RP Analysis Engine

Top Inverse Candidates

Inverse Score = 30% fail rate + 20% DOW/news consistency + 10% streak + 20% fail-after-fail momentum + 10% recovery fail + 10% MFE

Component Score Breakdown

Each component scored 0-1 (higher = better inverse candidate)

Aggregated Day-of-Week Win Rate

Per-Setup Day-of-Week Breakdown

News Tier Impact

SMA(20) Impact

Cross-Instrument Fail Clustering

When instrument A fails, how often does instrument B also fail?

Does Yesterday's Fail Affect Today's Outcome?

Compares win rate when the prior trade was a Fail vs a Pass, per setup. A positive delta means the setup tends to bounce back (mean reversion); negative means failures cluster (momentum).

Aggregated (All Setups)

Per-Setup Breakdown

Sorted by largest impact (|Δ|). Δ = After-Fail WR minus Overall WR.

Two Consecutive Failures

All trade models where the most recent 2 decided outcomes (Pass/Fail) are both Fail. Click a row to see setup outcome details, or select to submit via the order bar above.

Last Session Failures (All Trade Models)

Every trade model whose primary outcome on the last session was a Fail. Prior outcomes alone do not indicate edge; once the SKEW filter is applied, the likelihood of multiple sequential failures drops. The SKEW column shows whether the failed trade was Aligned (with daily trend vs SMA(20)) or Counter-trend. Click a row to see details, or select to submit via the order bar above.

🔥 FF Recovery Trades (Fail + Recovery Fail)

Setups where the most recent session had both the original trade AND the recovery trade hit their stop loss (FF outcome). Click a row to see details, or select to submit via the order bar above.

All FF Models from Last Session

Best Cross-Instrument Pairs

Click a row to select both setups, or use Select on portfolio cards below

Optimal Multi-Instrument Portfolios

Recovery vs Primary Win Rate by Day of Week

Recovery vs Primary Win Rate by Instrument

Per-Setup Recovery Performance

Sorted by recovery advantage (delta). Only setups with 3+ recovery trades shown.

Last Globex Session — Expired Trades

Trades whose entry filled but neither TP nor SL was hit by the end of the most recent Globex session. These positions carry open into the next session.

Resolution by Session Count

How many sessions after expiry until TP or SL was hit. Cumulative % shows how quickly expired trades resolve.

Expired by Day of Week (Origination)

Which day of the week do trades most often expire, and how do they resolve?

Expired Resolution by Instrument

Per-Setup Expired Resolution

Setups with 2+ expired trades. Adj WR includes resolved expired trades in the win rate calculation.

Advanced Setup Metrics

EV (R) = expected value per trade in R-units. Kelly = optimal bet fraction. Sharpe = risk-adjusted return. Sorted by EV.

EV / Kelly / Sharpe

Streaks & Drawdown

After 2W/2L = WR after 2 consecutive wins/losses. Max DD = largest cumulative drawdown in ticks.

Time Decay (Bar Duration)

Quick = trades resolving at or below median bar duration. Slow = above median.

Formation (Compression vs Extension)

Compression = fills at/below median formation bars. Extension = above median. Shows whether quick or slow setups perform better.

Post-Entry Follow-Through

How price behaves in bars 1–3 after fill. Follow-through = % with MFE > 0 after bar 1. Chop = % where MAE > MFE after bar 3.

ME Performance (Managed Execution Simulation)

Simulates the ME exit/re-entry strategy using each setup's MREC as the control level. When MAE exceeds MREC% of SL, ME would trigger an exit. Save Rate = % of losing trades where ME caps the loss early. Cut Rate = % of winning trades where ME exits prematurely (re-entry may recover). ME EV = adjusted expected value assuming saved losses capped at MREC level and cut winners go flat.

Regime Detection (Rolling 20-Trade Window)

Tracks WR over a rolling window to detect improving, degrading, or stable performance. Sparkline shows recent trajectory.

Monthly Seasonality Heatmap

Win rate by month of year. Cells colored green (high WR) to red (low WR). Gray = insufficient data (<3 trades).

Conditional Probability Chains

Win rate under specific combinations of DOW + News Tier + Prior Outcome. Only shows deviations ≥8% from overall WR.

DOW × News Tier Matrix (Aggregated)

Per-Setup Condition Chains (DOW × News × Prior Outcome)

Strategy Simulator

Monte Carlo simulation: given a trade model with recovery trades, how fast can you reach your profit target without busting the trailing drawdown?

For every setup, sweeps candidate risk ratios from 1:1 to 1:4 and finds the best one by the chosen metric. Search is analytical (uses stored MAE) and restricted to tighter-than-current ratios. In-sample optimum is picked on the first 70% of trades; out-of-sample (OOS) columns validate on the last 30%. Recommended ratio = plateau midpoint (the midpoint of the contiguous range where the metric stays within 5% of its peak) — prefer this over the raw argmax since it is more robust.