Rule Collider

Backtests · DJI30 · m5_gap_atr · long

Does m5_gap_atr below 0 work on DJI30? 2023 to 2026 backtest

DJI30 long when Open gap (ATR) is below the threshold, tested at five thresholds taken from the feature’s own distribution on DJI30 over this window: its 10th, 25th, 50th, 75th and 90th percentiles. Window 2023-09-30 to 2026-09-30, the last 30 percent held out of sample; stop 2.5 ATR, target 3.5 R, held overnight, max spread 50 pips, one-minute confirmation 0.2 ATR, no hour filter. Fills and spread as measured on a demo fleet (how a result is judged). The direction and the exit are the ones the research catalogue carried for this pair; the thresholds are not.

Five thresholds, one rule shape

PercentileConditionIn sample R / tradesOut of sample R / tradestTrades to confirmLast quarter R / tradesVerdict
p10m5_gap_atr < -0.05426-336.4 R / 2150-131.3 R / 881-2.4603-46.7 R / 270Active, net negative
p25m5_gap_atr < -0.02265-347.7 R / 2753-193.9 R / 1172-3.2457-64.3 R / 371Active, net negative
p50m5_gap_atr < 0-429.1 R / 2917-215.0 R / 1260-3.4427-85.1 R / 418Active, net negative
p75m5_gap_atr < 0.02246-390.2 R / 2959-209.9 R / 1330-3.2506-78.0 R / 436Active, net negative
p90m5_gap_atr < 0.054-407.7 R / 3018-215.6 R / 1349-3.3493-84.3 R / 452Active, net negative

R is the risk on one trade. The out-of-sample split carries the verdict: below 100 trades it is “no edge shown” whatever the t; a positive mean with t below 2 is “net positive, not yet distinguishable from chance”; the last column is how many trades that mean would need to reach t = 2.

Each threshold in full

m5_gap_atr < -0.05426 (p10)

Honesty panelActive, net negative

Traded 881 times out of sample, net negative beyond chance (-131.3 R, -0.149 R per trade, t = -2.4); still active in the last quarter (-46.7 R over 270 trades).

In sample
2023-09-30 to 2025-11-05
-336.4 R
2150 trades · 27% wins
-0.156 R per trade · t = -3.9, beyond chance
about 560 trades to confirm this mean (-0.156 R per trade) at t = -2 · shown: 2150
spread 0.105 R per trade
Out of sample
2025-11-05 to 2026-09-30
-131.3 R
881 trades · 27% wins
-0.149 R per trade · t = -2.4, beyond chance
about 603 trades to confirm this mean (-0.149 R per trade) at t = -2 · shown: 881
spread 0.082 R per trade
Last quarter
last 91 days of the window
-46.7 R
270 trades · 26% wins
-0.173 R per trade · t = -1.6, within chance
about 439 trades to confirm this mean (-0.173 R per trade) at t = -2 · shown: 270
spread 0.098 R per trade
Costs. Spread 0.098 R per trade at 1.25 times the bar’s minimum. Fill offset 0.14 stop-distances on every entry, measured on 931 demo fills. At 1.5 times the bar’s minimum the out-of-sample figure would read -145.7 R (-0.165 R per trade over 881 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 3,062 signals placed, 3,031 filled, 31 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 554.2 pips, 2.50 ATR); target 3.5 R; exits 71% at the stop, 21% at the target, 8% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (9,093 bars); max spread 50 pips (1,272 bars over it); M1 confirmation 0.2 atr (1,071 signals rejected).

