The Forex Factory "Monthly Profit - Daily Trading using SMA" rule on EUR/USD daily candles (spec sma5_spec.md). Run 2026-10-29T09:27:26.492Z; 500 control runs per variant. Results in R (fill to the initial stop) after 1 pip a round trip. data: 8326700 minutes 2004-01-01 00:03 .. 2026-09-28 15:46 New York; 5908 trading days (17:00 to 17:00 New York; 6 with fewer than 600 minutes) first failed condition at each day close: warm-up 19, signal 824, open not across the SMA 2003, candle colour 1778, flat SMA 814, band touched 216, not clearly crossed 254; signals long 437, short 387 === SMA-1D: closed at the end of the entry day (the first position) === signals: traded 822, fill beyond the stop 2 DEV n= 358 net -9.8 R avg -0.027 R PF 0.89 t -0.79 VAL n= 244 net +3.5 R avg +0.014 R PF 1.06 t 0.35 HOLDOUT n= 220 net +0.2 R avg +0.001 R PF 1.00 t 0.02 ALL n=822 net -6.2 R avg -0.007 R a trade (gross +0.003 R, cost 0.010 R); net pips -1538.30; PF 0.97 t -0.34; drawdown on closed trades -25.1 R winners 46.6% (95% 43.2-50.0%); ended by day close 736 (89.5%), stop 86 (10.5%) risk median 106.90 pips; days held median 0.00; the 10 best trades +24.5 R, without them -30.7 R stop: normal candle n=816 avg -0.006 R | short candle (two lows) n=6 avg -0.245 R longs n=435 avg -0.018 R net -7.9 R | shorts n=387 avg +0.005 R net +1.8 R cost scenarios: 1 pip -0.007 R | 1.5 pips -0.013 R | 2.5 pips -0.023 R; financing A (2% a year) -0.014 R, B (today's swap) -0.007 R; trades held overnight 0 descriptive, without the Bollinger condition: n=1038 avg -0.001 R by year (n, net R): 2004 46 -6.7 R | 2005 40 -2.0 R | 2006 35 -5.9 R | 2007 48 -6.8 R | 2008 41 +0.4 R | 2009 45 +2.9 R | 2010 33 +12.1 R | 2011 37 -2.2 R | 2012 33 -1.6 R | 2013 42 -3.6 R | 2014 36 -2.0 R | 2015 26 -0.6 R | 2016 35 +3.4 R | 2017 38 -1.0 R | 2018 33 +5.2 R | 2019 34 +2.0 R | 2020 30 -1.0 R | 2021 35 -2.1 R | 2022 29 +2.7 R | 2023 39 +1.2 R | 2024 33 -5.6 R | 2025 26 +4.9 R | 2026 28 +0.2 R control (trades moved by whole trading days, 500 runs, 4 s): mean trades 822.00, median avg -0.007 R, 5-95% -0.050 R .. +0.036 R; real avg -0.007 R at percentile 48.20 [FAIL] 1 PF >= 1.2 in validation: VAL PF 1.06 [PASS] 2 >= 50 trades in each period: 358 / 244 / 220 [FAIL] 3 t >= 2.5 on development: DEV t -0.79 [FAIL] 4 above 95th pct of the control: percentile 48.20 [FAIL] 5 positive without top-10 trades: -30.7 R [PASS] 6 holdout net positive (flagged used): HOLDOUT net +0.2 R PF 1.00 VERDICT Buried (2/6 gates) rule 3: 13 trades open across a known gap; without them avg -0.009 R, gates 1/6 -> Buried (verdict unchanged) === SMA-3R: target 3R, stop to breakeven after +1R, opposite signal reverses === signals: traded 593, same direction (ignored) 230, fill beyond the stop 1 DEV n= 253 net -1.6 R avg -0.006 R PF 0.99 t -0.07 VAL n= 180 net +0.8 R avg +0.005 R PF 1.01 t 0.05 HOLDOUT n= 160 net -23.7 R avg -0.148 R PF 0.71 t -1.64 ALL n=593 net -24.4 R avg -0.041 R a trade (gross -0.031 R, cost 0.010 R); net pips -4293.50; PF 0.92 t -0.79; drawdown on closed trades -41.7 R winners 25.0% (95% 21.6-28.6%); ended by stop 224 (37.8%), reverse 185 (31.2%), breakeven 116 (19.6%), target 68 (11.5%) risk median 109.00 pips; days held median 4.00; the 10 best trades +32.1 R, without them -56.5 R stop: normal candle n=591 avg -0.045 R | short candle (two lows) n=2 avg +0.982 R longs n=310 avg -0.010 R net -3.2 R | shorts n=283 avg -0.075 R net -21.3 