EURUSD forecast & Chaos theory

I made the backtest (trading without any margin on daily time-frame) for GBP-contained pairs (as they are very popular) for last year. I think these results could be useful as they correctly value profitableness of these pairs independently of my system.

Aleksey, thank for your information. I have researched a little about chaos in tha last days because your approach is very interesting and mathematical founded. So, some question have arised from my lectures and your commentaries:
You said you have designed a model of the market (defined by diff equations) and then I suppose you resolved numericaly (your computer ) this model (I assume that your model is non-linear because the non-linearity is a pre-requisite for chaos). Your graphics have showed the chaotic behavoiur of your model, but there is controversial about if financial time series are or not caotic. There are several papers that want to demostrate that the behavoiuor of the financial time series are stochastic, at least, not caotic. But your model have great results so one can think that there is caos in financial time series, at least in the Forex pairs you have tested.
So, the questions:
Have you test the Forex time series you use in your investigation with the caracteristic test to determine if chaos is present, like estimation of Lyaponuv exponent and others?
Have you possitive Lyaponuv exponent (because there are report of negative exponents in Forex pairs, although some think that chaos has not to be excluded because the the present of low dimensional chaos - even with negative exponets)?
What is your forecast horizon because it can not be very long if chaos is present, not ?
How you obtained the volume in a desentralizaded market like Forex, can you tell us where to obtain this volume?
Can you tell me in which Journal your paper will be published?
Please can you give us more information about you, you are in a University for example (because your skills looks high)? Thinks like that. And please PM me your web page.
Thanks

1. Yes, of course, I made estimation of an attracter dimension. It is d=2.66.
2. The forecast horizon of the nonlinear part of the system is one day. It only initiates (overturns) a position and "forgets" about that in a day. Further accompaniment is going on by using second linear part.
3. I like SQG data.
4. I will place the article here after publication.
Thanks for good questions.

Aleksey, thanks for your responses.
Here are some more questions:
Can you explain a little more about the linear system to achieve more longer forecast horizon?
How you determine your StopLoss and Profit Targets?
Can you give us links for source of information about Chaos and time series analysis: papers (first, because they have the newest information avaible), books, etc? (I started to read about Chaos Theory 2 days ago and I want to know if I am in the correct way in the source of information).
For the graph in 3D you have posted, which soft you are using? (because I have found some free chaos 3D soft but i am not tested yet).
I am very excited with this approach because I have found someone that take the direction I think/fell is the correct (maths, I love them!). I hope do more question with the advance of my study in this topic.

Regards.

La Totona

Thankyou Aleskey, fxbroker, la totna and zu si fu. now we are getting somewhere In this discussion. Lets keep the inquiry going so we are all get an understanding of chaos theory prior to the release of Aleskeys' system.
Here's my conclusions incorporating chaos, quantum physics and my personal theosophical viewpoint
1 there is a harmonious structure determining all systems (chaos)
2 Chaos is very cleverly designed to appear random thus completely disguising itself
3 all outcomes are predetermined by chaos. alternative outcomes to that already predetermined are not possible.
4 Time is not linear. the past does not precede the future. The real reason why the chaos determined outcome is guaranteed is that, in reality, the outcome has already occurred. Appearances of sequence, complexity and randomness are an illusion caused by the hidden underling structure called chaos. It's all a trick, a slight of hand.

La Totona, the linear part is based on methods of digital filtration. It also determines levels of stoploss and takeprofit.
It is difficult to recommend information directly connected with my research, but I think you should read books about unlinear systems describing biologic populations and any information about chaos theory will also be useful.
Those pictures were made by using Maya source. I do not use their to analysis, only for visualizing of the market state, not more. Analysis is made on the basis of two indicators. First indicator initiates or overturns a position, second one accompanies and stops a trade.

Aleksey,
Can you tell us more about your assuptions in the development of the non-linear system of diff equations. I am not preteding that you posted the system of course (i know you want to wait until your publcation), but maybe you can tell us more about which principles/assuption take your model.
For example, in growth populations, in the logistic equation http://mathworld.wolfram.com/images/.../equation1.gif
the assuptions are that the rate of growth (dN/dt):
1- is proportional to the number of individuals (N)
2- is inversally proportional to the nuber of individuals in relation with the max capacity of the enviroment to contain the population (there are limited recurses, for example material for feed the population).
The points 1 and 2 are the assupntions for the model, can you tell us more about your assuptions.
Maybe you want to keep this information until the publication, I understand that, but I hope you can answer in some way this question to know more about your system .

Regards.

La Totona

La Totona, I meaned more complex models, for example "predator-sacrifice" or "predator-sacrifice-environment ". I use three parameters C, V, I so second model is more preferable.To build my model it is necessary to connect these three parameters on base of economic ties between them.

I am totally in agreement with you here Aleksey. I also shy away from answering questions like 'how do you calculate this or that."

Instead, maybe there should be new thread on general chaos theory - to deal with basic concepts and how poeple think chaos can be used to solve financial market problems. People can share whatever they think they want to share or exchange with other people.

