Technical Analysis

Overview

The technical analysis module processes raw market data through multiple analytical lenses to identify high-probability trading setups within the 30-minute hold constraint.

Analysis Categories

1. Trend Indicators

Indicator	Parameters	Signal
EMA Cross	9/21 periods	Bullish/Bearish cross
MACD	12/26/9	Histogram divergence
ADX	14 periods	Trend strength > 25
Parabolic SAR	0.02/0.2	Trend reversal dots

2. Momentum Indicators

Indicator	Parameters	Signal
RSI	14 periods	Oversold < 30, Overbought > 70
Stochastic	14/3/3	%K/%D crossover
CCI	20 periods	Extreme readings ±200
Williams %R	14 periods	Reversal zones

3. Volatility Indicators

Indicator	Parameters	Signal
Bollinger Bands	20/2σ	Band squeeze/expansion
ATR	14 periods	Position sizing, stop distance
Keltner Channels	20/1.5	Breakout confirmation

4. Volume Analysis

Indicator	Parameters	Signal
VWAP	Session	Price vs fair value
OBV	Cumulative	Divergence from price
Volume Profile	Session	Key levels, POC

Pattern Recognition

Candlestick Patterns (1-5 bar)

• Reversal: Hammer, Shooting Star, Engulfing, Morning/Evening Star
• Continuation: Three White Soldiers, Rising/Falling Three Methods
• Indecision: Doji, Spinning Top, Harami

Chart Patterns (Multi-bar)

• Double Top/Bottom
• Head & Shoulders
• Ascending/Descending Triangles
• Bull/Bear Flags
• Wedges

Confluence Scoring

Signals are scored based on confluence — multiple indicators agreeing:

def calculate_confluence_score(signals: list[IndicatorSignal]) -> float:
    """
    Score 0.0 - 1.0 based on how many indicators agree.
    
    Weights:
    - Trend alignment: 0.30
    - Momentum confirmation: 0.25
    - Volume confirmation: 0.20
    - Pattern match: 0.15
    - Key level proximity: 0.10
    """
    weights = {
        'trend': 0.30,
        'momentum': 0.25,
        'volume': 0.20,
        'pattern': 0.15,
        'level': 0.10
    }
    
    score = sum(
        weights[s.category] * s.strength 
        for s in signals 
        if s.direction == consensus_direction(signals)
    )
    return min(score, 1.0)

Minimum confluence threshold for signal generation: 0.65

Timeframe Analysis

Given the 30-minute max hold constraint:

Timeframe	Purpose
1-minute	Entry timing, micro-structure
5-minute	Primary signal generation
15-minute	Trend context, S/R levels
1-hour	Bias direction (trend filter)

Backtesting Requirements

• Minimum 30 days historical data
• Walk-forward optimization
• Out-of-sample validation
• Maximum drawdown tracking
• Sharpe ratio > 1.5 target