Tri-Star Candlestick Pattern: Bullish and Bearish Setups Tested

The tri-star candlestick pattern is a three-candle reversal pattern made of three consecutive dojis, the middle one gapping away from the other two. After a downtrend it is a bullish tri-star; after an uptrend, a bearish tri-star. It is uncommon: on daily charts our bench found one bullish tri-star every 20,020 bars.

This guide explains how to identify both versions, what they tell traders, how to trade them, and what happened after them on real data. Our test bench scanned full price history with a strict rule and detected 389 bullish and 551 bearish on daily charts across stocks, ETFs, indices and commodities, and compared the outcome with random days on the same assets.

What is the tri-star candlestick pattern?

A tri-star is a pattern that consists of three consecutive doji candles. A doji is a candlestick whose opening and closing prices are equal or almost equal, so its body is a thin line: a sign of indecision. One such candlestick is common. Three in a row, with the middle one set apart by a gap, are not: the market hesitates three sessions in a row at the end of a trend.

The doji page covers the single candle and its types. A tri-star extends the doji star, where a cross gaps away from a long candlestick, by replacing the long candlestick with a cross and adding a third one.

How to identify the tri-star formation

Our bench uses a rule close to the open-source TA-Lib library (function CDLTRISTAR): it finds 76% of the bullish cases and 75% of the bearish cases that function flags. Each candlestick qualifies when its body is at most 10% of the average range of the ten previous candles, the same test used on our doji pages.

The first candle: the initial doji

The trend is established: price is at least one average true range below its level of 10 bars earlier for the bullish version, above it for the bearish version, and in the matching half of its recent range. Then a first doji appears. On its own it only says that the trend paused for a session.

The second candle: the gap and indecision

The second doji gaps away in the direction of the trend: its body sits entirely below the first body in a falling market, entirely above it in a rising market. The gap is measured on the bodies, not on the shadows, so it can be very small. This candlestick is the star: the trend pushed price one step further, and the session still ended where it began.

The third candle: the completion of the pattern

The third candlestick comes back toward the first: its body is above the body of the second in the bullish version, below it in the bearish version. The three bodies draw a V (bullish) or an inverted V (bearish). It only exists once the third candlestick has closed.

Bullish vs. bearish tri-star patterns

Bullish tri-star: signaling a potential reversal at the bottom

After a downtrend, three doji candles with the middle one below the other two. Sellers pushed the price lower one more time, then could not hold it. We test a buy stop above the highest high of the three candles, with a stop under the lowest low and a target at twice the risk.

Bearish tri-star: signaling a potential reversal at the top

After an uptrend, three such candlesticks with the middle one above the other two. Buyers pushed the price higher one more time, then the advance stalled. We test a sell stop below the lowest low, with a stop above the highest high and a target at twice the risk.

These charts are not drawn by hand: they are the median shapes of the cases our bench detected, with the faded candlesticks showing the trend before them.

What the pattern tells traders

It is read as exhaustion. A strong trend needs one side in control; three sessions that end where they started show that neither side is. The gap of the middle candlestick shows the trend still tried, and the third one shows it failed. Classical candlestick analysis reads it as a turning point, but indecision is not a reversal by itself: the next candles decide.

Is the tri-star reliable? Our test results

Not proven on daily charts: the bearish 10-bar move is rated promising, but does not survive our correction for multiple testing. "Moved as expected" means higher 10 bars later for the bullish version, lower for the bearish one.

How to read the table: compare each column with the one next to it. On daily charts, 54.5% of bullish tri-stars were higher 10 days later, against 54.2% of random days on the same assets: the average stock rises a little more often than it falls, and the signal adds nothing to that. The 2:1 trade looks better (42.5% of targets hit against 36.2%, an expectancy of +0.29 R per trade before costs), but the gap in average result per trade does not pass our statistical test. On US stocks alone the answer is the same: rated No clear edge.

The bearish version shows the same picture on daily charts, with a negative expectancy (−0.11 R), slightly less negative than random entries (−0.16 R) but within the margin of error. Its 10-bar move is the one gap worth noting: bearish daily cases fell 0.67% more than their market on average, a gap rated promising that passes a simple test but not our correction, almost all of it on US stocks. On hourly crypto the sell trade reached its target 17.9% of the time against 29.7% on random dates, on only 67 trades, a gap rated weak, below chance: worse than chance on a simple test, not after our correction for multiple testing.

What the breakdowns show

These splits are descriptive and not corrected for multiple testing. Bullish cases on daily charts below their 200-day moving average reached the 2:1 target 41.0% of the time (+0.25 R), against 43.1% (+0.30 R) above it. High volume (more than 1.5 times the 20-day average) gave 43.8% against 42.6% on normal volume. The excess return over the rest of the market was −0.58% in the first half of the history and +0.20% in the more recent half: no stable edge.

