EA31337 released for free

Brief Story

After 3 years of development of my EA, MQL5 Ltd decided to remove my seller account along with the product right after I've reported scam brokers appearing on their CDN and app with no possibility for negotiation. Therefore I've decided to open source and give my EA for free.

I'm releasing it as it is, without promises and warranty. It is a very complex project and due to serious limitations of the MetaTrader 4/5 platforms which I've learned during development, and due to an unfortunate MQL5 decision, I've decided to not use the platform as the main one and move on to other interesting projects.

To clarify, the free code doesn't mean that scammers (which are everywhere) can steal the code and re-sell it. The code is enough secured to prevent that happen. If you decide to distribute copies of such a program, 'you must pass on to the recipients the same freedoms that you received'. Otherwise, expect some troubles.

Furthermore, I won't take any criticism or get into long opinion-based argues without any claims to support (as I've learned it's a common vibe in trading communities), so don't be surprised when you'd be ignored. If you believe you've found the bug, just raise the issue, and I'll try to help in my spare time. Any contributors are welcomed.

Due to the complexity of the system, don't expect to just work right after placing it on some random account, then being surprised after few weeks it doesn't work as expected without any effort to understand how the EA works. Please read the documentation thoroughly, get familiar with all the params, make some valid backtests, and adjust the params according to your needs.

EA Intro

The EA31337 trading robot is taking the currency trading on the next level by providing coordinated multi-timeframe algorithms which makes the trading 24/7 very simple experience.

EA Features

1. Over 30 optimized strategies to choose from, coordinated and working as one single algorithm to analyze market on multiple timeframes at the same time (M1-M30).
2. Over 30 major technical indicators analyzed in real time (on every single tick).
3. Risk Ratio manages automatically your risk given your current equity/balance and the market conditions.
4. An exceptional number of customizable user parameters (over 300).
5. Task scheduler which queues and re-trades the rejected trades due to high volatility or broker requotes.
6. Trade prioritiser aims to prioritise trading according to high demand or market limits.

EA Basic Params
The EA provides the following basic params:

1. Trade
   1. Max orders, Max orders per type - Sets the limit of total or per-strategy orders to open. Set 0 for auto, so it knows exactly how many trades to open given your balance, a number of trades, Stop Out levels and market conditions (see: Risk Ratio).
   2. Lot size - Sets a lot size of the order to trade. Set 0 for auto calculation based on your equity/balance ratio.
   3. InitNoOfDaysToWarmUp - A Warm-up period in days to reduce lot size. Helps to protect balance against unknown market conditions.
   4. Max take profit/stop loss - Limits when to close the order with profit or loss.
   5. Trend method- Method of calculating the current trend for the trades.
   6. Min pip gap- Min gap between trades per strategy.
   7. Tick ignore method- Sets method for ignoring certain ticks.
   8. CloseOrderAfterXHours - Sets a number of hours after profitable trades are closed (negative number includes non-profitable trades).

   
   

2. Profit and loss
   1. Max Take profit/Max Stop loss - Limits when to close the order with profit or loss (in pips).
   2. TrailingStop/TrailingProfit - Extra trailing stops
   3. TrailingStopAddPerMinute - Decrease trail stop per minute (pip/min)

   
   

1. Risk
   1. Risk margin per order/in total - Limits trade risk in %.

   
   

1. 1. Risk ratio - If set to 0, it automatically calculates your risk given your current equity/balance and market conditions.
   2. Risk margin - Sets the risk of the margin in percentage which you want to risk per each trade. Set 0 for auto.
   3. Min/Max profit factor per strategy- Sets min/max required profit factor of strategy to trade. If the limit is reached, it disables non-successful strategies.

   
   

2. Boosting
   1. Boosting: by trend factor, by the best daily/weekly/monthly strategy- Activates lot size boosting by its profit factor or the most successful strategies over a certain period.

   
   

3. Account conditions - Invokes condition-based user configurable actions on certain account or market conditions (e.g. close profitable orders on high equity or before margin call happens).
4. Strategy-based Signal Methods - Unique filtering system which enhances technical indicators and trades only on successful market conditions.

and many other params (over 300), for more details, check the documentation page .

