Index Enhancement Strategy Backtest System

Quick Start

# Run backtest with a strategy
bash run.sh baseline              # Baseline strategy
bash run.sh fixed_lot_baseline    # Fixed lot version
bash run.sh basis_timing_roll     # Basis timing roll

# List all available strategies
bash run.sh

# Run all strategies
bash run_all.sh

Available strategies:

• baseline, smart_roll, basis_timing, basis_timing_roll, spread_timing, liquidity_roll, aery_roll
• Fixed lot versions: fixed_lot_baseline, fixed_lot_smart_roll, fixed_lot_basis_timing, fixed_lot_basis_timing_roll, fixed_lot_spread_timing, fixed_lot_liquidity_roll, fixed_lot_aery_roll

Strategy configs are in configs/strategies/. Edit them to customize parameters (e.g., fut_code, start_date).

---

Project Overview

This is an OOP-based backtest system for equity index futures enhancement strategies. It captures excess returns over the benchmark by leveraging the tendency of a discounted futures price to converge to the spot index level as the contract approaches expiration.

Quick Start

1. Preprocess Data

python scripts/filesync_client.py # get data from filesync
python scripts/preprocess_data_from_wind.py

2. Configure Strategy

Edit config.toml to choose:

• fut_code: IC (CSI500), IM (CSI1000), or IF (CSI300)
• strategy_type: baseline or basis_timing
• parameters: roll days, leverage, thresholds, etc.

3. Run Backtest

python main.py                    # Use default config.toml
python main.py path/to/config.toml  # Use custom config

4. View Results

Results are saved to output/{strategy}_{fut}/:

• report.png - Visual dashboard with NAV, drawdown, metrics
• trade_log.csv - All trades with timestamps and prices
• nav_series.csv - Daily NAV for further analysis

5. Run Tests

pytest tests/ -v

Dependencies

• Python 3.11+
• polars, pandas, numpy
• matplotlib
• loguru
• pytest (for testing)

Documentation

• Design Document - System architecture and class design
• Backtest Flow - Step-by-step simulation process
• Examples - Interactive Jupyter notebooks

License

Academic use only.