ICT lab

ICT concepts put to the test on NQ history: FVG fills, previous day high and low sweeps, timing of the extremes, Asian range, OHLC mapping, reaction to releases. Observed frequencies, each with its sample size and period.

55%of 5-minute FVGs fully filled within the next 5 barsN = 2,841 · Jul 17 to Sep 25, 2026

71%of daily highs formed during New York, which covers 39% of the dayN = 559 · May 6, 2024 to Sep 25, 2026

67%of cash sessions with their high or low within the first 30 minutesN = 49 · Jul 17 to Sep 25, 2026

56%of days when New York takes the other side of the Asian range after London took only oneN = 400 · May 6, 2024 to Sep 25, 2026

Descriptive, not predictive. These statistics describe what happened on NQ; none of them says which way the market will go.

FVG fill rates, 5-minute bars

N = 2,843 FVGs · Jul 17 to Sep 25, 2026 · 5-minute bars

5-minute FVGs: 70% partially and 55% fully filled within the next 5 bars (N = 2,843, from July 17, 2026 to September 25, 2026).

Median size: 6.25 ptsmiddle 50% 2.75 to 13
Days covered: 49

Within the next bars	N	Partial	Full
1 bar	2,843	42%CI 40-44	24%CI 22-25
5 bars	2,841	70%CI 68-71	55%CI 53-56
20 bars	2,828	84%CI 83-86	75%CI 74-77

By direction, within the next 5 bars

Direction	N	Partial	Full
Bullish	1,488	68%CI 66-70	53%CI 50-56
Bearish	1,355	71%CI 69-74	56%CI 54-59

Within the next 5 bars

	N	Partial	Full
Session
Asia (18:00-02:00 NY)	960	71%CI 68-74	56%CI 53-59
London (02:00-08:00 NY)	792	69%CI 66-72	55%CI 51-58
New York (08:00-17:00 NY)	1,091	69%CI 66-71	53%CI 50-56
Size, at the current price
Smallest quarter0 to 2.9 pts	711	70%CI 66-73	67%CI 64-70
2nd quarter2.9 to 6.7 pts	710	72%CI 68-75	62%CI 58-65
3rd quarter6.7 to 13.7 pts	711	71%CI 68-74	53%CI 49-57
Largest quarter13.7 pts and above	711	66%CI 62-69	36%CI 33-40

Definition and source

Fair value gap (FVG) on 3 consecutive bars: bullish when the 3rd bar's low is above the 1st bar's high (zone = between them), bearish in the opposite case. Partial fill: a later bar trades into the zone (strictly). Full fill: a later bar reaches the far edge. Measured over the 1, 5 and 20 bars after the 3rd. Session assigned from the middle bar's time. Size quarters computed in % of price (converted to points at the current price). Non-consecutive bars (CME pause, data gap) and contract switch days excluded.

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract)

FVG fill rates, 1-hour bars

N = 2,600 FVGs · May 3, 2024 to Sep 25, 2026 · 1-hour bars

1-hour FVGs: 69% partially and 52% fully filled within the next 5 bars (N = 2,600, from May 3, 2024 to September 25, 2026).

Median size: 17.5 ptsmiddle 50% 6.75 to 40.25
Days covered: 586

Within the next bars	N	Partial	Full
1 bar	2,593	43%CI 41-44	23%CI 22-25
5 bars	2,586	69%CI 67-71	52%CI 50-54
20 bars	2,527	84%CI 82-85	73%CI 72-75

By direction, within the next 5 bars

Direction	N	Partial	Full
Bullish	1,459	68%CI 66-70	52%CI 49-54
Bearish	1,141	71%CI 68-73	52%CI 49-55

Within the next 5 bars

	N	Partial	Full
Session
Asia (18:00-02:00 NY)	856	72%CI 68-74	56%CI 52-59
London (02:00-08:00 NY)	692	80%CI 77-83	67%CI 63-70
New York (08:00-17:00 NY)	1,044	60%CI 57-63	39%CI 36-42
Size, at the current price
Smallest quarter0 to 9.3 pts	650	75%CI 71-78	72%CI 69-76
2nd quarter9.3 to 22.6 pts	650	70%CI 66-73	59%CI 55-63
3rd quarter22.6 to 51.8 pts	650	69%CI 65-72	47%CI 44-51
Largest quarter51.8 pts and above	650	63%CI 59-67	29%CI 25-32

