Theorem on the intersection of two MAs - page 3

New comment
 

Optimal machine parameters in the sense of the TC are at least two tasks:
1. elimination of false integral-differential MA signals
2. phasing of MA according to the market
The solution of the first problem is in itself the exclusion (e.g. filtering) of false signals,

cannot be solved correctly without the second problem - phasing by the market,
=>
to the flow of false signals of TS(MA) it would be possible to compose a functional, which in its turn would be useful to reduce to zero,
but all this is possible if only originally there was an optimal phasing of MA.
=the circle is closed on false and/ or MA signals.

 

For me, the perfect mash-up is the one that lags minimally behind the cotier and is as smooth as possible. You can't think of a better one! The function to minimize it is simple: (x[i]-y[i])^2+(y[i]-y[i-1])^2-->0

Solving it, we get a recursive expression for the ideal LPF. This will be the most delay-free and maximally smooth. Everything else is bogus.

 
That's true, it's mathematically clear,
but how can we trade if this MA has gone into the Open/Close,
- you can't see the forest in the undergrowth.
=the closer the MA is to the price, the greater the trading uncertainty.
i.e. the functionality would have to include something from the TS
 
diakin >> :

1. For any time interval it is possible to choose such parameters (optimization) that the Expert Advisor will give profit on their basis.

In other words, there is no interval, on which the optimization will not give results.

2) Any timeframe can be divided into a finite number of parts, so that after the optimization the Expert Advisor will be profitable at each part of it.

Hmmm...

Is any of this correct?

It's all true:)

 
Korey писал(а) >>
I mean, the functionality would have to include something from the TC as well.

What do you mean - an adaptive dash?

 
Adaptive by drawdown, spread, freeze levels....
 
It should be understood that the parameters listed are somewhat stationary... is that right? Otherwise, we get a vicious circle - because of non-stationarity we will have FZ throwing our machine over the decision horizon (the same eggs, side view).
 
we come to question the correctness of the decomposition of the TC task into subtasks: 1.invention of MA, 2. fitting in the TS
 

I think I'm slowly beginning to understand you!

You want to construct a function that minimizes the equity deviation from the straight line and simultaneously maximizes the tangent of the slope of that line. At the same time, equity is connected to the quotient via TC. Do you need to solve the task as stated above?

Then, we need to determine the optimal (based on what?) TS.

 
it may not be so global,
e.g. two joint MA extrema with spread < 3*spread break the following TS,
but they are not noticed in the problem of MA finding
1234567
New comment