Trading-strategy family
HOUR_SEASON_V1Time-of-day seasonalityFAIL
tested Oct 24, 2026 2 variants tried
What was tested
Find the 4-hour block of the UTC day that has the best average BTC return in the IS period, then trade that block every day (buy at the start, sell at the end). A second variant also shorts the worst block.
Kraken PF_XBTUSD hourly candles, same costs and R convention as TSMOM_D_V1. IS Oct 3, 2022 to Apr 1, 2025, OOS Apr 1, 2025 to Oct 23, 2026.
Pre-registered pass bar and result
The bar was written and hashed before the run. Results are on data the selection step never saw (out-of-sample).
| Criterion | Bar | Result | Verdict |
|---|---|---|---|
| Fee gate: IS gross captured move ≥ 2× the cost wall | ≥ 17.0 bp | 6.2 bp | FAIL |
| OOS mean R per trade (net) | > 0 | −0.058 | FAIL |
| OOS trades | ≥ 30 | 570 | PASS |
| OOS mean R above random-direction 95th percentile | > −0.034 | −0.058 | FAIL |
| IS mean R > 0 and gate passed | > 0 | −0.036 | FAIL |
| Monte Carlo pass probability ≥ MF4 baseline + 5 pp | ≥ 30.6% | 3.5% | FAIL |
Why it failed
The time-of-day drift was tiny: about 6 bp per 4-hour block in-sample, below the 8 bp round-trip commission alone. Out of sample it shrank to 0.8 bp. The strategy lost money in-sample (−0.036R, 910 trades) and out of sample (−0.058R, 570 trades).
Variants tried
2 variants: long the best block only (HS_L, long 21:00–01:00 UTC, selected on IS), or long best and short worst (HS_LS). IS / OOS (OOS information only): HS_L −0.036 / −0.058, HS_LS −0.045 / −0.049.
What this test cannot tell
Process slip, disclosed: the out-of-sample numbers were computed although the fee gate had already failed. They are information only.
Rules lock
Rules file SHA-256:
4657ac56e6f5c8eacec015fed9ff07d9f85ba84d06527695ed8d897c9a9c7b39
Locked:
“not recorded” means the number or time was not written down at the time, and we have not filled it in afterwards.
Submit a rule
Have a rule with numbers? Send it to us on X @ProjectAnchorHQ or Telegram t.me/ProjectAnchorHQ (market, entry, exit, stop). Free; we only need the written rule, never keys, accounts or funds. We publish the result pass or fail, and credit you only if you say yes.