Definition and source

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract)

Previous day high / low sweeps during New York · 1 h

N = 519 days · May 24, 2024 to Sep 25, 2026 · 1-hour bars

PDH untouched at the open: taken during the New York window in 30% of cases; after the sweep, back inside the range 25%, continuation 40% (N = 519 days, from May 24, 2024 to September 25, 2026).

	Previous day high (PDH)	Previous day low (PDL)
Already taken before the New York window	38%CI 34-42	27%CI 24-31
Taken during the window, when untouched	30%CI 25-35	22%CI 18-26
Outcome after the sweep
Continuation	40%CI 31-50	42%CI 32-53
Back inside the range	25%CI 17-34	34%CI 24-44
Undecided	35%CI 26-45	24%CI 16-34
Median extension beyond the level	83.5 ptsmiddle 50% 35 to 145 · 0.21 ATR	137.5 ptsmiddle 50% 44.8 to 241 · 0.32 ATR

Both sides taken during New York: 6.2%CI 4.4-8.6 · New York window: 09:00-16:00 NY (1-hour bars, so the window starts 30 min before the cash open)

Definition and source

Previous day high (PDH) and low (PDL) = previous CME trade date (18:00-17:00 NY). Only days where the level is still untouched when the New York window opens (not taken overnight). Sweep: the New York window trades beyond the level. Outcome at the end of the window: continuation if the close is beyond the level by at least 10% of the prior day's 14-day ATR, back inside the range if it closes back inside by at least 10% of the ATR, undecided otherwise. Extension = maximum excursion beyond the level.

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract)

Previous day high / low sweeps during New York · 5 min

N = 37 days · Jul 22 to Sep 25, 2026 · 5-minute bars

PDH untouched at the open: taken during the New York window in 25% of cases; after the sweep, back inside the range 40%, continuation 40% (N = 37 days, from July 22, 2026 to September 25, 2026).

	Previous day high (PDH)	Previous day low (PDL)
Already taken before the New York window	46%CI 31-62	24%CI 13-40
Taken during the window, when untouched	25%CI 11-47	21%CI 10-40
Outcome after the sweep
Continuation	40%CI 12-77	33%CI 10-70
Back inside the range	40%CI 12-77	50%CI 19-81
Undecided	20%CI 4-62	17%CI 3-56
Median extension beyond the level	119.5 ptsmiddle 50% 103.5 to 238.5 · 0.21 ATR	187.3 ptsmiddle 50% 71.1 to 260.1 · 0.35 ATR

Both sides taken during New York: 2.7%CI 0.5-13.8 · New York window: 09:30-16:00 NY (exact cash session)

Definition and source

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract)

Time of the daily high and low

N = 559 days · May 6, 2024 to Sep 25, 2026 · 1-hour bars

Daily high formed during New York (08:00-17:00) 71% of the time, low 61%, for 39% of the day (N = 559 days, from May 6, 2024 to September 25, 2026).

