Average daily journey in points by instrument. - page 6

 

Calculated in the new points.


Then drew the relationship ln(period) -> ln(spread). By the way, the weeks fit this dependence better if you count exactly the number of bars per week, i.e. if period_W1 is considered equal to 1440*5=7200. I did so (otherwise the line would collapse at the last point downwards).


As you can see, the line is almost perfect (against the blue line, if you look closely, a linear regression line is drawn). But the slope, i.e. the degree, is not exactly 0.5 (in fact - about 0.529).

After that I checked the Lanchester-Ainstein formula :) (last column, it should be close to 1). Especially strong deviations from the formula suggested earlier are between D1 and H4, and also around TFs up to 30 min, i.e. for the smallest TFs.

If now we calculate by the number of bars, it turns out to be even worse.

P.S. Similarly - by the modulus of Close-Open difference:

Still no particular difference: the degree is now 0.515. And still a strong deviation between D1 and H4.

 
Svinotavr:

At the end, I want to show you this trick. Throw the ATR indicator on the chart. This is one of the price volatility indicators. Set the period to 24. Set the time frame to H1 and look at the indicator. And then change the timeframe to M30 and look at the indicator again. There's a lot to think about. Have a good day.



Will you parry Alexey's answer?
 
Trololo:
Will you parry Alexey's answer?

What is there to argue about? I believe that Alexey's description of the pattern is correct. I would like to add that this is one of the most important laws that novice system developers and beginning traders need to know. It is almost impossible to create safe and highly profitable systems without knowledge of this law.

But, at the same time, we must admit that this law alone is not enough for the creation of such systems. All regularities of price movements are interconnected, as well as the laws of physics, for example, are interconnected with each other. Therefore, by getting a grip on one pattern, it is possible to pull out other patterns as well. If you have the right intention, of course :).

About randomness. The fact that many people including those who know mathematics very well for years have been working on developing trading systems proves that the part of a random component in price movements is significant. But studies show that it is not an absolutely random process and has its own regularities and limitations, which must be known and be able to be applied in theory and in practice.

---------------------------------------------------------------------------------

Then let's continue with the subject.

A question for you, Trololo and to the rest of the forum:

What should be the simplest figure whose one side is approximately proportional to the square root of the length of its other side (or the sum of the lengths of several other sides)? Is such a figure possible? Draw this figure, if it is possible.

(I may not have formulated the question quite correctly, let mathematicians correct me then, but the meaning is clear, I suppose).

 

Question for you,Trololo and to the rest of the forum:

What should be the simplest figure whose one side is approximately proportional to the square root of the length of its other side (or the sum of the lengths of several other sides)? Is such a figure possible? Draw this figure, if it is possible.

(I may not have phrased the question correctly, let the mathematicians correct me then, but I think the point is clear).



I keep wondering how a beginner (note that you said so yourself) can keep saying what's important and what's not, all with the guise of years of experience.

you define who you are. otherwise let's discuss something, you say in a year, I'm a beginner. and meanwhile you teach (no, I'm for communication, but I do not know why you have such a thing).

I've never categorized the quality and experience of people according to age, but the cathigoric nature of it makes me suspicious.

Here is your figure. it happens to look like a right triangle, it may not actually be so. from your terms.

 

Trololo, here is your chart. There are 2 ATR indicators, the lower one with a period of 12 hours and the upper one with a period of 24 hours. The TF is hourly. I think you will see the difference. This is especially for you, but I don't want to discuss the subject of ATR.

 
Trololo:

Jos, what triangles? It's more complicated than that. Think it over until tonight.
As for me, I offered you Skype communication via microphone and video, but you refused, it is your right. And to read and moreover to realize all this forum in a small period of time I simply am not able.

 
Svinotavr:
Jos, what triangles? It's more complicated than that. Think about it.



What is it like, what you asked, that's what you wrote.

One side is equal to the root of the other, no one wrote about the 3rd and subsequent sides (and their number).

 
Svinotavr:

What is there to argue about? I believe that Alexey's description of the pattern is correct. I would like to add that this is one of the most important laws that novice system developers and beginning traders need to know. It is almost impossible to create reliable and profitable systems without knowledge of this law.

The problem is that there is no regularity here. There is a very strong semblance of a random system.

I can add that this ratio is fulfilled with high accuracy for both Forex and stock instruments. There is some difference from one (0.8-1.15) for some commodity markets and for indices (S&P, DAX).

 
sand:
The problem is that there is no pattern here. There is a very strong semblance of a random system.
But, not completely random. See you tonight, I'm off to the sports centre.
 
Svinotavr:
But, not completely random. Until tonight.


The claim of not being totally random requires justification, and better yet, examples and facts. Until there is none, the system can be considered random, even if it has a pattern to it.