Accuracy — lstm_s0
validation only · 2.3M
Summary
| horizon | days | token-days | rank IC | t | LS decile | hit (alpha) | hit (price) | MAE (alpha) | MAE (price) |
|---|---|---|---|---|---|---|---|---|---|
| +1d | 265 | 179k | +0.0584 | 14.1 | +0.80% | 55.6% | 54.5% | 0.0356 | 5.781241 |
| +7d | 265 | 178k | +0.0779 | 20.3 | +2.57% | 61.0% | 61.2% | 0.0935 | 16.273686 |
| +30d | 265 | 177k | +0.1032 | 26.3 | +8.24% | 61.9% | 70.2% | 0.2083 | 32.555165 |
By month
| month | IC +1d | hit +1d | LS +1d | IC +7d | hit +7d | LS +7d | IC +30d | hit +30d | LS +30d | days |
|---|---|---|---|---|---|---|---|---|---|---|
| 2026-01 | +0.066 | 56.3% | +0.64% | +0.088 | 63.3% | +3.13% | +0.018 | 47.8% | +5.86% | 28 |
| 2025-12 | +0.044 | 56.3% | +1.11% | +0.078 | 60.5% | +4.21% | +0.092 | 64.5% | +8.10% | 31 |
| 2025-11 | +0.051 | 51.4% | +1.45% | +0.040 | 45.3% | +3.33% | +0.102 | 63.6% | +9.86% | 30 |
| 2025-10 | +0.046 | 59.4% | +0.83% | +0.062 | 66.4% | +1.07% | +0.078 | 44.8% | +14.00% | 31 |
| 2025-09 | +0.050 | 58.9% | +0.28% | +0.058 | 71.3% | +1.00% | +0.085 | 85.0% | +1.85% | 30 |
| 2025-08 | +0.066 | 49.9% | +0.36% | +0.079 | 49.3% | +1.82% | +0.125 | 58.0% | +5.33% | 31 |
| 2025-07 | +0.076 | 54.5% | +0.96% | +0.083 | 54.4% | +3.47% | +0.087 | 51.4% | +7.52% | 31 |
| 2025-06 | +0.050 | 59.7% | +0.46% | +0.082 | 74.5% | +1.28% | +0.125 | 64.1% | +4.96% | 30 |
| 2025-05 | +0.082 | 53.9% | +1.14% | +0.150 | 65.8% | +4.21% | +0.246 | 82.1% | +19.00% | 23 |
t-stats at +7d and +30d are optimistic: consecutive windows overlap, so the effective sample size is closer to the number of distinct days than the number of rows. +1d is the clean read. Long-short spreads are before trading costs and slippage, and the tail of the top-1000 by volume is thin.