How to Read a Crypto Candlestick Chart From Scratch
OHLC anatomy, what bodies and wicks actually mean, how timeframe and timezone change the candle you see, and an honest assessment of how reliable single-candle patterns really are.
On this page
- What a candlestick actually records
- Reading the body and the wicks
- Why your daily candle is not everyone's daily candle
- Choosing a timeframe
- Five single candles worth recognising
- Why candlestick patterns are less reliable than they look
- How to use candles without over-reading them
- Conclusion
- Frequently asked questions
- Sources and further reading
Quick read
A candlestick compresses a stretch of trading into four numbers: open, high, low and close. This lesson explains how to read the body and the wicks, what changing timeframe actually changes, which single candle patterns are worth recognising, and why those patterns are far less reliable than most tutorials admit.
What to remember
- A candle is four numbers and a clock boundary. It records where price opened, how far it travelled in each direction, and where it ended, but not the order in which any of that happened.
- The body is the net result of the period. The wicks are the price levels that were reached and rejected.
- The same trading data produces different candles depending on the timezone your chart anchors to, so a pattern can exist on one chart and not on another.
- Single-candle patterns are context-dependent hints about order flow, not standalone signals. Their published names imply far more precision than they carry.
- Use candles to locate levels where buying or selling was absorbed, and confirm with something that is not derived from price.
What a candlestick actually records
Every candlestick on a crypto chart is a summary of one slice of time, defined by exactly four prices. Exchanges publish this data directly. Binance's endpoint, for example, returns each candle as a record containing open time, open price, high price, low price, close price, base asset volume, close time, quote asset volume, number of trades, and taker buy volumes.
Those first five fields are the whole of the chart. Everything else a candlestick chart appears to show you is a visual encoding of them.
| Price | Definition | What it is evidence of |
|---|---|---|
| Open | The first traded price after the period begins | Where the period started. On a continuous 24/7 market this is simply the previous period's close |
| High | The highest price traded during the period | The furthest buyers were able to push before running out of willing sellers at higher prices |
| Low | The lowest price traded during the period | The furthest sellers were able to push before running out of willing buyers at lower prices |
| Close | The last traded price before the period ends | Where the balance settled at the arbitrary moment the clock ran out |
The critical omission is sequence. A candle does not record whether the high came before the low, how long price spent at each level, or how many times it visited them. Two completely different sessions can produce an identical candle. This is the root of most over-reading: traders infer a story about how the period unfolded from a summary that contains no such information.
Reading the body and the wicks
The body is the rectangle between open and close. It is coloured to show direction: conventionally green or white when the close is above the open, red or black when it is below. Its height is the net move over the period.
The wicks, also called shadows, are the thin lines above and below the body. The upper wick runs from the top of the body to the high; the lower wick runs from the bottom of the body to the low. They represent price that was reached and then given back.
Together they carry one genuinely useful piece of information: the ratio of net movement to total movement.
A worked candle
Take a daily candle with open 100, high 112, low 99, close 101.
- Net move: plus 1, a green body one unit tall.
- Total range: 13, from 99 to 112.
- Upper wick: 11. Lower wick: 1.
One candle, with its four prices marked
A daily candle with open 100, high 112, low 99, close 101. Illustrative values, not market data.
The worked candle: close above open (O 100, H 112, L 99, C 101)
Figure 1 makes the proportion visible: the body between 100 and 101 is a sliver, while the upper wick running to 112 is eleven times its height. So price travelled 12% above its open and retained less than one twelfth of that.
Why this happens. A long upper wick with a small body means buyers had enough force to lift price substantially, and sellers had enough size resting above to absorb all of it and push the close back down. That is a real, mechanical statement about where supply sat. The level near 112 is where someone was willing to sell in size.
What it does not tell you. It does not tell you whether the spike happened in the first hour or the last, whether it was one large order or thousands of small ones, or whether the sellers who absorbed it have anything left. A candle with a long upper wick is frequently described as "bearish rejection," which smuggles in a forecast that the data does not contain.
The actionable version. Treat the wick extreme as a level, not a signal. Note 112 as a price where selling appeared, and watch what happens if price returns there. If it is absorbed again, the level is real and worth trading around. If it goes straight through, the sellers are gone and the wick told you nothing about the future. Levels can be tested; patterns cannot.
Why your daily candle is not everyone's daily candle
This is the caveat that almost no candlestick tutorial mentions, and it undermines a great deal of pattern reading.
Crypto trades continuously, so there is no natural session boundary the way there is in equities. A "daily" candle is therefore whatever your chart provider decided a day starts at. Binance's kline endpoint makes this explicit: intervals default to UTC, and an optional timeZone parameter re-anchors them, with the documentation stating that if a timezone is provided, kline intervals are interpreted in that timezone instead of UTC.
