Home  /  The library  /  Range and volatility
Range and volatility

What true range actually measures

High minus low undercounts any bar that opens away from the previous close. True range is the correction, and it is the measurement this company is named after.


The obvious way to measure how far an instrument moved in a bar is high minus low. It is also wrong often enough to matter.

Consider a stock that closes at 50 and opens the next morning at 47 after news. It then trades between 46.50 and 47.50 all day. High minus low reports a one-point day. The instrument actually travelled three and a half points from the previous close, and anyone holding overnight felt every bit of it.

The correction

previous close previous bar next bar — opens 3 lower high − low |high − prev close| |low − prev close| TRUE RANGE = the largest of the three. Here, the third.
Schematic illustration of the three distances true range compares. Not market data.

True range takes the largest of three distances:

TR = the greatest of: high − low (the bar's own span) | high − previous close | (upward gap plus the bar) | low − previous close | (downward gap plus the bar)

On an ordinary bar that opens near the previous close, the first term wins and true range equals high minus low. On a gap, one of the other two wins and the gap is counted. The measure is self-correcting — you do not need to detect a gap first.

The idea comes from J. Welles Wilder's New Concepts in Technical Trading Systems (1978), where he also introduced the average of it. Wilder was working on commodity futures, where limit moves could produce bars with almost no range and enormous actual price change. True range was built for exactly that failure.

Two properties worth knowing

True range is never negative and never smaller than high minus low. It either equals the bar's span or exceeds it. There is no case where the correction makes a bar look quieter.

The first bar of a series has no previous close. Most platforms fall back to high minus low for it. On a short chart this is invisible. On a chart with only twenty bars it slightly biases the first reading downward.

What it does not tell you

Direction. A bar that fell three points and a bar that rose three points produce the same true range. This is not a limitation to be worked around — it is the definition. Range measures distance travelled, and distance has no sign.

It also says nothing about the path. A bar that ran straight up and a bar that whipped both ways before closing where it started can have identical true range. If path matters to you, range is the wrong measurement and you need to look at smaller bars.

Why measure it at all

Because a two-point move means something different in a market that typically moves one point than in one that typically moves six. Almost every useful thing you can say about the size of a move requires a denominator, and true range is how you build one. That is what the average is for.

Next in this group