Per-year gross Sharpe (M momentum vs R reversal)
| Year | M momentum | R reversal | bars |
|---|---|---|---|
| 2020 | -3.79 | +3.79 | 1771 |
| 2021 | -3.11 | +3.17 | 1764 |
| 2022 | -1.68 | +2.15 | 1757 |
| 2023 | -1.50 | +1.63 | 1750 |
| 2024 | -3.44 | +3.44 | 1764 |
| 2025 | -1.69 | +1.88 | 1750 |
| 2026 | -1.53 | +0.91 | 1309 |
Significance (bootstrap p that gross mean <= 0)
| Series | p (full) | p (OOS) | OOS Sharpe |
|---|---|---|---|
| M momentum | 1.000 | 1.000 | -2.23 |
| R reversal | 0.000 | 0.000 | 2.10 |
Turnover-reduction variants (reversal), gross + net
| Variant | turn/bar | gross SR | b/e bps | OOS SR @0.5bp | @1bp |
|---|---|---|---|---|---|
| R K=20 | 2.78 | 2.10 | 0.75 | 0.64 | -0.81 |
| R K=30 | 2.50 | 1.75 | 0.72 | 0.12 | -1.51 |
| R hysteresis .15/.30 | 2.54 | 1.91 | 0.78 | n/a | n/a |
Read: if the sign flips year to year and the reversal's break-even stays sub-spread, there is no stable hourly edge.