Dragonfly Doji — A Beautiful Candle That Statistically Doesn't Work
A dragonfly doji looks like a textbook buy signal: a long lower wick, zero body, a close right at the top of the candle. Supply pushed price down, demand bought it all back — a beautiful story. The problem is that it's been measured: across 4.7 million candles, the dragonfly reverses the trend 50% of the time. Exactly as often as a coin flip. This article doesn't teach "how to trade a dragonfly" — it teaches why you almost certainly shouldn't.
What a Dragonfly Doji Looks Like
A dragonfly doji is a single candle with three characteristics:
- Open, high, and close at (nearly) the same level — that's what makes it a doji. There's no body at all, or a negligible one.
- A long lower wick — all the action of the candle happened below the opening price: supply pushed price down, demand bought back the entire move.
- No upper wick (or a negligible one) — hence the "T" shape, resembling a dragonfly with a long tail.
The classic interpretation: sellers lost control, buyers took over the market, and at the bottom of a downtrend this is a reversal signal. Sounds logical. Let's see what's left of that logic after testing it.
[Chart coming soon: real BTC/USDT H4 chart from TradingView with the SRL indicator — a dragonfly doji at support: the long lower wick and absent body highlighted, plus an example of a dragonfly after which the decline continued]
What Happens Inside a Dragonfly
Walk through this candle hour by hour and you'll understand why its outcome is random. After the open, supply takes control and pushes price down — sometimes triggering stops below a local support level. Then demand steps in and recovers the entire move, but not a cent more: the close lands exactly where the open was. The session ends in a draw.
And that's the heart of the problem. A hammer tells a story with a punchline — demand not only recovered the losses but seized the initiative, closing the candle above the open. A dragonfly tells a story without an ending: both sides used their ammunition, the market returned to its starting point, and the question of "who wins tomorrow" remains open. A market that ends the day where it started hasn't made any decision — so the formation has nothing to predict. Hence the 50%.
What the Numbers Say — Not Opinions
Thomas Bulkowski tested candlestick formations on 4.7 million candles of US stocks. Results for the dragonfly doji:
| Measurement | Value |
|---|---|
| Trend-reversal success rate | 50% — pure randomness |
| Performance rank (1 = best of 103) | 98/103 |
| Frequency of occurrence | 44/103 |
Test conditions: US stocks, daily timeframe. On crypto (a 24/7 market, higher volatility), treat these numbers as directional guidance — market regime matters.
Read those two numbers again, because they're brutal:
- 50% reversal means the dragonfly is neither a reversal formation nor a continuation one. It's nothing. A coin has identical performance and doesn't require a $500 course.
- Rank 98 of 103 means that even when price does break out in one direction after a dragonfly, the move ranks among the weakest of all formations tested. Fifth from the bottom of the entire table.
Bulkowski himself comments directly: a dragonfly represents indecision, not reversal — and his advice is to simply ignore it. The only curiosities in his data: breakouts from a dragonfly more often go up (because the close is at the top of the candle, so an upward break is close by), and the best results come from dragonflies near yearly lows. Nuances — not an edge.
Notice the frequency too: 44/103 means a dragonfly appears on charts regularly. A formation that's both common and random is the worst possible combination — you get plenty of chances to make a mistake and zero edge. A rare, random formation at least doesn't tempt you every day.
Dragonfly vs. Hammer — Why the "Cleaner" Version Performs Worse
This is the most interesting paradox of this formation. A hammer looks almost identical: a long lower wick, a small body at the top. The difference is cosmetic — a hammer has a small body, a dragonfly has none.
Intuition (and a good chunk of trading courses) suggests: since in a dragonfly demand bought back everything, all the way to the open price, the signal must be stronger than in a hammer. The measurement says the opposite:
- Hammer: reversal 60% of the time.
- Dragonfly doji: reversal 50% of the time.
