Charts · Range
Reading a High-Low Range Chart
Two sessions can close at the same price after covering wildly different ground. Plotting the range says which was which, and plotting volume beside it finds the sessions where the two disagree.
What the range adds
A chart of closes records one number per session: where the market stopped. It cannot distinguish a quiet drift from a session that fell four per cent, recovered all of it, and closed unchanged. The high-low range records the second number (how much ground was covered), and the two together describe a session in a way either alone cannot.
range = high − low, and range % = (high − low) ÷ close × 100 for anything you intend to compare with another instrument or with a different era.
That percentage form matters more than it sounds. Ranges in points grow with the price level, so a long history of point ranges shows an expansion that is mostly the price of the instrument rather than the behaviour of the market.
The four configurations
| Range | Volume | What it describes |
|---|---|---|
| Wide | Heavy | The ordinary decisive session: a lot of stock changed hands and the price travelled. Common, and it adds nothing beyond what the price already shows. |
| Narrow | Heavy | Effort without result. Supply and demand were matched at that level, evidence of absorption, and no evidence about direction. |
| Wide | Light | Thin liquidity. A modest order moved the price a long way, which is the configuration most often mistaken for conviction. |
| Narrow | Light | A quiet session, and the least interesting state. Most sessions are here, which is worth knowing before building a rule on the other three. |
The second and third rows are the reason to plot both series, and they are the entire subject of volume spread analysis. What that method adds is a further question, where in the range the close sat, and what it shares with everything else in this reference is the honest limit: these are descriptions of a session, testable and checkable, and none of them is a forecast.
The one property of ranges that is statistically robust
Almost nothing in this reference comes with a strong empirical result attached. Range behaviour is the exception: the size of a session's range is autocorrelated. Wide sessions tend to follow wide sessions and quiet ones follow quiet ones, over horizons of days to weeks.
That is a much better-established finding than anything about direction, it holds across markets and across decades, and it is the reason the whole family of volatility models used in pricing and risk exists at all. Returns themselves are close to unpredictable; the magnitude of returns is not. A range chart makes the property visible without any modelling: the periods of expansion and compression are obvious to the eye, and they persist.
Two practical consequences follow, and both are about sizing rather than forecasting. A stop or a target set as a fixed number of points is loose in an expanded regime and tight in a compressed one, which is the entire argument for expressing such distances in ATR units. And a threshold calibrated on a quiet period will fire constantly in a volatile one, the same recalibration warning that every page in this library ends with, here with a documented mechanism behind it.
What the property does not support is the step people take next. Knowing that a compressed range is likely to be followed by more compression, and eventually by expansion, says nothing about the direction of the expansion. "Volatility is coiling" is a defensible statement about magnitude and an empty one about which way the market goes.
Where the raw range fails, and what to use instead
Gaps. A session that opens well away from the previous close and then trades quietly has a small high-low range and was a large move. True range fixes this by including the previous close in the calculation, which is why ATR rather than the raw range is the right input to anything being sized against volatility.
Single prints. On a thinly traded instrument the high or the low can be one small trade at an unrepresentative price, which widens the range without describing where business was actually done. A liquidity floor removes most of this, and a volume-weighted measure removes the rest.
Sessions that are not sessions. Half-days, holidays and the sessions around them have structurally smaller ranges, and they sit inside any twenty-session baseline for a month afterwards. The same window problem the average volume page describes applies here unchanged.
The market-wide version
Averaging the percentage range across a whole list of issues gives a volatility measure for the market rather than for one instrument, and it is a genuinely useful description of conditions: elevated through declines, compressed in quiet advances, and expanding sharply before it expands again at the end.
Two construction notes, both of which follow from everything above. Average the percentageranges rather than the point ranges, or the most expensive constituents dominate the figure. And restrict the list to common stock, because funds and preferred shares have structurally narrow ranges and their share of any listing has grown, so an unrestricted long history shows a compression that is composition rather than calm.
Read against the market’s breadth data, that average range is one of the more informative pairings available: participation says how many issues took part, and the average range says how far the ones that did actually travelled.
Frequently asked questions
What is a high-low range chart?
A chart of the distance between each session’s high and low, plotted as a series in its own right rather than as bars around a price line. It answers a question a close-only chart cannot: how much ground was covered inside the session, regardless of where it ended. A quiet session and a violent one that closed unchanged look identical on a line chart of closes and completely different here.
Should the range be in points or as a percentage?
As a percentage of price for any comparison between instruments or across a long history, because a two-point range means something different at ten dollars than at four hundred. In points it is only comparable with the same instrument’s own recent ranges, which is exactly what average true range does, and why ATR is expressed in the instrument’s own units and read against itself.
How is this different from average true range?
ATR is this quantity smoothed, with one addition: true range also accounts for gaps by including the previous close, so a session that opens away from yesterday’s close and trades in a narrow band is correctly recorded as a large move. The raw high-low range misses that. Use the raw range to see individual sessions and ATR when you want a stable level to size something against.
Why plot range and volume together?
Because they usually move together, and the exceptions are informative. Both respond to activity: a session where a lot of stock changes hands normally covers more ground. When the two part company — a wide range on ordinary volume, or heavy volume inside a narrow range — something structural is happening, and that pairing is the whole foundation of volume spread analysis.
What does heavy volume in a narrow range mean?
That a great deal of stock changed hands without the price going anywhere, which means supply and demand were closely matched at that level. Practitioners describe it as effort without result and treat it as evidence of absorption, someone large taking the other side. It is a genuine observation about the session and it says nothing reliable about direction: the same configuration appears at turning points and in the middle of ranges that continue.
And a wide range on light volume?
Usually thin liquidity rather than conviction. With few participants a modest order moves the price further, so the range expands without much stock changing hands. It is the configuration most likely to be mistaken for a decisive move, and the check is straightforward, compare the volume against the instrument’s own baseline before reading anything into the range.
Does range expansion mark the start or the end of a move?
Both, which is why it cannot be used alone. Ranges expand when a market begins to move and expand again when it is exhausted, and the widest single sessions in most histories sit near the ends of declines rather than at their beginnings. What range expansion reliably says is that the market has become more active; whether that activity is starting something or finishing it requires the price context around it.
What is the market-wide version of this chart?
The average range across a whole list of issues, which behaves like a volatility measure for the market rather than for one instrument, high in declines, low in quiet advances, and a reasonable description of conditions. It is worth computing as a percentage per issue and then averaging, rather than averaging point ranges, so that the largest and most expensive constituents do not dominate the figure.