The Sultonov Regression Model (SRM) - claiming to be a mathematical model of the market. - page 26

 
tara:

Well don't :)

To be or not to be, that is the question!

;)

 
avtomat:

To be or not to be, that is the question!

;)


To be or not to be. Budmo!
 
tara:

Everything is to be. Budmo!
And drink immediately... (с)
 
gpwr:

What is the problem? The conversation was about a normally distributed price, not random straying, which are two different things.

Random price straying will give you a normally distributed price in the end - that's for sure.

;)

 
avatara:

Randomly wandering the price will give you a normally distributed price in the end - that's for sure.

;)

figure... if you look more closely, the price has more of a Laplacian distribution... - Admittedly :)
 
Aleksander:
figure... if you look more closely, the price has more of a Laplace distribution... - Admittedly :)

Incidentally, the parapmeter t in (18) is nothing but a representation of time t in the Laplace transform, so, as shown earlier https://c.mql4.com/forum/2012/07/qwzi3.JPG, RMS perfectly describes the Laplace distribution http://dic.academic.ru/dic.nsf/enc_mathematics/2650/ЛАПЛАСА.

P(t) = P0 +D*Hammasp(t/t;n+1;1;1), as interpreted by Microsoft.

 
Aleksander:
figure... if you look more closely, the price has more of a Laplace distribution... - Adnazno :)

The real price distribution, the Laplace distribution and the normal distribution are three different things... )

And SB and normal distribution are the same thing. and as I wrote:

Random price straying will give you a normally distributed price in the end....

Or do you think otherwise?

Good thing the price does not wander randomly.

;)

 
avatara:

The real price distribution, the Laplace distribution and the normal distribution are three different things... )

And SB and the normal distribution are the same thing. and as written:

Or do you think otherwise?

It's a good thing that the price does not wander randomly.

;)


The random walk has price increments described by a normal distribution, not the price itself. Two different things. SB has no tendency to return to the mean and may trend away from its original value. A normally distributed price has a 100% guaranteed return to the mean.
 
gpwr:
A normally distributed price has a 100% guaranteed return to the average.
And where do we get it?
 
yosuf:

By the way, the parapmeter t in (18) is nothing but a representation of time t in the Laplace transform, so, as shown earlier https://c.mql4.com/forum/2012/07/qwzi3.JPG, RMS perfectly describes the Laplace distribution http://dic.academic.ru/dic.nsf/enc_mathematics/2650/ЛАПЛАСА.

P(t) = P0 +D*Hammasp(t/t;n+1;1;1), as interpreted by Microsoft.


Yes, the Gammarasp included in (18) describes the functions of the Laplace and many other distributions, but not the random variables themselves. This is a big difference that you apparently do not understand. Similarly, one could argue that Exp(-x^2/sigma) is the best white noise regression function because it describes its statistical (Gaussian) distribution. Bullshit!