What Power of 3 is
Power of 3 (often written Po3) is an ICT model that says every candle is built in three phases:
- Accumulation: price consolidates near the open. Positions build. Range is small.
- Manipulation: price makes a false move away from the open, against the eventual direction, taking liquidity on that side.
- Distribution: price makes the real move, expanding away from the open in the other direction, and closes near the far extreme.
People shorten it to AMD. The idea comes from looking at a finished candle. A bullish daily candle with a small lower wick and a close near the high can be read as: open, drop (manipulation forms the low), rally (distribution forms the body), close. A bearish candle reads the other way.
Traders use Po3 for two things:
- Bias framing. If you expect a bullish day, you look for weakness below the open early in the day as a buying opportunity, not as a reason to turn bearish.
- Timing. ICT ties the phases to sessions. The manipulation leg, called the Judas swing, is commonly expected in the London kill zone.
Po3 is a description of candle shape plus a story about intent. The shape is real and measurable. The story, that someone deliberately engineers the false move, is not something a retail chart can prove. Use the shape, test the timing, and stay skeptical of the story.
The reference open: midnight, 18:00 or 08:30?
Po3 needs an open to measure from. You need to choose one and stay consistent.
| Reference | New York time | Why traders use it |
|---|---|---|
| Midnight open | 00:00 | ICT’s preferred “true day” open. It sits after the Asian session has started and before London. |
| Gold futures daily open | 18:00 | The actual start of the trading day for gold after the 17:00–18:00 break. Matches the daily candle on many platforms. |
| 08:30 open | 08:30 | A New York session reference, used for a session-level Po3 around US data. |
| Weekly open | Sunday 18:00 | For a weekly Po3 frame. |
Your platform’s daily candle may open at 17:00 or 18:00 New York time, or at 00:00 server time, depending on the broker. That changes the shape of the daily candle. If you trade Po3, draw the midnight open as a horizontal line each day instead of relying on the candle.
How it behaves on gold (XAUUSD)
Gold’s day has a natural three-part rhythm, which is one reason Po3 is popular with gold traders:
| Session (New York time) | Po3 phase it is often mapped to | What gold often does |
|---|---|---|
| Asian, 20:00–00:00 | Accumulation | Builds a narrower range. Creates the Asian high and low, which become liquidity for later. |
| London, 02:00–05:00 | Manipulation | Often raids one side of the Asian range or the midnight open. This is where the Judas swing is expected. |
| New York, 07:00–11:00 | Distribution | US desks and 08:30 data add volume. The day’s main expansion frequently happens here. |
That mapping is a tendency, not a law. Here is how gold breaks it:
- News-driven days. On CPI, payrolls or Fed days, the real move often begins at 08:30 or 14:00 New York time. London may be the accumulation and New York may contain both the manipulation and the distribution.
- Trend days. In strong trends, gold may never trade meaningfully against the open. London simply continues the prior day’s direction. There is no Judas swing to buy or sell.
- Asian-driven days. Big moves sometimes start in Asia on Chinese demand headlines or geopolitical news over the weekend. The whole day may then be distribution from the start.
- Range days. Price swings both sides of the open and closes near it. AMD only fits in hindsight, if at all.
Gold’s sensitivity to the dollar and real yields matters too. A Judas swing that coincides with a sharp move in DXY may simply be gold following the dollar, not a liquidity raid. Check DXY and yields before you label a move.
How to identify Po3 step by step
- Pick your frame. Most traders frame the daily candle. Mark the midnight open line.
- Form a daily bias before London. Use the daily and H4 chart: where is the draw on liquidity? Is price in premium or discount of the higher-timeframe range? Are you expecting a bullish or bearish daily candle?
- Mark the accumulation range. The Asian high and low, plus the midnight open.
- Define what manipulation should look like. For a bullish bias: a move below the midnight open, ideally taking the Asian low or another obvious sell-side level. For bearish: above the open, taking buy-side.
