Automating Ichimoku Cloud Retracement Strategy in Python (Full Backtest)

Summary

This video demonstrates how to automate a trading strategy in Python that combines the Ichimoku Cloud with an EMA trend filter. The presenter walks through step-by-step coding, signal creation, and backtesting on 4-hour historical data, achieving annual returns up to 40% with low drawdowns. Key concepts include filtering trades by trend using EMA, identifying retracement entries via the Ichimoku Cloud, and managing trades with ATR-based stop-loss and take-profit. The video also explores optimizing parameters, visualizing results with heat maps, and emphasizes that while the automated system sends signals, manual discretion can enhance trading effectiveness, particularly on stocks rather than Forex.

Takeaways

• 📈 The trading strategy combines the EMA 100 trend filter with Ichimoku Cloud retracement entries to trade with the prevailing market trend.
• 🔹 Uptrend condition: Current candle plus at least five previous candles must be fully above the EMA; downtrend is the opposite.
• ☁️ Entry signal requires at least 5–7 of the last 10 candles to be above (or below) the Ichimoku Cloud, with the current candle opening inside the cloud and closing in the trend direction.
• 💻 The strategy is implemented in Python using libraries such as numpy, pandas, technical analysis, and backtesting.
• 🛠 Functions include fetching data from yfinance, computing custom Ichimoku to avoid lookahead bias, calculating EMA trend signals, and generating entry signals.
• 📊 Backtesting uses ATR-based stop-loss and risk/reward ratios; the strategy is highly selective, yielding fewer trades but higher probability setups.
• 📉 Basic backtest results on a 4-hour timeframe show annual returns around 28%, a Sharpe ratio of 1, maximum drawdown of -6%, and a 53% win rate.
• 🔍 Optimization using grid search on ATR multipliers and risk/reward ratios can improve returns up to 43% annually with a win rate of 69% and Sharpe ratio of 1.38.
• ⚖️ Practical observations: Works better on trending stocks than choppy Forex markets, and small leverage can enhance returns while keeping drawdown manageable.
• 💡 Manual flexibility is recommended: traders can adjust entry rules slightly, ignore overly long candles, or consider support/resistance levels to improve real-world performance.
• 📧 The strategy can be automated to send notifications rather than trading live, allowing time-efficient monitoring on a 4-hour timeframe.
• 🎯 Heatmap analysis shows optimal parameter clusters, emphasizing the balance between stop-loss distance and take-profit for maximizing trend-following profits.

Q & A

What is the main goal of the trading strategy presented in the video?

The main goal is to automate a trading strategy that combines the Ichimoku Cloud with an EMA trend filter in Python. The strategy aims to capture profitable trading opportunities by using these indicators to filter trades and enhance decision-making.

How does the EMA trend filter work in the strategy?

The EMA trend filter uses the EMA 100 to determine the prevailing market trend. For a long trade, the strategy checks if the current candle and the previous five candles are fully above the EMA curve. For a short trade, the opposite is true, with candles needing to be below the EMA curve.

What role does the Ichimoku Cloud play in the strategy?

The Ichimoku Cloud is used to identify potential retracements within the trend. For a long trade, the strategy checks if the price is above the cloud, then looks for a candle that dips into the cloud and closes above it, signaling a reversal back into the trend.

Why is it important to check for previous candles above or below the Ichimoku Cloud?

Checking previous candles ensures that the market is in a strong trend. When at least six out of the last ten candles are above or below the Ichimoku Cloud, it confirms the strength of the trend and increases the probability of a successful trade.

What is the purpose of using the Average True Range (ATR) in the strategy?

The ATR is used to set stop-loss and take-profit levels based on market volatility. The stop-loss is calculated as 1.5 times the ATR, and the take-profit is set at twice the stop-loss distance. This helps manage risk and ensures appropriate trade exits.

What kind of backtesting was done in the video, and what were the results?

The backtesting was conducted on historical data with a focus on the 4-hour time frame across multiple assets. The strategy produced around 40% yearly returns in some cases, with a positive performance across various stop-loss and risk-reward combinations, even when accounting for commission fees.

How are the signals for long and short trades identified?

Long signals are generated when the trend is confirmed by the EMA filter, and a retracement occurs with a candle dipping into the Ichimoku Cloud and then closing above it. Short signals are generated by reversing these conditions, with candles being below the EMA and a retracement into the cloud followed by a close below it.

What were the challenges faced during backtesting, and how were they addressed?

A challenge was the strategy's selectivity, resulting in fewer trades. The strategy was designed to be very strict with its entry rules, which may have limited trading opportunities. This was addressed by considering manual trading flexibility, such as accepting smaller candles or adjusting parameters.

How does the optimization process work in the strategy?

The optimization process tests different ATR and risk-reward ratio multipliers using a grid approach. The best parameters for each asset are found by running the backtest across various combinations of these values to maximize returns while maintaining a manageable risk level.

What insights can be drawn from the heat map of backtest results?

The heat map visually shows the impact of varying ATR and risk-reward ratios on returns. It reveals clusters of high returns when there is a low stop-loss distance paired with a high risk-reward ratio, highlighting the optimal parameter combinations for maximizing profits.