DNYL . SPACE

Which 10 currency pairs would you recommend for 10 rounds?

When selecting currency pairs for 10 rounds of testing, consider pairs with diverse volatility and market behavior. This allows for a comprehensive and effective strategy across different markets. Examples of 10 recommended currency pairs include:

EUR/USD – A popular, highly liquid pair
GBP/USD – High volatility
USD/JPY – A pair correlated with the stock market
AUD/USD – Sensitive to commodities
USD/CAD – Linked to oil
EUR/JPY – Cross-regional correlation
GBP/JPY – High volatility
EUR/GBP – A European pair
NZD/USD – Correlated with commodities, similar to AUD
USD/CHF – Often used as a safe haven asset during high market risk periods.

You can adapt this list of currency pairs to suit your needs and the market behaviors you’re interested in.

Monte Carlo process results

For the Monte Carlo process results that did not pass the walk-forward matrix in the Ending section, should we re-apply them to Genetic evo or not?

If a strategy passes the Monte Carlo process but fails the Walk-Forward Matrix (WFM) in the Ending section, it may indicate that the strategy is vulnerable to real market changes. Although it may be resilient to random sampling (Monte Carlo), it may not adapt well to changing times (Walk-Forward Matrix). Considering reapplying this strategy in a Genetic Evolution:

Benefits of reuse:

You may find good elements within the strategy, but they require adjustment to fit changing market conditions. Using Genetic Evolution may help find parameter sets that are more flexible and adaptable.

Drawbacks of reuse:

If the strategy completely fails the WFM, it may indicate a fundamental problem within the strategy, which adjustments may not resolve. It may be a waste of time to attempt to refine a strategy likely to fail in the long run.

Recommended Approach:

Analyze the cause: Before reapplying the strategy to Genetic Evolution, analyze why it failed the WFM, such as overfitting or inability to adapt to changing markets.
Apply only parts: If parts of the strategy show promise, these parts can be incorporated or improved upon in a new Genetic Evolution process.
Create new conditions: In the new Genetic Evolution round, new conditions or constraints could be added to ensure the strategy passes the WFM or becomes more resilient.

Re-applying a strategy that failed the WFM to Genetic Evolution can be a good approach if proper analysis and improvement are undertaken. This can give the strategy a better chance to adapt and succeed in a wider range of market conditions.

A. Building…

Create 10 strategies using Genetic Evolution.
1.1 Breakout Strategy
1.2 Day Trading Strategy
1.3 Grid Trading
1.4 Multi-Timeframe Divergence Strategy
1.5 News Trading
1.6 Oscillator Reversion and Momentum Burst Strategy
1.7 Range Trading Strategy
1.8 Scalping Strategy
1.9 Swing Trading
1.10 Trend Trading Strategy

Use TF H1/H4 – Opened Timeframe – Ranking Profit > 1.4 Ret/DD > 4 Min trade per month > 2 – No Cross check / Robustness test
** Market Entry Method: Enter at market / Reverse / Stop / Limit pending
*** Filter each step by 10 entries per market entry method, totaling 40 in one main group, and a total of 400.
**** Enable Exit types: MoveSL2BE / SL2BE add pips / Profit target / Stop loss / Trailing stop / Trailing activation / Exit rules
***** Trading option: Friday Exit 23:00
****** Money management: Risk fixed % of account

B. Melting…

Backtest all 400 results against 10 currency pairs, separating the timeframe into 5 TFs.
2.1 H1/H4 – M30/H4 – M15/H1 – M5/H1 – M15/M30
2.2 Aggregate the results and perform Automatic Retest by changing the Precision to 1-minute data tick simulation, but maintain the original trading settings.
2.3 Add Monte Carlo trades manipulation and Sequential optimization testing.
2.4 Save the results to the @Blender folder.
2.5 Perform another Automatic Retest using only the Walk-Forward Matrix.
2.6 Save the results to the @Melted folder.

C. Ending…

Load all files from @Blender into the Initial Population to create a new set of EA, repeating the process 10 times (10 currency pairs).
3.1 Create 3 new steps: Build Martingale / ATM / Everyday Exit
3.2 Automatic Retest with the original trading settings at 3 timeframes (H1/H4 – M30/H4 – M15/H1), using 2 Monte Carlo factors (trades manipulation and retest methods) and Sequential optimization.
3.3 Another Automatic Retest with 1-minute data tick simulation, using 2 Monte Carlo factors and Sequential optimization.
3.4 Another Automatic Retest using the Walk-Forward Matrix.

Considering removing unnecessary parts or reducing certain processes can help improve efficiency and reduce processing time. Here are some suggestions for possible cuts or reductions:

1. Number of strategies in the Building phase:

Currently: You create 20 different strategies.
Suggestion: Try reducing the number of strategies generated in each cycle, starting by selecting the most likely successful strategies. For example, choose 10 strategies that are popular in the current market conditions or have been tested and are more likely to yield good results.

2. Number of currency pairs in the Melting phase:

Currently: You test all 30 currency pairs.
Suggestion: Reduce the number of currency pairs in the initial test rounds. You may select major currency pairs with high liquidity, such as EUR/USD, GBP/USD, etc., which will reduce the overall testing time.

3. Timeframes in testing:

Currently: Testing is done on 10 timeframes.
Suggestion: Try reducing the number of timeframes used in the test, focusing on only relevant timeframes to your developed strategies, such as H1/H4 or M15/H1, which are more likely to give accurate results.

4. Automatic Retest phase:

Currently: You perform Automatic Retest multiple times.
Suggestion: Reduce the number of Automatic Retest cycles. Perform it only once after the Monte Carlo test to save time.

