To follow up - page 21
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
Happy New Year to everyone!
It seems that a certain number of authors find it easier to describe the state of the trading system in the market rather than the market itself.
Like there is a rocket and an approaching target. ;)
So in context, what kind of system do we see?
IMHO, the phase space need not be linear vector space, nor does it need to have not only two, but even one operation on its elements. I believe that there is only one requirement that must be taken into account: the state parameters of the system, which may be the coordinates of the phase space, must be continuous in time. This requirement ensures continuity of the trajectory of the system in phase space. If there is no such continuity, any attempt to draw any conclusions about the system's behaviour even in the near future turns into a meaningless effort.
On the other hand, if we have a continuous trajectory, then there must be a certain notion of proximity of points. It follows that the phase space must be metric. It is possible that the explicit form of this metric does not even play a role and the Euclidean metric can be used. However, it is unlikely to do anything without it.
Logically. See the article at the link just below.
Seconded. I just wouldn't fixate on mass wishful thinking. The arguments are this.
Crowd desire is known to set a positive feedback loop. The greater the desire, the clearer the crowd understands where to run, the more the crowd is drawn in that direction and the greater the desire becomes. The end of this trend occurs when there is no longer anyone to entrain, i.e. when desire is maxed out. Banks and big players take advantage of this and manipulate the crowd in this way. Therefore, I would consider it most promising to find instruments which reflect the mood of the banks and big players and which will be ahead of the crowd's desires. But this is not in opposition to "instruments that reflect this mass desire in the best way possible", but in addition, for balance, for completeness of the picture.
Speaking of crowds.
Here's a picture for reflection:
A time series of returns was split into two rows based on some feature. The length of the first series turned out to be a couple of percent shorter than the second series. Histograms of the resulting series are shown above. The first distribution is bimodal and tailed, while the second is unimodal and has almost no tails. This is clearly not the case for random returns.
Can you elaborate on that?
I partially developed the idea from this article: http://www.waset.org/journals/waset/v30/v30-62.pdf
I also assume now that we will be interested precisely in areas with minimal m.
About that I am not sure.
But can you give me some links to more effective algorithms?
Yes, I can. Besides, Google gives these links first, if you ask it nicely.
http://public.lanl.gov/jt/Papers/box-assisted-correlation.pdf // O(N*logN)
http://arxiv.org/ftp/arxiv/papers/0905/0905.4138.pdf // -||-
http://matematicas.uclm.es/abueno/webpdfs/ebaalgor2.pdf // -||-
http://www.pittsburgh.intel-research.net/people/gibbons/papers/ipl05.pdf // Intel algorithm with complexity O(N)
http://dissertations.ub.rug.nl/FILES/faculties/science/1998/s.a.borovkova/c2.pdf // good overview of methods
http://repository.tudelft.nl/file/333955/202549 // it's just something to think about
http://www.adamgrare.com/docs/Ning2008c.pdf // also interesting
Насчёт этого я не уверен.
What's wrong with too few degrees of freedom?
What is wrong with too few degrees of freedom?
As I understand it, the higher the number of degrees of freedom (~groups of agents with independent behaviour) - the more stable the price behaves. Accordingly, the price returns to some average and we can use the pullbacks to make a profit. // Second histogram from my previous post
When the number of degrees of freedom is small, the price tends to continue moving in any direction, but this movement is likely to be unstable as the change of behavior of one group can have a very strong impact on the price and simply turns it around.
Correct me if anything is wrong.
By the way, here is another useful link: http://www.mirkin.ru/_docs/kon_diser/diserfilatov.pdf (I was interested in Wege-Easing model from there).
It seems to be easier for a certain number of authors to describe the state of the trading system in the market rather than the market itself.
In fact, it's not clear which is easier. :)
The way of thinking is this:
In order to look for such tools, you need to have examples of the implementation of the given contexts at your disposal. That is, to be able to identify them in history. Mathematics comes to mind precisely as a means of finding realizations of contexts in the past.
And how do you see a practical approach to finding such tools?
There are many variations. It all depends on whose sanction we're tracing.
