forex software

Create and Test Forex Strategies

forex software
Forex Software
Products
Expert Advisor Studio Forex Strategy Builder Professional Express Generator CoinRang Trading Courses Historical Forex Data
Purchase Partners Repository Docs & Guides Forum
You are here: Docs & Guides » Expert Advisor Studio - User Guide » Monte Carlo
eas-guide:monte-carlo

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
eas-guide:monte-carlo [2018/01/09 13:04] – yavoreas-guide:monte-carlo [2021/01/14 06:56] (current) – Ilan Vardy
Line 2: Line 2:
  
  
-Monte Carlo is the best tool for testing the strategy robustness. When you create a strategy, you see its backtest statistics. However, there is a problem - the strategy might be over-optimized (a.k.a curve fitted). The goal of the Monte Carlo tool is to verify that the strategy is not over-optimized. This tool allows you to randomly change the market data, the execution of the strategy and the numeric parameters of the strategy indicators.+Monte Carlo is the best tool for testing the strategy robustness. When you create a strategy, you see its backtest statistics. However, the strategy might be over-optimized (curve fitted). The goal of the Monte Carlo tool is to verify that the strategy is not over-optimized. This tool allows you to apply random changes to the market data, the execution of the strategy and the numeric parameters of the strategy indicators.
  
  
-If you make minor changes to the strategy and its environment, and the strategy continues to have good profits, this means that the strategy has a good chance to make money in the real market.+{{youtube>Wsqh3n6LfqI?medium|}}
  
-On the other hand – if you make some minor changes and the strategy's profits crumble – it means that this strategy is over-optimized and thus, is a bad choice to trade.+If you make minor changes to the strategy and its environment, and the strategy continues to have good profits, this means that the strategy has a good chance of making money on the real market. 
 + 
 +On the other hand – if you make some minor changes and the strategy profits crumble, it means that this strategy is over-optimized and thus, it is a bad choice to trade.
  
  
 <WRAP center round tip 60%> <WRAP center round tip 60%>
-When you run the Monte Carlo tool, you actually do not run a single test but 20 (by default) tests. Each test is random can use one or more simulations.+When you run the Monte Carlo tool, you actually do not run just a single test but 20 (by default) tests. Each test is random. It can use one or more simulations.
 </WRAP> </WRAP>
 +
  
 ===== Simulations ===== ===== Simulations =====
Line 18: Line 21:
 Here you can choose the simulations that will be used in the Monte Carlo tests. You select optional parameters for each simulation from the **Options** tab. Here you can choose the simulations that will be used in the Monte Carlo tests. You select optional parameters for each simulation from the **Options** tab.
  
-{{ :eas-guide:monte-carlo-simulations.jpg |}}+{{:eas-guide:monte-carlo-simulations.jpg?nolink|}}
  
 **Market variations** **Market variations**
  
-    * **Randomize history data** - This will change the bar range of a certain number of bars. It can change the bar’s Low and High values in both directions. This will result in having higher or shorter bars.  +    * **Randomize history data** - The tool will take a certain number of random bars and change their range. It can change the bar's Low and High values in both directions. This will result in having higher or shorter bars.  
-    * **Randomize spread** - If this is enabled EA Studio can set the spread to a different value. It might vary within the borders set in Options.+    * **Randomize spread** - If this option is enabled EA Studio can set the spread to a different value. It might vary within the limits set in Options.
  
  
-**Execution problems** - Those are related to problems when executing the signals sent from the Expert Advisor to the broker.+**Execution problems** - These are related to problems when executing the signals sent from the Expert Advisor to the broker.
  
-    * **Randomize slippage** - Slippage is present on markets, where the prices change fast. It might happen that the price changes by one or two ticks, until the the order gets to the broker and gets executed. If you enable this option Monte Carlo will sometimes add a slippage in a direction contrary to the direction of the position. In the Options you can set how often to use slippage and how much the slippage should be.+    * **Randomize slippage** - Slippage is present on markets, where the prices change quickly. There may be a situation where the price changes by one or two ticks, before the the order gets to the broker and gets executed. If you enable this option, Monte Carlo will sometimes add a slippage in a direction contrary to the direction of the position. In the Options you can set how often to use slippage and how much the slippage should be.
     * **Randomly skip position entry** - Will skip opening some of the positions. This sometimes happens in the real world. It might occur due to problems on the broker's side, Internet connection issues or slowness in the machine executing the Expert Advisor.     * **Randomly skip position entry** - Will skip opening some of the positions. This sometimes happens in the real world. It might occur due to problems on the broker's side, Internet connection issues or slowness in the machine executing the Expert Advisor.
     * **Randomly skip position exit** - Will sometimes skip closing the position. This one of course can be very destructive to the strategy. **Not** opening a position is not so bad since we won't lose any money. But **not closing** a position might result in a loss.      * **Randomly skip position exit** - Will sometimes skip closing the position. This one of course can be very destructive to the strategy. **Not** opening a position is not so bad since we won't lose any money. But **not closing** a position might result in a loss. 
-    * **Randomly close position** - This happens only occasionally. Your broker might close your position for some reason. Of course brokers who close your positions when they find suitable are not brokers you should be trading with.+    * **Randomly close position** - This happens only occasionally. Your broker might close your position for some reason. Of course brokers who close your positions when they decide to are not brokers you should be trading with.
  
