Random Flow Theory and FOREX - page 60

 
benik >> :

I would also like to ask people: does anyone have a function that returns a value with normal distribution in the range (0,1)? I killed all day yesterday, but still haven't figured out how to implement it in mql.

Here's formula to turn uniform random, which MT does, into normal - https://en.wikipedia.org/wiki/Box-Muller_transform

 
FOXXXi >> :

What percentage of profitable trades in real trading and the ratio of average profit to average loss?

Do you understand the difference between a real process and a mathematical model that tries to simulate it?

If the process is a stationary one, then its parameters are well known, which means that there will be no losing trades at all, or exactly as many and as big as you want. Number and size of profitable trades depend on parameters of the model. For the first example with the normal distribution there will be a lot of deals. For the second example, AR(1), the number of trades depends on a, the more a, the less trades, the size of profit in each trade depends on parameters (st.dev.) of the process s(t).

Real losses and profits depend on how close the chosen model is to what you observe in real life. And of course depending on the parameters of the model as stated above.

 
benik >> :

You say that at a time.
Please don't be too lazy to write a script in mql that would simulate a winning strategy on a process with a normal distribution.

I think I am too lazy to write a script for now. And you, please explain how you can NOT create a winning strategy on a graph like this - it's a normally distributed process.



 
timbo >> :

I think I am too lazy to write a script for now. And you, please explain how you can NOT create a winning strategy on a graph like this - it's a normally distributed process.

Um. It's certainly very easy to work on such a chart, as long as it's a chart of the price itself.
The problem, however, is that all the pluses of such a chart tend to disappear if it's the price after the transformation. Suppose we somehow manage to reduce a real price chart to a process like the one you have in the figure. This process is quite easy to predict at some points. But in order to predict the real price we need to perform a transformation that is reversed from the one that was done in the beginning. This is what kills the benefits.
It's quite difficult to explain without revealing the details. And the details, you understand yourself, can't be laid out on a forum. Well, I will think over how to explain it so that "the wolves have been satisfied and the sheep intact". In the mean time just answer: Have you managed to create at least one perceptibly profitable trading strategy based on price transformation to a stationary form?

to Yurixx
I must have misunderstood you. You suggested to first load the P.D.F. values from outside, didn't you?

 
benik >> :

Hm. on such a chart it is of course very easy to work, if it is a chart of the price itself.

The original question was "how to create a strategy on a stationary process". The answer was "easy!" Precisely because the process is stationary.

Price is not a stationary process. One widely used model for the price process is random walk, a process which is guaranteed to be unpredictable. That is, one cannot make money on price. Or rather, someone will earn, someone will sell out at the same time, the first one will sell out later - there can't be stable earnings.

There are variants of stable earnings on occasional movement of price. Two men won the Nobel Prize for this idea, and everyone knows that Nobel prizes, especially in economics, are awarded to dullards. "A lot of fiddling" and repeatedly breaking into open doors, Yurixx thinks it's Timbov's "ninth wonder of the world". Case in point. "Kool hackers don't read manuals". Or rather, no manuals are written for demo traders.

I use this "miracle" to make 10-20% per month of the amount of attracted capital (not to be confused with the deposit).

 
timbo >> :

I think I am too lazy to write a script for now. And you, please explain how you can NOT create a winning strategy on a graph like this - it's a normally distributed process.



Absolutely impossible - NOT to create. Therefore, my words "who needs a flat?" only partially apply here. If you REALLY believe it's a flat, then your task is just to cut off emissions and open positions at significant deviations of the price from the horizontal line of "stationarity". Regarding the "amount of a flat" in the price flow - I have encountered different estimates - from 25% to 80%. In such a case of a flat price becomes similar to a random variable/process and here one can indeed apply some developments in matstatistics and probability. The question remains - how do you know if you are in a flat and how long will it last?

 
I'm having trouble connecting the stattrick dll. It writes: "TEST_Probability EURUSD,H4: dll calls are not allowed; 'probability.dll'-'bdtr'". What do you think it means?
 
Mathemat >> :

And why should I interfere when it's very good and quite dynamic here without me? But still I found out something interesting: it turns out that in modern mathematics there is no such a thing as probability.


