Power of Three (AMD) in Trading, Explained

The power of three, or AMD, is an ICT model that splits a candle or session into three phases: accumulation, a quiet range near the open; manipulation, a false move that sweeps liquidity; and distribution, the real expansion that follows. It describes one common shape a session takes, not a rule.

What is the power of three?

The power of three is a template for the internal structure of a candle or session, describing it as a three-part story rather than a single move. The name refers to the three phases, accumulation, manipulation, and distribution, that ICT teaches every meaningful period passes through. The point of the model is to reframe a candle: instead of seeing one green or red bar, you see a sequence of intentions playing out inside it.

The framing borrows the language of accumulation and distribution, classic ideas about where large participants build and unwind positions. ICT layers a manipulation phase in the middle, arguing that before the real move, price is pushed the wrong way to generate the liquidity and the trapped traders that fuel the eventual expansion. Read this way, the ugly false move at a session open is not noise but a designed part of the structure. That is the claim, and it is worth holding it as a claim rather than a fact.

What happens in accumulation, manipulation, and distribution?

Each of the three phases has a distinct job, and the model only makes sense when they are read in order. Accumulation sets the range, manipulation fakes a direction, and distribution delivers the real one. The transition from manipulation to distribution is the moment ICT traders care about most.

Accumulation is the opening balance, a relatively quiet range near the period's open where price consolidates and, in the theory, positions are built without moving the market far. Manipulation is the Judas swing: a false break out of that range, in the opposite direction to where the period will actually go, designed to sweep the liquidity resting beyond the range and trap traders who chase the breakout. Distribution is the payoff, the genuine expansion move that reverses out of the manipulation and runs toward the close, in the true direction. The candle chart below sketches the shape: a tight open, a false dip that sweeps below it, then a strong close higher.

The manipulation phase is where PO3 overlaps with the rest of the ICT toolkit, because that false move is precisely a liquidity grab that sweeps resting orders before the real move begins.

How does the manipulation phase trap traders?

The manipulation phase works by exploiting the most natural retail behaviour: chasing a breakout. When price pushes out of the accumulation range, breakout traders enter in the direction of the break and place stops on the other side, while traders already positioned correctly get stopped out as price runs against them. Both actions feed orders into the market at exactly the level ICT calls the manipulation extreme.

That is the point of the fakeout in the theory. The false move manufactures the liquidity the eventual distribution move needs, and it puts a cohort of traders on the wrong side, so that when price reverses, their stops add fuel to the real direction. The label draws on the broader idea of market manipulation, though ICT uses the word loosely to mean any engineered sweep rather than anything illegal. For a trader using the model, the manipulation low or high becomes a reference: the reversal out of it, ideally with displacement, is what confirms the accumulation range's true direction and sets up an entry, often via an optimal trade entry into the retracement that follows.

Does the power of three apply to every timeframe?

ICT teaches the power of three as fractal, meaning the same three-phase shape is said to appear on weekly, daily, and intraday session candles alike. A daily candle, in this view, opens into an accumulation range, prints a manipulation move early in the day, and then distributes toward its close, and the same story is claimed for a weekly candle built from five daily ones or a session built from hourly bars.

The practical use of the fractal claim is nesting: a trader might read the daily candle as being in its manipulation phase while watching a lower timeframe for the reversal that begins distribution, using the trading day structure and session opens as anchors. This is elegant, and it is also where caution is needed. A model that is claimed to apply at every scale is very hard to falsify, because a false move can always be relabelled as manipulation on some timeframe after the fact. The fractal framing makes PO3 flexible, but that same flexibility is what makes it easy to fit to any chart in hindsight.

Can an AI chart reader identify the power of three?

It can describe the shape, and that is about the limit. From a screenshot showing a session with its time axis, an image-based reader can point out where the range was tight, where a spike ran beyond it, and where the largest directional move occurred. If the session did unfold as AMD, that description will match.

The catch is that the phases are only unambiguous in hindsight, and a model asked to fit three labels onto a session will generally succeed whether or not the pattern is really there. That is not a defect of the software so much as a property of the model itself: any session with a wick and a trend can be narrated as accumulation, manipulation and distribution. It also cannot verify the session boundaries without a reliable timezone on the chart, and it cannot see whether the previous five sessions did the same thing. Use it to describe what happened, treat any live phase call as speculation, and keep the AI chart analysis work on the structural side where the evidence is in the picture.

Putting the power of three in context

The power of three is best understood as a narrative lens for reading how a candle or session unfolds, not as a mechanical signal. Its value is that it makes you expect a false move before the real one, which can keep you from chasing the manipulation breakout that traps less patient traders. Its weakness is that the labels are applied with hindsight, the fractal claim resists falsification, and the whole framework is anecdotal rather than evidenced.

Used honestly, PO3 sits alongside market structure and liquidity as one way to organise what you see, cross-checked against break of structure and confirmed only once distribution actually begins. Anchored to real risk control rather than a belief that every session must follow the script, it connects the ICT ideas together: the manipulation phase is the sweep of an inducement level, the reversal is the shift, and the distribution is the leg an SMT divergence or an OTE entry might help you read. As with all of technical analysis, the model describes tendencies, not certainties.