Running Backtests

Overview

Item	Description
Goal	Master the three backtest interfaces
Prerequisite	Writing Strategies

---

1. Running Backtests

1.1 run_strategy (Recommended)

One-stop backtest execution with automatic report generation.

result = run_strategy(
    initialize,
    start_date='2024-01-01',
    end_date='2024-12-31',
    starting_cash=100000,
    benchmark='000300.XSHG',
    securities=['601390', '600519'],
    report_dir='reports',
)

1.2 run_backtest (Fine-grained Control)

Runs the backtest only, without generating reports.

result = run_backtest(
    initialize,
    start_date='2024-01-01',
    end_date='2024-12-31',
    starting_cash=100000,
)

1.3 run_portfolio_backtest (Portfolio Backtest)

High-level interface for multi-stock portfolio backtesting.

from eqlib import StrategyConfig, run_portfolio_backtest

config = StrategyConfig(
    starting_cash=200000,
    securities=["601390", "600519", "000858"],
    benchmark="000300.XSHG",
    position_pct=0.33,
    start_date="2024-01-01",
    end_date="2024-12-31",
    report_suffix="momentum_v1",
)

def my_strategy(context):
    for sec in context.universe:
        hist = attribute_history(sec, 25, "1d", ["close"])
        if hist.empty:
            continue
        ma20 = hist["close"].tail(20).mean()
        price = hist["close"].iloc[-1]
        if price > ma20 * 1.02:
            order_value(sec, context.portfolio.available_cash)
        elif price < ma20 * 0.98 and context.portfolio.positions.get(sec):
            order_target(sec, 0)

result = run_portfolio_backtest(config, my_strategy, report_dir="reports")

1.4 Benchmark Comparison

Set the benchmark via the benchmark parameter (defaults to 000300.XSHG, the CSI 300 Index). The results will automatically calculate alpha, beta, and information ratio.

---

For complete parameter documentation, see Backtest Engine API Reference.