📘 Chapter 4: "How to Reliably Test Strategies? Complete Guide to Freqtrade Backtesting Command"

Backtesting is one of the most critical steps in strategy development. By simulating your strategy on historical market data, you can effectively evaluate its performance and determine whether it’s worth deploying in live trading.

This article provides a detailed guide to Freqtrade’s backtesting command, including usage, common parameters, data handling, result analysis, multi-processing acceleration, and Docker usage.

---

🔁 1. What is Backtesting and Why Is It Important?

Backtesting is the process of running your strategy on historical data as “simulated trades”, aiming to understand how the strategy would have performed in past markets.

A quality backtest can answer questions such as:

• Is the risk/reward ratio reasonable?
• Is the win rate stable enough?
• Is the strategy overfitted? (May fail in live trading)
• Which parameters most significantly affect profits?

---

🚀 2. Basic Backtesting Command Structure

freqtrade backtesting \
  --config user_data/config.json \
  --strategy MyStrategy \
  --timeframe 15m \
  --timerange 20220101-20230701

Parameter Details

Parameter	Description
--config	Path to configuration file (includes pairs, timeframe, etc.)
--strategy	Strategy class name
--timeframe	Timeframe, e.g., 15m, 1h
--timerange	Backtesting period (format YYYYMMDD-YYYYMMDD)

Optional Parameters:

Parameter	Description
--export	Export detailed trade data as CSV
--stats-file	Output result statistics as JSON
--processes	Run multiple processes in parallel to improve speed

---

💡 3. Pre-Backtesting Preparation

Backtesting isn’t just running a command. Make sure to:

1. ✅ Download historical data for the target timeframe:

freqtrade download-data --timeframes 15m --timerange 20220101-20230701

2. ✅ Place strategy files in user_data/strategies/ with correct class names.

3. ✅ Ensure config.json is correctly set, including:

   • Correct trading pairs
   • Correct exchange
   • stake_currency set to USDT, BTC, etc.

---

🧪 4. Backtesting Output Explanation

After backtesting, Freqtrade will log information including:

Metric	Meaning
Total profit	Strategy’s final net profit
Total trades	Number of trades executed
Win / loss ratio	Win rate (ratio of profitable trades)
Sharpe Ratio	Risk-adjusted return
Avg trade duration	Average holding period
Drawdown	Maximum drawdown
Profit factor	Profit factor (profit / loss)

---

🖼️ 5. Visual Backtesting (backtesting-show)

Freqtrade provides backtesting-show to visualize buy/sell points and strategy behavior:

freqtrade backtesting-show \
  --config user_data/config.json

📍 Displays the strategy’s equity curve, trade markers, positions, etc.

---

🧩 6. Common Backtesting Issues

Issue	Possible Reason
No trades executed	Strategy signals too strict, or missing data
Data not available	Data not downloaded or timerange does not cover
Indicators are NaN	Wrong indicator settings / not suitable for timeframe
Suspected overfitting	High backtest profit but live loss; use forward test

---

🧠 7. Multi-Processing for Faster Backtesting (--processes)

For multiple pairs or complex strategies, --processes can speed up backtesting:

freqtrade backtesting \
  --config user_data/config.json \
  --strategy MyStrategy \
  --processes 4

🧪 Generally, set to half to all of CPU cores. For example, on an 8-core machine, set 4–8.

---

🐳 8. Backtesting in Docker

If running Freqtrade in Docker:

docker compose run --rm freqtrade backtesting \
  --config /quants/freqtrade/user_data/config.json \
  --strategy MyStrategy \
  --timeframe 15m \
  --timerange 20220101-20230701

Ensure user_data/ is mounted to /quants/freqtrade/user_data/ in the container.

---

📊 9. Exporting Backtesting Results

To save detailed trades:

--export user_data/backtest_result.csv

Or export strategy performance stats as JSON:

--stats-file user_data/backtest_stats.json

---

✅ 10. Recommended Analysis Metrics

Focus on these after backtesting:

• Total profit: Core metric
• Drawdown: Risk exposure
• Win rate / Risk-reward ratio: Stability
• Average trade profit: Value per trade
• Profit factor: Worthiness of trading

Combine with chart analysis:

freqtrade plot-dataframe --config user_data/config.json --strategy MyStrategy --timerange 20220101-20230101

---

📌 Summary

Freqtrade’s backtesting system is powerful and flexible, suitable for a complete strategy research and validation workflow.

Step	Tool / Command
Download historical data	download-data
Write strategy class	new-strategy
Run backtest	backtesting
Visualize results	backtesting-show / plot-dataframe
Accelerate performance	--processes for multi-processing
Containerized use	Docker commands