Monthly R, m5_gap_atr < -0.05426 on DJI30One bar per month: the R earned that month; the held-out months in the full colour.2023-10: -12.6 R over 92 trades2023-102023-11: +24.5 R over 86 trades2023-12: +9.3 R over 72 trades2024-01: -30.1 R over 103 trades2024-02: -7.6 R over 90 trades2024-03: -2.9 R over 74 trades2024-04: -36.3 R over 86 trades2024-042024-05: +17.4 R over 89 trades2024-06: -19.5 R over 83 trades2024-07: -11.3 R over 91 trades2024-08: -0.9 R over 88 trades2024-09: -3.0 R over 81 trades2024-10: -48.1 R over 103 trades2024-102024-11: -11.5 R over 84 trades2024-12: -54.5 R over 86 trades2025-01: +8.8 R over 76 trades2025-02: -38.8 R over 82 trades2025-03: -23.7 R over 61 trades2025-04: +0.3 R over 61 trades2025-042025-05: -32.7 R over 92 trades2025-06: +20.1 R over 79 trades2025-07: -29.2 R over 104 trades2025-08: -1.1 R over 91 trades2025-09: -37.9 R over 94 trades2025-10: -12.6 R over 98 trades2025-102025-11: -25.5 R over 81 trades2025-12: -14.8 R over 93 trades2026-01: -16.5 R over 98 trades2026-02: -9.6 R over 87 trades2026-03: -13.1 R over 72 trades2026-04: +6.0 R over 74 trades2026-042026-05: -4.3 R over 41 trades2026-06: -9.2 R over 69 trades2026-07: -23.3 R over 86 trades2026-08: -10.8 R over 97 trades2026-09: -12.6 R over 87 trades+25 R-54 R

m5_gap_atr < -0.02265 (p25)

Honesty panelActive, net negative

Traded 1172 times out of sample, net negative beyond chance (-193.9 R, -0.165 R per trade, t = -3.2); still active in the last quarter (-64.3 R over 371 trades).

In sample
2023-09-30 to 2025-11-05
-347.7 R
2753 trades · 29% wins
-0.126 R per trade · t = -3.7, beyond chance
about 807 trades to confirm this mean (-0.126 R per trade) at t = -2 · shown: 2753
spread 0.083 R per trade
Out of sample
2025-11-05 to 2026-09-30
-193.9 R
1172 trades · 28% wins
-0.165 R per trade · t = -3.2, beyond chance
about 457 trades to confirm this mean (-0.165 R per trade) at t = -2 · shown: 1172
spread 0.069 R per trade
Last quarter
last 91 days of the window
-64.3 R
371 trades · 27% wins
-0.173 R per trade · t = -1.9, within chance
about 417 trades to confirm this mean (-0.173 R per trade) at t = -2 · shown: 371
spread 0.081 R per trade
Costs. Spread 0.079 R per trade at 1.25 times the bar’s minimum. Fill offset 0.14 stop-distances on every entry, measured on 931 demo fills. At 1.5 times the bar’s minimum the out-of-sample figure would read -210.0 R (-0.179 R per trade over 1172 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 3,956 signals placed, 3,925 filled, 31 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 710.3 pips, 2.50 ATR); target 3.5 R; exits 68% at the stop, 19% at the target, 13% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (2,948 bars); max spread 50 pips (1,067 bars over it); M1 confirmation 0.2 atr (1,415 signals rejected).

Monthly R, m5_gap_atr < -0.02265 on DJI30One bar per month: the R earned that month; the held-out months in the full colour.2023-10: -23.9 R over 126 trades2023-102023-11: +26.4 R over 100 trades2023-12: +15.9 R over 87 trades2024-01: -18.1 R over 126 trades2024-02: -9.7 R over 110 trades2024-03: -11.3 R over 72 trades2024-04: -35.0 R over 116 trades2024-042024-05: -4.2 R over 109 trades2024-06: -25.2 R over 109 trades2024-07: -9.1 R over 119 trades2024-08: +17.6 R over 103 trades2024-09: -14.9 R over 112 trades2024-10: -60.3 R over 133 trades2024-102024-11: +4.3 R over 99 trades2024-12: -51.2 R over 121 trades2025-01: +3.5 R over 112 trades2025-02: -34.1 R over 104 trades2025-03: -10.6 R over 77 trades2025-04: -19.4 R over 110 trades2025-042025-05: -29.3 R over 109 trades2025-06: +15.7 R over 101 trades2025-07: -23.0 R over 130 trades2025-08: -15.0 R over 113 trades2025-09: -24.5 R over 120 trades2025-10: -3.0 R over 125 trades2025-102025-11: -25.3 R over 91 trades2025-12: -14.1 R over 116 trades2026-01: -37.8 R over 118 trades2026-02: -9.5 R over 110 trades2026-03: -50.8 R over 113 trades2026-04: +6.0 R over 102 trades2026-042026-05: -8.2 R over 55 trades2026-06: +0.8 R over 106 trades2026-07: -13.4 R over 118 trades2026-08: -13.3 R over 119 trades2026-09: -37.5 R over 134 trades+26 R-60 R