R cost scenarios: 1 pip -0.041 R | 1.5 pips -0.046 R | 2.5 pips -0.056 R; financing A (2% a year) -0.087 R, B (today's swap) -0.064 R; trades held overnight 531 descriptive, without the Bollinger condition: n=645 avg -0.015 R by year (n, net R): 2004 31 -4.3 R | 2005 29 -6.6 R | 2006 28 -8.1 R | 2007 34 -8.4 R | 2008 30 +11.1 R | 2009 28 +5.0 R | 2010 19 +6.7 R | 2011 29 -1.4 R | 2012 25 +4.4 R | 2013 29 +3.8 R | 2014 24 +2.6 R | 2015 23 +3.6 R | 2016 28 -4.2 R | 2017 26 -0.5 R | 2018 24 +0.8 R | 2019 26 -5.3 R | 2020 23 +4.6 R | 2021 21 +4.8 R | 2022 24 -4.5 R | 2023 26 -8.9 R | 2024 27 -12.2 R | 2025 19 -3.4 R | 2026 20 -4.1 R control (trades moved by whole trading days, 500 runs, 21 s): mean trades 412.49, median avg -0.011 R, 5-95% -0.114 R .. +0.090 R; real avg -0.041 R at percentile 31.80 [FAIL] 1 PF >= 1.2 in validation: VAL PF 1.01 [PASS] 2 >= 50 trades in each period: 253 / 180 / 160 [FAIL] 3 t >= 2.5 on development: DEV t -0.07 [FAIL] 4 above 95th pct of the control: percentile 31.80 [FAIL] 5 positive without top-10 trades: -56.5 R [FAIL] 6 holdout net positive (flagged used): HOLDOUT net -23.7 R PF 0.71 VERDICT Buried (1/6 gates) rule 3: 33 trades open across a known gap; without them avg -0.061 R, gates 1/6 -> Buried (verdict unchanged) === SMA-TRAIL: stop trailed to the previous day low - 15 pips, exit when a day touches the opposite band, opposite signal reverses === signals: traded 742, same direction (ignored) 81, fill beyond the stop 1 DEV n= 328 net -20.3 R avg -0.062 R PF 0.83 t -1.36 VAL n= 215 net +4.7 R avg +0.022 R PF 1.06 t 0.36 HOLDOUT n= 199 net -14.1 R avg -0.071 R PF 0.80 t -1.16 ALL n=742 net -29.6 R avg -0.040 R a trade (gross -0.030 R, cost 0.010 R); net pips -4109.80; PF 0.89 t -1.27; drawdown on closed trades -34.1 R winners 37.2% (95% 33.8-40.7%); ended by trailing stop 411 (55.4%), band 222 (29.9%), stop 84 (11.3%), reverse 25 (3.4%) risk median 107.90 pips; days held median 1.00; the 10 best trades +34.0 R, without them -63.6 R stop: normal candle n=737 avg -0.043 R | short candle (two lows) n=5 avg +0.353 R longs n=389 avg -0.052 R net -20.1 R | shorts n=353 avg -0.027 R net -9.5 R cost scenarios: 1 pip -0.040 R | 1.5 pips -0.045 R | 2.5 pips -0.055 R; financing A (2% a year) -0.058 R, B (today's swap) -0.047 R; trades held overnight 568 descriptive, without the Bollinger condition: n=957 avg -0.026 R by year (n, net R): 2004 44 -10.2 R | 2005 38 -5.3 R | 2006 34 -5.3 R | 2007 44 -11.2 R | 2008 37 +1.9 R | 2009 37 +4.4 R | 2010 29 +13.1 R | 2011 33 -1.6 R | 2012 32 -6.1 R | 2013 37 +1.7 R | 2014 34 -4.7 R | 2015 23 -0.5 R | 2016 32 +2.9 R | 2017 33 -2.9 R | 2018 28 +5.3 R | 2019 28 +2.9 R | 2020 27 +1.5 R | 2021 32 -2.6 R | 2022 26 -0.4 R | 2023 31 -0.0 R | 2024 32 -12.4 R | 2025 25 +1.1 R | 2026 26 -1.2 R control (trades moved by whole trading days, 500 runs, 9 s): mean trades 649.92, median avg -0.025 R, 5-95% -0.085 R .. +0.040 R; real avg -0.040 R at percentile 34.60 [FAIL] 1 PF >= 1.2 in validation: VAL PF 1.06 [PASS] 2 >= 50 trades in each period: 328 / 215 / 199 [FAIL] 3 t >= 2.5 on development: DEV t -1.36 [FAIL] 4 above 95th pct of the control: percentile 34.60 [FAIL] 5 positive without top-10 trades: -63.6 R [FAIL] 6 holdout net positive (flagged used): HOLDOUT net -14.1 R PF 0.80 VERDICT Buried (1/6 gates) rule 3: 21 trades open across a known gap; without them avg -0.047 R, gates 1/6 -> Buried (verdict unchanged)