Power of Three: Accumulation, Manipulation, Distribution
PO3 shows up constantly in ICT-adjacent trading content, usually reduced to three words and a diagram: accumulation, manipulation, distribution. Most explainers stop there. What’s missing is why the model is built the way it is, where it’s meant to apply beyond a single daily candle, and — because this publication doesn’t skip the uncomfortable part — where it honestly doesn’t hold up.
Table of contents
- What the model actually claims
- Why it’s fractal
- The daily and weekly templates
- Where the model honestly breaks down
- Where this gets hard to do by hand
What the model actually claims
Power of Three, sometimes written AMD, is attributed to Michael Huddleston’s ICT methodology. The core claim: a trading range — over a day, a week, or another defined period — tends to move through three phases. First, accumulation: price consolidates, building a range as positions get built without much directional commitment. Second, manipulation: price pushes beyond one edge of that range, often described as a “Judas swing” — a deceptive move that looks like the start of a breakout in one direction. Third, distribution: the genuine directional move, frequently in the opposite direction from the manipulation leg.
That third phase — the idea that the manipulation push is often a fakeout before the real move — is the part that gets the most attention, because it’s the part that maps directly onto a specific trading decision: don’t take the manipulation leg at face value.
Why it’s fractal
The model isn’t presented as something that only applies to a single daily candle. It’s taught as fractal — the same three-phase idea repeating across different period lengths: a single session, a full day, a full week, each nested inside the next. A week can have its own accumulation-manipulation-distribution arc, and within that week’s manipulation phase, an individual day can run its own smaller AMD cycle. That’s part of why the framework gets applied so broadly across different trading styles and timeframes — it’s offered as a repeating structure, not a one-off daily pattern.
This fractal claim is also where the model asks the most of a trader applying it. If the same three-phase idea can, in principle, be found on a 1-hour chart, a daily chart, and a weekly chart simultaneously, deciding which timeframe’s AMD cycle is actually the relevant one for a given trade becomes a judgment call in itself — the framework describes a repeating shape, but it doesn’t tell you, on its own, which instance of that shape you should be paying attention to right now.
The daily and weekly templates
On the daily version, the phases are commonly mapped onto the standard forex sessions: the Asia session as the accumulation phase (building the range), London as the manipulation phase (the sweep beyond it), New York as the distribution phase (the genuine move). This is a widely taught mapping, and it’s worth being precise about what kind of claim it is: it’s a commonly-taught tendency, not a guaranteed sequence that fires the same way every single day.
There’s also a commonly-taught weekly version, mapping the same three phases onto specific days: Monday and Tuesday often framed as accumulation and manipulation, Wednesday frequently highlighted as a common reversal point, Thursday and Friday as the distribution leg. As with the daily mapping, this is a descriptive framework passed down through ICT-adjacent education, not a statistically validated rule — we haven’t found a disclosed, methodologically sound study establishing what percentage of weeks actually follow this exact template, and per our own editorial standard, we’re not going to invent one.
Underlying both templates is the idea of daily bias: before the current period’s own AMD cycle plays out, traders using this framework typically look at where resting liquidity sits relative to the prior period’s range, to form an expectation about which direction the eventual distribution phase is more likely to favor. That expectation is a starting bias, not a conclusion — the accumulation and manipulation phases still have to actually happen, and confirm or contradict it, before the framework is telling you anything you couldn’t have guessed in advance.
[Screenshot: SS-PO3-01 — a full annotated AMD day, accumulation through distribution]
Where the model honestly breaks down
Here’s the part worth connecting to a different piece we’ve published: the “manipulation” phase in PO3 is, mechanically, the same event as a liquidity sweep or stop hunt — price pushing beyond an obvious range edge. We’ve already covered, with sourcing from actual order-book research, that a swept level reverses far less reliably than trading folklore assumes; the documented, dominant pattern in that research is continuation, not automatic reversal.
That same caution applies directly here. Treating every manipulation-looking push as automatically destined to reverse into a distribution phase in the opposite direction has exactly the same weakness as treating every liquidity sweep as an automatic reversal signal — because they’re describing the same underlying mechanism. Sometimes the “manipulation” leg simply is the move, and there’s no distribution phase in the opposite direction waiting to happen. The model is a framework for organizing observation, not a sequence that resolves predictably on a fixed schedule.
None of this means the framework is worthless — it’s a genuinely useful way to organize what to watch for: has a range actually built, has price pushed convincingly beyond one edge of it, and — the part that actually decides the outcome — is there real structural evidence that the push has failed and reversed, or is structure instead confirming that the push is continuing. The letters A, M and D describe a shape worth recognizing. They don’t, by themselves, tell you which one you’re looking at until the third phase has already started to reveal itself.
[Screenshot: SS-PO3-02 — apparent manipulation leg continuing rather than reversing]
Where this gets hard to do by hand
Applying this framework properly means tracking which phase a given period is likely in, checking that against session context, and — critically, given the caveat above — waiting for genuine structural confirmation before assuming a manipulation leg is actually reversing rather than continuing. Doing all of that by eye, consistently, across the timeframes this model is meant to apply to, is a lot to hold at once.
One way to handle that: map session context and phase timing with SSM IQ Price Action with Sessions and PO3, and check whether a suspected manipulation leg is actually confirming structurally — rather than just continuing — with SSM IQ ICT. Neither tool tells you which outcome is coming. Both make the pieces of this framework, and the confirmation this article says you should be waiting for, visible together instead of asking you to track them separately.
The three letters were never the whole model. What happens after the manipulation leg — reversal or continuation — is the part that actually matters, and it isn’t decided until it happens.
Related SSM IQ Tools: SSM IQ Price Action with Sessions and PO3 · SSM IQ ICT Related Articles: London-New York Session Overlap, Explained · What Research Actually Says About Stop Hunts
SSM IQ tools are analytical software, not investment advice. Trading involves substantial risk. Author: SSM IQ Research · We Are Tech Sp. z o.o. · Updated: 2026-08-16