Rule Collider

Backtests · GOLD · macd_histogram · long

Does macd_histogram above -0.01282 work on GOLD? 2023 to 2026 backtest

GOLD long when MACD histogram (12, 26, 9) is above the threshold, tested at five thresholds taken from the feature’s own distribution on GOLD 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 R, held overnight, max spread 30 pips, one-minute confirmation 2 pips, 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
p10macd_histogram > -8.396-148.1 R / 3260-134.4 R / 1227-2.4886-54.1 R / 418Active, net negative
p25macd_histogram > -3.179-103.6 R / 3007-148.2 R / 1164-2.7651-67.9 R / 403Active, net negative
p50macd_histogram > -0.01282-147.5 R / 2746-113.4 R / 1086-2.1966-54.3 R / 363Active, net negative
p75macd_histogram > 3.195-17.8 R / 1913-105.1 R / 1020-2.0981-42.1 R / 329Active, net negative
p90macd_histogram > 8.424+10.6 R / 816-66.5 R / 867-1.41,764-18.7 R / 271No edge shown

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

macd_histogram > -8.396 (p10)

Honesty panelActive, net negative

Traded 1227 times out of sample, net negative beyond chance (-134.4 R, -0.110 R per trade, t = -2.4); still active in the last quarter (-54.1 R over 418 trades).

In sample
2023-09-30 to 2025-11-05
-148.1 R
3260 trades · 33% wins
-0.045 R per trade · t = -1.5, within chance
about 5,492 trades to confirm this mean (-0.045 R per trade) at t = -2 · shown: 3260
spread 0.106 R per trade
Out of sample
2025-11-05 to 2026-09-30
-134.4 R
1227 trades · 30% wins
-0.110 R per trade · t = -2.4, beyond chance
about 886 trades to confirm this mean (-0.110 R per trade) at t = -2 · shown: 1227
spread 0.065 R per trade
Last quarter
last 91 days of the window
-54.1 R
418 trades · 29% wins
-0.129 R per trade · t = -1.6, within chance
about 645 trades to confirm this mean (-0.129 R per trade) at t = -2 · shown: 418
spread 0.087 R per trade
Costs. Spread 0.095 R per trade at 1.25 times the bar’s minimum (more than the trades made before spread); 781 fills in measured hot hours (01:00 at 3.9 times). Fill offset 0.08 stop-distances on every entry, measured on 732 demo fills. Note: the symbol's quoting precision changed on 2025-08-04; the spread field is converted with the point of each era. At 1.5 times the bar’s minimum the out-of-sample figure would read -150.4 R (-0.123 R per trade over 1227 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 4,716 signals placed, 4,487 filled, 229 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 87.1 pips, 2.50 ATR, 30 raised to the 10-pip floor); target 3 R; exits 65% at the stop, 22% at the target, 13% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (1,415 bars); max spread 30 pips (1 bars over it); M1 confirmation 2 pips (952 signals rejected).

Monthly R, macd_histogram > -8.396 on GOLDOne bar per month: the R earned that month; the held-out months in the full colour.2023-10: +17.2 R over 138 trades2023-102023-11: -14.1 R over 137 trades2023-12: -0.7 R over 128 trades2024-01: -22.2 R over 165 trades2024-02: -28.3 R over 139 trades2024-03: +19.3 R over 102 trades2024-04: -28.3 R over 140 trades2024-042024-05: -33.4 R over 149 trades2024-06: -28.1 R over 140 trades2024-07: +8.9 R over 146 trades2024-08: -5.0 R over 146 trades2024-09: +6.6 R over 127 trades2024-10: +5.2 R over 143 trades2024-102024-11: -7.2 R over 117 trades2024-12: -29.2 R over 129 trades2025-01: +10.2 R over 125 trades2025-02: -2.0 R over 128 trades2025-03: +16.3 R over 70 trades2025-04: +4.5 R over 112 trades2025-042025-05: -3.3 R over 109 trades2025-06: -22.8 R over 127 trades2025-07: -19.4 R over 142 trades2025-08: -27.2 R over 145 trades2025-09: +35.0 R over 125 trades2025-10: +4.4 R over 122 trades2025-102025-11: +7.0 R over 115 trades2025-12: +8.9 R over 118 trades2026-01: +10.3 R over 103 trades2026-02: +20.7 R over 77 trades2026-03: -30.2 R over 81 trades2026-04: -26.4 R over 126 trades2026-042026-05: -11.3 R over 66 trades2026-06: -64.0 R over 132 trades2026-07: -30.9 R over 138 trades2026-08: +21.3 R over 129 trades2026-09: -44.5 R over 151 trades+35 R-64 R