Documentation

Documentation can be found at https://github.com/EA31337/EA31337/wiki

Project Page

The project page: https://github.com/EA31337/EA31337

Downloads

The compilation process could be a bit challenging, therefore you can find the release files at https://github.com/EA31337/EA31337/releases

Thank you so much for taking time out to do this.

However, the EAs are not working when dragging to chart???

Please refer to the Journal log for details. Make sure you have got an uncompressed version of it (e.g. EA31337-Lite-Full, v1.075 at the time of writing) as GitHub doesn't allow .ex4 release binaries anymore. Feel free to join Telegram or Gitter chat for any trouble, explaining what do you see exactly, so I can help you, as there can be a plenty of reasons.

Thanks for the advanced EA. Trying to figure out why most of the trades are being closed at a small loss. Not so easy because there are so many trailing options.

Backtest version fails to initialize:

This works for me:

Due to limitations of MT4 platform or lack of data, EA automatically disables strategies which are not able to read the data. The EA tries to trade on multiple timeframes at the same time as best as it can, given the limitations of the platform, so it's doing some initial checks and automatically disables strategies which failed to initialize.

So basically:


means that Envelopes technical indicator which trades on M15 is not available during backtesting for some reason. In other words, iEnvelopes() function returned zero, instead of proper data.

This may happen when you don't have HST files for all the timeframes. Sometimes it's because of the MT4 bug, where backtesting on lower timeframes, higher timeframes aren't activated for some reason. But based on the screenshot that all M1, M5, M15, M30, H1, H4 and D1 are on the list, could be a missing HST data for M15 for given period. For easier understanding the code, error number corresponds to line number from the source code file.

You can bypass the validation checks by disabling ValidateSettings param, but the results could be not as expected.

The EA has been optimized for EURUSD only, in unknown market conditions, the default params would need some tweaking.

For easier understanding what causing specific trades to close, hovering on the buy/sell arrows, or checking the comments of trades or the Journal log, should give you the indication which strategy was responsible for the trades. Another way is to run the full backtest and check the report (either in Journal log or text file which has been generated in files/ folder). Then identify the losing strategies and consider tweaking them or disabling. For simplicity, it's also good to backtest few strategies at the same time and choose the best. For optimization tweaks, it's good to focus only on one strategy at the same time.

In the recent version, I've also activated the condition (CloseOrderAfterXHours) which automatically closes trades after 24h. This is due to fact that normal market conditions on the next day are completely different from the technical condition (e.g. daily pivotal points, etc.) which were used to open specific trades so they cannot be taken into the account. In other words, profits from the trades which were opened for few days are merely a luck. Since EA opens the trades based on the technical indicators, the trades need to be closed after 24h to provide reliable trading results especially during backtesting, otherwise, one few lucky trades could pollute optimization results by tweaking the params based on few lucky shots which would never happen again (that's the main problem of many other EAs which are showing great results, but only during backtesting). I hope that makes sense. This is, of course, my point of view, feel free to disable CloseOrderAfterXHours, if you think that's the reason for trades being closed at a small loss. But it could be also something else. I could activate CloseOrderAfterXHours by default only during optimization or backtesting, but not for release files, but small changes after heavy optimization tests could have some drastic effects, so you would be using something that was not tested using the same params.

Found this set file from your repository:
It does not initialize on the 1.075 version but works on the 1.072 version:

This is a very complex and losing EA: http://en.zulutrade.com/trader/29357...0000&ref=57780
That's why it's free...

How do you add the set file to the EA??

You're posting something random without any context or support. I've invested some time into ZuluTrade and I wouldn't recommend to anybody using this service unless you want to blow your account, no matter how good is EA.

The facts are:

- This was barely an experiment in 2015 for the sake of curiosity how it goes using v1.061 and v1.062 unreleased (anywhere) versions. The first official release (~v1.066) was in 2016. And comparing to v1.075 is not appropriate, as it been changed over hundreds of times since then, and you may check the revision history.
- ZuluTrade is not designed to run high-frequency robots due to limits (such as opened orders per day, updates per minute, TP/SL updates per day).
- For example, ZuluTrade limits 30 orders per day for demo accounts and there is some limit on updating the trades as well. And the EA was generating at least hundreds of them per day at that time.
- As far as I remember, there were up to 1000 updates per day, which EA used the limit within first few hours as the early version of EA was generating around few thousand order modification requests per day.

