📘Chapter 5: "How to Optimize Strategy Parameters? Freqtrade Hyperopt Quick Start"

This article was produced by the Quantitative Trading Lab at https://www.itrade.icu. Visit for more benefits.

In strategy development, beyond constructing buy/sell logic, parameter settings often determine the final profit and risk ratio.

Freqtrade provides a powerful hyperopt function for automated search of optimal parameter combinations, greatly accelerating strategy iteration.

---

🧠 1. What is Hyperopt and When to Use It?

Hyperopt is an automated parameter optimization tool that can:

• Find the best RSI threshold

• Test optimal take-profit and stop-loss levels

• Automatically try multiple parameter combinations → compare results → find the best configuration

✅ Suitable scenarios:

• Strategies with multiple numerical parameters (e.g., RSI, MACD, Bollinger Band width, stop-loss ratio)

• Searching for combinations that perform best over historical periods

• Avoiding manual parameter tuning

---

🚻 2. Dependency Installation

Hyperopt requires additional modules:

pip install 'freqtrade[hyperopt]'

Tip

For full functionality including freqai or telegram support:

pip install 'freqtrade[all]'

---

🚀 3. Basic Command and Parameters

freqtrade hyperopt \
  --config user_data/config.json \
  --strategy MyStrategy \
  --hyperopt-loss SharpeHyperOptLoss \
  --timerange 20230101-20230701 \
  --epochs 100

Parameter	Description
--config	Configuration file path
--strategy	Strategy class to optimize
--hyperopt-loss	Optimization target function
--timerange	Backtest time range
--epochs	Number of iterations (more = more precise but slower)
--spaces	Which parameter spaces to optimize (default: buy, sell)

---

🎯 4. Common Hyperopt Loss Functions

Function Name	Meaning	Scenario
SharpeHyperOptLoss	Optimize Sharpe ratio	Balance risk and return
SortinoHyperOptLoss	Optimize Sortino ratio	Focus on downside risk
ProfitHyperOptLoss	Maximize total profit	Aggressive return-driven strategy
CalmarHyperOptLoss	Return / max drawdown ratio	Risk-aware preference
TrailingBuyHyperOptLoss	For trailing buy strategies

---

🧩 5. Defining Hyperparameters

Use the @parameter decorators in your strategy class:

from freqtrade.strategy import IStrategy, IntParameter, DecimalParameter

class MyStrategy(IStrategy):
    rsi_period = IntParameter(10, 30, default=14, space="buy")
    stoploss_value = DecimalParameter(-0.10, -0.01, default=-0.05, space="sell")

Freqtrade will automatically search for the best combination within the defined ranges.

---

⚠️ 6. Common Pitfalls

❗ Overfitting

Short time frames or single-scenario data may lead to strategies that only perform well historically but fail live.

Recommendations:

• Use longer time periods (6+ months)

• Run Hyperopt multiple times → compare parameter convergence

• Reserve part of data for forward testing

❗ Improper Loss Function

Optimizing only for profit may ignore risk, creating extreme strategies.

Recommendations:

• Prefer Sharpe or Sortino as default

• For low risk tolerance, use CalmarHyperOptLoss

❗ Large Search Space / Parameter Conflicts

Too many combinations increase search time and may cause conflicts.

Recommendations:

• Limit parameters to 3–6

• Use effective ranges (e.g., RSI from 10–50 instead of 1–100)

---

🛠️ 7. Example Strategy and Hyperopt Run

Example strategy:

from freqtrade.strategy.interface import IStrategy
from pandas import DataFrame
from freqtrade.strategy import IntParameter, DecimalParameter
import talib.abstract as ta

class MySimpleStrategy(IStrategy):
    timeframe = '15m'
    rsi_buy = IntParameter(10, 50, default=30, space="buy")
    stoploss_value = DecimalParameter(-0.1, -0.01, default=-0.05, space="sell")
    stoploss = -0.05

    def populate_indicators(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe['rsi'] = ta.RSI(dataframe, timeperiod=14)
        return dataframe

    def populate_buy_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe.loc[(dataframe['rsi'] < self.rsi_buy.value), 'buy'] = 1
        return dataframe

    def populate_sell_trend(self, dataframe: DataFrame, metadata: dict) -> DataFrame:
        dataframe['sell'] = 0
        return dataframe

Local run:

freqtrade hyperopt \
  --config user_data/config.json \
  --strategy MySimpleStrategy \
  --hyperopt-loss SharpeHyperOptLoss \
  --timerange 20230101-20230630 \
  --epochs 100

Docker run:

docker compose run --rm freqtrade hyperopt \
  --config /quants/freqtrade/user_data/config.json \
  --strategy MyStrategy \
  --hyperopt-loss SharpeHyperOptLoss \
  --timerange 20230101-20230630

---

🎯 Adjustable Parameters

Parameter	Type	Description
rsi_buy	IntParameter(10,50)	RSI buy threshold
stoploss_value	DecimalParameter(-0.1,-0.01)	Stop-loss ratio

Tips

1. Use .value to access actual parameter value

2. space="buy" / "sell" defines Hyperopt search space

3. default=... is the manually set default

4. Run backtesting first to confirm basic strategy logic

---

📊 8. Hyperopt Evaluation Metrics

Common metrics:

Metric	Description
Sharpe Ratio	Annualized return / volatility
Sortino Ratio	Annualized return / downside volatility
Calmar Ratio	Annualized return / max drawdown
Total Profit	Total profit
Drawdown	Max drawdown
Avg Trade Duration	Average trade duration

You can also define custom evaluation functions.

---

✅ 9. Recommended Workflow

1. Use new-strategy to define strategy and parameters
2. Run backtesting to validate basic logic
3. Use hyperopt for automated parameter tuning
4. Replace strategy parameters with optimal values
5. Run backtesting again to verify
6. Visualize results → decide whether to go live

---

📌 Summary

Freqtrade Hyperopt provides powerful support for parameter tuning, but correctly defining parameter ranges, loss functions, and data periods is key.

📍 The goal is not maximum profit, but stable, risk-resistant, and generalizable parameters.