Backtesting with Python: Step-by-Step Tutorial

# Create virtual environment and install quant stack
python -m venv quant_env
source quant_env/bin/activate  # On Windows: quant_env\Scripts\activate

pip install pandas numpy yfinance backtesting bokeh

import yfinance as yf
import pandas as pd

def fetch_data(symbol="SPY", start="2020-01-01", end="2025-12-31"):
    df = yf.download(symbol, start=start, end=end, progress=False)
    
    # Flatten MultiIndex columns if returned by yfinance
    if isinstance(df.columns, pd.MultiIndex):
        df.columns = df.columns.get_level_values(0)
        
    df = df[['Open', 'High', 'Low', 'Close', 'Volume']].dropna()
    print(f"Loaded {len(df)} bars for {symbol}")
    return df

data = fetch_data("SPY")
print(data.head())

from backtesting import Strategy
from backtesting.lib import crossover

def calculate_ema(series, period):
    return pd.Series(series).ewm(span=period, adjust=False).mean()

def calculate_rsi(series, period=14):
    delta = pd.Series(series).diff()
    gain = (delta.where(delta > 0, 0)).rolling(window=period).mean()
    loss = (-delta.where(delta < 0, 0)).rolling(window=period).mean()
    rs = gain / loss
    return 100 - (100 / (1 + rs))

class EmaRsiStrategy(Strategy):
    fast_period = 20
    slow_period = 50
    rsi_period = 14

    def init(self):
        # Register indicators for plotting and execution
        price = self.data.Close
        self.fast_ema = self.I(calculate_ema, price, self.fast_period)
        self.slow_ema = self.I(calculate_ema, price, self.slow_period)
        self.rsi = self.I(calculate_rsi, price, self.rsi_period)

    def next(self):
        # Current bar close and indicator values
        price = self.data.Close[-1]
        
        # Entry Logic: Fast crosses above Slow with RSI > 50
        if crossover(self.fast_ema, self.slow_ema) and self.rsi[-1] > 50:
            if not self.position:
                # Place buy order with 2.5% stop loss and 5.0% take profit
                sl = price * 0.975
                tp = price * 1.05
                self.buy(sl=sl, tp=tp)
                
        # Exit Logic: Fast crosses below Slow
        elif crossover(self.slow_ema, self.fast_ema):
            if self.position.is_long:
                self.position.close()

from backtesting import Backtest

# Instantiate backtest with $100k cash and 0.1% round-turn commission
bt = Backtest(
    data,
    EmaRsiStrategy,
    cash=100_000,
    commission=0.001,
    exclusive_orders=True
)

# Run the simulation
stats = bt.run()
print(stats)

# Generate interactive HTML tearsheet and candlestick plot
bt.plot(filename="backtest_result.html")

# Grid Search: evaluate fast EMA from 10 to 30, slow EMA from 40 to 80
stats_opt = bt.optimize(
    fast_period=range(10, 30, 5),
    slow_period=range(40, 80, 10),
    maximize='Sharpe Ratio',
    constraint=lambda p: p.fast_period < p.slow_period
)

print(stats_opt._strategy)
print(f"Optimal Sharpe: {stats_opt['Sharpe Ratio']:.2f}")