Therefore, due to limits, 80% of the trades were rejected by ZuluTrade system causing it to fail. Even after some tweaks (as I was attempting to create a separate ZuluTrade-friendly version), there is no way that EA, or any other high-frequency can work there. I've invested enough time into their service and after some learning, I have not used it anymore since then.

I"m sorry.. I don't belive EA can work long term..

Unfortunately, some SET files aren't back compatible. Most of the time the default settings included with the release files are the best.

To find the compatible SET files, you may go to EA31337-Lite-Sets repo and dig for some .set files (hitting T on the repo and typing ".set" can give you some list). The compatible version should be visible in the first line of the file after opening it. For SET file for the specific version, you may check the git history. As generating them based on the optimization tests costs a lot. My usual process was to execute hundreds of instances on AWS each testing headlessly certain components of EA, however to generate a reliable SET file for one configuration (pair+spread+year) costs over few thousands of dollars per release version and one tiny modification, mistake or tweak doubles the amount. That's the reason I no longer provide any SET files for the latest versions, as the process is a bit overkill.

You can always compare two files using some diff utility, check the differences and change the settings manually.

If you'd like to check the most recent charts, you can find them at https://github.com/EA31337/EA31337-L.../master/EURUSD
These are the best to my knowledge, however, each version could generate a different result.

To see how the result was generated and against which version, click on the image (at the repo page) and zoom in. Each chart should have the details how the test was performed. These results were generated by fully automated systems, not by manual means. That means there were generated by using repeatable/reproducible environment. Some of them could be out-of-date, so they need some work along with the EA while it evolves.

Hello all,

I think we just need to be appreciative that people come on FF to share hence, should thank people like Kenorb just for sharing.

Please remember those people that are busy making money in Forex will never reveal their skills, knowledge and strategy on sites like FF.

So when somebody does come along to share ideas, whatever they maybe, the least we can do is look into it to see if we can learn from it and help towards our own development/testing/strategy on Forex.

Because of the complexity of the EA it is sensible to back-test on low spread currency pair ranges. So if some of you can do this and report back would be greatly appreciated.

Raf

I didn't say it can work. Depending on the market conditions, it needs the right settings. The EA is fully customizable by the user with over 300 params. The challenge is to configure them right for the given market conditions. In the end, I'm not selling it anymore, and it's up to you if you want to use it or not.

A lot of parameters for optimization.
Tell me please, is there any advice on the initial choice of optimized parameters?

Backtest allways provide very good results. Do you have forward test results? Myfxbook?

Here is my suggestion. Run the backtest with the default settings for some longer selected period. Check the final report at Journey log or generated text file in tester/files/ folder (you can access from: File, Open Data Folder in a top menu bar). The file will given you the profit stats per strategy. Choose the most profitable strategy which you like, then set all Disable all Enable for xxM params apart from the strategy which works the best for you given the dataset. So you can either decide to enable one timeframe for the given strategy or group of timeframes for the strategy.

Then start by optimizing specific single strategy, depending on what's the problem. Describing the optimization process is a topic for a book. But here are some hints:
- If the strategy doesn't trade often, reset filters: 'Open condition' all to 0, re-test, if it's still the issue, reset 'Signal Method' to 0 as well.
- If the strategy trades too often, do the opposite, so optimize 'Signal method' (which acts like the 1st layer of filters), then 'Open condition' (which acts like a 2nd layer of filters).
- When optimizing 'Signal Method' or 'Open condition', check for the ranges which strategy supports. If it's too much, do step on every odd number (like 3, 5 or 7).
- If the amount of trades feels right, but there are some misses, then adjust the standard technical settings such as period, applied price, etc. depending on the params supporting by the strategy.
-If you test one strategy across multiple timeframes, and one timeframe is profitable, other not, try optimizing Period/Deviation ratio. This is kind of multiplier between timeframes.
- If the profit takes of the strategy aren't feeling right, optimize only 'Trail profit method'.
- If the stop losses of the strategy aren't feeling right, optimize only 'Trail stop method'.
- You may optimize both trail profit and stop methods at the same time, that would be an ideal, but it all comes to the power of your machine and size of the dataset.

It would be worth to write some optimization guide, but at the end, it's not a fixed process and there are many variables to take into the account.