5. Use of Monte Carlo:

Currently: Using multiple Monte Carlo methods.
Suggestion: Consider using only one method that gives the most appropriate result or use it only in the final round of testing to save time.

6. Walk-Forward Matrix:

Currently: You use the Walk-Forward Matrix in multiple steps.
Suggestion: Reduce the use of the Walk-Forward Matrix in some steps and use it only in the final step to confirm the best results.

7. Adding new steps in the Ending phase:

Currently: You add new steps in each cycle.

Suggestion: Try limiting the addition of new steps in certain cycles, or only when unsatisfactory results are found from previous testing.
These reductions will help reduce overall processing time and make your EA creation more efficient without sacrificing the quality of the results obtained.

Your current step-by-step process has both advantages and disadvantages to consider:

Advantages

Strategy diversity: Creating 20 strategies allows for greater trading diversity, increasing the chance of finding effective strategies in different market conditions.
Comprehensive testing: Backtesting with multiple currency pairs and different timeframes provides a comprehensive view of the strategy’s performance in various conditions, making the results more reliable.
Monte Carlo and Walk-Forward Testing: Increases confidence in the robustness of the strategy, as these tests simulate unexpected situations and market fluctuations.

Disadvantages

Time-consuming: Complex steps like backtesting and parameter adjustments for each strategy require significant processing time, which may hinder strategy development and improvement within limited time.
High resource consumption: Testing multiple iterations and multiple timeframes requires high computer resources, potentially causing delays on lower-performance machines.
Complexity of the process: The diverse and complex steps can make result analysis challenging and may increase the possibility of errors during processing.

Evaluation

Positive results if you can efficiently manage processing time and resources, and the results from comprehensive testing can be used to improve strategy effectiveness.
Negative aspects if the long processing time and high resource requirements prevent you from making timely improvements or developing, or causing the strategy to be delayed in responding to changing market conditions.

You can assess which steps take too long and consider reducing or simplifying those steps to improve overall efficiency.

These are the steps for creating our EA, based on our conversation. Please clarify and analyze which parts take too long. My complete EA creation steps:

A. Building..

1. Creating 20 strategies using Genetic Evolution.
1.1 Breakout Strategy
1.2 Carry Trade Strategy
1.3 Cycle-Based Trend Reentry Strategy
1.4 Day Trading Strategy
1.5 Dynamic Momentum and Price Rejection Strategy
1.6 Grid Trading
1.7 Market Chaos and Divergence Strategy
1.8 Mean Reversion on Trend Weakness
1.9 Multi-Timeframe Divergence Strategy
1.10 News Trading
1.11 Oscillator Reversion and Momentum Burst Strategy
1.12 Position Trading
1.13 Price Action Trading
1.14 Range Trading Strategy
1.15 Retracement Trading
1.16 Scalping Strategy
1.17 Sentiment and Momentum Reversal Strategy
1.18 Swing Trading
1.19 Trend Channel Momentum Strategy
1.20 Trend Trading Strategy

* Using TF H1/H4 – Opened Timeframe – Ranking Profit > 1.4 Ret/DD >4 Min trade per month > 2 – No Cross check / Rubustness test
** Entering always includes Enter at market / Reverse / Stop / Limit pending
*** Filtering each step with 25 entries per market entry type, totaling 100 entries per main category, and a total of 2000 entries.
**** Enabling Exit types: MoveSL2BE / SL2BE add pips / Profit target / Stop loss / Trailing stop / Trailing activation / Exit rules
***** Trading option: Friday Exit 23:00
****** Money management: Risk fixed % of account

A. Building Improved

• Create 20 strategies using Genetic Evolution, which takes time for filtering and refining.
• Filtering 25 strategies for market entry and multiple exit options increases processing time.
• Setting money management complicates calculations.

B. Melting..

2. Testing all 2000 results backtested with all 30 currency pairs, separating timeframes into 10 TFs.
2.1 H1/H4 – M30/H4 – M15/H4 – M5/H4 – M5/H1 – M15/H1 – M30/H1 – M15/M30 – M5/M30 – M5/M15
2.2 Combining the above results for Automatic Retest by changing Precision to 1-minute data tick simulate, but adhering to the original trading settings.
2.2 Adding Monte Carlo trades manipulation – Monte Carlo retest methods and Sequential optimization by changing values in the respective EAs.
2.3 Saving successful files in Folder @Blender
2.4 Automatic retest using only Walk-Forward Matrix.
2.5 Saving successful files in @Melted

B. Melting Improved

• Backtesting 30 currency pairs and various timeframes is time-consuming, especially with trading value changes and Monte Carlo testing.
• Automatic retests with 1-minute tick data and Walk-Forward Matrix require high computational power.

C. Ending..

3. Loading all files in @Blender into the Initial Population to create entirely new EAs, repeating 30 times (30 currency pairs).
3.1 Creating 5 new steps: Build Martingale / ATM / Everyday Exit / Risk % account, and using all Exit types.
3.2 Automatic retest with the original trading settings for 5 timeframes (H1/H4 – m30/h4 – m15/h4 – m5/h4 – m5/h1 – m15/h1 – m30/h1). Using the original settings for Opened timeframe. Using Monte Carlo, both (trades manipulation and retest methods), and Sequential optimization. Adding strategies that passed.
3.3 Automatic retest for all, another round, with 1-minute data tick simulation, with both Monte Carlo (2 versions) and Sequential optimize as before.
3.4 Automatic retest for all, another round, using walk-forward matrix.

C. Ending Improved

• Loading files in @Blender to create 30 new EAs is time-intensive, particularly with additional strategies.
• Multiple automatic retests using Monte Carlo and Sequential Optimization increase processing time.

Conclusion:
Suggestions include reducing strategies and pairs tested, considering partial parameter adjustments, and using more efficient processing tools to save time.