Dragonfly Doji at Support: How to Read It

Why does one candle attract so much attention?

Candlestick patterns are among the first chart tools that beginners learn, and the dragonfly doji is one of the most discussed. Its shape is distinctive: a long line hanging below a very small body, with almost nothing above it. Seen on a chart, it looks like a clear message that price was tested lower and rejected, and that feeling is why the candle is so often shown in educational material.

The distinctive shape is also the reason to be careful. A single candle is a small piece of information. It shows what happened inside one period, the open, high, low and close, and nothing about what will happen next. A trader who treats the shape as a signal without measuring it, or without checking what follows, is building on a foundation that cannot be tested.

This guide explains how to read a dragonfly doji at a support level, how to measure its wicks, why confirmation matters and how to log the pattern so the results can be judged. It does not promise that the candle will reverse price, and it does not claim that the shape works in every market. It treats the dragonfly as a hypothesis to test.

What is a doji, and where does the dragonfly fit?

A doji is a candle whose open and close are at or very near the same price. Its body is therefore very small, and the candle is defined mainly by its wicks, which are the thin lines above and below the body. A doji shows indecision in the period: buyers and sellers ended the period at about the same price, after pushing the price in both directions.

A dragonfly doji is a doji whose body sits at the top of the range, with a long lower wick and little or no upper wick. The lower wick shows that price moved down during the period, and then recovered to close near the high. Other doji shapes, such as the gravestone, have a long upper wick and a body at the bottom, and they carry the opposite message.

The dragonfly is therefore a rejection shape. It says that the market tried lower and failed to keep the lower price. Whether that rejection matters depends on where it occurs, what follows it and how often it appears in your sample.

How do you measure the wicks and the body?

Measure the candle in four steps. First, find the range, which is the high minus the low. Second, find the body, which is the absolute difference between the open and the close. Third, find the upper wick, which is the high minus the higher of the open and close. Fourth, find the lower wick, which is the lower of the open and close minus the low.

Express each measurement as a percentage of the range, so the same rule works for a candle of any size. A common dragonfly rule requires the body to be below ten percent of the range, the lower wick to be above sixty percent of the range, and the upper wick to be below ten percent. These figures are a starting point, and you should choose and record your own thresholds before testing.

The included log calculates these measurements for each candle you enter. It flags a dragonfly when all three conditions are met. The flag is a count, not a judgement. Whether a candle is a meaningful pattern depends on the level it forms at and the candles that follow.

What does “at a support level” mean?

Support is a price level where buying has repeatedly appeared in the past, causing price to stop falling and turn upward. A dragonfly at support is one that forms when price tests a support level and rejects it within a single candle. The reference level is the important part. A dragonfly in open space, with no level nearby, has no defined context to test.

Mark support levels before you look at the candle. Use a method you can repeat: a previous swing low, a round price, or a horizontal zone where price has reacted at least twice. Draw the level on the chart and note it in the journal with its date. If you draw the level after the dragonfly has formed, you are fitting the chart to the pattern, and the test loses its value.

The guide to identifying support and resistance explains how to draw levels consistently. For the dragonfly test, the rule is simple: the level is fixed before the candle, and the candle is logged against it.

Why does confirmation matter?

A dragonfly is incomplete evidence. It shows that price was rejected within its own period, but it does not show that the rejection will hold. The next candle is the first test. If price closes above the high of the dragonfly, the rejection has been followed by buying. If the next candle closes below the low of the dragonfly, the support has failed for now.

Some traders wait for a close above the dragonfly high before entering. Others require a close above a nearby swing high, or a move that breaks a short-term trendline. Each method is a different confirmation rule, and each has its own delay and cost. The rule you choose should be written in the journal before the sample starts.

Confirmation also protects against a common error: entering on the candle itself, before the period has closed. A dragonfly can look convincing in the middle of a candle and then close differently. Wait for the candle to close, and log the entry only after the close.

How do you set the stop and the target?

The stop for a dragonfly trade is usually placed below the low of the candle, because a move below that low means the rejection has failed. Some traders place the stop a fixed number of pips below the support level instead, which can be tighter or wider depending on the level and the candle.

The target depends on the approach you test. A simple method sets the target at twice the stop distance. A structured method sets the target at the next resistance level above the entry. Whichever you choose, measure the distance to the stop and the distance to the target before the trade, and make sure the reward is worth the risk.

The risk-reward guide explains how to make that comparison. A dragonfly with a long lower wick can produce a stop that is wide in pips, and a wide stop means a small lot size for the same risk. Do not let the size of the candle decide the size of the trade; let the risk per trade decide it.

What does a demo test of the dragonfly look like?

A fair demo test records every dragonfly that forms at a marked support level on the chosen pair and timeframe, using the thresholds you wrote. It also records a comparison group of ordinary candles at the same levels, so you can see whether the shape adds information beyond the level itself. Without the comparison group, a good result may simply reflect the level.

For each dragonfly you record the timeframe, the support level, the body, the wicks, the confirmation rule and the outcome. The outcome should be measured in R-multiples, using the stop and target set before the trade. Include the spread and any swap in the cost of each trade, so the result reflects what you would keep.