As a purely formal approach, you could put it like this:
IMHO the following variant is interesting: we consider only price changes.
K1*(C0-C1)*V1 + K2*(C0-C2)*V2 + ... KN*(C0-CN)*VN,
where С0-current price, СN- price of N bars ago, VN - volume of N bars ago, KN - coefficient of effect of N bars ago price
Actually we examine equity of some part of the crowd. When traders open a position, they make decisions based on price changes, not prices themselves. The trick is in the correct choice of KN ratios, it would be more logical to choose them depending on the trading methods of the initiated group of traders.
For example, if we want to get the aggregate equity of breakers, the ratios can depend on the price deviation from the breakdown level.
Further we analyze this equity. Both critical values of this curve and its critical changes (like drawdown) may be of interest. The main thing is to adjust the ratios to a certain target group.
Then let's consider the main trading methods of breakout/breakdown, medium widening, channeling. And obtain the interaction of these groups in the numerical form. Perhaps even for the groups with different horizon, but it may be too much.
From the standard indicators it will be distinguished by targeting the target group, not just mixing up all the prices. And discreteness - the depth of the analyzed history is not fixed, but depends on the time of the signal for entry for this group of traders. We focus on the exit of a certain group of traders, which can be both sl, tp and time.
http://forex.kbpauk.ru/showflat.php/Cat/0/Number/118170/page/0/fpart/1/vc/1
But this is a purely theoretical approach. In practice everything depends on random findings - testing different ideas and hypotheses, making sense of the results. And so the concept itself - the constant attempt to tie the patterns found to the market logic (and contex/sanction is an important part of it) is mainly useful in practice. Because the search is practically blind - nothing is reliably known about modern trading, what the participants are guided by, and how they make money in general apart from speculation. I am not going to try to find out what is going on in the market, but I am going to find out what is going on in the market.
I will therefore be pleased to see that I can use the EA to emphasize a certain context. Probably I will comprehend, investigate and find something interesting :)
I second that. I just wouldn't fixate on mass desire. The arguments are this.
Crowd desire is known to give positive feedback. The greater the desire, the clearer the crowd understands where to run, the more people are drawn in that direction and the greater the desire becomes. The end of this trend occurs when there is no longer anyone to entrain, i.e. when desire is maxed out. Banks and big players take advantage of this and manipulate the crowd in this way. I would therefore consider it most promising to find instruments which reflect the mood of the banks and big players and which will be ahead of the crowd's desires. But this is not in opposition to "instruments that reflect this mass desire in the best way possible", but in addition, for balance, for completeness of the picture.
I agree. I was not referring to the crowd as uninformed philistines, but to any group that is significant financially and acts in some sense stereotypically (similarly). Even if it is a single participant - for example a large market maker.
На самом деле непонятно, что легче. :)
So maybe define what is worthwhile.
After all, without a definition of the system, all arguments about phase spaces are "lattice"...
;)
Складывается впечатление, что определённому числу авторов легче описывать состояние торговой системы в рынке, а не сам рынок.
How do you yourself understand the context? How do you propose to describe a market without actually knowing anything about how it actually functions?
For ordinary consumers of quotes, the market can be nothing more than a stream of data coming from the Almighty (the DC where you trade). The speculation about the big players, Consortiums, market makers, even their option levels, means nothing until we have an accurate model describing their behaviour. In its input form the market for us is a black box with unknown stuffing, but with some known outputs.
I'll try to be more specific, as I see an attempt to see at least some semblance of the Context of the market. This attempt is not definitive, I'm just trying to bring the discussion down to earth.
Firstly, there seems to be no way for us to escape from the timeframe as a Context parameter anyway - simply because there are players with different horizons of action. The entries of a player holding positions for weeks are qualitatively different from those of a scalper catching a few pips. The timeframe does not necessarily unambiguously define the shape of the initial chart data. It is simply the minimum quantum of time a player is working with. It can be a chart with astronomically marked candlesticks, or Renko, or even ZZ.
