Theorem on the intersection of two MAs - page 4

New comment
 
Korey писал(а) >>
two joint MA extrema with a spread < 3*spread break the following TS,
but they are not observed in the problem of MA finding.

Korey, pardon me, but the language in your last post sounds like the vernacular of an old bumpkin. In short, explain yourself properly, preferably mathematically correct.

 
Neutron and I'm sorry, too: in the language of Chapaev and his friend Petka, it is much easier and more profitable.
- I was thinking about what prevents an educated person (not a chess-player) from being rich, (examples are massive((((.
...I watched long-lived traders in stock trading,
и ... I removed all higher mathematics from the computer as well as from trading, and forgot the rest a long time ago.
so now it is difficult for me to create a mathematically correct post.
 

Just kidding, though!

You, however, should think a bit (when you're a little more relaxed) about the design of the functionality in this formulation. Maybe there's some good sense in it... Otherwise we'll just keep squashing the wagons for nothing.

P.S. Mathemata seems to have lost internet again.

 
Korey писал(а) >> Is it possible to create a BP in which a system of two MAs would not have a profit in a finite spread?
- An example of such a segment is a long gently sloping slope with superimposed oscillations of double amplitude up to 3x spread.
no combination of MAs is catching.

That's an interesting problem statement. But the oscillations themselves should hardly be too periodic (that's if you set the problem the Chapaev way).

Mathemata seems to have lost internet again.

Yeah, the bastards haven't got it home yet. Only at work I can answer on occasion. Looks like the topic is starting to get interesting.

Here's your functionality, Neutron, it's kind of incomplete. Where will smoothness come from if there's no upper mathematics in its expression (i.e. derivatives)?

 
Mathemat писал(а) >>

Looks like the topic is starting to turn interesting.

Here's your functional, Neutron, it's kind of incomplete. How can it be smooth if there is no upper mathematics in its expression (i.e. derivatives)?

This, you, which of the two functionals mentioned above are you talking about? If you are talking about this one: (x[i]-y[i])^2+(y[i]-y[i-1])^2-->0, then the derivative of the waving is y[i]-y[i-1], and if about this: "You want to construct a functional that minimizes the equity deviation from a straight line and at the same time, maximizes the tangent of the angle of inclination of that line", then it still needs to be constructed.

 
Yeah, that's what it's about. I'm sorry, I just got it now. The first bracket is the deviation of mach from the price, the second bracket is the derivative of mach. Then you can try to set y[i] as some filter from price x[i] with unknown coefficients and then try to find them by such "ANC". The filter, of course, does not have to be linear. In principle, this problem can be solved in Excel using the "Find Solution" addon.
 

We don't need Yoksil-Moksil! Let's decide for ourselves. Why some? Once again, there is a certain TC that chops up the initial quotient at Mashka's beck and call. Mashka's parameters are all right - two or even more, picked up when deciding to minimise functionality. All that's left is to design it... I.e. our task is to harness a horse and a doe (TC and Masha) to one harness (functional).

 
Neutron писал(а) >>
So, theoretically?

Yes, directly in the Expert Advisor to programmatically calculate the optimal parameters using some "formula". Although, in general, it does not matter, because we can build the optimizer into the Expert Advisor that performs virtual deals. In both cases, the price series are manipulated and the result is obtained. The only difference may be the speed and the use of computer resources.

 
Korey писал(а) >>
Let's look at it from the other side:
Is it possible to create a BP in which a system of two MAs would not have a profit on the final spread?
- An example of such a segment is a long gently sloping slope with superimposed oscillations of double amplitude up to 3x spread.
no combination of MAs is catching.

A zone of insensitivity could be introduced, at least such that no transactions will take place.

-

The trades will be executed when the speed of the wheel passes through 0, those at the points of extremums where the reversal is expected.

To remove noise and false positives, smoothing on some period is required which leads to inevitable half-period delay. It may happen so that while the price is averaging the wopper will reverse and the trades will be executed against the movement, in the opposite phase. Basically, the loss may be only in this case.

 
registred писал(а) >>

That's all true.)

Well, here's how to prove it theoretically.

1234567
New comment