Guide · Candlesticks
How to read candlestick charts
The short answer
A candlestick chart draws each period of trading as one candle built from four prices: open, high, low, close. The body spans open to close, the period's net move; the thin shadows, or wicks, mark the extremes that traded but did not hold. Green means the close sat above the open, red below. A candle carries exactly the same data as a bar chart; it is the readability that differs.
Most candlestick guides are pattern catalogues: forty names, forty thumbnails, and the promise that recall equals foresight. This page works the other way. A candle is a record, a lossless compression of one period's auction between buyers at market and sellers at market, and it can be read the way any record is read: by knowing precisely how it was made. So we start with what the four prices are and how NSE actually constructs them (both ends of an Indian daily candle are about to change mechanism, with dates), then read single candles as auction statements, put context in its rightful place above name-recall, tell the Dojima-to-Nison history straight, and map the famous formations to their dedicated guides.
A candle is the record of one auction
Fix the period, say one trading day, and every trade in it collapses to four numbers. The open is the first price of the period; the high is the most any buyer paid; the low is the least any seller accepted; the close is where the argument stood when time ran out. The candle is just these four numbers drawn so that two quantities become visible at a glance: the net displacement of the period, which is the body from open to close, and the rejected excursions, which are the shadows. A shadow is not decoration. It is a region where trades actually happened, and where one side then produced enough pressure to reprice the market away before the period ended.
Be equally clear about what the candle discards. It does not record the order of events: a daily candle cannot tell you whether the high came before the low, whether the range was built in ten minutes or six hours, or how volume was distributed across it. Two sessions with completely different internal stories can print identical candles. That lost information is recoverable only by dropping to a lower timeframe, which is why timeframe nesting, covered at the end of this guide, is part of candle literacy and not an afterthought.
Candles and bar charts: same data, different legibility
Here is the sentence most candlestick pages will not print: a candlestick chart contains no information a bar chart lacks. A Western OHLC bar encodes the identical four prices, a vertical line from low to high, a left tick for the open, a right tick for the close. Every pattern in the candlestick canon can be defined, tested and traded off bars, because a pattern is a set of relationships between OHLC values, not a property of the drawing style.
What candles add is legibility, and the addition is genuine. Body width against shadow length is visible pre-attentively, before you consciously read anything, so regimes announce themselves: a march of full green bodies reads differently from a fence of long-shadowed spinners even at arm's length. But keep the implication honest. Since the drawing adds readability and not information, any predictive claim made for a candlestick pattern is really a claim about OHLC relationships, and it stands or falls on testing, not on the elegance of the form. That framing will matter twice more on this page: once when we read the history, and once when we read the evidence.
Where the four prices come from on NSE
A candle is only as trustworthy as the process that stamps its four prices, and on Indian exchanges that process is more engineered than most chart-readers realise. The continuous session runs 9:15 am to 3:30 pm. But the daily candle's open is not the first scramble of continuous trade: since 18 October 2010, NSE and BSE have run a pre-open call auction from 9:00 to 9:15, introduced first for the index constituents, in which orders are collected, matched at a single equilibrium price, and that auction print becomes the open. Overnight news is absorbed in one multilateral match instead of a rush of market orders hitting a thin book.
The close is even less "raw". The official closing price of an NSE equity is the volume-weighted average price of the final 30 minutes of trading, 3:00 to 3:30. The number your daily candle closes on is a computed average of the last half hour, not the day's final print, which is why the last tick on an intraday chart and the daily close routinely disagree by a few paise. The averaging exists for a reason: a single closing print would be cheap to push around, and too much hangs off the close, index values, derivative settlement marks, fund NAVs, for it to rest on one trade.
And this machinery is changing on a known date. By a SEBI circular dated 16 January 2026, India is introducing a Closing Auction Session in the equity cash segment: from 3 August 2026, stocks that have derivative contracts will close through a call auction run from 3:15 pm to 3:35 pm, the same order-collection-and-single-match logic as the morning, while stocks outside the framework stay on the half-hour average for now. Modifications to the pre-open session follow from 7 September 2026. For the affected stocks, both ends of the daily candle become auction prints. None of this changes how you read a candle; it changes what the candle's open and close actually are, which is exactly the kind of fact a chart-reader should hold.
Date your sources. Guides that call the close "the last traded price" were never right for NSE equities, and from August 2026 even the half-hour-average description goes stale for derivative-linked stocks. When a candlestick page cannot tell you where its four prices come from, weigh its other claims accordingly.
