Moneydesks

🕯️ Reading price bars: candles, bars, time frames, scales and adjusted data

Open this guide in the app →

🎯 In 30 seconds: A price bar squeezes a stretch of trading into four prices: open, high, low and close. That is useful, and it also throws away detail, such as which came first, the high or the low. Bars of different sizes nest inside each other, and today’s bar is unfinished until the session closes. Use a log scale when price has moved a lot, and make sure old prices are adjusted for splits and dividends before you draw anything on them.

What one bar keeps and what it drops

A bar is a summary. It records the first trade, the highest trade, the lowest trade and the last trade of its period, plus the total volume. Everything in between is gone.

$49$50$51$52Hourly bars of one sessionOpenCloseHighLowThe same day as one bar
Hypothetical. The daily bar takes the first hourly open, the highest high, the lowest low and the last close. The dip in the middle of the session disappears inside the lower wick.

Look at what the daily bar hides. Midway through the session the stock fell to $48.50 and many holders may have been nervous. The daily candle shows only a lower wick. It does not tell you whether the low came at 10 a.m. or 3 p.m.

That matters when you test rules. If a bar touched both your stop and your target, the bar alone cannot tell you which was hit first. Careful tests assume the worse outcome, or check a smaller bar size.

Where the close sits in the range

One of the most useful readings of a single bar is where it closed within its own range. A simple ratio does it:

Close position = (close minus low) / (high minus low)

The answer runs from 0 (closed at the low) to 1 (closed at the high). For the daily bar in the figure: ($52.20 minus $48.50) / ($52.50 minus $48.50) = $3.70 / $4.00 = 0.93. Buyers held control into the end of the day.

A close near 1 after a long fall says sellers ran out of push by the close. A close near 0 on a day that opened strong says the early buying did not last. One bar proves little, but a run of closes in the top or bottom quarter of their ranges deserves a line in your journal. The single-candle chapter builds on this.

Bar styles in the chart tool

StyleWhat it drawsGood for
CandlesticksA body from open to close, wicks to high and lowSeeing who won each period at a glance
Hollow candlesBody filled or hollow by close versus open, colored by close versus the prior closeSpotting days that rose from the open but still closed lower than yesterday
OHLC barsA vertical line with a tick left for the open and right for the closeDense charts where wide bodies get in the way
HLC barsThe same without the openMarkets where the open carries little meaning
Line and mountainCloses only, joined up (mountain shades below)Long histories and clean trend reading
BaselineCloses colored above or below a reference priceTracking a position against your entry price

All of these come from the same four prices. Switching between them is a cheap way to check a read. If a pattern is only visible in one style, be skeptical of it.

Time frames nest inside each other

A weekly bar is built from the week’s daily bars: Monday’s open, the week’s high and low, and Friday’s close. A daily bar is built from intraday bars the same way.

Bar sizeBars in a regular 6.5-hour U.S. sessionTypical use
5 min78Fine timing within one day
15 min26Intraday structure
Hourly7 (one of them covers only half an hour)Swing timing over a few days
Daily1The main view for most swing traders
Weekly and monthlya fraction of oneThe big picture and long trends

Two practical points follow. First, short bars are noisier: a bigger share of their movement is random back and forth. Second, every pattern on a short bar size sits inside some larger bar. Read the larger one first. The multiple time frames chapter turns this into a routine.

Watch out
The rightmost bar is still forming until its period ends. A daily candle at 11 a.m. can look like a strong reversal and finish the day as something else. A weekly bar on a Tuesday is only two days old. Base decisions on completed bars, or say clearly that you are acting on an unfinished one.

Linear or log scale

On a linear scale, each dollar takes the same height. On a log scale, each percentage change takes the same height.

Linear scale$10$20$40$80Each doubling looks biggerLog scale$10$20$40$80Each doubling is the same height
Hypothetical. A price that doubles at a steady pace, from $10 to $80. The linear scale stretches the later moves. The log scale gives every equal percentage move the same height.

A move from $10 to $20 and a move from $40 to $80 are both 100% gains. A log scale shows them as equal. A linear scale makes the second look four times larger, which can fool you into seeing a “blow-off” in what is really a steady trend.

A rough rule: when the chart covers a price change of more than about 50%, switch to log. For a few weeks of a quiet stock, the two look almost the same. Trendlines in particular can change meaning between the two scales, which the trendline chapter covers.

Adjusted data: splits and dividends

Companies change their share count and pay cash out. Both events shift the price for reasons that have nothing to do with supply and demand.

  • Splits. In a 2-for-1 split, a $100 share becomes two $50 shares. An adjusted history divides all earlier prices by 2 and multiplies earlier volume by 2, so the chart stays continuous.
  • Dividends. On the day a stock goes ex-dividend, its price usually drops by about the payout. Say a stock closes at $50.00 and then goes ex-dividend for $1.00. A dividend-adjusted series multiplies all earlier prices by ($50.00 minus $1.00) / $50.00 = 0.98. A $40.00 price from last year would show as $39.20.
Unadjusted$50$1002-for-1 split:a fake cliffAdjusted$50$100Split dayOld prices are divided by 2, so the trend reads correctly
Hypothetical. Owners lost nothing in the split: each old share became two half-price shares. Adjusting the history removes the false drop.

Adjustment has a side effect. Old adjusted prices are not the prices people actually paid. A level that “everyone remembers” at $40.00 may sit at $39.20 on an adjusted chart. For long histories that does not matter much. For a support level from a few months back, it can.

If a chart shows a sudden drop of exactly one half or two thirds with no news, suspect an unadjusted split before you suspect a crash.

Try it on the chart
In Technical Analysis, load one stock on daily bars over several years. Flip the chart type from candlesticks to hollow candles, then OHLC, then line. Next switch between Linear and Log and note how the slope of the last big move changes. Finally change the bar size to weekly and then hourly. Find one hourly swing and locate the single daily bar that contains it.
Check yourself
  1. A bar has high $30, low $26 and close $29. Where did it close in its range? Answer: ($29 minus $26) / ($30 minus $26) = 0.75, in the upper quarter.
  2. Why can’t a daily bar tell you whether a stop or a target was hit first? Answer: It keeps only four prices, not the order in which the high and low happened.
  3. When does a log scale help most? Answer: When the chart covers a large percentage change, because it shows equal percentage moves at equal heights.
  4. After a 3-for-1 split, how is an old $90 price shown on an adjusted chart? Answer: $30, since earlier prices are divided by 3.

Sources

  • Steve Nison, Japanese Candlestick Charting Techniques (1991)
  • John J. Murphy, Technical Analysis of the Financial Markets (1999)
  • Robert D. Edwards and John Magee, Technical Analysis of Stock Trends (first published 1948)
  • U.S. Securities and Exchange Commission, Investor.gov (investor education pages on stock splits and dividends)
  • Moneydesks Technical Analysis tool (chart types, bar sizes and scale settings)