Expert Insights: Inside Our Strategy Development Process
At CodeGuard, we combine advanced technology with market expertise to create trading strategies that perform in real-world conditions. In this section, we share an in-depth look at our strategy development process, from initial performance goals and trading ideas to rigorous testing and real-time incubation. By leveraging data, market insights, and cutting-edge techniques, we ensure our strategies are robust, adaptable, and designed to thrive in the ever-changing world of futures trading.
Performance goals
Before developing a new trading strategy, we first establish clear objectives to guide our process. This includes defining the time frame (intraday, swing, or long-term), selecting the specific futures contracts to trade, and setting performance targets such as expected returns, drawdown limits, and risk-adjusted profitability. By setting these foundational goals, we ensure each strategy is designed with precision and aligned with our overall trading vision.
Trading Idea
Our trading ideas come from a variety of sources, ensuring a well-rounded and adaptable approach to the markets. These can be derived from chart patterns, market cycles, technical indicators, and historical price action. We also analyze economic reports, news events, and macro trends to identify potential opportunities. By combining data-driven insights with market intuition, we develop strategies that can adapt to different futures market conditions.
Limited Feasibility Testing
To evaluate the viability of a trading idea, we conduct initial testing on a randomly selected two-year period. This approach ensures that we don’t “burn” through all available data too early in the process. By analyzing performance on a limited dataset, we can quickly assess whether the idea has potential before committing to more extensive backtesting and refinement.
Walk Forward testing
We design our strategies with walk-forward testing to ensure robustness and adaptability in changing market conditions. Our process involves optimizing strategy parameters across six different walk-forward periods, each using a distinct segment of historical data. Instead of relying on a single best-performing set of parameters, we select the average performer, ensuring the strategy remains stable and effective across different market environments.
Monte Carlo simulation
To assess the robustness of our strategies, we conduct two types of Monte Carlo simulations: one based on closed trade results and another using day-to-day results. We run between 2,500 and 10,000 simulations, shuffling trade outcomes to generate all possible equity curve variations. This process helps us understand potential performance fluctuations and identify worst-case scenarios. We use the median values from these simulations as a guiding benchmark, ensuring our strategies are prepared for real-world market conditions.
Diversification and Position Sizing
In our strategy development, we place a strong emphasis on diversification rather than focusing solely on position sizing. By spreading risk across multiple asset classes, strategies, and timeframes, we create a more balanced and resilient approach. While position sizing is important, we believe that diversifying our trades helps reduce the impact of market volatility, improving the overall stability and long-term performance of our strategies.
Incubation
Every strategy undergoes a two-phase incubation process to validate its real-world performance before deployment:
Out-of-Sample Testing: During strategy development, we intentionally exclude the last 9 months of data. Once the strategy is finalized, we test it on this “unseen” data to assess its robustness and adaptability.
Live Market Tracking: After passing the initial test, the strategy is monitored on real-time data for 1 to 6 months. This phase helps us evaluate execution efficiency, slippage, and overall stability under live market conditions before full implementation.