Wyckoff accumulation and distribution: reading a range while it is still forming

Open any book or course that teaches Wyckoff and you will meet the same diagram: a trading range with a dozen events labelled in sequence, arrows showing where price went next, and an air of total inevitability about the whole thing. The diagram is not wrong. It is a reasonable summary of what a completed accumulation range often looks like. The trouble is what it does to the reader, which is to suggest that the job is recognition. Find the shape, match the labels, act. That is not the job, and the diagram cannot show you why, because a diagram of a completed range has already answered the only question that was ever difficult.

The difficulty is not identifying a range after it has resolved. That is trivial, and it is trivial in a specific way worth naming: once you know the outcome, the labels become forced. If price left the range upward, then the last probe below support was a spring, the pullback before the advance was a last point of support, and the wide bar that cleared the ceiling was a sign of strength. Every one of those names is assigned by looking at what happened next. Run the same exercise on a range that broke down and you will find the identical bars accept a completely different set of names, equally confidently.

The figure above is the whole argument of this page in one image, and it is worth sitting with. Both panels contain exactly the same bars. The left panel shows all of them, and the labelling is easy, obvious and completely uncontroversial. The right panel stops at the session on which a real decision would have had to be made, and at that moment the probe below support is consistent with at least three futures: a spring that resolves upward, a breakdown that keeps going, and a range that simply refuses to resolve for another two months. Nothing visible at that right edge distinguishes them. The green path is the one that in fact occurred, and it is not marked as more likely, because at that moment it was not.

This is not an argument that Wyckoff analysis is useless. It is an argument about what kind of tool it is. The schematic is a description of a mechanism, and mechanisms are worth understanding because they tell you what evidence to look for. It is not a classifier, and treating it as one produces a very particular failure: an analyst who is fluent in the vocabulary, who can label any historical chart in seconds, and whose real time record consists mostly of ranges that were confidently called accumulation right up until they were not.

Most chart vocabulary treats a sideways range as an absence: consolidation, indecision, a pause, the market catching its breath. Wyckoff treats it as the opposite, as the period in which the thing that matters actually happens. A range is where ownership changes hands. The trend before it and the trend after it are consequences; the range is the cause.

The reasoning is arithmetic rather than psychological, and it is worth stating precisely because the psychological version of this idea has been repeated into meaninglessness. Suppose a participant wants a position worth many multiples of what the stock turns over in an average session. They cannot buy it in a day. If they try, they will lift every offer in the book, the price will run away from them, and their average cost will be dreadful. The only way to acquire that position at a sensible average price is to buy where stock is being offered: into declines, on bad news, at the lower end of a range, from holders who have decided to get out. That takes time, and it takes a price that does not run away while it is happening.

Once framed that way, several features of a range stop being arbitrary and start being predictions. Supply has to be exhausted, so somewhere near the start there should be a session where a great deal of stock changes hands at falling prices and the decline nonetheless stops. Sellers thin out over time, so activity across the range should generally recede rather than build. The float has changed hands, so the range should eventually stop making lower lows, not because a line held but because the people who were going to sell have sold. And because the buyer needs to know whether any supply is still hiding below the obvious support, a probe under that support becomes a sensible thing to expect rather than a decorative flourish.

The same logic in reverse gives distribution. A large holder cannot exit a position of that size into a falling market without destroying their own price. They need enthusiasm, volume and a rising or at least stable price to sell into, which is why distribution tends to form after an advance, at highs, while sentiment is good and the news is favourable. The range is where the position is worked off against public demand.

This is also where the broader idea of supply and demand zones connects, and the connection is worth being careful about. A Wyckoff range and a supply and demand zone are both attempts to mark where significant transacting occurred, but Wyckoff insists on a sequence and a volume signature rather than simply a rectangle drawn around a prior reversal. That extra demand is the point of the framework.

Wyckoff analysis divides a range into five phases, labelled A to E. The phases are not equal in length, in difficulty or in usefulness, and one of them does almost all of the work while offering almost nothing to look at. Each event within a phase is a claim about the balance between supply and demand, and the value of the vocabulary is that it forces you to say which claim you are making.

Phase A stops the decline. It begins with preliminary support, the first point at which buying appears in enough size to slow the fall, usually visible as a widening of both spread and volume that does not yet turn the trend. Then comes the selling climax, the session that does the real work: panic supply is dumped into the market and absorbed, spread widens dramatically, volume reaches the highest level in the sequence, and price closes well off its low. The low of that session usually defines the floor of the range for months afterwards. What follows is the automatic rally, and it is more diagnostic than it looks. Price rises sharply not because demand has become powerful but because supply has temporarily been exhausted, so even modest buying meets little resistance. The height of that rally is therefore a measurement of how thin supply has become, and its high sets the ceiling of the range. Phase A closes with the secondary test, a return toward the climax low that should arrive on narrower spread and materially lighter volume than the climax itself. If the test comes on heavy volume and makes a decisively lower low, the phase A reading is in trouble.

Phase B builds the cause. This is the long, tedious middle, and it can run for weeks or many months. Price crosses and recrosses the range, rallies fail near the ceiling, declines hold near the floor, and nothing about any individual session looks significant. That is exactly what accumulation is supposed to look like, because the buyer's whole objective is to acquire stock without moving the price. The one broadly readable feature of phase B is that activity should recede across it: the supply available for sale is being used up, so the volume required to move price around inside the range should generally diminish. Occasional sharp declines within phase B, sometimes called secondary tests in phase B, are normal and are read as further checks for hidden supply.

Phase C is the test. The best known form is the spring, a probe below the range floor that fails to attract continued selling and returns inside the range. Its function is informational: it establishes whether any meaningful supply remains under the obvious support level, and it does so cheaply, by offering the market one last opportunity to sell. A probe that trades on light volume and recovers quickly supports the accumulation reading. A probe that trades heavily and keeps going does not. It should also be said plainly that phase C is optional. Plenty of ranges that resolve upward never probe below support at all, which means the absence of a spring is not evidence against accumulation, and waiting for one can mean waiting through the entire advance.

Phase D is where demand takes over. The marker is the sign of strength, a move that clears the range ceiling on a wide spread and a clear expansion in volume. The volume condition is not decoration: a breakout on thin trade is a breakout nobody is behind. What usually follows is the last point of support, a higher low on light volume, frequently formed above the level that used to be resistance. Wyckoff's own metaphor for this pair is a creek: the range ceiling is a stream of supply, the sign of strength is the jump across it, and the last point of support is the point where price comes back to the far bank and finds it dry. Phase E is simply the markup outside the range, where the position built during phase B is carried.

Distribution is accumulation performed by a seller, and most of the structure carries over directly. A range forms after an advance rather than after a decline. The climax happens at the top rather than the bottom, and it is a buying climax: an exhaustive surge of public demand, on the heaviest volume of the sequence, into which a large holding is sold. The automatic reaction replaces the automatic rally and measures how thin demand has become. Secondary tests return toward the highs on lighter volume. Phase B is again long and dull, and again does the real work. The test in phase C happens above the range rather than below it, and it is called an upthrust, or an upthrust after distribution when it occurs late in the sequence. Phase D produces a sign of weakness and a last point of supply, and phase E is markdown.

The mirror is genuinely useful, and it is also the point at which a careless reading of the schematic starts to cost something, because the two sides are not symmetric in practice. Ranges that form after a decline tend to build slowly, under fear, with poor news flow and thin participation, and they can take a very long time. Ranges that form after an advance frequently resolve faster and more sharply, because the emotional character of the two situations differs: a holder who is frightened sells at once, while a holder who is comfortable holds until they are not. The consequence is that a distribution range flipped upside down out of a textbook tends to understate how abruptly the resolution can arrive, and analysts who are used to reading bottoms often find they are given less time to act at tops.

There is a second asymmetry worth stating. The events of accumulation are frequently accompanied by genuinely bad news, which makes the reading psychologically hard but evidentially cleaner: heavy volume at a low with an obvious reason for the selling is easy to interpret. Distribution often forms while the story is excellent, which makes the reading psychologically easy to dismiss and evidentially muddier, because strong demand and heavy volume near a high are exactly what a continuation would also produce. The upthrust is supposed to resolve that ambiguity, and it does, but only after the fact.

The single most common failure in applied Wyckoff analysis is naming an event from its geometry. A low that pokes below the range gets called a spring, because a spring is a low that pokes below the range. That reasoning is circular, and it is circular in a way that is very hard to notice from the inside, because the resulting chart looks exactly like the diagram in the book. The label has been assigned from the shape, and the shape was never the evidence.

What makes a probe a spring is the relationship between effort and result. Effort, on a chart, is volume, the amount of transacting that took place. Result is what the price did with it. A break below support is an effort by sellers to establish lower prices. If that effort is substantial and the result is a decisive move lower, the market is telling you supply is winning. If the effort is unremarkable and the result is a swift return inside the range, something is absorbing what is being sold, and the accumulation reading strengthens. The law itself is derived on the companion method guide; what matters here is that it is the instrument capable of saying no.

