Nonparametric Sweep Regime Engine [PhenLabs]

• Nonparametric percentile ranks replace brittle fixed volume and wick thresholds
• Liquidity sweeps are scored by participation, wick extremity, and closing reclaim
• A trend-expansion gate avoids fading statistically extreme directional conditions
• Confirmed pivots create objective buy-side and sell-side liquidity references
• Targets adapt to setup risk and use opposing liquidity when it offers a valid objective
• A single 0–100 score compresses multiple confirmation layers into a readable decision aid

• Liquidity Pivot Engine: confirms swing highs and lows, extends active liquidity levels, and retires them after a raid
• Percentile Rank Engine: ranks volume, directional wick size, and EMA-spread expansion over a rolling sample
• Sweep Confirmation Gate: requires a close back through the raided level plus configurable score, volume, candle, and regime conditions
• Projection Engine: places an ATR-buffered stop and two risk-multiple targets, substituting opposing liquidity for the extended target when appropriate
• NSRE Dashboard: displays the latest direction, score, volume rank, trend rank, regime, stop, and targets

• Adaptive thresholds make the same logic portable across futures, crypto, FX, equities, and indices
• One-use liquidity states prevent repeated signals from the same pivot
• Confirmed pivots avoid lookahead in live signal logic
• Dashed liquidity and invalidation levels keep structure distinct from dotted target projections
• Independent bullish and bearish alert conditions support automation workflows
• All calculations use robust NA and zero-division guards

• Teal triangle: confirmed bullish sell-side liquidity sweep
• Magenta triangle: confirmed bearish buy-side liquidity sweep
• Dashed horizontal levels: active liquidity, swept reference, and invalidation
• Dotted horizontal levels: first and second projected objectives
• Top-right dashboard: latest signal state, percentile context, regime, and exact projected prices

• Pivot strength — Default: 5 — Range: 2–20 — Lower values react faster and create more liquidity references; higher values isolate more significant structure
• ATR length — Default: 14 — Range: 5–100 — Controls volatility normalization and the stop buffer baseline
• Require directional reclaim candle — Default: true — Requires the sweep bar to close in the intended reversal direction
• Percentile lookback — Default: 100 — Range: 30–500 — Shorter samples adapt faster; longer samples produce more stable ranks
• Minimum sweep score — Default: 65 — Range: 50–95 — Raise for fewer, more selective signals
• Minimum volume percentile — Default: 55 — Range: 0–100 — Sets the minimum relative participation required
• Maximum trend-expansion percentile — Default: 85 — Range: 40–100 — Lower values reject more countertrend sweeps during expansion
• Fast EMA — Default: 21 — Range: 2–100 — First component of the normalized trend-expansion metric
• Slow EMA — Default: 55 — Range: 5–250 — Second component of the normalized trend-expansion metric
• Stop ATR buffer — Default: 0.15 — Range: 0–2 — Adds volatility-adjusted space beyond the sweep extreme
• Target 1 risk multiple — Default: 1.0 — Range: 0.5–5 — Controls the first objective relative to setup risk
• Target 2 risk multiple — Default: 2.0 — Range: 1–10 — Controls the fallback extended objective
• Projection length — Default: 40 — Range: 10–200 — Sets how far the latest stop and targets extend

• Intraday reversal setups around established swing liquidity
• Filtering ICT and SMC sweep concepts with adaptive statistical context
• Comparing signal quality across instruments with different volume and volatility scales
• Locating risk-defined entries after stop runs in futures, indices, crypto, and FX

• Pivot levels require right-side confirmation and therefore appear after the structural turning point
• Percentile ranks need the selected lookback to warm up before signals can qualify
• Volume quality depends on the data supplied for the selected market
• Projected targets are analytical references and do not model slippage, commissions, or order execution

• Distribution-aware confirmation: every sweep is judged relative to recent market behavior rather than universal constants
• Regime-sensitive rejection: extreme trend expansion can invalidate an otherwise attractive countertrend sweep
• Liquidity-aware targeting: the extended objective can snap to opposing confirmed liquidity when that level is structurally valid

1. Confirmed swing highs and lows become active buy-side and sell-side liquidity references
2. Price must raid an active level and close back through it to form a raw sweep
3. The engine percentile-ranks volume, directional wick size, and trend expansion over the rolling sample
4. Volume, wick, and reclaim inputs produce a composite 0–100 sweep score
5. The score, participation, directional candle, and regime gates must all pass on the sweep bar
6. A valid signal projects an ATR-buffered invalidation level and two risk-normalized objectives

💡 Note:
Start with the default settings, then adjust the percentile lookback and minimum score to the instrument’s tempo. Higher-timeframe liquidity can improve context when using NSRE on lower execution timeframes. This tool is an analytical aid, not financial advice.