Judge the sample only after it is large enough to be meaningful. Thirty trades is a sensible minimum for a first look, and a sample that is split by timeframe or by pair will need more. A few good dragonflies can look like a strategy until the sample grows.

What are the common mistakes with dragonflies?

The first mistake is calling any long lower wick a dragonfly. The shape must meet the thresholds, and a candle that has a body of twenty percent of the range is not a dragonfly under a ten percent rule. Measure every candle you call a dragonfly, and log the ones that almost qualify separately.

The second mistake is trading the dragonfly without a level. A long lower wick in the middle of a trend, far from any support, is much less informative than one at a level that has held before. Record whether the candle formed at a level or in open space, and compare the results.

The third mistake is ignoring the candle that follows. Many dragonflies look convincing until the next candle closes below the low, and then the trader wonders why the signal failed. The next candle is part of the signal, and the rule for it must be written before the trade. The fourth mistake is moving the stop after entry to avoid being stopped out on a wick. A stop is a decision made before the trade, and moving it removes the test.

How does the dragonfly compare with other reversal shapes?

Many reversal shapes exist, including the hammer, the inverted hammer, the engulfing candle and the morning star. Some look similar to the dragonfly, and the difference is often in the context: where the candle forms, what came before it and what follows. A hammer usually appears after a decline, with a body near the top of the range, and the rules for it are often written differently.

Compare the dragonfly with random candles of a similar range. If a dragonfly at support performs no better than an ordinary candle at the same support, the shape is not adding value, and the level is doing the work. That finding is useful, and it should change how you use the pattern.

Reversal shapes are better tested in batches than individually. A sample of one pattern type on one pair and timeframe will show more clearly whether it works than a mixed collection of shapes. Start with one shape, measure it carefully, and only then compare it with others.

How do dragonflies behave across trading sessions?

A dragonfly formed during a quiet session can mean something different from one formed during the open of a busy one. Thin trading in the early hours can produce long wicks because a few orders move the price, and those wicks may not reflect a real change in the balance between buyers and sellers. A dragonfly formed when two major sessions overlap has a larger pool of orders behind it.

Record the session for each candle in your log. A simple split into three groups, such as Asian, European and US hours, is enough for a first test. Compare the outcomes between groups after the sample is large enough. If the dragonflies in one session perform very differently from the others, that is a finding worth investigating, not a reason to trade only that session before the evidence is in.

Session effects can also come from scheduled news. A candle with a long lower wick that forms in the minutes around a release may be the product of a spike rather than of rejection at a level. Mark any dragonfly that forms within thirty minutes of a high-impact release, and keep those rows separate in the summary, so the news effect is visible.

What does a worked dragonfly example look like?

Imagine a dragonfly on the one-hour chart of EURUSD. The candle opens at 1.0912, rises to 1.0921, falls to 1.0884 and closes at 1.0915. The range is 37 pips. The body is 3 pips, which is about 8% of the range and passes a body threshold of ten percent. The lower wick runs from 1.0884 up to the lower of the open and close, 1.0912, so it is 28 pips, about 76% of the range. The upper wick is 6 pips, about 16% of the range, which fails an upper threshold of ten percent.

On this measurement the candle does not qualify under the rule set out in this guide. The upper wick is too long, and the log will show it as a near-miss, not a dragonfly. A trader who had not measured it might have called the candle a clear rejection and entered on the next candle. The measurement prevents that error, and the near-miss goes into the separate list for later comparison.

Now consider a second candle with the same range of 37 pips, a body of 2 pips, a lower wick of 31 pips and an upper wick of 4 pips. The body is about 5%, the lower wick about 84% and the upper wick about 11%. The upper wick still narrowly fails a ten percent threshold. The rule you chose decides the result, and the sample will show whether the rule is too strict. The point of the example is that a rule must be written in numbers before you look at the candle, or each decision becomes a matter of mood.

How do you review a month of dragonfly records?

At the end of each month, review the log in three passes. In the first pass, count the candles you entered, the dragonflies that met the rule, and the near-misses. If the near-misses are numerous, your thresholds may be too strict for the instrument, but change them only at the start of a new sample, never in the middle of one.

In the second pass, check the outcomes. For each dragonfly that met the rule and reached confirmation, record whether price closed beyond the high or the low, and the R-multiple of the trade. Compare the results with the comparison group of ordinary candles at the same levels, because a dragonfly with a good result but an ordinary comparison group is not showing an edge.

In the third pass, read the losing trades. For each one, ask whether the level was clearly marked before the candle formed, whether the confirmation rule was followed, and whether the stop sat at the invalidation point. The answers tell you whether the problem is the pattern, the level or the execution. Write the conclusion in the journal in one or two sentences, so the next review starts from a clear record.

How do you judge whether a dragonfly adds information?

The central question of the test is whether the shape adds anything beyond the level it forms at. If price reacts to a support level whether or not a dragonfly appears, the level is doing the work, and the candle is only a marker of an event that was already likely. The test is designed to separate those two explanations.