Colleague, please write more clearly from now on: either you are sarcastically mocking me, or you are really aware of the matter. Otherwise it is not clear whether I should elaborate and give references. Just in case, here's a quote from Wikipedia:


The word probability does not have a consistent direct definition. In fact, there are two broad categories of probability interpretations, whose adherents possess different (and sometimes conflicting) views about the fundamental nature of probability:

  1. Frequentists talk about probabilities only when dealing with experiments that are random and well-defined. The probability of a random event denotes the relative frequency of occurrence of an experiment's outcome, when repeating the experiment. Frequentists consider probability to be the relative frequency "in the long run" of outcomes.[1]
  2. Bayesians, however, assign probabilities to any statement whatsoever, even when no random process is involved. Probability, for a Bayesian, is a way to represent an individual 's degree of belief in a statement, given the evidence.

http://en.wikipedia.org/wiki/Probability

Of course, it's just ridiculous that "probability science" hasn't been able to decide on the most basic definition for 200 years and they all don't know exactly what they're doing, so some of them are engaged in INTERPRETATION of the underlying word:

http://en.wikipedia.org/wiki/Probability_interpretations

 

AlexEro, thanks for the links (without steering). I know it. But you yourself have an idea of the difference between a concept and its interpretation. I do not see anything funny in the lack of definition of the central concept of the field of science. This is a contradiction of our whole science (at least its strictest part - mathematics), as it is constructed according to known axiomatic principles, and no one really knows how to change the state of affairs.

Do you know that in geometry the concepts of point, line and plane are undefined?

And natural numbers, which, according to the well-known expression, were invented by God? Here I am frankly floating: I do not know their strict definition, which seems to have been invented by Peano or someone else. But in the theory of numbers (natural numbers), the natural series itself is not a defined concept. And why define it if it's already an obvious concept? :)

And set (you must know at least a little bit about the problems of naive set theory)?

And gravitation in physics is also something extremely vague and certainly not strictly defined?

All this stuff is undefined and has a lot of interpretations, but this fact doesn't prevent them all from being practical and not purely theoretical concepts.

P.S. My snide remark was referring to the following words of yours on p. 49 of this thread:

начав разбираться с определением "вероятности" (а оно напрочь ОТСУТСТВУЕТ в современной математике)

Well, yes, we are essentially talking about the same thing, I screwed up, I didn't read carefully. The probability itself is there, but there is no definition. And it is very good that Kolmogorov's axiomatics allowed to finally clarify what we can define and what we cannot.

 
timbo писал(а) >>

The original question was "how to create a strategy on a stationary process". The answer was "easy!" Precisely because the process is stationary.

Price is not a stationary process. One widely used model for the price process is random walk, a process which is guaranteed to be unpredictable. That is, one cannot make money on price. Or rather, someone will earn, someone will sell out at the same time, the first one will sell out later - there can't be stable earnings.

There are variants of stable earnings on occasional movement of price. Two men won the Nobel Prize for this idea, and everyone knows that Nobel prizes, especially in economics, are awarded to dullards. "A lot of fiddling" and repeatedly breaking into open doors, Yurixx thinks it's Timbov's "ninth wonder of the world". The case

The illustration of a stationary process you cited has nothing to do with price or stationarity. If you want to know what a stationary process is, see page 57 of this thread for the definition given by AlexEro.

Speaking of terms. Once again you don't understand what you're being told. Random walk is a well defined term of physics and mathematics. The SB process has a normal distribution and is stationary in both the broad and narrow sense. You cannot make money from this process - that is the mathematical result. My assertion was that the price series is not a random walk. To quote myself for the stupid:

Yurixx wrote >>

I was investigating a one-dimensional equal probability random walk. Exactly what is described in wikipedia. Not experimentally, as you might think, but theoretically. Obtained certain, absolutely rigorous mathematical results. Comparison of those results with "random price wandering" shows that it is definitely not random wandering. The reliability is 100%.

Note the highlighting. The words are in quotes because they are your words. There is no such thing in mathematics, it is your invention. Therefore the meaning of what I have said may be formulated in the following form: since the price series is not a random walk, the possibility of earning on it is not excluded.

And now I ask for the names of these two men and a link to their work. Enough with the unsubstantiated assertions already.