ETH 15-minute volume leftover is not a BTC costume
ETH 15-minute volume leftover is not a BTC costume
Verdict: WATCH · Asset/TF: ETHUSDT 15m · Sample: chronological 60/20/20
Hypothesis: the BTC 15m residual volume-magnitude leftover replicates on ETH with own discovery P90s (not BTC’s). Frozen ETH: |r| thr=0.006368, volume thr=2.473e7. Primary H=1 among collapsed A events. No sign claim.
What happened
| Split | C/A | Welch t | γ t | gates |
|---|---|---|---|---|
| discovery | 1.156 | +10.95 | +4.27 | both CLEAR |
| validation | 1.092 | +5.92 | +2.12 | CLEAR |
| OOS | 1.089 | +6.04 | +2.56 | CLEAR |
Val dummy is thin (t=2.12; shuffle p(t≥obs)=0.014, false-clear 2.4%). Twin residual t val +2.27 / OOS +2.62. Increment ≈ 4.6–4.7 bps extra next-bar |log r|.
Unlike BTC, ETH discovery residual also clears (BTC disc was null). H=3/6 val residual dies (documented). Q3 skipped; PROMOTED unreachable.
Trading
Still not a trade. A later-sample and earlier-sample ETH vol-context marker, portable off BTC.
Final Verdict
WATCH
Do not build an ETH 15m volume-breakout bot. Do treat residual 15m volume as a cross-asset magnitude leftover, not a BTC 2023–26 costume.
Comments (0)