00:00

01:00

02:00

03:00

04:00

05:00

06:00

07:00

08:00

09:00

10:00

11:00

12:00

13:00

14:00

15:00

16:00

17:00

18:00

19:00

20:00

21:00

22:00

HighLowUniform baseline 4.3%Paris time

The figures

New York	Paris	High	Low
18:00 NY	00:00	7.2%CI 5.3-9.6	12%CI 10-15
19:00 NY	01:00	2.9%CI 1.8-4.6	3.6%CI 2.3-5.5
20:00 NY	02:00	3.2%CI 2.1-5.0	4.1%CI 2.8-6.1
21:00 NY	03:00	1.8%CI 1.0-3.3	1.4%CI 0.7-2.8
22:00 NY	04:00	1.4%CI 0.7-2.8	1.4%CI 0.7-2.8
23:00 NY	05:00	0.5%CI 0.2-1.6	1.1%CI 0.5-2.3
00:00 NY	06:00	0.4%CI 0.1-1.3	0.9%CI 0.4-2.1
01:00 NY	07:00	1.4%CI 0.7-2.8	2.0%CI 1.1-3.5
02:00 NY	08:00	1.6%CI 0.9-3.0	1.3%CI 0.6-2.6
03:00 NY	09:00	3.0%CI 1.9-4.8	3.0%CI 1.9-4.8
04:00 NY	10:00	1.8%CI 1.0-3.3	3.2%CI 2.1-5.0
05:00 NY	11:00	0.9%CI 0.4-2.1	2.3%CI 1.4-3.9
06:00 NY	12:00	1.1%CI 0.5-2.3	1.1%CI 0.5-2.3
07:00 NY	13:00	2.0%CI 1.1-3.5	1.4%CI 0.7-2.8
08:00 NY	14:00	5.6%CI 3.9-7.8	3.6%CI 2.3-5.5
09:00 NY	15:00	10%CI 8-13	9.1%CI 7.0-11.8
10:00 NY	16:00	8.9%CI 6.9-11.6	11%CI 8-14
11:00 NY	17:00	6.8%CI 5.0-9.2	5.2%CI 3.6-7.4
12:00 NY	18:00	3.8%CI 2.5-5.7	4.5%CI 3.1-6.5
13:00 NY	19:00	6.1%CI 4.4-8.4	5.2%CI 3.6-7.4
14:00 NY	20:00	5.4%CI 3.8-7.6	5.2%CI 3.6-7.4
15:00 NY	21:00	14%CI 11-17	7.7%CI 5.8-10.2
16:00 NY	22:00	11%CI 8-13	9.5%CI 7.3-12.2

	High	Low	Either	Share of the day
Sessions
Asia (18:00-02:00 NY)00:00-08:00 Paris	19%CI 16-22	27%CI 23-30	45%CI 40-49	35%
London (02:00-08:00 NY)08:00-14:00 Paris	10%CI 8-13	12%CI 10-15	23%CI 19-26	26%
New York (08:00-17:00 NY)14:00-23:00 Paris	71%CI 67-74	61%CI 57-65	98%CI 96-99	39%
Killzones
Asian killzone19:00-22:00 NY · 01:00-04:00 Paris	7.9%CI 5.9-10.4	9.1%CI 7.0-11.8	17%CI 14-20	13%
London open killzone02:00-05:00 NY · 08:00-11:00 Paris	6.4%CI 4.7-8.8	7.5%CI 5.6-10.0	14%CI 11-17	13%
London Silver Bullet03:00-04:00 NY · 09:00-10:00 Paris	3.0%CI 1.9-4.8	3.0%CI 1.9-4.8	6.1%CI 4.4-8.4	4.4%
New York open killzone07:00-09:00 NY · 13:00-15:00 Paris	7.5%CI 5.6-10.0	5.0%CI 3.5-7.1	13%CI 10-16	8.7%
New York AM Silver Bullet10:00-11:00 NY · 16:00-17:00 Paris	8.9%CI 6.9-11.6	11%CI 8-14	20%CI 17-23	4.4%
London close killzone10:00-12:00 NY · 16:00-18:00 Paris	16%CI 13-19	16%CI 13-19	31%CI 27-35	8.7%
New York PM Silver Bullet14:00-15:00 NY · 20:00-21:00 Paris	5.4%CI 3.8-7.6	5.2%CI 3.6-7.4	11%CI 8-13	4.4%

Paris times: New York time + 6 h (+ 5 h during the weeks when only one of the two countries has changed clocks).

Definition and source

For each complete CME trade date (23 one-hour bars), time of the first bar reaching the day's high and low. Breakdown by hour, session and killzone. “Share of the day” = share expected if the extreme fell at random in the day (window length / 23 h), to compare with the result.

