Whether there is a process whose analysis of one part does not allow predicting the next part. - page 9
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
can I explain to those who don't understand?
What are the plots in the past? and I take it that it also includes the present...
And in the sense of Cauchy mode will become the mean?
can I explain to those who don't understand?
What are the plots in the past? and I understand that the present is also present...
And in the sense of Cauchy the mode will become the mean?
Well, I was just giving an exaggerated example to show that lack of MO and variance and non-stationarity is not a reason to consider the process as unpredictable. The key phrase is the ability to identify areas of predictability, meaning in time.
Of course, I was lying in the heat of the moment about MoD.
Where is the stationarity on "ordinary SB"?
Where is the "ideal RMS"?
P.S. You need to be more specific about what you are talking about. If you're talking about returns, yes.
Unfortunately, any forecast can only rely on a deterministic component. On rows that do not have this component, any prediction, and therefore earnings, becomes impossible.
How the team views such considerations.
1. Prediction is possible if there is a deterministic component.
2. The deterministic component is differentiable not only on the left but also on the right at the last bar.
3. Differibility to the right (until the next bar arrives!) is provided by the type of the smoothing function. I saw somewhere that cubic splines at the junction remain differentiable.
It is possible to predict undifferentiated functions as well.
Prediction is also possible in the absence of a deterministic component.
We should not associate differentiability with predictability. It is like comparing warm and soft;
This is not an answer, but a question to you regarding your own delusions. I give an example to refute them.
A non-stationary process with density 1/pi*1/(1+(x-x0)^2), and expectation x0 is also a random variable, let it be for complete uncertainty - with unknown distribution (stationary or not - also unknown). And let the correlation time of the process be non-zero, i.e. the integral of the product of ACF(tau,t)*tau is greater than 0 for any t.
What do we know about the process:
a) Its variance is always infinite (calculate the integral if you don't believe).
b) It is non-stationary both in the narrow and almost probably broad sense. The first follows actually from the definition of stationarity in the narrow sense, as the density of the process is not constant, the second follows from the unknown properties of the process x0.
Nevertheless, despite all the aggravating circumstances, under certain conditions, namely, in those sections where the correlation time (it may not be constant - the process is non-stationary!) exceeds some threshold value, we can make predictions with a perfectly acceptable finite variance. It is the condition of good (exceeding some threshold, which in principle can be calculated) correlation of the process at least at some moments, and our ability to identify these moments are sufficient conditions for the possibility of prediction. However, the facts of non-stationarity and lack of variance are not important in themselves.
The error can vary as it wants, and our job is to be able to calculate it. If we can do that, why can't it be different for different points in time? Your fatal error is that you do not distinguish between the variance of the forecast and the variance of the predicted process, which are completely different things and not rigidly related to each other. The presence and depth of the relationship between them depends on many factors, including the amount of knowledge we have about the process, the forecasting methods we have in our arsenal, and only lastly on the properties of the forecasting process itself. The example above confirms this.
It is true that you are not the only one who is fixated, because people tend to err not on their own, but on the advice of authorities.
It is not about authority.
The fallacy of your reasoning is typical of people with a mathematical background (maybe you don't have one, but the fallacy of mathematicians) who are very trusting of mathematical calculations.
In statistics it is very easy to get just about any justification, which is easily refuted by simple reasoning, which I am very fond of.
Uncertainty of variance is a determinant of prediction and referring to historical data is not appropriate, whatever formulas it may be covered up with.
A simple example. Shooting at a target. I was taught that normal law rules and we can judge the probability of hitting 10, 9, 8, etc. and estimate the quality of the shooter. The basis is the variance value, which we have calculated from historical data. But if any shooter is blindfolded, put in a chair and spun, the whole story along with the variance will go into oblivion.
To me, this is what is a sign of non-stationarity. The past says nothing. And it takes some effort to use the past.
A prediction is a random variable, i.e. the figure we calculate is a realisation from a range, and it is the range boundary and the level of confidence in the calculated range boundary that is fundamental. There is nowhere without variance. what if it is a variable? ARCH models in particular try to model this variance, clarify the uncertainty of the variance and by getting some insights into the behaviour (not a constant, but a behaviour) of the variance more definitely make a statement about the prediction.
If your post talks about being able to work with non-stationary VR - then I completely agree with you. But always in model it is necessary to specify how this problem is solved, by what method, what will be solved and what not, as I don't know complete solution of non-stationarity problem. There will always be areas with some unsteady characteristics that our model does not take into account, TC will merge and we will never get the balance line as a straight line.
I'll write later, OK? I can't...
The uncertainty of the variance is crucial to the prediction and no historical data is appropriate, no matter what formulas are used to hide it.