Light martingale expert - gives good results (expert code attached) - page 2

 
Sart:
On the other hand, it is not difficult to trace the change of the trend, but what should I do? Stop martingale, close as it is and start a new trading cycle in another direction, but in that case the losses on the unfinished martingale may occur, and the martingale itself is very dangerous. I had a plan to organise parallel work on the same account and the same EA chart with a different magic number, but the first tests turned out to be very unfavourable......

Yes, I see your point, I too am primarily puzzled by the "off-track" thing. It's hard to disagree with you there. As for suggestions, I am not ready to make them yet, I need time to visually inspect the system, test it and optimise all parameters. That is why rushing is not the best decision here.

Regards, Vyacheslav.

 
iNatlami:

If you need, I can send you the source code for victoria.

Frankly, I'd rather be acquainted with the description of the strategy, and if "mona", provide a link.

Sincerely, Vyacheslav.

 

Good afternoon!

I may have some crazy ideas! But even delusional ideas sometimes produce good results.

On this forum, in the article "Unusual Tactics", Leonid533, suggested a non-standard approach to the subject of reducing risks in adverse conditions. It may help?

Regards!

 
Vyacheslav:
Sart:
This is the point - when the signal changes, the Expert Advisor will stupidly draw a light line until the end of the open trade cycle. Martingale is facilitated in the sense that due to constant modification of orders, we only need for the price to return to our side by at least one level in order to breakeven - then all open orders at TakeProfit will trigger. One of these orders (the last one opened) will be profitable and compensate losses (or even be profitable, depending on parameters settings) on all other orders that were closed in losses.

Questions:

1) Is the number of levels the result of your optimisation for this currency (we are talking about Noveylander)?

2) Is the numeric value of levels also a result of optimization?

As for the first thoughts that came up, I'm expressing, purely intuitively without any practical confirmation. Have you tried to limit the number of levels, say 4 ( the first - with 0,1 lot, the second - 0,2, the third - 0,4 and the fourth - 0,8) and in proportion to the lot size to distribute levels. We should boundaries with the opening price of the first market order and the price at which one of the two furthest lines of the indicator (the Tenkan or the Qun) from the order opening price is located. Thus, all planned levels should be placed between the opening price of the first market order and the price with one of the two furthest lines of the indicator. However, it is important to consider the distance between these prices and plan the placement of levels within the range of 2 to 4 from this size.

Regards, Vyacheslav.

You understand, this is my very first eXpert. I was testing Victoria's eXpert and I did not like the martingale trade cycle entries there - all the solid timeframe indicators starting from 4H are screaming sell, sell, and Victoria starts the buying cycle. Even a simple entry on the Ishimoku gives better results. The idea of the game is quite simple - we should choose such volumes and distances between levels that a pullback at one level would give more profit than total losses on all other orders. To ensure that all orders trigger in this rebound level, we should modify the Take Profit of all levels by the open price value of this rebound level (it is the second-to-last open level) and take the rebound level to Breakeven. The modification is carried out when the next Limit order triggers. That is all. I just borrowed all these figures from Victoria for the first time. I have not optimized anything.

Regards - S.D.
 
iNatlami:
Sart:
The Expert presented in the attached file was written by me based on Victoria.
I had it hanging around the clock for a fortnight on NZDUSD 4H, initial account size 2000.00 . This instrument was chosen because of the small margin.
The average profit is $40 per day.
It works very well on the flat.
This Expert Advisor has the ability to interact with another Expert Advisor using Magic.
It would be good to organize the support by another expert in case of unsuccessful entering of the first expert into a prolonged slight pulling trend.... - can someone give me an idea?
This Expert Advisor uses codes of well-known programmers on this forum - the links to them have been saved.

Regards - S.D.

If you need, I can send you the source code of victoria.
And what I'll see there is new? As Reshetov said, "it's clear as day". And the guys, of course, are cheeky. They don't just sell, they rent it out for a year,
and then you have to pay again. It is unclear what for? Without insurance against entering a prolonged irreversible trend, a normal person wouldn't work with this program.
This Expert Advisor is my first practical experience and it works better than theirs. So with their prof. level (or attitude to clients - they'll cheat...) everything is clear....