That single parameter changes what patterns exist on your chart.
Working it through
Consider a 24-hour stretch of trading, described by its actual price path rather than by a candle:
- Price is 103 at 16:00 UTC on the previous day.
- It drifts down to 100 by 00:00 UTC.
- Between 00:00 and 12:00 UTC it sells off hard to 92, then recovers.
- By 16:00 UTC it is back at 99.
- It closes the UTC day at 101, having peaked at 101.
Now draw the daily candle two ways.
| Chart anchor | Open | High | Low | Close | What it looks like |
|---|---|---|---|---|---|
| UTC day, 00:00 to 24:00 | 100 | 101 | 92 | 101 | Small green body at the top of a long lower wick. Textbook hammer |
| UTC plus 8 day, 16:00 to 16:00 | 103 | 103 | 92 | 99 | Red body closing well below the open with a long lower wick. Not a hammer |
Why the two disagree. Nothing about the market changed. The only difference is where the clock cut the data. The UTC candle happens to end at the top of the recovery, so the close sits near the high and the pattern qualifies as a hammer. The UTC plus 8 candle ends eight hours earlier, before the final leg up, and it also starts eight hours earlier at a higher price, so the same recovery reads as a failure to reclaim the open.
What this actually tells you. A candlestick pattern is not a property of the market. It is a property of the market and an arbitrary clock offset. Any pattern whose classification depends on where the close falls relative to the range is exposed to this, which is most of them: hammers, shooting stars, dojis and engulfing patterns all depend on close position.
The actionable version. Check what timezone your charting tool uses before you rely on any daily pattern, and be aware that a large share of the crypto market watches UTC while a large share of retail flow watches local exchange time. If a pattern only appears on one anchor, it is not evidence. Levels, by contrast, survive re-anchoring: 92 was the low and 103 was the high regardless of how you slice the day, which is another reason to prefer levels over patterns.
Choosing a timeframe
Changing does not reveal more information. It changes how much is aggregated away.
| Timeframe | What survives aggregation | What is lost | Typical use |
|---|---|---|---|
| 1 minute to 15 minutes | Individual bursts of order flow and short-lived liquidity gaps | Any sense of the prevailing direction | Execution timing on an already-decided trade |
| 1 hour to 4 hours | Intraday swings and the levels that held during them | Micro noise, and also genuine short-lived events | Swing entries and stop placement |
| 1 day | The structure most participants and most published analysis reference | Everything about how the day unfolded | Trend context and level identification |
| 1 week and above | Only the largest structural moves and multi-month ranges | Almost all tradable detail | Position framing and cycle context |
The practical rule is that shorter timeframes have a worse ratio of signal to noise, not more signal. On a one-minute chart, a single moderately sized can create a candle that looks identical to a meaningful rejection on a daily chart. The pattern vocabulary is the same; the information content is not.
A second consequence matters for anyone using patterns: shorter timeframes generate far more pattern occurrences, which feels like more opportunity and is actually more false positives. If a pattern has weak predictive power, seeing it forty times a day does not improve it. This is the covered in how to read crypto market signals.
Five single candles worth recognising
These are worth learning because they describe order flow compactly, not because they predict. Each definition below states what must be true of the four prices, and each is followed by what it actually implies.
Doji
Definition: open and close are approximately equal, so the body is very small or absent, with wicks on one or both sides.
What it implies: the period ended where it started despite trading in a range. Buyers and sellers were closely matched at the boundary moment. Nothing more.
Doji
Open 100, close 100.1, with wicks on both sides. Illustrative values.
Doji: close above open (O 100, H 104, L 96, C 100.1)
The honest caveat: "approximately equal" has no standard threshold. Whether a candle is a doji depends on the tolerance your platform or your eye applies. In fast crypto markets, a body of 0.3% may be visually indistinguishable from zero on a chart while representing a meaningful move. Figure 2 is drawn at the unambiguous extreme; the ones you will actually meet sit much closer to the line.
Hammer
Definition: small body near the top of the range, long lower wick, little or no upper wick, appearing after a decline.
What it implies: price traded significantly lower during the period and buyers absorbed it, returning the close to near the high. The low is a level where buying appeared.
Hammer
Two declining candles, then a small body at the top of a long lower wick. Illustrative values.