Reading one candle as an auction statement
Now read a candle the way the market wrote it. Take a stock near ₹500 on the daily: it opens at ₹500 on the auction print, sells off through the morning to ₹484, finds buyers who absorb the entire decline, runs to ₹512 in the afternoon and closes at ₹508. Every measure a candle-reader uses is arithmetic on those four prices.
| Measure | Value | Derived how |
|---|---|---|
| Net move (body) | +₹8 | Close ₹508 minus open ₹500, a net displacement of 1.6 percent of the open |
| Range | ₹28 | High ₹512 minus low ₹484, the full width of the day's fight |
| Lower shadow | ₹16 | Open ₹500 minus low ₹484: the depth sellers reached before the entire excursion was bought back |
| Upper shadow | ₹4 | High ₹512 minus close ₹508: the small give-back at the top |
| Close location | 0.86 of range | Close ₹508 minus low ₹484, divided by the ₹28 range: the close sits 86 percent of the way up the day |
The statement this candle makes: sellers controlled the morning and drove price ₹16 below the open; buyers absorbed all of it and paid up into the close; four rupees of rejected downside for every rupee of rejected upside, and a settlement near the top of the range. That is a hammer-shaped candle, but notice that the name added nothing the arithmetic had not already said. Single candles make a small vocabulary of such statements, and each one is worthless until the chart around it supplies a subject.
| Shape | Anatomy, precisely | What the auction did | Before it means anything |
|---|---|---|---|
| Marubozu (full body) | Body spans nearly the whole range; shadows near zero | One side held control from the first trade to the last | A trend state and a location; a full body into no level is just a wide day |
| Hammer form | Small body near the top; lower shadow at least about twice the body | A deep probe down was fully bought back | A prior decline and a level beneath it; mid-range the same shape is noise |
| Shooting-star form | Small body near the low; upper shadow at least about twice the body | A push higher was fully sold back | A prior advance or overhead supply; otherwise it is a failed spike, not a signal |
| Doji | Open and close within a whisker; body near zero | A round trip: effort with no displacement | A trend to interrupt; inside chop, a doji is simply more chop |
| Spinning top | Small body with long shadows on both sides | Both sides probed; neither could hold the extreme | The next candle; a contested bar only resolves forward |
Context is the meaning
The candlestick canon quietly concedes this point inside its own naming. The hammer and the hanging man are anatomically identical candles; the name, and the implied reading, flips entirely on whether the prior trend was down or up. The taxonomy itself is telling you that the candle does not carry the meaning. The location does.
So the reading order that works runs opposite to the way candlesticks are usually taught. Trend state first: is this market advancing, declining or ranging on the timeframe above the one you trade? Location second: is price at a level where the auction has previously turned, the territory mapped in our guide to support and resistance on the Nifty, or in the middle of nowhere? Participation third: did the candle print on expanded or contracted volume relative to its recent sessions? Only then does the candle itself get a vote, as the timing device that says the level is being defended now. Training that reading order, trend first, level second, participation third, candle last, is exactly what the method we teach is built around.
The history, told straight, and what the evidence says
Candlesticks come with the best origin story in technical analysis, and most of it deserves a closer look. The setting is real: rice traders were licensed at Dojima in Osaka in 1697, and in 1730 the Tokugawa shogunate authorised trading in rice bills on the books of the exchange, which makes Dojima the world's first organised futures market, complete with clearing, margining and price information relayed across Japan by flag signal. Munehisa Homma (1724 to 1803), the rice merchant from Sakata whom the legend crowns as the father of candlesticks, was real too: he traded at scale, and his 1755 text on the psychology of the rice market, the Sanen Kinsen Hiroku, survives as one of the earliest written treatments of market sentiment anywhere.
What the record does not support is the chart. There is no evidence Homma drew candlestick charts; the research, including Steve Nison's own in the book that popularised candles, concludes that the candlestick form most likely developed in Japan's Meiji era, the late 1800s, a full century after Homma. The honest lineage is: an eighteenth-century futures market and a trader who wrote brilliantly about crowd psychology, a nineteenth-century drawing style, and then a long domestic tradition that stayed unknown to the West until Nison's December 1989 article and his 1991 book, Japanese Candlestick Charting Techniques, after which candles became the default rendering on trading platforms worldwide, India included.
The modern evidence deserves the same honesty. When researchers test candlestick patterns the way the pattern books implicitly promise they work, mechanically, in isolation, buy on the signal and hold for a fixed period, the results are sobering: the best-known study, Marshall, Young and Rose (2006) in the Journal of Banking and Finance, ran the standard single- and multi-candle signals across Dow stocks from 1992 to 2002 against bootstrapped price series and found they did not create value, and later replications in other markets, including Japan's, reached similar conclusions. State it qualitatively and precisely: isolated candlestick patterns have weak standalone predictive power. That is not a reason to discard candles. It is the reason the weight in any serious process sits where this guide has been putting it all along: in context, in the levels and trend state that select which candles deserve attention, and in risk control that survives the reads that fail.
