ICT Power of Three AMD: accumulation, manipulation and distribution explained

The ICT Power of Three (AMD) model describes the market as a three-phase cycle: institutional accumulation of orders, manipulation of retail liquidity levels, and directional distribution toward the institutional target. Developed by Michael Huddleston (Inner Circle Trader), this framework explains why price consistently appears to hunt stop losses before reversing: this is not random, but a structural consequence of how large institutions fill their orders at scale. According to the European Securities and Markets Authority (ESMA), between 74% and 89% of retail investors lose money on leveraged products. Understanding institutional mechanics is therefore a practical necessity, not a theoretical exercise.

What is the ICT Power of Three (PO3)?

The Power of Three is one of the foundational frameworks in the ICT (Inner Circle Trader) methodology. Its premise is straightforward: markets do not move linearly. Instead, price follows a repeating pattern of consolidation, false breakout, and then true directional movement.

This pattern is not random. It reflects the mechanical reality of large-scale institutional order execution.

Origin: how Michael Huddleston developed the AMD model

Michael Huddleston, the trader and educator behind the ICT methodology, developed the AMD model as an extension of his theory on market maker behavior. His core argument is that major institutions and central banks must execute enormous orders without creating unfavorable price displacement against their own positions.

To buy massively without pushing price up during accumulation, they need sellers on the other side. Manipulation creates exactly that: by pushing price toward the stop losses of long traders (below key support levels), institutions trigger sell orders that allow them to buy at favorable prices before launching the distribution move higher.

The model draws conceptual roots from Richard Wyckoff's accumulation/distribution cycle from the 1930s. The key difference: the ICT AMD model integrates modern elements specific to electronic markets, including trading sessions, kill zones, and liquidity levels tied to institutional order flow.

For a comprehensive overview of the full ICT methodology, read our guide on the ICT method by Michael Huddleston.

Why price moves in three phases

Markets require liquidity to function. Every order execution requires a counterparty. The three-phase structure responds to this mechanical necessity:

1. The accumulation phase creates a consolidation zone where institutional orders build silently, outside the visibility of standard retail analysis tools.
2. Manipulation generates the liquidity needed for execution by triggering the stops of traders positioned in the wrong direction.
3. Distribution allows institutions to deliver price toward their target, where the next pool of opposing liquidity resides.

According to the Bank for International Settlements (BIS), the foreign exchange market averaged $7.5 trillion in daily turnover in 2022. At this scale, institutional order execution requires structured, predictable liquidity. The AMD model describes exactly how that liquidity is generated and consumed.

Phase 1: Accumulation - understanding the consolidation range

Accumulation is the phase where price moves within a tight range, typically following a significant prior move. This is where institutions fill their order books before the next major move.

Identifying the accumulation range

A valid accumulation range shows several characteristics:

• Clearly defined highs and lows (visible equal highs and equal lows)
• Compressed volatility: candle bodies are shorter than the recent average
• Stable or slightly declining volume (institutions absorb orders quietly)
• A minimum of 3 to 5 candles on the analyzed timeframe

In intraday trading, the accumulation range often corresponds to the Asian session. Price consolidates during Asia (low volatility, low volume), creating the reference levels that London and New York sessions will use for manipulation and distribution.

On daily or weekly timeframes, the accumulation phase can last from several days to several weeks.

Institutional limit orders forming within the range

During accumulation, institutions place limit orders (buy limits for long positions, sell limits for short) inside the range. These orders are not visible on standard retail charts, but their effects are observable: whenever price attempts to break out of the range and returns inside it, this often reflects institutional absorption.

An indirect indicator: price levels where the market bounces repeatedly (equal highs or equal lows). These repeated levels signal the presence of accumulated liquidity, both as institutional limit orders and as retail stop losses placed at these obvious technical levels.

Phase 2: Manipulation - the false breakout (stop hunt)

Manipulation is the most distinctive phase of the AMD model. It corresponds to the move that appears to break the accumulation range, triggers retail stops, then reverses sharply back inside the range before moving in the opposite direction.

Recognizing manipulation vs. a genuine breakout

Distinguishing manipulation from a real breakout is the core skill of the AMD framework. Signals of manipulation include:

• The breakout move is fast (one or two candles) and does not hold
• Volume during the breakout candle is low (no institutional conviction)
• Price returns quickly inside the range after the initial break
• The move does not confirm with a valid structural change (no CHOCH or BOS)
• The breakout occurs during an ICT kill zone (London open, NY open) when manipulations are most frequent

A legitimate breakout, by contrast, holds above (or below) the range, forms a new structural high (or low), and is often accompanied by above-average volume.

