Volume by Price

The whole instrument follows from one observation, and it is the opposite of what most people assume heavy volume implies.

For a large amount to trade at a price, buyers and sellers have to agree that the price is reasonable — one side willing to pay it, the other willing to accept it, repeatedly, for a long time. That is what a high-volume price is: a price both sides were comfortable transacting at.

And a price both sides are comfortable with is a price that does not need to move. So the places where the most volume traded are the places where price stalls, not where it accelerates.

The reverse follows immediately. A price level with very little volume is one where the two sides could not agree — somebody had to move fast to find a counterparty. Price passed through quickly the first time, and it will pass through quickly again, because nothing has changed about the disagreement.

High volume is acceptance and price sticks. Low volume is rejection and price travels. The profile is a map of which is which.

The vocabulary, which is only four words

Point of control

The single price with the most volume. The market's own answer to "what was this thing worth over this period" — not an average, but the price the largest number of participants agreed on.

Value area

The band around the point of control containing roughly 70% of the period's volume. Conventional, borrowed loosely from one standard deviation, and useful as a definition of "normal price" for that period. Outside it is where the market itself treated price as expensive or cheap.

High-volume node

Any local peak in the profile. A shelf of agreement. Advances stall here and declines find footing here, because a large population transacted at this price and has an opinion about it.

Low-volume node

Any local trough — a price range the market passed through rather than traded in. These are the gaps that produce fast moves, and the most actionable feature on the whole profile.

Exhibit 1 · The profile, annotatedPOC · value area · nodes

The most useful thing on that chart is not the point of control. It is the hollow between the two shelves — the range the market crossed quickly and never came back to settle in. That band is where the next fast move will happen, in whichever direction it is entered.

This is the claim worth testing rather than believing, so here it is measured directly.

Exhibit 2 · Volume against time spentTwo readings of the same price buckets

So when price re-enters a low-volume band, the expectation is not "support at the midpoint." It is that price covers that band quickly, in one or two bars, and only decelerates when it reaches the next shelf. On the chart in Exhibit 1 that is exactly what happens on the way back down: a slow grind inside the upper shelf, a fast drop through the hollow, then an abrupt stop at the old lower shelf.

The profile's silhouette says something about what kind of period it was. Four are common enough to name.

Exhibit 3 · Profile silhouettesWhat each shape is made of

The practical use is not classification for its own sake. It is that a fat shape says the market found value and is likely to keep trading around it, while a thin, stretched shape says the market was in transit and has not yet decided where it belongs. Transit periods are where trends live; balanced periods are where they die.

A point of control from a prior period that price has never traded back to has a name — a naked or virgin point of control — and a reputation as a magnet.

The reputation is probably half deserved, and it is worth separating the mechanism from the folklore. The mechanism: a price where a lot of trade happened, which price then left quickly, is a place where a large population has unfinished business. People who were filled there and want out at the same price, people who wanted to be filled there and missed. Both groups leave resting orders, and resting orders are what price gravitates toward, because a market seeks the prices where it can transact in size.

The folklore: that price must return to it. It does not have to, and often does not for years. The honest version is narrower — when price does approach an untested high-volume level, it tends to reach it rather than stop just short, because the liquidity is at the level rather than in front of it.

This is the same argument as the falling wedge page's overhead supply, and it generalises: the useful question about any level is not "is it support" but "how many people transacted there and what do they want to do now."

• The lookback window changes every level. A profile of the last month and a profile of the last year have different points of control, and neither is more correct. Choose the window from the question you are asking — a composite since the last major low, not "the default setting."
• Most platforms approximate it. A true profile needs every trade's price and size. Many charting tools spread each bar's volume evenly across its range instead, which is an estimate, and on a bar with a long wick it can be a poor one. Lower timeframes make the estimate better.
• It says nothing about direction. Exactly like raw volume, and for the same reason — every one of those transactions had two sides. The profile tells you where the population is, not which way it wants to go.
• Old inventory decays. A shelf built two years ago is made of holders who have since sold, died, been redeemed, or forgotten. Recent nodes are much stronger than old ones, and no platform's shading reflects that.

• Anchor the window to an event, not to a round number of days. Since the last major low, since the earnings gap, since the range began — the profile then answers a question about a specific population.
• Find the hollows first. Low-volume nodes are the most actionable feature and the one most people ignore in favour of the point of control. They tell you where price will move fast.
• Do not put stops inside a hollow. There is nothing there to slow price down, which is the whole reason the hollow exists.
• Expect deceleration at shelves and use them for entries, exits and stop placement — they are the prices at which you can actually transact in size.
• Read the shape before the levels. Balanced means rotation and mean reversion. Stretched means transit and trend. The same level means different things in each.
• Prefer recent structure. When an old node and a recent one conflict, the recent one is made of people who still exist.

A description of market mechanics, not trading advice.