The practical method is to compare two groups. The first group contains dragonflies at marked support levels, with confirmation applied. The second group contains ordinary candles that also touch the same levels and close in the same way, but do not meet the wick and body rule. If the first group shows a clearly better outcome, in R-multiples and in consistency, the shape may be adding value. If the two groups look the same, the shape adds little.

Be careful with the size of the groups. A comparison based on ten trades in each group will move a great deal with a single result, and it will not separate a real effect from chance. Aim for a sample large enough that one or two trades cannot change the conclusion. If the groups are still too small after several months, the honest answer is that the question is not yet answered.

What does a single dragonfly trade plan look like?

A written plan for one dragonfly trade has seven lines. The first line names the pair, the timeframe and the support level, with the date the level was marked. The second line gives the candle measurements, the body, the lower wick and the upper wick, with the rule each one meets. The third line states the confirmation rule and the price at which it is met.

The fourth line sets the entry, which is the open of the candle after confirmation. The fifth line gives the stop, which is placed at the invalidation point, and the sixth gives the target, with the reward and the risk in pips. The seventh line sets the lot size, calculated from the risk per trade in your plan and the stop distance in pips.

Write the plan before the entry and record the outcome after the exit. A plan written in this form makes the review straightforward, because every trade has the same fields. It also stops a common error, where a trader enters on the feeling that the candle looks strong and then chooses a stop afterwards to suit the entry.

What happens when the support finally breaks?

Every support level eventually breaks, and the test has to account for that. A level that has held several times can give way on a larger move, a change in the outlook for the currencies, or a release that the market did not expect. A dragonfly at such a level can be the last rejection before the break, and the trade that follows will then lose. The test should record those losses in full, not treat them as anomalies.

A sensible response when a level breaks is to stop trading that level. Note the date and the reason in the journal, and mark the level as broken. A broken support can become a resistance, and a dragonfly formed below it on a later retest has a different meaning from one formed above it. Record the change in the level’s status, so you do not treat a broken support as if it still held.

The break is also useful information for the test. If dragonflies at a particular type of level, such as a round number, fail more often than dragonflies at a previous swing low, that difference should show in the summary. Splitting the sample by level type is one of the most useful checks available, and it is easy to do if the level type is recorded for every entry.

How does the trend change a dragonfly’s meaning?

A dragonfly at support during a steady downtrend is a different event from one formed during a range or an uptrend. In a downtrend, the long lower wick may be a brief pause before the decline continues, and the level may give way on the next wave. In a range, the same candle may be a clean rejection of the lower boundary, and the opposite boundary becomes the natural target.

Tag each entry in the log with the trend on a higher timeframe at the moment the candle closes. A simple tag of up, down or range is enough. Over time, the tags show whether the dragonfly performs differently in each condition. A pattern that works only against a strong trend is still useful information, but the condition belongs in the rules for the method.

Be careful not to use the trend tag to explain away every loss. If you find yourself saying that a dragonfly failed because the trend was strong, check whether the same reason was available before the trade. A condition that you can see only after the outcome is not a condition, and it will not help you choose trades in the future.

What does this guide not promise?

This guide does not promise that a dragonfly doji will reverse price, and no single candle can make that promise honestly. A support level can fail, a news release can overrun any pattern, and a trend can carry price through a level that appeared to hold. A candlestick is a description of one period, and the market does not need to follow any description.

The guide also does not describe a complete trading plan. A plan that includes a dragonfly needs the level definition, the candle thresholds, the confirmation rule, the stop, the target, the position size, the news rule and the review routine. Each one affects the result, and leaving any of them out makes the method hard to judge.

What is the offer in this guide?

The offer is the dragonfly doji measurement log, a free spreadsheet that calculates the range, the body and the two wicks as percentages for up to one hundred candles. It flags a dragonfly when the thresholds you set are met, and it has a column for the confirmation outcome. You can change the thresholds in the formulas if your rule differs from the one described here.

Use the log for every candle you consider, including the ones you reject. Enter the support level and the timeframe, and record the next candle’s close against the dragonfly’s high and low. Keep the file with your demo journal, and review it after every thirty entries.

The log is educational. It measures the candles you enter and cannot tell you which pattern will work. Check each calculation against your own arithmetic the first time you use it, and check the thresholds against the examples in this guide.

How do you start this week?

Start by writing your dragonfly rule on paper. Choose the body threshold, the lower wick threshold and the upper wick threshold, and write the confirmation rule in one sentence. Keep these fixed for the whole test.

Next, pick one pair and one timeframe. On a historical chart, mark five support levels using your method, then scan the candles at those levels for dragonflies. Enter each candle in the log, even if it does not qualify, and note the ones that almost qualify. Do this before you open any demo trade.

Then open a demo account with the spread and lot size you plan to trade live. Trade only the dragonflies that meet your rule, and only after the confirmation candle has closed. Record every trade, and judge the pattern after at least thirty entries. For a wider view of how the market moves before you test any shape, read the pip guide, and the forex basics guide for the terms used in the log.

For more free tools and checklists, visit the resource library.