macd_histogram > -3.179 (p25)

Honesty panelActive, net negative

Traded 1164 times out of sample, net negative beyond chance (-148.2 R, -0.127 R per trade, t = -2.7); still active in the last quarter (-67.9 R over 403 trades).

In sample
2023-09-30 to 2025-11-05
-103.6 R
3007 trades · 34% wins
-0.034 R per trade · t = -1.1, within chance
about 9,534 trades to confirm this mean (-0.034 R per trade) at t = -2 · shown: 3007
spread 0.110 R per trade
Out of sample
2025-11-05 to 2026-09-30
-148.2 R
1164 trades · 30% wins
-0.127 R per trade · t = -2.7, beyond chance
about 651 trades to confirm this mean (-0.127 R per trade) at t = -2 · shown: 1164
spread 0.063 R per trade
Last quarter
last 91 days of the window
-67.9 R
403 trades · 28% wins
-0.169 R per trade · t = -2.1, beyond chance
about 375 trades to confirm this mean (-0.169 R per trade) at t = -2 · shown: 403
spread 0.084 R per trade
Costs. Spread 0.097 R per trade at 1.25 times the bar’s minimum (more than the trades made before spread); 733 fills in measured hot hours (01:00 at 3.9 times). Fill offset 0.08 stop-distances on every entry, measured on 732 demo fills. Note: the symbol's quoting precision changed on 2025-08-04; the spread field is converted with the point of each era. At 1.5 times the bar’s minimum the out-of-sample figure would read -162.9 R (-0.140 R per trade over 1164 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 4,401 signals placed, 4,171 filled, 230 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 88.8 pips, 2.50 ATR, 30 raised to the 10-pip floor); target 3 R; exits 64% at the stop, 22% at the target, 14% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (1,875 bars); max spread 30 pips (2 bars over it); M1 confirmation 2 pips (901 signals rejected).

Monthly R, macd_histogram > -3.179 on GOLDOne bar per month: the R earned that month; the held-out months in the full colour.2023-10: +11.9 R over 128 trades2023-102023-11: -15.5 R over 135 trades2023-12: +12.3 R over 114 trades2024-01: -18.5 R over 149 trades2024-02: -26.7 R over 135 trades2024-03: +9.1 R over 96 trades2024-04: -22.8 R over 123 trades2024-042024-05: -23.9 R over 141 trades2024-06: -21.0 R over 128 trades2024-07: +0.0 R over 145 trades2024-08: -2.2 R over 130 trades2024-09: +6.1 R over 117 trades2024-10: +5.1 R over 126 trades2024-102024-11: -5.7 R over 108 trades2024-12: -14.6 R over 110 trades2025-01: -0.8 R over 121 trades2025-02: +4.8 R over 110 trades2025-03: +15.8 R over 64 trades2025-04: +13.5 R over 103 trades2025-042025-05: -16.8 R over 112 trades2025-06: -20.4 R over 113 trades2025-07: -16.5 R over 135 trades2025-08: -22.3 R over 125 trades2025-09: +28.4 R over 123 trades2025-10: +19.7 R over 109 trades2025-102025-11: +14.7 R over 112 trades2025-12: -3.6 R over 110 trades2026-01: +25.6 R over 88 trades2026-02: +7.4 R over 82 trades2026-03: -35.7 R over 85 trades2026-04: -23.1 R over 103 trades2026-042026-05: -10.7 R over 70 trades2026-06: -57.3 R over 118 trades2026-07: -33.4 R over 134 trades2026-08: +12.8 R over 126 trades2026-09: -47.3 R over 143 trades+28 R-57 R

macd_histogram > -0.01282 (p50)

