The General Settings for the Automatic Strategy Builder (ASB) allow users to configure key parameters that influence how strategies are generated, optimized, and managed. These settings encompass conditions for strategy building, trade types, genetic algorithm configurations, and management of the evolutionary process.

 

Conditions to Generate

The "Conditions to Generate" settings in the Automatic Strategy Builder (ASB) allow traders to define the foundational rules and parameters that govern how trading strategies are constructed. These settings ensure that the strategies are built based on precise entry and exit conditions, leveraging indicators and signals in a structured manner. Here is a detailed breakdown of each setting:

 

Global Indicator Period Setting
  • Description: This setting establishes the default period for all indicators used in the strategy generation process. The period is a key parameter that affects how indicators interpret market data, influencing the sensitivity and responsiveness of the indicators to price movements.
  • Usage: By setting a global indicator period, users can ensure consistency across all indicators, which simplifies the strategy development process and ensures that all indicators operate over the same time frame.
  • Example: If the default period is set to 14, all indicators like SMA (Simple Moving Average), RSI (Relative Strength Index), and ATR (Average True Range) will use a 14-period calculation by default unless specifically overridden.

 

Number of Entry Conditions
  • Description: This parameter defines the number of conditions that will be used to construct the entry rules for a trading strategy. Entry conditions are the criteria that must be met for a trade to be initiated.
  • Usage: Specifying the number of entry conditions allows users to control the complexity of the strategy. More conditions can lead to more refined and specific entry signals, while fewer conditions might result in broader, more frequent signals.
  • Example: If set to 3, the strategy might require that the RSI is below 30, the price is above the 200-day SMA, and the MACD line has crossed above the signal line before entering a trade.

 

Number of Exit Conditions
  • Description: This setting determines how many conditions will be used to build the exit rules for a trading strategy. Exit conditions are criteria that signal when to close a trade.
  • Usage: Defining the number of exit conditions helps manage how trades are concluded, potentially optimizing profit-taking and loss-cutting processes. Multiple conditions ensure that exits are well-timed and based on comprehensive analysis.
  • Example: If set to 2, the strategy might exit a trade when the price falls below the 50-day SMA or when the ATR indicates increased volatility beyond a set threshold.

 

Number of Exit Signals
  • Description: This parameter sets the number of distinct exit signals that the strategy will use. Exit signals are specific triggers or events that cause the strategy to close an open position.
  • Usage: By setting a specific number of exit signals, traders can fine-tune their strategies to respond to multiple market conditions and risk scenarios, improving the robustness and adaptability of the strategy.
  • Example: If set to 3, the strategy could include signals such as a trailing stop being hit, a predefined profit target being reached, or a reversal signal from a key indicator like the MACD.

 

Summary

The "Conditions to Generate" settings in the ASB are critical for defining how trading strategies are formulated. By carefully configuring the global indicator period, the number of entry and exit conditions, and the number of exit signals, traders can create well-rounded and effective trading strategies. These settings provide the foundation for strategies that are responsive to market conditions, ensuring they can enter and exit trades based on precise, rule-based criteria.

 

 

Trade Type

The "Trade Type" settings in the Automatic Strategy Builder (ASB) determine the kinds of trades the generated strategies are allowed to execute. These settings ensure that the strategies align with the trader's preferred trading style and market approach. Below is a detailed explanation of each trade type option:

 

Both (Long and Short)
  • Description: This setting allows the strategy to open both long and short positions. Long positions are taken when the strategy anticipates the price of an asset will rise, while short positions are taken when the strategy expects the price to fall.
  • Usage: Enabling both long and short positions ensures that the strategy can take advantage of various market conditions, whether bullish or bearish. This flexibility is crucial for maximizing trading opportunities and achieving balanced performance across different market cycles.
  • Example: A strategy with this setting might buy (go long) an asset when the price crosses above a moving average and sell (go short) the same asset when the price crosses below the moving average.

 

Long
  • Description: This setting restricts the strategy to only open long positions. A long position is initiated with the expectation that the price of the asset will increase over time.
  • Usage: Choosing only long positions is ideal for traders who prefer to capitalize on upward market trends and may be particularly suitable for markets or assets that generally appreciate in value. This setting also aligns with certain regulatory or risk management policies that may limit short selling.
  • Example: A strategy configured to only open long positions might buy an asset when it breaks out above a resistance level, without ever taking a position that profits from a price decline.

