Research Watchlist OBSERVATIONAL

Observational watchlist. Forward performance starts at inception (q=10); historical research evidence is shown separately. Logging-only; no real orders. Generated 2026-07-09T12:24:25Z. Full performance breakdown →
CodeNameAssetAdded (UTC)Fwd tradesFwd net PnLResearch test PnLBaseline testΔnull pSample daysStatus
FTE-N5H6FTE-14 NO Near-Strike Hour GateBTC2026-07-01T20:45:00Z400-210.077102.6017-1.7583104.360.03527.37WATCHLIST
EME-N25P40ETH Near-Moneyness Low-pYES GateETH2026-07-01T20:45:00Z143-39.933778.8182-148.7548227.570.09.01WATCHLIST
Research evidence is historical and separate from forward performance. Forward PnL starts at zero at inception.

FTE-N5H6 — FTE-14 NO Near-Strike Hour Gate WATCHLIST

Added 2026-07-01T20:45:00Z · run run_20260701T230600Z_07cc273_q10grid · asset BTC

Forward since inception (q=10): 400 trades, net -210.077 (period 2026-07-01T20:45:33.737+00:00 → 2026-07-09T12:15:44.033+00:00)

Research evidence (historical, separate from forward performance): locked-test PnL 102.6017, same-universe baseline -1.7583, null p=0.035, stability passed, sample 27.37 days.

Promotion status: watchlist only (forward sample below 30 days). Promotion review requires a longer live window and passing robustness checks. Research evidence is historical and separate from forward performance.

Latest verification (run run_20260702T042653Z_1f1776c_weekly, 27.76d window): locked-test q10 108.7017, candidate-level p=0.035, stat gates cleared.
Robustness: under review · Search-adjusted check: not yet passed (search-adjusted p=0.455) · Promotion status: watchlist only

EME-N25P40 — ETH Near-Moneyness Low-pYES Gate WATCHLIST

Added 2026-07-01T20:45:00Z · run run_20260701T230600Z_07cc273_q10grid · asset ETH

Forward since inception (q=10): 143 trades, net -39.9337 (period 2026-07-01T22:04:32.637443Z → 2026-07-06T17:34:32.435801Z)

Research evidence (historical, separate from forward performance): locked-test PnL 78.8182, same-universe baseline -148.7548, null p=0.0, stability passed, sample 9.01 days.

Promotion status: watchlist only (forward sample below 30 days). Promotion review requires a longer live window and passing robustness checks. Research evidence is historical and separate from forward performance.

Latest verification (run run_20260702T042653Z_1f1776c_weekly, 9.9d window): locked-test q10 30.5473, candidate-level p=0.0, stat gates not re-cleared.
Robustness: under review · Search-adjusted check: not yet run · Promotion status: watchlist only

Research Watchlist — ML OBSERVATIONAL

Model-score-gated candidates from the ML discovery grid. Forward evidence starts at dashboard activation and is never backfilled from historical research evidence, shown separately below. Near-duplicate threshold variants are collapsed into one public watchlist entry per distinct tradeable idea. Not an execution candidate. Paper-only; real orders disabled.
CodeNameAssetAdded (UTC)Fwd tradesFwd net PnLTracking statusStatus
XGB-GATE-20XGB Risk Gate — q20BTC2026-07-07T17:37:28Z135-27.845ACTIVEWATCHLIST

XGB-GATE-20 — XGB Risk Gate — q20 WATCHLIST

Added 2026-07-07T17:37:28Z · run phase2_20260707T045123Z · asset BTC · Top 20% model-score filter / risk-gate candidate

XGB-GATE-20 is an observational ML risk-gate candidate selected from the Phase 2 feature-family grid. Historical research evidence is shown separately; forward watchlist PnL starts at dashboard activation. Paper-only; real orders disabled.

Forward since inception: 135 trades, net -27.845

Forward tracking: Feature-lineage review passed. ML forward scoring is active; forward PnL starts at dashboard activation and is not backfilled from historical research tests.

Research evidence (historical, source PR #18 Phase 2 grid, separate from forward performance): locked-test PnL 315.55, same-universe baseline -125.1, Δ 440.65, search-adjusted p=0.01, stability passed_ex_best_day_and_market, 501 trades across 373 markets.

Promotion status: Research Watchlist only (execution review requires a separate human approval process). Execution review would require a separate human approval process, longer forward evidence, risk-limit review, and no change to real_orders_enabled=false. Research evidence is historical and separate from forward performance.