m5_gap_atr < 0 (p50)

Honesty panelActive, net negative

Traded 1260 times out of sample, net negative beyond chance (-215.0 R, -0.171 R per trade, t = -3.4); still active in the last quarter (-85.1 R over 418 trades).

In sample
2023-09-30 to 2025-11-05
-429.1 R
2917 trades · 29% wins
-0.147 R per trade · t = -4.5, beyond chance
about 580 trades to confirm this mean (-0.147 R per trade) at t = -2 · shown: 2917
spread 0.079 R per trade
Out of sample
2025-11-05 to 2026-09-30
-215.0 R
1260 trades · 28% wins
-0.171 R per trade · t = -3.4, beyond chance
about 427 trades to confirm this mean (-0.171 R per trade) at t = -2 · shown: 1260
spread 0.066 R per trade
Last quarter
last 91 days of the window
-85.1 R
418 trades · 26% wins
-0.204 R per trade · t = -2.4, beyond chance
about 293 trades to confirm this mean (-0.204 R per trade) at t = -2 · shown: 418
spread 0.076 R per trade
Costs. Spread 0.075 R per trade at 1.25 times the bar’s minimum. Fill offset 0.14 stop-distances on every entry, measured on 931 demo fills. At 1.5 times the bar’s minimum the out-of-sample figure would read -231.8 R (-0.184 R per trade over 1260 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 4,212 signals placed, 4,177 filled, 35 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 750.4 pips, 2.50 ATR); target 3.5 R; exits 69% at the stop, 18% at the target, 13% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (1,903 bars); max spread 50 pips (1,061 bars over it); M1 confirmation 0.2 atr (1,398 signals rejected).

Monthly R, m5_gap_atr < 0 on DJI30One bar per month: the R earned that month; the held-out months in the full colour.2023-10: -36.5 R over 136 trades2023-102023-11: +23.0 R over 103 trades2023-12: +10.3 R over 95 trades2024-01: -18.3 R over 133 trades2024-02: -9.6 R over 117 trades2024-03: -6.4 R over 75 trades2024-04: -50.5 R over 133 trades2024-042024-05: -5.2 R over 109 trades2024-06: -23.5 R over 121 trades2024-07: -1.9 R over 123 trades2024-08: +8.7 R over 110 trades2024-09: -12.9 R over 123 trades2024-10: -56.0 R over 139 trades2024-102024-11: +3.8 R over 104 trades2024-12: -61.2 R over 122 trades2025-01: +5.7 R over 116 trades2025-02: -40.7 R over 112 trades2025-03: -32.2 R over 82 trades2025-04: -21.6 R over 127 trades2025-042025-05: -32.8 R over 118 trades2025-06: +15.9 R over 111 trades2025-07: -23.0 R over 128 trades2025-08: -19.2 R over 115 trades2025-09: -24.0 R over 124 trades2025-10: -15.9 R over 130 trades2025-102025-11: -24.0 R over 98 trades2025-12: -13.6 R over 121 trades2026-01: -29.7 R over 121 trades2026-02: -12.1 R over 122 trades2026-03: -30.8 R over 93 trades2026-04: +7.6 R over 105 trades2026-042026-05: -12.2 R over 63 trades2026-06: -20.1 R over 130 trades2026-07: -20.6 R over 147 trades2026-08: -15.1 R over 126 trades2026-09: -49.4 R over 145 trades+23 R-61 R

m5_gap_atr < 0.02246 (p75)

