CQCharting Queen
CONFIDENTIAL
PROPRIETARY

📡 Signal Generation & Validation Engine

End-to-end pipeline from raw data to executable trading signals

Signal Generation Pipeline

6-stage processing architecture

1

Data Ingestion

Real-time market data from multiple sources — price, volume, options flow, VIX, yields, breadth

2

Feature Engineering

6-dimensional observation vectors computed: log returns, volatility, VIX, SMA distance, volume ratio, yield slope

3

Strategy Evaluation

All 7 strategies independently evaluate current market state against their entry/exit criteria

4

Ensemble Consensus

5-brain voting system produces weighted regime classification and confidence score

5

Risk Validation

Signal passes through 7 validation gates (see below) — any failure blocks emission

6

Signal Emission

Validated signal with confidence score, position size, stops, and targets emitted to execution engine

7 Validation Gates

Every signal must pass all gates

G1

Regime Alignment

Signal direction must align with current HMM regime classification

G2

Multi-Timeframe Confluence

15-minute and 1-hour signals must agree — no single-timeframe entries

G3

Risk Budget Check

Position size within per-trade and portfolio-level risk limits

G4

Correlation Filter

New position correlation with existing portfolio below threshold

G5

Volatility Gate

Current VIX within acceptable range for strategy type

G6

Liquidity Check

Sufficient volume and tight spreads for clean execution

G7

Circuit Breaker Clear

No active circuit breakers or system-level halt conditions

Confidence Scoring

0–100 composite score

Each signal receives a composite confidence score from 0 to 100, computed as a weighted average of: regime confidence (40%), strategy signal strength (30%), multi-timeframe alignment (20%), and volume confirmation (10%).

80–100

HIGH

Full position size

60–79

MEDIUM

50% position size

40–59

LOW

25% size or skip

Signal Lifecycle

From generation to close

Generated
Validated
Emitted
Executed
Monitored
Closed

Signals are tracked through their complete lifecycle. Each stage transition is logged with timestamp, confidence score, and any modifications. Post-close analysis feeds back into strategy optimization.