- Wait for a shift. After the raid, look for displacement back toward and through the open, with a market structure break on M5 to M15 and a fair value gap.
- Enter in the direction of distribution. Use the FVG, an order block or the OTE band of the displacement leg.
- Target the far side. The previous day’s high or low, or another external liquidity pool, sized to what gold’s daily range can realistically deliver.
- Invalidate on the extreme. If price breaks beyond the manipulation extreme after your entry, the Po3 read is wrong.
| Bias | Manipulation should be | Distribution should be | Invalid if |
|---|---|---|---|
| Bullish day | Below the midnight open, often taking the Asian low | Expansion above the open toward buy-side liquidity | Price breaks the manipulation low after the shift |
| Bearish day | Above the midnight open, often taking the Asian high | Expansion below the open toward sell-side liquidity | Price breaks the manipulation high after the shift |
The Judas swing in detail
The Judas swing is ICT’s name for the manipulation leg. Common features traders look for:
- It happens early, usually between about 00:00 and 05:00 New York time.
- It trades against the expected daily direction and away from the midnight open.
- It takes an obvious level: the Asian high or low, the previous day’s level, or equal highs or lows.
- It reverses with displacement, not a slow drift.
The hard truth: a Judas swing is only obvious afterward. In real time, a London drop below the open could be the Judas swing, or it could be the start of a bearish distribution. Your bias, your levels and your invalidation rule are what separate the two. Nothing in the candle itself tells you.
Variations and settings: from beginner to pro
Beginner
- Frame only the daily candle with the midnight open.
- Only trade when the higher-timeframe bias is clear.
- Wait for the Asian range to be taken and for a structure shift back through the open before entering.
- Use a fixed small risk and one target.
Intermediate
- Add a session Po3: treat the New York session as its own candle with the 08:30 open, looking for a raid shortly after the release and expansion afterward.
- Use the London kill zone and Silver Bullet windows from the kill zones guide for timing.
- Track the ratio of the manipulation leg to the average daily range. A “Judas” that already covers most of the day’s typical range is suspicious.
Pro
- Weekly Po3: some traders expect the weekly high or low to form early in the week in a trending market, then the rest of the week to expand away. Treat this as a hypothesis to test, not a rule.
- Nested Po3: a session AMD inside a daily AMD inside a weekly AMD. It can sharpen timing, but it also makes it easy to find a pattern anywhere. Keep strict rules.
- Opening range checks: compare the midnight open, the 08:30 open and the 09:30 equity open. When price sits below all of them on a bullish-bias day, the discount is deeper.
- Correlation: use SMT divergence between gold and silver at the manipulation extreme as an extra filter.
Three setups traders use
1. London Judas below the midnight open (bullish day)
| Part | Detail |
|---|---|
| Context | Daily bias bullish; price in discount of the higher-timeframe range; no major data until 08:30. |
| Trigger | During the London kill zone, price trades below the midnight open and the Asian low, then displaces back above the Asian low with an M5 or M15 structure shift. |
| Entry | Retrace into the FVG or order block formed by the displacement, ideally still at or below the midnight open. |
| Stop | Below the manipulation low plus a spread buffer. |
| Targets | First: the Asian high. Second: the previous day’s high or a buy-side pool. |
| Management | Consider reducing risk ahead of 08:30 data. Take partials before the release if the trade is in profit. |
2. New York session Po3 around the 08:30 open
| Part | Detail |
|---|---|
| Context | Clear daily bias; data at 08:30. London has not delivered a clean move. |
| Trigger | After the 08:30 release, price spikes against the bias, takes a nearby level, then displaces back through the 08:30 open. |
| Entry | FVG or OTE retracement of the displacement, after spreads normalize. |
| Stop | Beyond the spike extreme. |
| Targets | The session high or low on the far side, then the daily draw. |
| Management | Never enter during the first seconds of the release. Expect slippage. See the news trading guide. |
3. The defensive rule: failed manipulation
| Part | Detail |
|---|---|
| Context | You expected a bullish day and price drops below the midnight open. |
| Trigger | The bounce stalls at or under the midnight open, and price makes a new low below the supposed manipulation low. |
| Entry | No long. Some traders flip to a short on a retest of the open from below; others simply stand aside. |
| Stop | For any flip trade, above the failed retest high. |
| Targets | The next sell-side liquidity pool. |
| Management | Log it as a failed Po3. Count these days. They tell you how reliable your bias process really is. |
Worked example (hypothetical)
Account: $25,000. Risk per trade: 0.4% = $100.
