Claim check · NQ futures
ICT Power of 3 test on NQ futures daily candles: 2,419 sessions, 2016–2026
Summary
ICT’s Power of 3 daily candle shape, counted on 2,419 NQ futures sessions from 2016-09-26 to 2026-08-27: the strict shape appeared on 8.6% of sessions, and on 8.3% when the same days’ 1-minute bars were shuffled inside 30-minute blocks. The loose shape appeared on 56.1% of sessions against 55.6% after the shuffle, with no trades or costs involved. Tested by CE Market Solutions (cemarketsolutions.com), a backtesting service for CME futures strategies.
The claim tested
ICT (Inner Circle Trader), the Power of 3 (accumulation, manipulation, distribution) daily candle model.
A daily candle first moves against its eventual direction, often through an old swing level, then runs the other way and closes in the direction of the day.
Rules exactly as tested
- Session: one NQ daily candle, 18:00 ET to the 16:59 ET close.
- Low-first shape (S1): on an up day, the low came before the high. Down days are the mirror.
- Loose Power of 3 (S2): the low came first and the day closed in the top quarter of its range. Down days are the mirror.
- Strict Power of 3 (S3): the low came first, the day closed in the top quarter, it opened in the lower half of its range, and the low went beyond the old 15-minute swing low. Down days are the mirror.
- Old swing level: the most recent confirmed 15-minute swing (2 bars each side) formed before 18:00 on the same contract, beyond the opening range.
- M1: of the days where the low came first, the share that went beyond the old 15-minute swing low. M2: the share that went beyond the previous session’s low (previous day’s high for down days).
- Nulls: Null A shuffles all of a day’s 1-minute bars. Null B shuffles the bars only inside each 30-minute block. Both keep the day’s open and close exact. Each real day is compared with 20 shuffles.
Assumptions where the rules were ambiguous:
- Down days are mirrored and pooled with up days in the combined rows.
- The accumulation range is the bars before the first move of 0.25 of the day’s range away from the open.
- Intervals are day-bootstrap, 10,000 draws, paired between the real days and the shuffled days.
Data
- Instrument
- NQ (E-mini Nasdaq-100) futures, continuous contract
- Bar size
- 1 minute, with 15-minute swings
- Date range
- 2016-09-26 to 2026-08-27
- Sample size
- 2,563 sessions, of which 2,419 valid (1,342 up days and 1,077 down days). 39 roll days, 3 late starts, 97 short days and 5 flat days were set aside.
- Costs
- Not applicable: descriptive counts, no trades.
- In-sample or out-of-sample
- Whole-history descriptive statistics. Nothing was fitted, so the split does not apply.
Results
| Measure | Real | Null A (full shuffle) | Null B (shuffle in 30-minute blocks) | Real minus Null B, points (95% interval) |
|---|---|---|---|---|
| S1 low first | 87.8% | 88.6% | 88.0% | −0.2 (−0.8 to +0.4) |
| S2 loose Power of 3 | 56.1% | 52.4% | 55.6% | +0.5 (−0.3 to +1.3) |
| S3 strict Power of 3 | 8.6% | 7.6% | 8.3% | +0.2 (−0.4 to +0.9) |
| M1 old swing low taken | 26.8% | 24.6% | 25.6% | +1.1 (+0.3 to +2.0) |
| M2 previous-session low taken | 32.5% | 30.7% | 31.6% | +0.9 (+0.3 to +1.6) |
| Median minutes after 18:00 to the first 0.25-range move | 504 | 155 | 508 | −4 (−12 to +3) |
| Measure | Real | Null A | Null B | Real minus Null B, points (95% interval) |
|---|---|---|---|---|
| Up days (1,342): S1 | 86.8% | 89.9% | 87.1% | −0.3 (−1.2 to +0.6) |
| Up days: S2 | 64.2% | 55.6% | 62.8% | +1.4 (+0.4 to +2.4) |
| Up days: S3 | 9.8% | 7.5% | 9.4% | +0.5 (−0.5 to +1.4) |
| Down days (1,077): S1 | 89.0% | 87.0% | 89.0% | −0.0 (−0.9 to +0.8) |
| Down days: S2 | 46.1% | 48.4% | 46.7% | −0.6 (−1.8 to +0.6) |
- The low of the day fell between 09:00 and 10:00 ET on 12.3% of up days. The within-block shuffle gave 12.2%; the full shuffle gave 1.8%.
- M1 and M2 ran 1.8 points above Null B on up days only, about 21 of 1,165 low-first up days. On up days the mean gap from the open down to the low was 2.03 points (3.2%) larger than in the within-block shuffle (95% interval +1.36 to +2.69).
Questions and answers
How often does the ICT Power of 3 pattern appear on NQ daily candles?
The strict shape appeared on 8.6% of 2,419 valid NQ sessions from 2016-09-26 to 2026-08-27, and the loose shape on 56.1%.
How was the Power of 3 tested?
Each daily candle (18:00 to 16:59 ET) was labelled for low-first, loose and strict Power of 3. Those rates were compared with the same days after their 1-minute bars were shuffled, either fully or only inside 30-minute blocks.
What sample size was used for the Power of 3 test?
2,419 valid sessions: 1,342 up days and 1,077 down days.
Does the Power of 3 shape appear more often than in shuffled data?
Against the within-block shuffle the strict shape differed by +0.2 points (95% interval −0.4 to +0.9) and the loose shape by +0.5 points (−0.3 to +1.3).
Related reports
- ICT Silver Bullet backtest on NQ futures: 426 trades, 2016–2026
- Asia range sweep reversal backtest on NQ futures: 3,622 trades, 2016–2026
- ICT opening range gap 50% level test on Nasdaq futures: 732 sessions, 2022–2025
Read more: R multiples and expectancy, explained · How many trades does a backtest need?
Want your own strategy tested the same way? Every version is $19 at cemarketsolutions.com.
More tests like this on Instagram and TikTok: @cemarketsolutions.