Where candle reading breaks. The anatomy assumes a liquid, continuous, two-sided auction. On thin counters a handful of trades can paint long shadows that reflect the bid-ask spread, not rejection. A stock pinned at a circuit limit prints candles with no real auction inside them at all. And gaps mean the next open owes nothing to the last close, so a "confirmed" pattern can be underwater before you can act. A pattern traded mechanically, without a level, a trend state and a pre-set stop-loss, is a coin toss wearing a Japanese name.
The pattern map: families, not flashcards
Once single candles read as auction statements, the named formations organise themselves into a handful of families, each encoding one idea at increasing length. That is the useful way to hold the taxonomy: four ideas, not forty flashcards. Each family below links to its dedicated guide, where the anatomy, the failure modes and the context requirements are worked in full; the multi-candle structures that build from these, flags, triangles, heads and shoulders, live in our guide to chart patterns on Indian stocks.
| Family | Formations | What the family encodes | Deep dive |
|---|---|---|---|
| Single-candle indecision | Doji; long-legged, dragonfly and gravestone variants | Displacement near zero despite a traded range: the auction ended unresolved | What is a doji candlestick? |
| Single-candle rejection | Hammer, hanging man, inverted hammer, shooting star | One extreme probed and refused; the shadow is the evidence | What is a hammer candlestick? |
| Two-candle control shift | Bullish and bearish engulfing, harami, piercing line, dark cloud cover | One period's body overwhelming or hollowing out the prior period's | What is a bullish engulfing pattern? |
| Three-candle sequence | Morning star, evening star, three white soldiers, three black crows | Control, pause, handover: a reversal narrated across three auctions | Chart patterns on Indian stocks |
Every entry in that table obeys the same grammar: define the anatomy from OHLC, translate it into what buyers and sellers did, then demand a location and a trend state before acting. A reader who can do that derivation no longer needs the flashcards; the names become an index, not a skill.
Timeframes: the same auction, nested
Every candle on a higher timeframe contains the candles below it, by construction. The aggregation rule is mechanical: the higher candle's open is the first lower candle's open, its close is the last one's close, and its high and low are the maximum and minimum across the set. An NSE session runs 375 minutes, so one daily candle swallows seventy-five 5-minute candles whole. The ₹500 hammer from earlier is the compressed form of an entire morning of selling and an afternoon of repair; on the 5-minute chart that day looks like a collapse, a base and a recovery rally, three different stories that are all the same data at different resolutions.
| Timeframe | One candle spans | Candles per session | What it answers |
|---|---|---|---|
| 5-minute | 5 minutes | 75 | Where inside the day the fight happened; execution detail |
| 15-minute | 15 minutes | 25 | Intraday structure without tick-level noise |
| 1-hour | 60 minutes | 6 full, plus a 15-minute stub whose placement depends on the platform's anchor | Session rhythm across two or three days |
| Daily | The whole session | 1 | The reference record: auction-set open, official close, the day's verdict |
| Weekly | Usually 5 sessions | 1 per week | Position context: where the larger fight is settling |
Two practical consequences follow. First, lower timeframes recover sequence but multiply noise: each 5-minute candle is a 5-minute auction, thinner and more accident-prone than the day's. Second, the daily candle is the cleanest single record an Indian chart offers, opened by one auction and closed by an official computed price, which is why beginners should learn to read dailies fluently before earning the right to drill down. The hourly stub in the table is a small honesty worth keeping: 375 minutes does not divide by 60, so "the 1-hour chart" is already a platform convention, not a fact of the market. This guide's Hindi companion, candlestick chart kaise padhe, covers the same ground for Hindi-first readers.
Where does that leave the candle in a working process? Exactly where a record belongs: as the final check, not the idea. The idea comes from structure, the level comes from the map, the size comes from a risk budget, and the candle answers one narrow question, is the auction confirming, right here, right now, that the level is being defended? Read that way, a candlestick chart is one of the most information-dense displays ever designed for markets. Read as a deck of magic shapes, it is the fastest route to overtrading yet invented.
Common Questions
Frequently Asked Questions
What is a candlestick chart?
+A candlestick chart plots each period of trading, a day, an hour, five minutes, as one candle built from four prices: the open, high, low and close. The thick body spans open to close and shows the period's net move; the thin shadows above and below mark the extremes that traded but did not hold. A green candle closed above its open, a red one below. Reading a candlestick chart means reading a sequence of these compressed records of buying and selling.
What do the wicks or shadows on a candlestick mean?
+A wick, or shadow, is the part of the period's range that traded but could not hold to the close. The upper shadow spans the high down to the top of the body: buyers pushed price there and sellers repriced it away. The lower shadow spans the bottom of the body down to the low: sellers pressed it there and buyers absorbed the selling. Long shadows record probes and rejections; short shadows record a session that closed near its extremes.