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract) · innercircletrader.net, “ICT Killzones” · ictkillzone.com, “ICT Kill Zones” · chartwhisperer.ca, “ICT Silver Bullet & Killzones”

Time of the cash session high and low

N = 49 sessions · Jul 17 to Sep 25, 2026 · 5-minute bars

Cash session: the high or the low forms in the first 30 minutes on 67% of days (N = 49, from July 17, 2026 to September 25, 2026).

Cash session high or low formed within the first 30 minutes: 67%CI 53-79one 30-minute slot is 7.7% of the session

15:30

16:00

16:30

17:00

17:30

18:00

18:30

19:00

19:30

20:00

20:30

21:00

21:30

HighLowUniform baseline 7.7%Paris time

The figures

New York	Paris	High	Low
09:30-10:00 NY	15:30-16:00	35%CI 23-49	35%CI 23-49
10:00-10:30 NY	16:00-16:30	4.1%CI 1.1-13.7	22%CI 13-36
10:30-11:00 NY	16:30-17:00	4.1%CI 1.1-13.7	6.1%CI 2.1-16.5
11:00-11:30 NY	17:00-17:30	8.2%CI 3.2-19.2	4.1%CI 1.1-13.7
11:30-12:00 NY	17:30-18:00	8.2%CI 3.2-19.2	6.1%CI 2.1-16.5
12:00-12:30 NY	18:00-18:30	4.1%CI 1.1-13.7	2.0%CI 0.4-10.7
12:30-13:00 NY	18:30-19:00	2.0%CI 0.4-10.7	0%CI 0-7
13:00-13:30 NY	19:00-19:30	4.1%CI 1.1-13.7	2.0%CI 0.4-10.7
13:30-14:00 NY	19:30-20:00	2.0%CI 0.4-10.7	2.0%CI 0.4-10.7
14:00-14:30 NY	20:00-20:30	4.1%CI 1.1-13.7	4.1%CI 1.1-13.7
14:30-15:00 NY	20:30-21:00	0%CI 0-7	2.0%CI 0.4-10.7
15:00-15:30 NY	21:00-21:30	8.2%CI 3.2-19.2	4.1%CI 1.1-13.7
15:30-16:00 NY	21:30-22:00	16%CI 9-29	10%CI 4-22

Paris times: New York time + 6 h (+ 5 h during the weeks when only one of the two countries has changed clocks).

Definition and source

Cash session 09:30-16:00 NY (at least 76 of 78 five-minute bars), 30-minute slot of the first bar reaching the session high and low.

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract)

Asian range sweeps by London then New York

N = 559 days · May 6, 2024 to Sep 25, 2026 · 1-hour bars

Asian range (20:00-00:00 NY): London takes exactly one side 72% of the time, both 25%; when London takes only one, New York takes the other 56% of the time (N = 559 days, from May 6, 2024 to September 25, 2026).

Median Asian range (20:00-00:00 NY): 81.5 ptsmiddle 50% 53.4 to 131.9 · 0.35% of price

When London takes only one side, New York takes the other: 56%CI 51-60

	London02:00-08:00 NY · 08:00-14:00 Paris	New York08:00-16:00 NY · 14:00-22:00 Paris
High only taken	41%CI 37-45	27%CI 24-31
Low only taken	31%CI 27-35	20%CI 17-24
Both sides taken	25%CI 21-28	52%CI 48-56
Neither side taken	3.8%CI 2.5-5.7	0.4%CI 0.1-1.3

Definition and source

Asian range = high and low from 20:00 to 00:00 NY (previous evening; ICT “Asian range”, wider than the 19:00-22:00 Asian killzone used in the plan). London = 02:00-08:00 NY, New York = 08:00-16:00 NY (00:00-02:00 not counted). A side is taken when the window trades strictly beyond it. The New York row is measured independently of London.

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract)

“OHLC Statistical Mapping” versus realized prices

N = 2,415 days · Oct 26, 2016 to Sep 25, 2026 · daily bars

With the final direction known in hindsight, the low (up day) or the high (down day) stays inside the manipulation zone on 59% of days and the distribution target is reached on 43% of days. Without knowing the direction: once the lower manipulation level breaks, the day closes up 35% of the time, versus 56% on average (N = 2,415 days, from October 26, 2016 to September 25, 2026).

Original concept and indicator: “OHLC Statistical Mapping” by Joshuuu, toodegrees collection on TradingView. The test replicates its calculation rules exactly.

20-day average manipulation, median: 62.1 ptsmiddle 50% 32.3 to 86.3
20-day average distribution, median: 178.6 ptsmiddle 50% 94.8 to 253.2
Up days: 56%CI 54-58

With the final direction known (hindsight)

	Up days	Down days	All days
Opposite extreme stayed inside the manipulation zone	60%CI 58-63	57%CI 54-60	59%CI 57-61
Distribution target reached	39%CI 37-42	48%CI 45-51	43%CI 41-45

Without hindsight (observable during the session)

	Frequency	Baseline
Upper distribution target touched	23%CI 22-25
Lower distribution target touched	24%CI 22-26
Both manipulation levels touched	36%CI 35-38
Up close after the lower manipulation level breaks	35%CI 32-37	56% of all days
Down close after the upper manipulation level breaks	27%CI 25-30	44% of all days

82 days excluded (quarterly expiries and the following days).

Definition and source

Exact replica of the site's indicator (daily timeframe, N = 20 closed bars): Distribution = average extension in the final direction (up: high - open; down: open - low), Manipulation = average wick against the final direction. Manipulation zone of an up day = [open - manipulation, open] (down day: symmetric). Distribution target = open +/- distribution. Measures “with the direction known” use the day's close (hindsight); “without hindsight” measures only use what is observable during the session. Yahoo daily bars (open at 18:00 NY), expiry day and next day excluded (contract change).

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract) · original concept: “OHLC Statistical Mapping” indicator by Joshuuu, toodegrees collection (TradingView)

NQ reaction to macro releases

N = 206 releases · 5-minute and 1-hour bars

CPI inflationBLS · 08:30 NY · 14:30 Paris
1 release at the same minute as: Retail sales
Median 1-hour reaction · window 08:00-10:00 NY121 ptsordinary day 86 · N = 28 · May 15, 2024 to Sep 11, 2026
Over 5 min31 ptsordinary day 13 · N = 2
Employment report (NFP)BLS · 08:30 NY · 14:30 Paris
2 releases at the same minute as: Retail sales
Median 1-hour reaction · window 08:00-10:00 NY122 ptsordinary day 86 · N = 25 · May 3, 2024 to Sep 4, 2026
Over 5 min100 ptsordinary day 13 · N = 2
PPI producer pricesBLS · 08:30 NY · 14:30 Paris
4 releases at the same minute as: Retail sales
Median 1-hour reaction · window 08:00-10:00 NY69 ptsordinary day 86 · N = 28 · May 14, 2024 to Sep 10, 2026
Over 5 min69 ptsordinary day 13 · N = 2
JOLTS job openingsBLS · 10:00 NY · 16:00 Paris
Median 1-hour reaction · window 10:00-11:00 NY130 ptsordinary day 80 · N = 26 · Jun 4, 2024 to Sep 1, 2026
Over 5 min11 ptsordinary day 29 · N = 2
US GDPBEA · 08:30 NY · 14:30 Paris
8 releases at the same minute as: Personal income and outlays (PCE)
Median 1-hour reaction · window 08:00-10:00 NY123 ptsordinary day 86 · N = 27 · May 30, 2024 to Aug 26, 2026
Over 5 min44 ptsordinary day 13 · N = 2
Personal income and outlays (PCE)BEA · 08:30 NY · 14:30 Paris
8 releases at the same minute as: US GDP
Median 1-hour reaction · window 08:00-10:00 NY112 ptsordinary day 86 · N = 22 · May 31, 2024 to Aug 26, 2026
Over 5 min44 ptsordinary day 13 · N = 2
Personal income and outlays (PCE)BEA · 10:00 NY · 16:00 Paris
Median 1-hour reaction · window 10:00-11:00 NY116 ptsordinary day 80 · N = 4 · Nov 27, 2024 to Jan 22, 2026
Over 5 minn/a
Retail salesCensus · 08:30 NY · 14:30 Paris
7 releases at the same minute as: CPI inflation, Employment report (NFP), PPI producer prices
Median 1-hour reaction · window 08:00-10:00 NY83 ptsordinary day 86 · N = 27 · May 15, 2024 to Sep 16, 2026
Over 5 min22 ptsordinary day 13 · N = 2
Fed decision (FOMC)Federal Reserve · 14:00 NY · 20:00 Paris
Median 1-hour reaction · window 14:00-15:00 NY158 ptsordinary day 37 · N = 19 · Jun 12, 2024 to Sep 16, 2026
Over 5 min62 ptsordinary day 11 · N = 2