Context: the daily bias is bullish. The previous day’s high is 4,168.0. Overnight, gold builds an Asian range from 4,146.6 to 4,153.6. The midnight open is 4,150.4.
Manipulation: around 02:30 New York time, gold drops to 4,138.2, below both the open and the Asian low. An M15 candle then rallies hard back above 4,146.6, breaking the last M5 lower high and leaving an FVG at 4,143.0–4,145.2.
Plan: limit buy at 4,144.0. Stop at 4,137.0, below the manipulation low with a buffer. Stop distance: $7.00. First target: the Asian high at 4,153.6. Final target: 4,167.5, just under the previous day’s high.
Position size: $100 ÷ $7.00 = $14.29 per $1 move, so 0.1429 lots, rounded down to 0.14 lots. Actual risk: 0.14 × 100 × $7.00 = $98.
Management: take half off at 4,153.6 (+$9.60 × $7 = $67.20). Move the stop on the remainder to 4,143.0 only after a new M15 higher low forms. If the rest reaches 4,167.5, that half earns $23.50 × $7 = $164.50.
Invalidation: a trade below 4,137.0 means the drop was not manipulation. It may have been the start of a bearish distribution day, like the failed Po3 chart above.
This example has a clean outcome because it is drawn to explain the model. In live gold trading, many London drops below the open just keep going. Use the position size calculator and accept that the stop will be hit regularly.
Combining it with other concepts
- Kill zones and the Silver Bullet: give the timing for the manipulation and distribution phases.
- Liquidity sweeps: the manipulation phase is a liquidity sweep with a time expectation attached.
- Asian range breakout: the opposite reading of the same range. A breakout trader buys the break; a Po3 trader fades the first break. Knowing both helps you understand who gets trapped.
- Wyckoff: a Spring is a larger-scale cousin of the bullish Judas swing.
- Premium and discount: a bullish manipulation leg that ends in discount of the daily range is a better location.
- Multi-timeframe analysis: the daily bias decides which way you expect the manipulation to go.
Best timeframes and sessions
- Framing: daily candle (most common), weekly for swing traders, a single session for intraday traders.
- Reading the phases: M15 to H1.
- Entries: M1 to M5 inside the London or New York kill zones.
- Sessions: the Asian session for marking accumulation; London for the most common manipulation; New York for the most common distribution. On news days, expect the whole sequence to compress into New York.
Common mistakes
- Labeling every early move as a Judas swing. Fix: require a liquidity raid and displacement back through a key level.
- No daily bias. Fix: without a bias, you can’t know which direction is “manipulation.” Write the bias before London opens.
- Mixing reference opens. Fix: pick midnight or 18:00 and use it every day.
- Fading a trend day. Fix: if price never trades against the open and keeps expanding, there is no Po3 trade.
- Moving the stop when the “Judas” extends. Fix: the manipulation extreme is fixed once you enter. If it breaks, you were wrong.
- Ignoring data. Fix: if 08:30 data lands mid-trade, reduce size or take profits before it.
Limitations and honest caveats
Po3 encourages buying weakness and selling strength. When the bias is wrong, that means buying into a falling market. Always keep the stop beyond the manipulation extreme, keep risk small and read XAUUSD risk management.
