What Is Volume Analysis?
Volume analysis is reading how much traded rather than where price went. A volume bar has no direction in it - every contract has a buyer and a seller - so it measures participation, and what counts as high participation depends entirely on the baseline period chosen.
Before anything else, a disclosure that shapes this whole page: the volume bars on these charts are illustrative and derived from each bar’s range, so no claim about how volume relates to price is made from them. What follows is either definitional or a statement about method.
How it works
A volume bar counts how many units changed hands in that period. That is all it is.
And every one of them had a buyer and a seller. There is no such thing as more buying than selling in a completed trade — the quantities are identical by definition, because a trade is two parties.
So a volume bar has no direction in it. Whatever else volume can tell you, it cannot tell you which side was in control, and the phrase “buying volume” describes an inference, not a measurement.
What the colours actually encode
Platforms colour the volume bar green or red. That colour is taken from the candle: green if the close was above the open, red if below.
It is the same information as the candle, drawn twice. Nothing about the volume bar changes when its colour does, which is worth knowing before reading a wall of green bars as evidence of demand.
Order flow tools do try to split the two sides — by tagging whether each trade printed at the bid or the offer, which is a real measurement and a different data feed. The order flow page covers what that adds and what it costs.
“High” needs a baseline
A bar is only heavy compared with something. Take the same 144 bars and flag every bar more than twice its recent average:
| Baseline | Bars flagged |
|---|---|
| Ten-bar average | 8 |
| Fifty-bar average | 3 |
| Flagged by both | 2 |
Six of the eight are only high against the short baseline.
This is a statement about method, not about markets — and that is why it holds even on illustrative data. The lookback is a choice, it is rarely stated, and it decides which bars you call significant.
Which is exactly the argument the volume profile page makes about the window. A level, or a bar, described without the period it was measured against is an opinion with a number attached.
The three claims worth testing yourself
Volume analysis is usually taught as three rules. All three are testable, and none of them is tested in the material that teaches them.
“A breakout on high volume holds; on low volume it fails.” Testable: take every breakout on your instrument, split them by volume against a fixed baseline, and compare what happened next. The breakout page has the counting method.
“Volume should rise with the trend.” Testable, and the harder one, because “the trend” has to be defined before the volume is looked at or the test is circular.
“A volume spike marks exhaustion.” Testable, and the one most likely to disappoint, because exhaustion is defined by what happened afterwards.
Every one of those is a two-column count — the condition, and the outcome — which is the same exercise the do indicators work page runs against a control. Do it on the instrument you actually trade, because the answer will not be the same on a stock and on spot foreign exchange.
What volume genuinely adds
Where participation happened. Price tells you the range; volume tells you which prices inside it were busy, and that is what a volume profile turns into a level.
How much of the market was present. A move on very little participation and the same move on a great deal are different events, whatever you then do with that.
And a check on a level. If a level is meant to matter because people transacted there, the volume column is the only place on the chart that says whether anybody did.
Notice what is missing from that list: an entry signal. Every use above is a filter or a measurement, which is the same conclusion the do indicators work page reaches for indicators generally.
A worked example
Write down your baseline. Ten bars or fifty — either is defensible, and choosing after seeing the chart is not.
Then use volume to disqualify rather than to trigger. A breakout into a price region nothing has traded in behaves differently from one into a busy one.
And ignore the colours. They are the candle again.
The original data
Across our study of 24,971 trading videos, 55 cover volume analysis. The median one gets 26,981 views, 55% never pass 50,000, and the median length is 13.8 minutes.
That median is roughly twice the broader volume topic’s 13,581, on a field one twelfth the size — 55 videos against 683.
The corpus carries description text for only 18 of those 55, which is too thin to say anything about how the topic is written, and this page does not.
When it fails
You read direction into it
“Heavy selling” is a description of the candle, not of the volume bar. The failure is not that the reading is wrong — it is that it adds nothing the candle had not already told you.
The baseline moved with the answer
Choosing a ten-bar average because it flags the bar you were interested in is the whole right-hand-edge problem applied to a setting. Fix the baseline first.
The instrument does not report it properly
Spot foreign exchange has no central exchange, so the volume your platform shows is your broker’s own flow, not the market’s. It is a sample of unknown size — usable for shape, not for a threshold.
You labelled it afterwards
A heavy bar that leads a move is called a breakout. The same bar at the end of one is called a climax. The bar is identical; only the label is assigned later, and at the moment it prints there is nothing on the chart that distinguishes the two.
Related
Volume is the basic reading — what the bar is and how it sits under the candle.
Volume profile is the one thing this data does that price cannot: say where the participation happened.
And order flow is what it takes to actually separate the two sides of a trade.
The only way I use volume is as a check rather than a signal: if a level is doing something and nothing traded there, I trust it less. I have never been able to make a volume rule that fires on its own, and I have stopped trying, because everything I built ended up measuring the size of the bar.
— Michael Whitman, from this video
This page is educational, not financial advice. Test every idea on your own charts before risking money.