Honesty panelActive, net negative

Traded 1330 times out of sample, net negative beyond chance (-209.9 R, -0.158 R per trade, t = -3.2); still active in the last quarter (-78.0 R over 436 trades).

In sample
2023-09-30 to 2025-11-05
-390.2 R
2959 trades · 29% wins
-0.132 R per trade · t = -4.0, beyond chance
about 731 trades to confirm this mean (-0.132 R per trade) at t = -2 · shown: 2959
spread 0.076 R per trade
Out of sample
2025-11-05 to 2026-09-30
-209.9 R
1330 trades · 28% wins
-0.158 R per trade · t = -3.2, beyond chance
about 506 trades to confirm this mean (-0.158 R per trade) at t = -2 · shown: 1330
spread 0.065 R per trade
Last quarter
last 91 days of the window
-78.0 R
436 trades · 28% wins
-0.179 R per trade · t = -2.1, beyond chance
about 390 trades to confirm this mean (-0.179 R per trade) at t = -2 · shown: 436
spread 0.076 R per trade
Costs. Spread 0.073 R per trade at 1.25 times the bar’s minimum. Fill offset 0.14 stop-distances on every entry, measured on 931 demo fills. At 1.5 times the bar’s minimum the out-of-sample figure would read -227.2 R (-0.171 R per trade over 1330 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 4,321 signals placed, 4,289 filled, 32 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 756.9 pips, 2.50 ATR); target 3.5 R; exits 69% at the stop, 18% at the target, 13% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (1,730 bars); max spread 50 pips (1,010 bars over it); M1 confirmation 0.2 atr (1,427 signals rejected).

Monthly R, m5_gap_atr < 0.02246 on DJI30One bar per month: the R earned that month; the held-out months in the full colour.2023-10: -27.3 R over 132 trades2023-102023-11: +33.7 R over 95 trades2023-12: +11.2 R over 96 trades2024-01: -19.6 R over 135 trades2024-02: -15.7 R over 120 trades2024-03: -14.8 R over 89 trades2024-04: -54.4 R over 138 trades2024-042024-05: +4.5 R over 110 trades2024-06: -24.4 R over 117 trades2024-07: -4.2 R over 125 trades2024-08: +9.9 R over 117 trades2024-09: -9.4 R over 123 trades2024-10: -51.5 R over 138 trades2024-102024-11: +2.0 R over 106 trades2024-12: -67.7 R over 118 trades2025-01: +15.5 R over 114 trades2025-02: -33.1 R over 111 trades2025-03: -24.9 R over 77 trades2025-04: -14.9 R over 125 trades2025-042025-05: -28.7 R over 125 trades2025-06: +6.4 R over 117 trades2025-07: -26.9 R over 132 trades2025-08: -16.1 R over 120 trades2025-09: -19.5 R over 132 trades2025-10: -12.6 R over 134 trades2025-102025-11: -19.8 R over 106 trades2025-12: -24.7 R over 134 trades2026-01: -27.9 R over 125 trades2026-02: -16.8 R over 125 trades2026-03: -33.3 R over 102 trades2026-04: +13.3 R over 114 trades2026-042026-05: -14.4 R over 61 trades2026-06: -16.2 R over 140 trades2026-07: -16.3 R over 152 trades2026-08: -21.1 R over 134 trades2026-09: -40.6 R over 150 trades+34 R-68 R

m5_gap_atr < 0.054 (p90)

Honesty panelActive, net negative

Traded 1349 times out of sample, net negative beyond chance (-215.6 R, -0.160 R per trade, t = -3.3); still active in the last quarter (-84.3 R over 452 trades).

