🇨🇳 中文版

🥇 Gold Quant Project

XAUUSD Quantitative Trading System — Multi-Timeframe Multi-Strategy Channel Analysis + Backtesting

A Python-based quantitative trading project for gold, covering the complete pipeline: data acquisition → technical indicators → multi-strategy channel analysis → multi-period backtesting → live trading interface.

---

📁 Project Structure

Gold_Quant_Project/
├── analysis/                    # Channel analysis module
│   ├── channel_analyzer.py      # Multi-timeframe analyzer (aggregates strategies + consensus)
│   └── strategies.py            # 4 channel analysis strategies
├── backtest/                    # Backtesting engine
│   └── engine.py                # Backtrader-based engine wrapper
├── data/                        # Data layer
│   └── data_fetcher.py          # Data fetchers (YFinance / MT5)
├── execution/                   # Live execution
│   └── mt5_trader.py            # MetaTrader5 trading interface (Windows only)
├── factors/                     # Factor computation
│   └── technical_indicators.py  # RSI, MACD, SMA, ATR, etc.
├── strategies/                  # Trading strategies
│   ├── dual_ma_strategy.py      # Dual moving average strategy
│   ├── enhanced_ma_strategy.py  # Enhanced MA strategy
│   ├── swing_strategy.py        # Swing trading strategy
│   ├── optimized_swing.py       # Optimized swing V1
│   └── optimized_swing_v2.py    # Optimized swing V2
├── output/                      # Analysis output directory (.gitignore)
├── main.py                      # Main entry point
├── run_channel_analysis.py      # Multi-timeframe channel analysis script
├── run_backtest.py              # Basic backtest
├── run_4h_backtest.py           # 4H timeframe backtest
├── run_daily_backtest.py        # Daily timeframe backtest
├── run_long_backtest.py         # 25-year long-term backtest
├── run_swing_backtest.py        # Swing strategy backtest
├── run_optimized_backtest.py    # Optimized strategy backtest
├── run_optimized_v2_backtest.py # Optimized V2 backtest
├── test_integration.py          # Integration tests
├── requirements.txt             # Python dependencies
└── .gitignore

---

⚙️ Environment Setup

1. Create Conda Environment

conda create -n gold_quant python=3.11 -y
conda activate gold_quant

2. Install Dependencies

pip install -r requirements.txt

Dependencies:

Package	Purpose
pandas	Data processing
numpy	Numerical computation
ta	Technical indicators (ADX, Bollinger Bands, Donchian, etc.)
backtrader	Backtesting engine
yfinance	Yahoo Finance data source
python-dotenv	Environment variable management

Note: MetaTrader5 is Windows-only. On Linux, yfinance is used as the data source.

3. Environment Variables (Optional)

Create a .env file for configuration (e.g., MT5 account credentials):

cp .env.example .env  # Edit as needed

---

🚀 Features

1. Multi-Timeframe Multi-Strategy Channel Analysis

Analyzes gold price data across 1H / 4H / Daily / Weekly timeframes using 4 independent strategies, then aggregates results into a multi-strategy consensus.

4 Analysis Strategies

Strategy	CLI Name	Core Method	Best For
Linear Regression	regression	Regression slope + R² + ADX	Trend strength quantification
Bollinger Bands	bollinger	SMA(20) ± 2σ, bandwidth + %B	Volatility + overbought/oversold
Donchian Channel	donchian	N-period highest high / lowest low	Breakout confirmation
Trendline	trendline	Pivot high/low regression	Closest to manual chart drawing

Channel Types

Type	Condition
📈 Uptrend Channel	Positive trend indicators + sufficient strength
📉 Downtrend Channel	Negative trend indicators + sufficient strength
↔️ Sideways / Ranging	No clear trend or insufficient strength
🔄 Transitioning	Direction emerging but strength not confirmed

