Many traders have noticed the phenomenon where price moves in the opposite direction during the London session before running in its true direction. Power of Three (PO3), a concept put forward by ICT (Inner Circle Trader), systematizes this movement into three phases: Accumulation, Manipulation, and Distribution. This article organizes the definition, session timing guidelines, concrete examples, and common misconceptions, drawing on relevant data.
Definition and Mechanics of Power of Three
Power of Three is a framework that breaks down the process by which a candle on a given timeframe forms into three stages. It is one of the core concepts of ICT Methodology, systematized by Michael J. Huddleston (known as ICT) from the early 2010s and freely published on YouTube and other platforms.
① Accumulation: The period during which smart money (institutional investors and large players) quietly builds up positions. Price moves within a narrow range and volatility is relatively low. The Asian session (approx. 0:00-9:00 JST) most commonly corresponds to this phase.
② Manipulation: Price temporarily moves to either side of the range, triggering stop-losses placed by retail traders. This is often recorded as a wick on the chart and is also known as a “false break” or “Judas Swing.” This phase most often aligns with the early London session (approx. 16:00-18:30 JST).
③ Distribution: Price moves significantly in the direction opposite to the Manipulation phase — the true direction. Smart money takes profit on accumulated positions while pushing price higher (or lower). The New York session (approx. 22:00-2:00 JST next day) most commonly corresponds to this phase.
Within SMC, this framework functions as a hypothesis explaining why false breaks repeat. That said, ICT is ultimately one interpretive model, and it is not possible to verify from the outside whether institutional investors actually move in accordance with this scenario.
Concrete Example: Reading the Three Phases on a Daily EUR/USD Chart
The following is an illustrative example for understanding the daily PO3 concept. Actual price action varies case by case, but please use this as a reference for how the framework is applied.
- Open: 1.0820
- Accumulation Phase: 0:00-9:00 (JST), price moves in a range of 1.0810-1.0830
- Manipulation Phase: After the London open, price drops to 1.0795 (hunting stops at the recent low)
- Distribution Phase: Price rises to 1.0870 during the NY session and closes there
In this pattern, the Manipulation phase low of 1.0795 is the “Judas Swing” — a move opposite to the true direction (upward). From a PO3 perspective, even when observing a London session breakout, it is always important to consider the possibility that it may be the Manipulation phase.
Phase-to-Session Reference Table
| Phase | Session | Approximate Time (JST, no DST) | Key Characteristics |
|---|---|---|---|
| Accumulation | Asia | 0:00-9:00 | Tight price movement; liquidity converges to a single level |
| Manipulation | Early London | 16:00-18:30 | Frequent false breaks and Judas Swings |
| Distribution | New York | 22:00-2:00 (next day) | Volatility expands; primary trend forms |
On days with major economic data releases, session timing can shift significantly. In addition, the dominant session varies by currency pair, making pair-specific backtesting essential (see also: Profit Factor Benchmarks and Strategy Evaluation Criteria).
Common Pitfalls for Beginners
- Misreading the Manipulation Phase as an “Entry Signal”: Attempting a reversal entry when price moves in the opposite direction during the London session is dangerous. PO3 is a “filter for suspecting that a move may be a fake” — it is not a standalone entry trigger. Whether the Manipulation phase has ended cannot be judged in real time; an entry should only be considered after multiple confluences such as Order Blocks, Fair Value Gaps, and liquidity levels all align.
- Applying It Mechanically Across All Timeframes: Daily PO3 and weekly PO3 do not necessarily align. Trying to capture accumulation on a lower timeframe when the higher timeframe is in the Distribution phase puts your multi-timeframe (MTF) analysis in the wrong direction. While the basic approach is to use PO3 to pass context from higher to lower timeframes, timeframe conflicts occur frequently in practice.
- Confirming It in Hindsight and Projecting It Forward: Reading a chart from left to right, “that wick was the Manipulation” may look obvious — but in real time there is no objective standard for confirming when Manipulation has ended. Holding a position simply because it “looked like the pattern” carries high risk, and when combined with inadequate stop-loss rules, can lead to significant drawdowns (see also: Why Martingale EAs Develop Deep Drawdowns).
- Moving to Live Trading Without Backtesting: PO3 is a qualitative framework. Without converting rules into numbers, statistical verification is impossible. An intuitive sense of “it seemed to work” at a small sample size carries the risk of survivorship bias. In our lab’s work incorporating PO3 into AI EAs, quantifying conditions and running large-scale backtests is set as a mandatory step.
FX AI Lab View
Power of Three offers reference value as a framework that intuitively explains how market liquidity is utilized. However, as of this writing, AI backtesting conducted at our lab shows that the statistical edge of strategies based on quantified PO3 conditions is still under verification — no definitive conclusion has been reached. We are continuing live experiments combining PO3 with our AI autonomous trading on an HFM demo account; those who wish to track and compare results are welcome to refer to How to Open an HFM Demo Account. For questions, please use our contact form at any time.
Related Links
This article is provided for informational purposes only and does not constitute a solicitation to invest in any specific financial product. FX trading carries the risk of losses exceeding your initial investment. Please make all investment decisions at your own discretion. For details, please refer to our Risk Disclosure.