Both panels of that figure contain the identical thirty four candles. Bar for bar, the price geometry is the same, the range is the same, the probe below support is the same. Only the volume differs, and it reverses the conclusion completely. In the left panel activity thins across the range and the probe trades at roughly seven tenths of its own recent average, which is what absorption looks like: the sellers who wanted out are gone, and the break attracts nobody. In the right panel activity builds across the range and the probe trades at nearly twice its recent average, which is what a genuine breakdown looks like: supply is being met with more supply. An analyst reading price alone would label both of those charts identically, and would be wrong about one of them.

This is also why the volume pane is not an optional accessory to the schematic. Every event in the accumulation and distribution vocabulary carries a volume condition, and stripping those conditions out leaves a set of shapes that can be fitted to almost any chart. The general question of what volume can and cannot tell you, including its real limits, is covered in the guide to volume. The specific point for range reading is narrower: volume is the only variable in the standard schematic that is capable of refusing a setup the structure has already approved.

Two cautions on using it. First, volume comparisons must be relative to the instrument's own recent history rather than to an absolute figure, which is why the figure above expresses the probe as a multiple of its own twenty session average rather than as a raw number. Second, volume in derivatives and volume in the cash market are different quantities measuring different things, and a range read on one while volume is taken from the other is not a coherent analysis.

There is no diagram in the Wyckoff canon called the range that did not work. Every schematic in every textbook is a schematic of a process that completed, which means the entire visual literature of the method is a survivorship sample. Study it for long enough and you acquire a confident sense of what accumulation looks like, assembled exclusively from examples where accumulation is what it turned out to be.

Ranges that look like accumulation break down regularly. The probe below support that was going to be a spring keeps going, and there is nothing wrong with the analysis up to that point, because a breakdown and a spring are the same event until one of them resolves. Calling it a failed spring afterwards is a way of preserving the vocabulary at the expense of the lesson. It was a break of support. It broke.

The figure above is drawn deliberately: the range is textbook, the probe below the floor is exactly what a spring looks like, and the four session recovery afterwards is exactly the reclaim a spring is supposed to produce. Then it fails anyway. The point of the ledger beneath the chart is not the specific percentages, which are illustrative, but the structural fact they expose. None of those exit levels came from the schematic. The schematic named the events; it did not name a price at which the reading should be abandoned. Every one of the four rows is a rule an analyst brought to the chart from outside, and the bottom row measures what happens to someone who brought none.

That is the honest structural criticism of the framework, and it is not fatal so long as it is understood. Wyckoff analysis is a way of forming a hypothesis about what is happening to supply and demand. It is not a trading system, it contains no position sizing, and it specifies no invalidation. Those have to be supplied, and they have to be supplied before the range resolves, because afterwards is when everyone knows where the line should have gone.

If the retrospective label is worthless, the practical question is what can honestly be said at the right edge of a chart. The answer is more than nothing and considerably less than the diagram implies, and the useful discipline is to sort the vocabulary by the moment at which each term becomes assignable rather than by the order the events occur in.

Some events are recognisable close to the time. A selling climax can be identified within a few sessions, because everything that defines it, the volume expansion, the wide spread, the close off the low, is present on the session itself and does not depend on what follows. An automatic rally can be named once it has run and stalled. A secondary test can be named when it arrives on lighter volume near the climax low. These are descriptions of what has happened, and they are legitimate.

Other terms are verdicts rather than descriptions, and they are not available in real time at all. Spring is a verdict. It asserts that the probe below support was a test that failed to find supply, and that assertion is only settled by what price does over the following sessions. The same is true of a last point of support, which claims to be the last one, and of an upthrust after distribution, which claims that distribution has largely been completed. Using a verdict word before the verdict exists is the specific mechanism by which hindsight analysis leaks into live analysis.

Two habits follow from that table. The first is linguistic and it is not a trivial matter of style: describe what you can see and reserve the verdict words until they are earned. A note that reads probe below support, volume did not expand, watching for a reclaim within three sessions is a statement that can turn out to be wrong in a specific way. A note that reads spring has quietly assumed the conclusion and left nothing to test.

The second is to keep competing readings alive rather than picking one early. At any point inside a range there is usually an accumulation reading, a distribution or breakdown reading, and a reading in which nothing has been decided. Each has evidence for and against, and the useful work is tracking which pieces of evidence would move you between them. The moment the analysis collapses to a single story, the chart stops being able to surprise you, and the schematic in your head starts filling in bars that have not printed.