Usage

# Run all strategies (default)
conda run -n gold_quant python run_channel_analysis.py

# Select a single strategy
conda run -n gold_quant python run_channel_analysis.py --strategy bollinger

# Select multiple strategies (comma-separated)
conda run -n gold_quant python run_channel_analysis.py --strategy regression,donchian

# Analyze a different symbol
conda run -n gold_quant python run_channel_analysis.py --symbol SI=F  # Silver

Sample Output

╔══════════════════════════════════════════════════════════════╗
║         XAUUSD Multi-Timeframe · Multi-Strategy Report       ║
╠══════════════════════════════════════════════════════════════╣
║  📊 1H (Intraday)                                            ║
║     Regression  ↔️ Sideways   Pos:59%  5298/5062              ║
║     Bollinger   ↔️ Sideways   Pos:41%  5237/5176              ║
║     Donchian    ↔️ Sideways   Pos:50%  5237/5163              ║
║     Trendline   🔄 Transition Pos:68%  5225/5149              ║
║     ── Consensus: 🔄 Mixed (0 up / 0 down / 4 other)         ║
║  ...                                                         ║
║  📊 Weekly (Long-term)                                       ║
║     ── Consensus: 📈 Bullish (3/4 strategies agree)          ║
╚══════════════════════════════════════════════════════════════╝

Results are saved as timestamped JSON files in output/:

output/channel_gcf_20260226_152500_all.json

---

2. Strategy Backtesting

Built on the Backtrader engine, supports multiple trading strategies across different timeframes.

Available Strategies

Strategy	Description	Script
Dual MA	SMA golden/death cross	run_backtest.py
Enhanced MA	MA + momentum filter	run_optimized_backtest.py
Swing Trading	Trend following + swing capture	run_swing_backtest.py
Optimized Swing V2	Improved stops + short selling	run_optimized_v2_backtest.py

Backtesting Timeframes

Script	Timeframe	Description
run_backtest.py	1H	Quick short-period verification
run_4h_backtest.py	4H	Medium-period swing trading
run_daily_backtest.py	Daily	Medium-to-long-term trends
run_long_backtest.py	Daily (25 years)	Long-term robustness validation

Usage

# Run a backtest (e.g., swing strategy)
conda run -n gold_quant python run_swing_backtest.py

# View results
cat output/gold_swing_backtest.json

Backtest reports include: return rate, Sharpe ratio, max drawdown, trade count, win rate, and other key metrics.

---

3. Data Acquisition

• YFinance (default): Fetches GC=F (Gold Futures) OHLCV data via yfinance. Supports 1m to 1wk intervals. Cross-platform.
• MT5 (optional): Fetches XAUUSD real-time data via MetaTrader5 API. Windows only.

from data.data_fetcher import YFinanceDataFetcher

fetcher = YFinanceDataFetcher(symbol="GC=F")
df = fetcher.fetch_ohlcv(period="1y", interval="1d")

---

4. Technical Indicators

factors/technical_indicators.py provides one-call computation of all indicators:

• Trend: SMA (multiple periods), EMA, MACD
• Momentum: RSI
• Volatility: ATR, Bollinger Bands
• Trend Strength: ADX

---

5. Live Trading Interface (Windows Only)

execution/mt5_trader.py interfaces with MetaTrader5:

• Account info query
• Market / pending order placement
• Position management & closing
• Stop-loss / take-profit configuration

⚠️ Warning: Live trading involves real financial risk. Always test thoroughly on a demo account first.

---

📊 Quick Start

# 1. Clone the repository
git clone git@github.com:Br0wn2000/Gold_Quant_Project.git
cd Gold_Quant_Project

# 2. Setup environment
conda create -n gold_quant python=3.11 -y
conda activate gold_quant
pip install -r requirements.txt

# 3. Run channel analysis
python run_channel_analysis.py

# 4. Run strategy backtest
python run_swing_backtest.py

---

📄 License

This project is for educational and research purposes only. It does not constitute investment advice.