Low Risk SPX Day Trading Strategy [Part 2]

Key Concepts

• Flat Fly Strategy: An options trading strategy utilizing an Iron Butterfly with short strikes pinned at the previous day’s closing price and $10 wings.
• Iron Butterfly: A neutral options strategy that involves selling an at-the-money call and put, and buying an out-of-the-money call and put.
• Standard Deviation (SD) Filter: A statistical filter used to exclude trades when the market gaps beyond a specific volatility threshold.
• Profit Taking (PT): A predefined exit point to lock in gains before expiration.
• Drawdown: The peak-to-trough decline in the value of a trading account.
• Backtesting: The process of testing a trading strategy on historical data to determine its viability.

1. The Flat Fly Strategy Overview

The "Flat Fly" is designed to capitalize on mean reversion. By pinning the short strikes to the previous day's close, the strategy bets that the underlying asset will remain relatively flat throughout the trading day.

• Risk/Reward Profile: The strategy is characterized by low risk per contract ($75–$100) with the potential for high-reward payouts (often 10x the risk).
• Performance: Historical data suggests the underlying price reverts to the previous day's close approximately 25% of the time.
• Live Results: Initial live testing showed a streak of seven losses followed by two significant wins ($2,000 and $2,280 profit on risks of $200 and $225, respectively).

2. Optimization and Backtesting Methodology

Jack Slocum utilized the Option Alpha backtester to refine the bot’s performance, focusing on two primary variables: market entry filters and profit-taking thresholds.

Market Entry Filter (Standard Deviation)

Slocum tested various filters to avoid trading on days with extreme overnight gaps (e.g., a 2.5% move).

• Methodology: He implemented a 1.1 Standard Deviation filter. If the market opens more than 1.1 SD away from the previous close, the bot skips the trade.
• Result: This filter improved the win rate by 0.6%, slightly reduced the maximum drawdown, and increased overall profitability.

Profit Taking (PT) Thresholds

Slocum compared holding until expiration versus automated profit-taking.

• 50% PT: Increased the win rate to 29% but resulted in a higher drawdown and a lower average win amount.
• 70% PT (Selected Strategy): This was identified as the optimal setting. It allowed for higher average wins while maintaining the same average loss profile as the original strategy.

3. Comparative Backtest Results (3-Year Period)

Using a $500 position size, the backtest results for the optimized strategy (1.1 SD filter + 70% PT) compared to the original strategy are as follows:

• Total Profit: $103,000 (an increase of $8,000 over the original).
• Max Drawdown: Decreased by approximately $200.
• Efficiency: The optimized version showed a higher win rate and a lower frequency of total losses.

4. Key Arguments and Perspectives

• Automated Decision Making: Slocum argues that while manual intervention (closing trades at 3:55 p.m.) worked in the short term, automating these decisions removes emotional bias and ensures consistency.
• Risk Management: The strategy is presented as superior to "long" options (calls/puts) because it does not require a large directional move (2–4%) to be profitable; it only requires the market to remain relatively stable.
• Community Collaboration: Slocum emphasizes the importance of community-driven ideas, specifically crediting a community member for the suggestion of using standard deviation as a filter.

5. Synthesis and Conclusion

The "Flat Fly" strategy represents a high-reward, low-risk approach to daily options trading. Through rigorous backtesting, Slocum demonstrated that applying a 1.1 Standard Deviation filter to avoid volatile market gaps and setting a 70% profit-taking target significantly improves the strategy's robustness. The updated bot, incorporating these parameters, is designed to minimize drawdown while maximizing the potential for large, mean-reverting gains. Slocum encourages users to continue testing and sharing results to further iterate on the bot's performance.