 

Short
  • Description: This setting restricts the strategy to only open short positions. A short position is taken with the expectation that the price of the asset will decrease.
  • Usage: Selecting only short positions is suitable for traders looking to profit from declining markets or assets. This approach can be particularly effective during bearish market conditions or for assets that exhibit regular price declines.
  • Example: A strategy set to only open short positions might sell an asset short when it drops below a support level, aiming to buy it back at a lower price later.

 

Summary

The "Trade Type" settings in the ASB are essential for defining the directional bias of the trading strategies. By choosing between both long and short, only long, or only short positions, traders can tailor the strategy generation process to fit their market outlook, risk tolerance, and regulatory constraints. These settings ensure that the generated strategies align with the trader's overall trading philosophy and are equipped to capitalize on anticipated market movements.

 

Genetic Algorithm Settings

 

The "Genetic Algorithm Settings" in the Automatic Strategy Builder (ASB) control how the genetic optimization process is conducted. Genetic algorithms are used to iteratively improve trading strategies by simulating the process of natural selection. These settings allow traders to fine-tune the optimization parameters to enhance strategy performance. Below is a detailed breakdown of each setting:

 

Number of Maximum Generations
  • Description: This setting specifies the maximum number of generations the genetic algorithm will create before terminating. Each generation represents a cycle of evaluation, selection, crossover, and mutation to produce new strategy variants.
  • Usage: Setting an appropriate number of generations ensures that the optimization process has enough iterations to converge on an optimal or near-optimal strategy. Too few generations might result in under-optimized strategies, while too many could lead to overfitting.
  • Example: If set to 100, the genetic algorithm will evolve and evaluate strategies across 100 generations before stopping.

 

Island Population Size
  • Description: The genetic algorithm uses an island architecture, where multiple populations (islands) evolve in parallel. This setting determines the number of optimizing object instances created per island.
  • Usage: A larger population size can increase genetic diversity and improve the algorithm's ability to explore the solution space. However, it also requires more computational resources.
  • Example: If set to 50, each island will start with 50 strategy variants.

 

Crossover Probability
  • Description: This setting defines the probability that two optimizing object instances will exchange their genes during the crossover phase. Crossover combines parts of two parent strategies to produce offspring with potentially better performance.
  • Usage: A higher crossover probability encourages more frequent genetic mixing, which can help discover new and effective strategy combinations. Conversely, a lower probability favors retaining existing successful strategies.
  • Example: If set to 0.7, there is a 70% chance that two selected strategies will exchange genetic material to create new strategies.

 

Mutation Probability
  • Description: This setting specifies the likelihood that the genes of an optimizing instance will mutate. Mutation introduces random changes to strategy parameters, promoting genetic diversity and helping to escape local optima.
  • Usage: A higher mutation probability increases the chance of exploring new strategy configurations, which can be beneficial for finding innovative solutions. However, excessive mutation can disrupt converging towards an optimal strategy.
  • Example: If set to 0.1, there is a 10% chance that any given strategy parameter will mutate.

 

Number of Islands
  • Description: This setting determines the number of islands in the genetic architecture. Each island evolves independently but can exchange individuals during migration phases.
  • Usage: More islands can enhance parallel exploration of the solution space and prevent premature convergence. It also allows for diverse subpopulations that can evolve different strategies.
  • Example: If set to 5, the algorithm will run 5 separate populations (islands) in parallel.

 

Interval of Migration
  • Description: This setting defines the number of generations that pass before optimization instances migrate between islands. Migration allows sharing of genetic material across islands, promoting diversity and potentially leading to better overall solutions.
  • Usage: Setting an appropriate migration interval ensures that islands remain diverse while still benefiting from successful strategies developed on other islands.
  • Example: If set to 10, every 10 generations, a portion of the population will migrate between islands.

 

Rate of Population Migration
  • Description: This setting specifies the percentage of the island population that will migrate during each migration event. Migration helps spread beneficial genetic traits across different populations.
  • Usage: A balanced migration rate ensures that islands benefit from genetic exchange without overwhelming their unique evolutionary paths.
  • Example: If set to 20%, one-fifth of each island’s population will migrate during a migration event.

 

Summary

The "Genetic Algorithm Settings" in the ASB are crucial for controlling the optimization process of trading strategies. By configuring the number of generations, island population size, crossover and mutation probabilities, and migration parameters, traders can fine-tune how strategies are evolved and improved. These settings enable the genetic algorithm to effectively explore the solution space, maintain genetic diversity, and converge on robust trading strategies.