In sample
2023-09-30 to 2025-11-05
-407.7 R
3018 trades · 29% wins
-0.135 R per trade · t = -4.2, beyond chance
about 700 trades to confirm this mean (-0.135 R per trade) at t = -2 · shown: 3018
spread 0.076 R per trade
Out of sample
2025-11-05 to 2026-09-30
-215.6 R
1349 trades · 28% wins
-0.160 R per trade · t = -3.3, beyond chance
about 493 trades to confirm this mean (-0.160 R per trade) at t = -2 · shown: 1349
spread 0.065 R per trade
Last quarter
last 91 days of the window
-84.3 R
452 trades · 27% wins
-0.186 R per trade · t = -2.2, beyond chance
about 359 trades to confirm this mean (-0.186 R per trade) at t = -2 · shown: 452
spread 0.076 R per trade
Costs. Spread 0.073 R per trade at 1.25 times the bar’s minimum. Fill offset 0.14 stop-distances on every entry, measured on 931 demo fills. At 1.5 times the bar’s minimum the out-of-sample figure would read -233.2 R (-0.173 R per trade over 1349 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 4,401 signals placed, 4,367 filled, 34 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 752.7 pips, 2.50 ATR); target 3.5 R; exits 69% at the stop, 18% at the target, 13% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (1,685 bars); max spread 50 pips (1,046 bars over it); M1 confirmation 0.2 atr (1,419 signals rejected).

Monthly R, m5_gap_atr < 0.054 on DJI30One bar per month: the R earned that month; the held-out months in the full colour.2023-10: -18.1 R over 134 trades2023-102023-11: +32.9 R over 104 trades2023-12: +13.6 R over 100 trades2024-01: -24.6 R over 140 trades2024-02: -9.7 R over 123 trades2024-03: -19.7 R over 92 trades2024-04: -49.0 R over 133 trades2024-042024-05: -5.1 R over 111 trades2024-06: -24.5 R over 120 trades2024-07: -6.6 R over 123 trades2024-08: +5.5 R over 124 trades2024-09: -10.9 R over 133 trades2024-10: -55.2 R over 142 trades2024-102024-11: +2.4 R over 111 trades2024-12: -70.7 R over 121 trades2025-01: +12.6 R over 118 trades2025-02: -33.7 R over 111 trades2025-03: -25.3 R over 77 trades2025-04: -18.8 R over 126 trades2025-042025-05: -30.6 R over 127 trades2025-06: +9.4 R over 121 trades2025-07: -26.9 R over 133 trades2025-08: -18.4 R over 118 trades2025-09: -15.8 R over 128 trades2025-10: -13.2 R over 135 trades2025-102025-11: -19.0 R over 110 trades2025-12: -23.4 R over 129 trades2026-01: -33.0 R over 126 trades2026-02: -14.5 R over 126 trades2026-03: -35.9 R over 105 trades2026-04: +21.1 R over 113 trades2026-042026-05: -18.6 R over 62 trades2026-06: -15.4 R over 139 trades2026-07: -16.4 R over 156 trades2026-08: -24.2 R over 140 trades2026-09: -43.7 R over 156 trades+33 R-71 R

What m5_gap_atr measures

Open gap (ATR) (atr), on M5 bars. This bar's open minus the previous close, in ATR multiples.

Test your own variant

Rule Collider runs one rule on one symbol over years of broker M5 data with the fills and the spread a demo fleet measured, and shows the out-of-sample figure first. Early access goes to the waitlist in signup order.

Join the waitlist

Historical simulation on broker M5 data with a measured fill offset, a measured spread factor, swap and a one-minute confirmation model. Past results do not predict future results. This is a research tool, not investment advice. Hours are broker time, the chart time in MT5. Built 2026-11-04 from runs ending 2026-09-30; refreshed monthly.