Prior: close below open (O 108, H 109, L 103, C 104) · Prior: close below open (O 104, H 104.5, L 99, C 100) · Hammer: close above open (O 99.5, H 100.5, L 93, C 100)
The honest caveat: the "after a decline" condition is doing enormous work and is almost never specified precisely. Figure 3 draws that condition as the two faded candles, and they are the part that is usually missing in practice. The identical shape appearing mid-range is not a hammer in the traditional sense, but charts are full of the shape, and the classification is applied retrospectively once the outcome is known.
Shooting star
Definition: the hammer inverted. Small body near the bottom of the range, long upper wick, appearing after an advance.
What it implies: price was pushed up and sold into. The high is a level where supply appeared.
Shooting star
The hammer inverted, after an advance rather than a decline. Illustrative values.
Prior: close above open (O 92, H 97, L 91.5, C 96) · Prior: close above open (O 96, H 101, L 95.5, C 100.5) · Shooting star: close above open (O 101, H 108, L 100.5, C 101.3)
The honest caveat: identical to the hammer's. Figures 3 and 4 are the same geometry reflected, which is worth pausing on: the shape alone carries no direction, and everything distinguishing the two names lives in the faded candles beside them. It is also the single most common shape produced by a brief on a thin book, where a small order momentarily lifts price with no meaningful supply story behind it.
Engulfing candle
Definition: a candle whose body completely covers the previous candle's body in the opposite direction. Bullish engulfing means a green body covering a prior red body; bearish engulfing is the reverse.
What it implies: the period reversed the entire net result of the previous period. On its own that is just a large move in the opposite direction.
Bullish engulfing
Day 2's body fully covers Day 1's body in the opposite direction. Illustrative values.
Day 1: close below open (O 104, H 104.8, L 100.8, C 101) · Day 2: close above open (O 100.6, H 105.6, L 100.2, C 105)
The honest caveat: in crypto, engulfing candles are extremely common because volatility is high and consecutive periods routinely exceed each other. A pattern that occurs constantly cannot carry much information per occurrence. Note also how modest the move in Figure 5 actually is, about 4%, which on a daily crypto candle is unremarkable.
Marubozu
Definition: a large body with no meaningful wicks. Price opened at one extreme and closed at the other.
What it implies: genuinely one-directional trading with no material pushback at any point. Of the five, this is the least ambiguous statement about order flow.
Marubozu
Open at the low, close at the high, no wicks at all. Illustrative values.
Marubozu: close above open (O 100, H 107, L 100, C 107)
The honest caveat: it describes what happened, not what happens next, and it is frequently the last candle of a move rather than the first candle of one. Compare Figure 6 with Figure 1: the same four numbers describe both, and the entire difference in what they tell you is where the close sits relative to the range.
Why candlestick patterns are less reliable than they look
Four structural problems apply to all of them, and they compound.
The definitions are subjective. How small is a small body? How long is a long wick? How far back does "after a decline" extend? Every one of these thresholds is a free parameter, and a pattern with three free parameters can be tuned to fit whatever past data you point it at. This is the data-mining problem described in how to read crypto market signals.
The classification is usually retrospective. In live trading you do not know a candle is a hammer until it closes, and you do not know it "worked" until price moves. Reviewing a chart afterwards, the eye finds the hammers that preceded rallies and slides past the identical shapes that preceded further declines. Nothing about this feels like bias while you are doing it.
They depend on a clock boundary that is arbitrary in a 24/7 market. As the timezone example above shows, the same trading produces different patterns under different anchors. Patterns inherited from Japanese rice markets and adapted to equities assume a session with a genuine open and close, a structure crypto does not have.
The base rate problem applies with full force. Common patterns occur constantly. If a bullish engulfing candle appears on a daily chart dozens of times a year and major rallies happen a handful of times a year, most occurrences necessarily precede nothing.
How to use candles without over-reading them
Steps
Read the range before the pattern
Look at the high and the low first. Those are the two levels where the market found a boundary, and they survive changes of timezone, platform and pattern definition. The body colour and the pattern name do not.
Ask which exchange drew the candle
Wicks that appear on one venue and not others usually reflect that venue's thin liquidity rather than a market-wide event. Before treating an extreme as a level, check whether a second exchange recorded it too.
Check the same candle one timeframe up
A dramatic pattern on a 15-minute chart frequently disappears into the middle of an unremarkable 4-hour candle. If the higher timeframe shows nothing, you are reading noise with a name attached.
Confirm with something that is not price
Volume, open interest, or funding are separate measurements with different failure modes. Two price-derived observations agreeing is one observation. A candle plus a genuine second dimension is two.
Trade the level, not the shape
Instead of entering because a hammer printed, mark the hammer's low as a level and decide in advance what you will do if price returns to it. This converts an untestable pattern claim into a testable level claim.