  
Line 37: Line 40:
  
  
-    * **Randomize indicator parameters** - EA Studio comes with this option off by default. This is because usually strategies do not change. If enabled EA Studio will choose some of the indicators in the strategy and change their numeric values by a certain percentage (only for the test). If you have an over-optimized strategy, even minor changes in its parameters will make the profits drop significantly or the strategy will even lose money.+    * **Randomize indicator parameters** - EA Studio comes with this option off by default. This is because usually strategies do not change. If enabled EA Studio will choose some of the indicators in the strategy and change their numeric values by a certain percentage (only for the test). If you have an over-optimized strategy, even minor changes in its parameters will make the profits drop significantly or the strategy may even lose money.
     * **Randomize backtest starting bar** - will randomize starting bar of the test. Sometimes starting at a different point in time might bring your account to zero in a very short period.     * **Randomize backtest starting bar** - will randomize starting bar of the test. Sometimes starting at a different point in time might bring your account to zero in a very short period.
  
Line 43: Line 46:
 ===== Simulations chart ===== ===== Simulations chart =====
  
-{{ :eas-guide:monte-carlo-simulations-chart.jpg |}}+{{:eas-guide:monte-carlo-simulations-chart.jpg?nolink|}}
  
 By the default running the Monte Carlo tool will run 20 tests with randomized data. The randomization of the data will be done according to the **Simulations** checklist and the values in the **Options** tab. By the default running the Monte Carlo tool will run 20 tests with randomized data. The randomization of the data will be done according to the **Simulations** checklist and the values in the **Options** tab.
  
-Each of the tests will be drawn on the **Simulations chart** with a colored line. You can see how the lines are grouped and what is the end result of each backtest. This can easily show how robust the strategy was and how destructive for it the randomized environment and data turned out to be.+Each of the tests will be drawn on the **Simulations chart** with a colored line. You can see how the lines are grouped and what the end result is of each backtest. This can easily show how robust the strategy was and how destructive the randomized environment and data turned out to be for it.
  
 ===== Confidence table ===== ===== Confidence table =====
  
-{{ :eas-guide:monte-carlo-confidence-table.jpg |}}+{{:eas-guide:monte-carlo-confidence-table.jpg?nolink|}}
  
 This table shows different statistics from the testing. This table shows different statistics from the testing.
Line 59: Line 62:
 The first column shows the percentage of tests that showed results better than the current row. “Confidence” column shows the probability for the profit to be higher than the “Net Balance” value. The first column shows the percentage of tests that showed results better than the current row. “Confidence” column shows the probability for the profit to be higher than the “Net Balance” value.
  
-For example in the screenshot above, you see the lowest row shows 100% confidence. This "assures" us that if we have traded this strategy, we could expect it to make at least 351 over the given period. This of course if based on 20 random tests and has a great risk of being incorrect as soon as we run the tests or trade the strategy on a live account.+For example in the screenshot above, you see the lowest row shows 100% confidence. This "assures" us that if we have traded this strategy, we could expect it to make at least 351 over the given period. This of course if based on 20 random tests and has a great risk of being incorrect as soon as we run the tests anew or trade the strategy on a live account.
  
-The tool is only named "Confidence Table" to be easily to recognized by traders who used it in other software. However neither here, nor in other software is the Confidence Table about confidence. It's only a tool that groups the results from the **random** tests in the **past** and shows us how many of them succeeded and how much did they succeed. This means that if you rerun the Monte Carlo tests almost certainly won't get the same results. +The tool is only named "Confidence Table" to be easily recognized by traders who used it in other software. However neither here, nor in other software is the Confidence Table about confidence. It's only a tool that groups the results from the **random** tests in the **past** and shows us how many of them succeeded and how much the succeeded. This means that if you rerun the Monte Carlo tests, you almost certainly won't get the same results. 
  
