Obtaining a stationary BP from a price BP - page 3
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
Do a little more thinking :) Just because the MO of a random variable=0 doesn't mean that CB itself can be replaced by zero, as you deftly do :) :)
And from what point do you believe the direction of the tail in ZZ? It's clear what we're modelling here, how to use the confidence interval as well. And what are you doing? Predicting the price on the next candle, Why? What does this have to do with a stationary or non-stationary series. No purpose, no means, you learn it in vocational school and start chasing words around the forum.
And at what point is the direction of the tail in ZZ to be believed? It is clear what we are modelling here, how to use and the confidence interval. What are you doing? Predicting the price on the next candlestick. Why?
Who told you I do that? It's Reshetov's extrapolation, I have no reason to do it :)
What does this have to do with stationary or non-stationary series? No purpose, no means, you learn it in a vocational school and start chasing words around the forum.
Who told you that I do that? Reshetov is extrapolating, I have no reason to :)
I don't struggle with unsteadiness at all - that's your point ;) And the goal seems to be the same for everyone - sustained profit.Converting a non-stationary series into a stationary series is an exercise that has nothing to do with profit. Profit can come from a specific idea, such as a trend reversal. Then from all non-stationarity problems we get the problems of finding reversals in non-stationary VR. There may be other TC ideas. But always fighting non-stationarity is for a specific purpose. Everything else is in the Nobel forum.
Why don't we pay attention to the ponytail in ZZ?
What do you mean? I mean the way to check if the row is white noise, and you mean ZZ. I'm not interested in ZZ, I'm interested in the topic of the thread.
It's a prediction for sure.
No one is stopping you from treating ZZ as a prediction.
And you have yet to prove that you have one.
This phrase of yours I did not understand at all (neither what it refers to, nor its meaning).
What do you mean? I'm talking about the way to check whether a row is white noise, and you're talking about ZZ. I'm not interested in ZZ, I'm interested in the topic of the thread.
No one is stopping you from treating ZZ as a prediction.
There were such people in the Middle Ages - alchemists. They too transformed shit into gold.
The topic of the thread: transforming unsteady BP into stationary. Why? I wrote above about profit. You don't have to convert anything. You have to process what you have under the TC idea.
What for? I wrote above about profit. You don't need to convert anything.
Well, that's what you think.
Well, that's what you think.
I'm not the only one. When we transform something into something, the main question is 'does the result have anything to do with the original phenomenon'. Not answering that question is alchemy.
Converting a non-stationary series into a stationary series is an exercise that has nothing to do with profit. Profit can come from a specific idea, such as a trend reversal. Then from all the problems of non-stationarity we get the problems of finding reversals in non-stationary BP. There may be other TC ideas. But always the struggle with non-stationarity is for a specific purpose. Everything else is in the Nobel forum.
There is a real practical problem to analyse non-stationarity and find the parameters of the extrapolation model. This is, for example, the analysis of system performance. For example, a model that defines equity as the sum of the trend component and the noise component. Accordingly, we can analyze the residuals - this noise component - for the absence of the above-mentioned dependencies. It may allow us to determine the trend component if it is really present. It seems to be clear what it is for.
I.e. there may be practical applications, but it is not a price extrapolation. At least, not on a long and continuous price series. Perhaps some local segments of time series. Perhaps, again, identification of some local trends and their patterns. Maybe someone will find it useful? The alchemy is not in the theory but in its inappropriate application.
there is a real practical task for analysing non-stationarity and finding parameters for an extrapolation model. This is, for example, the analysis of system performance. For example, a model that defines equity as the sum of the trend component and the noise component. Accordingly, we can analyze the residuals - this noise component - for the absence of the above-mentioned dependencies. It may allow us to determine the trend component if it is really present. It seems to be clear what it is for.
I.e. there may be practical applications, but it is not a price extrapolation. At least, not on a long and continuous price series. Perhaps some local segments of time series. Perhaps, again, identification of some local trends and their patterns. Maybe someone will find it useful? Alchemy is not in the theory, but in its inappropriate application.
Alchemy is a methodology where we ignore everything that has been done before, we do not identify the problem, we do not specify attempts to solve this problem. all further discussion is about nothing. You can assume what you have assumed, you can fancy something else. Absence of equity from the noise component shows only that we cannot get equity from the noise component.
There were several attempts to analyze non-stationary series on the forum, but the question of its transformation into a stationary series was always raised. The closest to a serious conversation about non-stationarity was on the branch about LPF and others. As soon as anyone put forward the idea of identifying the transition from one pseudo-stationary section to another - there were always returns and everything died.
Alchemy is a methodology where we ignore everything that has been done before, we don't identify the problem, we don't specify attempts to solve that problem. all further discussion is about nothing. You can assume what you have assumed, you can fancy something else. Absence of equity from the noise component shows only that we cannot get equity from the noise component.
There were several attempts to analyze non-stationary series on the forum but the question of its transformation into a stationary series was always raised. The closest to a serious conversation about non-stationarity was on the branch about LPF and others. As soon as anybody expressed an idea of identification of transition from one pseudo-stationary section to another - there always appeared returns and all died.
So start a branch and discuss what interests you.