Making an EA that work

New comment
 

I created several EA some of them where making good profit  for few months, some of them gave me millions of $ with only 100$ starting but I had to enter at a certain time and date like 2017/01/06. I'm still figuring out the basic or a method to create an EA that will go up consistently in the backtesting. A lot of EA on the market are showing positive gain for several years and signal is showing very good result. How do they the idea or the concept to make somewhat successful EA. I'm using EAbuilder which is a visual scripting tool to make EA. Its very powerful but I'm not sure if it can compete with Neural network EA and all those new things... 

<Deleted>

- Is it everyone using BB?

- Is it you need a lot of indicators mix up?

- Does people who makes a good EA have a way to create EAs?


I know that making an EA based on the basis of a certain regularity will never work long term. But making an EA based on irregularity will not make something good long term. So what is the path?

I'm so confused on how to create an EA that will works for several years.

Neural Networks: From Theory to Practice
Neural Networks: From Theory to Practice
  • www.mql5.com
Nowadays, every trader must have heard of neural networks and knows how cool it is to use them. The majority believes that those who can deal with neural networks are some kind of superhuman. In this article, I will try to explain to you the neural network architecture, describe its applications and show examples of practical use.
 
carl_carl101:

I created several EA some of them where making good profit  for few months, some of them gave me millions of $ with only 100$ starting but I had to enter at a certain time and date like 2017/01/06. I'm still figuring out the basic or a method to create an EA that will go up consistently in the backtesting. A lot of EA on the market are showing positive gain for several years and signal is showing very good result. How do they the idea or the concept to make somewhat successful EA. I'm using EAbuilder which is a visual scripting tool to make EA. Its very powerful but I'm not sure if it can compete with Neural network EA and all those new things... .
..
- Is it everyone using BB?
- Is it you need a lot of indicators mix up?
- Does people who makes a good EA have a way to create EAs?
I know that making an EA based on the basis of a certain regularity will never work long term. But making an EA based on irregularity will not make something good long term. So what is the path.
I'm so confused on how to create an EA that will works for several years.

Hello,

I understand the difficulty you mean,
- the good news is you are not alone.
because .. millions of people also experience such difficulties.

Just as on a passenger plane to be controlled by at least 2 pilots (pilot + co-pilot) and a ground crews,
maybe you should not work alone but form a teamwork to design
and build the best EA that can be created.

Otherwise you can try https://www.mql5.com/en/job

Good luck.

Kind regards.

Trading applications for MetaTrader 5 to order
Trading applications for MetaTrader 5 to order
  • 2022.08.23
  • www.mql5.com
The largest freelance service with MQL5 application developers
 
carl_carl101:

I created several EA some of them where making good profit  for few months, some of them gave me millions of $ with only 100$ starting but I had to enter at a certain time and date like 2017/01/06. I'm still figuring out the basic or a method to create an EA that will go up consistently in the backtesting. A lot of EA on the market are showing positive gain for several years and signal is showing very good result. How do they the idea or the concept to make somewhat successful EA. I'm using EAbuilder which is a visual scripting tool to make EA. Its very powerful but I'm not sure if it can compete with Neural network EA and all those new things... 

<Deleted>

- Is it everyone using BB?

- Is it you need a lot of indicators mix up?

- Does people who makes a good EA have a way to create EAs?


I know that making an EA based on the basis of a certain regularity will never work long term. But making an EA based on irregularity will not make something good long term. So what is the path?

I'm so confused on how to create an EA that will works for several years.

Some tips and tricks I got while coding EAs, some might not agree but here are my suggestions:


- BBs are very important, as trading during a range isn't really good usually, so I use them as a confirmation of trend / volatility.

- Not specifically, however, something important that I noticed: I used to make my entries based on technical indicators, but it wasn't really working. I noticed price action (candlestick patterns / S&R) seems to provide better results. Therefore, I base myself on price action to open, and indicators to confirm that I should open a position, and not the opposite. (indicators to open, and PA to confirm).

- When you create your EA, if you focus on backtesting only a certain period of time, and try to change / add indicators to fit that period of time, you will notice it won't work on the long run. The more indicator you use, the more likely you are just trying to fix it for a little and have conditions that won't work.

- Also check that great article to check for risk management: https://www.mql5.com/en/articles/4233

How to reduce trader's risks
How to reduce trader's risks
  • www.mql5.com
Trading in financial markets is associated with a whole range of risks that should be taken into account in the algorithms of trading systems. Reducing such risks is the most important task to make a profit when trading.
 
Yohana Parmi #:

Hello,

I understand the difficulty you mean,
- the good news is you are not alone.
because .. millions of people also experience such difficulties.

Just as on a passenger plane to be controlled by at least 2 pilots (pilot + co-pilot) and a ground crews,
maybe you should not work alone but form a teamwork to design
and build the best EA that can be created.

Otherwise you can try https://www.mql5.com/en/job

Good luck.

Kind regards.

Ohh yes, this could be a good place to get different thought. Thanks!

 
Zaky Hamdoun #:

Some tips and tricks I got while coding EAs, some might not agree but here are my suggestions:


- BBs are very important, as trading during a range isn't really good usually, so I use them as a confirmation of trend / volatility.

- Not specifically, however, something important that I noticed: I used to make my entries based on technical indicators, but it wasn't really working. I noticed price action (candlestick patterns / S&R) seems to provide better results. Therefore, I base myself on price action to open, and indicators to confirm that I should open a position, and not the opposite. (indicators to open, and PA to confirm).

- When you create your EA, if you focus on backtesting only a certain period of time, and try to change / add indicators to fit that period of time, you will notice it won't work on the long run. The more indicator you use, the more likely you are just trying to fix it for a little and have conditions that won't work.

