Skip to content

Absorption in order flow

By Author of ORDER FLOW — Market Mechanics

If delta measures how much effort one side puts in, absorption measures what happens to that effort when it reaches a level where someone is willing to contain it.

It is read on the order flow record, and specifically on the footprint. This page covers the signature, the three things it is most often mistaken for, and why even a genuine absorption can carry no useful information.

Effort and result

Price reaches a level with aggression from one side, and that side cannot move it in proportion to its effort. Someone on the other side, with enough size, is taking every contract with limit orders that do not give way, and price is stuck.

Absorption is therefore a relationship between two things: high volume in one direction, and a price displacement that does not match that volume. Neither one alone is enough to call it absorption.

On a footprint it has a recognisable signature: large numbers in one column, across several consecutive levels, and a candle extreme that does not move.

Price does not advance despite buying delta 21,302.25 18 210 21,302.00 22 265 21,301.75 31 240 21,301.50 45 90 21,301.25 60 40 level under attack Δ +669 Candle delta
Schematic of an NQ candle carrying an absorption signature. Values are illustrative.

The delta of that candle is strongly positive. Buyers arrived with urgency, and price still did not advance. The signal is that mismatch between effort and result.

Absorption or a void

This is the most common false positive, and also the easiest to avoid.

If price stops at a level but the footprint shows low volume, with cells nearly empty on both sides, what you are looking at is a lack of interest. Nobody is trading at that level.

Absorption 24 268 31 240 40 255 High volume, price still. Someone is holding it. Void 2 3 1 0 0 4 Low volume, price still. Nobody is there.
Schematic comparison of an absorption signature and a lack-of-interest signature. Values are illustrative.

On the chart price does the same thing in both cases: it stands still. The implications are opposite. In the first, a participant with size is defending a price. In the second there is no information of any kind, and the level will give as soon as an order with real volume arrives.

The working rule: absorption requires high volume with price failing to advance. On low volume, a stalled price only tells you that interest is missing.

Absorption or exhaustion

The void has a close relative. The book's glossary defines exhaustion as the loss of aggression's ability to keep moving price, and it distinguishes two forms. Both end with price stalled at an extreme, as absorption does, and neither has the cause absorption has.

Exhaustion through withdrawal. The side that was pushing gradually steps back. Candle by candle, volume and delta fall and the imbalances in the direction of the move get smaller, and price stops, or drifts on a little into the vacuum, because nobody is carrying it. There is no wall here. What is missing is the push.

Exhaustion through withdrawal close volume · delta 2,600 +820 2,000 +540 1,400 +300 900 +120 500 +30 Volume falls candle by candle mean 1,480 Absorption close volume · delta 2,400 +760 2,800 +880 2,700 +810 2,900 +900 2,600 +740 Volume stays high mean 2,680
Schematic of two five-candle sequences at the same NQ extreme, with each candle's close, volume and delta. Values are illustrative.

In the schematic, volume in the withdrawal falls from 2,600 to 500 contracts and delta from +820 to +30, while the close advances 2.75 points and stops. In the absorption the close barely moves, 0.25 points between the highest and the lowest, on an average of 2,680 contracts per candle against 1,480. What separates the withdrawal from the void in the previous schematic is time: a void shows in a single candle, and a withdrawal only in the sequence.

Climactic exhaustion. This is the case most often mistaken for absorption, because volume spikes instead of falling. It comes after an extended directional leg, when late buyers, shorts closing at a loss and whoever uses the rush to sell all arrive at once. Effort is at its peak and sustained displacement is small.

Delta does not separate them. In chapter 7 the book warns that absorption and climactic exhaustion both produce a high delta heavily skewed toward the aggressors, and that looking for the difference in its size is a common trap. The difference lies in how volume is spread inside the candle and in the leg that comes before it.