Are candlestick charts better than bar charts?
+They carry identical information. A bar chart encodes the same open, high, low and close with a range line and two ticks; nothing in a candle is missing from a bar. What candles add is readability: the filled, colour-coded body makes conviction and indecision visible across dozens of candles at a glance. That is a real advantage for scanning and for teaching, but it is not extra data, and it gives candlestick patterns no predictive power a bar chart would lack.
What is the difference between a doji and a hammer?
+Both are single candles defined by their shadows, but the anatomy differs. A doji opens and closes at nearly the same price, so it has almost no body: the session ended in a stalemate. A hammer has a small body near the top of its range and a lower shadow at least about twice the body: the session probed sharply lower, the selling was absorbed, and price closed back near the high. A doji records indecision; a hammer records a probe down that failed.
Who invented candlestick charts?
+The popular story credits Munehisa Homma, an eighteenth-century rice merchant from Sakata who traded at Osaka's Dojima exchange, the world's first organised futures market. Homma was real and wrote on market psychology in 1755, but there is no evidence he drew candle charts; research, including Steve Nison's own, places the candlestick form in Japan's Meiji era, the late 1800s. Nison introduced candles to Western traders through a 1989 article and his 1991 book, Japanese Candlestick Charting Techniques.
Do candlestick patterns work on Indian stocks?
+The candle form works on any auction market, and NSE charts are no exception. What the evidence does not support is treating isolated patterns as signals: academic tests that applied candlestick rules mechanically, such as the 2006 Journal of Banking and Finance study on Dow stocks, found they did not create value on their own. A pattern earns meaning from context: the trend it interrupts, the level it forms against and the participation behind it. Read Indian charts that way, pattern last, with risk defined before entry.
Which timeframe should a beginner read candlestick charts on?
+Start on the daily chart. One candle per session keeps the record clean: the open comes from the exchange's opening auction, the close from its official closing price, and the day's whole fight is summarised in one candle. Lower timeframes multiply candles, and noise, without adding new kinds of information; a 5-minute chart of an NSE session prints 75 candles that all live inside that day's single daily candle. Learn to read one candle well on the daily, then drill down when a specific question needs intraday detail.
Why is the close of an NSE daily candle not the last traded price?
+Because the exchange computes it. The official closing price of an NSE equity has been the volume-weighted average price of the final 30 minutes of trading, so the daily candle's close is an average, not the day's last print, and the two usually differ slightly. From 3 August 2026, under a SEBI circular dated 16 January 2026, stocks with derivative contracts will instead close through a closing auction session run from 3:15 pm to 3:35 pm; other stocks stay on the VWAP method for now.
Can I trade using candlestick patterns alone?
+No. A candle is a record of one period's auction, not a forecast, and tests of patterns applied in isolation show weak standalone value. A workable process puts the pattern last: establish the trend state, mark the levels that matter, check the participation behind the move, and only then let a candle time the decision, with the stop-loss and position size fixed before entry. Candles compress information honestly; they do not remove the need for context, risk control or a tested method.
Where the facts come from
Sources
- SEBI circular, 16 January 2026. Introduction of the Closing Auction Session in the equity cash segment: from 3 August 2026, stocks with derivative contracts close through a 3:15 pm to 3:35 pm call auction in place of the last-half-hour average, with pre-open modifications following from 7 September 2026. sebi.gov.in
- NSE market mechanics. The pre-open call auction, introduced alongside BSE on 18 October 2010 for index constituents, sets the session's opening price; the official closing price of an NSE equity is the volume-weighted average of the final 30 minutes of trading.
- Marshall, Young and Rose (2006). "Candlestick technical trading strategies: Can they create value for investors?", Journal of Banking and Finance 30(8): candlestick rules tested mechanically on Dow stocks, 1992 to 2002, against bootstrapped price series did not create value, the anchor of the standalone-power caveat this guide states. sciencedirect.com
- Steve Nison, Japanese Candlestick Charting Techniques (1991). The book that carried candles to the West after Nison's December 1989 article, and the source of the finding that candle charts likely date to Japan's Meiji era rather than to Homma himself.
- Dojima rice market records. Rice traders licensed at Dojima, Osaka in 1697; the Tokugawa shogunate authorised book trading in rice bills in 1730, making Dojima the world's first organised futures exchange. jpx.co.jp
Educational note. This guide explains how a candlestick chart encodes price behaviour and how its prices are formed on Indian exchanges. It is not a recommendation to trade or invest, and it is not investment advice. Bharath Shiksha is an educational publisher, not a SEBI-registered investment adviser or research analyst.