From theory to practice - page 242
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Alexey, if you have reviewed Shelepin's work, can you explain the meaning of parameter C? Is it a constant?
In other works of the same authors, I have seen contradictory information:
1. The parameter itself has no meaning, only the value of C/lambda is defined as the jump frequency.
2. The parameter C is a constant, by analogy with the speed of light in a vacuum.
Now I'm setting C is a constant, = 0.0001, but I doubt it...
In my opinion, it's still some calculated parameter and it's different for different currency pairs.
If, at your leisure, take a look at it and give your opinion, I would be extremely grateful.
For the first time the relativistic electron flutter effect was analytically "discovered" by Schrödinger and, further on, was considered by W. Pauli and Dirac. Analytically it is quite simply found, for example, in I.V. Polubarinov" About coordinate operator in quantum field theory", preprint, Dubna, JINR, 1974. From this relativistic jitter already effects similar to diffusion, i.e. appears similar analytic description of the phenomena. I have not read Shelepin, but it describes something similar.
I do not think it is appropriate to describe the market by means of these equations, but it is my opinion.
I believe that one should simply study the market directly statistically and find the statistical patterns that are characteristic of it, which only then should be modelled analytically (the cart should not roll before the horse).
There is nothing to think about. In Shelepin's book, C is a constant = speed of light in a vacuum. If we quantise the relativistic equations, then at the Planck wavelength scale there appears as a relativistic electron shuddering effect, i.e. the electron seems to move by chaotic small jumps with the speed of light.
For the first time the relativistic electron flutter effect was analytically "discovered" by Schrödinger, and further considered by W. Pauli and Dirac. Analytically it is quite simply found, for example, in I.V. Polubarinov" About coordinate operator in quantum field theory", preprint, Dubna, JINR, 1974. From this relativistic jitter already effects similar to diffusion, i.e. appears similar analytic description of the phenomena. I have not read Shelepin, but it describes something similar.
I do not think it is appropriate to describe the market by means of these equations, but that is my opinion.
I believe that one should simply study the market directly statistically and find exactly the statistical regularities that are characteristic of it, which only then should be modelled analytically (the cart should not roll ahead of the horse).
I wholeheartedly agree.
Now divide that velocity by the negative lambda, well....
we will have to spit on the formula - the charge will turn out to be meaningless
But if we put the sense into it, the barrier will not be penetrated, the particle will bounce back into the negative lambda and will lose the speed of light, but is it possible...?
At max minima.)) The level is usually a flat, and it is unknown where it will go next. And on max-minas you know - in the opposite direction.)
I confirmed it during the day, and having suffered with it a fair bit, I'm coming to the conclusion that it's not quite
not working!
with this approach it's either paltry profits or livantos
i'll try to make a profit no matter how hard they try to hide the line
Wholeheartedly agree.
Now divide that speed by the negative lambda, well and....
We'll have to spit on the formula - it's a mess, it's a no-brainer.
But if we put the sense into it, the barrier will not break through, the particle will bounce back (negative) lambda and will lose the speed of light, but is it possible...?
It's me personally who gives the meaning of lambda a bit different. And Shelepins lambda is the average of the spikes and it is positive. There's no need to be picky. They've got it right.
There is nothing to think about. In Shelepin's book, C is a constant = speed of light in a vacuum. If we quantise the relativistic equations, then at the Planck wavelength scale there appears as an effect of relativistic electron shuddering, i.e. the electron seems to move by chaotic small jumps with the speed of light.
For the first time the relativistic electron flutter effect was analytically "discovered" by Schrödinger, and further considered by W. Pauli and Dirac. Analytically it is quite simply found, for example, in I.V. Polubarinov" About coordinate operator in quantum field theory", preprint, Dubna, JINR, 1974. From this relativistic jitter already effects similar to diffusion, i.e. appears similar analytic description of the phenomena. I have not read Shelepin, but it describes something similar.
I do not think it is appropriate to describe the market by means of these equations, but that is my opinion.
I believe that one should simply study the market directly statistically and find exactly the statistical regularities characteristic of it, which only then should be modeled analytically (the cart should not roll ahead of the horse).
Thank you!