- Also check that great article to check for risk management: https://www.mql5.com/en/articles/4233

Thanks for the article, I will check it out in details.

I see you have released some products and you have a signal that look like its working fine. The pictures showing gains from your EA are what I would like to reproduce. I understand that some strategies dies at one point and having a profitable EA in real life for several years its rare. If I could make an EA that is profitable for 1 years then it would be good.


When you work on your EA do you have certain way of making them? Like do you always use trail stop lost for example? How much time does it take you to make an EA that you are willing to sell? Could this take week? or you have to work several months?

 
carl_carl101 #:

Thanks for the article, I will check it out in details.

I see you have released some products and you have a signal that look like its working fine. The pictures showing gains from your EA are what I would like to reproduce. I understand that some strategies dies at one point and having a profitable EA in real life for several years its rare. If I could make an EA that is profitable for 1 years then it would be good.


When you work on your EA do you have certain way of making them? Like do you always use trail stop lost for example? How much time does it take you to make an EA that you are willing to sell? Could this take week? or you have to work several months?

It’s complicated, I don’t really use trailings, even though the option is available is some of my EAs.
However, keep in mind some EAs don’t work during certain situations. My EA for instance, should be turned off during news events or gradual bollinger bands expansions. But that’s totally fine.

Usually, I can pull out a good working EA in about a week, the hard part is finding a viable strategy.

I sometimes create first as an indicator using EAbuilder or something, then code the full EA myself, with parts from EABuilder
 
Zaky Hamdoun #:
It’s complicated, I don’t really use trailings, even though the option is available is some of my EAs.
However, keep in mind some EAs don’t work during certain situations. My Stochastica martingale trader for instance, should be turned off during news events or gradual bollinger bands expansions. But that’s totally fine.

Usually, I can pull out a good working EA in about a week, the hard part is finding a viable strategy.

I sometimes create first as an indicator using EAbuilder or something, then code the full EA myself, with parts from EABuilder

Do you always make an EA on EAbuilder then code it yourself after? or its possible to make an EA only on EAbuilder?

 
carl_carl101 #: Do you always make an EA on EAbuilder then code it yourself after? or its possible to make an EA only on EAbuilder?

EABuilder generates very badly coded results with many bugs. So I would not recommend you use such code for live trading. I would not even trust it for back-testing either.

I recommend you learn to code an EA yourself. Doing so will also educate you on the many intricacies of trading which you may not even be aware of, that affects your trading.

 
Fernando Carreiro #:

EABuilder generates very badly coded results with many bugs. So I would not recommend you use such code for live trading. I would not even trust it for back-testing either.

I recommend you learn to code an EA yourself. Doing so will also educate you on the many intricacies of trading which you may not even be aware of, that affects your trading.

What kind of bugs you can encounter? Because for me it has worked well. What kind of complicated things that I could not be aware of?


Thanks for your reply

 
carl_carl101 #: What kind of bugs you can encounter? Because for me it has worked well.

Forum on trading, automated trading systems and testing trading strategies

About Shifting Value Adjustment

William Roeder, 2022.07.10 13:43

EA builder, EA Builder Pro, EATree, Etasoft forex generator, Forex Strategy Builder, ForexEAdvisor STRATEGY BUILDER, ForexRobotAcademy.com, FX EA Builder, fxDreema, Forex Generator, FxPro, Molanis, Octa-FX Meta Editor, Online Forex Expert Advisor Generator, Strategy Builder FX, Strategy Quant, Visual Trader Studio, MQL5 Wizard, etc., are all the same. You will get something quick, but then you will spend a much longer time trying to get it right, than if you learned the language up front, and then just wrote it.

  1. Since you haven't learned MQL4/5, therefor there is no common language for us to communicate.
    If we tell you what you need, you can't code it.
    If we give you the code, you don't know how to integrate it into yours.
    We are willing to HELP you when you post your attempt (using Code button) and state the nature of your problem, but we are not going to debug your hundreds of lines of code. You are essentially going to be on your own.

  2. EA builder makes bad code counting up while closing multiple orders.
    EA builder makes bad code Bars is unreliable (Max bars in chart), volume is unreliable (miss ticks.) Always use time.
    EA builder makes bad code, not adjusting for 4/5 digit brokers, TP/SL and slippage.
    EA builder makes bad code, not adjusting for ECN brokers. pre-Build 500)
    EA builder makes bad code, not checking return codes.

  3. EATree uses objects on chart to save values — not persistent storage (files or GV+Flush.) No recovery (crash/power failure.)

  4. FX EA Builder makes bad code, not checking return codes.
    FX EA Builder makes bad code, loosing open tickets on terminal restart. No recovery (crash/power failure.)
    FX EA Builder makes bad code, not adjusting stops for the spread.
    FX EA Builder makes bad code, using OrdersTotal directly.

  5. FOREXEADVISOR STRATEGY BUILDER makes bad code, non-updateing global variables.
    FOREXEADVISOR STRATEGY BUILDER makes bad code, compilation errors.
    FOREXEADVISOR STRATEGY BUILDER makes bad code, not checking return codes.

Learn to code it, or pay someone (Freelance) someone to code it.
          Hiring to write script - General - MQL5 programming forum #1 (2019)


 
carl_carl101 #: What kind of complicated things that I could not be aware of?

You have only to go through the many posts on the forum to see the large amount of doubts people have for all sorts of things.

From incorrectly calculating risk, misunderstanding about leverage and margin, not knowing which prices are used for entry and exit of different orders types, and many more things newbie traders are just not knowledgeable enough about.

It takes may years (not months) to fully grasp all those aspects and even then you still keep learning new things when you least expect it.

12
New comment