Absorption 21,302.25 40 520 21,302.00 60 660 ◂ close 21,301.75 70 150 21,301.50 55 45 ◂ open Delta +1,150 Volume 1,600 Two busiest levels 80% Range 3 ticks Close over open 2 ticks Climactic exhaustion 21,318.50 20 95 21,318.25 25 170 21,318.00 35 190 21,317.75 40 200 21,317.50 45 215 21,317.25 50 180 21,317.00 50 160 ◂ close 21,316.75 55 140 21,316.50 50 120 ◂ open 21,316.25 55 105 Delta +1,150 Volume 2,000 Two busiest levels 25% Range 9 ticks Close over open 2 ticks Outlined cell: diagonal buying imbalance at the common 3:1 reference.
Schematic of two NQ footprint candles with the same delta. At each level, aggressive selling (bid) is on the left and aggressive buying (ask) on the right. Values are illustrative.

Both candles in the schematic have a delta of +1,150. In the absorption, the two busiest levels hold 80% of the volume and the candle spans 3 ticks: there is a wall at one point. In the climactic exhaustion, the two busiest levels hold only 25%, there are 6 buying imbalances in a row, and the candle still closes 2 ticks above its open after spanning 9. It travelled a long way inside the candle and ended up almost where it started.

At the end of a climactic move the two readings often coexist: absorption concentrated at the upper levels of the candle, and exhaustion across the move as a whole. The first says where the wall is; the second, that the impulse has run its course.

Neither form of exhaustion is confirmed in a single candle. Withdrawal needs the sequence, and a climax needs the leg before it and what happens afterwards. A high-delta breakout candle raises the same question, and the answer again comes from the candles that follow the break.

When both sides absorb

The book describes a third type of absorption that is more a market state than a turn: bilateral absorption. Two blocks of liquidity contain each other inside a range of three to five ticks. The footprint shows clusters of heavy volume on both sides, delta swings without building a bias, and the CVD runs sideways.

Trying to guess the direction while the fight lasts is the trap. The useful reading arrives when one side runs out and the range breaks with volume and delta pointing the same way. Chapter 13 uses it to read a rejection at a high volume node of the volume profile, where price is expected to get caught in balance.

Why a real absorption can mean nothing

The signature can be genuine, the volume real and the containment obvious, and the reading can still be useless. Three situations come up again and again.

Absorption in a structural void. It happens at a level with no prior reference, far from any zone with memory in the volume profile, with no concrete reason for a large position to want that price. The volume is real, but no narrative explains it. It could be distribution, accumulation, mechanical hedging or high-frequency noise, and the signature is the same in all four.

Absorption without exhaustion. Seeing the containment does not tell you how much ammunition the participant holding the level has left, or how long they mean to hold it. An absorption can last far longer than seems reasonable. Acting the moment you see it assumes the wall is permanent, and often turns you into the liquidity the other side needed to finish building its position.

Continuation absorption. The hardest to recognise in real time. What looks like a turn is a participant building a position in the direction of the move before it carries on. Price stops, aggression piles up against the move, and the footprint shows the full signature. Once the accumulation is complete, price continues the way it was going.

Reversal absorption zone The defence holds and price turns. Continuation absorption zone It was accumulation: price carries on. Same footprint signature. Opposite outcome.
Schematic of two possible outcomes after one and the same absorption signature. Simplified representation.

Both cases in the schematic read the same on the footprint. What separates them sits outside the candle: where it happens, what the market was doing beforehand, and what follows once the aggression runs out.

What turns absorption into information

What it takes, then, is context from outside the footprint.

The level must have structural relevance, marked before anything happened. Volume has to be clearly high, so that absorption is not confused with a void. You have to wait for the aggressive side to run out, which has a signature of its own: the attacking side's volumes shrink and delta flattens. And a response has to appear, some sign that containment has turned into initiative. That response is usually read as a run of imbalances on the side opposite the one that was attacking.

Exhaustion also leaves a mark at the extreme of the move. When price stops because aggression fades, rather than because a rejection appears, the cell at the high or the low tends to keep volume in both columns. That cell is the unfinished auction, and it has a reading of its own.

None of those four conditions can be read in the absorption candle. All of them depend on a framework built before price got there, which is why absorption is at once one of the most useful concepts in order flow and one of the most error-prone.

The type of day is part of that framework. The same absorption on a pullback comes before a turn in a rotational session and is only a pause in a trend day, and the candle does not say which of the two you are in.