And the topic I've started in the hope that someone would share their experience on the interaction between EAs in terms of mutual protection ...
 
Sart:
The point of the game is quite simple - we should choose such volumes and distances between levels that a pullback at one level would give more profit than total losses of all other orders. To ensure that all orders triggered at this level, we modify the Take Profit of all levels by the open price value of this level (it is the second-to-last open level) and reclassify the level to Breakeven. The modification is executed when the next Limit order triggers. That is all.


This is what I wanted to hear. Thank you. As for the past 24 hours of my acquaintance with your system, I can say that after studying and analyzing the code I guessed what currency is needed for this system and maybe I got a confirmation. Of course, it is too early to say for sure, because it is ridiculous in one day trading, but there is such a thing as intuition and some experience. So, pay attention to the yen-dependent crosses. I think the main thing for me is to study the system operation in visual mode on reversals when the signal changes. By the way, from my question about a safety cushion. You have already fixed the second magic number in your code, why not use it immediately following the principle one signal == one magic number. That is, for a signal +1 use one magic number and for a signal -1 use the second, then in the case of a signal in the system will simultaneously work two magic numbers with their levels. Disadvantage perhaps one and very important, you need the appropriate depot, although you can try, especially if you limit the number of levels to 3-4 for each magician.

Regards, Vyacheslav.

 

Among other things, my main concern is whether DC will allow us to trade with this system.

Such an abundance of pending orders, especially those that do not work, is a significant drawback. I will certainly check this question next week on my real account with obligatory visual inspection of the EA.

Regards, Vyacheslav.

 
Vyacheslav:

Among other things, my main concern is whether DC will allow us to trade with this system.

Such an abundance of pending orders, especially those that do not work, is a significant drawback. I will certainly check this question next week on my real account with obligatory visual inspection of the EA.

Regards, Vyacheslav.

Even with my lack of experience I am 100% sure there will be no problems with AC. The program works very delicately with the trade server - for one tick there is only one request to the trade server (if this request is necessary), all other necessary work is postponed until the next tick. As for pending orders, there are only two of them. My understanding is that the setting/deleting of pending orders is done automatically on the trade server, without the involvement of a broker.

However, I must say that the program is absolutely crude in terms of software glitches, and in such an ideological form as it is - even malicious. Not having solved the issue of safety in case of low-return trends I would not put this program on a real account. But, because my rating
as a newbie is still zero, none of the old guard has shown interest in my open topic....

Regards - S.D.
 
Sart:
Without having solved the issue of hedging in case of a low-return trend, I would not put this program on the real account.


For me the most important thing is that the mechanism itself of placing, deleting and modifying, works without fail. And as for entering, I will naturally choose the right time, place and currency to run on real, and only under my visual control. That's all! As for the safety net, the answer must be sought in the reserves of the system itself. And that reserve may well be what I told above in a couple of posts, even without reducing the number of levels.

Regards, Vyacheslav.

 
Good day.
I have a very small experience, but I'm 100% sure there will be no problem with DC. The program works very delicately with the trade server - for one tick there is only one request to the trade server (if this request is necessary), all the other necessary work is postponed until the next tick. As for pending orders, there are only two. As I understand it, the setting/deleting of pending orders is done automatically on the trade server, without the involvement of a broker.

The thing is that some brokerage companies have restrictions in terms of trading conditions just for the number of orders per day (brokerage companies do not like "graders").
This point can be avoided by introducing "virtual orders", but even with this approach we have problems. The dependence on the quality of communication, and on the broker who executes these orders at the moment of switching from a virtual to a real state for the order, sharply increases.

.
As a matter of fact, I opened the thread in the hope that someone would give me a tip. ... But since my rating
as a newbie is still zero, no one from the old guard has shown any interest in the topic I opened....

The matter is not in the rating but in the fact that we are a bit tired of martingale. Many variations, but the essence is ultimately the same. I have never suggested any more or less practical way out of the situation of a set of orders with a negative balance...
Nevertheless, I find the theme interesting since using this technique in a more complex system looks very tempting.
+ possibility to go further and try Portfolio trading to reduce risks.

Regards, Alexander.
Reason: