Distribution of price increments - page 14
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
The fractal nature of risk distribution across all timeframes is right on point. The creator of fractals himself and many others wrote about it.
https://ru.wikipedia.org/wiki/Фрактальный_анализ_рынка
What's funny about it - originally I wrote this page and there was a link to my site, which no longer exists :)) But Almazov got in and rewrote it all for himself, it's funny :))
Originally I did it because it was well positioned in search engines, and there were direct links to the site. I'm laughing :)
It has not been an easy task. We have to find an invariant statistical parameter, which would not change upon increasing/decreasing the volume of tick data sampling. This parameter turned out to be nonparametric skew coefficient (nonparametric skew). Perhaps there are some others, but that is enough to prove it.
A dynamic FIFO-type tick data buffer was used in the calculations. EURJPY was analyzed on the general data set of 1.500.000 quotes i.e. 1.500.000 sequential samples with 1 quote difference were analyzed. We have received the following results for the average value of skew taken modulo for different volume of samples.
0.186043748375457
0.18560474492056
0.184953481402386
0.184985234902438 etc.
Simply put - for any sample size of tick data, the non-parametric asymmetry coefficient remains constant.
The conclusion is as follows: indeed, small TFs show the same processes as large ones, and a trading system operating on one TF will operate on the other and vice versa.
But what is interesting is that we get a quite mystical thing - it turns out that some distribution with a strange mean (I emphasize - mean) nonparametric coefficient of skewness = 0.185 (modulo) "walks" in Forex. I personally do not know such a distribution... Maybe someone can help me determine it?
I.e. in a simple way - at different moments of time this distribution is like "born", "formed" and "dies", and the process starts all over again. At different points in time this distribution has different skew, but on average this distribution is skewed with coefficient = 0.185 and it is invariant.
Until I understand what kind of distribution it is in its average form - there's no point in exploring it further...
Respectfully,
Alexander.
There are 4 questions and one answer. The questions are:
1. "dynamic FIFO type tick data buffer" - this is the second time in your messages, I think it's time to clarify. Did I understand correctly that you originally used the accounting term "first come, first go" of the method of processing document stacks, which with the appearance of PCs and stack organization of data became applicable to the "stack" memory area, where functions working with this memory each separately have no information about what lies above and below the memory area accessible to this function in the stack?
If so, why? Everyone here is familiar with the notion of moving average, and the sequence of handling courses in it is unambiguous for everyone. And at the same time, everyone knows that it is not a stack but a series where all the rates are known - an array that allows processing not only the extreme elements. Perhaps the accounting terminology should be removed for the sake of clarity? Honestly, after sampling from time series "with exponentially distributed time", I have doubts whether I understand correctly that the data was processed using sliding sampling of 10, 11...14 thousand elements in a row over time.
2. about the "general population of 1,500,000" - after all you have already written that this is a sample of roughly a month, why mislead people? Wiki:
General population (from Latingeneris- general, generic) (in English terminology - population) - the totality of all objects (units), in relation to which it is supposed to draw conclusions in the study of specific tasks.
Conclusions you draw at once for Forex...
3. "Nonparametric skew" - am I right to understand that you are saying that this parameter =0.185 modulo? Translation from google:"nonparametric skew." Your title is "nonparametric skewness coefficient". I couldn't find both Russian versions on google, and since I'm not sure of the correspondence, I took the formula from https://en.wikipedia.org/wiki/Nonparametric_skew. It is the ratio (median - mean)/(standard deviation) - did I guess correctly what you were talking about?
4. I cannot understand how an indicator that has positive and negative values, in addition to "distributions that at different points in time seem to be "born", "formed" and "die" and the process begins again", has a modulo average value, in addition to being "invariant". What does this mean? Can you give a graph or other illustrative material?
As an apology for my annoying questions, I offer an answer to your question about the distribution that walks on forex "there is some 1 distribution that "walks" on forex with a strange mean (I emphasise - mean) non-parametric slope ratio = 0.185 (modulo)."
The first thing that comes to mind is that the resulting value of 0.185 is one of the settings (parameters) of the filtering algorithms in the forex company from which you obtained the raw data, in the month to which it applies. Extremely briefly, as I present one of the (relatively fair and short) possible schemes by which quotes are generated at this company:
- Banks sell snapshots of the last real (not gambling, no obligation to "close the open transaction back") made in them without specifying the volume of transactions to agencies like Reuters, Bloomberg;
- your company buys these snapshots from the agencies;
- the company averages the prices of the latest deals by space (by value of rates) and by time, spreads the resulting spread estimate to its needs (or even shifts it, if it has shifted the spread to a commission) and sends it to you in the terminal. Each company has its own know-how of setting up these algorithms, and in addition, each of the company's dealers has the individual authority to configure these algorithms according to the list of currency pairs for which it is responsible.
P.S. Yes, usually also for each type of real account company sets a different spread level, and informs about it on its site. Of course, according to some algorithms that also set parameters. So, analyzing ticks we analyze not Forex, but the properties of quotes generation algorithms of the given brokerage company for the given pair on the specified account type at the selected time period. And here we can detect a lot of miracles. For example, serving shaggy (roughly speaking, unfiltered) or even purposely hacked (for example, by "overregulation") quotes on demo accounts as a way to lure clients to real accounts. Or such signs of a company's "youthfulness" when it allows a lot of arbitrage (which you probably noticed when you were talking about 7 sigma outliers) already on real accounts.There are four questions and one answer. Questions:
1. "dynamic FIFO type tick data buffer" - this is the second time in your posts, I think it's time to clarify. Did I understand correctly that you originally used the accounting term "first come, first go" of the method of processing document stacks, which with the appearance of PC and stack organization of data became applicable to the "stack" memory area, where functions working with this memory each separately have no information about what lies above and below the memory area accessible to this function in the stack?