 

Substitute Underperforming Strategies

 

The "Substitute Underperforming Strategies" settings in the Automatic Strategy Builder (ASB) are designed to enhance the genetic evaluation process by continually improving the population of strategies. By identifying and replacing underperforming strategies, these settings help maintain a high level of performance and diversity within the strategy pool. Below is a detailed explanation of each setting:

 

Identify and Substitute Identical Strategies
  • Description: This setting allows the algorithm to detect and replace identical strategies within the population. Identical strategies do not contribute to genetic diversity and can hinder the optimization process.
  • Usage: Enabling this setting ensures that the strategy pool remains diverse by continually introducing new strategies. This can prevent stagnation and promote the discovery of innovative solutions.
  • Example: If two or more strategies in the population are identical, they will be identified and replaced with new, randomly generated strategies or variations of existing ones.

 

Every X Generations Substitute Y% of the Least Effective Strategies
  • Description: This setting specifies that a certain percentage of the least effective strategies will be replaced with new strategies every X generations. This helps to continuously improve the population by discarding underperforming strategies and introducing new ones.
  • Usage: By regularly replacing the weakest strategies, the algorithm ensures that the population remains competitive and adaptive. This can lead to a more efficient optimization process and better overall performance of the generated strategies.
  • Example: If set to replace 10% of the least effective strategies every 5 generations, the algorithm will evaluate the performance of all strategies and substitute the bottom 10% with new candidates every 5 generations.

 

Summary

The "Substitute Underperforming Strategies" settings are vital for maintaining a dynamic and effective strategy pool within the ASB. By identifying and replacing identical strategies, and regularly substituting the least effective strategies, these settings ensure that the genetic algorithm remains focused on improving performance and discovering new, high-potential trading strategies. This approach helps prevent stagnation, promotes diversity, and enhances the overall quality of the generated strategies.

 

Evolutionary Process Management

 

The "Evolutionary Process Management" settings in the Automatic Strategy Builder (ASB) oversee the ongoing operation and optimization of the genetic algorithm. These settings determine how the evolutionary process is managed, ensuring continuous improvement and effective strategy generation. Below is a detailed explanation of each setting:

 

Start Again When Finished (Continuous Repeating Evolution)
  • Description: This setting allows the genetic algorithm to restart automatically after it completes the specified number of generations. This continuous loop ensures that the strategy generation process is ongoing, continuously seeking to improve and adapt strategies over time.
  • Usage: Enabling this setting is beneficial for traders who want the strategy builder to run indefinitely, constantly generating and optimizing strategies without manual intervention. This helps in keeping the strategy pool fresh and updated with new ideas.
  • Example: Once the algorithm finishes 100 generations, it will automatically restart and begin a new cycle of 100 generations, repeating this process continuously.

 

Restart Evolution if Fitness Stagnates for X Generations
  • Description: This setting triggers a restart of the genetic algorithm if there is no improvement in the fitness of the population over a specified number of generations. Fitness stagnation indicates that the current population is no longer evolving effectively.
  • Usage: This setting is crucial for preventing the algorithm from getting stuck in local optima. By restarting the evolution process after a period of stagnation, the algorithm is given a fresh start to explore new solutions and improve overall fitness.
  • Example: If set to 10 generations, the algorithm will restart if it does not find any improvement in the best strategy’s fitness score for 10 consecutive generations.

 

Automatically Send Generated Strategies to Result Analysis
  • Description: When this setting is enabled, all generated strategies are automatically sent to the Result Analysis module for evaluation and reporting. This integration ensures that traders can easily review and analyze the performance of new strategies without manual export.
  • Usage: This setting streamlines the workflow by automatically transferring generated strategies to the analysis phase, allowing traders to focus on reviewing results and making informed decisions about which strategies to implement.
  • Example: After each generation, the best-performing strategies are sent directly to the Result Analysis module, where detailed performance metrics and reports are generated for trader review.

 

Summary

The "Evolutionary Process Management" settings are essential for overseeing and maintaining the efficiency of the genetic algorithm in the ASB. By enabling continuous evolution, restarting the process when fitness stagnates, and automating the transfer of strategies to the Result Analysis module, these settings ensure that the strategy generation process is dynamic, adaptive, and seamlessly integrated into the overall trading workflow. This approach helps maintain a high level of innovation and effectiveness in the generated trading strategies.

Cookies Notice

We use cookies to improve your experience, personalize content, and analyze our traffic. By clicking "Accept All Cookies," you agree to the storing of cookies on your device. You can manage your cookie preferences at any time by visiting our Cookie Settings.