Our guide on the Market Structure Shift (MSS) covers in detail how to identify a genuine structural change.

Liquidity sweep within the AMD model

Manipulation almost always corresponds to a liquidity sweep: price briefly exceeds equal highs (bullish manipulation before bearish distribution) or equal lows (bearish manipulation before bullish distribution) to trigger accumulated stops.

This liquidity sweep serves two simultaneous functions:

1. It captures the liquidity institutions need to execute their orders in the opposite direction.
2. It eliminates traders positioned on the wrong side, reducing counter-pressure during distribution.

Phase 3: Distribution - the true directional move

Distribution is the phase where institutions deliver price toward their target. This is the only moment when an ICT trader seeks to enter in the direction of the move.

Entering after distribution confirmation

Entry during distribution comes after confirmation that manipulation is complete. The most commonly used confirmation criteria:

• A Market Structure Shift (MSS) or Change of Character (CHOCH) on the entry timeframe
• An Order Block or Fair Value Gap formed during the manipulation phase
• Price returning inside the range following the liquidity sweep

The principle: never enter during manipulation (too risky), but wait for the reversal to be confirmed by market structure. Entry on the retest of the order block or FVG created during the sweep is the classic AMD setup.

Price targets and position management in distribution

The natural distribution targets are the next liquidity levels in the direction of the move:

• Previous high or low (previous daily high/low, weekly high/low)
• An unfilled Fair Value Gap on the higher timeframe
• An equilibrium zone (50% of the last major move)

Position management during distribution is straightforward: stop beyond the extreme of the liquidity sweep (for a long) or above it (for a short), partial take profit at the first liquidity level, final take profit at the primary target.

AMD across timeframes and trading sessions

The Power of Three operates across all timeframes. The intraday version is the most widely used by ICT traders, as it aligns with the natural cycle of global trading sessions.

Daily AMD (Asia accumulation, London manipulation, New York distribution)

The most documented intraday AMD cycle maps directly onto the three major trading sessions:

• Asian session (Accumulation): price consolidates in a narrow range. Volume is low. This is the accumulation phase. The high and low of this session (Asia High and Asia Low) serve as the reference levels for what follows.
• London open (Manipulation): at the London open (08:00-10:00 GMT), price breaks violently out of the Asian range, typically in the direction opposite to the intended true move. This stop hunt on the Asia High or Low is the most documented AMD manipulation.
• New York open (Distribution): following the London manipulation, price reverses and moves in the institutional direction for distribution. The London-New York overlap session (13:00-17:00 GMT) concentrates the highest volume and the strongest directional moves.

Our guide on ICT kill zones and trading sessions provides the exact timing windows for these opportunities. The Asian session range and liquidity traps article covers the accumulation phase in even greater depth.

AMD on weekly levels

The AMD model also applies on higher timeframes. At the weekly scale:

• Monday/Tuesday: weekly accumulation (weekly range forming)
• Wednesday/Thursday: potential manipulation (false breakout of previous week's high or low)
• Thursday/Friday: distribution and delivery toward the weekly target

This weekly cycle is less mechanical than the intraday version, but it provides a useful framework for identifying directional bias for the week. The ICT Market Maker Model (MMXM) complements this multi-timeframe reading.

Backtesting the AMD model with Backtrex

What separates a skeptical ICT trader from a confident one is backtesting. Understanding the AMD model conceptually is not sufficient. You need to know whether this model has a positive mathematical expectancy on your specific instruments and session windows.

Backtrex allows you to define AMD model rules in no-code format and run a backtest on 5 to 10 years of historical data in under 30 seconds. The output gives you real statistics:

• What percentage of valid AMD setups (confirmed manipulation and MSS entry) are profitable?
• How does performance differ between AMD setups on London session vs. New York session?
• On which currency pairs or instruments is the model most robust?
• What is the historical maximum drawdown if you trade this setup systematically?

These questions cannot be answered through theoretical study of the model. They require data.

Discover how to configure and run this type of AMD backtest on Backtrex, our visual no-code backtesting platform designed for ICT and SMC traders.

Conclusion

The ICT Power of Three (AMD) model is one of the most coherent frameworks for understanding the institutional mechanics of financial markets. Accumulation, manipulation, and distribution are not abstract concepts: they correspond to real observable execution phases visible on any liquid market chart.

But theoretical understanding is not enough. To know whether this model is tradeable on your instruments and sessions, you need data. Backtest your AMD setups rigorously, measure the real statistics, and calibrate your approach accordingly. This is the difference between an attractive concept and an operational trading strategy.