The potential yield of the instrument. - page 4

 
Neutron, your hypotheses about the differences between the real and ideal market are not true in this context. Yes, I took real market data as input. But you will get the same result if you input any data at all. There is only one condition: all input prices must be a multiple of one point.
 

Then, I don't know. Maybe there's a problem with the algorithm for building the ZZ...

In the context of our argument this is not relevant (small second order). You argued that there would be an increase in profit when you deviate from Hopt to the left. There isn't.

I argued that if you deviate from Hopt by at least a point to the side - there will be a decrease. For the analytical solution I investigated this is the case (see figure blue line). The fact that for the numerical simulation when deviating by 1 point from the optimum we do not observe a change says most likely that there is an error in the ZZ-construction (e.g. in rounding or the like).

What next?

P.S. It is very interesting to know why it is that my reasoning " about the differences between the real and ideal market is not true in this context" . What is this vocal assertion? What kind of context is it that doesn't correspond to reality? Like - there is in general, but not in the context! That's bullshit. Would you kindly answer for the bazooka? A few posts above, where I mentioned that there is rigorous proof of my assertion, I had to provide proof at your request. Now it's your turn to show me where my hypothesis is wrong. And please, no generalities.

 

The next step is to call an ambulance.

 
In a loop:-)
 
Neutron писал(а) >>

Then, I don't know. Maybe there's a problem with the algorithm for building the ZZ...

..

What's next?

...

Next thing seems simple to me. We take the same zodiac. And we change the spread. Because the spread is different on different symbols. Let the fluctuations remain the same, but the spread changes. It is like checking ourselves.

We obtain the results and analyze them. Neutron, maybe I'm wrong, check it. Check it. What spread we have chosen is the maximal one. If we do. Then agree it is not logical.

I don't know what I'm thinking.

 

So, if the algorithm for building a zone is not correct, then you may as well run with the spread.

Privat, I am not interested at all in searching for a software blunder. If there is a discrepancy between the analytical solution and the experiment, it shows an error in the numerical model or a subtle effect associated with the real kotir (discrepancy - just a little). About the error in the algorithm - I'm not interested, about the subtle effect I have already said. Talking about an error in the analytics is silly. There are three formulas and one prism!

 
Neutron писал(а) >>

If there is a discrepancy between the analytical solution and the experiment...

Here it shows that your analytical solution is inadequate. Maybe you didn't understand the problem at all.

 

There is no problem with the ZigZag algorithm. You can verify this by writing your own. And note that there are only integer addition operations in the algorithm, so there can be no rounding errors here.

You said that the profit at min knee equal to spread is greater than at (spread + point). This is incorrect. As these values will always be equal. And I don't know why this is not obvious to you. I have given you an example of why it is so. There are no errors there. You have the opportunity to experiment.

Again I wrote above that the points of ZigZag extrema for min knee n (n < N) contain absolutely all ZigZag points with min knee N. You can see this through the MathCad file given above.

It follows from this statement that the sum of the knees of ZigZag_N will always be at most the sum of the knees of ZigZag_n.

When it comes to taking into account the spread and not just the sum of ZigZag knees, but the profit including the spread, things change slightly: the ZigZag will increase when the knee is reduced to N=Spread (max).

But due to the fact that the profit for knees of exactly Spread length will be zero, the ZigZag_Spread profit will always be equal to the ZigZag_Spread+1 profit.

This is a detailed statement.

But, once we get away from the condition that price is a multiple of a point (i.e. prices can change by any amount), the ZigZag_Spread profit will be higher than ZigZag_Spread+1. Failure to account for price multiples of a point in your analytical modelling is what leads you to your result.

 
Integer писал(а) >>

Here it shows the inadequacy of your analytical solution. Maybe you didn't understand the problem at all?

That's possible.

Except that statements in such a form do not make you look good. Making such a statement, you need to back it up with facts, e.g. a mistake in a formula or derivative... Although, in the post above you complained about the very complicated mathematics in those three formulas. Consequently you can't understand what is reflected there and you are making a statement...

What's that called in plain language? That's right - chit-chat, flubbing. Why would you do that?

 
mql4com писал(а) >>

Not taking into account the multiplicity of the price point in your analytical modelling and leading you to your result.

Sounds about right. Need to make sense of it.