overnight sweep context, midpoint reclaim trigger, EMA trend filter, 1.5R bracket plotting
Tier B funded/eval demo research
Midday Overnight Sweep Reversal
A 5-minute MES/MNQ long-only midday reversal candidate built around overnight liquidity sweep behavior, trend context, and fixed 1.5R bracket exits. It passed the low-risk 10,000-path stress read, but still requires forward paper fills before any live automation label.
Chart5mMES/MNQ execution view
Trades1410.91 per week baseline
Win rate53.2%baseline tradebook
Target RR1.5Rpositive-R bracket
Expectancy+0.196Rper baseline trade
Profit factor1.44baseline gross wins/losses
Max DD7.96Rhistorical baseline path
Eval median46dp25 34 / p75 61
Eval risk$125recommended demo profile
Funded risk$80survival-first setting
Historical Equity Curveclosed-trade R path, baseline tradebook
Stress comparison
The candidate survived the low-risk pass, but not as a final live bot.
The best practical setting is $125 eval risk and $80 funded risk. It keeps the drawdown profile controlled and preserves funded survival, but it is still too new for real-money unattended automation until forward paper logs confirm fills.
| Stress | Trades | Win | ExpR | PF | Max DD | Eval pass | Eval fail | Median | Funded payout |
|---|---|---|---|---|---|---|---|---|---|
| Baseline | 141 | 53.2% | +0.196R | 1.44 | 7.96R | 88.1% | 3.1% | 46d | 99.8% |
| Miss every 7th fill | 121 | 54.5% | +0.219R | 1.50 | 8.47R | 88.9% | 2.1% | 46d | 99.8% |
| Miss every 7th fill + 0.05R cost | 121 | 54.5% | +0.169R | 1.37 | 10.39R | 75.0% | 5.8% | 52d | 98.7% |
Operating profile preview
Entry, stop, target, and governor logic
The public page shows the research-grade operating shape. Exact thresholds, alert code, and broker automation files stay gated until unlock and forward validation are ready.
| Item | Validated public rule | Execution note |
|---|---|---|
| Chart | 5-minute execution chart | Do not assume the same result on 1m, tick, range, or volume charts without retesting. |
| Market | MES/MNQ micro futures | Micro sizing is preferred for prop-rule control and fill drift monitoring. |
| Direction | Long only | No short-side version survived the same filter stack at comparable quality. |
| Session | Midday RTH window after overnight range context is known | Model is not designed for blind open impulse chasing. |
| Setup | Overnight sweep/reclaim behavior with trend-context filter | Exact threshold map is locked; the public page intentionally does not expose a turnkey recipe. |
| Entry | Closed-bar signal, next-bar/market style execution in the approximation | Forward paper testing must verify real fills and alert timing. |
| Stop | ATR/structure bracket, attached immediately | No widening, no averaging down, no manual rescue logic. |
| Target | Fixed 1.5R take-profit | Positive-R target, slower than a negative-R scalp but cleaner for funded survival. |
| Governor | $125 eval risk, $80 funded risk, stop after lockout conditions | Profile optimizes pass probability per drawdown, not raw profit. |
TradingView package
PineScript approximation not staged yet.
This candidate is promoted to the site first as research. PineScript, screenshots, and automation alert templates should be generated only after one more forward/demo replay check confirms the exact alert timing.
Will require buy signal, stop, target, trade ID, risk dollars, and kill-switch state in each alert.
Research page and PDF are staged. TradingView source is not public-published.
Sources and testing protocol
What this validation is based on
The page is a research summary, not audited live performance. These are the internal files and rule assumptions behind the published metrics.
| Area | Source | Test note |
|---|---|---|
| Market data | Normalized MES/MNQ continuous-contract intraday bars from the local Databento validation folder. | Coverage used by this research family spans 2023-06-01 through 2026-05-29 where full replay exists. |
| Tournament | work/prop_alpha_state_vwap_compression_5m_pulse1_20260628/EXPANDED_PROP_ALPHA_REPORT.md | 500 focused state/VWAP/compression specs; 38 survivors moved into prop simulation. |
| Tradebook | work/prop_alpha_state_vwap_compression_5m_pulse1_20260628/top_trades/01_X_state_overnight_sweep_reversal_4b61efa3.csv | Closed-trade R path behind the equity curve and baseline stats. |
| Stress sweep | work/candidate_stress_sweep_overnight_midday_4b61efa3_10k_20260628/CANDIDATE_STRESS_SWEEP.md | 10,000 Monte Carlo prop paths with baseline, missed-fill, added-cost, and combined stress modes. |
| Rule template | Topstep-style 50K prop simulator in the local lab. | Refresh from official firm rules before any real challenge; do not hardcode stale prop assumptions. |
| Costs | Simulator includes commission/slippage assumptions and explicit extra-cost stress. | Live fills must be audited before sizing beyond micro demo risk. |
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.