- Hindsight bias is strong. Almost every finished candle can be described as AMD. The model’s real test is whether you can call the phases before they finish.
- No verified statistics. There is no reliable public figure for how often gold days follow Po3. Claims online are rarely auditable.
- Intent is assumed. “Manipulation” implies a deliberate actor. A false move may simply be early traders being wrong, stops triggering or the dollar moving.
- Candle shape depends on your broker’s day start. Two traders can see different daily candles for the same day.
- Macro dominates. A data surprise can turn the planned distribution into a reversal at any point.
Pro tips
- Count the failed days. Your journal should record how often the expected manipulation turned into the real move. That number tells you how much to trust your bias process.
- Measure the manipulation leg in ATR terms. If the “Judas” has already used most of the day’s typical range, the distribution may not have room to pay you.
- Prefer raids that take two levels at once, such as the Asian low and the midnight open together, or the Asian low and a round number.
- Watch the return to the open. In a valid bullish Po3, price often reclaims the midnight open with conviction. A slow drift under it is a warning.
- On Mondays, be careful. The Sunday open can gap, and the Asian range can be distorted. Many traders skip Monday Po3 setups.
- Use the close as feedback. At 17:00 New York time, check whether the daily candle actually closed in the direction of your bias. That simple tally is one of the best checks on your bias skill.
Quick checklist
- Reference open chosen and drawn (midnight or 18:00).
- Daily bias and draw on liquidity written down before London.
- Asian range high and low marked.
- Manipulation seen in the expected direction, taking a real level.
- Displacement and structure shift back through the level.
- Entry at an FVG, order block or OTE, ideally on the discount (long) or premium (short) side of the open.
- Stop beyond the manipulation extreme; size calculated from it.
- Target set at a realistic liquidity pool, checked against the day’s typical range.
- Data times checked and a plan for them in place.
- Outcome logged, including whether the daily candle closed as expected.
Frequently asked questions
What is the ICT Power of 3?
Power of 3 (Po3) is an ICT model that describes a price candle in three phases: accumulation, manipulation and distribution, often shortened to AMD. On a bullish day, price opens, first trades below the open to take sell-side liquidity, then rallies and closes near the high. On a bearish day the order is reversed.
What is the Judas swing?
The Judas swing is the manipulation phase: an early move in the session that runs against the day’s eventual direction, often taking an obvious high or low before reversing. In ICT teaching it is most often expected in the London kill zone, roughly 02:00–05:00 New York time.
What is the midnight open in ICT?
It is the price at 00:00 New York time. ICT uses it as the reference open for the trading day. In a bullish Po3 day, traders look for price to trade below the midnight open before rallying; in a bearish day, above it before falling. Some traders use the 18:00 New York gold futures open instead, so state which one you use.
Does Power of 3 work on gold?
Gold’s session structure, with a quieter Asian range followed by more active London and New York sessions, often produces days that look like AMD in hindsight. There is no verified figure for how often the pattern plays out, and many days do not follow it. Test the rules on your own XAUUSD data.
What is the difference between Power of 3 and Wyckoff?
Both describe accumulation, a shakeout and a markup or markdown. Wyckoff works over weeks or months with defined phases and events like the Spring. Power of 3 applies the idea to a single candle, usually the daily or a session, and adds time-based expectations such as the London Judas swing.
Which timeframe should I use for Power of 3?
Most traders apply Po3 to the daily candle and read it on M15 to H1 charts. It can also be applied to the weekly candle or a single session. Pick one candle to frame, then use a lower timeframe for entries.
Sources and further reading
- CME Group — Gold futures overview — Contract specifications and trading hours for COMEX gold
- CFTC — Commitments of Traders — Weekly futures positioning, including gold
- Investor.gov — Stop order — How stop orders work, and their limits in fast markets
External sites are independent of XAUUSD Global. Check dates and data yourself.