Honesty panelActive, net negative

Traded 1086 times out of sample, net negative beyond chance (-113.4 R, -0.104 R per trade, t = -2.1); still active in the last quarter (-54.3 R over 363 trades).

In sample
2023-09-30 to 2025-11-05
-147.5 R
2746 trades · 33% wins
-0.054 R per trade · t = -1.7, within chance
about 3,815 trades to confirm this mean (-0.054 R per trade) at t = -2 · shown: 2746
spread 0.095 R per trade
Out of sample
2025-11-05 to 2026-09-30
-113.4 R
1086 trades · 31% wins
-0.104 R per trade · t = -2.1, beyond chance
about 966 trades to confirm this mean (-0.104 R per trade) at t = -2 · shown: 1086
spread 0.061 R per trade
Last quarter
last 91 days of the window
-54.3 R
363 trades · 29% wins
-0.150 R per trade · t = -1.8, within chance
about 473 trades to confirm this mean (-0.150 R per trade) at t = -2 · shown: 363
spread 0.080 R per trade
Costs. Spread 0.086 R per trade at 1.25 times the bar’s minimum (more than the trades made before spread); 571 fills in measured hot hours (01:00 at 3.9 times). Fill offset 0.08 stop-distances on every entry, measured on 732 demo fills. Note: the symbol's quoting precision changed on 2025-08-04; the spread field is converted with the point of each era. At 1.5 times the bar’s minimum the out-of-sample figure would read -126.6 R (-0.117 R per trade over 1086 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 4,071 signals placed, 3,832 filled, 239 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 91.0 pips, 2.50 ATR, 18 raised to the 10-pip floor); target 3 R; exits 64% at the stop, 21% at the target, 15% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (2,108 bars); max spread 30 pips (4 bars over it); M1 confirmation 2 pips (920 signals rejected).

Monthly R, macd_histogram > -0.01282 on GOLDOne bar per month: the R earned that month; the held-out months in the full colour.2023-10: -0.7 R over 120 trades2023-102023-11: -26.7 R over 119 trades2023-12: +13.9 R over 112 trades2024-01: -26.0 R over 125 trades2024-02: -21.8 R over 107 trades2024-03: +7.6 R over 85 trades2024-04: -11.3 R over 111 trades2024-042024-05: -43.6 R over 138 trades2024-06: -14.6 R over 104 trades2024-07: +2.4 R over 131 trades2024-08: -1.8 R over 118 trades2024-09: +7.2 R over 113 trades2024-10: +9.5 R over 116 trades2024-102024-11: -2.8 R over 91 trades2024-12: -20.7 R over 96 trades2025-01: -5.5 R over 115 trades2025-02: -6.7 R over 106 trades2025-03: +16.5 R over 62 trades2025-04: +2.5 R over 105 trades2025-042025-05: -3.4 R over 102 trades2025-06: -18.7 R over 105 trades2025-07: -14.0 R over 121 trades2025-08: -16.9 R over 113 trades2025-09: +25.8 R over 110 trades2025-10: +5.8 R over 113 trades2025-102025-11: +17.5 R over 102 trades2025-12: -0.5 R over 115 trades2026-01: +27.8 R over 80 trades2026-02: +5.7 R over 85 trades2026-03: -32.1 R over 71 trades2026-04: -20.7 R over 98 trades2026-042026-05: -13.8 R over 65 trades2026-06: -46.6 R over 115 trades2026-07: -27.0 R over 121 trades2026-08: -0.2 R over 123 trades2026-09: -27.1 R over 119 trades+28 R-47 R

macd_histogram > 3.195 (p75)

Honesty panelActive, net negative

Traded 1020 times out of sample, net negative beyond chance (-105.1 R, -0.103 R per trade, t = -2.0); still active in the last quarter (-42.1 R over 329 trades).

In sample
2023-09-30 to 2025-11-05
-17.8 R
1913 trades · 34% wins
-0.009 R per trade · t = -0.3, within chance
about 122,228 trades to confirm this mean (-0.009 R per trade) at t = -2 · shown: 1913
spread 0.057 R per trade
Out of sample
2025-11-05 to 2026-09-30
-105.1 R
1020 trades · 31% wins
-0.103 R per trade · t = -2.0, beyond chance
about 981 trades to confirm this mean (-0.103 R per trade) at t = -2 · shown: 1020
spread 0.055 R per trade
Last quarter
last 91 days of the window
-42.1 R
329 trades · 30% wins
-0.128 R per trade · t = -1.4, within chance
about 648 trades to confirm this mean (-0.128 R per trade) at t = -2 · shown: 329
spread 0.074 R per trade
Costs. Spread 0.056 R per trade at 1.25 times the bar’s minimum (more than the trades made before spread); 294 fills in measured hot hours (01:00 at 3.9 times). Fill offset 0.08 stop-distances on every entry, measured on 732 demo fills. Note: the symbol's quoting precision changed on 2025-08-04; the spread field is converted with the point of each era. At 1.5 times the bar’s minimum the out-of-sample figure would read -116.4 R (-0.114 R per trade over 1020 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 3,097 signals placed, 2,933 filled, 164 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 106.1 pips, 2.50 ATR); target 3 R; exits 63% at the stop, 20% at the target, 17% at the day end.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (3,436 bars); max spread 30 pips (4 bars over it); M1 confirmation 2 pips (475 signals rejected).

Monthly R, macd_histogram > 3.195 on GOLDOne bar per month: the R earned that month; the held-out months in the full colour.2023-10: +7.5 R over 59 trades2023-102023-11: +3.8 R over 46 trades2023-12: +1.6 R over 60 trades2024-01: -21.2 R over 69 trades2024-02: -4.0 R over 38 trades2024-03: +23.5 R over 53 trades2024-04: -2.6 R over 80 trades2024-042024-05: -26.9 R over 90 trades2024-06: -5.4 R over 69 trades2024-07: +14.5 R over 73 trades2024-08: -2.4 R over 83 trades2024-09: +15.3 R over 70 trades2024-10: -3.4 R over 78 trades2024-102024-11: +0.1 R over 69 trades2024-12: -24.5 R over 79 trades2025-01: +10.0 R over 74 trades2025-02: -9.3 R over 82 trades2025-03: +10.2 R over 49 trades2025-04: +8.0 R over 101 trades2025-042025-05: -3.1 R over 96 trades2025-06: -14.0 R over 93 trades2025-07: -29.0 R over 116 trades2025-08: +1.1 R over 84 trades2025-09: +37.9 R over 87 trades2025-10: -2.0 R over 110 trades2025-102025-11: +15.8 R over 87 trades2025-12: +6.8 R over 106 trades2026-01: +27.8 R over 80 trades2026-02: +2.5 R over 88 trades2026-03: -37.4 R over 76 trades2026-04: -28.4 R over 100 trades2026-042026-05: -13.7 R over 56 trades2026-06: -39.9 R over 103 trades2026-07: -19.2 R over 114 trades2026-08: +1.5 R over 110 trades2026-09: -24.4 R over 105 trades+38 R-40 R

macd_histogram > 8.424 (p90)

Honesty panelNo edge shown

No edge shown out of sample: -66.5 R over 867 trades is -0.077 R per trade, within what chance produces (t = -1.4); about 1,764 trades would be needed to confirm this mean at t = -2, shown: 867.