OK, let's take the TF H1 and look at our favourite Euro. The first thing we need to do in this chart is to set some reference points, anchors from which we will have to move further on. These anchors should not be ideal markup, detached from reality (like the ZZ markup), but signals of a real algorithm that works in real time (without looking into the future). This, by the way, is why I don't like the ZZ, as the moments of signal registration on it are quite distant from the moments of "ideal inputs".
There is complete arbitrariness in the choice of this real algorithm. It does not have to be a trading system. It can be an arbitrary combination of historical prices, volumes, levels of bids, etc. The only important thing is that this algorithm is real-time.
Let's start with something primitive to illustrate the idea. The basic system - "crossover of the chart with the period of 10 bars with the first bar close price". Let's start it for testing. This system marks the chart into [Open_position; Close_Position] segments. Let's call this system of bars as "10-point chart".
Further we similarly build N-charts for N = 2...500. What are we doing it for? We do "convolution by parameter N": we get rid of arbitrary parameter N, which obviously has nothing to do with Context. What we achieve is that some logical function from N-units (symmetric in arguments, i.e., markups) gives us some Ultimate markup in the end. Say, if this function is an intersection of markings, then the Marginal marking is the conjunction of all N markings. I want to point out that it doesn't have to be a conjunction, but it's the conjunction that seems the most logical so far.
It is clear that in the case of our basic system the Ultimate markup will not exist. It will be empty. Choosing the right basic system is a kind of art. It needs to be learned.
By the way, if the base system has two parameters (say, the system "intersection of two M, N masks"), then the conjunction will have two parameters.
Suppose we do manage to find a good basic system which, as a result of a convolution operation on all its parameters, gives us a non-empty Ultimate Partitioning. This is only the first step.
The Marginal Partitioning is already something stable for this Basic Parametrization. It is on this Marginal Partitioning that the search for the Context can be built. That's it, let's rest for now.
P.S. Already now I see flaws in the proposed one. But I'm not going to correct this version. Let it remains what has turned out.
Как Вы сами понимаете контекст? Как Вы предлагаете описывать рынок, не зная по сути ничего о том, как он на самом деле функционирует?
... для нас рынок - это черный ящик с неизвестной начинкой, но с несколькими известными выходами.I am closer to the definition of the context as a frame of reference, where
Система отсчёта — это совокупность тела отсчёта, системы координат и системы отсчёта времени, связанных с этим телом, по отношению к которому изучается движение (или равновесие) каких-нибудь других материальных точек или тел.
The frame of reference could be an "indifferent observer" - sitting on the fence watching. ;)
The price moves in time in relation to it, then near it... then really up... and sometimes down.
(there will also be questions about the error of dividing such a scale to significant levels of "running", but sooner or later we will have to think about that too) the topicstarter's idea of allowing extremum 33 to be used as a reference point seems to me appropriate.
First, apparently, we still can't get away from the timeframe as a Context parameter - simply because there are players with different horizons. The entries of a player holding positions for weeks are qualitatively different from those of a scalper catching a few pips. The timeframe does not necessarily unambiguously define the shape of the initial chart data. It is simply the minimum quantum of time a player is working with. It can be a chart with astronomically marked candles, or a Renko, or even a ZZ.
here I see an attempt to solve this very problem of meaningful partitioning of the coordinate axis for our as yet indifferent body.
To begin with we need to set some reference points, anchors, from which we have to dance further on.
Exactly. Then you can notice in context the state of the market.
These anchors should not be perfect markups detached from reality (e.g. ZZ markups), but signals from a real algorithm running real-time (without looking ahead). This, by the way, is why I don't like the ZZ, as the moments of signal registration on it are quite distant from the moments of "ideal inputs".
.... for now we rest.I didn't expect the concept to be switched. 33 - gives not the signal, but the place where the observer sits. Sometimes a dilapidated fence. Sometimes a citadel.
And if an indifferent observer notices the dilapidated fence, he or she immediately proceeds to the next, more favourable - in his or her estimation - place.
(I would hypothesise that his aim is to minimise the frequency of shifting his head, up or down.
We, who are not indifferent ("not bystanders":), should teach him to behave in such a way, and from his behaviour and strategies to build.
;)