TP/SL VS TIMEOUT-ONLY
Est-ce que les barrières TP/SL aident vraiment ?
12 configs comparées: mêmes signaux, mêmes caps/consensus, sortie par TP/SL/timeout ou sortie uniquement au timeout H. Frais = 0.
Best robust40.12
Val/Test56.16/57.22
Test gain/DD104.8% / -31.7%
Avg/trade0.050%
Timeout rate73.5%
Classement
| Config | Seuil | Val | Test | Robust | Gain | DD | Avg/trade | Trades | Open avg | TP | SL | Timeout | Recent score | Recent gain | Recent DD |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1h H=42 k=2.25 ranked_cap5_barrier barrier · ranked cap=5 agree=1 | p70 | 56.16 | 57.22 | 40.12 | 104.77% | -31.70% | 0.050% | 1538 | 4.65 | 14.1% | 12.4% | 73.5% | -61.18 | -21.67% | -26.34% |
| 1h H=42 k=2.25 ranked_cap10_barrier barrier · ranked cap=10 agree=1 | p60 | 52.01 | 46.80 | 31.46 | 92.98% | -30.78% | 0.022% | 3104 | 9.45 | 14.2% | 11.9% | 73.9% | -65.63 | -21.13% | -29.67% |
| 1h H=36 k=2.25 cons3_cap3_barrier barrier · consensus cap=3 agree=3 | p60 | 180.08 | 33.18 | 23.64 | 75.91% | -28.49% | 0.070% | 912 | 2.42 | 12.4% | 10.1% | 77.5% | -53.95 | -16.82% | -24.76% |
| 1h H=36 k=2.25 cons3_cap3_timeout timeout · consensus cap=3 agree=3 | p60 | 19.93 | 22.48 | 11.41 | 59.53% | -24.70% | 0.065% | 832 | 2.47 | 0.0% | 0.0% | 100.0% | -47.21 | -14.57% | -21.76% |
| 1h H=42 k=2.25 cons4_cap3_barrier barrier · consensus cap=3 agree=4 | p60 | 64.39 | 14.49 | 5.83 | 46.91% | -21.61% | 0.064% | 700 | 2.17 | 13.1% | 11.0% | 75.9% | -32.95 | -8.94% | -16.01% |
| 1h H=48 k=2.0 cons5_cap3_barrier barrier · consensus cap=3 agree=5 | p80 | -8.21 | 36.84 | -10.70 | 62.97% | -17.42% | 0.152% | 347 | 1.15 | 17.9% | 14.4% | 67.7% | -14.41 | -4.51% | -6.61% |
| 1h H=48 k=2.25 cons5_cap3_barrier barrier · consensus cap=3 agree=5 | p80 | -7.01 | 73.29 | -10.79 | 95.84% | -15.03% | 0.209% | 340 | 1.17 | 17.4% | 10.6% | 72.1% | -8.25 | -2.01% | -4.16% |
| 1h H=48 k=2.25 cons5_cap3_timeout timeout · consensus cap=3 agree=5 | p70 | -16.04 | 36.52 | -20.78 | 65.34% | -19.21% | 0.129% | 426 | 1.69 | 0.0% | 0.0% | 100.0% | -22.93 | -7.90% | -10.02% |
| 1h H=42 k=2.25 ranked_cap5_timeout timeout · ranked cap=5 agree=1 | p70 | -24.32 | 90.83 | -33.36 | 118.75% | -18.62% | 0.061% | 1361 | 4.71 | 0.0% | 0.0% | 100.0% | -20.01 | 6.81% | -17.88% |
| 1h H=48 k=2.0 cons5_cap3_timeout timeout · consensus cap=3 agree=5 | p80 | -28.92 | 26.10 | -34.58 | 51.24% | -16.76% | 0.145% | 312 | 1.24 | 0.0% | 0.0% | 100.0% | -14.83 | -4.67% | -6.77% |
| 1h H=42 k=2.25 ranked_cap10_timeout timeout · ranked cap=10 agree=1 | p60 | -38.54 | 57.53 | -51.36 | 89.15% | -21.08% | 0.024% | 2773 | 9.60 | 0.0% | 0.0% | 100.0% | -42.38 | -7.79% | -23.06% |
| 1h H=42 k=2.25 cons4_cap3_timeout timeout · consensus cap=3 agree=4 | p60 | -42.19 | -1.57 | -59.78 | 45.81% | -31.59% | 0.070% | 639 | 2.22 | 0.0% | 0.0% | 100.0% | -27.19 | -6.13% | -14.04% |
Timeout-only ignore les high/low TP/SL pendant le trade et sort au close cumulé à H. Les labels/signaux restent ceux du modèle triple-barrier original, donc ce test isole surtout la mécanique de sortie.