How often does it occur?

On daily charts, once every 20,020 eligible bars for the bullish version and once every 14,134 for the bearish one. A single doji appears about once every 7 daily bars. Forex does not produce it in our data: no occurrence of either version, which is why the tables have no forex row, probably because currency pairs trade around the clock and gaps between bodies seldom form.

How to trade the tri-star pattern effectively

Setting entry points and stop-loss levels

On daily charts, 67.4% of bullish setups triggered the entry; in 32.4% of cases price broke the stop side first and the setup was canceled; in the few remaining cases, the order was never filled: neither level was touched within the 20 bars the order stays active, or the data ended first. The whole pattern spans a median of 3.41% of price (1.09 times the average true range), so the stop sits about one average daily range away. Winning bullish trades took a median of 8 days to reach the target, losing ones 4 days to hit the stop.

Using volume and indicators for confirmation

One option is to wait for a confirmation candlestick that closes beyond the pattern, or to combine it with other technical indicators: an oversold RSI under a bullish setup, a moving average above it as resistance. We have not tested an RSI filter. The volume and 200-day average splits above are the confirmation filters we measured, and neither turned it into a tested edge.

Tri-star vs. similar candlestick patterns

• Doji star: a long candle, then one doji that gaps away. On 1,839 daily closed trades, the bullish doji star reached its 2:1 target 37.7% of the time against 36.7% on random dates, rated "No clear edge": its average result per trade (+0.18 R against +0.16 R) is within the margin of error. It is far more frequent than the tri-star, with 6,448 bullish daily occurrences.
• Morning star and evening star: the third candlestick is a long one in the new direction, so the reversal is shown, not just suggested.
• Three stars in the south: three black candlesticks shrinking after a decline, with real bodies. It is much rarer: with its strict definition, our bench finds no daily case at all on our markets.
• Abandoned baby: a single doji isolated by gaps on both sides, shadows included.

Advantages and limitations of the tri-star pattern

It is easy to spot, it works the same way in both directions, and its compact size keeps the stop close to the entry. Its limits are in the data: neither version's average result per trade beat random dates on daily charts by a margin that passes our statistical test; on hourly crypto both versions reached their target less often than random dates (the bearish gap rated weak, below chance, the bullish one no clear edge); the bearish 10-bar gap on daily charts passes a simple test but not our correction; and the gap between bodies can be tiny.

Case studies from real charts

The two galleries in the "Real examples" section, one per version, show real tri-stars picked by our bench, with their actual outcome. The four cases described here come from our bench, on daily charts: gold futures and a Brazil equity ETF for the bullish version, a US financial sector ETF and a US stock for the bearish one, at adjusted prices. Only the EWZ case is also a gallery card. On daily crypto charts the bench finds no tri-star at all, bullish or bearish, with the current rules. They show how setups end, not whether the pattern works.

Bullish case: canceled

Gold futures printed a bullish tri-star from January 11 to 13, 2021. The buy stop sat at the pattern high (1,864.0), with a stop at 1,817.1 and a 2:1 target at 1,957.8. Price never traded above the entry: it fell through the stop level on January 19, down to 1,800.8, so the setup was canceled without a trade, and gold was 0.9% lower 10 days after the pattern.

Bullish case: stopped out

The Brazil equity ETF EWZ printed one from June 21 to 25, 2013. The entry at 23.95 triggered the next day, the ETF rose to 24.59 on June 27, then fell through the stop at 22.49 on July 5: a 1 R loss. EWZ was 3.3% lower 10 days after the pattern.

Bearish case: target reached

The US financial sector ETF XLF printed a bearish tri-star from August 6 to 10, 2010. The sell stop below the pattern low (8.85) filled at the open of August 11, at 8.80, with a stop at 9.05 and a target at 8.32. On August 24 the ETF opened at 8.24, below the target, so the target order filled at that open: a gain of 2.3 R. XLF was 8.3% lower 10 days after the pattern.

Bearish case: canceled

Brown-Forman (BF-B) printed one from November 24 to 26, 2025. Price rose above the stop level (28.11) on November 28 before it traded under the entry (27.35), so the setup was canceled without a trade, and the stock was 8.7% higher 10 days after the pattern.

Frequently asked questions about the tri-star pattern

Summary

• A tri-star is three dojis in a row, the middle one gapping away: bullish after a downtrend, bearish after a rally.
• On 389 daily bullish tri-stars, 54.5% were higher 10 days later, against 54.2% for random days.
• The bullish 2:1 trade hit its target 42.5% of the time against 36.2% on random dates, but its average result per trade (+0.29 R against +0.07 R) is within the margin of error.
• On the 2:1 trade, the bearish version reads No clear edge on daily charts and weak, below chance on hourly crypto.

This guide is educational, not investment advice: you can lose money on any setup.

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 detects each setup with fixed, published rules on 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.