Discussion of article "The correct way to choose an Expert Advisor from the Market"

 

New article The correct way to choose an Expert Advisor from the Market has been published:

In this article, we will consider some of the essential points you should pay attention to when purchasing an Expert Advisor. We will also look for ways to increase profit, to spend money wisely, and to earn from this spending. Also, after reading the article, you will see that it is possible to earn even using simple and free products.

Compound Strategy

All waves have been smoothed-out due to counter-waves on other graphs. The same will happen during real trading. Stronger Expert Advisors will support weaker ones, and vice versa, when a drawdown occurs on a stronger one, the weaker one starts to act as an assistant. This keeps alpha and beta within a narrow range. Also, the number of trades has increased to acceptable levels. This is the result of joining only nine individual strategies. But there can be more of them" the more strategies, the smoother the graph, because it always works.

Even with a bunch of free EAs, it is possible to achieve such results, provided that you carefully analyze each individual EA before adding them to the Portfolio. If you want to achieve impressive results, do not be afraid of purchasing paid EAs, as many of them are really worth the money. You can also efficiently combine free and paid EAs to create successful portfolios.


Author: Evgeniy Ilin

 
This is a very good article and the same conclusion I have come to after spending over two years in practice.
 

Bull, but I still won't screen Trending EA's, all the EA's I see are short term high frequency trades, scalping, pure martin additions, and that results in quick death, quick death, quick death.

Eager to recommend a few. I'll be grateful for that!

 

Thank you Evgeniy for sharing this great article. I have a question for you. What's the meaning of these numbers in C1[] in the AUDUSD 60 EA? They look like random, but are very important in the PolinomTrade() function.

double C1[] = { 1,0.822133022277678,0.315994176511425,0.30644298990423,-0.261842065069115,0.158169820998708,-0.051703465015978,0.0137189652920811,0.00882785013014115,0.00130415798969633,-9.61281860378625E-06,2.75142549605097E-06,-1.04660808141669E-06,-8.37363464590955E-07,7.38001346413789E-07,-2.72705244338816E-07,1.01022722238735E-08,3.80702601586341E-09,-4.33358120845453E-10 };//array for coefficients

 
lycenet #:

Thank you Evgeniy for sharing this great article. I have a question for you. What's the meaning of these numbers in C1[] in the AUDUSD 60 EA? They look like random, but are very important in the PolinomTrade() function.

double C1[] = { 1,0.822133022277678,0.315994176511425,0.30644298990423,-0.261842065069115,0.158169820998708,-0.051703465015978,0.0137189652920811,0.00882785013014115,0.00130415798969633,-9.61281860378625E-06,2.75142549605097E-06,-1.04660808141669E-06,-8.37363464590955E-07,7.38001346413789E-07,-2.72705244338816E-07,1.01022722238735E-08,3.80702601586341E-09,-4.33358120845453E-10 };//array for coefficients

Hello, this is really an array selected using random, based on the brute force approach. I have a program that generates these advisors. You can use any tool... One of the variations of machine learning. In short, if I take the program, press 1 button, I’ll wait, I’ll come and I already have ready-made robots .. You can say I can do the robots by weight))
 
In this paragraph: "To calculate the final profit in this formula, the starting balance is deducted from the initial backtest balance. Maximum drawdown, both in equity and in balance, is available in any backtest in MetaTrader 4 and MetaTrader 5. After the expected payoff in points, this variable is as important, as the profit factors, which is also provided in strategy tester reports". In the phrase "this variable is as important", what variable does it mean? ¿ Total Profit?
 
Daniel #:
In this paragraph: "To calculate the final profit in this formula, the starting balance is deducted from the initial backtest balance. Maximum drawdown, both in equity and in balance, is available in any backtest in MetaTrader 4 and MetaTrader 5. After the expected payoff in points, this variable is as important, as the profit factors, which is also provided in strategy tester reports". In the phrase "this variable is as important", what variable does it mean? ¿ Total Profit?

No, i mean "Maximum drawdown, both in equity and in balance". I'm sorry for not answering for a long time, I don't come here often lately

 

It is high time to stop providing EAs for testing only in the tester.

For the tester, you can stitch trades into the bot so that there are far fewer losing trades and show how good it is....

It is better to provide EAs, at least for a demo account with a time limit, for example 1-2 months.

Here is a screenshot from the tester, where you can see how the bot trades before the date of publication in the Market and how it trades after this date:

Who needs, I can provide the investor's password to the account on which such a "Miracle Bot" is trading.... There is just a slow drain going on....

Many sellers will now start shouting that if you provide it for demo account, you can copy trades from demo account to your real one....

Yes, I agree, this option is possible, but also in this case, if your bot is so profitable, what are you afraid of?

