PO3: Power of Three
Also known as: PO3, power of 3, AMD, accumulation manipulation distribution
What is it?
Power of Three is the idea that a single candle - usually a daily or weekly one - is built in three phases: price accumulates in a tight range, manipulates by pushing out of one side, then distributes by running hard the other way. Read a EUR/USD daily candle that opened at 1.0850, printed a low of 1.0821, and closed at 1.0932. Flat as a single bar it just says 'up day'.
How one daily candle is built: accumulate, manipulate, distribute
-
1
Accumulation
Price coils in a tight band around the 1.0850 daily open while the session is quiet. No trade here - you are marking the range.
-
2
Manipulation
The London open pushes below the range to 1.0821, filling the sell stops. The day's low forms here, not at the end.
-
3
Distribution
Price reclaims the 1.0850 open and runs to 1.0932 through New York. The real move only starts after the flush.
Broken into three phases it says much more: price accumulated between 1.0845 and 1.0858 through the Asian hours, dropped to 1.0821 at the London open to take out the sell stops sitting under that range, then reversed and distributed to 1.0932 across the New York session. The low of the day was manipulation, not weakness, and the 1.0850 open is the line that separates the two readings. The practical value is timing.
If you expect the daily candle to close up, you also expect the low to form early, which tells you to look for longs after a flush below the open rather than before it. Not every candle follows the sequence - ranging days accumulate and never distribute, and a candle that closes near its open never resolves at all. Your capital is at risk on trades taken from it.
Why it matters: Reading a daily candle as accumulate, manipulate, distribute tells you the day's low often comes before its run - so you buy the flush instead of fearing it.
It sets when you expect the day's extreme to form, which changes entry timing rather than the size of the position you take.
Real-world example
A EUR/USD daily candle opened at 1.0850, accumulated near 1.0845-1.0858, flushed to 1.0821 at the London open, then distributed up to close at 1.0932.
How SignalBots handles it
SignalBots timestamps every signal in UTC and keeps the session it fired in on the signal card, so you can see whether a setup came from the flush or from the run that followed it. See /risk-warning.
Pro tip
Use the daily open as the dividing line - price pushing below it and then reclaiming it is the cleanest tell that the manipulation phase is over.
Common pitfalls
Assuming the manipulation leg always lands at the London open; on days driven by US data the flush arrives in the New York session instead.
Frequently asked questions
What does PO3 stand for?
Power of Three, the three phases of a candle: accumulation, manipulation and distribution. It is often written AMD after those phases, and the two names describe exactly the same model.
Which candle should I apply it to?
The daily candle is the standard choice because its phases line up with the Asian, London and New York sessions. The weekly candle works the same way with Monday as accumulation, and lower timeframes get noisy fast.
How is Power of Three different from a Judas swing?
The Judas swing is the manipulation phase seen on its own, right after a session open. Power of Three is the wider frame that puts that swing between an accumulation range before it and a distribution leg after it.
How do I decide which way the candle will close?
That comes from outside the model - higher timeframe structure, where the draw on liquidity sits, and whether price is in premium or discount. Power of Three tells you how the move will be built, not which direction it takes.
Does Power of Three work on crypto?
The phases still appear, but without a daily close and session opens the boundaries are blurrier. Many crypto traders anchor the model to the 00:00 UTC daily candle and treat the equity session opens as the manipulation windows.
Trading involves substantial risk of loss. Historical and backtested results do not guarantee future performance. Read the full risk warning.