 <WRAP center round important 60%> <WRAP center round important 60%>
Line 69: Line 72:
  
  
-===== Options =====+===== Settings =====
  
-{{ :eas-guide:monte-carlo-options.jpg |}}+{{:eas-guide:monte-carlo-settings-2.png?nolink&800|}}
  
   * **Count of tests** - choose how many tests you want to run.   * **Count of tests** - choose how many tests you want to run.
Line 81: Line 84:
   * **Count of changed bars %** - what percentage of the bars should be changed.   * **Count of changed bars %** - what percentage of the bars should be changed.
   * **Data range change ATR %** - how much the bars should change in percent based on Average True Range.   * **Data range change ATR %** - how much the bars should change in percent based on Average True Range.
-  * **Maximum spread (points)** - what should be the maximum spread when using  **Randomize spread**+  * **Maximum spread (points)** - what should the maximum spread be when using  **Randomize spread**
  
 **Execution problems** **Execution problems**
Line 87: Line 90:
   * **Maximum slippage (points)** - how much can the maximum Random slippage be in points   * **Maximum slippage (points)** - how much can the maximum Random slippage be in points
   * **Skip entry probability %** - For each new entry there will be a probability (default 2%) to skip opening the position.   * **Skip entry probability %** - For each new entry there will be a probability (default 2%) to skip opening the position.
-  * **Skip exit probability %** - For each exit there will be a probability (default 2%) to skip closing the position. +  * **Skip exit probability %** - For each exit there will be a probability (default 2%) to skip exiting the position. 
-  * **Close position probability %** - What is the chance to close a position in each bar.+  * **Close position probability %** - What is the chance of a position being closed in each bar.
  
 **Strategy variations** **Strategy variations**
  
-  * **Indicator change probability %** - chance to change an indicator parameter value. By default only 20% of the parameters will be changed and 80% of the parameters will remain unchanged +  * **Indicator change probability %** - chance of changing an indicator parameter value. By default only 20% of the parameters will be changed and 80% of the parameters will remain unchanged. 
-  * **Indicator max change %** - shows how much the changed parameter’s value can fluctuate. +  * **Indicator max change %** - shows how much a changed parameter's value can deviate from the original value. 
-  * **Minimum deviation range** - by default 20 steps. This setting is meant for indicators that have parameters with very low values where changing the value by 20% won't do much difference. In such cases the **Minimum deviation range** will be used and the indicator will be changed by 20 steps (by default) or however much you set.+  * **Minimum deviation range** - by default 20 steps. This setting is meant for indicators that have parameters with very low values where changing the value by 20% won't make much difference. In these types of cases, the **Minimum deviation range** will be used and the indicator will be changed by 20 steps (by default) or however much you set.
  
  
  
-  * **Reset** - resets all the fields within Simulations and Options to their default values.+  * **Reset** button - resets all the fields within Simulations and Options to their default values.
  
 ===== Validation ===== ===== Validation =====
  
-{{ :eas-guide:monte-carlo-validation.jpg |}}+{{:eas-guide:monte-carlo-validation.jpg?nolink|}}
  
 When running the [[eas-guide:reactor|Reactor]] or [[eas-guide:validator|the Validator]] Monte Carlo testing is present in the Robustness testing section.  When running the [[eas-guide:reactor|Reactor]] or [[eas-guide:validator|the Validator]] Monte Carlo testing is present in the Robustness testing section. 
Line 108: Line 111:
 The **Validated tests** setting uses the **Validation** settings in the Monte Carlo tool here. The **Validated tests** setting uses the **Validation** settings in the Monte Carlo tool here.
  
-{{ :eas-guide:reactor-validator-monte-carlo.jpg |}}+{{:eas-guide:reactor-validator-monte-carlo.jpg?nolink|}}
  
 These are very similar and work in the same way as the general [[eas-guide:acceptance-criteria|Acceptance Criteria]] for the [[eas-guide:collection|collection]] works. These are very similar and work in the same way as the general [[eas-guide:acceptance-criteria|Acceptance Criteria]] for the [[eas-guide:collection|collection]] works.
  
-In the above example a strategy will be considered passing the Monte Carlo validation only if at least 80% of the tests pass.+In the above example a strategy will be considered to have passed the Monte Carlo validation only if at least 80% of the tests pass.
  
 ~~DISQUS~~ ~~DISQUS~~
Copyright © 2006 - 2024, Forex Software Ltd.