FR H-Volatility - page 30
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
I think the variable lags between ticks contain information, i.e. I'm only talking about the one that we ourselves remove by an equilateral transformation.
If one chooses what to predict, I choose precisely the "continuous worldwide quoting process", and the DC will ultimately get nowhere from it.
I am merely trying to follow the tradition of scientific enquiry, established over the centuries since the Renaissance, which proposesan analysis of a phenomenon whose essential part is to break it down into logically separable parts (not necessarily independent at all, by the way; what matters is that these dependencies between the parts are described by clear concepts). One can still try to study a phenomenon in its absolute totality, the way the medieval scholastics liked to argue about the properties of a stone, without trying to influence or feel it. In my view, for a phenomenon as complex as s.p. market quotations, it would make sense to first divide it into parts, study each of them, then go back again, but on a qualitatively new basis - to study it as a whole, knowing the properties of its parts and, perhaps, the dependencies between them.
2 Neutron and Mathemat zip cannot be attached either. I guess it's a problem with the site. Here is the link where you can download the data
My bad. I attached rar, and it was necessary to zip.
Interesting series! Let us plot the distribution of the modulus of gestation (Fig. on the left). You can see that the centre of gravity of the distribution is located in the area m=0.7 Now let us build an artificial series from the sum of constants=m taking into account the sign of the real increment (see Fig. right) where the red line shows the original series and the blue line shows the artificial series.
It would seem that if the price increments are independent, the sum=increment_sign*sonstant will yield a trajectory that lies in the corridor between two curves y=+-m*SQRT(t) (black colour). But this is not the case. Maybe the signs of the increments are dependent? -No, the correlation coefficient between the neighboring increments is -0.05 i.e. almost zero. So the growth is not determined by the "herd effect" and most likely is not accidental.
The conclusion is as follows: someone or something monotonically pushes the index upwards all the time (the blue curve), and the fact that the index is in no hurry to go in this direction, says that someone seldom but aptly collapses the index!
What else is there to add? Probably, to check yourself by building the same, but for the currency instrument:
Everything is fair here - no one is pulling anyone anywhere :-)
It seems on the previous page of this thread it was discussed that it would be good to find a way to convert the exponential nature of the distribution of increments on small TFs into a normal one. Why this is needed I'm not quite sure... But there is a way.
Take a look at the figure on the left. The red line shows EUR/USD minute bars, and the blue one - a model series that preserves directions of initial price increments, but the amplitude is strictly defined by RMS with normal distribution law and zero MO. We can see that all movements are strictly repeated, but with a "different" amplitude.
The right figure shows the distribution of the EUR/USD series increments (red) and the model one (blue). Hooray! We have managed to get away from the hated "non-normal" distribution and have one of the realizations of the initial series with a normal distribution, see fig:
One can immediately notice areas where the initial and the model series move in different directions! How can it be? It means that the directional motion in the real series on the chosen site is determined not by small and frequent steps of the timid crowd, but by strong and rare strokes of the power ones!
Here. Perhaps this information is new to many and there is some, as yet, hidden potential in it. What do you think, colleagues?
Amplitude is rigidly given by RNG with normal distribution law and zero MO.
HNG? decipher please + if it is some kind of normal distribution generator, you need the value of s.c.o. for its full description.
And if I understood correctly, all your constructions, you intuitively came to the model which is a special case of system of stochastic differential equations.
Amplitude is rigidly given by RNG with normal distribution law and zero MO.
HNG? decipher please + if it is some kind of normal distribution generator, you need the value of s.c.o. for its full description.
And if I understood correctly, all your constructions, you intuitively came to model which is a special case of system of stochastic dif. equations.
The GCF is a random number generator (although I may be wrong).
HSCH? please decipher it + if it is some kind of normal oscillator you need the s.c.o. value to fully describe it.
And if I understood correctly, all your constructions, you intuitively came to model which is a special case of system of stochastic dif. equations.
Hi Sergey!
Are we on speaking terms again? Yes, you're absolutely right - this is a random number generator with the normal distribution and zero expectation. In my example, s.c.o.=m. And unfortunately I understand nothing about stochastic control systems.
Here everything is simple. SSDU (a system of stochastic differential equations). A system means that there may be many of them, the simplest case is one. The equations are clear here like y(x)=a*x+b. Differential (derivatives, increments) i.e. derivative on the left, i.e. dV/dt=a(t) - the derivative of velocity equals acceleration. Stochastic (random) remains, which means that there is a random process on the right. The derivative of price is BGS with can=0 and sko=1. The solution to these equations is to take the integral.
This is what we were talking about a few pages ago with mech.matts, how to solve them using an ITO or Stratonovich notation. pg. 18 there are some simple models (economists) I posted the attached file, look it up equations 8.1-8.6. For military radio engineers they (models) are more complicated.
Z.U. Do not be offended if you or you, okay. I do not want to be put in the oven with a pot :-). I just get confused a lot, it's hard to switch sometimes. Especially on Mondays and Fridays, I talk to too many people. I have three jobs.
One can still try to study the phenomenon in its absolute totality, the way the medieval scholastics liked to argue about the properties of a stone, without trying to influence it or feel it. In my opinion, for a phenomenon as complex as the s.p. of market quotes, it would make sense to first divide it into parts, study each of them, then go back again, but on a qualitatively new basis - to study it as a whole, knowing the properties of its parts and perhaps the dependencies between them.
That's quite scientific :). And did you notice that the discussion was not on the proposed test but instead of it? :) Although, for a person who works with tics bars, such a test is a matter of minutes. If I had a tick history in my terminal, I would have done it even before writing that post, but I won't search and download it for the sake of it.
to Neutron
Seryoga hello. Explain, please, where did you get this from:
I am interested in the formula y=+-m*SQRT(t) itself, how did you get it, where did you get it from? The approximation of the law of repeated logarithm for trajectories of a Wiener process cannot be either, I give a short form:
For a Wiener process W(t) with probability one holds:
All trajectories of the Wiener process remain inside the expanding "pipe" between the curves
At the same time, with probability 1 the trajectories infinitely often jump out of the boundary pipe
Not that it matters at all, just interesting. By the way, I've established through experiments that this law, to put it mildly (if I calculate it correctly of course), does not work on quotes, hence it may be interpreted as an indirect confirmation that the market is not random, something like that. :о)