Where does this difference come from? Most likely from the fact that a hammer's body — a close clearly above the open — is proof that demand not only recovered the losses but seized the initiative. A dragonfly is a draw: the market returned to its starting point and that's it. A draw is indecision, and indecision has no direction. That's why pixel-level precision ("is this still a hammer or already a doji?") has real statistical consequences.
The same pattern repeats on the bearish side of the market: a shooting star (small body) reverses the trend 59% of the time, while its bodyless version — the gravestone doji — does so only 51% of the time. Two pairs of candles, the same pattern: the version with a body has a (modest) edge, the doji version is a coin flip. It's hard to find cleaner evidence that it's the closed body, not the length of the wick, that carries information about who took control.
How (Not) to Trade a Dragonfly Doji
The honest answer is: on its own — not at all. A formation with the performance of a coin flip and a strength rank of 98/103 is not the basis for any setup. This isn't a matter of opinion or "trading style" — it's arithmetic: an entry with 50% odds and a symmetrical stop and target has zero expected value before costs, and negative expected value after spread and commissions. If you still want to use a dragonfly as part of the bigger picture, treat it at most as an observation candle:
1. Only at a strong level
A dragonfly in the middle of a range is pure noise. A dragonfly at strong support from a higher timeframe at least tells you the level was tested and (for now) held. Information about a level — not a buy signal.
2. It doesn't exist without confirmation
If the next candle closes decisively above the dragonfly's high, that candle is the signal — the dragonfly was only the backdrop. You're then trading confirmation and a level, and the doji is decoration.
3. Stop loss and expectations
If you do build an entry from this: stop below the wick's low with a buffer, target at the nearest resistance. And sober expectations — rank 98/103 means that after the breakout, price statistically loses momentum fast. Don't plan the trade of your life after a dragonfly.
⚠ The most common mistake: buying "because a long wick = demand." A long lower wick shows demand defended itself — it doesn't show demand will win the next candle. Half of all dragonflies end in continued declines. Exactly half.
Where a Dragonfly Actually Comes in Useful
To be fair, the formation isn't entirely useless — it has two honest uses, neither of which involves trading it alone:
- As a component of multi-candle formations. A doji (including a dragonfly) can be the middle candle of a morning doji star (76% reversal) or an element of an abandoned baby (70%). These setups have measured, real performance — but the edge there comes from the entire three-candle sequence with confirmation, not from the doji alone. The same candle that's a coin flip solo can, in the right surroundings, be part of something valuable.
- As a liquidity map. A long lower wick shows where the market reached for stops and from where it got bought back. That's a clue about where a defended level sits — useful for planning stops and entries from other signals, even if the dragonfly itself doesn't predict direction.
On crypto there's one more caveat: on low timeframes, dragonfly-style wicks are produced by liquidation cascades and thin order books — they appear dozens of times a day and mean even less than on the stocks in the test. If you're going to look at dragonflies at all, look on D1/H4.
Myth vs. Measurement
Myth: "A dragonfly doji is one of the strongest bullish reversal signals — a close at the top of the candle proves buyers took control."
Measurement: reversal 50% of the time (randomness), move strength after the formation ranked 98/103 (fifth from the bottom). A dragonfly doesn't predict direction — at most it shows that some level was tested. Its bearish mirror, the gravestone doji, performs almost identically. If you're looking for a rejection candle with a real (if modest) statistical edge, trade a hammer with confirmation — leave the dragonfly to the course sellers.
FAQ
Is a dragonfly doji a buy signal?
What's the difference between a dragonfly doji and a hammer?
If a dragonfly is random, why bother learning it at all?
Trader and founder of Strefa Tradingu. He’s been breaking crypto down on YouTube for years — no hype, no signals, with a focus on market structure and risk management.
🎁 Grab Strefa’s free TradingView indicators
Drop your email — we’ll send you links to our free TradingView indicators plus a no-fluff starter kit. Zero spam.
You’re joining the Strefa Tradingu list. Unsubscribe with one click, anytime.Check your inbox (and the Spam/Promotions folders) and add us to your contacts.