Writing an effective advisor - page 37

New comment
 
Vitaly Murlenko #:
This is the second time I've read here that trades are oversubscribed. :) Hilarious. Let me explain.

If the SL is clearly set, is not moved further from the price, and its value can be compensated for less than 10 average trade results - then we cannot talk about "over-exposure". If there is no SL or its value is more than 10 average trade results - this is called "over-exposure".

Another indication of oversubscription is a significant distance of the equity graph from the balance graph.

 
Vitaly Murlenko a group of orders was working, it was necessary to close this group in time. I waited until the group would go to the profit and closed the positions with the only objective: to reduce the load on the deposit. Even closing a group in the red is good here. If the trend is correctly predicted, the next groups will eliminate all the drawdown. If there are too many orders in a group, we can make a mistake on the horizontal trend in case it is not closed (for example, if we wait for the trend to continue). The deeper into the trend in this case (sideways), the more orders we have. And as a consequence, when the counter-order wall of the corridor is touched, the bigger the drawdown. This may lead to the possibility of reaching the Frosty Kohl's. Therefore, it would be better to close the oversized group as soon as possible. This will increase the deposit and reduce the risks. The fact that we have to get a new group of orders again should not be a problem. If the main trend is calculated correctly, the group will profit anyway. It is better not to be greedy.
How do you determine the trend?
 
The classic: 2 highs + 2 lows.
 
Georgiy Merts #:

If the SL is clearly set, is not moved further away from the price, and its value can be compensated for less than 10 average trade results - then there can be no "over-exposure". If there is no SL or its value is more than 10 average trade results - this is called "over-exposure".

Another indication of oversubscription is a significant distance of the equity graph from the balance graph.

There is no overexposure in the algorithm. Unless the straight line is reversed.
 
Vitaly Muzichenko #:

That's right, it's a way of showing something you're selling.

Nothing was being sold here, it was looking for like-minded people.

A like-minded person is someone who works in parallel and sees results, rather than waiting for something ready-made with all the proofs on a platter.

Do you feel the difference?

Well, there are people who are less mercenary than you. They use signals to demonstrate their ideas. They use signals as an argument. Do you feel the difference?

 
a007 #:

Well, there are people who are less mercenary than you. Demonstrate their ideas with a signal. They use signals as an argument. Can you tell the difference?

An idea is a thought, an intention, a plan.

Explain to me how you can demonstrate something that has not yet been created.

And even if it were, it is not the subject of the topic, we are looking for something else, not a way to sell it.

 
Uladzimir Izerski #:

The sequence of ticks is random))

The sequence of bars is also random. But at a certain stretch we see this:

77

It's the same with ticks. It is the robot's job to identify such a section and make a buck.

 
Vitaly Muzichenko #:

An idea - a thought, an intention, a plan.

Explain to me how you can demonstrate something that has not yet been created.

Besides, even if it was, that's not what the topic is about, it's about something else, not a way to sell it.

Your idea, your plan, your intention existed before you wrote your post?

But why are you again getting attached to the man? This topic is about something else!

 
zvezdocheet #:

Did your thought, idea, plan, intention exist before you wrote your post?

However, why are you picking on the man again? That is not what we are discussing in this thread!

Of course it is!

Forum on trading, automated trading systems and testing of trading strategies

How to write an efficient Expert Advisor

Vitaly Murlenko, 2021.11.15 15:55

I have always had the idea of creating an efficient Expert Advis or, which would bring real profit. The idea is the following: I start it and give the reasons why, then I make a start code, explain all the nuances that need to be known, you test and watch the robot, and if you find any flaws, you tell me. We fix it, test it again, search... And so on, until we find the "non-clearer".

I will name my Expert Advisor "Binarnic". The file name will look as follows: Binarnic_1.0.

I want to stress this point and tell you why: since no one is allowed to make changes in my EA's code, I want my code versions not to be mixed up with versions of other forum users. For example, a user named Ivan wants to change my code. He makes changes, but calls his file Binarnic_Ivan_1.1. With this approach, it becomes clear that he took my file Binarnic_1.0, made changes to it and the result is the next version of the file - " Binarnic_Ivan_1.1".

I think everyone would agree with these simple rules, for they are reasonable.


 
To write an effective Expert Advisor, you need to have a deep knowledge of the financial markets. To make money in the market, you need not programming skills, but a clear, tested on a demo account TS (not necessarily algorithmic, you can also trade manually). The main thing is the right idea, which you can then program after checking. It is even needed, because trading should be done with the help of robots. But the main thing is the right idea, based on scientific knowledge.
1...303132333435363738394041
New comment
Reason: