For those who are convinced that all EAs with a martin are losing out. - page 6

 
If the EA is not earning a fixed lot, then no amount of position size management, including MM, will help.
 

;))) And martingale again...

http://procapital.ru/showthread.php?t=10728&page=28&p=304073&viewfull=1#post304073

There are several pages of different options discussed here. starting with #418

Maybe you'll find it useful... ;)

 
avtomat:

or the fall will be much bigger and more rapid.
The fall is bigger in relation to what? If compared to the variant without reinvestment, it is not clear, the deposit is growing there and does not fall in any way during the whole quote history.
 
Yur, why did you open the topic: to demonstrate/promote the tester-state or to share something useful. If you want to share something, then share it. We already know that there is no such thing as a perpetual motion machine. Good luck. Personally, I'm interested in the logic of opening/closing positions
 

As for Expert Advisors, I use them for a long time. I totally agree with the fact that all EAs will fail sooner or later.
However, I prefer to draw a picture in which the EA will 100% fail. And look for it in the "past" on the pair and settings where you want to use the EA.
I look for a year. If there are no glitches (the tester may do the same, for me the drawing itself is important), then I trust this EA for a month.

In any case, you can't just leave any algorithm unattended. Recently, my Expert Advisor from Invest-systems developed on EUR-EUR pair with maximal security I lost 30$ in 3 days, and I have been working normally for half a year.

As for the opening/closing algorithm, the advantage of the Expert Advisor is that it continuously monitors the market, but it lacks intelligence. Therefore, the simpler the algorithm, the better. I wish the expected payoff would be greater than zero)

 
khorosh:
Is the drop more in relation to what? If compared to the variant without reinvestment, it is unclear, the deposit is growing and does not fall in any way throughout the quotes history.


Everything depends on the reinvestment coefficient and we should also know how to select it: the larger the coefficient is, the higher is the probability of quotes; the lower the coefficient is, the closer the system is to the variant with a fixed lot and therefore the lower is the effect of reinvestment itself. This is not the most difficult task, but it is not trivial either.

Generally, with drawdowns like the one in the first post, I wouldn't risk reinvesting more than 5-7%.

 
moskitman: And now that we are talking about 31% drawdown, it will not take long for equity to hit zero.

Well, not 31, but more, see the picture in the first post. The equity drawdown is right there:

Relative drawdown 58.29% (7166.17).

2 khorosh: This all looks depressing, you have been stagnating in the same place for several years already. Don`t you have anything new?

OK, let's say 31% since 2008. And you seriously believe that it will always be less?

You seem to be an adult, you know what equity is and what balance is, and yet you keep banging your head against the wall...

 
Mathemat:

OK, let's say 31% since 2008. And you seriously think it will always be less?

... But you are still banging your head against the wall...

1. I don't, I allow rare cases of depo draining, but I hope to earn more than the amount drained between drains.

2. I will not go back to the drawing board, I will not go back to the drawing board again, I will not go back to the drawing board again.

My aim is not to try this version but to prove that EAs with martin may work without losing money on the entire quote history. I have 2 more profitable versions on the real account.

 
khorosh: 1. I do not count, I allow a rare case of losing a depot, but I hope to earn more than the amount I have lost.

This is understandable, you have said it before. It makes sense, but you have to be clear about it at once - not to admit it on the tenth page of a topic somewhere.

The test results are just a proof of the fact that Expert Advisors can work without losing money over the entire quote history.

Well, it's not so much to rush, but it's good to reconsider previous approaches, too.

I am not going to use this version for real, and the test results are just a proof that EAs with martin can work without losing on the whole quotes history. There are 2 more profitable versions on the real account.

So show me the more profitable ones, what is stopping you? You don't need to show the backtest, just show us the real quotes history.

I really want to know what you want to achieve by showing beautiful pictures with martin for the second time, don't you? General and unconditional recognition of the martin - or something more substantial for yourself?

 
Mathemat:

That's understandable, you've said it before. It makes sense in principle, but it should always be said explicitly and immediately - not admitted to somewhere on page ten of the thread.

Well, it's not so much to rush, but it's good to reconsider previous approaches, too.

Well, show the more profitable ones, what's stopping you? Don't do the backtest, just show me the real story.

I really want to know what you want to achieve by showing beautiful pictures with martin for the second time, don't you? General and unconditional recognition of the martin - or something more substantial for yourself?

1. I told you, that is why I decided that it makes no sense to repeat myself, even on 1 or 10 pages.

2. I review approaches, but my Expert Advisors always contain some Martin. 3.

I will not show them. 3. One of my EAs has been working since last year, but I traded manually on the same account, and I am not fond of posting statements from real account.

4. 4. If you have read carefully, it was clearly enough said " .I brought the test results simply as a proof that EAs with a Martin may work without losing money on the entire quote history". I did not want to do anything for myself, I just decided to post it to expand my knowledge about EAs capabilities with martin. I thought it would be interesting for those who deal with this subject. If I wanted to share my EA with some other EAs, I would not expect them tobe interested in this kind of information.

Reason: