Backtesting and Optimizing a Pair Trading Strategy

Summary

This example demonstrates a portfolio-strategy backtest for a pair trading strategy on two Dalian Commodity Exchange continuous contracts. It configures minute data over a specified historical interval and supplies commission rates, slippage, contract sizes, tick increments, and starting capital. The strategy uses Bollinger-band window and deviation settings; the example then loads data, runs the backtest, calculates results and statistics, and displays a chart.

It also shows two parameter-search methods: genetic-algorithm optimization and brute-force optimization, both targeting the Sharpe ratio. The example is a concise workflow rather than a research report: it gives no performance results, data-quality checks, out-of-sample validation, or discussion of transaction-cost realism. Its zero commission and slippage assumptions may make the illustrative backtest less representative of live trading.

Key ideas

• The example configures a minute-level pair strategy backtest on two commodity futures contracts.
• Backtest inputs include dates, costs, contract specifications, capital, and Bollinger-band parameters.
• The workflow calculates results and statistics and can display a chart.
• Genetic and brute-force searches optimize the strategy settings against Sharpe ratio.
• The example reports no outcomes and uses zero commission and slippage assumptions.

Tags

# backtesting demo


```python
from datetime import datetime

from vnpy_portfoliostrategy import BacktestingEngine
from vnpy.trader.constant import Interval
from vnpy.trader.optimize import OptimizationSetting

from vnpy_portfoliostrategy.strategies.pair_trading_strategy import PairTradingStrategy
```

```python
engine = BacktestingEngine()
engine.set_parameters(
    vt_symbols=["y888.DCE", "p888.DCE"],
    interval=Interval.MINUTE,
    start=datetime(2019, 1, 1),
    end=datetime(2020, 4, 30),
    rates={
        "y888.DCE": 0/10000,
        "p888.DCE": 0/10000
    },
    slippages={
        "y888.DCE": 0,
        "p888.DCE": 0
    },
    sizes={
        "y888.DCE": 10,
        "p888.DCE": 10
    },
    priceticks={
        "y888.DCE": 1,
        "p888.DCE": 1
    },
    capital=1_000_000,
)

setting = {
    "boll_window": 20,
    "boll_dev": 1,
}
engine.add_strategy(PairTradingStrategy, setting)
```

```python
engine.load_data()
engine.run_backtesting()
df = engine.calculate_result()
engine.calculate_statistics()
engine.show_chart()
```

```python
setting = OptimizationSetting()
setting.set_target("sharpe_ratio")
setting.add_parameter("boll_window", 10, 30, 1)
setting.add_parameter("boll_dev", 1, 3, 1)

engine.run_ga_optimization(setting)
```

```python
engine.run_bf_optimization(setting)
```

Shown in full with attribution under the source's licence. Licence: MIT

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.