OpenCode/Beam Paper Trading Monitor

Primary candidate — luxalgo_confluence SHORT (CRYPTO)

CANDIDATE (downgraded from PROVEN after independence/cost audit)

Metric	Value
Symbols	XRPUSDT, ETHUSDT, BTCUSDT, STXUSDT
n (hard-resolved)	404
Win rate (aggregate)	56.9%
Profit factor (aggregate)	1.57
Bootstrap P(Expectancy>0)	1.0000
Bootstrap 95% CI	[+0.283, +0.880]%/trade
Strict gate (INCEPTION)	4 cells pass (non-artifact, test n≥10, PF≥1.15, WR≥0.5, train PF>1)
One-position-at-a-time	n=104, PF 1.29 gross → 1.20 (0.10% cost) → 1.13 (0.20% cost)
Monthly consistency	Jul WR 0.47, Aug avg -0.29 (inconsistent)

Certificate: winner_certificate_luxalgo_confluence_short_portfolio.json

Per-asset-class best candidates (strict non-artifact gate + bootstrap P>0≥0.95)

Asset class	Strategy	Symbol	n	Train PF	Test PF	P(>0)	Status
CRYPTO	luxalgo_confluence SHORT	XRP/ETH/BTC/STX	404	1.30-2.39	1.08-1.47	1.0000	Candidate
CRYPTO	regime_accumulation	USDCAD=X	204	3.88	2.74	1.0000	Candidate
CRYPTO	prediction_market_consensus	BNBUSDT	59	1.94	1.97	0.9954	Candidate
FOREX	cta_fx_multifactor	USDJPY=X	75	1.96	1.79	0.9930	Candidate
CRYPTO	high_trust_momentum	BTCUSDT	64	1.66	1.99	0.9730	Candidate
CRYPTO	luxalgo_confluence	ETHUSDT	132	1.48	1.47	0.9728	Candidate
CRYPTO	regime_accumulation	AAPL	76	1.90	5.93	1.0000	Suspicious (test PF jump)
FUTURES	commodity_term_cot	HG=F	815	2.21	2.41	1.0000	Artifact-flagged (751/815 overlap)

Devil's advocate audit (truth check)

• No strategy is PROVEN. All candidates fail at least one strict test (independence, cost, or monthly consistency).
• luxalgo_confluence SHORT: aggregate P=1.0 but one-position-at-a-time PF drops to 1.13 after 0.20% cost; monthly inconsistent.
• commodity_term_cot HG=F: bootstrap P=1.0 but 751/815 signals overlap earlier positions; peak 53 simultaneous LONGs; one-position-at-a-time PF 1.21 → -0.0277%/trade after cost. Downgraded.
• smart_money: 72.1% WR / PF 8.32 — suspiciously high; likely soft resolution.
• multi_asset_copytrader: 88.8% WR / PF 8.02 — likely soft/zero-loss resolution.
• Predictions: PF 4.37 inflated by single +370,850% outlier.
• Data quality: 635 artifact cells (frozen-exit), 9.4% null pnl in CRYPTO, 26.8% zero pnl, only AlphaEngine + mega_mutation are live pipelines.
• Consensus: only CRYPTO and EQUITY have resolved consensus picks; FOREX/COMMODITY have 1 each.

Verdict: No true winner yet. Best candidate is luxalgo_confluence SHORT but requires 100+ forward TP/SL outcomes and independence verification before promotion.

Paper trading status

Status JSON: opencode_paper_trading_status.json

Update job: python3 tools/opencode_paper_trading_update.py (run hourly via cron/GHA)

Validation harness: python3 tools/validate_winner.py --strategy <name> --source AlphaEngine [--symbol SYM]

NFA — research only. Not financial advice.