Indian equity markets offer one piece of evidence that most markets do not make available to retail participants, and it bears directly on the question a range is asking. The exchanges publish security wise delivery position data, giving the deliverable quantity as a percentage of the traded quantity for each stock each day. Volume alone tells you how much changed hands; the delivery share tells you roughly how much of that was taken into a demat account rather than squared off within the session.

The relevance to range reading is direct. Absorption, in the Wyckoff sense, means someone is taking ownership. If a range shows heavy volume but the delivery share stays low, a great deal of the activity is intraday churn that ends the day flat, which is turnover without transfer. If price moves sideways while the delivery share rises, that is a more interesting combination, because it suggests the volume is producing actual changes in ownership. The full derivation of why this is a genuinely useful second witness, and how it interacts with the three laws, is set out on the companion Wyckoff method guide and is not repeated here.

Two further Indian specifics are worth noting for anyone reading ranges here. Circuit limits can truncate exactly the sessions a Wyckoff reading depends on most, since a climax that hits a band stops printing the spread and the volume that would have defined it, leaving a bar that looks orderly for a reason that has nothing to do with absorption. And in stocks with concentrated ownership, a very large share of the issued capital may simply not be available to trade, so a range can absorb what looks like a small quantity relative to the company and still represent a substantial share of what actually circulates.

Set against everything above, it would be easy to conclude that the schematic should be discarded. That would be the wrong conclusion, and it would be wrong for a reason worth stating carefully: the mechanism the diagram describes is real even though the diagram is a poor classifier. Large positions genuinely cannot be built or unwound quickly. Absorption genuinely does take time and genuinely does leave traces. A range in which the character of the volume changes is genuinely different from one in which it does not.

The schematic earns its place as two things. It is a vocabulary, which lets you say precisely what you think is happening rather than gesturing at a chart, and precision is what makes a view capable of being wrong. And it is a checklist of evidence, because each event in the sequence comes attached to a condition, and running through those conditions is a structured way of asking whether the story you like is actually supported. Used that way the framework does something genuinely valuable: it slows down the leap from picture to conviction.

What it is not, and cannot be made into by adding more labels, is a system. It has no entry rule you have not supplied, no invalidation you have not chosen, no position size, and no way of telling you how often the reading works. Those belong to a separate layer of decision making, and the honest version of the method we teach keeps that layer distinct rather than allowing a chart reading framework to masquerade as a complete process. A range that breaks out is a hypothesis that survived; the schematic was the language you used to state it, and the risk rules were what kept the cost bounded while you waited to find out.

If you take one thing from this page, make it the pair of panels near the top. The completed range and the range at its right edge contain the same information, and one of them is easy while the other is the job. Everything that is difficult about applying Wyckoff analysis lives in the gap between those two pictures, and no amount of additional vocabulary closes it. What closes it, partially and imperfectly, is demanding evidence for each claim, writing the invalidation down before it is needed, and keeping an honest record of what you actually said at the time. For the mechanics of trading a range boundary once you have a view, see the guide to breakouts.

• Wyckoff, Studies in Tape Reading (1910). Published under the pen name Rollo Tape, this is the early written record of reading the footprint left by large orders on the tape, and the origin of the absorption reasoning this page relies on. archive.org
• The Wyckoff course (1931) and the Stock Market Institute. Wyckoff issued a formal correspondence course, The Richard D. Wyckoff Method of Trading and Investing in Stocks, and founded an institute to teach it. The phase structure and the event vocabulary used on this page descend from that material. en.wikipedia.org
• The accumulation and distribution schematics as taught today. Standard contemporary Wyckoff reference literature documents the phases A to E and the event abbreviations (PS, SC, AR, ST, spring, LPS, SOS, and their distribution twins) in the form used here. Note that schematic variants differ in detail between sources, and the event sequence is a common summary rather than a fixed rule. chartschool.stockcharts.com
• Delivery position data. The National Stock Exchange of India publishes security wise delivery position reports giving deliverable quantity as a percentage of traded quantity, per security, per trading day. Report formats, file locations and definitions change over time; verify at source before relying on them. nseindia.com
• No peer reviewed validation of the schematic. The Wyckoff accumulation and distribution schematics have not been validated as predictive structures in peer reviewed research; they are practitioner descriptions of a market mechanism. This guide explains their logic and their evidential structure, and makes no claim about how often either pattern completes or about the results of trading them.
• On the figures. Every chart on this page is generated from a synthetic price series written to illustrate a specific point about method. The prices, volumes, levels and percentages are illustrative. They do not describe any real instrument, any real period or any real outcome, and the computed values in them are outputs of the drawn series rather than measurements of a market.