Conclusion
A candlestick is an efficient compression of a period of trading into four prices, and it is genuinely worth learning to read. The body tells you the net result, the wicks tell you which levels were reached and given back, and the ratio between them tells you how much of the period's movement was retained. That is real information about where buying and selling appeared.
What a candle cannot do is tell you the sequence of events inside the period, and it cannot tell you what comes next. Two entirely different sessions produce the same candle, and the pattern vocabulary attaches confident directional names to shapes that are consistent with many underlying stories.
The crypto-specific problem is sharper still. Because the market never closes, the daily candle depends on an arbitrary timezone anchor, and the same trading data yields a textbook hammer on one anchor and an ordinary red candle on another. Any pattern defined by where the close sits within the range inherits this fragility, which is most of them. Levels do not: the high and the low are the same numbers however you cut the day.
The practical conclusion is to demote patterns and promote levels. Use candles to identify prices where supply or demand actually appeared, mark those levels, and decide in advance what you will do when price returns to them. Confirm with volume or a structural measure rather than with a second price-derived indicator. And when a pattern name feels like it is telling you what happens next, remember that it is describing what already happened, in a shape whose definition you chose.
Frequently asked questions
Read the four prices first. The body spans the open and close and shows the net result of the period, coloured by direction. The wicks extend to the high and the low and show the prices that were reached and given back. The ratio of body to total range tells you how much of the period's movement was retained by the end of it.
Open, high, low and close. These are the four prices that define every candle: the first trade of the period, the highest and lowest traded prices during it, and the last trade before it ends. Exchange APIs return these together with volume and trade count for each period.
That price reached that level and was pushed back before the period ended. A long upper wick means buyers lifted price and sellers absorbed it; a long lower wick means the reverse. The wick extreme is best treated as a price level where opposing size appeared, not as a directional prediction.
There is no universally best one, and shorter timeframes contain more noise rather than more information. Daily candles are the most widely referenced structure. Hourly and four-hourly are common for swing entries. Minute charts are mainly useful for timing an execution once the decision is already made.
Each exchange builds candles from its own trade history, so the open, high, low and close differ between venues. On thin pairs one exchange can print a large wick that no other venue recorded. Charts also differ by timezone anchor, which changes where each daily candle starts and ends.
Less reliable than their names imply. Pattern definitions contain subjective thresholds, classification is usually applied retrospectively, common patterns occur far too often to carry much information per occurrence, and in a 24/7 market the pattern itself depends on an arbitrary timezone boundary. Treat them as descriptions rather than forecasts.
A candle whose open and close are approximately equal, leaving a very small or absent body. It says the period ended where it started despite trading in a range. There is no standard threshold for how small the body must be, so whether a candle counts as a doji depends on the tolerance you or your platform apply.
They are the same shape inverted. A hammer has a small body near the top of its range with a long lower wick and appears after a decline. A shooting star has a small body near the bottom with a long upper wick and appears after an advance. Both depend heavily on the preceding context, which is rarely defined precisely.
Not on its own. A bullish engulfing candle means one period's green body fully covered the previous period's red body, which in a volatile market simply means a large move in the opposite direction. Because crypto volatility makes consecutive periods routinely exceed each other, the pattern occurs too frequently to carry much weight alone.
Candles for analysis and lines for context. A line chart plots only the closing price, which hides the range entirely and therefore hides every level where price was rejected. Candles show the same closes plus the highs and lows, which is where most of the useful information sits.
Sources and further reading
Primary sources:
Related CoinBeaver articles:
- How to read crypto market signals
- What trading volume tells you
- What the RSI indicator can and cannot tell you
- How to read a crypto order book
- Crypto order types explained
This article is educational and is not financial advice. The candle values used in the worked examples are illustrative numbers chosen to demonstrate how the four prices combine, not observations of any real market. Peer-reviewed studies of candlestick pattern profitability exist in the equities literature, but their full texts sit behind publisher paywalls that could not be opened and verified at the time of writing, so no specific win rates or study results are quoted here. Treat any pattern reliability percentage you encounter elsewhere as unverified unless it states the exchange, the timezone anchor, and the exact numeric thresholds used to classify each pattern.
Keep learning
Recommended next reads based on this lesson.
- How to Read Crypto Market Signals Without Fooling YourselfThe three families of crypto signals, why a signal with an 80% hit rate is right only 31% of the time it fires, and how to combine signals honestly.
- RSI in Crypto: What It Measures and Its Real LimitsWhy RSI 70 means a trend exists rather than an exhausted one, why RSI carries no volatility information, and why your reading differs from everyone else's.