People with mathematical or programming skills can formalise those ideas into problem solving system of equations or software by themselves and do whatever they want with their own efforts. In this way a specific methodological thread like this one will not be overwhelmed with questions not relating to the specific method or what one do not want to give out as public information.

To FXBroker, my questions pointed to 2 ways:
1-questions about general information in chaos and non linear systems
and
2-questions related specifically with the Aleksey System.

I am agree that maybe we can open a thread about general chaos theory and non linear systems to answer of the first type of question, but because this type of thread wasnt at that moment of questions, I thought it was better make them in this thread at that moment.
The questions related to Aleksey System are not the type of questions "how do you calculate this or that", they are questions to know about the basis of his System, with some specific questions like Leypunov exponent or more interesting like the assuptions to develop his model of diff equation.
I asume that Aleksey wants to share with us his knowlegment because he open this thread and answer my question. I hope nobody take my question in a bad form, I only want to know more about specific points of Aleksey System before the publicacion of his paper and to know in which way I have to align my research.

Regards

La Totona

Aleksey, you have pointed pray-predator models.
I have read some about that a time ago. I remember the Lokta-Volterra model, the most famous predator-pray model, that is based in the following diff equation:

dR/dt = rR(1 − R/K) − aRN ,

dN/dt = caRN − dN,

But, you have pointed more more complex model, predator-sacrifice and predator-sacrifice-enviroment models. I suppose these model incorporate terms to take in account other features (for the information in his names). Can you point out this new terms/assuptions in that models or tell us where can get more information about that (I have searched in the web but i dont found any consistent information about that).

Regards.

La totona

Talking about complexity, I meaned a three-component system, for example "prey-predator1-predator2" in general terms "prey-predator-anybody/anything". Unfortunately web has a little information about this subject but the system you noted is enough to start learning. I use differential equations of first order too.

Aleksey, I know your system have 3 components, so your system must have 3 diff equation for determination of infinitesinal change of component in relation with time (in the Lokta-Volterra model we have 2 component and hence 2 diff equation, point out behind). Thanks for the clarification about "prey-predator-anybody/anything", I understand better now how is structured your model.
I have impressed that you have achived a great accuracy model with only 3 component and with first order equations. I think the most important think to model a complex system is found the correct component and the relationship between them. And if you can model this relationship in a simple form, you will arrive to a simple model that can represent well the complex system, and it seems that you have take this way succesfully.
I appreciate your answers, they have very useful information, thanks.

Regards.

La Totona.

First destination is reached...


Dear members,
I have made decision to continue my thread after publication of my article because for now I can not give more information than I have given. Afterwards I will be able to give detailed description of the method and some indicators for MT4 and Trade Station 8.xx and each reader will be able to test their abilities and will ask more concrete questions. Good luck!

I understand that you stop releasing more information until publication.
I hope we can discuss more about your system soon , I am very interesting in it. I will be waiting for any news you can give us in the next months.
When you webpage will be enable, please PM the adress.
Thanks for the great information you have provided us.

Regards.

La Totona.

PS: I have searched deeper in the web, especially in biological population models. Anyone interested can achieve very good information in scientific papers of especializated journals.
Regards to everyone.

La Totona.

A little useful information about hybrids...
I have lately been using this simple but very effective hybrid constructor with noted normalization of a lots size. It simplifies analysis and gives comparable results with trading in 105 hybrids.

This is my forecast, eur-usd will not break the high of 23-11-2007, maybe it can break but only for some 5-10 pips (false breakout), before continue the uptrend, it will go down at 1.3930 level, now.. it seems a good point to short here.

http://img208.imageshack.us/img208/8...castkn9.th.jpg

So, i decided to short right here (1.4800), sl 1.5000, target 1.3950 .

Risk: 200pips (4,2% of account)
Reward: 850 pips (18% of account)

Risk/Reward: 1-4.25

It seems to make some sense to me

Tunera, I saw more strong potential in 23.01.08 CADCHF buy=1.0625, stoploss=1.0361, destinations=1.1000,1.1200 plus
24.01.08 GBPAUD sell=2.2395, stoploss=2.3073, destinations=2.1800,1.1500 so I have lately been staying away from EURUSD.
I will make deep (daily, weekly, monthly) analysis of EURUSD tomorrow to value probable future moves.

For now on daily time-frame I do not see any intentions of the market for EURUSD with the exception of the attempt of raising of the up-trend continuation but weekly time-frame shows risk of falling so I am out.

Maybe it sounds a little crazy, but i also see a H&S forming, image attached.

http://img153.imageshack.us/img153/9084/hsgk7.th.jpg

Also, the oil price is becoming low, the swap is positive for short positions, i don't know maybe this drop can be a probable scenario, i will remain in my position for some days, and see what will happen..

Let's take a look also to the indicators, everyone is showing some kind of divergence, image attached.

http://img176.imageshack.us/img176/3160/divertj5.th.jpg

I'm absolutely not sure about this drop, but.. i'm not sure about the opposite, let's come the next week..!