Forget random quotes - page 52
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
There are no problems with detrending. Lots of qualitatively different methods. The most elaborate tool.
such as? least squares method?
The ISC is an estimation method, not a detrending tool.
It is the use of estimation methods, of which there are many, that qualitatively distinguishes statistics from TA.
It is not their aphorism. Once needed systems theory, studied it.
It is almost an axiom there: 'The question of model adequacy in systems theory is incorrect, because the very structure of the model is determined by the subjective division of the system into an object of influence and an environment.
The problem of the adequacy of a model to its object is pretty well worked out, but I haven't come across it on the market. It seems to me that a standard scheme is used: some verbal assumption is made about the marketplace, then this assumption is decomposed into components and these components are written down analytically. But there is always a tail that drives the whole model. But I cannot recall such works whether the obtained model is adequate to the market itself.
Although modeling always calculates the error of the model. Can this be it?
I haven't thought about it. Although the problem is obvious.
The problem of the adequacy of the model to its object is pretty well worked out, but I haven't seen it in the marketplace. It seems to me that a standard scheme is used: some verbal assumption is made about the marketplace, then this assumption is decomposed into components and these components are written down analytically. But there is always a tail that drives the whole model. But I cannot recall such works whether the obtained model is adequate to the market itself.
Although modeling always calculates the error of the model. Can this be it?
I haven't thought about it. Although the problem is obvious.
The model is the TS. Tailings are forecasting errors in the form of losses. Drawdowns are the accumulation of error in a series of trades. Traders analyse them through such indicators as MAKSDD, PF, FS, Sharpe ratio, etc. Most assess the adequacy of the model/TS purely visually - by the smoothness of the curve and maximum drawdowns. Some by imperial rules - like PF>2 etc.
Just different TS have a lot of nuances in the analysis of results to reduce everything to a single mathematical formula.
The model is a TC. The tail is forecasting error in the form of losses. Drawdowns are the accumulation of error in a series of trades. Traders analyse them through metrics such as MAKSDD, PF, FS, Sharpe's coefficient, etc. Most assess the adequacy of the model/TS purely visually - by the smoothness of the curve and maximum drawdowns. Some by imperial rules - like PF>2 etc.
There are just a lot of nuances in analysing the results of different TCs in order to reduce everything to a single mathematical formula.
This is all understandable. If we add to your list of thoughts about the statistical characteristics of the residual (error), we get a fairly complete list.
But my understanding of Mathematician is that he is talking about model evaluations in systems theory. In TAP for example. These are all extremely advanced. Their models fly as far as Mars, target US cities themselves, etc. It's about that which doesn't apply in the marketplace.
The ISC is an estimation method, not a detrending tool.
It is the use of estimation methods, of which there are many, that qualitatively distinguishes statistics from TA.
If there is a trend assessment, how is it not a detrending tool? and TA is a consequence of this, is it not?
This is all understandable. If you add to your list thoughts about the statistical characteristics of the residual (error), you get a fairly complete list.
But my understanding of Mathematician is that he is talking about model evaluations in systems theory. In TAP for example. These are all extremely advanced. Their models fly as far as Mars, target US cities themselves, etc. It's about that which doesn't apply in the marketplace.
there deal with models with stationary error - they do not lose their relevance over time. The laws of physics always apply, although there are also unaccounted for factors, but they are insignificant. In the market everything changes and participants adjust to changing rules and to each other. So it is not known in advance how long the model/TS will be relevant/robust. Therefore the relevance needs to be constantly monitored, not just once in a century as in some other areas. The lag in determining the relevance of the TS is critical.
There deal with models with stationary error - they do not lose their relevance over time. The laws of physics always apply, although there are also unaccounted for factors, but they are insignificant. In the market everything changes and participants adjust to changing rules and to each other. So it is not known in advance how long the model/TC will be relevant/robust. Therefore, the relevance needs to be constantly monitored, not once in a century as in some other areas. The lag in determining the relevance of the TS is critical.
Yes, I missed that, it just slipped my mind.
Part of our object we are trying to model is human, and this leads to the object's behaviour being non-stationary and to everything having qualitative jumps. I just forgot. I was once taught that objects come in deterministic, stochastic (a mixture of the two) and indeterminate, i.e. one whose behaviour can be deterministic at one interval, then stochastic and, often with a jump from one to the other. This is a characteristic of systems of which humans are a part.
Damn, everything slips my mind. They taught me that back in college.
The story of when Soros brought down the British pound is well known.