Strategies

One canonical inventory of every strategy TradeRep has tested, across both campaign engines, plus the prospective live-shadow candidates and the pre-registered AV2 items that have not run. Every row carries its mechanism family, market, timeframe, research gate, Phase A/B/C state, metrics, exact stop reason and integrity signals — and links to a record page with its full history. Nothing here re-decides a verdict: stage labels are the campaign runner’s own, quoted.

Dashboard Research Status (card view)Live Audit Research History

Current

Completed

Passed

Failed

Nominal tests

Unique strategies

Effective independent

—

The Strategy Factory artifact tree (research-output/strategy-factory/) is not in this checkout, so A6+ campaigns are not listed below. That is a missing-artifact condition, not a finding: those campaigns are produced by the multi-market research lane and appear here as soon as their artifacts are present.

ViewCurrent 13 Completed 58 All 71

Enginestrategy lab strategy factory pre campaign shadow preregistration only

MarketMNQ

Mechanism familyVolatility expansion / contraction 13 Order flow / aggressor imbalance 11 Unclassified 11 Overnight / prior-day structure 7 Regime-conditioned 6 Breakout / failed breakout 5 Momentum / trend 5 Market structure / FVG / liquidity 5 Session / time-of-day 4 Mean reversion 2 Auction / value profile 2“inf” = classified by the flagged keyword heuristic rather than a curated mapping.

Completed strategies

5 of 5 shown

Strategy	Family	Campaign	Market · TF · side	Phase	Gate	Outcome	Screen n	Screen exp	Val exp	Holdout	Stop reason
A12 — Efficiency Ratio Momentum a1-efficiency-ratio-momentum	Momentum / trend a1-efficiency-ratio-momentum	a1 strategy lab	MNQ MNQ 1-minute OHLC · BOTH	Phase A	FINAL_AUDIT	retired	9,975	+0.044R	+0.032R	+0.130R · n=100	Survived Holdout but the Final Research Audit flagged unresolved risk(s): Longest observed losing streak 23 trades -- confirm this is tolerable before any further stage.; Holdout expectancy diverges materially from Validation's own -- treat as a fresh, not fully explained, open question..
a2-a12-short-efficiency-ratio-momentum a2-a12-short-efficiency-ratio-momentum	Momentum / trend a2-a12-short-efficiency-ratio-momentum	a2 strategy lab	MNQ MNQ 1-minute OHLC · BOTH	Phase A	EXECUTION_STAT	fail	3,690	+0.049R	—	—	Execution+Statistical Gate: Bootstrap 95% CI [-0.0054, 0.0992]R includes zero -- the raw edge is not statistically distinguishable from noise at this sample size.
a2-failed-trend-persistence a2-failed-trend-persistence	Momentum / trend a2-failed-trend-persistence	a2 strategy lab	MNQ MNQ 1-minute OHLC · BOTH	Phase A	LIGHTNING	fail	906	-0.078R	—	—	Lightning Screen: Aggregate expectancy -0.0784R over 906 resolved trades is negative/flat (PF=0.887), with 4/4 qualifying years (n>=15) agree with the aggregate sign.
a3-multi-session-trend-persistence a3-multi-session-trend-persistence	Momentum / trend a3-multi-session-trend-persistence	a3 strategy lab	MNQ MNQ 1-minute OHLC · BOTH	Phase A	LIGHTNING	fail	49	-0.060R	—	—	Lightning Screen: Aggregate expectancy -0.0600R is negative/flat, and the sign is inconsistent across years (0/1 qualifying years (n>=15) agree with the aggregate sign).
a5-14-impulse-pullback-velocity-asymmetry a5-14-impulse-pullback-velocity-asymmetry	Momentum / trend a5-14-impulse-pullback-velocity-asymmetry	a5 strategy lab	MNQ MNQ 1-minute OHLC · BOTH	Phase A	LIGHTNING	fail	3,105	+0.005R	—	—	Lightning Screen: Aggregate expectancy is positive (0.0050R) but the top 10 winning trades alone account for 127.7% of total R -- the apparent edge is fragile.

Read from research-output/strategy-lab/campaigns/*/campaign_state.json · research-output/strategy-factory/campaigns/*/report.json · research-output/registry/mechanism-families.json · lib/liveStatus/strategyRoster.ts

Campaigns

Both engines, newest last

Campaign	Engine	State	Submitted	Effective tests	Holdout opened	Markets
a0	lab	CLOSED	5	—	0	MNQ
a1	lab	CLOSED	10	—	1	MNQ
a2	lab	CLOSED	10	—	1	MNQ
a3	lab	CLOSED	10	—	2	MNQ
a4	lab	CLOSED	5	—	0	MNQ
a5	lab	CLOSED	19	—	0	MNQ

Research rules in force

Roadmap Part 0 — quoted, not restated

✕Merge strategy/research branches into main without explicit approval.
✕Alter approved/frozen V1 strategy semantics without explicit approval.
✕Promote any research result into V1 automatically.
✕Change live READY behavior based on unvalidated research.
✕Change risk sizing in production.
✕Rewrite historical results.
✕Reuse discovery data and later present it as unseen validation.
✕Quietly discard failed/negative experiments.
✕Optimize only for win rate.
✕Invent new strategy mechanics because a test performed poorly.
✕Skip, disable, or hide a failed replication.

Any strategy-affecting proposal stops at CANDIDATE RULE — HUMAN REVIEW REQUIRED. Research may continue around it; the candidate itself is never promoted automatically.

Read from docs/TRADE_RESEARCH_ROADMAP.md

Holdout policy

One-shot, sealed, chronological, never retried

1,079,881 sealed MNQ candles (2019-05-05..2022-06-02), handed out one chronological slice at a time and frozen before evaluation. 4 Holdout(s) spent. 0 strategies are Holdout-eligible and awaiting explicit human authorisation — nothing in this UI can grant it. Full state on the Monitor.

Read from research-output/strategy-lab/sealed-holdout-bank/SEALED_holdout_bank_manifest.json