How to Read a Candlestick: Open, High, Low, Close and What Color Means

Michael Whitman · Trading since 2017, full-time since 2020 · YouTube creator, 137K subscribers · Published 2026-09-05 · Updated 2026-10-27 · Reviewed 2026-10-27

Covered on this page: TradingView.

A candlestick shows four prices for one period: the open, the high, the low and the close, with a thick body from open to close and thin wicks out to the high and low. Green means the period closed above its own open, not above the previous close, and on 17.29% of SPY trading days since 1993 those two disagreed.

Candlesticks are what every price chart is built from, so it is worth being exact about what one actually contains, and about the things people read into them that are not there.

How to read one

A candlestick is four prices for one slice of time.

The open is where the period started. The close is where it ended. The high and low are the furthest price reached in each direction. The thick part is the body, spanning open to close, and the thin lines are wicks, reaching out to the high and the low.

If you are on a one-hour chart, each candle is one hour of trading compressed into those four numbers. Everything else on the chart is drawn on top.

Green and red: color measures open to close, not up or down

Green means the close was above the open. Red means it was below. That is the entire meaning, and it is worth stating because beginners routinely read color as strength or as a signal.

It is not the same as the day being up. A market that gaps down at the open and then climbs all session prints a green candle on a down day. One that gaps up and drifts lower prints a red candle on an up day.

A long wick is a rejection

A long wick points at a price the market visited and would not stay at. It tells you where the rejection happened, which is useful, and it says nothing at all about what comes next.

A candle with almost no body covered its range and ended where it started. On SPY’s daily chart about one day in ten looks like this, with a body no bigger than a tenth of the range, so it is an ordinary sight rather than an event.

The candle on the right is still moving

The rightmost candle is not finished. On a one-hour chart it is being built for the whole hour. Its high and low only ever grow, and its close is wherever price happens to be right now.

That means a candle can look like a perfect rejection at minute forty and close as something else entirely at minute sixty. This is the concrete reason a break needs a close beyond a level rather than a touch.

A candle is an arbitrary box

Nothing in the market knows when your candle starts or ends. The boundaries come from the timeframe you happened to open. A dramatic wick on the five-minute chart may not exist at all on the one-hour, because the hour closed somewhere else entirely.

So a candle is not an event. It is a summary, and you chose the length of the summary. The same session is one daily candle or seventy-eight five-minute candles, and every indicator built on them, from a moving average of closes to an RSI, changes with that choice.

Where it printed matters more than its shape

Two identical candles, one at a level that already mattered and one in the middle of nowhere. The shape is the same and the meaning is not. A rejection candle at a level is a rejection of that level; the same candle floating in open space is a period of trading.

This is why candle shapes learned on their own tend not to survive contact with a real chart, and why price action, reading a chart from the bars alone with no indicator on it, is harder than it sounds: the bars do not carry their own context.

Volume says how much actually traded. The same candle on heavy volume and on light volume are different events, and the volume panel is right there under the chart by default.

What counts as heavy changes over time: SPY’s typical session saw 256.7 million shares change hands in 2008 but only 54.6 million in 2026, as the trading volume page shows year by year.

A worked example

Picture a candle with a long lower wick (often called a hammer) and read it honestly.

What you know. Price opened, fell a long way, came back, and closed near its open. That is not interpretation; it is the four numbers.

What you cannot infer. Anything about the next candle. The rejection is a fact about the period that just finished, not a statement about the one that has not started.

What makes it usable. The level it rejected from, whether that level existed before today, and whether the volume was unusual. Three checks that live outside the candle itself.

One real candle: SPY on 18 September 2026

SPY opened at $761.31, traded up to $762.00 and down to $757.97, and closed at $761.69. The range was $4.03. The body, $0.38, was 9.4% of it; the lower wick, from $761.31 down to $757.97, was $3.34, or 82.9% of the range; the upper wick was $0.31, or 7.7%. By shape, a textbook hammer, and green.

The day before had closed at $762.60, so the fund finished $0.91 lower on a green candle. Both facts are true at once: buyers won the session from the open, and the day still lost ground from yesterday’s close.

The original data

How SPY’s daily candles look

We measured every SPY daily candle from 29 January 1993 to 25 September 2026: 8,472 bars, all of them valid (high at or above the open and close, low at or below both, high above low). Prices are Yahoo Finance’s unadjusted daily open, high, low and close.

4,431 candles were green (52.3%), 3,956 red (46.7%), and 85 opened and closed at the same price. The median body filled 46.9% of its candle’s range, and 854 candles (10.1%) had a body of a tenth of the range or less.

Long lower wicks said nothing measurable about the next day. We defined one as a lower wick of at least 60% of the range with an upper wick of at most 10%, which picked out 383 candles (4.5%). The next close was higher after 202 of them (52.7%), against 4,540 of 8,471 (53.6%) after every day. Five days on it was 236 of 383 (61.6%) against 57.9%. Neither gap is distinguishable from chance (two-sided binomial p = 0.76 and 0.15). The year-by-year counts are in the SPY candle table.

