CopyTrader Intel Assessment — 2026-05-15

Status: Partially Wired | Mixed Performance | High Potential if Fixed

Current State

• copy_trader_intel/ module exists and is referenced in alpha_engine/score_booster.py.
• Scoring contributions (as of latest code):
• copy_trader_intel: +25 boost, 59.6% WR on 47 closed trades (best of the group).
• copy_trader_consensus: Only +5 (was previously +45 — heavily reduced).
• copy_trader_highscore: +10.
• copy_trader_clones: -10 (actively penalizing).
• These are mostly used as scoring adjustments on picks generated by other systems, not as primary high-volume signal generators.

Key Issues

• Volume from CopyTrader sources is low compared to quan_engine, goldmine, and incubator scanners.
• Some CopyTrader categories are net negative and should be disabled or heavily filtered.
• The "smart money" CopyTrader edge is not being turned into standalone high-conviction picks at scale.

90-Day Recommendation

• P0: Disable or zero-out copy_trader_clones and any negative categories.
• P1: Promote copy_trader_intel as a primary signal source for non-crypto (especially EQUITY and COMMODITY) with proper elite filtering.
• P2: Build a dedicated CopyTrader Consensus emitter that surfaces top traders' positions with trust_score weighting.
• Target: Make CopyTrader intel contribute ≥15-20% of high-trust non-crypto emissions by end of 90 days.

Verdict: Good data exists. Execution is incomplete. Fixing the wiring and volume would meaningfully lift win rates in EQUITY and COMMODITY.