Magnitude without direction, in points at the current price (medians converted from %). Light bar: release; gray bar: ordinary day.

Definition and source

Magnitude only, never direction (it depends on the surprise). 5-minute bars: reference price = close of the bar ending at the release time; absolute move at 5, 30 and 60 min and maximum excursion over the window. 1-hour bars: from the close of the bar before the release to the close of the first bar ending at least 60 min after (8:30: 08:00-10:00 window). “Ordinary day” = same measure, same time, on business days without any collected release (Thursdays excluded at 8:30: jobless claims). Values in points at the time, in % and in points at the current price (NQ went from about 18,000 to over 30,000 points over the period).

Source: Yahoo Finance, chart API v8 (NQ=F, continuous contract) · official calendars (BLS, BEA, Census Bureau, Federal Reserve)

Limitations

Yahoo NQ=F continuous contract: the intraday series switches contract a few days before the quarterly expiry (sometimes mid-session, sometimes flipping back and forth), the daily series after it. Switch days are detected by comparing both series and excluded; older levels are shifted by the jump (additive method) when compared with the current contract.
Yahoo data quality: gaps (mostly Sunday evenings and holidays), corrected bad prints (open far away that snaps back), sometimes duplicated volumes. Prices delayed by about 10 minutes.
Short windows: 60 rolling days of 5-minute bars at Yahoo (the local cache extends them by one day per day of collection), about 2 years of 1-hour bars. Confidence intervals are wide on small samples.
Descriptive statistics of the past, not forecasts. None of them says which way the market will go.

Method

Data and cleaning

Yahoo Finance NQ=F continuous contract: 5-minute bars over about two months (the history grows by one day per day of collection), 1-hour bars over about two years, daily bars over ten years. Days when the series switches to the next contract are detected by comparing the intraday and daily series, then excluded; bad prints (an open far away that snaps back) are corrected.

CME trade date, sessions and times

A CME trade date runs from 18:00 the day before to 17:00, New York time. Sessions: Asia from 18:00 to 02:00, London from 02:00 to 08:00, New York from 08:00 to 17:00 (New York time). The statistics are computed in New York time; charts and tables add Paris time, six hours later, five during the weeks when only one of the two countries has changed clocks.

Proportions and confidence intervals

Each proportion is a number of cases (k) out of a sample size (n). The 95% confidence interval (Wilson method), marked CI, gives the range compatible with the data: it is wide when the sample is small. Two proportions whose intervals largely overlap are not really different. Hover over a figure to see k and n.

Why this lab

ICT concepts circulate as rules and hand-picked examples. This lab counts them, over every available day, with definitions written in advance and published. The result describes what happened: it says neither when to enter nor in which direction. The program runs every night on our server, without AI: the sentences are templates filled with the figures. This is not investment advice.

Today's session plan

Computed:: Sep 26, 2026, 01:22 Paris
5-minute bars:: Jul 17 to Sep 25, 2026 · 50 days
1-hour bars:: May 3, 2024 to Sep 25, 2026 · 615 days
Daily bars:: Sep 27, 2016 to Sep 25, 2026 · 2,435 days
Excluded days:: 15 (contract switches, bad data)
Refresh:: every night, Monday to Friday