Color against change, timeframe by timeframe

The new question: how often does a candle’s color disagree with the change from the previous candle’s close, and does it depend on the timeframe? We counted every SPY bar whose color pointed the other way from its change: green on a lower close, or red on a higher one. Bars run to 26 Oct 2026; the weekly and monthly series stop at the last complete week and month.

The finding: the disagreement lives in the gap between one bar’s close and the next bar’s open, and the daily chart is where that gap matters most. Inside a session one bar opens about where the last one closed, so color and change almost always agree: on the hourly chart the color disagreed on just 17 of 4,308 bars (0.39%) that followed a bar from the same session, against 224 of 708 first bars of a session (31.64%), which open after the overnight gap. On the 5-minute chart the split was 62 of 4,620 (1.34%) against 23 of 59 (38.98%). On the daily chart every bar starts after an overnight gap, and the rate there, 17.29%, was more than twice the weekly 7.56%.

It has also grown. On the daily chart the color disagreed on 13.90% of days from 1993 to 1999, 15.23% in 2000-2009, 20.19% in 2010-2019 and 19.50% from 2020 to 26 Oct 2026.

The check. Nasdaq’s own SPY daily history, 26 Oct 2016 to 26 Oct 2026, gives 508 of 2,510 days (20.24%) against 507 of 2,510 (20.20%) from Yahoo’s bars on the same dates. The two files classify 7 days differently, all in 2017, where Yahoo’s open differs from Nasdaq’s. The 20 Apr 2026 row in Nasdaq’s file, open, high, low and close all $710.14 with no volume, is not a real session, so it and the 21 Apr comparison across it were left out. Weekly candles rebuilt from the daily bars give the same 133 of 1,760. The limits: one fund and 60 sessions of 5-minute bars. Yahoo’s hourly file skips 13 and 14 Jul 2026, is short on 30 Jan, 2 Feb, 9 Jun and 10 Jul 2026, and has no 12:30 bar on half-day sessions; the bars of those sessions, and the first bar of any session whose previous session is one of them or is missing, are left out. Every count is in the candle color by timeframe file.

What the videos teach

Across our study of 24,971 trading videos, 645 have candlestick or candle in the title. The median one gets 7,167 views, 67% never pass 50,000, and the median length is 12.6 minutes. The corpus carries description text for 312 of those 645, and across those 312, 4 mention invalidation, failure, or what a bad read looks like.

Body, wick, and which one is telling you more

A candle that is nearly all body settled the period: price went one way and stayed there. A candle that is nearly all wick traveled a long way in both directions and finished roughly where it began. Same range, opposite meanings.

The two get conflated constantly. A long lower wick and a long green body are read as the same bullish signal when they describe different things, one about where price failed and one about where it finished.

Patterns are claims about the bars, not about the market. An engulfing candle describes a shape that already printed; the candlestick patterns page takes the named shapes one by one. Heikin Ashi candles go the other way and average the four prices, which smooths the picture and stops each candle showing the real open and close.

When it fails

The textbook shape leads nowhere

A long-lower-wick candle can look exactly like the one in the worked example and still be followed by price going the other way. Nothing about the candle distinguishes the two, which is the honest position and the one the 308 descriptions in the data above that never mention a failed read leave out.

You read the color as the day’s result

A green candle on a day that closed lower is not a contradiction; it is two different comparisons. Check the change from the previous close before calling a day up or down, especially on the daily chart, where the two disagreed on about one day in six.

The wick was a single print

In a thinly traded instrument the far end of a wick can be one small order crossing a wide spread. Check the volume bar underneath before reading a struggle into a long wick on something that trades rarely.

Questions people ask

What does a green candle mean?

That the period closed above its own open, nothing more. It does not mean the price rose from the previous close: on SPY’s daily chart 686 green candles since 1993 came on days that closed lower.

Why is a candle red when the price went up?

Because the price opened above the previous close and then fell back, but not all the way. On SPY’s daily chart that happened on 782 days since 1993, and it is far rarer inside a session, where each bar opens about where the last one closed.

Which timeframe shows the truest candle color?

None is truer; each compares its own open and close. The color disagreed with the change on 17.29% of SPY daily candles, 7.56% of weekly and 3.22% of monthly ones, because for a weekly or monthly candle only the gap at its start can split color from change, and that gap is small next to a week’s or a month’s move.

How many candles make up a trading day?

On a 5-minute chart of a US stock fund, 78 for a full session from 9:30 a.m. to 4 p.m. Eastern; on Yahoo’s hourly bars, seven, the last one half an hour long.

Related

Support and resistance is the next thing to learn, because a candle only starts meaning something once you know what it printed at.

Market structure is how single candles become a readable pattern of highs and lows.

And an order block is one specific candle picked out of a sequence, so it is worth being comfortable reading one first. For the step-by-step version with a chart open, see how to read a candlestick chart, and candlesticks vs Heikin Ashi sets the two chart types side by side.

This page is educational, not financial advice. Test every idea on your own charts before risking money. Some links on this page earn a commission if you buy through them. It costs you nothing and it does not decide what appears here or in what order — how these pages are made is set out in our methodology.