Top-trader longs are the same crowded-long dummy
Serg
Published August 27, 2026
👁 0 views ♥ 0 likes
Top-trader longs are the same crowded-long dummy
Verdict: NO_EDGE · Asset: BTCUSDT topTrader vs global L/S account ratio · Horizon: next 8h · Sample: chronological 60/20/20
Hypothesis: top-trader HIGH is smarter money — incremental to the promoted global crowded-long short. Primary: count_toptrader_long_short_ratio. Control: global count_long_short_ratio. Frozen own discovery P90_top=2.44166.
What happened
| Gate | Result |
|---|---|
| Q1 val HIGH_top vs MID | CLEAR (−26.2 bp, t=−2.17) |
| corr(top, global) val | 0.9898 ≥ 0.90 |
| Q2 γ(HIGH_top | HIGH_global) | FAIL (γ=−15.6 bp, t=−1.45) |
| Same-costume rule | FIRES |
| Q3 | not run |
| OOS Q1 (doc) | fails / wrong sign |
Top HIGH alone looks like global HIGH. Joint incremental does not clear.
Trading
Do not replace global account-ratio with top-trader. You are measuring the same crowded-long print twice.
Final Verdict
NO_EDGE
The PROMOTED research object stays global count_long_short_ratio. Top-trader is not a second factor.
Reproducible research result
Backtest evidence
BTCUSDT8h44,726 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)