A person will put on a demo account, will copy trades to the real, will earn and in 99% of cases (tested), will want to buy your bot, if it is really so good!

It is not difficult at all to specify in the Expert Advisor restrictions only for a demo account and set the date of termination of its work (this I already refer to MQL).

Especially recently there have appeared EAs at a VERY expensive (and even rapidly increasing) price with fashionable descriptions in the style that it is based on Artificial Intelligence or made with the use of ChatGPT.

Don't fall for this scam!!! If the bot is on AI, it will trade only in the tester and only on the trained period. In real trading it will at best trade 50% by 50%, and in the worst case it will be a slow drain....

So if the seller does not provide a version for a demo account - DO NOT BUY FROM THAT SELLER!!!!

Who has nothing to hide - he will give a version for demo account with a time limit and will be able to attract a potential buyer....

 
Oleg Pavlenko #:

It's high time to stop providing advisors for tester-only testing.

For the tester, you can stitch trades into the bot so that there are much less unprofitable ones and show how good it is....

It is better to provide Expert Advisors, at least for a demo account with a time limit, for example 1-2 months.

Here is a screenshot from the tester, where you can see how the bot trades before the date of publication in the Market and how it trades after this date:

Who needs, I can provide the investor's password to the account on which such a "Miracle Bot" is trading.... There is just a slow drain going on....

Many sellers will now start shouting that if you provide it for demo account, you will be able to copy trades from demo account to your real account....

Yes, I agree, this option is possible, but also in this case, if your bot is so profitable, what are you afraid of?

A person will put on a demo account, will copy trades to his real account, will earn and in 99% of cases (tested), will want to buy your bot, if it is really so good!

It is not difficult at all to specify in the Expert Advisor restrictions only for a demo account and to set the date of termination of its work (this I already refer to MQL).

Especially lately there have appeared EAs at a VERY expensive (and even rapidly increasing) price with fashionable descriptions in the style that it is based on Artificial Intelligence or made using ChatGPT.

Don't fall for this scam!!! If the bot is on AI, it will trade only in the tester and only on the trained period. In real trading it will at best trade 50%/50%, and in the worst case it will be a slow drain....

So if the seller does not provide a version for demo account - DO NOT BUY FROM THAT SELLER!!!!

Who has nothing to hide - he will give a version for demo account with a time limit and will be able to attract a potential buyer....

You simply do not have the necessary skills and criteria for evaluation, no proper methodology of testing, selection, or you are just too lazy to look and do not believe that something can be found. Everything you have written is absolutely unnecessary (not necessary). The existing MQL5 functionality is more than enough, especially since there are a lot of goods on the site. If a person needs something, he will find a way to find it. I can do it. About the negative aspects, yes, they are there and everything is true, your knowledge and experience decide here, you will find what you need if you want.

 

A. I would like to draw your attention to the fact that the drawdown on funds for mql5.com signals is displayed only from the date of opening the signal, not from the date of the beginning of trading

And if in the above example it seems to be ok as the whole 2021 signal hangs and there is no data about the drawdown only for the part of 2020, then mostly in the market I observe the situation when the signal shows that trading is going on for 50-100-150 weeks, but the signal itself was loaded a couple of weeks ago and actually Alpha is excellent and the chart is also excellent, although it is not true.


B. do you have any articles on points 17-22 mentioned but not disclosed in this article:

  • Knowledge of basic mani-management mechanisms and their impact on trading
  • Understanding the differences between backtests and real automated trading
  • Understanding the differences between a demo account and a real account
  • Knowledge and understanding of the most popular trading strategies
  • Knowledge of maths and in particular probability theory as it applies to the market
  • Knowing the truth about the market (you could call it experience)
 
Yauheni Shauchenka #:

A. I would like to draw your attention to the fact that the drawdown on funds for mql5.com signals is displayed only from the date of opening the signal, not from the date of the beginning of trading

And if in the above example it seems to be ok as the whole 2021 signal hangs and there is no data about the drawdown only for the part of 2020, then mostly in the market I observe the situation when the signal shows that trading is going on for 50-100-150 weeks, but the signal itself was loaded a couple of weeks ago and actually Alpha is excellent and the chart is also excellent, although it is not true.


B. do you have any articles on points 17-22 mentioned but not disclosed in this article:

  • Knowledge of basic mani-management mechanisms and their impact on trading
  • Understanding the differences between backtests and real automated trading
  • Understanding the differences between a demo account and a real account
  • Knowledge and understanding of the most popular trading strategies
  • Knowledge of maths and in particular probability theory as it applies to the market
  • Knowing the truth about the market (you could call it experience)

There are parts scattered in different articles. It is just knowledge that is complex, which is hard to sew into one article. Only pieces can be placed somewhere... Where it's appropriate. You can ask questions in PM if you have them.