Triple Liquidity Sweep Engine
- Sandra Wakefield
- 10 hours ago
- 3 min read
Access the Full Whitepaper:
1. Executive Summary
PSRC-LSMSS is a multi-timeframe Pine Script v6 AI Engine that identifies 3 key institutional liquidity raids during the New York AM session and confirms them against a structural market-shift model before generating a tiered, alert-ready signal. It was built following an explicit Quant-First, Code-Second discipline: the edge hypothesis, timeframe stack, and risk model were defined before a single line of entry logic was written, and every stage of the resulting pipeline has been individually stress-tested against three months of live tick data across six markets.
This manual documents what the indicator does, why its architecture reflects institutional development standards, how to configure it correctly, which assets and timeframes it is validated for, and — just as importantly — what it has not yet proven. Section 8 (Validation Status) should be read in full before any live use.
2. Core Methodology
2.1 The Edge Thesis
Institutional order flow requires liquidity to fill size. Retail stop-losses and breakout orders cluster at obvious levels — session highs and lows, the prior day's high/low, and equal highs/lows (double tops/bottoms). Smart money routinely engineers price through these levels to access that resting liquidity before reversing in the intended direction. The New York AM session (approx. 08:30–11:00 EST) is the highest-probability window for this behavior because it carries the volume and volatility to sweep multiple stacked pools — left over from the Asian and London sessions — in a single move, and because it frequently sets the day's true directional bias.
2.2 Market Logic
The core hypothesis is confluence-weighted: the more distinct liquidity pools are swept in the same raid, the larger the implied institutional fill, and the higher the probability that the subsequent reversal is genuine rather than a shallow stop-run that continues in the original direction. The indicator scores this numerically (0–4) rather than relying on a single swept level, and only arms a setup once a configurable minimum score is met.
2.3 The Multi-Timeframe Stack
Every arm decision passes through four independent timeframe roles, each with a distinct job:
• Bias (Daily + H4): Directional filter. A setup is only considered valid when the higher-timeframe structure agrees with the direction of the reversal being traded.
• Confirmation (H1): A secondary bias check, displayed for context and optionally enforced as a hard gate.
• Detection (M5): Where the actual liquidity sweep and Change-of-Character (CHoCH) / Market Structure Shift (MSS) are identified.
• Entry Refinement (M1): Where the retracement into a Fibonacci OTE zone and the final micro-structure confirmation are timed. The indicator must run on the M1 chart for this layer to function correctly.
3. Why This Is Institutional-Grade Engineering
“Institutional-grade” in this document refers to engineering methodology — the discipline and rigor applied to building and validating the system — not a claim about proven returns. Five design properties distinguish this build from a typical retail indicator:
3.1 Non-Repainting Architecture
All higher-timeframe structure is derived from confirmed pivots (a pivot only exists once price has moved a defined number of bars past it), and all cross-timeframe data requests use lookahead-disabled security calls. The Previous Day High/Low reference uses only the prior, fully-closed daily bar. Nothing in the signal path depends on a bar that has not yet closed at the relevant timeframe.
3.2 Liquidity Confluence Scoring
Rather than a binary “price broke a level” trigger, the indicator maintains independent swept/unswept state for four liquidity pool types per session and requires a configurable minimum score before arming. This directly encodes the edge thesis in Section 2.2 rather than trading every isolated sweep.
3.3 Tiered Signal Classification
Entries are not treated as pass/fail. A retracement into the 62–79% Fibonacci zone (“Premium”) sits close to the invalidation line and offers materially better risk/reward than a 50–61% retracement (“Standard”). The indicator labels each fired signal by tier so position sizing and conviction can be adjusted per trade rather than treating every alert identically.
3.4 ATR-Relative Adaptive Logic
Structural pivot tolerances, volatility filtering, and invalidation buffering are all expressed as multiples of Average True Range rather than fixed price distances. This allows the same configuration to operate correctly on a $4,100 gold contract and a 1.14 EUR/USD rate without per-instrument hand-tuning.
3.5 Built-In Diagnostic Transparency
The indicator ships with an optional Signal Funnel panel that records, in real time, exactly how many candidate setups pass each individual gate — session timing, bias resolution, direction match, confluence, volatility filter, zone-tap, and final confirmation — plus an ATR-normalized overshoot distribution for every setup that reaches a zone. This is the same instrumentation used internally during development to diagnose and correct three separate defects before this release. Institutional systems are audited, not trusted blindly; this indicator exposes its own internals so you can audit it too.
Access the Full Whitepaper: