Opening Failed-Break Portfolio Add-On
A long-only opening failed-break add-on layered onto the active-RTH plus prior-day sweep portfolio. It adds 80 unique trades, lifts the combined book to 4.32 trades/week, and cuts modeled median evaluation pass time to 21 trading days in the 5,000-path baseline.
Why this is the current faster-pass branch
The failed-break sleeve increases frequency and improves modeled median pass speed while keeping funded payout survival near 100% in the test harness. The main weakness is heavy cost sensitivity: +0.10R cost drops eval pass to 91.7% and failure to 6.0%, so paper fill quality matters.
| Stress | Trades | Win | ExpR | PF | Max DD | Eval pass | Eval fail | Median | p95 DD |
|---|---|---|---|---|---|---|---|---|---|
| Baseline | 672 | 61.6% | +0.327R | 1.97 | 5.03R | 99.1% | 0.8% | 21d | $1,447 |
| +0.05R added cost | 672 | 61.3% | +0.277R | 1.78 | 5.38R | 97.0% | 2.3% | 25d | $1,739 |
| +0.10R added cost | 672 | 60.9% | +0.227R | 1.61 | 6.60R | 91.7% | 6.0% | 28d | $1,894 |
| Random 10% miss + 0.03R | 602 | 61.5% | +0.306R | 1.89 | 6.29R | 98.3% | 1.3% | 24d | $1,547 |
| Skip every 7th + 0.05R | 576 | 61.3% | +0.278R | 1.78 | 8.32R | 96.5% | 2.5% | 26d | $1,730 |
Entry, exit, and source trail
| Item | Public rule | Source note |
|---|---|---|
| Market | MES/MNQ futures micros under one portfolio governor | Combined anchor tradebook has 672 accepted trades after adding this sleeve. |
| Direction | Long-only failed-break/reclaim during the opening window | Candidate source: work/prop_alpha_expanded_or_failed_break_20260625/top_trades/07_X_or_failed_break_605ebdc8.csv |
| Entry | Closed-bar failed-break confirmation, intended next-bar execution | No intrabar repainting; cancel stale entries. |
| Stop | Immediate structure/bracket stop | No averaging down or stop widening. |
| Target | Positive-R bracket, portfolio target family 1.0-1.5R | Do not manually stretch targets unless retested. |
| Governor | $200 eval risk, $90 funded-risk model, global daily stop and profit lock inherited from anchor | One global governor must prevent correlated sleeve stacking. |
| Validation files | work/or_failed_break_combined_anchor_validate_5000p_20260629 | 5,000-path baseline, added-cost, missed-fill, and funded-mode simulations. |
Risk note
Strategy research is educational software and trading research, not financial advice. This is a paper/demo research candidate, not a live-funded approval. Live markets can break backtests through fees, slippage, latency, platform behavior, missed fills, rule changes, and regime shifts.