Short answer: power of three ict is the idea that any candle, on any timeframe, is built from three phases. Price first accumulates in a quiet range near the open. It then manipulates by running past that range to take liquidity, and finally distributes in the true direction toward the close. Traders use the model to sell the false high of a bearish candle or buy the false low of a bullish one. The stop sits just beyond the manipulation extreme.
Who is this page for? Traders who already know what a liquidity sweep is but still get caught on the wrong side of the open. By the end you should be able to name the phase you are in and size a trade inside it. You should also be able to tell a real power of three day from a trend day that never looks back.
The model as one candle: open, wick, body, close
Take a bullish daily candle. It opened at 1.0800, printed a low at 1.0770, and closed at 1.0900. The open sits near the bottom third. The lower wick is small. The body is long and points up. That shape is the whole ICT power of three in one picture. The open is where accumulation sits, the wick below the open is the manipulation, and the body is the distribution.
A bearish candle is the mirror image. The open sits near the top, the upper wick is the fake rally, and the body runs down into the close.
The practical consequence: on a bullish day the low tends to print below the open and early, and the reverse holds on a bearish day. If you expect a bullish close, you should want to see price trade under the open first. A jittery trader sees that dip as a failure. A patient one sees it as the phase that creates the entry.
The three phases in plain terms
| Phase | What price does | What it feels like | Where the pressure comes from |
|---|---|---|---|
| Accumulation | Tight, unremarkable range around the open | Boring, tedious, easy to ignore | Orders build on both sides of a narrow box |
| Manipulation | A sharp push past the range against the real direction | Convincing, sometimes violent | Stops and breakout entries get filled at the extreme |
| Distribution | A sustained move the other way into the close | Obvious in hindsight, awkward in real time | Trapped positions unwind, momentum follows |
Accumulation is the coarse, choppy stretch where the market decides nothing. In a forex pair this is often the Asian session, roughly 20 to 40 pips wide on a normal day. Nothing consequential happens, which is exactly the point: the range only exists to gather resting orders.
Manipulation is the lopsided break. Price leaves the box in one direction, often around the London or New York open, and takes liquidity from stops sitting beyond the range. It scarcely matters whether anyone intends this. Concentrated orders at an obvious level get hit; that is enough.
Distribution is the phase most traders actually want but seldom catch, because they are either stopped out during manipulation or too skittish to enter after it. It is the long body of the candle.
A worked example on a 10,000 account
Assume a 10,000 account and a rule of risking 1 percent per trade, so 100 per attempt.
- Daily bias: bullish, from a higher timeframe structure of rising lows.
- Accumulation: the Asian range on a major pair sits between 1.0790 and 1.0820, a 30-pip box.
- Manipulation: at the London open, price drops to 1.0765, 25 pips under the box, then closes back inside it within two 15-minute candles.
- Entry: a buy at 1.0800 once price reclaims the open.
- Stop: 1.0750, just under the manipulation low, a 50-pip stop.
- Position size: 100 divided by 50 pips is 2 per pip, so 0.2 lots on a standard contract.
- Target: 1.0900, the prior day's high and a clean pool of buy-side liquidity, 100 pips away or 2R.
If the distribution phase delivers, the trade earns 200. If the reclaim was false and price makes a new low, the loss is 100 plus spread and commission cost. That cost is not trivial on a 50-pip stop. A 1-pip round trip is 2 percent of the risk budget on every attempt. It adds up when you are stopped out twice on the same idea.
Results vary, and this is one planned trade rather than a promise. The point is the structure: the stop sits where the model is plainly wrong, and the target sits where liquidity is plainly waiting.
A decision framework: is this a power of three day?
Before every session, I run a short checklist. If fewer than four of the six boxes tick, I treat the day as unclassified and stay flat.
- Is there a higher timeframe bias, so I know which side of the open the manipulation should form on?
- Did a clear accumulation range form, tidy enough that its extremes are obvious to everyone?
- Did the manipulation push reach an obvious pool of liquidity, such as a prior day extreme or a round number?
- Did price get back inside the range quickly, rather than stalling stubbornly beyond it?
- Is there a punctual reason for the move, a session open or a scheduled release, instead of a random midday drift?
- Is the reward to the opposite liquidity at least twice the stop distance?
The framework grew out of the broader ICT trading strategy, which treats time and liquidity as the two inputs that matter. Our guide to the judas swing covers the session-level version of the manipulation phase. The power of three is the candle-level frame that contains it.
How the model compares with other ICT setups
| Model | Scale | What it answers | Where it is weakest |
|---|---|---|---|
| Power of three | One candle on any timeframe | Which phase am I in? | Ambiguity about when accumulation has ended |
| Judas swing | One session open | Which direction was the fake? | Trend days that never come back |
| Silver bullet | One fixed hour | When to act? | Windows that produce nothing |
| Plain breakout | Any range | Where did momentum start? | It buys the manipulation phase |
The power of three is the most forgiving of these, because it does not demand a specific clock time. It is also the most brittle, because it explains every candle after the fact. Each of the others adds a filter that the power of three lacks, so they are stronger used together. The underlying logic of all four comes from smart money concepts, where liquidity is treated as the reason price moves rather than a side effect.
Limitations and common mistakes
Bluntly, the model has a hindsight problem. Every candle has an open, a high, a low and a close, so every candle fits once it has finished. That fit is worthless if you could not name the phase while it was forming.
The mistakes I see most often:
- Labelling every dip as manipulation. A dip under the open in a downtrend is mostly distribution rather than a fake. Without a bias, the label is a guess.
- Entering during the push. Buying the manipulation low while it is still printing is a knife catch. Wait for the reclaim.
- Using a stingy stop. A stop inside the manipulation wick gets hit by ordinary noise. Put it beyond the extreme or do not take the trade.
- Trading the model on a disorderly day. When a release moves price violently and the range never forms, there is no accumulation to work from.
- Ignoring the timeframe. A 5-minute candle and a daily candle both have three phases, but the daily one contains dozens of 5-minute versions. Pick one frame and stay on it.
- Squandering the second half of the day. Distribution can decelerate into a sluggish grind. Holding for a full-range target when momentum has faded is how a winner becomes flat.
The other honest caveat is that the phases are not equal in size. Accumulation and manipulation can be brief while distribution is lumpy and long, or the reverse on a stale, low-volume day.
Where the model meets evaluation rules
A prop firm evaluation adds a daily loss limit and a total drawdown to the trade plan. The power of three interacts with both in a specific way. The manipulation phase is where most of your losses will cluster, because that is where the market is designed to look wrong. Two failed reclaims at 100 each put a 10,000 account 200 down before distribution even starts.
That is why a pre-set attempt limit matters more here than in most setups. Pivex states its full rule set ahead of the first trade. That includes a one-step evaluation, a daily loss allowance that renews each session, and a fixed total drawdown that does not trail your equity. Knowing those numbers up front lets you decide, in advance, that two manipulation-phase losses end your session. The fees and costs of a failed attempt are then bounded by a decision you made calmly rather than one you made jittery at the open.
Next step: take one major pair and log the open, the manipulation extreme and the close for twenty sessions. Note which phase you would have entered in each time. Once that log shows how often the reclaim actually held, set one attempt per session against limits you already know. Pivex fixes its daily loss limit and total drawdown before you place the first trade, so those limits are never a surprise.
Related guides: ICT trading strategy and the judas swing.
Before participating, review Pivex's current program rules and decide whether a simulated proprietary trading evaluation fits your experience and risk limits. Participation does not guarantee a reward, income, or any particular trading result.