Accuracy — lstm_s6
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 | +0.0644 | 13.2 | +0.82% | 55.9% | 54.5% | 0.0356 | 5.802690 |
| +7d | 265 | 178k | +0.0728 | 14.7 | +1.68% | 61.3% | 61.5% | 0.0935 | 16.264935 |
| +30d | 265 | 177k | +0.0854 | 18.3 | +5.86% | 61.6% | 69.7% | 0.2088 | 32.341035 |
By month
| month | IC +1d | hit +1d | LS +1d | IC +7d | hit +7d | LS +7d | IC +30d | hit +30d | LS +30d | days |
|---|---|---|---|---|---|---|---|---|---|---|
| 2026-01 | +0.061 | 56.0% | +0.63% | +0.065 | 62.7% | +0.36% | +0.013 | 47.2% | +3.23% | 28 |
| 2025-12 | +0.044 | 56.2% | +1.07% | +0.072 | 59.9% | +3.66% | +0.075 | 64.0% | +3.63% | 31 |
| 2025-11 | +0.066 | 51.3% | +1.71% | +0.056 | 45.2% | +3.09% | +0.067 | 62.9% | +6.39% | 30 |
| 2025-10 | +0.068 | 59.9% | +0.77% | +0.061 | 65.5% | -0.56% | +0.114 | 45.3% | +10.74% | 31 |
| 2025-09 | +0.053 | 58.8% | +0.19% | +0.038 | 71.2% | -0.10% | +0.044 | 84.8% | -3.91% | 30 |
| 2025-08 | +0.065 | 49.8% | +0.50% | +0.051 | 49.3% | +0.99% | +0.057 | 58.0% | +4.46% | 31 |
| 2025-07 | +0.067 | 53.6% | +0.79% | +0.074 | 53.9% | +2.67% | +0.058 | 51.0% | +4.20% | 31 |
| 2025-06 | +0.065 | 59.7% | +0.56% | +0.077 | 73.7% | +0.68% | +0.114 | 63.9% | +5.36% | 30 |
| 2025-05 | +0.099 | 58.3% | +1.20% | +0.187 | 73.8% | +5.04% | +0.265 | 81.8% | +22.28% | 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.