In sample
2023-09-30 to 2025-11-05
+10.6 R
816 trades · 37% wins
+0.013 R per trade · t = 0.2, within chance
about 56,806 trades to confirm this mean (+0.013 R per trade) at t = 2 · shown: 816
spread 0.038 R per trade
Out of sample
2025-11-05 to 2026-09-30
-66.5 R
867 trades · 31% wins
-0.077 R per trade · t = -1.4, within chance
about 1,764 trades to confirm this mean (-0.077 R per trade) at t = -2 · shown: 867
spread 0.046 R per trade
Last quarter
last 91 days of the window
-18.7 R
271 trades · 31% wins
-0.069 R per trade · t = -0.7, within chance
about 2,247 trades to confirm this mean (-0.069 R per trade) at t = -2 · shown: 271
spread 0.061 R per trade
Costs. Spread 0.042 R per trade at 1.25 times the bar’s minimum (more than the trades made before spread); 153 fills in measured hot hours (01:00 at 3.9 times). Fill offset 0.08 stop-distances on every entry, measured on 732 demo fills. Note: the symbol's quoting precision changed on 2025-08-04; the spread field is converted with the point of each era. At 1.5 times the bar’s minimum the out-of-sample figure would read -74.5 R (-0.086 R per trade over 867 trades).

Entry: stop order 2 pips past the signal close, cancelled after 4 bars; 1,735 signals placed, 1,683 filled, 52 expired; trigger state, an entry on every bar the conditions hold while flat.

Exit: stop 2.5 ATR (mean 140.1 pips, 2.50 ATR); target 3 R; exits 60% at the stop, 21% at the day end, 19% at the target.

Filters: signals on every weekday; no entries 22:00 to 01:00 broker time (7,198 bars); max spread 30 pips (7 bars over it); M1 confirmation 2 pips (187 signals rejected).

Monthly R, macd_histogram > 8.424 on GOLDOne bar per month: the R earned that month; the held-out months in the full colour.2023-10: -0.9 R over 16 trades2023-102023-11: +0.5 R over 11 trades2023-12: +3.0 R over 12 trades2024-01: -5.5 R over 9 trades2024-02: +3.9 R over 7 trades2024-03: +0.6 R over 18 trades2024-04: -3.5 R over 35 trades2024-042024-05: -13.8 R over 29 trades2024-06: -12.4 R over 21 trades2024-07: -2.0 R over 23 trades2024-08: -6.6 R over 30 trades2024-09: +6.9 R over 26 trades2024-10: -0.9 R over 24 trades2024-102024-11: +1.7 R over 28 trades2024-12: -3.3 R over 30 trades2025-01: -0.9 R over 19 trades2025-02: +6.4 R over 38 trades2025-03: -0.6 R over 33 trades2025-04: +6.0 R over 72 trades2025-042025-05: +11.7 R over 61 trades2025-06: -12.3 R over 59 trades2025-07: +8.9 R over 35 trades2025-08: +0.1 R over 32 trades2025-09: +25.8 R over 58 trades2025-10: +0.2 R over 86 trades2025-102025-11: +14.4 R over 69 trades2025-12: -1.1 R over 84 trades2026-01: +24.8 R over 73 trades2026-02: +9.3 R over 78 trades2026-03: -34.8 R over 73 trades2026-04: -15.7 R over 82 trades2026-042026-05: -13.1 R over 52 trades2026-06: -33.9 R over 89 trades2026-07: -0.0 R over 89 trades2026-08: +0.1 R over 100 trades2026-09: -18.7 R over 82 trades+26 R-35 R

What macd_histogram measures

MACD histogram (12, 26, 9) (pips), on M5 bars. MACD line minus its 9-bar signal line, in pips.

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.