Wavelet Trading Strategy Backtester

A Python-based trading strategy that uses wavelet decomposition to analyze price trends and generate trading signals.

Overview

This project implements a trading strategy that:

• Uses wavelet analysis to decompose price data into trend and cycle components
• Backtests the strategy across multiple 30-day intervals
• Generates performance metrics and visualizations
• Performs Monte Carlo simulations using bootstrap resampling

Key Files

• wavelet.py - Main strategy implementation and backtesting code
• interval_performance_metrics.csv - Detailed performance metrics for each interval
• interval_returns_plot.png - Visualization of returns across intervals
• train_vs_test_returns.png - Performance comparison between training and test periods
• monte_carlo_simulations_bootstrap.png - Monte Carlo simulation results

Requirements

backtesting
pandas
numpy
pywt
matplotlib
arch
empyrical
scipy

Usage

1. Place your data files in the data/gold directory as CSV files with columns:

   • Volume
   • Date
   • Open
   • High
   • Low
   • Close

2. Run the strategy:

python wavelet.py

Key Features

• Wavelet decomposition of price data
• Rolling 30-day interval backtesting
• Performance metrics calculation including:
  • Returns
  • Sharpe ratio
  • Maximum drawdown
  • Win rate
  • SQN
  • Profit factor
• Monte Carlo simulation with bootstrap resampling
• Train/test period analysis
• Risk metrics including VaR and CVaR

Output

The script generates:

• Trading logs and performance metrics in CSV format
• Performance visualization plots
• Risk analysis and Monte Carlo simulation results
• Train