The market is a controlled dynamic system. - page 13
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
Firstly, I am not claiming anywhere that the coefficients are constants.
Not you, but the author of the book you are referring to.
Secondly, if you had given yourself the trouble to look into the formulation of the problem, you would have noticed that initially the problem was set in the class of stochastic differential equations
The problem of "stochasticity" is a problem in itself and the market is not conceivable without taking this factor into account. But the characteristics of the random variables in your diffusers are very important. The main issue is stationarity. I didn't notice any reasoning on this topic in your posts.
And again, would you be so kind as to provide a link to a textbook on econometrics that describes the use of dynamical systems of random structure. Otherwise, it would be a good idea to apologise, as I was not assessing your knowledge and you, for some reason, assessed mine as "hearsay"
Firstly, I am not claiming anywhere that the coefficients are constants.
Not you, but the author of the book you are referring to.
Secondly, if you had given yourself the trouble to look into the formulation of the problem, you would have noticed that initially the problem was set in the class of stochastic differential equations
The problem of "stochasticity" is a problem in itself and the market is not conceivable without taking this factor into account. But the characteristics of the random variables in your diffusers are very important. The main issue is stationarity. I didn't notice any reasoning on this subject in your posts.
And again, would you be so kind as to provide a link to a textbook on econometrics that describes the use of dynamical systems of random structure. Otherwise, it would not be a bad idea to apologise, as I was not assessing your knowledge and you, for some reason, assessed mine as "hearsay"
Hmmm.... you seem to be completely out of touch...
The coefficients of the stochastic equations and the G and L matrices -- are by no means constants, but current estimates. And they remain unchanged on the current observation interval -- again, not because they are constants, but because they are so-called "slow processes" and on the current small observation and control interval they can be viewed as "frozen coefficients" -- well that is if you are familiar with the technique of dividing processes into fast and slow.
.
Next. In your formulation of the problem the main question is stationarity. But my formulation of the problem is such that initially processes are considered as non-stationary, i.e. in the class much wider including stationarity as a special case, not distinguished by anything special.
.
And finally, about the reference to the econometrics textbook... Here again you are mistaken -- we are not talking about a textbook on econometrics. It's about systems engineering applied to time series, which are data sets of any kind -- stock quotes, forex, cardiograms, encephalograms, solar flares.....
And if I have hurt your ego so much, well, I'm sorry, I really didn't know, and now I don't know the level of your knowledge of econometrics.
The coefficients of the stochastic equations and the D and L matrices are by no means constants, but current estimates.
This is already more interesting. I can't resolve the "slowness" of these coefficients. I suspect that you haven't dealt specifically with the behaviour of these coefficients. Let's estimate coefficients of some equation and you can see from the standard error value that we can't speak of slowness (it reaches 30% of the coefficient's value).
Next. For you, or rather, in your formulation of the problem, the main issue is stationarity. My formulation of the problem is that initially the processes are considered as non-stationary, i.e. in a much wider class including stationarity as a special case that does not stand out in any way.
In my formulation of the problem, on the contrary, the initial process is non-stationary and its non-stationarity gives rise to a host of problems when building the model. The first problem is the presence of trends that "crowd out" the stochasticity of the process. So I would separate economic data from solar flares.
hmmm.... I can see you're not in the loop at all...
Finally, about the link to the econometrics textbook... Here again you are mistaken -- we are not talking about an econometrics textbook. It's about systems engineering as applied to time series.
I'm right on top of it. There is a science of measuring economic data, and you have come to the econometrics monastery with a systems engineering charter. You still have to prove that it can be done and productively. Without comparison with established science, you cannot justify a new application.
You and I speak different languages... However, we need to dot our i's and cross our t's. This branch (my "monastery", so to speak) was conceived and created by me exactly as a system-technical branch. The object of research here is time series. The purpose of research is definition of a control (or otherwise, a setting function) for the investigated time series. No economic theories with their productive forces and other components are considered and NOT mentioned here.
But you barge in here with your econometric charter, and with the nerve to accuse me of my wrongdoing. But your econometric monastery, as a monastery, does not interest me at all.
They are completely different approaches, tasks, methods and results.
You and I speak different languages... However, we need to dot our i's and cross our t's. This branch (my "monastery", so to speak) was conceived and created by me exactly as a system-technical branch. The object of research here is time series. The purpose of research is definition of a control (or otherwise, a setting function) for the investigated time series. No economic theories with their productive forces and other components are considered and NOT mentioned here.
But you barge in here with your econometric charter, and with the nerve to accuse me of my wrongdoing. But your econometric monastery, as a monastery, does not interest me at all.
These are completely different approaches, objectives, methods and results.
Before you open threads, learn how to respond to posts on their merits.
Farewell. I leave your manners to you.
Before you open threads, learn how to respond to posts on substance.
Goodbye. I leave your manners to you.
avtomat:
... but there can be perceptual inconsistencies ;)
and this is a prime example of such perceptual inconsistencies...
I am unable to resolve the "slowness" of these coefficients.
For a first glimpse, to get at least a general idea of the "slowness" in question, look here.
This article is by no means sufficient, but it will help to get an idea.
The theory itself is very broad and beautiful. It has many branches as well as many applications.
The problem of 'stochasticity' is a problem in its own right and the market is not conceivable without taking this factor into account. But the characteristics of the random variables in your diffusers are very important. The main issue is stationarity. ...
And again, would you be so kind as to provide a link to a textbook on econometrics that describes the use of dynamical systems of random structure. Otherwise, it wouldn't be a bad idea to apologize, as I wasn't assessing your knowledge and you, for some reason, assessed mine as "hearsay.
The use of dynamical systems of random structure in a textbook on econometrics may be described somewhere, although I have not come across it, but usually the descriptive apparatus you are interested in the literature on statistical mechanics or on statistical dynamics, another name. I.e. dynamical systems, but describing in dynamics exactly the probabilistic characteristics of processes.
p.s. I hope, the author of the topic does not mind a short remark, faa1947 does not understand that his favourite package is not intended for the analysis of market series with unstable probability characteristics.
Here's a picture - it looks like an ordinary quotation, in fact it's a sum of sine waves. I wonder if it can be identified as a stationary series?