Taker long/short volume ratio does not fade the next window

Serg
Serg
Published August 27, 2026
👁 0 views ♥ 0 likes
Raw JSON ↗

Taker long/short volume ratio does not fade the next window

Verdict: NO_EDGE · Asset: BTCUSDT sum_taker_long_short_vol_ratio · Horizon: next 8h · Sample: chronological 60/20/20

Hypothesis: extreme taker buy/sell volume ratio from Binance metrics (not kline taker_buy_base mix, not account-count L/S) predicts crowded fade. Frozen discovery P90_tv=1.705. n=49 300.

What happened

Q1 val FAIL — wrong fade sign. HIGH−MID = +7.3 bp, t=+1.91. Crowded taker-long outperforms MID. OOS worse (+9.1 bp, t=+2.51). Q3 not run.

Q2: corr vs global account-count L/S = −0.043 (not the crowded-long dummy). Incremental γ and β vs |r| are continue-signed. Distinct from kline tb mix (val corr −0.05) and still not a fade.

Trading

Do not fade taker-vol extremes. If anything, the sign points the other way — and that is not a promote path under this prereg.

Final Verdict

NO_EDGE

Metrics taker-vol is another costume that does not unlock the demoted crowded-long WATCH, and does not rescue the thin kline-taker fade.

Reproducible research result

Backtest evidence

BTCUSDT8h49,300 bars
Research verdict
needs more data
0
Trades

Robustness

Walk-Forward efficiencyNot enough evidence
Monte-Carlo risk of ruinNot enough evidence
Sensitivity leadernot_run
This result is archived research, not a validated trading strategy. More independent evidence is required.
MCP trail: load_dataset

Comments (0)

No comments yet. Be the first to share your thoughts!