Is there any profit in forex? - page 3

 
Александр Князев:
Strangely enough after writing this thread, my EA in the tester (right after changing "<" sign to the opposite) showed 44% profit in last 12 runs with drawdown less than 10%. This is a grid martin putting out pending orders with complex step (about 50 pips). However it differs from usual martin (apart from the step) in that it opens orders in the opposite direction to the previous order. well. At least something.... My first robot with profit and no loss. I think it's not bad for owl with 50 pips increment. Although I've seen similar messages on the forum more than a hundred times....

"A robot with profit without losing is when you manage to make a profit at least on the demo. I've seen a lot of robots that were showing great results in the tester, but stopped working right away when I put them on the demo. Moreover, I have seen a lot of them that at first worked well on a demo, and then all of a sudden stopped working, went deep down and never rose again, constantly showing losses.

 
Александр Князев:
Strangely enough after writing this thread, my EA in the tester(right after changing the "<" sign to the opposite) showed a profit of 44% on a run for the last 12 years with a drawdown of less than 10%.

Do you think there is a connection? Then write, write all the time! :)

 
Aleksey Nikolayev:

If switching between systems is done according to some pre-defined algorithm, then you just get another system, albeit a more complex one, which can also work with varying success.

Or are we talking about some kind of manual (intuitive, creative) approach to switching between systems?

Statistics are hard to come by in this system. In addition, EAs are periodically optimized. Expert Advisor is a method, a kind of method and a tool. There are over 600 of them in total. I looked, but could not even make a comparison logic to distinguish more or less the same periods). But I have the material to work with) and is gaining.

 
Valeriy Yastremskiy:

Statistics are hard to come by in this system. In addition, EAs are periodically optimised. An EA is a method, a type of method and a tool. There are over 600 of them in total. I looked through them, but could not even make a comparison logic to distinguish more or less equal periods). But I have the material to work with) and is gaining.

Generally speaking, it's a task of optimizing a portfolio of systems. The problem is not new, but not to say that it is completely solved)

 
Aleksey Nikolayev:

Generally speaking, this is the problem of optimising a portfolio of systems. The problem is not new, but not to say that it is completely solved)

The problem is usually solved for a set where each system is initially profitable in order to maximize the recovery factor.

For a portfolio in which there are no profitable systems the problem does not make sense.

 
Andrei Trukhanovich:

The problem of optimising a portfolio of systems is usually solved for a set in which each system is initially profitable in order to maximise the recovery factor.

For a portfolio with no profitable systems the problem does not make sense.

Let's assume that systems can be shorted and the portfolio can be frequently recalculated and rebuilt(sliding optimization), then everything will surely work out)

 
Aleksey Nikolayev:

Let's assume that systems can be shorted

I don't know, that's a very bold assumption. for me, if a profitable system initially runs at a loss, it makes no sense to flip it.

The main point is that the problem of finding profit should be solved at the system level, not at the portfolio level.

 
Andrei Trukhanovich:

I don't know, that's a very bold assumption. For me, if a system is initially profitable, it makes no sense to flip it.

If equity is a global trend up, then there is no point in flipping it. But if it forms some kind of channel, then why not try?

Andrei Trukhanovich:

The main point is that the profit seeking task should be done at the systems level, not at the portfolio level.

I agree in general, but one possible way to develop a 'league table' is some sort of boosting analogue of MO, where you can try to build something that works from a bunch of bad parts.

 
Aleksey Nikolayev:

... when you can try to build something that works from a bunch of bad parts.

The reliability of a system assembled from unreliable parts deteriorates many times over.

 
Aleksey Nikolayev:

If equity is global trending upwards, then there is no point in a flip. And if it forms something like a channel, then why not give it a try?

according to Occam's razor principle, you are guaranteed to complicate the system by essentially shifting the trading system from instruments to synthetic derivatives.

i.e., to the unsolved problem, the assumption that you can find profit on synthetics, even though you could not find it on the basic instruments, is attached.

And the synthetics are not fundamentally sound.

___

Using Georges as an example.

He is looking for a sacred way to switch between strategies that will raise his deposit to the sky and which doesn't exist as see above.

The trick is that with normal TS any adequate switching system will work well (plus\minus) and vice versa.