 587.1. The target was hit on September 21, when BNB traded up to 590.

A bullish setup that failed

From January 13 to 15, 2022, UNI/USDT formed the same setup: a bearish candle from 16.29 to 15.45, a short candle within it, then a close at 16.47. The buy stop at 16.72 was triggered on January 16, with a stop at 15.13. Three days later, on January 19, UNI traded down to 15.10 and hit the stop: a full 1 R loss. Ten days after the pattern, UNI was 35.8% lower.

A bearish setup that was canceled

From March 29 to 31, 2026, TRX/USDT printed a three inside down, closing at 0.3133, below the first open of 0.3166. The sell stop sat at the bottom of the pattern (0.311) and the protective stop at the top (0.3246). The market never traded below the bottom before breaking the top on April 15, so the setup was canceled without a trade. Ten days after the pattern, TRX was 1.9% higher: the entry order was never reached, so the trader stayed out of a losing short.

Frequently asked questions

Summary of key takeaways

• The three inside up is a bullish harami confirmed by a third candle that closes above the first open, after a downtrend.
• The three inside down is its bearish mirror after an uptrend.
• On 6,665 daily occurrences, 54.0% of up cases were higher 10 days later, against 54.2% for random days.
• The 2:1 setup's share of targets hit was 35.8%, against 37.6% for the same setup on random dates: no clear edge.
• Volume and the 200-day moving average changed the results only slightly.

If you use these patterns, treat them as a location for a position, not as a reason for one: define the stop at the other end of the pattern, size the position on that distance, and keep the amount at stake per position modest. This page is educational content to keep you informed, not investment advice, and leveraged products can lose more than the stake.

About the author and the data. Written by Amaury, founder of PatternsWizard, who is passionate about finance, quantitative trading and data analysis. He built the PatternsWizard test bench, which applies the same published rules to full market history: crypto, forex, US stocks, ETFs, indices and commodities. Trades and win rates are compared with random bars on the same assets, excess returns with the rest of the market, with error margins and a correction for multiple testing, and the whole pipeline is checked on random-walk prices, where nothing can be predicted. Statistics on this page were generated on November 5, 2026. The full methodology explains every step.