If so, why? Everyone here is familiar with the notion of moving average, and the sequence of handling courses in it is unambiguous for everyone. And at the same time, everyone knows that it is not a stack but a series where all the rates are known - an array that allows processing not only the extreme elements. Perhaps the accounting terminology should be removed for the sake of clarity? Honestly, after sampling from time series "with exponentially distributed time", I have doubts whether I understand correctly that the data was processed using sliding sampling of 10, 11...14 thousand elements in a row over time.
2. about the "general population of 1,500,000" - after all you have already written that this is a sample of roughly a month, why mislead people? Wiki:
General population (from Latingeneris- general, generic) (in English terminology - population) - the totality of all objects (units), in relation to which it is assumed to draw conclusions in the study of specific tasks.
Conclusions you draw at once for Forex...
3. "Nonparametric skew" - do I understand correctly that you are saying that this parameter =0.185 modulo? Translation from google:"nonparametric skew." Your title is "nonparametric skewness coefficient". I couldn't find both Russian versions on google, and since I'm not sure of the correspondence, I took the formula from https://en.wikipedia.org/wiki/Nonparametric_skew. It is the ratio (median - mean)/(standard deviation) - did I guess correctly what you were talking about?
4. I can't understand how the index having positive and negative values, moreover "distribution which at different moments is "born", "formed" and "dies", and the process begins again", has an average value modulo, moreover "invariant". What does this mean? Can you give a graph or other illustrative material?
Good morning everyone!
Answers:
1. Vissim has a buffer block. It works according to the FIFO principle - "first in, first out". I.e. when I receive tick data sequentially, I am typing an array of a certain size - let's say 10,000. Then a new tick comes in and takes the place of the very first tick out of 10,000 and so on. It turns out. I have analysed over 1,000,000 different consecutive arrays with a difference of 1 tick. A giant statistic and can be believed, otherwise "statistics is not a science" and I don't agree with such a statement.
2. Of course, for forex, the general population is infinite. But, in this case I applied this term, because I couldn't find a better one. After all, my sample volume is 10.000, 11.000 etc., and I took 1.500.000 and called it GS :)))))).
3. Yes, that's exactly what it is.
4. No graphs - the arrays are generated dynamically and they are gigantic in size - I only saved the results. Basically, those interested can repeat my experiments in VisSim or MathLab (not sure in this system, as I haven't worked with it).
Here's what I was thinking.
If the statement that nonparametric skew for Forex distribution is invariant and equals +-0.185 is true, it can mean (without mysticism:))))) only one thing.
Note that for a normal distribution, its half (so-calledHalf-normal distribution) has a nonparametric skew=0.36279.
In this case we have on average a kind ofHalf-unknow distribution which has nonparametric skew=0.185, and if we look at it from both sides, we will see a symmetrical normal-like distribution.
I'm afraid to suggest that on average we're just dealing with my "favourite" non-standardised t2 distribution, the Student's distribution. Formed at the level of increments, it doesn't disappear, it transforms in dynamics, but when averaged, it appears more or less clearly.
Actually this confirms my hypothesis that the probability distribution of price in the Forex market is a superposition ("mix") of non-standardized t2-distributions.
Now we only need to learn to "see" this distribution in dynamics and the problem is solved.
How to see it? I stand by my opinion - by averaging some parameters taking into account quantiles of this distribution.
The fractal nature of risk distribution across all timeframes is right on point. The creator of fractals himself and many others wrote about it.
https://ru.wikipedia.org/wiki/Фрактальный_анализ_рынка
What's funny about it - originally I wrote this page and had a link to my site that does not exist any more :)) But Almazov got in and rewrote it all for himself, it's funny :))
Initially, I built it because it was well positioned in search engines and had direct links to the site. I'm laughing :)
Good day, Maxim! My special respect to you, because without your first comments, which led me to seriously engage in research, this theme would not have been possible.
The first option was sold around 1690. Terver and statistics go even further back in time. Do you really think that all this is not a waste of time? That you are more intelligent and smarter than at least Cardano?
No. Although I am not the worst educated and most experienced, I am far from thinking that everything is already clear and understandable. I am even surprised by the results and would like someone to independently verify them based on just their experience and knowledge.
No. Although I am not the worst educated and most experienced, I am far from thinking that everything is already clear and understandable. I am even surprised by the results and would like someone to independently verify them on the basis of their own experience and knowledge.
Moreover - I think this topic should be published on physics and mathematics forums. However - why should it, if I communicate with such people every day and they are simply not interested in this topic. It is considered that this is not a serious topic, exclusively for young people and there is nothing to waste time on it. I, on the other hand, just got into it out of curiosity.
Moreover - I think this topic should be published on physics and mathematics forums. However - why should it, if I communicate with such people every day and they simply are not interested in this topic. It is considered that this is not a serious topic, exclusively for young people and there is nothing to waste time on it. I just got into it out of curiosity.
Actually this is exactly what I wanted to ask - why should an experienced physicist, statistician (or whatever else you are) be interested in this subject? Wouldn't financiers be better to deal with finances? Everyone should mind his own business. And if there is none, it makes you wonder.
Or is being a physicist a vocation, as Mr Medvedev used to say... If you want money, go into business. If you want to lose money, go into financial markets...