I assure you that of all the models I've looked through for Brownian motion processes, this one describes the market most accurately. Well, I'm just experimentally convinced of that. And probability densities- the product of the exponent on the Bessel function and variance calculation etc. I've only slightly edited formulas, giving the physical parameters a market sense. The constant C has been "picked up", it's certainly not the speed of light here... etc.
Went through VERY many models, only this one works.
And doubts about C were and still are because I was the one who "adjusted" its meaning. Since my school days, I think that fitting it to the right answer is the wrong solution. Well, let it be what it is.Hello, everybody. I am interested in this topic, for some time I tried to translate laws of statistics and time series prediction to price movement, but in this topic I am also interested in volume of discussion - 2400 posts in 3.5 months. Did you manage to get theory proved in practice and earn something? Or are only theoretical statements growing exponentially so far?
So far, that's it for 1.5 months.
Not everything in this thread has to be read, of course. Most importantly books by Orlov, Shelepin and some selective posts.
It is necessary to bring order to the branch, but there is no time - money to be made.
I'm sure that in time someone will come along who will formalize everything and write a good article, while all the other articles on this forum can be thrown out.
So far so good in 1.5 months.
Thanks to the kind Bath moderators - saves a lot of time. Keep up the good work)
Since I asked for the stats myself, I can't help but write)
Basically, a good start for a full automatic. Strongly similar (PF etc) to Bollinger's trading results (which is not surprising).
The bad thing is that the average plus is significantly less than the average minus - a clear sign of overlapping losses (so I understand that trading is done without stops).
Although, for better statistics it would be better to try it for half a year to be sure.
For euphoria purposes - imagine, Alexander, that the beginning of testing will be on the 34th trade of the chart, or, God forbid, on the 15th trade. At this point (the test user) will have doubts about the adequacy.
Now imagine that such a state of "market" (as in the middle and end) would repeat, or even worsen.
This is called "fooled by chance". Without the history test - not a scientific approach (for the algo-systems, of course) . That's where you should have started. And don't say you weren't offered one.
And so, on the whole, so far, so good))
IMHO, of course.
Thanks to the kind Bath moderators - saves a lot of time. Keep up the good work)
Since I asked for the stats myself, I can't help but write)
Basically, a good start for a full automatic. Strongly similar (PF etc) to the trading results from Bollinger (which is not surprising).
The bad thing is that the average plus is significantly less than the average minus - a clear sign of overlapping losses (so I understand that trading is done without stops).
Although, to get good statistics I would need to test it for half a year - to be sure.
For euphoria purposes - imagine, Alexander, that the beginning of testing will be on the 34th deal of the chart, or God forbid, on the 15th deal. Then the doubts would appear (the test tester's doubts) - in adequacy.
Now imagine that such a state of "market" (as in the middle and at the end) will be repeated, or even aggravated.
This is what is called "fooled by chance". Without a history test - not a scientific approach (for algo-systems of course) . That's where you should have started. And don't say you weren't offered one.
And so, on the whole, so far so good))
IMHO, of course.
There's not much euphoria either.
I'm interested in checking out the story myself. I'll probably post a freelance task soon. It is necessary to convert the tick archive to the required series. Bend it a little with an exponent :)) Tell me what you want, but it is necessary to read tick data in exponential intervals. Time is the cornerstone of this task. Equal deltaT-->0 between observations works, but not so well. Generally, now I really understand that the market is a very non-trivial task. You have to be in this non-linear space-time continuum yourself...
I have already written. I'm not working by OnTick, but by OnTimer = 300 ms.
I don't know if it is an error or not in MQL, but it happens very rarely that several trades are opened, despite the strict condition of OrdersTotal()=0.
It is a very unpleasant thing. That is why I very strictly adhere to MoneyManagement and take my time moving to larger lots.
If according to the strategy there may be strictly one open trade at a time, then perhaps it is necessary to fight for it.
The first thing that comes to mind is to hang the terminal global variable when starting the handler OnTImer in the "Expert Advisor is working" state, and therefore, before hanging it in the same handler, check if it is already hanging. If it is - do not process anything. The trades should be opened with synchronous OrderSend, because asynchronous ones can open as much as they want.
It's all "if by strategy, at one moment there can be strictly one open deal".
Or we can assume that the strategy allows more than one simultaneously open trades. But make up your mind!